Optimization And Simulation Of Multi-Objective Problem For Train Based On Swarm Intelligence

In recent years,the big cities in our country have successively constructed the subway system to improve the traffic efficiency.The total mileage of the subway has rank the first in the world.At the same time,the subway system operating costs and energy consumption also caused more and more attention.Therefore,how to reduce the running energy consumption of trains by using advanced intelligent algorithms effectively has become the research focus in this field and that has important significance.The main research work of this paper is divided into the following four points:(1)Take Fuzhou metro line 1 as the research object and collect various types of data and information of line 1 train by myself,which researched on the modeled.The model was key on train power system,resistance system and line condition and the basic control unit.In addition,this article also analyzes and researches the operation needs of the train,clear and designed the five subway-train operation evaluation system,which made the subway automation operation closer to the actual operating requirements.(2)In order to solve the multi-objective optimization problem of train operation,a swarm intelligence algorithm based on particle swarm and cuckoo proposed in this paper,called hybrid swarm intelligence-algorithm.The algorithm has the following improvements:1)According to the operational characteristics of train operation control and swarm intelligence algorithms,improve the train control unit,which makes the train more controllable.2)The particle swarm optimization algorithm is prone to appear that population concentration is too concentrated,which caused the loss of population diversity.We designed a random elimination mechanism with probability P random elimination of particles,which does not comprise the optimal particles,and use a new value to replace that,so that not only save the best individuals,but also increase the diversity of the particle swarm optimization algorithm.3)This paper makes use of the Levy flight strategy from the cuckoo algorithm.This strategy can make the particle search step more random and avoid the premature convergence of the algorithm.(3)In this paper,experiments carried out on the algorithm.Using MATLAB as the platform and the multi-section route of Fuzhou metro line 1 as the test line,the particle swarm algorithm,cuckoo algorithm and hybrid swarm intelligence-algorithm tested experimentally.Compared with the previous algorithms,the hybrid swarm intelligent algorithm has the highest comprehensive efficiency:1)the computing time is saved by about 7%,2)the energy consumption is saved about 3.5kw/h per 100 seconds.(4)Based on the research,the control and operation simulation for the whole section of Fuzhou metro line I designed and implemented on industrial control platform IAP.The simulation.system can realize the automatic driving of train,automatic monitoring and shows the result of simulation,which indicate that the improved algorithm can be well transplanted to the real industrial system and a feasible solution is put forward for the practical use in the future.