Research On The Optimization Method Of Operating Curve Of Heavy-haul Train On The Long Steep Downward Slope Sections

Heavy haul railway has the characteristics of large transportation capacity and high transportation efficiency,so it plays an important role in the transportation of bulk goods.With the increase of freight volume,the capacity of a heavy haul train in China has increased from 10,000 tons to 20,000 tons.The driving control of long heavy haul trains with traditional air braking systems under complex operating conditions has ushered in a new challenge.In order to further reduce the labor intensity of drivers and improve the transportation capacity of the heavy haul railways,it is particularly important to study the automatic driving control of the heavy haul trains.The following characteristics of the heavy haul railway and train operation make it difficult to realize the automatic driving of heavy haul trains in China.First of all,the marshalling of the heavy haul train is diverse and the length is long.The trains use the synchronous control system to realize the remote control of locomotives.The braking system includes the traditional air braking system and electric braking.Secondly,there are some long downward slope sections of the heavy haul railways in China because of the large altitude difference,even if the train adopts the maximum electric braking,the speed will increase.In order to improve operation efficiency and reduce maintenance costs,heavy haul trains usually adopt a cycle braking strategy on the long downward slope sections.Considering the complex conditions such as neutral zones and curves,more challenges have been brought to the calculation of driving strategies.Taking the Shuohuang Railway as the research background,this article studies the driving curve optimization problem of heavy haul trains on the long steep downward slope sections based on the model-driven(artificial bee colony algorithm,mixed integer linear programming method)and data-driven(approximate dynamic programming algorithm)methods with considering the operating characteristics and the corresponding constraints of heavy haul trains.The optimization effects of different algorithms are analyzed by the simulation results.The research work of this article mainly includes the following aspects:(1)This article analyzes the operation characteristics of heavy haul trains.An optimal control model of heavy haul train on the long steep downward slope sections is constructed with considering the constraints of air-recharging time and regimes conversion with the goal of reducing air braking time and improving train operation efficiency.(2)In this article,the existing optimization control model is transformed into the problem of solving the switching points of the train’s regimes.Based on the artificial bee colony algorithm,which is one of the heuristic algorithms,an optimization method is designed to solve the driving strategy of the heavy haul train.Besides,an initial solution generation method combining the line conditions is proposed,which reduces the search range of the optimal solution and improves the efficiency.(3)Based on the numerical method,this article linearizes the nonlinear problem and uses the mixed integer linear programming(MILP)method to solve it.Compared with the heuristic algorithm,the result obtained by MILP method is more accurate.The simulations are carried out to illustrate the effectiveness of the MILP method,and the influences of the parameters on the optimization effect are analyzed.(4)Additionally,the driving strategy of the heavy haul trains is solved based on a data-driven method.The optimization problem is transformed into a Markov decision process.Combined with the specific optimization problem,the state,action and cost function in each stage is defined,and the value function is approximated by the basis function.The approximate dynamic programming(ADP)method is used to solve the optimization problem,and the effectiveness of the ADP method is verified based on some experimental simulations.