本文在研究协同神经网络梯度动力学过程的基础上,针对学习过程收敛速度缓慢的缺点,提出了一种改进的基于梯度动力学的协同神经网络学习算法。-In this paper, the study of Synergetic Neural Networks gradient dynamic processes on the basis of the learning process for the shortcomings of slow convergence speed, an improved grad