Multi-Function Radar Working Mode Recognition Based on Entropy Map
In order to address the problem that traditional work mode recognition methods cannot adapt to the complex and varied waveform characteristics of multi-functional radar,a detection method of the pulse signal se-quence of multi-functional radar state switching points was proposed in this paper,and the pulse signal sequence was segmented into single work mode samples based on wavelet transform.Then,from the perspective of time series analysis and data mining,a multi-functional radar work mode recognition method based on entropy map was realized.This method extracted approximate entropy,permutation entropy,sample entropy,and fuzzy en-tropy of different inter-pulse parameter sequences to form an entropy map,and used a convolutional deep neural network model to achieve intelligent recognition of multi-functional radar work modes.Simulation results showed that the proposed state switching point detection approach achieved a switching point detection accuracy of 85%when the false pulse rate or the missing pulse rate was 25%,and the operating mode recognition accura-cy was above 83%when the false pulse rate,the missing pulse rate and the parameter error were 20%and 8%.Respectively,the recognition performance of the proposed method was superior to the two comparative literature methods,which verified the effectiveness of the algorithm.
multi-function radarworking mode recognitionentropy mapstate switch point detectionCNN
NETL
NSTL
万方数据
