摘要：为了不需要精确数学模型实现对超声马达运动的快速精确控制，在基于DDS的超声马达驱动控制平台上，运用单神经元自适应PID控制策略，分别以低频脉宽调制的占空比和频率控制字作为控制变量，对超声马达速度响应特性进行实验研究，并与传统PI控制策略进行比较，结果显示，采用单神经元自适应PID控制策略以后，超声马达对阶跃给定转速的响应速度加快，超调量减少，稳态运行精度有所提高。超声马达单神经元自适应PID控制系统的控制器结构简单，控制算法融合了神经网络控制和传统PID控制的优点,控制系统具有较强的自适应性，能够有效的提高超声马达转速跟踪控制的动静态性能。
关键词：超声马达；神经网络；单神经元；PID；自适应；运动控制
Abstract: In order to realize the fast and precise control of ultrasonic motor without precise mathematical model, the single neuron adaptive PID control strategy was used on the ultrasonic motor drive control platform based on DDS. The duty cycle and frequency control word of low frequency pulse width modulation were used as control variables to control the speed of ultrasonic motor. The experimental results show that the single neuron adaptive PID control strategy can speed up the response speed of the given step speed, reduce the overshoot and improve the steady state motion control performance. In this single neuron adaptive PID control system ,the controller structure is simple, the control algorithm combines the advantages of neural network control and traditional PID control, the control method has strong selfadaptability .The control system can improve the dynamic and static performance of ultrasonic motor speed tracking control effectively.
Key words: ultrasonic motor; neural network; single neuron; PID; adaptive; motion control