ANN combined HGA makes the weight and the deviation converge to the global solutions,which escapes local optima of back-propagation network.

• 将混合遗传算法用于人工神经网络,训练出全局最优的权值和偏差,解决了反向传播网络收敛于局部极值的问题。