Energy-saving Optimization Of Recommended Speed Curve Based On Artificial Bee Colony Algorithm

Urban rail transit has become the mainstream means of transportation in cities because of its large volume,fast speed,dense shift and high punctuality,and the number of lines and operating mileage are still increasing.Therefore,as one of the industries with the largest energy consumption in the national economy,energy-saving operation of trains is particularly important.Among them,the traction energy consumption of trains accounts for half of the total energy consumption of urban rail transit.Therefore,it is of great practical significance to reduce the traction energy consumption by optimizing the recommended speed curve.Combining with the current situation that genetic algorithm and particle swarm optimization are easy to fall into local convergence,an artificial bee colony algorithm with high precision,fast convergence and strong global optimization ability is proposed to optimize the recommended speed curve of trains.The main work is as follows:(1)The force of train in operation is analyzed and the dynamic model is established.Based on the dynamic equation and the maximum principle,the Hamilton function is established,and the Lagrangian operator is introduced to analyze the optimal control strategy of the train.Taking energy consumption and running time errors of stations as optimization objectives,a multi-objective optimization model is established,and the optimization problem of recommended speed curve is transformed into a path planning problem with the lowest energy consumption and running time errors in speed-distance two-dimensional space.(2)The performance of the artificial bee colony(ABC)algorithm is verified by comparing the algorithms.Particle swarm optimization(PSO)and differential evolution(DE)are compared.Three algorithms are used to simulate the single-peak single-extremum function Sphere Model,multi-peak multi-extremum function Griewank and respectively generalized Rosenbrock from low-dimensional to high-dimensional in MATLAB.Each simulation runs 30 times,and the maximum,minimum and average values are generated into tables for analysis.A total of 1110 simulation comparisons verify the absolute superiority of ABC algorithm in search accuracy and stability.By comparing the fitness evolutionary curves of the three algorithms It is proved that the convergence speed of ABC algorithm is relatively faster.It is proved that the artificial bee colony algorithm can optimize the recommended speed curve.(3)Two methods are proposed for the optimization of recommended speed curve.One is to transform multi-objective optimization into single-objective optimization by using run-time errors as constraints.The other is to improve the artificial bee colony algorithm.According to the principle of Pareto domination,the method of calculating the fitness value of the algorithm is changed,and the Pareto optimal solution is stored in the external file,and according to the congestion distance to maintain the external archives,at the same time,improve the neighborhood search formula,the solution in the external archives can guide the search.The two methods are modeled and optimized respectively.(4)Based on the line data of Beijing Metro Yizhuang Line,the optimization model is simulated and validated by using MATLAB.The advantages of multi-objective optimization method are verified by comparing single-objective optimization with multi-objective optimization.The validity of sub-section is validated by comparing the simulation between equal-section and non-equal-section.The simulation results obtained by setting different parameters of control strategy are simulated.The real results verify the influence of ATO control strategy on energy-saving optimization.