Research On Predictive Maintenance Support Resource Matching Algorithm For Active Radar

Radar maintenance resources play a crucial role in ensuring the combat effectiveness of radar equipment.In actual work,in order to ensure that radar equipment has high reliability,the use of troops will often be equipped with a large amount of maintenance support resources based on experience,which will often lead to serious waste of resources and increase maintenance costs.In order to avoid the problems of overcapacity and shortage of maintenance capacity,the dissertation studies the maintenance support for a certain type of active radar equipment,and uses big data technology to analyze and mine radar operating data,usage data,failure data,environmental data,and maintenance resource data.Firstly,the internal correlation between radar fault feature parameters,fault reason and maintenance resources is analyzed.Then radar faults are predicted based on radar operating parameters,and then the types and quantities of maintenance resources are determined according to radar fault types.Research work has important theoretical and practical significance for improving radar maintenance resource planning capabilities.Main work and characteristics of the thesis are as follows:1?Taking a certain type of active radar laser range finder as the research object,by collecting the operating parameters of the laser range finder,a hierarchical structure model of the maintenance resources under fault is constructed,and the laser range finder operating parameters,causes of failure and Fault type association mapping.2?The wavelet modulus maxima algorithm was used to denoise the fault feature parameters of the laser range finder.The rough genetic reduction algorithm was used to reduce the dimension of the fault data.Data preparation for subsequent data mining work.3 ? Based on the multivariable non-equal interval GM(1,m)model,the fault feature parameters of a certain type of active radar laser range finder are predicted,and the change value of the fault value of the range finder can be obtained for a given time interval,which can be used as a radar fault prediction.4?Based on the combined model of multivariable unequal GM(1,m)model and fuzzy neural network model based on genetic algorithm optimization,a fault prediction model of laser range finder was established to realize the prediction of the probability of occurrence of radar faults in future time intervals.Then,based on the correlation of radar fault feature parameters,fault types and maintenance resources,the types and quantities of required maintenance resources are obtained.Due to the time,the research results of the thesis only stay in the stage of theoretical analysis and cannot be applied in engineering.The validity of the algorithm still needs to be verified by the test results.However,as a solution,the method can play a role in aiding decisionmaking for the planning of equipment maintenance resources.