基于人工鱼群优化BPNN的AUV目标识别方法

为了削弱复杂恶劣的环境对水下成像造成的不利影响及满足水下机器人目标识别任务实时性的需求,提出了基于人工鱼群算法(AFSA)优化BP神经网络的水下目标识别算法,通过构造组合不变矩对水下目标进行特征提取,提高了目标的聚类性能.引入具有全局寻优能力的AFSA,其在增加单纯神经网络收敛速度的同时避免算法陷入局部最优,进而建立了完整的基于人工鱼群神经网络的水下目标识别系统.在不同的水下目标中对该系统进行实验,通过比较提取的不同的目标图像,结果表明所建立系统具有较优的聚类性能和较高的识别精度.该方法用于水下目标识别是可行的、有效的.

AUV Target Recognition Based on BPNN Optimized by Artificial Fish-Swarm Algorithm

In order to weaken the negative effects brought by the particularity and complexity of imaging environment,and satisfy the real-time need of the underwater task,the underwater target recognition algorithm to optimize BP neural network based on artificial fish-swarm algorithm (AFSA) is proposed.Combination of moment invariants is constructed for extracting recognition features.The clustering performance of targets is improved.AFSA with global optimization capability has been introduced,which can speed up converging velocity of pure BP neural network,and while avoid falling into local optimum,thereby a complete underwater target recognition system based on neural network of artificial fish is established.In the experiments,different kinds of targets images are used to test in the proposed recognition system,so as to some other optimized algorithm trained recognition system.Experimental results show that the proposed system has well clustering performance and high recognition accuracy.The method is feasible and effective for underwater target recognition.