Sound targets recognition based on hybrid algorithm of an improved adaptive particle swarm optimization and BP neural network

Object recognition using BP neural network is a common method nowadays. However, BP neural network algorithm is easy to fall into local extremity and exists shortcomings such as the slow training process. This paper proposes a sound targets identification method for battlefield multi-target detection environment. This method can improve BP neural network using the adaptive particle swarm optimization (APSO) and increase the convergence speed as well as the training accuracy of BP network. Experiment using sound targets show that the identification and recognition result of this method is better than the traditional BP algorithm recognition result.