The Clustering Analysis Method Of Electromagnetic Effect Data Of UWB On Radio Fuze

The electromagnetic environment that radio fuze confronted with is becoming moreand more complicated in modern battlefield.Especially the energy interference waveform ofhigh power microwave weapon, produced a huge threat to its viability in battlefield. As thenumber one killer, the ultra wideband electromagnetic pulse has a combination of factors tothe fuze.There is a direct relationship between and the pulse width, frequency, peak powerand other parameters, which has attracted wide attention and research.A lot of experimental study of ultra wideband electromagnetic effect had carried out athome and abroad, and a large number of data had wonned.The key step in radio fuze ultrawide band electromagnetic environment effect evaluation is to identify and extract therelevant parameters effectively, which is helpful to obtain the relationship between ultrawideband signal with the parameters.But because of the characteristic of UWB signals withmulti parameters determines the high dimensional and complex electromagnetic effect data.There is no mature techniques and methods to analysis these data very good at present.Andbased on this, the clustering analysis algorithm is proposed in this paper.The main works are as follows:Firstly, based on the traditional analysis of the electromagnetic effect data, aiming atthe existing problems and shortcomings, and combined with the features of ultra widebandelectromagnetic effect data,the clustering data analysis algorithm is proposed.Secondly, the related theory about clustering analysis is detailed introduced, and thegeneral process of clustering is clear. Two classic fuzzy clustering algorithm are introduced:K means clustering algorithm and fuzzy C means clustering algorithm. In addition, thecauses of selecting of fuzzy C means in this paper is explained by comparing the two kindsof algorithm.Finally, the simulation of ultra wideband electromagnetic effects data on a certain typeradio fuze is carried out with the method of fuzzy C mean clustering algorithm.Theidentification and classification of effect data between two parameters is realized intwo-dimensional space using FCM algorithm. And the simulation results is verified furtherwith KFCM algorithm, which can realize the classification in three-dimensional spaceeffectively.The best threshold injury of the main parameters of UWB electromagnetic pulse on the model of radio fuze are obtained by the experimental results, which proves the validityof clustering algorithm. The design results are then explored to be used in the assessment ofradio fuze UWB electromagnetic effect.