Transit Network design Methods

By,K.Vinay203113023

Introduction• transit network design problem is one of the most significant problems

faced by transit operators in the world if involving more than one mode and diverse vehicle fleets.

• For that transit network design we are opting for many Computerised models which have been developed for this purpose are often highly specialised but very expensive

• Discussing actual transit network design + Bee Optimisation Approach• In this paper all the main constraints are considered

1. maximize the number of satisfied passengers 2. minimize the total number of transfers and 3. minimize the total travel time of all served passengers

Why we have to go for transit design??

• Urban Road Consequences resulted in development of Park and Ride, HOV facilities, Congestion pricing etc.. But still raising of traffic is needed to be concentrated seriously

• A part from the above approaches even a Improperly designed public transit network can significantly increase public transport mode share like long waiting times of passengers where as for shorter times we require more public vehicles (Un economical)

• Public transit network design problem is almost faced by all authorities

Method1: Practical Transit Network Design

• adaptation of the Emme/2 transport network model• In the 1st Part - basic inputs to the Emme/2 model are the general

transport network, and the auto and transit demand matrices• 1st we give Auto Assignment Option in Emme/2 -Prevaling Traffic

modified volume delay functions (VDFs) are used to "focus" transit demand onto routes with sufficient demand volume

• initially a "free," or "user optimal" assignment of transit trips, is increasingly constrained in subsequent iterations

• Requirement of the Model is the network model for the planning of a transit system must reflect all possible transit facilities like for example railway station as a node

• For an Instance it is assumed that all roads and dedicated transit facilities in the modelled network are available to transit then

• For more in detail1. Vehicle occupancy2. Volume Delay Functions3. Generalized costs4. Other modes

• Vehicle Occupancy: • The travel time on a network link is a function of the link

capacity and the number of vehicles using the link.• The number of vehicles is dependent on the number of

people travelling and the average vehicle occupancy rate.• For the roads VolumeDelay functions Vehicle Occupancy is

considered- Note: More Occupancy vehicles create more delays

• Volume Delay Function• If transit only route design we no need to consider this

volume delay • Calibrated from parameters Possible trravel time,No of

vehicles

• Generalised costs• Sum of costs due to waiting/travel time + Toll and other direct costs• Function on link length and speed limit• Transit vehicles choose shorter routes

Method-2 Bee Optimisation Technique

Problem Statement• Let us consider the road network shown in the fig We denote this network

by G = (N,A), where, N is the set of nodes represent potential bus stops , A is links (street segments). • Any path used by transit passengers is defined by a sequence of nodes,

and links

• Main Intention is In the given Network transit we will select R links from the total A links and bring these links into fixed transit routes by using a Indicator

• Indicator we use is Total travel time spent by the users of transit service ‘T’

Note: Since Transfers make back passengers to come close transfers may be reduced by optimizing the transit route

• Solution of finding best transit route is by Bee Colony Optimization

Bee Optimization Technique:In this bees behavior is collected during the nectar collection process. Artificial bees utilize the principles utilized by bees1. Artificial bees explore through the search space looking for the feasible solution (i.e

honey)2. During the Search Every bee by accepting and improving solutions3. Finally after getting solutions they will go back to the nest4. During their return all the bees collaborate and exchange the information and decide

whether to abandon the created solution and become the follower(based on the probability)

5. Every follower choose a new search operation and this process goes on iteratively

BCO approach to the transit Network design Problem1. We will generate initial feasible solution (old route / less distance)2. Artificial bees try to investigate the solution space in the neighbourhood

of current solution and try to improve the solution

Step1: Initial Solution

• Bus line could be used by passengers that enjoy direct service as well as passengers with two transfers

• Passengers who enjoy direct service is given by dsij

• Initial Solution can be generated using certain algorithm .• Process 1:Prescribe the total number of bus lines NBL in the network. • Process 2:Find the pair of nodes that has the highest dsij value. Let

this pair is the pair of nodes (a,b). The nodes aand b are the terminals of the new bus line. Find the shortest path between these two nodes

Step2: Checking Alternate Solutions • We solve the transit network design problem by using two sets of

artificial bees. type 1 and type 2. • The type 1 artificial bees differ from the type 2 artificial bees only in

the way in which they modify the solutions. • When making decisions about the loyalty, as well as decisions to join

some of the recruiters both bees of type1, and bees of type 2 behave in the same way.

2.1 Solution modification by Bee1Type 1 bee chooses one line from the set of bus lines according to probability. Probability of taking line l can be calculated as:

for example, Step(a) type1 bee chose line whose terminals are i and j. Step(b) In the next step, type 1 bee chooses one among two terminals. Suppose i selected. Type 1 bee destroys selected bus line. New bus line will be created between terminal i and new terminal k, which could be chosen with the probability

• 2.2 Solution modification by Bee2• Type 2 bee chooses one bus line in the same manner as the type 1

bee and then in random selects one of the node for example ‘j’• In the next step, bee decides to destroy this terminal and creates a

new route as shown in fig

Analysis of Solutions generated by the beesLet Tb (b = 1, 2, … , B) the total travel time of bth bee. Ob =normalized value of the total travel time Tb, i.e.:Pb= probability that bth bee (at the beginning of the new forward pass) is

loyal to the previously generated solution

• where Tmin and Tmax are respectively the smallest and the largest total travel time in all transit networks generated by all bees.

d0 – the percentage of demand satisfied without any transfers,d1 – the percentage of demand satisfied with one transfer,d2 – the percentage of demand satisfied with two transfers,dun – the percentage of demand unsatisfied,ATT – average travel time in minutes per transit user (mpu).

Conclusion

• Transit Network design is large combinational problem which can be optimally solved by this New approach effectively

• In this BCO method author tried to maximize the number of served passengers, to minimize the total in-vehicle time of all served passengers, and to minimize the total number of transfers in the network

Drawbacks

• In BCO author has not considered about any other non-transit vehicles and

• he has not showed how effectively this technique works in reality

• Author also not considered the practical facilities and mainly design of approach roads to the transit network

• References• Transit network design by Bee Colony

Optimization by Miloš Nikolic´ , Dušan Teodorovic´ ⇑ (Volume 40, Issue 15, 1 November 2013, Pages 5945–5955)

• Feeder transit Network by Ashish Verma• A Practical Transit Network Design Method by

W R Duff-Riddell

Thank you..