| Since the emergence of self-service terminals,maintenance issues have been outstanding.In the traditional equipment inspection and maintenance mode,the staff can arrange the inspection sequence of the equipment by intuition and experience,which leads to the inefficiency of the inspection work.The device inspection problem in this paper is essentially a TSP problem and is a classic combinatorial optimization problem.Under normal circumstances,using an accurate algorithm to solve practical problems has certain limitations.Therefore,heuristic algorithms such as genetic algorithm and ant colony algorithm are increasingly applied to solve the approximate optimal solution problem of combinatorial optimization problems.This paper mainly uses genetic algorithm to solve the planning problem of equipment inspection sequence in Lanzhou self-service terminal project.The main research work is as follows:(1)Analyze the inspection problem in this paper and obtain its mathematical model and objective function.And its NP completeness is proved.(2)Combining with the actual situation in this paper,three improvement strategies are proposed for classical genetic algorithm,namely: initial population generation strategy based on greedy thought,mutation operator of historical memory retention and selection operator of elite retention combined ratio selection.Improve the convergence speed of the algorithm.(3)Designed and developed the Lanzhou self-service terminal inspection system,using the improved genetic algorithm to plan the inspection sequence of the equipment,and presented it to the inspector in the form of map markers.At the same time,the statistics and recording functions of the work were completed.According to the experimental data analysis,the equipment after planning with the genetic algorithm can greatly improve the efficiency of the inspection.The completion of each round of inspection can reduce the distance by 61 kilometers and reduce the time by 3 days. |
