Cellular Automaton Rafting Simulation Subtitle

C.A.R.S.:Cellular Automaton Rafting Simulation

Subtitle

Control#15878

13February2012

Abstract

The Big Long River management company o?ers white water rafting tours along its225mile long river with Y uniformly distributed camp sites.

We were tasked to optimize the rafting schedule for this company and to

formulate a method for determining the river’s carrying capacity.We repre-

sent the river as a one-dimensional array populated by objects representing

rafting parties of various trip length and propulsion methods.Thus the

model we propose can be treated as a cellular automaton-like structure pro-

vided there exists rules that govern how each cell behaves for each iteration.

Rules for assigning raft destinations are iterated daily,allowing our model

to dynamically respond to operating conditions such as new rafting trips,

inclement weather,and load-dependent degradation of camp sites,while si-

multaneously maintaining critical levels of a variety of visitor satisfaction

https://www.sodocs.net/doc/9b17164116.html,ing this dynamic model,camp site con?icts were eliminated,the

carrying capacity of the river was determined to be1102visitors per season,

with more than99%of visitors completing trips on-time and rejecting less

than2%of trips.

Contents

1Introduction3 2Plan of Attack3 3Our Assumptions4 4Important De?nitions5

4.1Visitor Satisfaction (5)

4.2Carrying Capacity (6)

5De?ning a Good Model6 6Scheduling:Static vs Dynamic Approaches7

6.1The Static Strategy (7)

6.2The Dynamic Strategy (7)

7The Model:A Cellular Automaton-like Strategy9

7.1The Visitor Object (10)

7.2Governing Rules for the Model (10)

7.3An Ideal Implementation of the Model (11)

7.4Control Implementations (12)

8Clarifying the Metrics12 9Results14

9.1Number of Interactions (15)

9.2Number of Visitor Rejections (16)

9.3O?-Schedule Completions (17)

9.4Carrying Capacity (17)

10Further Investigations19

10.1Environmental Degradation (19)

10.2Mitigating Storms (20)

11Conclusion20 12Memo to Rafting Management22

1Introduction

White water rafting has become one of the fasting growing outdoor leisure activities in the United States1,with participation in rafting activities increasing151%from1987levels.2 Multi-day trips of di?erent length,with opportunities for both rafting and traditional camping,have captured the interest of a public seeking a wilderness experience.As a result,park managers have sought to discover methods to accommodate more boating trips in a rafting season without loss of visitor satisfaction.

Previous models have emphasized social norm theory as an analytic tool for understand-ing the enjoyability of a camping/rafting trip.3Stankey argued that the enjoyment of a wilderness experience is inversely related to the number of interactions that occur with members of other parties during the trip,thus setting a social carrying capacity to the system.4Combining this Wilderness Simulation Model with trip logs,several researchers have constructed programs to generate optimal seasonal rafting schedules for various river systems.5

However,in order to reduce interactions between campers,a number of these models have generated seasonal schedules with trips of only one or two durations o?ered at any given time.These models often fail to take into account customer demand,and reduce the number of options in trip varieties to be o?ered to campers.To compensate,some rafting companies have set up multiple trips over di?erent rivers;however,this is clearly not practical for all rafting companies,and only partially mitigates the problem.Thus,we aim to develop a dynamic model to provide customers with a high quality rafting/camping trip of desired length.

2Plan of Attack

We will develop our model based on The Big Long River,a225mile long white water river system with Y camp sites distributed evenly along its banks.Rafting trips on this river are characterized by both their duration(ranging from6-18days)and by the raft type used(oar-propelled or motor).Our goal is to propose a method to allow as many visitors to raft through the river during the six month rafting season while maintaining an enjoyable experience for the visitors.We also wish to be able to provide a variety of di?erent trip options to the customers.The breadth of these requirements demands that we break down each component of the problem and analyze it separately;we will thus seek to restate the problem mathematically and clarify our goals.We will proceed by:

?Stating Assumptions.By stating our assumptions,we will narrow the focus of our approach towards the problem and provide some insight into the nature of the white water river rafting system.

?De?ning Evaluation Criteria.To make clear what the aims of our model should be,we must state what factors are involved in determining the success of the model.

?Describing a Good Model.A powerful model should be practical in terms of ease of application and ability to cope with extreme circumstances.By highlighting the characteristics of a good model,we will be able to direct our e?orts in the correct direction.

Once we have stated all of our goals,assumptions,and criteria,we will delve into our strategy to model the white water rafting system.We will:

?Present Our Model.We develop and test a dynamic model to simulate the river as a one dimensional cellular automaton-like structure with cells representing camp sites.Each cell will be occupied by visitors with unique properties.

?Develop Several Implementations.We will construct and compare several rules to govern loading visitors and movement of visitors along our cellular automaton-like structure.

?Evaluate the Model.We will postulate and test the behavior of our model under various limiting circumstances.We will then evaluate the strengths and weaknesses of our model and our model’s applicability to di?erent situations.

3Our Assumptions

To fully understand the whitewater river rafting system and to develop a model to simu-late this system,we must?rst make some assumptions.Since these assumptions will be essential in assisting us to construct our system,we will seek to formulate assumptions that are logical and reasonable.

?There are?xed ranges for both the duration of time visitors spend raft-ing as well as for the velocity of the rafts.We have estimated these ranges based on literature information such as the average velocity of white water rapids and sample rafting itineraries.67These values are listed in the table below.

Oar-Propelled Raft Motor Raft

Minimum Velocity3mph3mph

Maximum Velocity5mph10mph

Duration of Rafting4-8hours4-8hours

Table1:Constant values used in our model based on industry standards.

?While rafting,visitors do not dramatically change their velocity over the course of the day.We assume that visitors will not extensively dawdle or backtrack along the river.A large majority of camp sites send experienced tour guides out with visitors to ensure that the boat keeps moving at the necessary speed.As a result,motion of visitors along the river is a state function-it only depends the initial and?nal positions of the visitors along the river.

?Camp sites are equivalent and uniformly distributed alongside the river.

Thus,to evaluate the distance between camp sites,we simply need to consider the length of the river(225miles)divided by Y,the number of camp sites.

?Camp sites are located a signi?cant distance away from the river bank.

Due to safety,construction,and privacy issues,it is very unlikely that the sites are built immediately on the banks of the river.

?Visitors will travel and explore within a small distance of the camp site once o?the raft.While hiking is a very common outdoor leisure activity,part of the constraints of the Big Long River is that the river is inaccessible to hikers.

For this reason,we assume that no visitor will travel signi?cantly far from the camp site-a reasonable assumption,if we consider that a large portion of campers seek rest and relaxation.8

?The camp sites are properly maintained on a routine basis by the river management.We assume that at all of the camp sites will remain functionally operative throughout the season.We also assume that the river management will frequently service camp sites to clean garbage and to ensure availability of and access to clean water and bathroom facilities.

?The mean number of visitors that are loaded onto the river on any day remains constant.Most rafting companies o?er a set number of trips on a daily basis,and thus this assumption is in accordance with standard industry practices.

?Visitors will schedule their rafting trips at least24hours in advance of their desired launch date.Given industry standards,this assumption is reasonable,as most white water rafting companies require advanced booking of trips.

Our assumptions allow us to simplify the complexities of real life white water rafting systems,as well as provide us with a better understanding of our problem.We will frequently reference these assumptions to help us develop our dynamic model.

4Important De?nitions

To evaluate our model,it will be necessary for us to de?ne some of the primary criteria of our model:

1.Visitor Satisfaction-How do visitors perceive their wilderness experience?

2.Carrying Capacity-How many visitors can participate in white water raft-

ing/camping in a rafting season compared to the absolute capacity of the river system?

While these criteria,particularly visitor satisfaction,are somewhat subjective,we will seek to use both the literature as well as assumptions derived from common sense to justify our de?nitions.

4.1Visitor Satisfaction

A signi?cant portion of research on leisure activities in camping and rafting systems has revolved around the role of social norms.Based on these studies,we propose the criterion of visitor satisfaction,and break this criterion down into several components:?Visitor satisfaction is inversely correlated to the number of interactions

a visitor has with other groups of people.9

For many,wilderness excursions are meant to be privately shared and enjoyed;

it is no surprise that the literature on public perception towards camping trips frequently revolves around overcrowding.Since visitors in our model will either spend their time rafting or camping,there are three modes of interaction4that we will consider:

–Raft to Raft Interactions,which occur when one raft passes another on the river

–Raft to Camp Interactions,which occur when one raft passes a group at

a camp site

–Camp to Camp Interactions,which occur when individuals walking around their camp sites encounter one another

?Visitor satisfaction decreases if a trip is completed o?-schedule.

The amount that visitors pay for a trip is generally proportion to the duration of the trip;thus,if a trip is completed early,then the visitors may feel cheated out of their money.On the?ip side,most of the visitors only have a limited amount of leisure time,and if a trip is completed late,the visitors may be irate due to lost time.

?Visitor satisfaction decreases if visitors are unable to participate in de-sired trips.

Visitors all have di?erent personal schedules and commitments,and may desire a rafting trip that is uniquely customized to their needs.Thus,general satisfaction with a white water rafting company may decrease if the company is unable to accommodate customer demands.

Having broken down visitor satisfaction into these various components,we will seek to maintain satisfaction at a high level throughout the development and testing of our model. As these evaluation criteria for visitor satisfaction have been developed in the literature,our model can now be compared to established standards for white water rafting and camping.

4.2Carrying Capacity

We?rst consider the concept of absolute capacity,or the maximum number of visitors that can be accommodated in our white water river system.It is apparent that in our rafting system,the absolute capacity is:

6(months/season)×30(days/month)×Y=180Y(visitor groups per season) However,we must be cautious with interpreting absolute capacity-to state that a white water system can allow180Y visitor groups to be accommodated per season is akin to stating that the number of people that can?t in a house is equal to the number of people that can be squashed into every inch of the house.

We thus de?ne a criteria that could be used to better model white water river rafting systems:

?The carrying capacity of a model is the number of visitors that a model allows to participate in rafting/camping activities in any particular season while maintaining high standards of visitor satisfaction.

5De?ning a Good Model

Having developed a better understanding of the problem in terms of assumptions and criteria,we seek to describe the characteristics of a good model.These qualities will guide us towards constructing our model.

?The model should maximize the carrying e?ciency of the white water river rafting system while maintaining high levels of customer satisfac-tion.While the model’s goal is to allow greater numbers of visitor groups to participate in white water rafting than current industrial models allow,a good model will also minimize visitor interactions and o?-schedule trip completions.

?The model should satisfy customer demands in terms of o?ered trips.

A good model can allow customers to participate in trips suited to their personal

needs and interests.The model should seek to minimize the number of individuals whose trip demands cannot be satis?ed.

?The model should perform e?ectively under a wide range of variable operating conditions.The model should provide satisfactory results for a range of values of Y and the number of people who are loaded into the river each day.

