本文讨论了目标识别的协同方法在不平衡注意参数条件下的动力学行为 ,并提出了不平衡注意参数条件下的遗传协同学习算法 (GSLA) .该算法利用遗传算法的全局最优搜索能力 ,对协同神经网络的注意参数进行全局优化 .对从“车牌识别系统”中得到的数字样本的实验证明 :新算法能有效地在注意参数空间搜索全局最优解 ,挖掘出协同方法在目标识别方面的最大潜能 .另外 ,本文还将新算法与利用奖惩学习算法的协同学习算法进行了全局优化能力的比较 ,发现新算法具有收敛快和全局最优搜索能力强的特点 In this paper, we discuss the dynamic behavior of the collaborative method of target recognition under unbalanced attention parameters and propose a GSLA algorithm under the condition of unbalanced attention parameters.This algorithm uses the global optimal search ability of genetic algorithm, The parameters of the collaborative neural network are optimized globally.Experiments on digital samples obtained from the License Plate Recognition System show that the new algorithm can effectively search the global optimal solution with attention to the parameter space and find out the cooperative method in target recognition In addition, the new algorithm is also compared with the global optimization ability of the cooperative learning algorithm using the reward-punishment learning algorithm, and found that the new algorithm has the characteristics of fast convergence and global optimal search ability