Model And Algorithm Of Regional Bus Scheduling Problem With Uncertain Information

The regional bus scheduling,roughly consists of three steps: timetable generation, thegeneration of vehicle schedules and the generation of crew schedules, is the future trend ofdevelopment of urban public transport. This study, involving in operations research, appliedmathematics, network analysis, graph theory, computer applications and transportinterdisciplinary etc., can maximize the interests of both of passengers and bus companies.Hence, it's one of the hot issues. A efficient vehicle schedules, involving in a given set oftime-tabled trips to be assigned to vehicles, is one of the most effective measures to minimizecapital and operating costsConsidering uncertainty in real traffic, such as congestion, weather and accidents etc.,this paper aims at optimize regional vehicle scheduling problem,which is the core of theregional bus scheduling, to explore a range of suitable uncertainty bus scheduling models andalgorithms for public transport enterprise. The main research work and results are summarizedas follows:(1) This paper presents a few of the mathematical model of random, fuzzy and graytravel time RBSP to aim at maximizing the efficiency of utilization of vehicles, where someconstraints such as the depot capacity etc are considered. To resolve the uncertain model,RBSPUTT will be transformed into the certainty model,which is suitable for computerprocessing, or directly to a solution based on the uncertainty simulation technology.According to the characteristics of the problem, a variety of improved intelligent algorithm,such as: genetic algorithm, ant colony algorithm and particle swarm optimization, etc. areproposed to solve these problems. Through a simple numerical example, we have verified theeffectiveness of model and algorithm, and give the best bus scheduling schemes to theseproblems.(2) Learning from online VRP, this paper constructs a class of fuzzy random travel timeORBSP mathematical model, by assuming that uncertainty factors to affect RBSP are met oneby one. Based on the greedy strategy, a collaborative ant colony algorithm, integrating withthe uncertainty simulation technology, is presented to resolve for the RBSP in realtime togenerate a feasible schedule to adapt to a changing environment. (3) Reliability theory is firstly used in the RBSP in the paper. Concepts to the bus taskreliability and reliability of bus scheduling scheme are defined. Based on these concepts, thispaper constructs a mathematical model of the single-objective and multi-goal RBSP. Atheorem is given to prove that the probability of a vehicle completing adjacent trips on time isgreater than a certain value, and a high reliability solution can be constructed on this basis.This theorem can be used to the design corresponding heuristic algorithm to solve theproblem. Give an example for solving bus scheduling problem under different reliabilitycircumstance,and results show that: the cost and reliability directly proportional relationship,that is, the higher the reliability the higher the cost of the program.(4) This paper uses uncertain bi-level programming theory to explore the organic linksbetween a numbers of factors of the urban public transport scheduling, such as bus scheduleand its purchase plan, multi-modal bus timetables and its bus schedule, bus schedule and itscrew plan. Based on feature on the bi-level programming model, solution algorithms aredesigned. Defining concept of the solution satisfaction, the lower-level programmingsolutions with a group of satisfaction are generated for the upper planning. Thereby, the bestlower solutions and the corresponding upper one are generated. The study results can beday-to-day operations of the auxiliary bus companies.