Research And Simulation On Multi-objective Control Strategy Of Train Automatic Operation

For a long time, transportation of the railway in our country plays an important role. Dueto railway transportation energy consumption is huge, the domestic and foreign scholars havecarried out an extensive and in-depth research for train energy saving problem. With thetransport pressure increasing, in addition to reducing train energy consumption as much aspossible, the requirements for punctuality, stop accuracy and comfort level also constantlyimprove. This makes the train operation process into a multi-objective optimization problem,and the goal is to get a set of control strategy, so that each goal can be as much as possible toachieve an optimal value. So, this paper conducted the following research:On the basis of traction computing theory, the paper builds train dynamics model andcombines with multi-objective optimization theory and methods to establish a multi-objectivemodel of train operation process. Through solving the model, it has got the necessaryconditions for the optimal control of train and the four phases in train operation process, anddiscussed existence of the optimal train control strategy.Control strategy of train varying on local lines such as straight lines, gradient lines andbend lines, but similar control strategies shows on same local line. Based on this idea, paperstudies states of moving train on local lines. Combining with the actual situation to obtaintrain control strategy on straight level lines, gradient lines and bend lines, and confirm thelocation of phase transition point. Through reasonably composes every control strategy oflocal lines, getting overall control strategy of train operation.Using improved multi-objective genetic algorithm optimize train operation strategy bymapping train control strategy which including phase translation point position intochromosome with code designing. To improve the performance of the algorithm, using chaosoptimization method with tent map in the process of population’s generation, and theastringency and diversity of population was improved effectively. Designed selection operator,crossover operator and mutation operator of the optimization algorithm, and advanceddesigned simulation module of train operation process and fitness function, realizedmulti-objective improved genetic algorithm for optimizing train control strategy. Throughsimulation of actual example, the results show energy consumption has significantly reductionwith the comfort target and punctuality target and stop accuracy target in a satisfying situation,and this can provide reference to train’s efficient operation.Finally, a system for simulation optimization of train operation process was achieved.System can accomplish the input of data, generation and optimization of control strategy. Thesystem using human-computer interaction mode with simple operation for optimization oftrain control strategy and simulation of train operation process.