OPERATIONAL ANALYSIS OF URBAN BUS TRANSPORTATION SYSTEM

Abstract

Consequent to the increase in passenger demand on urban bus system, comprehensive bus transportation plans are required for several cities in India. A methodology needs to be developed for optimizing transit systems and route design. The methods presently used, however, do not sufficiently incorporate factors like engineering and policy changes and their influence on the system components. Thus, the main objective of this research is to formulate a simulation model to represent the bus operations on an urban route in Delhi. The model should be capable of studying the effect of changes in system variables on the operation of bus systems, thereby assisting the evaluation of new schemes for improving the urban bus service. The simulation models are the representation of a real life system by replication of a succession of events that take place in the system according to logical and mathematical relationships. In the case of urban bus system, most of these relationships are available in literature and others are developed by the use of logical rules after a close study of the bus traffic phenomenon (ii) in Delhi city. Development of some relationships, however, needed data input from the field survey which was conducted in Delhi Metropolitan area. The relationships that are formulated from the field data by statistical induction are: (i) "bus passenger service time i.e., time required to load and unload passengers at bus-stops, (ii) bus speeddensity relationship for mixed flow conditions of Delhi traffic S3Qd (iii) acceleration and •deceleration characteristics of buses. In formulating the above relationships, variabi lity of the behaviour of drivers, operational efficiency of buses and queue discharge characteristics of the vehicles are taken into consideration. This forms the necessary background information to create a logical structure of the model. The computer programme is written in FORTRAN IV language. The model is tested in IBM 360/44 available at the University of Delhi and in IBM 370/145 belonging to the Institute of Petroleum Exploration of Oil and Natural Gas Commission, Dehra Dun. The inputs for the model are as followst (i) For the traffic system - vehicle discharge headways from a queue, traffic demand for the main and cross streets, opposite traffic (iii) volume, turning probabilities, expected lane distribution etc. (ii) For the bus system - the normal bus speed, acceleration and deceleration rates, sche duled arrival times of buses, bus length and capacity, regression coefficients for bus passenger service time equations, passenger demand at various bus-stops etc. (iii) For the roadway system - number of blocks or links, number of" lanes, length of each block, bus-stop characteristics, traffic control devices at intersections etc. Some fundamental measures of effectiveness of urban bus transportation system are brought out as output of the model. These are: (i) Bus travelling speed (ii) Total and average passenger waiting time (iii) Arrival and departure times of buses at each bus-stop (iv) Number of passengers waiting to be loaded at each bus-stop (v) Number of passengers boarded and alighted for each bus etc. (iv) The general traffic is simulated macroscopically at constant time intervals and represented according to the inputs of the traffic system already described. But the bus traffic is represented microscopically i.e., indivi dually and processed only when significant bus events (such as arrival at bus-stops or stop lines and departure from them, passenger boarding and alighting operations at busstops) take place. The buses arrive at the first block or link as per the scheduled arrival times, modified to include the stochastic variation. Their movement is governed by the normal travelling speed of the bus, its acceleration and deceleration characteristics, the impedance caused by the traffic and traffic control devices at intersections and the passenger servicing at bus-stops. The simulation model,after a long and tedious debugging, is subjected to repeated tests for verifying its performance as per the logical rules. Model calibration is done for the Delhi operational environments by incorporating the relationships obtained from field data analysis and data observed in real life. The sensitivity analysis and vali dation of the model are done with respect to the magnitudes represented by certain elements of the model to the real life conditions of the system simulated. ' (v) These are: (i) Time-distance diagram of bus service (ii) Irregularity of service (iii) Passenger waiting time (iv) Bus running time The effectiveness of the model is demonstrated by providing an exclusive bus lane by computer simulation on a portion of the Trans-Jamuna Road in Delhi and also by com paring it with the values collected from the field obser vations. The usefulness of the model is brought out by conducting operational analysis on different alternative bus systems involving engineering and policy changes. The stepwise improvement of the urban transport that is possible in Delhi environment,which also has relevance to the developing economies of other Third World countries, is clearly brought out.