论文部分内容阅读
针对水声信号的特性和无源声呐目标识别的特点，提出了一种有师自组织神经网络分类算法。该算法主要针对水声信号的样本不完备的问题，在前债神经网络中引入了多重神经元激活模型和自组织竞争学习算法，使无源声呐分类系统的泛化性能有了明显的提高。该算法采用分层学习策略，有效地节省了训练时间，同时减少了陷入局部最优解的概率。通过对实录海上无源声呐目标信号的分类实验，检验了算法的识别能力和泛化能力，实验结果责明该算法具有良好的泛化能力同时保持了较高的识别率．
Aiming at the characteristics of underwater acoustic signals and the characteristics of passive sonar target recognition, a self-organized neural network classification algorithm is proposed. The algorithm is mainly aimed at the incomplete sample of underwater acoustic signals. The multi-neuron activation model and self-organization competition learning algorithm are introduced into the former debt neural network, which improves the generalization performance of the passive sonar classification system. The algorithm adopts hierarchical learning strategy, which effectively saves the training time and reduces the probability of falling into the local optimal solution. The experiments on the recognition of real-time passive sonar target signals show that the algorithm has good generalization ability and high recognition rate.