Research And Analysis Of Target Location Interference Model Based On Big Data Mining

As a large number of electronic equipment applying to the future battlefield,the electromagnetic environment in the battlefield are becoming more and more complicated.It is more and more difficult to obtain accurate information.All countries attach great importance to the study of the complex electromagnetic environment,and the domestic research has just started in this area.There are still many problems to be solved.In view of this situation,this paper refers to the domestic and foreign research results,We have an intensive study in complex electromagnetic environment based on data mining in the target location.This topic is also the first time to apply the theory of data mining to the analysis of complex electromagnetic problems.This paper studies some key technologies of the target location interference model,and provides an optimization method for the analysis of electromagnetic model.This paper studies acceleration method on the simulation computation of target localization interference model.Aiming at the feature of the Finite Difference Time Domain method(FDTD),We put forward the parallel computing model and the implementation process based on MPI+OpenMP mixed-technology.The efficiency of parallel algorithm is analyzed.The acceleration effect is obvious.Then,this paper mainly studies the Monte Carlo positioning method based on particle filter,and puts forward the realization scheme of the localization algorithm.We compare it with other methods.This method can realize the localization more efficiently.The key technologies of interference analysis model are studied can be applied to the target location,the main achievements are as follows.(1)The scheme of MPI+OPENMP hybrid FDTD parallel computing.is proposed.The constrains of the parallel algorithm is studied.We Complete parallel algorithms on the platform of Visual C++,and do the actual electromagnetic simulation.According to the computing and analysis,the parallel computing is efficient.(2)A model surface modeling and occlusion interference method based on NURBS method is proposed.Research on the surface sampling point coordinate parameters,tangential unit vector and normal unit vector calculation method.By MATLAB programming,we do the actual calculation on relevant examples.We do research on parameters of NURBS surface model source occlusion judgment method,field points,and tracing algorithm based on NURBS surface reflection.The calculated results were compared with reference,which demonstrates the feasibility of the proposed algorithm.(3)This paper first puts the application of data mining rough set theory algorithm to the complex electromagnetic analysis model.We focus on analysis accelerated effect based on neighborhood model of forward greedy numerical reduction.The algorithm is applied to the reduction calculation of large data sets generated by different radiation sources.The analysis results show that the proposed data mining algorithm is feasible in the analysis of complex electromagnetic problems.