Scheduling Research On Urban Bus Stop-skipping In View Of The Multi-objective Optimization

With the rapid development of the national economy,the speeding up of urbanization,the rapid growth of motor vehicle ownership,and the increasing saturation of road traffic,the traffic problems of large cities are becoming more and more serious.The priority to the development of public transport has become an important strategy to alleviate the congestion of urban roads.Bus scheduling,which is an important content impacting the bus operating system operating costs,efficiency and the level of service capabilities.Based on the previous research results,this paper focuses on the study of bus transit station in the middle of the scheduling.The paper analyzes the mechanism of conventional bus stop,and studies the running status of conventional bus.The paper analyzes the characteristics of bus transit time,including bus stop time,vehicle running time between stations and bus transit time.When the bus transit station happens,the bus runs through the site in the normal state,which does not produce docking time at the site and slowing down and accelerating the outbound than the normal travel time.Therefore,it is very important to analyze the stopping time of vehicles and the time of vehicles entering and leaving the station.On this basis,this paper focuses on the in-depth analysis and comparison of the bus stop time in the site.Through the investigation of the bus stop time and the corresponding number on and off,the paper calculates the bus stop time at the site by the average time passengers on and off by the TCQSM manual recommended respectively.According to the multiple regression analysis on the existing data,the paper builds the multiple linear regression model of the number of passengers on and off and bus stop time.Using the machine learning algorithm,this paper mainly uses the RBF neural network algorithm to carry on the excavation analysis to the existing data.Through the comparison of the results obtained by the three methods,the paper selects the RBF neural network algorithm,which is the method with the least error,to carry on the research of the transit vehicles.In this paper,the number of vehicles studied is extended to n vehicles in order to explore the optimization of the whole system when the number of vehicles increases.Secondly,the weight of the passengers’ waiting time,passengers’ time and vehicles’running time are analyzed,and a target with the greatest impact on the whole system is found.Finally,the problem is discussed and solved by using the cellular genetic algorithm,Monte Carlo simulation,RBF neural network and so on.The results show that the cellular genetic algorithm is more efficient and the speed is improved by 10.3%.Besides,as the proportion of the number of vehicles moving to the total number of vehicles increases,the efficiency of the system increases gradually and increases to 1/2(n is even).As the number of vehicles increases,the system efficiency increases while the value is gradually increasing.In addition,the passenger car time in the three goals is at the dominant position,while the passenger site waiting time on the system the least impact.For the bus departure interval,with the increase in the start interval,the implementation of the effect of the station will be a rapid decline in scheduling,when the departure interval of 25 minutes or so,the implementation of the station will make the overall system scheduling efficiency is lower than the station.