The model should also present a contingency plan for conditions that impede the ?ow of visitors down the river,such as weather.

?The model must be easy to implement.A good model should be easy to operate,provide interpretable results,and can be utilized on a day-to-day basis in real-life situations.

6Scheduling:Static vs Dynamic Approaches

We have up until now not considered the concept of“scheduling”visitors to the white water river park.Our goal of getting visitors through the river breaks down into providing each visitor group an itinerary of what camp site the group must

reach by the end of each day.There are some key concerns that must be addressed in scheduling:

?To ensure that every group is assigned a camp site for every day of their trip

?To reduce con?icts-that is,to ensure that no two groups are assigned to the same camp site on any given day

?To e?ectively load new visitor onto the river

?To ensure that visitors complete their trip on schedule

To proceed,we must evaluate two di?erent strategies to schedule visitors to camp sites-the static approach and the dynamic approach.

6.1The Static Strategy

Static approaches are the most widespread of scheduling strategies,and are extensively used by white water rafting companies across the nation.In the static approach, the rafting company selects a small number of trips to o?er to the public. These trips are selected and constructed such that con?icts in camp site scheduling can never occur.Furthermore,unless extenuating circumstances like storms occur,visitors will always complete their journey exactly on time.Visitors can pay to participate in one of these speci?c trips.Trips are seasonally o?ered and rotated on a regular basis. While static models prevent the concern for con?icts in scheduling,they present several shortcomings for white water rafting systems.

?Static approaches fail to take into account customer demand.To prevent con?icts in scheduling,the static approach necessarily reduces the number of trips o?ered.By rotating trips seasonally,the static approach also reduces the amount of time that any one trip is o?ered.One can thus imagine a white water river rafting company losing customers due to lack of?exibility in trips o?ered.

?Static approaches reduce the number of visitors who can participate in rafting activities during the rafting season.Since the success of the static approach is built on preventing scheduling con?icts,a necessarily smaller number of visitors are allowed to take rafting trips at any time.While the static strategy provides optimal scheduling for any one type of rafting trip,it can only present optimal scheduling for multiple trips if fewer people participate in those trips. 6.2The Dynamic Strategy

Unlike the static system,the dynamic system emphasizes maximizing the num-ber of trip options o?ered to visitors.Such a system allows visitors to entirely cus-tomize their trip to suit their needs.The dynamic system requires a more complex con?ict resolution strategy than the static approach,but ensures that customers can select any trip they desire at any point in the rafting system.The dynamic approach works in the following manner:

1.As per our assumption,the river management will always know the

desired trips of visitors the day before the visitors are to be loaded.

2.On the?rst day of the trip,the visitor is assigned his/her?rst destina-

tion camp site.This assignment is based on the current occupancy of the camp sites on the river,as well as the desired duration of the trip and the raft type.

3.During this?rst day,the rafting company receives information for all of

the rafters to be launched the next day.

4.At the end of the?rst day,the visitor is given his second destination

camp site.This assignment is made in the same manner as the?rst,except that the model will also consider the other rafters who are to be loaded on to the river the next day.

5.Every day,the visitor receives the next destination site until the journey

is completed.

The dynamic method thus necessarily maximizes the number of trip options that visitors are given for any river system.The dynamic approach recalculates the route of each visitor at the end of each day to keep the visitor moving towards the anticipated complete date of the trip,but also considers how occupied the camp sites along the river are.The dynamic model does have some potential weakness,however, which we will consider here:

?The possibilities of con?ict and of o?-schedule arrivals is not necessarily zero in the dynamic model.However,with careful scheduling,the chance of either of these issues are occurring can be made negligible.

?The dynamic model may seem unsavory to river managers as it seemingly operates one day at a time.In an abstract sense,this is true;the dynamic model thus requires that the camp management has some way of remaining in contact with visitors throughout their trip.In a real-world scenario,however,rafting companies will be able to plan far in advance of one day.There are two immediate consequences to this assumption:

–With enough advanced planning,a dynamic model can provide itineraries for visitors that are just as“set-in-stone”as those gen-

erated by static models.

–Dynamic models can allow river managers and rafting companies to predict whether a visitor’s desired trip plan can be accommo-

dated given the current river occupancy.If a visitor’s trip results in

scheduling con?icts,the rafting company can inform the visitor to resched-

ule at a di?erent time.Though these“rejections”are generally undesirable,

the predictive ability of the model allows it to function e?ectively in the real

world.

Given the robust ability of the dynamic model,we propose a dynamic system for scheduling visitors down Big Long River.Our strategy will seek to eliminate chances of con?ict and minimize o?schedule arrivals and visitor rejections.

Static Model Dynamic Model Types of Trips O?ered Few Options Maximum Options

Frequency of O?ered

Trips Low,Trips are

Seasonal

High,Trips are O?ered

Seasonwide

Maximum Possible

Visitors

Fixed,Low High

Chance of Con?ict Zero Very Low with Careful

Scheduling

Chance of Early or Late Completions Zero

Low with Careful

Scheduling

Table2:A comparison of static and dynamic strategies of scheduling visitors to raft down the river.While the static model o?ers a more predictable output,with guarantees for no delays or mistimed trip completions,it also sacri?ces a great deal of?exibility in terms of trips o?ered to visitors.The dynamic model o?ers?exibility as well as potential for more visitors participating in rafting/camping,though care and attention must be given to careful scheduling.

7The Model:A Cellular Automaton-like Strategy We here describe our dynamic model of white water river rafting scheduling over the course of a rafting season.We proceed by constructing a cellular automaton-like structure to represent the river,and de?ne the movement rules for objects contained within the cells of the automaton.The de?nition of a cellular automaton has many variants,though it is generally accepted to be a grid of cells that can take on discrete values,or states.These cells evolve over discrete time steps based on speci?c rules,and can be iterated for in?nitely many desired time steps.10

Our model possesses certain similarities to a one-dimensional cellular automaton,as well as some di?erent characteristics that make it uniquely suited to our problem:?We describe the river as a one-dimensional array containing Y cells.

Each cell represents a camp site,and can be in one of two states-occupied by one and only one visitor or unoccupied by a visitor.The cells will be numbered from 1to Y,with larger numbers representing a camp site that is geographically more downstream along the river.

?We will de?ne a visitor object that can occupy cells in the river array.

Visitors will be characterized by unique identi?ers which we will develop in detail.

?There are a discrete number of visitor objects that can be constructed.

?Every iteration will be considered one day in real time.

?From day to day,we will apply a series of rules designed to move each visitor object in the river array to an unoccupied space in the array that is downstream from the visitor’s current position.

?Every day,new visitors will also be added to the array based on the governing rules of the automaton-like structure.

?To account for the discrete nature of the array,continuous distances are expressed as a function of the number of camp sites Y.

Figure1:A visual demonstration of the one-dimensional cellular automaton-like structure representing the river.Each cell represents a camp site,and is either occupied with a visitor or empty.Visitors are shown with two of the several de?ning characteristics.Note:The data shown in this picture was arbitrarily selected and not related to any output generated by our model.

The cellular automaton-like model is highly?exible in its ability to test a variety of conditions.Since the model is de?ned by a series of governing rules,we can swap out di?erent sets of rules to produce di?erent implementations of our model. These various implementations will give us a sense of how to most ideally model the white water river rafting system.Having considered the general characteristics of the model in terms of cellular automaton terminology,we will now de?ne the variables associated with the visitor object.We will then formulate the general rules to govern visitor movement from iteration to iteration,as well as consider speci?c implementations of the rules.We will?nally de?ne metrics to evaluate our results.

7.1The Visitor Object

The visitor object is a central aspect of our model,as the movement rules from iteration to iteration of our one-dimensional automaton-like river will be determined by the unique characteristics of each object.We must thus state all of the relevant variables of the visitor object.

?The type of raft to be used-oar-powered or motor.The raft type deter-mines the minimum and maximum distance that a visitor can move in any given day.

?The anticipated trip length,D a.The anticipated trip length sets the initial move conditions of the visitor and is used to determine whether the visitor is on schedule upon?nishing the trip.

?The current position of the visitor,y,where1

?The number of days traveled,D i,and the number of days left to complete the trip,D r.In the instance that D r is0but y

?The bump coe?cient,β.In the instance of con?icts,βis used to determine which visitor is placed in the desired camp site and which visitor is placed in a di?erent site.βdepends on how far the visitor must travel to reach the end of the river as well as the number of days left in the trip.

Bump Coe?cient.β=Y?y

D r

βis thus of equivalent form to the ideal number of camp sites that a visitor should travel in a day.However,βis continuous,while ideal travel distance is rounded up to produce a discrete value.

?The bump direction.Bump direction can either represent the upstream direction or the downstream direction,and plays a role in resolving con?icts.The value of bump direction is set based on the speci?c rules governing the model.

7.2Governing Rules for the Model

We here consider the general outline of rules for scheduling of visitors in our cellular automaton-like model.This will serve as a template for generation of implementations of the model.Due to the highly?exible nature of the model,variants on rules can be quickly substituted to produce other implementations,which we will also test. In general,in any given iteration of the model,the following procedures are followed:

1.Queueing and Loading:The visitors to be loaded on to the river are de?ned

(based on raft type and D a)and input into the model.To generate sample data for our model,we used a Poisson distribution with meanλ.Di?erent values ofλwere tested,withλranging from1to15.

2.Moving:For each visitor,y?is calculated based on the move rules of the imple-

mentation.If this distance is between the moving constraints of the visitor(as de?ned in Table1),the visitor is moved y*camp sites downstream in the river ar-ray.Otherwise,the visitor is moved according to the absolute minimum or absolute maximum moving distance,based on the speci?c scenario.

3.Resolving Con?icts:For all cells with more than two visitors,the bump rules

of the implementation are used to resolve con?icts.These rules are implemented until no con?icts remain.

4.Perform Calculations:The desired metrics are calculated.If any visitors have

y>Y,they are dequeued from the model.

Figure2:A schematic for the generalized rules governing our cellular automaton-like river model.Note that both the Moving and Resolving Con?icts steps can be quickly replaced with variants to test desired conditions.

7.3An Ideal Implementation of the Model

Having considered the generalized rules for the model,we develop a speci?c implementa-tion of the model.This implementation utilizes an ideal bumping procedure to ensure that scheduling con?icts are minimized.To fully describe this implementa-tion,we de?ne the move rules and the bump rules for the implementation:?Move:At the beginning of every turn,y*is calculated for each visitor with the formula y?=(Y?y)/D r This ideal moving distance is then used to move each visitor down the river.We stress that in this implementation,y*is calculated every iteration-that is,movement is self-correcting.

?Bump:In case of scheduling con?ict,the visitor with the highestβremains at the originally assigned camp site.All other visitors search for the nearest unoccupied camp site within their movement range in their bump di-rection.If such a camp site does not exist,then the visitor searches for the nearest unoccupied camp site within their movement range opposite the bump direction.

