Evaluation And Analysis For Intelligent Optimization Of Urban Rail Train Operation Curve

With the continuous development of urban rail transit,the traction energy consumption of a train has gradually become a major concern.Train operation curve optimization is one of the main methods to reduce the train traction energy consumption,and its essence is to solve a complex multi-objective optimization problem.Among the existing solutions,traditional analytical and numerical methods have shortcomings such as difficulty in solving and long calculation time.As for intelligent optimization algorithm,it has strong adaptability.Moreover,since a high-precision mathematical model is not required,the modeling is relatively simple as well.And therefore,great optimization results have been obtained in solving the optimization problem of train operation curve.Most of the existing researches directly apply intelligent algorithms to solve the train operation curve,lacking the review of performance evaluation indices of intelligent optimization algorithms and the performance comparison and analyzation of the algorithm.Meanwhile,the applicability of the new intelligent optimization algorithm,appearing with the development of intelligent optimization algorithm,on the curve optimization problem remains to be studied.In view of the existing problems,the main research work of this paper is listed as follows:(1)Combining the different scene requirements of the rail transit industry for the optimization curve with the search characteristics of the intelligent optimization algorithm and starting from the algorithm results and the algorithm process,the algorithm performance evaluation indices for solving the optimization problem of the train operation curve is reviewed.In order to objectively evaluate the overall performance of the algorithm,the objective weighting method-entropy method is used to calculate the weight of each index,which can evaluate the overall performance of the algorithm(2)Based on the typical methods of traditional intelligent algorithms,such as genetic algorithm,ant colony algorithm and particle swarm optimization algorithm,the train operation curve optimization algorithm is designed.A heuristic setting method considering line speed limit and train running time is proposed,which greatly reduces the complexity of the problem solving.Based on the actual line information and train data of Beijing Maglev Line S1,the influence of algorithm parameters on the performance of the algorithm is investigated by numerical experiments and the optimal parameter settings of genetic algorithm,ant colony algorithm and particle swarm algorithm for solving the problem are obtained.On this basis,the performance of the three traditional intelligent optimization algorithms is evaluated,and the effectiveness of the performance evaluation indices and evaluation methods of the intelligent optimization algorithm is verified.(3)Taking the gray wolf optimization algorithm as an example,the applicability of the new intelligent optimization algorithm in train operation curve optimization is explored from the perspectives of algorithm performance and application scenarios.In response to the needs of different optimization curve scenarios,the applicability analysis of the algorithm for online computing and offline computing is carried out based on the combination of subjective and objective weighting method.The related results show that the overall performance of the gray wolf optimization algorithm is better than that of the traditional intelligent optimization algorithm,and it has strong applicability in offline computing scenarios.(4)Based on the WPF development framework,using the hybrid programming technology of MATLAB+C# is used to develop intelligent optimization algorithm performance analysis software,which has strong computing power and good visualization effect.Algorithm experiments under different parameter settings can be carried out for different line sections and different intelligent optimization algorithms,and relevant performance evaluation indicators can be automatically calculated and visualized.For needs of different scenarios,users can design the weights of various performance indicators in real time,and then comprehensive evaluations of intelligent optimization algorithms in different scenarios can be obtained.70 pictures,37 tables,59 references.