Bus Scheduling Optimization Research Based On Fuzzy Decision-making

Balancing passenger flow distribution as well as using vehicle capacity effectively is a challenge problem in timetable scheduling of public transit,the key of solving the problem is to optimize the time interval between two successive bus services.Other than the idea in the major researches in scheduling optimization that built corresponding multiobjective optimal decision making model under crisp constraints: vehicle size and cost coefficients.This thesis starts from the operation reality of different public transit problem,formalizing the single crisp constraint of vehicle capacity into two fuzzy constraints with practical meaning by using fuzzy membership functions.Fuzzy decision making models are build to solve timetable scheduling for single time state and continued time states.The main content includes the following parts:1.Passenger flow mining and analysis for public transit.According to the historical data from public transportation system(passenger taping records,the GPS(Global Position System)records of vehicle location and the basic bus line data).Developed the boarding,alighting bus stop matching models,and passenger multi-days missing values completion model to extract the integrated origin and destination of passenger flow.According to the vary of passenger flow in different time stage,estimating the arriving rate,leaving rate of passengers at bus stops and the number of passengers on-board,and prepared for the case study to investigate the following models of the bus timetable optimization.2.Designing fuzzy constraints and constructing the frame of fuzzy decision making.Take the fixed vehicle capacity as entry point,build the membership functions to reflect the satisfaction degree of passengers and usage degree of vehicles,and constructing the frame of fuzzy decision making model with fuzzy constraints base on this,and the optimal time interval is searched by the objective function that has the maximum intersection values of all fuzzy constraints.After this,the frame is used to optimize and adjust the time intervals in timetable with fixed bus service number,to solve the unbalanced vehicle capacity usage in even time interval departure schedule.The optimized timetable avoids the drawbacks of overloading and low vehicle usage effectively,and the frame sets research foundation for applying fuzzy decision making into timetable scheduling optimization of public transit at different time state.3.In the single time state,the corresponding methods of dynamic timetable scheduling are proposed to solve the timetable scheduling problems under the constraints of fuzzy level set and mixed vehicle capacity,respectively.Designing models for public transit timetable scheduling with fuzzy decision making,expanding the decision vector into decision space,to decide the optimal time interval set for each time state.Compared the optimized timetables with the two timetable optimization models with crisp constraints respectively.The results show that timetables optimized under fuzzy environment are more adjustable to passenger flow,and reducing passenger's waiting time and travel time.4.In the continues time state,a new approach of load balancing timetable optimization is proposed which treat the timetable as a whole to be optimized.Extended the reverse flow technique of fuzzy multi-stage decision making,set objective function as the maximize satisfaction degree of fuzzy constraints of the whole timetable,to built the model of fuzzy decision making for public transit timetable scheduling.Model optimization is not limited on a single departure,but on continues bus departures,and searching the optimal time interval sequence.Compared timetables from different bus lines and models,the results show that the timetables optimized under fuzzy environment can optimize the number of bus services and equilibrate the distribution of passenger flow,and provide the time interval sequence with strong stability,comfortable ride and efficient resource usage.The advantage of fuzzy decision making in timetable optimization is obviously showed by studying,modeling,solving,and comparing the above problems,that it can fits fluctuates of passenger's,balances vehicle loads,therefore improving passenger satisfaction degree and vehicle usage.This thesis engaged in both engineering practices and scientific research,expectorating timetable optimization for public transit,the solving approaches are not only significant in timetable intelligence of public transit,but also consultative in fuzzy optimization of bus lines and networks,and related field.