–If such a camp site does not exist,then the visitor is removed from the model and a note is made of this error.In our real world predictive model,

where visitor data is known well in advance of the visitor launch

date,this action corresponds to the river management informing a

customer that his desired trip is not available.

?Bump Direction:When each visitor is loaded into the model,the visitor is assigned a random bump direction.After this initialization,the bump direction is Figure3:A schematic for a speci?c implementation of our automaton-like model.In this particular implementation,y?is the ideal moving distance of each visitor.Though the chart points out that it is possible that any given implementation does not o?er all of the trips,in our data testing we allowed visitors to select any trip.

reversed every time the visitor is bumped.This reduces the likelihood that a visitor is bumped twice in the same direction,thus assisting in keeping the visitor on schedule.

7.4Control Implementations

To be able to assess the capacity of our ideal implementation of our model,we must have some standards of comparison.We thus generate variants of our ideal imple-mentation to serve as control implementations.These variants will constructed to highlight the unique characteristics that make our ideal implementation optimal.We will consider two major experiments:

1.We determine the role of both recalculating y?every turn and the use

of an ideal bump procedure on our model.Our ideal implementation recal-culates y?from iteration to iteration so as to be a self-correcting approach;we will thus create a variant implementation where y?is only calculated upon loading the visitor into the river.We will also create an implementation that does not bump visitors in case of scheduling con?ict.

2.We will consider how the initialization of bump direction as well as the

reversing of bump direction following a bump a?ect our model.In our ideal implementation,bump direction is initially set randomly;bump direction is then reversed every time a visitor is bumped.We will thus test variants in which the initial bump direction is always the same,as well variants in which bump direction is not reversed following a bump.This experiment will allow us to consider the self-correcting nature of our ideal bump rule.

We summarize all of the implementations to be tested in the table below.

Implementation y?recalculated?Initial Bump

Direction

Unidirectional

Bumping?

A(Ideal)Yes Random No

B(Exp.1Control)No Random No

C(Exp.1Control)Yes N/A N/A

D(Exp.1Control)No N/A N/A

E(Exp.2Control)Yes Downstream Yes

F(Exp.2Control)Yes Upstream Yes

G(Exp.2Control)Yes Random Yes

Table3:The ideal implementation as well as several control implementations.Since our model allows for a great variety of rules to de?ne movement,each unique aspect of our ideal implementation can be compared to a control implementation.

8Clarifying the Metrics

Having developed our dynamic cellular automaton-like model for white water river rafting systems,it is appropriate that we rede?ne our criteria in terms of our model.We will here seek to construct metrics that can be used to evaluate the results of our model. Our metrics should serve as a tool in allowing us to compare the various implementations of our automaton model;they will also allow us to evaluate our results based on industry standards.

1.Raft-to-Raft Interactions.According to our assumptions,when on the river,

visitors do not dramatically change velocity while travelling down the river.As

a result,motion down the river is a state function,and only depends on initial

and?nal conditions.In our model,where camp sites were represented by cells in an array,we need only consider which camp sites two visitors start and end at to determine whether an interaction has occurred.

Figure4:Our assumption that visitors maintain a constant velocity throughout the day allows us to treat movement as a state function.Thus,we need only consider which camps visitors start the day and end the day at to determine whether interactions have occurred.

De?nition.Let us consider the set of visitors V that are in the river grid.For some v1,v2∈V,a raft-to-raft interaction occurs between v1and v2if and only v1begins the day at a camp upstream from v2and ends the day at a camp downstream from v2.

2.O?Schedule Completions.O?schedule completions can be broken down into

both early and late completions,both of which are weighted equally when consid-ering o?-schedule completions.

De?nition.Let us consider the set of visitors V f who have completed their trips successfully.The number of early completions for the river system is equal to the number of v∈V f such that v a>v i.The number of late completions for the river system is equal to the number of v∈V f such that v a

3.System Rejections.In some rare cases,our dynamic model was unable to resolve

a scheduling con?ict,and thus removed the visitor with the lowerβfrom the system.

In a static case,an analogy would be turning away visitors whose demands were not accommodated.Our ideal implementation of the model should seek to minimize system rejections to make the model more powerful.

De?nition.Let us consider the set R of visitors who have been rejected from the model owing to unresolved scheduling con?icts(both prior to and after loading).

The number of rejections for system is equal to|R|.

4.Carrying Capacity.The carrying capacity of a model or implementation of a

model is,as per our previous discussion,the maximum number of visitors that can be accommodated in a white water rafting system per season while maintaining high standards of visitor satisfaction.We now present a more developed de?nition of carrying capacity in terms of the other metrics used to evaluate the model.

De?nition.Let us consider a model or implementation of a model M.The carrying capacity of M is the maximum number of visitors that the model allows to complete a trip in one rafting reason over all values of Y andλsuch that three standards are met:1)System rejections are fewer than10%of the visitors that travel down the river throughout the season,2)The number of visitors who complete the trip o?schedule is less than10%of the total visitors for the season,and3)There are fewer than10boat-to-boat interactions per person per day.

5.Camp-to-Camp Interactions.Our assumptions stated that visitors will only

move within a short distance of their camp site.Given that our visitors will have spent a signi?cant period of time rafting,it is unlikely that they will follow up this rafting with extensive hiking;furthermore,for reasonable values of Y,the probability of two visitors from even adjacent camp sites interacting is very low.

As a result,we state that the number of camp-to-camp interactions in our model is zero.

6.Raft-to-Camp Interactions.As per our assumptions,camp sites are located at

a distance from the river shore.As a result,we state that the number of

raft-to-camp interactions in our model is zero.

9Results

Implementation No.of

Visitors

%Rejec-

tions

%O?

Schedule

%Early%Late

A(Ideal)756 3.8 2.10.4 1.7

B(Constant y?)753 5.542.330.312

C(No Bump

Rule)

18275.5000 D(Constant y?,

No Bump Rule)

19074.627.527.50

Table4:Results of our?rst experiment,in which two variations were considered:1)y?is not recalculated at the start of every iteration and2)the bump rule is removed.We tested these implementations with Y=100andλ=5.

Implementation No.of

Visitors

%Rejec-

tions

Interactions

per Visitor

%O?

Schedule

A(Ideal)754 4.867.10 2.2

E(Unidirectional

Downstream

Bumping)

7198.737.34 2.2

F(Unidirectional

Upstream Bump-

ing)

51635.0 4.79 1.44

G(Unidirectional

Random Bump)

19074.627.527.5

Table5:Results of our second experiment,in which we investigated unidirectional bump-ing.We tested these implementations with Y=100andλ=5.

We present here the comparison of our ideal implementation(with idealized movement and bump rules)against control implementations designed to high-light the key components our the ideal implementation.From examining the results in Tables4and5,we note the following:

?Implementations with bumping rules increase the maximum number of visitors that completed trips during the season by nearly400%over implementations without bumping rules.However,as seen in implementation B,without recalculating y?at the start of each iteration of the model,the number of visitors that?nish o?-schedule also increases.

?The implementations that lack bumping move the visitors down the river in a manner analogous to that of a static model.In these implementations, each visitor has an ideal camp site that they travel to everyday regardless of other parameters.Therefore,everyone arrives on time,but the number rejected visitors is very high.

?Recalculating y?has very little impact on the total number of visitors but nearly eliminates the number of visitors who arrive o?schedule.

?Bumping in only one direction lowers the number of visitors who com-plete trips.This can be explained from the fact that when bumping in only one direction,there are less ways to resolve con?icts at camp sites and therefore less visitors can be on the river at any one time.

?While bumping in only one direction has the potential to reduce average boat interactions compared to bidirectional bumping implementations,unidirectional bumping implementations also reject a greater percentage of visitors.

?Bidirectional bumping rules increase the number of ways to resolve con?icts at camp sites.As a result,such implementations allow more visitors to complete trips during the season,and reject fewer visitors.

?While bidirectional bumping may increase the percentage of visitors who complete their trip o?schedule,this increase is o?set by a signi?cant diminishing of the number of rejected visitors.

These experiments have allowed us to demonstrate that Implementation A is in fact the ideal implementation of our model.Implementation A performs optimally from both a customer satisfaction and a business point of view.Not only is the number of visitors who participate in rafting throughout the season maximized,but the number of visitors whose desired trips are not accommodated is minimized.We will now test Implementation A over a wide variety of Y values andλto determine how successful our model is at scheduling white water river rafting systems.

9.1Number of Interactions

In order to ensure high standards of customer satisfaction,we want to ensure that the

Figure5:Surface and contour plot for average number of boat interactions per person with respect toλand Y.

number of boat-to-boat interactions per visitor is lower than the standard industry value. According to literature,customer satisfaction decreases if visitors experience more than 10interactions per day.For Implementation A of our model,however,the total number of interactions per visitor over the course of the season is less than10 interactions.Furthermore,values of boat-to-boat interactions are generally constant across variable values ofλand Y,showing the powerful nature of our model in maintaining high customer satisfaction standards.This observation also implies that carrying capacity of the model is not impacted signi?cantly by boat-to-boat interactions.

9.2Number of Visitor Rejections

As one of the key aspects of our model is that it accommodates for the demands of the customer,we sought to minimize the number of rejected visitors from our model.Through visual inspection of the surface,it is clear that the value for%rejections drops o?rapidly as a function of Y and is largely independent ofλat large values of Y.Thus,at large values of Y,Implementation A of our model is able to greatly maximize the number of visitors accommodated into the white water river rafting system.

Figure6:Surface and contour plot for the percentage of visitors who are rejected over the course of the season with respect toλand Y.

9.3O?-Schedule Completions

Figure7:Surface and contour plot for the percentage of visitors who complete their trip early or late with respect toλand Y.

We here consider the number of visitors who completed their trip o?schedule.We note that when visitors in our implementation of our model are o?schedule,they are often o?schedule by only one day.By examining the surface,we see that the percentage of o?schedule visitors drops o?rapidly as Y increases and becomes independent ofλat large values of Y.This observation is similar to the change in the percent of visitor rejections with respect toλand Y.Furthermore,we note that the%of o?schedule visitors approaches zero with increasing Y,implying that it is possible with enough camp sites to e?ectively eliminate the chance that a visitor will arrive early or late.

9.4Carrying Capacity

Based on the data from our various tests of Implementation A of our model, we can now consider a carrying capacity for our model.

Our de?nition of carrying capacity was based on determining the maximum number of visitors who could complete trips in a rafting season based on certain constraints of visi-tor satisfaction.Having analyzed the data,we can see that the carrying capacity for

Figure 8:Surface and contour plot for the number of visitors who are able to complete trips in our white water river rafting system over a rafting season of six months.

