Abstract:

Traditional sonobuoy polyline interception arrays suffer from low search probability and low search efficiency. Based on this, a back propagation (BP) neural network optimized by genetic algorithm (GA) is employed to enhance the performance of conventional buoy polyline interception arrays. Improving the inherent problems of BP neural network, firstly, using dynamically adjusted learning rate to solve the problem of fixed learning rate, and improving its loss function. The second is to use GA to optimize BP neural network and solve the problem of initial weight and random threshold in BP neural network. The proposed method is compared with the traditional GA optimized buoy polyline array and the traditional buoy polyline array method. The simulation results show that the proposed method not only improves the search probability, but also greatly reduces the planned buoy polyline interception array deployment time. Therefore, the GA-BP based buoy polyline interception array deployment planning method can effectively improve the submarine search efficiency.

Key words: buoy deployment, genetic algorithm, neural network, submarine search probability