Research On Multi-objective Optimization And Control Of Urban Rail Train Running Speed Profile

Since the 21 st century,rail transit has gradually become the artery of urban transportation due to its many advantages such as high efficiency and convenience.As one of the important ways to ease urban traffic congestion,urban rail transit not only solves the traffic congestion problem of big cities,but also optimizes the urban structure and promotes economic and social development,which is getting more and more attention and favor from cities and residents.The rapid construction of rail transit has brought huge socio-economic benefits to urban development,and while solving the huge demand of urban residents for travel,it has also brought huge energy consumption problems.Among them,traction energy consumption is the main factor affecting the total energy consumption of train operation,so there is a lot of room for optimization of traction energy consumption.Research aimed at reducing the energy consumption of train operation has also received increasing attention.The research in this paper mainly includes:(1)In the train optimization modeling,the functions and control principles of the automatic train operation(ATO)system are firstly introduced in detail to clarify the connection between the optimization of the target speed curve and the optimization of train energy saving,and to establish the research objective of off-line optimization of the target speed curve.Then,the forces during the train movement are elaborated,and the train dynamics model is established with a single mass point of the train as the research model,while the advantages and disadvantages of various train operation control strategies and the optimization performance index of the ATO system are introduced.Finally,the solution idea of multi-objective optimization problem is discussed in detail,so as to provide theoretical support for further optimization.(2)In the study of train speed profile optimization,it is proposed to solve the multi-objective optimization problem of train operation speed profile by using the weighted summation approach and the intelligent algorithm of multi-objective optimization,respectively.On the one hand,in order to prove that the improved hybrid control strategy has better energy efficiency,the optimization results of different control strategies under the same algorithm are compared respectively;on the other hand,in order to verify the superiority of the intelligent optimization algorithm based on the idea of Pareto optimal solution,the optimization effects of the hybrid algorithm combining the improved particle swarm algorithm and the simulated annealing algorithm and the multi-objective quantum particle swarm algorithm under the same control strategy are discussed respectively..The advantages and disadvantages of the joint optimization algorithm and the multi-objective quantum particle swarm algorithm are verified by the actual metro line data.(3)In the train operation control research,the principles of classical decision tree algorithm and integrated learning algorithm are firstly introduced,focusing on the widely used gradient boosting tree algorithm.Then the overall framework of the intelligent train control algorithm is introduced,based on the inference rules by expert knowledge and online adjustment using the regression model of the Light GBM algorithm.Finally,the multi-objective problem of subway train operation is solved by using expert knowledge in this field and historical driving data,offline optimized speed profile and the advantages of ATO system,rational use of train state information,combined with machine learning methods,thus optimizing train control commands.