The paper proposes a hierarchical method for learning rate adjustment of BP neural network algo-rithm by analyzing the learning rate effect on the BP algorithm. And the PID controller with the improved al-gorithm is applied in brushless DC motor system. The results of MATLAB simulation present that the PID controller based on improved BP neural network has characteristics of fast response, high accuracy and strong stability etc.%通过分析学习速率对BP算法的影响，提出一种分层调整学习速率的改进BP网络算法，并把该方法设计成PID控制器应用在无刷直流电机控制系统中，仿真结果验证了基于改进的BP网络的PID控制器具有响应速度快、精确性高和稳定性强等特点。
展开▼
