Intelligent Optimization Research On Integration Of Train Control And Train Scheduling

The rapid development of the current society and the huge difference between the urban and rural economy have caused more and more people to rush into the city to find opportunities.The rapid increase of the urban population has brought tremendous pressure on urban transportation.As an important member of public transportation,rail transit is indispensable for slowing down urban traffic pressure.It is an inevitable trend to vigorously develop urban rail transit and improve the quality of urban traffic.Along with this,train intelligent control,train automatic driving system,train scheduling and other related issues have become a research hotspot.In this environment,the intelligent optimization research on the integration of train control and train scheduling has carried out two different levels of optimization of single train multi-objective and multi-train multi-objective.The main contents are as follows:Single train multi-objective optimization: To satisfy the design requirement of speed profile of urban rail train,and to meet the safety needs as well as kinds of constraints,a multi-objective optimization model is established.This model combines with multiple driving patterns,aiming at shortening traveling time,reducing energy consumption and obtaining accurate stopping.To achieve this goal,a co-evolution based multi-objective chaotic particle swarm optimization(CMOCPSO)algorithm is proposed with the principle of Pareto.The lower layer of basic group launches the global search with goal guiding method,making the solution space evenly distributed while finding out the edge solutions to each target.Meanwhile,the upper layer of elite group completes the local searches carefully by enforcing disturbances.This strategy uses two external archives to forge two-way communication between two layers and set up an efficient search group with all-dimensional cooperation by setting proper communication cycle.Therefore the performance of algorithm is improved.The simulation experiment of urban rail train operation shows that the convergence and diversity of the proposed algorithm are superior to the multi-objective particle swarm optimization(MOPSO).Finally,the optimal driving pattern of automatic train operation speed profile of urban rail train was chosen through the fuzzy membership method.Multi-train multi-objective optimization: Under the various constraints of traindriving and tracking,the multi-target optimization model of multi-train tracking is established with the goal of passenger travel time and total train energy consumption.Using the immune clone particle swarm optimization algorithm,combined with time margin,to adjust the departure interval and stop time,optimizing the passenger travel time,meanwhile increasing the overlap time between acceleration and braking of adjacent trains as much as possible,and improve the utilization rate of regenerative-braking energy,so as to reduce the total energy consumption of train operation.In the optimization process of the algorithm,the dynamic grid ordering external archive update strategy is used to help the algorithm speed up the finding of excellent train driving curves and train schedulings.