Implementation A of our model is 1102visitors per season.This carrying ca-pacity occurs when the system has 180camp sites and is loading an average of 15visitors a day into the model.At this carrying capacity,we note that:

?The percentage of visitors who are rejected by the model is 1.79%.Thus,our model accounts for the trip desires of 98.21%of visitors.

?The percentage of visitors who complete their trip o?schedule is 0.67%.Thus,our model completes 99.33%of visitors trips perfectly on schedule.

?The average number of interactions per person per season is 10.35.We note again that this value is signi?cantly lower than the established standard for white water river rafting systems.

However,we make the observation that the carrying capacity of our model occurs when the river system has 180camp sites,which from an economic standpoint may be very costly to maintain.We note that a more economic decision would be to sacri?ce some of the carrying capacity and reduce the number of camp sites along the river.For example,at Y =100and λ=15,the capacity is 1025visitor groups per season ,which is remarkably high and more cost e?cient than the actual carrying capacity.

10Further Investigations

Based on the data collected so far,it is clear that Implementation A of our dynamic model performs excellently in idealized conditions;however,the real world performance of implementation A should also be analyzed.This can be accomplished by breaking some of the assumptions used in making the model and seeing if the model can perform adequately and if not,what this means and how poor performance in the real world can be recti?ed.

10.1Environmental Degradation

In our model,we assumed that camp sites would remain functional throughout the season. However,this is not realistic.Currently research has shown that frequency of use of camp sites is a major factor in soil quality and vegetation at the camp site. After a certain point of degradation,camp sites can even become unusable. We can thus model a more realistic system based on the literature for the river sys-tem by de?ning the number of operational camp sites on any given day in the river system as:

[Y?αlog(λ)]

Based on this formula,we present environmental degradation data from our model:

Number of

λ

Visitors

α=07699109731071

10755880900830

20680751570525

30556497323298

%O?

λ

Schedule

α=0 1.8 2.3 3.5 4.0

10 3.7 6.18.010

209.1102418

3012283354

%Rejected

λ

Visitors

α=00.50.7 1.9 2.9

10 2.6 3.515.323.6

209.919.245.749.6

3028.347.771.075.1

Table6:We consider the possibility that environmental degradation could cause camp sites to become unusable.Here,we present data for our environmental degradation model, where Y=110camp sites.

From these tables it is clear that very high values ofαgreatly e?ect the per-formance of Implementation A.Whenαis equal to zero,the number of camp sites

remains unchanged;for larger values ofαthe number of available camp sites greatly de-creases.Furthermore,the%rejected and%o?-schedule visitors are greatly perturbed by increasingα.It can be concluded that for locations that are prone to factors that can render camp sites unusable should expect to have higher%rejected visitors and%o?-schedule visitors for a lower number of visitors total for any given Y.The e?ects of high alpha can be mediated in two ways:

1.The rafting service can build more camp sites to mask the e?ect ofα.

2.The rafting company can take measures to protect the environmental

integrity and maintenance of the camp sites.

10.2Mitigating Storms

In addition to the assumption that camp sites do not degrade or get damaged over time, our initial implementation was not designed explicitly to handle events that interrupt the normal?ow of visitors down the river.However,in the real world,happenings such as storms can interrupt this?ow and any good model must be able to account for these events.Therefore,we added a rule to implementation A that is called whenever?ow down the river is prevented.This rule prevents movement and the loading of new visitors for N number of days.For each day that this rule is active,each visitor has their D i incremented,e?ectively putting them behind schedule to re?ect the fact that they were unable to move to the next camp site that day.This rule allows for the simulation of any event that occurs at a?xed probability and lasts for a ran-dom number of days distributed according to a Poisson distribution of givenμ,whereμis the average length of the event.For testing purposes,we treated this random event as a storm with a?xed chance of occurring in any given trial and an average length of1 day.For4values ofλ,4di?erent probabilities of a storm occurring were tested at Y=110.

Based on the results of these additional calculations,we noted that for low values of the%chance of storm the model performed very well.The e?ect upon the number of visitors total is minimal.Furthermore,for values of%chance of storm5%,the %of o?schedule visitors is only marginally above the10%cut-o?for customer satisfac-tion.This is acceptable as visitors would likely understand that the weather is beyond the control of the rafting company.

11Conclusion

Using our dynamic model,we have demonstrated that it is possible to maintain high carrying capacities on the Big Long River while continuing to o?er a wide range of trip options and maximizing customer satisfaction.Furthermore,using this dynamic model, con?icts for camp sites can be eliminated ensuring that each rafting group has its own camp site.

In order to evaluate the model’s performance we de?ned a number of customer satisfaction metrics such as the percentage of visitors unable to book,the percentage of visitors that do not?nish the trip on schedule and the average number of boat to boat interaction over the entire season.

After analyzing the values of these metrics at varying values of lambda and Y we discovered that the only metric that limits the carrying capacity of the river is the percent of visitors that are unable to book.When Y?110,all the metrics indicate that customer satisfaction is su?ciently high,regardless of the value for lambda.

The carrying capacity of the river was de?ned to be the maximum number of visitors that could be served during a season.The carrying capacity was equal to1102.8and occurred

防火墙设备技术要求 一、防火墙参数要求： 1性能方面： 11网络吞吐量

防火墙设备技术要求 一、防火墙参数要求： 1.性能方面： 1.1网络吞吐量>10Gbps，应用层吞吐率>2Gbps，最大并发连接数>400万（性能要求真实可靠，必须在设备界面显示最大并发连接数不少于400万），每秒新建连接＞15万。 1.2万兆级防火墙，网络接口数量不少于12个接口，其中千兆光口不少于4个、千兆电口不少于6个、万兆光接口不少于2个，另外具有不少于2个通用扩展插槽。 1.3冗余双电源，支持HA、冗余或热备特性。 1.4具备外挂日志存储系统，用于存储防火墙日志文件等相关信息，另外支持不同品牌网络设备、服务器等符合标准协议的日志格式，日志存储数量无限制。 1.5内置硬盘，不小于600G，用于日志存储。 2.功能方面（包含并不仅限于以下功能）： 2.1具有静态路由功能，包括基于接口、网关、下一跳IP地址的静态路由功能。 2.2具有OSPF动态路由功能，符合行业通用的OSPF协议标准。 2.3具有网络地址转换功能，支持一对一、多对一、多对多的网络地址转换功能。

2.4具有OSI网络模型三层至七层访问控制功能，可基于IP、端口号、应用特征、数据包大小、URL、文件格式、内容、时间段等进行安全访问控制和过滤。 2.5具有基于资源和对象的流量分配功能，包括基于单个IP、网段、IP组、访问源地址及目标地址、应用等的流量管理、分配。 2.6完善的日志及审计系统，防火墙内所有功能均具备相应的日志可供查看和审计。 2.7具有内置或第三方CA证书生成、下发及管理功能。 2.8具有身份认证、身份审计功能，支持用户ID与用户IP 地址、MAC地址的绑定，支持基于CA证书、AD域、短信、微信等多种身份认证方式。 2.9具有入侵检测模块，支持入侵防护、DoS/DDoS防护,包含5年特征库升级服务。 2.10具有上网行为管理模块，支持上网行为检测、内容过滤等功能，包含5年应用特征库升级服务。 2.11具有病毒防护模块，可包含5年病毒库升级服务。 2.12具备基于WEB页面的管理、配置功能，可通过WEB 页面实现所有功能的配置、管理和实时状态查看。 2.13具有SSL VPN模块，含不低于50人的并发在线数。针对手机和电脑，有专门的SSL VPN客户端。 3.质保和服务

华为USG6000系列防火墙性能参数表

华为USG6000系列下一代防火墙详细性能参数表

能，与Agile Controller配合可以实现微信认证。 应用安全●6000+应用协议识别、识别粒度细化到具体动作，自定义协议类型，可与阻断、限流、审计、统计等多种手段自由结合在线协议库升 级。注：USG6320可识别1600+应用。 ●应用识别与病毒扫描结合，发现隐藏于应用中的病毒，木马和恶意软件，可检出超过500多万种病毒。 ●应用识别与内容检测结合，发现应用中的文件类型和敏感信息，防范敏感信息泄露。 入侵防御●基于特征检测，支持超过3500漏洞特征的攻击检测和防御。 ●基于协议检测，支持协议自识别，基于协议异常检测。 ●支持自定义IPS签名。 APT防御与沙箱联动，对恶意文件进行检测和阻断。 Web安全●基于云的URL分类过滤，支持8500万URL库，80+分类。 ●提供专业的安全URL分类，包括钓鱼网站库分类和恶意URL库分类。 ●基于Web的防攻击支持，如跨站脚本攻击、SQL注入攻击。 ●提供URL关键字过滤，和URL黑白名单。 邮件安全●实时反垃圾邮件功能，在线检测，防范钓鱼邮件。 ●本地黑、白名单，远程实时黑名单、内容过滤、关键字过滤、附件类型、大小、数量。 ●支持对邮件附件进行病毒检查和安全性提醒。 数据安全●基于内容感知数据防泄露，对邮件，HTTP，FTP，IM、SNS等传输的文件和文本内容进行识别过滤。 ●20+文件还原和内容过滤，如Word、Excel、PPT、PDF等），60+文件类型过滤。 安全虚拟化安全全特性虚拟化，转发虚拟化、用户虚拟化、管理虚拟化、视图虚拟化、资源虚拟化（带宽、会话等）。 网络安全●DDoS攻击防护，防范多种类型DDoS攻击，如SYN flood、UDP flood、ICMP flood、HTTP flood、DNS flood、ARP flood和ARP 欺骗等。 ●丰富的VPN特性，IPSec VPN、SSL VPN、L2TP VPN、MPLS VPN、GRE等。 路由特性●IPv4：静态路由、RIP、OSPF、BGP、IS-IS。 ●IPv6：RIPng、OSPFv3、BGP4+、IPv6 IS-IS、IPv6RD、ACL6。 部署及可靠性透明、路由、混合部署模式，支持主/主、主/备HA特性。 智能管理●支持根据应用场景模板生成安全策略，智能对安全策略进行优化，自动发现冗余和长期不使用的策略。 ●全局配置视图和一体化策略管理，配置可在一个页面中完成。 ●可视化多维度报表呈现，支持用户、应用、内容、时间、流量、威胁、URL等多维度呈现报表。 标准服务●USG6300-AC：SSL VPN 100用户。 ●USG6300-BDL-AC：IPS-AV-URL功能集升级服务时间12个月，SSL VPN 100用户。 可选服务●IPS升级服务：12个月/36个月 ●URL过滤升级服务：12个月/36个月●反病毒升级服务：12个月/36个月 ●IPS-AV-URL功能集：12个月/36个月 注：√表示为支持此项功能，—表示为不支持此项功能。

