The learning algorithm is a process of chaos optimization, which can make the network avoid the local minima problem and false saturation phenomenon. 网络的学习过程是一种混沌优化算法，可有效避免普通神经网络的局部极值和假饱和现象的发生。
At the same time it can make the solution escape local extreme value and cause the overall optimization capacity when the solution approximates the minimum point. 当解逼近当前极小点的后，动态方程又能使解逃逸局部极值，使其具有全局寻优能力。
