Adaptive Frequency Point Optimization Technique for Frequency Agility Radar Tracking Low Altitude Targets on the Sea Surface

‍ ‍In this paper， the elevation measurement error caused by multipath effect was studied when radar tracked low elevation targets on the sea surface. Firstly， the multipath reflection effect and frequency agility radar system model of sea surface signal were established， and the elevation error model of frequency agility radar under multipath effect was solved by bringing the target into the model. A low angle measurement error improvement method based on adaptive frequency agility optimization was proposed. The method used wideband frequency agility to remove the correlation between target signal and multipath reflected signal， was developed to reduce the multipath effect， and controlled the error range of elevation measurement by selecting the central frequency point. Then， Mesh Adaptive Direct Search Algorithms （MADS） was used to optimize frequency points with low elevation errors， which were automatically selected according to the multipath effects generated by different distances， altitudes and sea conditions. On this basis， a strategy was developed to combat the multipath of the low altitude target on the sea surface through the adaptive frequency agility of radar. The simulation results showed that the distance and height of the target， the frequency point of the transmitting signal and the sea condition had great influence on the elevation measurement error. When the parameters of the target and the sea surface remained unchanged， the elevation measurement error could be controlled by selecting the frequency point of the transmitting signal. The frequency point optimization algorithm based on mesh adaptive direct search could automatically select low error frequency points according to different conditions， and controlled the low Angle measurement error of each distance within $\pm \mathrm{2}\mathrm{m}\mathrm{r}\mathrm{a}\mathrm{d}$.Compared with the traversal frequency-point method and genetic algorithm， the CPU calculation time of MADS was reduced by nearly 6 times and 2 times， and the number of iterations was reduced by 20 times and 2 times， respectively. The algorithm greatly improve the frequency selection speed， reduce the number of iterations， and adaptively optimize the frequency points in different occasions according to the multipath effect， so as to reduce the influence of the multipath effect on the target elevation measurement， which shows the effectiveness of the optimization algorithm.