改性沥青防水卷材性能指标及执行标准

改性沥青防水卷材执行标准 SBS弹性体改性沥青防水卷材是以SBS（苯乙烯-丁二烯-苯乙烯）热塑性弹性体改性沥青为浸涂材料，以优质聚酯毡、玻璃纤毡、玻璃增强聚酯毡为胎基，以细砂、矿物粒料、PE膜、铝膜等为覆面材料，采用专用机械搅拌、研磨而成的弹性体改性沥青防水卷材。 APP塑性体改性沥青防水卷材是以APP（无规聚丙烯）或APAO、APO（聚烯烃类聚合物）改性沥青为浸涂材料，以优质聚酯毡、玻纤毡为胎基，以细砂、矿物粒（片）料、PE膜为覆面材料，采用先进工艺精制而成的塑性体改性沥青防水卷材。 凯鑫防水改性沥青防水卷材性能指标： SBS执行标准GB18242-2008 序号项目 ⅠⅡ PY G PY G PYG 1 可溶物含量 （g/㎡）≥ 3mm 2100 * 4mm 2900 * 5mm 3500 试验现象* 胎基不燃* 胎基不燃* 2 耐热性 ℃90 105 ≤mm 2 试验现象无流淌、滴落 3 低温柔度℃-20 -25 无裂纹 4 不透水性30min 0.3Mpa 0.2Mpa 0.3Mpa 5 拉力最大峰拉力 （N/50mm） ≥ 500 350 800 500 900 次高峰拉力 （N/50mm） ≥ * * * * 800 试验现象 拉伸过程中，试件中部无沥青涂盖层开裂或 与胎基分离现象 6 延伸率最大峰时延 伸率%≥ 30 * 40 * * 第二峰时延* * 15

伸率%≥ 7 渗油性张数≤ 2 APP执行标准GB18243-2008 序号项目 ⅠⅡ PY G PY G PYG 1 可溶物含量 （g/㎡）≥ 3mm 2100 * 4mm 2900 * 5mm 3500 试验现象* 胎基不燃* 胎基不燃* 2 耐热性 ℃110 130 ≤mm 2 试验现象无流淌、滴落 3 低温柔度℃ -7 -15 无裂纹 4 不透水性30min 0.3Mpa 0.2Mpa 0.3Mpa 5 拉力最大峰拉力 （N/50mm） ≥ 500 350 800 500 900 次高峰拉力 （N/50mm） ≥ * * * * 800 试验现象 拉伸过程中，试件中部无沥青涂盖层开裂或 与胎基分离现象 6 延伸率最大峰时延 伸率%≥ 30 * 40 * * 第二峰时延 伸率%≥ * * 15

防火墙技术参数

防火墙技术参数： 网络端口8GE+4SFP VPN支持支持丰富高可靠性的VPN特性，如IPSec VPN、SSL VPN、L2TP VPN、MPLS VPN、GRE等 入侵检测超过5000种漏洞特征的攻击检测和防御。第一时间获取最新威胁信息，准确检测并防御针对漏洞的攻击。可防护各种针对web的攻击，包括SQL注入攻击和跨站脚本攻击。管理预置常用防护场景模板，快速部署安全策略，降低学习成本 自动评估安全策略存在的风险，智能给出优化建议 支持策略冲突和冗余检测，发现冗余的和长期未使用策略，有效控制策略规模 一般参数 电源选配冗余电源100-240V，最大功率170W 产品尺寸442×421×44.4mm 产品重量7.9kg 适用环境温度：0-45℃(不含硬盘)/5℃-40℃ (包含硬盘) 湿度：5%-95%(不含硬盘)/ 5%-90% (包含硬盘) 其他性能产品形态：1U 带宽管理与QoS优化：在识别业务应用的基础上，可管理每用户/IP使用的带宽, 确保关键业务和关键用户的网络体验。管控方式包括：限制最大带宽或保障最小带宽、应用的策略路由、修改应用转发优先级等 数据防泄漏：对传输的文件和内容进行识别过滤，可准确识别常见文件的真实类型，如Word、Excel、PPT、PDF等，并对敏感内容进行过滤 应用识别与管控：可识别6000+应用，访问控制精度到应用功能，例如区分微信的文字和语音。应用识别与入侵检测、防病毒、内容过滤相结合，提高检测性能和准确率 接口扩展：可扩展千兆电口/千兆光口/万兆光口，支持BYPASS插卡 软件认证：ICSA Labs: Firewall，IPS，IPSec VPN，SSL VPN CC：EAL4+ NSS Labs：推荐级 硬件认证：CB，CCC，CE-SDOC，ROHS，REACH&WEEE(EU)，C-TICK，ETL，FCC&IC，VCCI，BSMI

防水材料取样标准与要求

SBS改性沥青防水卷材、APP改性沥青防水卷材: 同一厂家、规格不大于10000㎡为一批，外观合格卷材中，每批随机取 1卷。 聚氯乙烯（PVC）防水卷材、氯化聚乙烯防水卷材：同一厂家、规格不大于5000㎡为一批，任取 1卷，在距端部300mm处截取约3m作为试样。 三元乙丙卷材、橡胶防水卷材：同一厂家、规格不大于5000m2为一批，每批抽取 2m。 防水早已不再是一个新鲜的话题，不论是隧道、桥梁、堤坝、道路、建筑还是我们日常居住的房屋，防水都不容忽视，同时，我们日常生活中所使用的电器产品也存在必须直面的防水防潮问题。 用于防水工程的防水材料，其技术性能指标应符合相应的技术规范要求，如地下工程使用的SBS卷材应选用Ⅱ型，抽样检查的项目及技术指标应符合规范要求。这是工程对材料的要求。 防水材料标准的作用，针对生产厂家，提供相应材料的技术指标要求及检测判断方法，针对施工现场材料检验，提供检测及判断方法。 从严格意义来说，技术指标要求应符合规范规定，检测方法及判定按材料标准进行。 三、取样要求 1、石油沥青纸胎油毡 批量：以同一品种、标号、等级的产品每1500卷为一批，不足1500卷者亦按一批计。 抽样：在每批产品中抽取10卷进行卷重检验，全部达到规定时即为合格。在重量检查合格的10卷中取重最轻的，外观面积合格的无接头的一卷作为物理性能试样，若最轻的一卷不符合抽样条件时，可取次轻的一卷，切除距外层卷头2.5m后，顺纵向截取0.5m长的全幅卷材两块，一块作物理力学性能用，另一块备用。 2、弹性体沥青防水卷材 批量：以同一类型、同一规格10000m2为一批，不足10000m2时亦可作为一批。 抽样：在每批产品中随机抽取5卷进行卷重、面积、厚度与外观检查，在上述检查合格的卷材中随机抽取1卷，切除距外层卷头2.5m后。顺纵向切取长度为0.8m的全幅卷材两块，一块用物理力学性能检测用，另一块备用。

防火墙技术指标

技术要求采购数量：1台

服务要求 1、提供标的产品免费全国范围内现场安装服务和售后服务； 2、提供厂商出具的标的产品三年硬件保修服务证明，备品备件保障证明； 3、投标人至少有2名以上工程师，并指定专人工程师为项目接口工程师，提供标的产品的技术服务支持； 4、硬件故障处理 a)乙方提供三年产品保修服务。若出现硬件故障问题，从甲方报修开始至厂维修完毕返还给用户应不超过10个工作日，维修产品的运费及运输保险费由乙方承担； 5、技术咨询服务 b）乙方免费为用户提供产品咨询服务，协助用户进行新需求规划； c）甲方在系统相关环境上进行研发测试过程中，遇到技术难题时，提供技术咨询服务，包括远程电话支持和现场研讨支持； 6、系统配置管理 d) 建立所有维保设备的配置统计文档，在维保期间随时更新设备硬件或系统软件配置变更的情况。 7、不间断服务 e) 7×24小时响应的电话、传真、E-MAIL方式等日常售后技术服务；提供7×24小时的支持服务，设备出现故障进行实时电话响应； f）设备出现故障，如电话、远程等方式不能解决，全国范围内6小时内赶到现场，12小时内排除由设备问题造成的网络故障； 8、疑难问题升级处理 g) 就维保设备在用户日常管理中遇到的疑难杂症进行升级处理，疑难问题不限于设备故障； 9、定期巡检 h) 提供产品定期巡检服务：乙方在服务期内提供12人次/年的定期巡检服务（每月1次），对本次合同购买的设备的运行状态、安全策略等进行检测，确保其安全稳定的运行，巡检后3个工作日内向甲方提交检查总结报告； 10、系统升级、补丁安装服务

i)提供三年软件版本及补丁免费升级服务，特征库升级授权，供用户自行升级以及提供技术支持服务； 11、系统及相关环境重建服务 j) 设备维保期间，提供系统及相关环境的重新搭建服务； 12、重要事件服务 k）服务期内若甲方有特殊需求（如网络架构调整需求等），乙方技术支持工程师须在甲方提出需求的8个工作小时内到达甲方现场配合讨论方案和进行现场配置和调试； 13、紧急处理服务 l）对于紧急支持服务需求（例如系统瘫痪等严重问题），乙方须在接到甲方服务要求后2个小时内作出反应，在4个小时内到达甲方现场； m）服务期内，若因设备硬件/系统问题影响甲方正常生产环境，乙方须在4个小时内调拨备机到甲方现场，并确保备机的策略/运行状态正常，提供的备机服务应符合现在管理平台的要求，提供的备机应为同等档次或高于目前使用的型号； 14、硬件设备移机服务 n) 根据甲方实际需求提供硬件设备的物理搬迁服务。在搬迁过程中，提供相当于搬迁设备同样配置的备机在甲方本地(单次搬迁设备超过4台时，提供一半以上的备机，保证对故障设备进行及时替换)，保障设备移机前后的正常运行；搬迁过程中如设备有任何损坏，其造成的损失在得到甲方认可的前提下，全部由乙方负责赔偿； 15、辅助故障定位服务 o) 在甲方使用主流品牌的硬件、软件产品出现兼容性问题时，积极配合，与有关硬件、软件厂商和采购人接洽，及时定位问题原因、寻求解决方案； 16、产品关联服务 p) 提供其他与产品相关联的服务，如产品过期EOL(End of life)，提前一年通知甲方。 17、培训服务 q）提供标的产品2人/次原厂技术培训（非现场培训）； 18、续约 r) 维保服务合同到期后1个月内，如果甲方提出需求，继续提供上述技术服务；18、提供厂商出具的五年内备品备件保障证明；

