Research On Cross-sectional Area Of High Frequency Ground Wave Radar

High Frequency Surface Wave Radar(HFSWR)is one of the main means used for long-range maritime detection.The Radar Cross Section(RCS)characterizes the intensity of the reflected radar signal from the target.Accurately predicting the RCS of a target can improve the accuracy of target detection and identification.The HFSWR is interfered by sea clutter when detecting targets of ships at sea.Therefore,in order to improve the detection probability of HFSWR for marine vessel targets,sea clutter RCS and target RCS are researched in this paper.The main research contents are as follows:1.First-order and second-order sea clutter frequency spectrum formulas for shore-based HFSWR and first-order sea clutter frequency spectrum formulas for shipborne HFSWR are derived.Simulation experiments were used to further analyze the effects of wind speed,wind direction,and radar operating frequency on the RCS of sea clutter.In addition,the effect of ship speed on the first-order sea clutter RCS of shipborne HFSWR was also analyzed.2.An algorithm for estimating the vessel target RCS using Back Propagation(BP)neural network is proposed.In the algorithm,a BP neural network with a three-implicit-layer structure is used to predict the RCS of a large cargo ship target.The measured HFSWR vessel target data are correlated with the AIS point traces and combined with the inverse RCS data to form the model training set.It is verified by the measured data that the algorithm can better accomplish the estimation of vessel target RCS values.3.An algorithm is proposed to estimate the vessel target RCS with a long short-term memory(LSTM)model.The algorithm uses a three-layer network structure containing an input layer,an implicit layer and an output layer to achieve target RCS prediction for large cargo ships.The mean square error between the model output and the inverse RCS is minimized by adjusting the hyperparameters in the model.After verified by the measured data,the vessel target RCS prediction accuracy of this algorithm is better than that of BP neural network.Therefore,the trained LSTM model is used to estimate the target RCS in the region to be detected.Then,the estimated target RCS is substituted into the radar equation to derive the average power of the target echoes.In the subsequent area detection,the calculated average power is used as the detection threshold,and the echo data larger than the threshold is considered as the target.This detection method is combined with the Cell Averaging Constant False Alarm Rate(CA-CFAR)algorithm to improve the detection probability of HFSWR for maritime vessel targets.