Models And Algorithms For The Integrated Optimization Of Train Rescheduling And Train Control For High-speed Railway

The high-speed railway system is an important component of the comprehensive transportation system for China,which develops rapidly in recent years.Moreover,the demands and service quality requirements of passengers for the high-speed railway continue to increase.In the daily operation of the high-speed railway,many factors from the internal and external conditions may influence train operations.Thus,integrated optimization of train rescheduling and train control is an important breakthrough point,to ensure efficient high-speed rail operation with good quality under disturbances.However,there are many challenges to develop and solve the integrated optimization model,e.g.,it is difficult to couple the decisions of macroscopic train rescheduling and microscopic train control;it is difficult to model multiple decisions of integrated optimization;it is difficult to analyze and model the influence of disturbance on train operations.Moreover,the integrated optimization model is typically non-linear,strongcoupling,multi-constraints and multi-objectives,which is difficult to obtain the solutions with good quality in a reasonable computational time.Recently,there are a few research focuses on the integrated optimization model of train rescheduling and train control.However,the existing methods consider limited integrated coupling elements and disturbance scenarios,which is difficult to ensure the scalability,accuracy and enforceability of the solutions for train rescheduling and train control.Therefore,this paper studies the methods of modelling,solving and extending for the integrated optimization models.In addition,this paper develops an algorithm for the large-scale problem,to obtain optimal solutions with reasonable computational time.The integrated optimization methods of train rescheduling and train control have theoretical and practical significance to ensure the efficient and high-quality operations of the high-speed railway system under disturbances,to improve the quality of passenger services.This paper studies the integrated optimization models and algorithms for train rescheduling and train control,which focus on the following four issues.(1)This paper studies the theoretical model of the integrated optimization of train rescheduling and train control for the high-speed railway.First,we analyze the problems of train rescheduling and train control.Based on the analysis of the traditional methods for optimizing the two problems,we study the steps of the iterative optimization methods in the theoretical research.Second,we address the definition of the integrated optimization method.In details,we analyze the key elements of the integrated optimization and the internal correlation mechanism among them.And then,based on the analysis of the key points and difficulties of modeling the integrated optimization problem,we develop a theoretical integrated optimization model.Finally,we analyze the complexity of the integrated optimization model.(2)This paper studies the basic model and the solution method for the problem of integrated optimization of train rescheduling and train control with high-precision requirements.With considering the train characteristics and line conditions,the train traction calculation is carried out to improve the precision of train control.In addition,we use the microscopic model to describe the high-speed railway line.First,we establish two mixed-integer nonlinear programming models for the integrated optimization of train rescheduling(i.e.,the arrival and departure times of trains at each block section)and train control(i.e.,the train speed trajectories).The objective is to minimize the total deviation times of the trains.The train operation process is discretized based on the discrete-space method and the discrete-time method,respectively,for the two models.Second,we analyze the non-linear constraints of the models and design the reformulation methods for the two nonlinear models.Finally,the two nonlinear models are transformed into mixed-integer linear programming models.The proposed models could accurately describe the traction calculation process of high-speed trains.Compared with the model based on the discrete-time method,the proposed integrated optimization model based on the discrete-space method is better in terms of the problem scale,computational efficiency,scalability and applicability.(3)Considering multiple train rescheduling decisions,this paper studies the model and solution method for the problem of integrated optimization of train rescheduling and train control.Based on the basic integrated optimization constraints,we study the integrated optimization method of re-ordering,re-routing,re-timing and re-speeding.The objectives are to minimize the total delay times and total energy consumptions.A mixedinteger linear programming model with two objectives is proposed to solve the integrated optimization problem.Two algorithms based on the epsilon-constraints and weightedsum method are proposed to obtain the Pareto frontier for the bi-objective optimization problem.A two-step approach is designed to speed up the solution process.The proposed bi-objective integrated optimization model could obtain the rescheduled timetable and train speed trajectories at once,with less total delay times and lower energy consumptions.(4)This paper studies the integrated optimization method of train rescheduling and train control under complex disturbances.In addition,a solution method is proposed to solve the large-scale problem.This paper analyses the influence of the complex disturbances on train operations and train control.Based on the constraints for the integrated optimization of train re-ordering,re-routing,re-timing and re-speeding,an integrated optimization model of train rescheduling and train control under complex disturbances is proposed.The model is a mixed-integer linear programming model.In the model,we consider two typical complex disturbances: temporary segment blockage and temporary speed restriction.A method based on the knowledge and rules is proposed to reduce the scale of the feasible region for the integrated optimization problem.In addition,a heuristic algorithm based on the Lagrangian multiplier is used to obtain a feasible solution in a short time.Furthermore,we proposed two iterative adjustment methods for interacting two problems of train rescheduling and train control,to be compared with the integrated optimization method.The integrated optimization method outperforms two iterative adjustment methods for the enforceability and solution quality.This thesis includes 51 figures and 12 tables,and refers to 199 literatures.