SMA驱动柔性机械臂BP神经网络PID控制方法研究  

Method of BP neural network PID control for flexible manipulator driven by SMA

作　　者：王扬威 吕佩伦 郑舒方 张铂轩 李三平[1] 李兴东 WANG Yangwei;LÜ Peilun;ZHENG Shufang;ZHANG Boxuan;LI Sanping;LI Xingdong(Shool of Electrical and Mechanical Engineering,Northeast Forestry University,Harbin 150040,China)

机构地区：[1]东北林业大学机电工程学院,黑龙江哈尔滨150040

出　　处：《现代电子技术》2022年第10期176-181,共6页Modern Electronics Technique

基　　金：国家自然科学基金项目(52075089);黑龙江省自然科学基金项目(LH2019E008);东北林业大学中央高校基本科研业务费(2572018BF03)。

摘　　要：针对形状记忆合金(SMA)丝驱动软体仿生柔性臂弯曲角度难以精确控制的问题,文中提出一种BP神经网络PID控制算法。该方法基于仿生柔性臂的运动学模型和驱动模型,利用电阻映射的弯曲角度反馈,采用BP神经网络优化PID参数实现位置控制。然后,搭建仿生柔性臂实验测试平台,利用不同驱动电压条件下弯曲实验测试数据确立初始PID参数,分别采用传统PID方法、模糊PID控制方法和BP神经网络PID控制算法对仿生柔性机械臂进行位置控制实验。结果表明,与其他两种方法相比,BP神经网络PID控制方法能在5 s内达到期望角度,调节时间更短,位置误差仅为2°,并且在外界扰动条件下2 s内可完成自适应调节,具有较强的鲁棒性,能够更好地补偿SMA相变过程中的迟滞问题,证明了所提方法的有效性。In allusion to the problem that it is difficult to accurately control the bending angle of the soft bionic flexible arm driven by shape memory alloy(SMA)wire,a BP neural network PID control algorithm is proposed. This method is based on the kinematic model and driving model of the bionic flexible arm,and the BP neural network is used to optimize PID parameters to realize position control by means of the bending angle feedback of resistance mapping. An experimental test platform for the bionic flexible arm was built,and the initial PID parameters were established using the test data of bending experiment under different driving voltage conditions. The position control experiment of the bionic flexible manipulator was carried out by means of traditional PID,fuzzy PID control method and BP neural network PID control algorithm,respectively. The research results show that in comparison with other two methods,the BP neural network PID control can reach the desired angle within 5 s,the adjustment time is shorter,the position error is only 2°,the adaptive adjustment can be completed within 2 s under external disturbance conditions,has strong robustness,and can better compensate the hysteresis in the SMA phase change process,which proves the effectiveness of the method.

关 键 词：仿生柔性臂 BP神经网络PID 弯曲角度 位置控制 形状记忆合金丝(SMA) 电阻反馈 

分 类 号：TN711-34[电子电信—电路与系统] TP24[自动化与计算机技术—检测技术与自动化装置]