Train Cooperative Operation Method Based On Model-Free Adaptive Control

In China,urban rail transit is favored because of its convenience and speed.However,with the rapid growth of the population of China,rail transit is overwhelmed which urgently needs to be resolved.At the same time,the development of modern communication technology also provides the necessary conditions for the exchange of information between trains.The popularity of multi-agent formation control has remained high in recent years.It aims to efficiently complete tasks through the cooperation of various agents in the system.Therefore,in this paper,the problem of multi-train cooperative control under the framework of multi-agent systems is considered,and two distributed multi-train cooperative controllers based on data-driven control are proposed,that is,Model-free Adaptive Control(MFAC)algorithm and Model-free adaptive iterative learning control(MFAILC)algorithm.The main research contents are as follows:(1)Firstly,MFAC,MFAILC and several classic formation control methods are studied.Secondly,the advantages and disadvantages of several formation control methods are analyzed.Finally,based on the graph theory method,this paper proposes a distributed cooperative controller based on MFAC for nonlinear multi-train systems and theoretically proves the convergence of the proposed controller.(2)Considering the repeatability of train operation tasks and train operation dynamics,based on the graph theory method,combining model-free adaptive control algorithm with iterative learning control,a distributed collaborative controller based on MFAILC is proposed,and the convergence of the controller on the time axis and the iteration axis is given.(3)Based on Matlab simulation software,for a multi-train simulation system containing three trains,the traditional PID controller and PD-type iterative learning controller are used as comparison methods for simulation experiments,and the simulation results are respectively compared with the MFAC-based distributed collaborative controller and the MFAILC-based distributed cooperative controller.The simulation results show that the distributed collaborative controller based on MFAC makes the convergence of train speed to the desired speed faster,and the distributed collaborative controller based on MFAILC makes the train speed error converge faster on the iterative axis.It illustrates that the two data-driven cooperative controllers have better control performance.