防火墙的性能评估

评价防火墙的功能、性能指标 (2011-06-03 23:47:16) 转载▼ 标签： 分类：网络技术 防火墙 性能 参数 吞吐量 最大连接数 新建连接数 丢包率 it 一、功能指标 1、访问控制：根据数据包的源/目的IP地址、源/目的端口、协议、流量、时间等参数对数据包进行访问控制。 2、地址转换：源地址转换(SNAT)、目的地址转换(DNAT)、双向地址转换(IP映射)。 3、静态路由/策略路由： 静态路由：给予目的地址的路由选择。 策略路由：基于源地址和目的地址的策略路由选择。 4、工作模式：路由模式、网桥模式(交换/透明模式)、混杂模式(路由+网桥模式并存) 5、接入支持：防火墙接口类型一般有GBIC、以太网接口等；接入支持静态IP设置、DHCP、PPOE(比如ADSL接入)等。 6、VPN：分为点到端传输模式(PPTP协议)和端到端隧道模式(IPSec、IPIP、GRE隧道)，支持DES、3DE、Blowfish、AES、Cast128、Twofish等加密算法，支持MD5、SHA-1认证算法；VPN功能支持NAT穿越。 7、IP/MAC绑定：IP地址与MAC地址绑定，防止IP盗用，防止内网机器有意/无意抢占关键服务器IP。 8、DHCP：内置DHCP Server 为网络中计算机动态分配IP地址；DHCP Relay的支持能为防火墙不同端口的DHCP Server和计算机之间动态分配IP地址。 9、虚拟防火墙：在一台物理防火墙设备上提供多个逻辑上完全独立的虚拟防火墙，每个虚拟防火墙为一个特定的用户群提供安全服务。 10、应用代理：HTTP、FTP、SMTP等协议应用代理，大多数内容过滤通过应用代理实现。 11、流量控制：流量控制，流量优先级，带宽允许条件下的优先保障关键业务带宽。

衡量防火墙性能的参数指标有哪些

衡量防火墙性能的参数指标有哪些 导读：我根据大家的需要整理了一份关于《衡量防火墙性能的参数指标有哪些》的内容，具体内容：防火墙借由监测所有的封包并找出不符规则的内容，可以防范电脑蠕虫或是木马程序的快速蔓延。对于防火墙的性能，我们应该怎么衡量比较呢?这篇文章主要介绍了衡量防火墙性能的几个重要参数指标,需要的... 防火墙借由监测所有的封包并找出不符规则的内容，可以防范电脑蠕虫或是木马程序的快速蔓延。对于防火墙的性能，我们应该怎么衡量比较呢?这篇文章主要介绍了衡量防火墙性能的几个重要参数指标,需要的朋友可以参考下 防火墙主要参考以下3种性能指标： 整机吞吐量：指防火墙在状态检测机制下能够处理一定包长数据的最大转发能力，业界默认一般都采用大包衡量防火墙对报文的处理能力。 最大并发连接数：由于防火墙是针对连接进行处理报文的，并发连接数目是指的防火墙可以同时容纳的最大的连接数目，一个连接就是一个TCP/UDP的访问。 每秒新建连接数：指每秒钟可以通过防火墙建立起来的完整TCP/UDP连接。该指标主要用来衡量防火墙在处理过程中对报文连接的处理速度，如果该指标低会造成用户明显感觉上网速度慢，在用户量较大的情况下容易造成防火墙处理能力急剧下降，并且会造成防火墙对网络攻击防范能力差。

并发连接数 并发连接数是指防火墙或代理服务器对其业务信息流的处理能力，是防火墙能够同时处理的点对点连接的最大数目，它反映出防火墙设备对多个连接的访问控制能力和连接状态跟踪能力，这个参数的大小直接影响到防火墙所能支持的最大信息点数。 并发连接数是衡量防火墙性能的一个重要指标。在目前市面上常见防火墙设备的说明书中大家可以看到，从低端设备的500、1000个并发连接，一直到高端设备的数万、数十万并发连接，存在着好几个数量级的差异。那么，并发连接数究竟是一个什么概念呢?它的大小会对用户的日常使用产生什么影响呢? 要了解并发连接数，首先需要明白一个概念，那就是"会话"。这个"会话"可不是我们平时的谈话，但是可以用平时的谈话来理解，两个人在谈话时，你一句，我一句，一问一答，我们把它称为一次对话，或者叫会话。同样，在我们用电脑工作时，打开的一个窗口或一个Web页面，我们也可以把它叫做一个"会话"，扩展到一个局域网里面，所有用户要通过防火墙上网，要打开很多个窗口或Web页面发(即会话)，那么，这个防火墙，所能处理的最大会话数量，就是"并发连接数"。 像路由器的路由表存放路由信息一样，防火墙里也有一个这样的表，我们把它叫做并发连接表，是防火墙用以存放并发连接信息的地方，它可在防火墙系统启动后动态分配进程的内存空间，其大小也就是防火墙所能支持的最大并发连接数。大的并发连接表可以增大防火墙最大并发连接数，允许防火墙支持更多的客户终端。尽管看上去，防火墙等类似产品的并

TPO防水卷材的材料性能指标应满足下表规定

TPO防水卷材的材料性能指标应满足下表规定 序号项目指标 H L P 1 中间织物上面树脂层厚度/mm≥﹣0.4 2 拉伸 性能最大拉力/N/cm≥﹣200 250 拉伸强度/Mpa≥12.0 ﹣﹣最大拉力时伸长率/≥﹣﹣15 断裂伸长率%/≥500 250 ﹣ 3 热处理尺寸变化率/%≤ 2.0 1.0 0.5 4 低温弯折性﹣40℃ 无裂纹 5 不透水性0.3Mpa，2h不透水 6 抗冲击性能0.5kg*m,不渗水 7 抗静态荷载﹣﹣20kg不渗水 8 接缝剥离强度/（N/mm） 4.0或卷材破坏 3.0 9 搭接强度/N≥﹣290 10 直角撕裂强度/N/mm 60 ﹣﹣ 11 梯形撕裂强度/N/mm ﹣245 450 12 吸水率（70℃，168h）/%≤ 4.0 3.0 13 热老化 （115℃）时间672h 外观无起泡、裂纹、分层、粘结和孔洞 最大拉力保持率/%≥﹣90 拉伸强度保持率/%≥90 ﹣﹣最大拉力时伸长率保持率/%≥﹣﹣90 断裂伸长率保持率/%≥90 90 ﹣低温弯折性﹣40℃ 无裂纹 14 耐化学性外观无起泡、裂纹、分层、粘结和孔洞 最大拉力保持率/%≥﹣90 90 拉伸强度保持率/%≥90 ﹣﹣最大拉力时伸长率保持率/%≥﹣﹣90 断裂伸长率保持率/%≥90 90 ﹣

低温弯折性﹣40℃ 无裂纹 15 人工气候 加速老化时间1500h 外观无起泡、裂纹、分层、粘结和孔洞 最大拉力保持率/%≥﹣90 90 拉伸强度保持率/%≥90 ﹣﹣最大拉力时伸长率保持率/%≥﹣﹣90 断裂伸长率保持率/%≥90 90 ﹣低温弯折性﹣40℃ 无裂纹 a 抗静态荷载仅对用于压铺屋面的卷材要求 b 单层卷材屋面使用产品的人工气候加速老化时间为2500h 尺寸偏差--厚度允许偏差 厚度/mm 允许偏差/% 最小单值/mm 1.20 -5，+10 1.05 1.50 -5，+10 1.35 1.80 -5，+10 1.65 2.00 -5，+10 1.85 规格： 宽度：2.0m/2.05 m/3.05m 长度：20 m /30.5m 或其他长度 厚度：1.2mm 1.5mm 1.8mm 2.0mm 5发展情况

互联网出口防火墙性能指标

互联网出口防火墙性能指标 标准1U机箱，标配4个千兆MBASE-T接口，2个千兆SFP 接口；整机吞吐率(bps)≥8G，最大并发连接数≥180万，每秒新建连接数≥8万，内置2对电口BYPASS。1. 具备2-7层的网络安全防护功能，具备防火墙、APT检测、VPN、IPS、WAF、网页防篡改、WEB风险扫描等功能模块；2. 要求能够实现对蠕虫/木马/后门/DoS/DDoS攻击探测/扫描/间谍软件/利用漏洞的攻击/缓冲区溢出攻击/协议异常/ IPS逃逸/VPN隧道攻击等的防护，支持安全防护策略智能联动，可智能生成临时规则封锁攻击源IP，临时封锁时间可自定义，要求提供设备界面截图证明；3. 漏洞特征库: 4000+，支持自动或者手动升级，需提供设备界面截图证明；具备专业攻防团队每1-2周进行特征库和紧急漏洞更新，需提供官网更新数据，处理动作支持“云联动”，支持自动拦截、记录日志、上传灰度威胁到“云端”，提高分析检测能力，提供设备界面截图证明；4. 要求能够实现SQL注入、XSS攻击、WEBSHELL攻击、CSRF跨站请求伪造、系统命令注入、文件包含攻击、目录遍历攻击和信息泄露攻击等的等web攻击防护，为保障功能可靠性，★要求提供OWASP认证和NSS Labs测试报告证明；5. 要求支持APT攻击检测，实现已感染病毒、木马、蠕虫、僵尸网络等终端的外发恶意流量阻断，保障内网终端安全，需提

供设备界面截图证明；6. 要求支持服务器、客户端的漏洞风险评估功能，能够对目标IP进行端口、服务扫描，同时支持多种应用的弱口令评估与扫描，要求提供设备界面截图证明；支持手动/定时的web风险扫描功能，预定义快速、完整的扫描模板，且支持自定义扫描模板；7. 为保障软件成熟度及软件可靠性，要求厂家为微软MAPP计划合作伙伴并具备国家互联网应急响应中心应急服务支撑单位资质，要求所投品牌厂商是公安部第二代防火墙标准起草单位之一；8. 支持脚本过滤和ActiveX过滤，并能识别并封堵含有恶意插件的网络访问行为，必须能识别并封堵含有恶意脚本、挂载木马等危险网页访问行为，要求提供自主知识产权证明，支持恶意链接检测功能，内置恶意链接地址库，对于可疑的地址链接，设备能够同云端安全分析引擎进行联动，可疑威胁行为在云端进行沙盒执行检测并返回行为分析报告，需提供设备分析报告截图；9. 要求支持网关型网页防篡改，篡改检测需要支持精确匹配和模糊匹配的方式，检测到篡改能提供短信、邮件等报警方式，要求提供设备界面截图证明；10. 要求支持双向内容检测，内置数据泄密防护识别库，能够识别身份证、手机号码、银行卡号、邮箱等敏感信息，并且阻止此类敏感信息被泄密，需提供设备界面截图证明；11. 支持针对重要业务系统的管理员后台URL/Telnet/SSH登录页面扩展短信强认证的功能，防止口令爆破或是社会工程学攻击，要求提供设备界面截图证明

