基于免疫进化神经网络的机械手逆运动控制

将免疫进化算法用于机械手逆运动神经网络控制，基于生物免疫系统的细胞克隆选择学说和生物进化过程中的变异思想构造了自适应变异算子，使系统能够根据环境条件自适应地确定各抗体的变异强度；通过亲和力抑制相似抗体生存并动态地产生新的抗体，以维持抗体种群的多样性。通过对机械手进行仿真试验，并与快速BP算法及标准遗传算法训练的结果进行比较，表明用免疫算法训练的神经网络具有良好的泛化能力，可大大提高机械手逆运动学求解精度。

Manipulator Inverse Kinematics Control Based on Immune Evolutionary Neural Network

An immune evolutionary algorithm was adopted for manipulator inverse kinematics neural network control. The algorithm constructed a self-adaptive mutation operator based on the cellular clonal selection theory of biology immune system and the mutation ideas of biology evolution process, which made the system decide the mutation intensity of each antibody according to their environment, similar antibodies were suppressed by computing the affinity of antibodies and new antibodies were produced dynamically to maintain the diversity. The simulation results of manipulator verify that the trained neural network by the immune evolutionary algorithm has better generalization ability than that of fast BP algorithm and standard genetic algorithm. It improves considerably the inverse kinematics solutions for maninulator.