中文摘要：

提出一种采用无刷直流电动机设计的左心室辅助装置,通过对自然心脏的搏动规律的分析进行建模仿真。先采用传统PID控制,获得速度环的偏差、偏差变化率以及转速的数据,然后将这些数据在MATLAB的自适应神经模糊工具箱进行训练来设计模糊控制器。结果表明：训练后得到的模糊控制器要比传统PID控制器设计简单,控制响应速度快,鲁棒性好。

英文摘要：

The left ventricular assist device adopting the brushless DC motor was designed.The model built on the basis of the throbbing regular analysis of the natural heart was simulated with the traditional PID control to get the data of the speed deviation,the deviation rate of the speed and the speed in the feedback of the speed.Then these data are input into the adaptive neural fuzzy toolbox in MATLAB for training to design the fuzzy controller.The results show that the design of the trained fuzzy controller is simpler than the traditional PID controller,the response speed is faster and the robustness is better.