Research On Optimization Of Urban Rail Transit Crew Planning Based On Improved Genetic Algorithm

As the last step of urban rail transit transportation plan,crew planning is a key step to ensure train operation and system operation.Under the current situation of rapid development of urban rail transit construction,how to efficiently integrate resources and use existing train operation data? It is of great and far-reaching significance to design and compile a crew plan that satisfies labor safety rules,balanced task distribution,high working efficiency and low operating cost.In this thesis,the research on crew planning is divided into crew scheduling problem and crew shift problem.Considering multiple constraints involved in the preparation process,optimization objectives are set and the scheduling model and shift model are constructed.The model is further optimized based on problem scenarios in actual operation,and an improved genetic algorithm is designed to efficiently solve the scheduling and shift model.Finally,the model algorithm is applied with Xuzhou metro operation data,and the crew planning results are analyzed and evaluated.The main research contents of this thesis are as follows:(1)The crew segment is divided according to the train diagram,and the crew work chain is generated according to the constraints of continuous duty duration.The indirect continuation time of the crew work chain is taken as the scheduling cost to construct the objective function.Constraints such as task duration,departure and return time,meal duration and rotation and continuation are comprehensively considered,and the penalty function for constraints such as rest duration is added as a part of the scheduling cost.Organize and form crew scheduling model;The crew shift model was formed by taking the scheduling result as the input and the workload distribution balance and average workload as the objective function of shift cost.Considering the continuous constraints of task types under the four-shift and three-shift operation mode,the load coefficients of different task types were set and the product of task duration was taken as the workload size.Combined with the problems encountered in actual operation,model constraints under three special scenarios of flight attendants’ leave,standby vehicle duty and shift at critical times were designed,and the newly added scenario constraints were summarized into the crew planning model.(2)Genetic algorithm for multi-objective optimization problem was designed to solve scheduling and shift problems.On the basis of traditional genetic algorithm,elite retention strategy was used to improve the gene selection process,adaptive crossover and mutation operators were used to improve the gene crossover and mutation process,and algorithm steps were designed for crew scheduling and shift model.(3)The operation data of Xuzhou Metro Line 1 were selected and substituted into the model algorithm to obtain the scheduling and shift results when the scheduling cost and shift cost converged,and to verify the application of the three scenario constraints designed in this thesis in the results of crew planning.(4)With working efficiency,task duration,cost size,workload variance and other parameters as evaluation indicators,it is proved that compared with existing operation schemes,the model designed in this thesis can achieve higher working efficiency,more balanced personnel rotation,and reduce the cost loss caused by non-labor time.