Design And Implementation Of Online Bus Vehicle Scheduling Based On Controller

In the field of public transportation operation planning,bus vehicle scheduling problem is vital for the quality of public transportation service.However,the traffic environment is unstable in the actual bus operation.Unexpected traffic congestion and vehicle breakdowns increase the travel time of the vehicles.If a vehicle cannot arrive on time,the next trip of the vehicle will not be able to depart on time,which decreases the quality of bus service.How to generate high quality vehicle scheduling scheme and how to ensure the online bus operation service is an important problem.Existing methods can be divided into exact and heuristic methods.The exact method takes too much time,which is difficult to be used in practice.The heuristic method is hard to ensure the effect of the result.For the bus vehicle scheduling problem under uncertain environment,the existing methods include robust vehicle scheduling method and dynamic vehicle scheduling method.For the former method,the vehicle utilization rate is low,resulting in higher operating costs.The latter method takes too much time,which is difficult to meet the online requirements.In this paper,first we proposed a controller-based bus scheduling method for the first time,and then we improved the controller using evolutionary neural network.In this paper,we designed a vehicle selection controller.Each departure time point is regarded as a decision point,and the controller selects a vehicle at each departure time point.A schedule is obtained when all time points are covered by vehicle trips.The controller consists of a vehicle type converter and a vehicle selector.The vehicle type converter dynamically converts the vehicle state and type in real time to improve the vehicle utilization.The vehicle selector selects an appropriate vehicle to issue.In addition,in the online vehicle scheduling scenario,in order to send vehicle scheduling commands in time,a forward detection method is designed improve the decision.The controller contains some manually set parameters.In this paper,a particle swarm optimization is used to optimize these parameters to improve the controller.To further improve the decision quality,we designed an evolutionary neural network instead of the manually designed controller.We design a three-layer radial basis function neural network,and a particle swarm optimization is used to optimize the neuron parameters.The experimental results show that the controller method performs better than the comparison method and artificial results,and the evolutionary neural network method also has an improvement on complex constraint problems.In the online scenario,controller method takes very short time to calculate and meets the online requirements.Finally,based on the above research,this paper designs and implements a bus scheduling system based on Qt.The system can be applied to the bus scheduling task in offline and online scenarios,which significantly improves the bus service quality and reduces the bus operation cost.