Research On Power Line Detection Technology For Helicopter Millimeter Wave Radar

Power line detection for millimeter-wave radar of helicopters is the import research direction in the area of preventing collision.The power line will cause the threat to helicopter because of the helicopter flying at low-altitude.This article will base on the problem of power line detection for millimeter-wave radar,analyzing the radar cross section of the power line and the tracking algorithms and the single pulse angle measurement and the classification algorithms that how to used in the algorithms of the detecting of the power line.The main work is listed as follows:1.This article analysis the radar cross section of the power line on the millimeter-wave radar.Analyzing the features of bragg pattern of the radar cross section by comparing the difference between infinitely long conductor cylinder model and periodic structural model.And analyzed statistical characteristics of the ground clutter,simulating the signal of the power line behind the ground clutter in the high SCR.2.Pre-treating on the actual data,including the pulse compression,coherent accumulation.The SNR is improved by these algorithms.High threshold and low thresholdare used in the constant false rate.First using the high threshold to detect a strong scattering point to ensure that the power line frame does not appear undetected,and then using the low threshold to detect a weak scattering point,the power line signal does not appear undetected.On the basis of CFAR detection,the ?? filter and kalman filter are respectively used for tracking and filtering to obtain the approximate trajectory of the power line.Firstly,the filtering process of the two filter algorithms on the power line target distance and extrapolation slope is simulated,and the root mean square error between the filter distance,extrapolation slope,true distance and extrapolation slope is obtained.Simulation results show that kalman filtering is better than ?? filtering,and the two filtering algorithms are used to process the measured data,and the better results are obtained.3.Using the difference between the angle caused by the height difference between the power line frame and the ground clutter,the single-pulse angle measurement technique is used to distinguish and reduce the false alarm rate caused by the CFAR detection.This section analyzes the effects of signal-to-clutter ratio and signal-to-noise ratio on the accuracy of the measuring angle in the phase-difference and single-pulse phase contrast technique.In the measured data,the error of the strong scattering point voltage CFAR detection,thereby reducing the false alarm rate.4.The clutter parameters of the measured data are estimated using geometric arithmetic average estimation,maximum likelihood estimation,least square estimation and moment estimation respectively.The first three algorithms are better than the others,and the fitting result of moment estimation algorithm is better difference.The common supervised learning classification algorithms are analyzed,including support vector machine(SVM)algorithm and neural network algorithm.The signals of power line echo are simulated respectively under different signal-to-noise ratios and classified by two classification algorithms.The simulation results show that the SVM classification effect is better than the neural network classification under the low SCR.Because supervised learning classification requires a large number of samples and can not be applied to the environment of the measured data itself,the k-means clustering algorithm which takes unsupervised learning uses the difference between the two echo signals to correct the linearity and clutter of power lines straight line classification.