Research On Case Base Maintenance Technology Of Energy-saving Oriented Rail Transit Metro Diagram

With the development of social economy,the concept of green energy conservation and emission reduction is gradually receiving more attention.Therefore,more and more people choose to ride on urban rail transit with fast speed and high punctuality.However,due to the large volume of urban rail transit,its energy consumption is also large.It is urgent to study the energy-saving technology of subway.This research group has cooperated with the subway operation enterprises for more than ten years and found that there must be the "optimal operation diagram" with the lowest energy consumption for the subway operation in a certain situation.However,it usually takes several days to calculate the optimal diagram through the simulation system,which cannot meet the needs of the subway in special situations.Therefore,this research group applies case-based reasoning technology to the design of energy-saving oriented case base construction of metro diagram.However,due to CBR's incremental learning characteristics,only by continuously adding new cases to the case library can the system's ability to solve problems be improved.With the passage of time,the number and types of metro diagram generated by rail transit line operations are increasing,which not only enhances the ability of the case-based reasoning system to solve problems,but also increases the time of case retrieval.When the time reaches a certain limit,the overall performance of the CBR system will decrease.Therefore,how to control the scale of the case library,ensure the ability of case reasoning to solve problems,and realize the maintenance of the case library are the main research goal of this paper.By summering the factors that affect the energy consumption of rail transit metro diagram and the related technologies of case base maintenance,this paper maintains the metro diagram case base oriented energy-saving from case base structure adjustment and redundant case deletion.First,the case of metro diagram is represented by the frame representation method and the triple method.A comprehensive weight determination method is proposed based on gray correlation and AHP.A high-dimensional similarity calculation method is adopted to solve the low discrimination between case similarity,and the method is verified by experiment.The density-based clustering method DBSCAN is used to adjust the structure of the case base.The method for determining the distance parameters of the DBSCAN algorithm is improved,The distance parameters is determined using the slope and mean theory,which solved the problem that common methods cannot cover all the data sets.In order to select representative case for every sub case base,the 2K distance outlier algorithm is used to calculate the density value DNR of every case,and the representative case of each sub case base is calculated based on this basis.In order to retain important cases in the case base,Construction Time and Retrieval Times are added to the case,and the concept of case retrieval ratio is proposed.The deletion function is established based on case coverage,case retrieval ratio,and case energy ratio.Prior to case deletion,redundant cases are classified through redundancy detection.The validity of the function is evaluated from the energy consumption and the precision of the retrieval results.Based on the above methods,the energy-saving oriented rail transit metro diagram is improved and maintained,and the validity of the maintenance method be verified from the energy consumption and retrieval time.