Font Size: a A A

Application Of Neural Network And Support Vector Machines In Effectiveness Evaluation Of UAV

Posted on:2019-01-11Degree:MasterType:Thesis
Country:ChinaCandidate:N K HuFull Text:PDF
GTID:2322330542956362Subject:Control engineering
Abstract/Summary:PDF Full Text Request
In modern warfare,unmanned aerial vehicle(UAV)will play a more and more important role in protecting national airspace security.Although UAV plays such an important role,in order to better excavate the operational potential of UAV,it is necessary to evaluate the combat effectiveness of UAV.The evaluation of combat effectiveness is to evaluate the completion degree of the operational task,and the reliable effectiveness evaluation plays a positive feedback role in improving the operational plan.Therefore,the in-depth study of the combat effectiveness of UAV possesses a very significant engineering value and theoretical significance for guiding warlike operations.In this dissertation,adaptive particle swarm optimization(APSO)is combined with back propagation neural network(BPNN),support vector machine(SVM),and wavelet neural network(WNN)to construct new evaluation methods,and then evaluate the operational effectiveness of the ground attack UAV,the reconnaissance UAV and the electronic war UAV by the above methods.First of all,in view of the engineering problems of combat effectiveness of air-to-ground attack UAV,this paper presents a method for effectiveness evaluation which is based on BPNN optimized by APSO.Then the proposed method is compared with the PSO-BPNN and BPNN,the simulation results show that the proposed method has more reasonable initial weights and thresholds,so that the information fusion ability is further enhanced,and the evaluation accuracy is higher.Secondly,aiming at the problems of the combat effectiveness of reconnaissance UAV,this paper proposes the APSO-SVM to manage this kind of engineering problems.Then the proposed method is compared with the PSO-SVM and SVM,and the simulation results show that the proposed method gets more reliable evaluation results of combat efficiency,and provides a reliable basis for rational collocation of reconnaissance equipment.Finally,in order to provide more bases for formulating a combat scheme,this paper uses the APSO-WNN to evaluate the combat effectiveness of electronic warfare UAV.Then the proposed method is compared with the APSO-BPNN and APSO-SVM,and the simulation results show that the proposed method can accurately judge the combat effects of the UAV and get more reliable evaluation results of the combat effectiveness.
Keywords/Search Tags:combat effectiveness evaluation, adaptive particle swarm optimization(APSO), back propagation neural network(BPNN), support vector machine(SVM), wavelet neural network(WNN)
PDF Full Text Request
Related items