Multi-objective Optimization Of Operating Curve Of Heavy Haul Train Based On MOEA/D

Heavy haul train is a large-scale power-concentrated freight train marshalling with a huge load.The carriages are connected by couplers,and there is a large hysteresis and nonlinearity in the transmission of power.At present,the manual driving used in heavy haul transportation in our country is difficult to operate and the operation line is complicated.With the increase of the traction load of the train,safety problems such as disconnection and decoupling in the operation process will emerge.Therefore,studying the optimization of the target operating curve during the operation of heavy haul trains is of great significance for ensuring the safe,stable,energy-saving and punctual operation of trains.In order to solve the above safety problems and achieve the optimal operation of the heavy haul train,a multi-particle dynamic model of the heavy haul train is established in this paper,and two different coupler-buffers models are selected as the constraints to optimize the running curve of the train.It can achieve the purpose of ensuring the safety of train operation,improving the stability of the running process,reducing the energy consumption of the locomotive and improving the on-time rate of the station.The specific research is as follows:1.Analyze the characteristics of the running process of a heavy haul train,establish a multi-particle longitudinal dynamics model,analyze the force of each particle,and use the "Zhai method" to analyze the train operation data to solve the established dynamic equation.2.Based on the established dynamic model,taking into account the constraints of the coupler-buffer model and the actual line conditions(ramp,curve,line speed limit,excessively equal),a multi-objective optimization model is established considering the four indicators of safety,stability,energy saving and punctuality.The multi-objective evolutionary algorithm based on decomposition(MOEA/D)is used to optimize the operation process,and the optimization results are compared with the vector-evaluated genetic algorithm(VEGA).It is found that the use of MOEA/D can comprehensively and balancedly ensure the safety,stability,energy saving and punctuality of train operation.3.For the MOEA/D,the method of generating the weight vector and the decomposition method used are improved,and the ε dominant idea is introduced,which is named as theε-IMOEA/D.The four objectives of safety,stability,energy saving and punctuality are also considered.In the case of considering the same model and constraints,ε-IMOEA/D is used for optimization simulation,and the optimized train operation curve,various operation data and the values of the four evaluation functions are analyzed,ε-IMOEA/D has better optimization effect than the previous MOEA/D,which can better ensure the safe,stable,energy-saving and punctual operation of heavy haul trains.In this paper,the operation process of the heavy haul train is studied.First,the overall multi-particle longitudinal dynamic model of the train is established.Taking the constraints of the coupler-buffer model and the actual driving rules as constraints,four factors of safety,stability,energy saving and punctuality are considered.The optimization model is established based on the index,MOEA/D and its improved algorithm are used for optimization simulation,and the target operation curve with better optimization index and more in line with the requirements of heavy-duty train driving operation is obtained.In this paper,various models are established based on real data,and the simulation results have certain practical significance for the application of heavy haul train driving optimization.