| The ZPW-2000 A track circuit independently developed by China has been widely used in various railway lines to realize multiple functions such as train occupancy supervision and transmission of train information.It is one of the important basic equipment to ensure the efficient and safe operation of trains.At present,the maintenance strategy for the ZPW-2000 A series of track circuits is mainly based on the service time and relies on manual experience to formulate maintenance plan.Such maintenance plan will cause insufficient or excessive maintenance of the equipment,hidden safety hazards and waste of maintenance resources.Therefore,formulating a set of scientific and effective maintenance strategies to ensure the safe and reliable operation of the track circuit while reducing maintenance costs is of great theoretical and practical significance for achieving refined and systematic maintenance.The research content of this thesis is as follows:(1)According to the composition characteristics and functional structure of the ZPW-2000 A track circuit,an improved fuzzy analytic hierarchy process is used to evaluate the importance of the components,and the key factors affecting the importance of the components are proposed from the perspectives of reliability,maintainability and economy.The global weight of each key factor is calculated by the improved fuzzy analytic hierarchy process and the fuzzy comprehensive scoring method is used to determine the weight of different components on the evaluation set(important components,general components),a "productsum" operator is introduced to realize the importance evaluation of components.(2)Aiming to maintain the important components of the track circuit,a preventive combination maintenance strategy that takes into account non-maintenance,minor repairs,preventive maintenance and preventive replacement in each flexible maintenance cycle of the system is proposed.The GO-FLOW method is used to analyze the dynamic reliability of the system and components during the service period.From the differences in component-level maintenance,construct preventive maintenance and replacement metrics to analyze the component’s choice of maintenance methods during the service period.Contrapose to the incompleteness of preventive maintenance and combined with the dynamic reliability of components,the dynamic degradation process of components including the service life regression factor and failure rate increase factor is deduced.Through comprehensive analysis of the economic correlation between system reliability and various maintenance methods,establish the preventive combined maintenance model with the goal of minimum unit time cost,which based on whether the reliability of system and components meet the constraint conditions.(3)The maintenance decision scheme for the important components in ZPW-2000 A track circuit is a non-linear optimization problem with multiple constraints.To solve the model well,a crossover operator of multi-strategy search direction and adaptive combinatorial mutation operator are proposed for the improved genetic algorithm.By analyzing the difference of the population,the proposed algorithm develops different cross strategies and mutation selection for the excellent individual and the disadvantaged individual,which improves optimization ability of the algorithm.To verify the effectiveness of the preventive combined maintenance model,a comparative analysis with various maintenance strategies is proposed to show that the combined maintenance model can provide more optimized maintenance strategy support and significantly save maintenance costs.Compared with the improved genetic algorithm in other literatures,the results show that the improved genetic algorithm mentioned in this thesis has better local and global search capabilities. |
PDF Full Text Request