弹性体改性沥青防水卷材材料性能指标

弹性体改性沥青防水卷材材料性能指标 1、分类 (1)、按胎基分为聚酯毡（PY）、玻纤毡（G）、玻纤增强聚酯毡（PYG）。 (2)、按上表面隔离材料分为聚乙烯膜（PE）、细砂（S）、矿物粒料（M）。按下表面隔离材料分为聚乙烯膜（PE）、细砂（S）。 (3)、按材料性能分为Ⅰ型和Ⅱ型。 2、规格 (1)、卷材公称宽度为1000mm。 (2)、聚酯毡卷材公称厚度为3mm、4mm、5mm。 (3)、玻纤毡卷材公称厚度为3mm、4mm。 (4)、玻纤增强聚酯毡卷材公称厚度为5mm。 (5)、每卷卷材公称面积为7.5m2、10 m2、15 m2。 3、标记 弹性体改性沥青防水卷材产品按名称、型号、胎基、上表面材料、下表面材料、厚度、面积和标准编号顺序标记。如10㎡面积、3mm厚，上表面为矿物粒料、下表面为聚乙烯膜聚酯毡I型弹性体改性沥青防水卷材标记为：SBS I PY M PE 3 10 GB18242—2008。 4、用途 (1)、弹性体改性沥青防水卷材主要适用于工业和民用建筑的屋面和地下防水工程。（2）、玻纤增强聚酯毡卷材可用于机械固定单层防水，但需通过抗风载试验。 （3）、玻纤毡卷材适用于多层防水中的底层防水。 （4）、外露使用采用上表面隔离材料为不透明的矿物粒料的防水卷材。 （5）、地下工程防水采用表面隔离材料为细砂的防水卷材。 5、单位面积质量、面积及厚度。弹性体改性沥青防水卷材的单位面积质量、面积及厚度应 （1）、成卷卷材应卷紧卷齐，端面里进外出不得超过10mm。 （2）、成卷卷材在（4～50）℃任一产品温度下展开，在距卷芯1000mm长度外不应有10mm 以上的裂纹或粘结。 （3）、胎基应浸透，不应有未被浸渍处 （4）、卷材表面应平整，不允许有孔洞、缺边和裂口、疙瘩，矿物粒料粒度应均匀一致并紧

防火墙的关键参数

2) 连接数评估 连接在状态防火墙中是一个很重要的概念，与连接相关的性能指标对评估防火墙非常重要。这些指标包括并发连接数、新建连接速率。 l 并发连接数的测试 并发连接是一个很重要的指标，它主要反映了被测设备维持多个会话的能力。关于此指标的争论也有很多。一般来说，它是和测试条件紧密联系的，但是这方面的考虑有时会被人们忽略。比如，测试时采用的传输文件大小就会对测试结果有影响。例如，如果在传输中应用层流量很大的话, 被测设备将会占用很大的系统资源去处理包检查，导致无法处理新请求的连接，引起测试结果偏小；反之测试结果会大一些。所以没有测试条件而只谈并发连接数是难以定断的。从宏观上来看，这个测试的最终目的是比较不同设备的“资源”，也就是说处理器资源和存储资源的综合表现。 目前市场上出现了大家盲目攀比并发连接数的情况。事实上，并发几十万的连接数应该完全可以满足一个电信级数据中心的网络服务需求了，对于一般的企业来讲, 甚至几千个并发连接数还绰绰有余。并发连接总数能由仪表自动测试得出结果，减少了测试所用的时间和人力，这类仪表目前很多，常见的有Spirent的 Avalanche、IXIA的IxLoad以及BPS等。 l 新建连接速率 这个指标主要体现了被测设备对于连接请求的实时反应能力。对于中小用户来讲，这个指标显得更为重要。可以设想一下，当被测设备可以更快的处理连接请求，而且可以更快传输数据的话,网络中的并发连接数就会倾向于偏小，从而设备压力也会减小，用户感受到的防火墙性能也就越好。Avalanche、IXLOAD以及BPS等测试工具都可以测试新建连接速率，帮助使用者搜索到被测设备能够处理的峰值，测试原理基本都是相同的。 2. 模拟真实应用环境进行性能指标测试 如果能够100%模拟用户的实际应用环境对防火墙性能进行测试，那么防火墙选型这类活动将变得非常简单，而且防火墙性能指标将变得更加有意义。但是模拟真实应用环境并不是简单的事情。主要是因为用户环境的复杂性和多变性导致真实环境的模拟几乎不可能实现。这里讨论的模拟真实应用环境测试，只是将用户环境进行抽象，使得模拟环境在满足测试条件的情况下最大限度的贴近真实应用环境。 1) 多应用协议吞吐量测试 前面提到goodput是衡量防火墙吞吐量的重要指标，基线测试中，一般采用HTTP协议作为应用层协议进行测试。而在实际应用环境中，应用层的流量并不是纯粹的HTTP，还有其他协议。如果用HTTP协议代替其他应用层协议测试应用层吞吐量，显然是不合适的。因此需要针对不同的应用场景，设计典型的应用层流量分布模型，按照不同的比例分配带宽。如图3所示，是一个典型的某场景应用带宽分布。

各种防水材料检测指标要求

各种防水材料检测指标 要求 公司标准化编码 [QQX96QT-XQQB89Q8-NQQJ6Q8-MQM9N]

各种防水材料指标要求 一、防水混凝土膨胀剂： 执行《地下工程防水技术规范》（GB50108-2008）与《混凝土膨胀剂》（JC476- ⑵、检验时A、B两法均可使用，仲裁检验采用A法。抗折和抗压强度均为A法数据； ⑶、增加3d、14d限制膨胀率的测定龄期，用于限制膨胀率规律的判定，水中养护期间，限制膨胀率的规律应满足28d＞14d＞7d＞3d，并且组分中不得添加其他外加剂如减水剂； ⑷、“※”为非规范中项目，但必检，以证实防水混凝土无后期收缩； ⑸、掺量为8～12%，检测方法见《混凝土膨胀剂》（JC476-2001）； ⑹、检测报告应由省级以上检测机构出示； ⑺、应按《混凝土外加剂应用技术规范》（GB50119）附录B的方法对各搅拌站供应的混凝土的限制膨胀率性能指标进行抽检。 二、单面自粘复合高分子防水卷材： 执行《地下工程防水技术规范》（GB50108-2008）与《高分子防水材料第一部分片材》（）标准要求； 1、材料组成：单（挂）面无纺布≥400g/m2，中间防水板厚≥1.0mm，总厚度≥1.8mm （不含无纺布厚度），并且厚度偏差≤-10%； 2、卷材幅宽≥2.0m，便于隧道施工，减少搭接；

三、自粘聚合物改性沥青防水卷材的物理力学性能指标：

四、钢边止水带主要物理力学性能指标： 执行《铁路隧道设计规范》（TB10003-2005）与《高分子防水材料第二部分止水 五、水泥基渗透结晶型防水涂料主要物理力学性能指标：

六、遇水膨胀止水胶主要物理力学性能指标： 执行《地下工程防水技术规范》（GB50108-2008）及《轨道交通地下工程防水技术规 七、不锈钢弹簧管骨架可重复注浆管主要物理力学性能指标： 八、400g/m2的短纤维针刺非织造土工布（无纺布缓冲层）性能指标：

防水材料性能表

防水卷材 1防水卷材的分类、基本特点及适用范围见表1 卷材的分类、基本特点、适用范围表

高聚物改性沥青防水卷材的主要规格尺寸 2．3 低密度聚乙烯 (LDPE)土工膜、 乙烯一醋酸乙 烯(EV A)土工膜 以低密度聚乙烯树脂或乙烯一醋酸乙烯树脂为主要原料，添加多种化学助剂，经 造粒和吹塑成型等工序加工制成可卷曲的防渗材料。该材料具有拉伸强度较高、 延伸率较大，可焊性好、耐腐蚀、耐根系刺穿性能优良等特点，适用于一般和中 档的隧道、洞库、堤坝等土木建筑工程作防水层，也适用于种植屋面作耐根系刺 穿的防水层。 2 防水卷材的性能指标及规格 2．1 防水卷材性能指标见各自相应的执行标准。 2．2 高聚物改性沥青防水卷材的主要规格尺寸 见表 2．2 2．3 合成高分子防水卷材的主要规格尺寸 见表 2．3 合成高分子防水卷材的主要性能指标及规格尺寸 表 3 建筑设计选用要点 3．1 SBS 改性沥青防水卷材的选用要点 1)Ⅰ型的聚酯胎或玻纤胎 SBS 改性沥青防水卷材，具有一定的拉力，且低温柔度较好， 适用于一般和较寒冷地区一般建筑工程作屋面的防水层。单层使用时，厚度不应小于 4mm ，双层使用时，每层厚度不应小于 3mm 。当采用板岩片(彩砂)或铝箔覆面的卷材作外

露屋面防水层时，无需另作保护层；若采用聚乙烯或细砂等覆面的卷材作外露屋面防水层时，必须涂刷耐老化性能好的浅色涂料或铺设块材、抹水泥砂浆、浇筑细石混凝土等作保护层。 2)Ⅱ型的聚酯胎SBS改性沥青防水卷材，具有拉力较高，低温柔度好，延伸率较大，以及耐腐蚀、耐霉烂等性能，适用于一般和寒冷地区的中、高档建筑作屋面或地下工程的防水层。单层使用时，厚度不应小于4mm，若对防水要求较高时，应选用双层卷材作防水层，且每层卷材厚度不应小于3mm。 3)Ⅱ型的玻纤胎SBS改性沥青防水卷材，具有一定的拉力，低温柔度较好以及耐腐蚀、耐霉烂等性能，但无延伸率，仅适用于一般和寒冷地区且结构刚度好的一般建筑作屋面或地下工程的防水层。单层使用时，厚度不应小于4mm，双层使用时，每层卷材厚度不应小于3mm，并可采用一层玻纤胎和一层聚酯胎的SBS改性沥青卷材作复合防水层。 4)厚度小于3mm的SBS改性沥青防水卷材，严禁采用热熔法施工。 3．2APP(APAO)改性沥青防水卷材的选用要点 1)Ⅰ型的聚酯胎或玻纤胎APP(APAO)改性沥青防水卷材，具有耐热度较高和耐腐蚀、耐霉烂等性能，但低温柔度较差，适用于非寒冷地区作一般建筑工程的屋面防水层。其它要求与Ⅰ型SBS卷材的内容相同。 2)Ⅱ型的聚脂胎APP(APAO)改性沥青防水卷材，具有拉力较大，延伸率较高，耐热度高、低温柔度较好和耐腐蚀、耐霉烂等性能，适用于一般和较寒冷地区的中、高档建筑作屋面、地下或道桥工程的防水层。其它要求与Ⅱ型SBS卷材的内容相同。 3)Ⅱ型的的玻纤胎APP(APAO)改性沥青防水卷材，具有一定的拉力，低温柔度较好，耐热度高和耐腐蚀、耐霉烂等性能，但无延伸率，适用于一般和较寒冷地区且结构刚度好的一般工程作屋面或地下工程的防水层。其它要求与Ⅱ型SBS卷材的内容相同。 4)厚度小于3mm的APP(APAO)改性沥青防水卷材，严禁采用热熔法施工。 3．3三元乙丙橡胶防水卷材的选用要点 三元乙丙橡胶(EPDM)防水卷材，具有拉伸强度较高、延伸率大，对基层伸缩或开裂变形的适应性强等特点，适用于炎热或严寒地区的中、高档建筑作屋面或地下工程的防水层。单层使用时，厚度不应小于1．5mm，双层使用时，每层厚度不应小于1．2mm。施工时，必须选用与该卷材相容且与其配套的专用胶粘剂、胶粘带、密封材料等进行卷材接缝的粘结密封处理。胶粘剂的粘结剥离强度浸水168h后，其保持率应大于70％，否则不允许采用。对施工完的防水层，必须及时做好成品保护。 3．4氯化聚乙烯——橡胶共混防水卷材的选用要点 该卷材具有耐高、低温性能较好，拉伸强度较高、延伸率较大，对基层伸缩或开裂变形的适应性较强等特点，适用于一般和寒冷地区的一般或中档建筑作屋面或地下工程的防水层。其它要求与3．3的内容相同。 3．5氯化聚乙烯(CPE)防水卷材的选用要点 该卷材具有拉伸强度较高、延伸率较大、耐腐蚀性能较好和便于粘结等特点，适用于一般地区的一般建筑作屋面或地下工程的防水层。其它要求与3．3的内容基本相同。3．6聚氯乙烯(PVC)防水卷材的选用要点 该卷材具有拉伸强度高、延伸率较大、耐腐蚀和耐刺穿性能较好，其接缝可进行焊接施工等特点，适用于一般地区的一般或中档建筑作屋面或地下工程的防水层，也可用作种植屋面作耐根系刺穿的防水层。单层使用时，卷材厚度不应小于1．5mm。 3．7高密度聚乙烯(HDPE)土工膜的选用要点 该材料具有拉伸强度很高、延伸率大、耐腐蚀和耐刺穿性能很好，其接缝可进行焊接施工等特点，适用于中、高档建筑的地下工程作初期支护与二次衬砌混凝土和种植屋面防

