Research On Multi-objective Optimization And Intelligent Control Algorithm Of Subway Automatic Operation

With its advantages of large passenger capacity,fast speed and high comfort,subway has become an important means of transportation to ease urban traffic congestion and improve travel efficiency.At present,the subway is transitioning from manual driving mode to automatic operation mode.Automatic subway operation instead of manual driving is the embodiment of scientific and technological progress,and also the general direction of subway development in the future.The automatic operation of subway needs to meet the five performance indexes of energy saving,safety,passenger comfort,parking accuracy and punctuality at the same time.At present,most of the indexes studied and considered are not perfect,and the multi-objective optimization is changed to the single-objective optimization by means of weight accumulation.Aiming at the problems existing in the above research,this paper solves them from the following aspects.Fully understand the operation process of ATO system,analyze the operating environment of subway,establish indicators of energy consumption,parking accuracy,passenger comfort,safety and punctuality,and bulid a multiobjective optimization model based on the multi-objective optimization theory.The non-dominant sorting genetic algorithm(NSGA-Ⅱ)was selected to solve the multi-objective optimization model.In view of the disadvantages of the multi-objective optimization algorithm,such as the inconsistency of each subobjective and low efficiency of finding the optimal solution,energy-saving,passenger comfort and punctuality were selected as the optimization objectives,and safety and parking precision were selected as the constraints.In order to increase the chance of finding the optimal solution,and improve the efficiency of the algorithm,a penalty function on safety and parking accuracy was established.According to the value of the penalty function,whether genetic modification should be carried out was determined,which not only optimized the multiple indexes of the automatic operation of subway,but also improved the ability of the algorithm to find the optimal solution.According to the PID control accuracy is not high,poor adaptive ability and the hysteresis of ATO system,choose the grey prediction algorithm combined with fuzzy control,and according to the characteristics of metro,the grey forecasting model of grey action from constants to variable,fuzzy PID controller is designed based on improved grey prediction,make the subway accurately track the actual speed of target curve.Finally,the route from Lvshun Xingang to Tieshan Town of Dalian Line 12 was selected as the simulation route.The solution of multi-objective optimization model and the establishment of ATO simulation control system were completed in MATLAB/Simulink.In order to improve the authenticity of the simulation,the controlled object used S-Function module to simulate the subway operation process.The simulation results verify the feasibility and superiority of the improved NSGA-Ⅱ algorithm and the improved grey predictive fuzzy PID control algorithm in the aspect of subway speed curve optimization and speed curve tracking.