硬件防火墙六大关键指标

硬件防火墙关键指标 1、吞吐量 2、时延 3、扩展性 4、并发连接数 5、丢包率 6、最大安全策略数 一、网络吞吐量 当CIO在企业中部署了企业级别的防火墙之后，企业进出互联网的所有通信流量都要通过防火墙，故对于防火墙的吞吐量就有比较高的要求。以前有些企业是通过ADSL拨号上网的，这时由于带宽的限制，可能防火墙还可以应付。可是现在大部分企业可能已经都采用了光纤接入，带宽本来就很大。此时就给防火墙带来了一定的压力。 因为防火墙是通过对进入与出去的数据进行过滤来识别是否符合安全策略的，所以在流量比较高时，要求防火墙能以最快的速度及时对所有数据包进行检测。否则就可能造成比较长的延时，甚至发生死机。所以网络吞吐量指标非常重要，它体现了防火墙的可用性能，也体现了企业级别的防火墙的重要性。如果资金充裕，那么吞吐量当然是越大越好。吞吐量一个基本的原则就是至少要跟企业现有的互联网接

入带宽相当。 二、协议的优先级（时延） 第二个指标就是这个协议的优先性。现在视频在实际应用中越来越广泛。如视频会议系统、语音电话等等都很普及，而这些应用都会占用比较大的带宽。但增加带宽需要花费比较大的投资。最理想的解决方案是对企业的通信流量进行管理，通过防火墙把一些关键应用的流量设置为比较高的优先级。在网络传输中，要首先保障这些通信流量能够优先通过。这就可以明显改善语音通话等视频应用的效果。 另外这还可以用来约束员工的网络行为。如有些员工喜欢利用emule等工具下载电影。但是这些工具的话会占用很大的带宽，因为他们在从网络上下载的同时，也提供别人进行下载。故耗用的带宽比较多。 三、具有一定的扩展性 网络不可能永远的一成不变。随着政府服务型网站的规模扩大，网络会不断的升级，各地各部门之间要开通业务网络互联互联，成为网络平台的一个节点，此时就遇到一个问题，如何把各单位的网络与政府的网络连接起来。为此通过VPN来连接无疑是一个不错的方案。但是现有的防火墙能否支持VPN技术呢?企业很有可能以前在选购防火墙的时候没有注意到这个问题。

防水卷材施工规范

内容提要 《地下工程防水技术规范》在修编过程中，修编组经过广泛地调查研究和收集资料，在总结我国地下工程防水近年来实践经验的基础上，参考有关国际标准，并广泛征球全国有关单位的意见，对《地下防水工程施工及验收规范》(GBJ208-83)、《地下工程防水技术规范》(GBJ108-87)中设计、施工方面的内容进行了修订。这次修订的主要内容有：在整体结构上，按地下工程结构主体防水、细部构造防水、排水的思路重新划分章节，进行改写;在设计内容中，增加了对防水等级标准进行量化、对采用不同施工方法的地下工程制定相应防水设计方案等内容;对常用防水方法和材料进行了较大的修改，增加了“塑料防水板防水层”的内容，对选用的材料提出了相应的技术性能指标，对防水混凝土抗渗等级的选取提出了新的规定;在细部构造防水内容中，增加了“预留通道接头”、“桩头”的防水做法，对变形缝的设计、施工补充了有关内容;增加了“逆筑结构”有关防水做法。经修订，原规范(GBJ108-87)10章32节179条现为10章36节285条，这将为保证地下工程防水质量发挥重要作用。 1 总则 2 术语 3 地下工程防水设计 4 地下工程混凝土结构主体防水 5 地下工程混凝土结构细部构造防水 6 地下工程排水 7 注浆防水 8 特殊施工法的结构防水 9 地下工程渗漏水治理 10 其他规定 附录A 安全与环境保护 经中华人民共和国住房和城乡建设部第172号公告批准，《地下工程防水技术规范》（GB50108-2008）为新的国家标准。该标准将于2009年4月1日起实施。其中，第3.1.4、3.2.1、3.2.2、4.1.22、4.1.26、(1.2)、

防火墙的性能指标

防火墙的性能指标 吞吐量是衡量一款防火墙或者路由交换设备的最重要的指标，它是指网络设备在每一秒内处理数据包的最大能力。吞吐量意味这台设备在每一秒以内所能够处理的最大流量或者说每一秒内能处理的数据包个数。设备吞吐量越高，所能提供给用户使用的带宽越大，就像木桶原理所描述的，网络的最大吞吐取决于网络中的最低吞吐量设备，足够的吞吐量可以保证防火墙不会成为网络的瓶颈。举一个形象的例子，一台防火墙下面有100个用户同时上网，每个用户分配的是10Mbps的带宽，那么这台防火墙如果想要保证所有用户全速的网络体验，必须要有至少1Gbps的吞吐量。 吞吐量的计量单位有两种方式：常见的就是带宽计量，单位是Mbps(Megabits per second)或者Gbps (Gigabits per second)，另外一种是数据包处理量计量，单位是pps (packets per second)，两种计量方式是可以相互换算的。在进行对一款设备进行吞吐性能测试时，通常会记录一组从64字节到1518字节的测试数据，每一个测试结果均有相对应的pps数。64字节的pps数最大，基本上可以反映出设备处理数据包的最大能力。所以从64字节的这个数，基本上可以推算出系统最大能处理的吞吐量是多少。 很多路由设备的性能指标有一点就是标称xxMpps，所指的就是设备处理64字节的pps数。比如64字节的pps为100000pps，吞吐量通过换算为100000×(64+20)×8/1000000=67.2Mbps，拿这个结果计算1518字节的数据为100000×(1518+20)×8/100000=1230.4Mbps。其中的20字节是指12字节的帧间距(IPG)以及8字节的前导码(7字节同步+1字节起始)，测试每一个字节的吞吐量都需要将这20字节计算在内。通过前面的算式可以看出，我们即使不测试1518字节，也能够大致推算出设备最大的吞吐量是多少。但最终的结果只能小于这个结果。为什么会小于呢?因为很多设备因为接口数或者内部器件带宽的原因，限制了最大的带宽，这样设备的最大吞吐量就会远远低于推算的数值。不过既然得出了64字节的pps数，那么只要厂商标称的最大吞吐量低于推算的带宽，就基本可以认为这个标称值是可信的。 时延(Latency) 时延是系统处理数据包所需要的时间。防火墙时延测试指的就是计算它的存储转发(Store and Forward)时间，即从接收到数据包开始，处理完并转发出去所用的全部时间。在一个网络中，如果我们访问某一台服务器，通常不是直接到达，而是经过大量的路由交换设备。每经过一台设备，就像我们在高速公路上经过收费站一样都会耗费一定的时间，一旦在某一个点耗费的时间过长，就会对整个网络的访问造成影响。如果防火墙的延时很低，用户就完全不会感觉到它的存在，提升了网络访问的效率。

防水材料性能指标

精心整理聚酯胎SBS改性沥青防水卷材主要物理性能 项目聚酯毡胎 Ⅰ型Ⅱ型 可溶物含量（g/㎡）≥3mm厚2100 4mm厚2900 不透水性压力(Mpa）≥0.3 保持时间（min）≥ 30 耐热度（℃）90 105 无滑动、流淌、滴落 拉力（N/50mm）≥纵向 450 800 横向 最大拉伸时延伸率（%）≥纵向 30 40 横向 低温柔度（℃）-18 -25 无裂纹 聚乙烯丙纶复合防水卷材主要物理性能 项目指标 树脂类 FS1 FS2 断裂拉伸强度/(N/cm) 常温≥100 60 60℃≥40 30 扯断伸长率/% 常温≥150 —20℃≥10 10 撕裂强度/N≥20 20 不透水性(0.3MPa,50min) 无渗漏无渗漏低温弯折温度/℃≤—30 —20 加热伸缩量/mm 延伸≤ 2 2 收缩≤ 2 4 热空气老化(80℃×168h) 断裂拉伸强度保持率/%≥80 80 扯断伸长率保持率/%≥70 70 耐碱性(质量分数为10%的Ca(OH)2溶液,常温×168h) 断裂拉伸强度保持率/%≥80 80 扯断伸长率保持率/%≥80 80 臭氧老化(40℃×168h),200×10-8—— 人工气候老化断裂拉伸强度保持率/%≥80 80 扯断伸长率保持率/%≥70 70 粘结剥离强度(片材与片材) N/mm(标准实验条件)≥ 1.5 1.5 浸水保持率(常温×168h)/% ≥ 70 70 复合强度(FS2型表层与芯层)/(N/mm)≥— 1.2 注1：人工气候老化和粘合性能项目为推荐项目； 注2：非外漏使用可以不考虑臭氧老化、人工气候老化、加热伸缩量、60°断裂拉伸强度性能