基于爬楼轮椅座椅的神经网络优化PID控制研究

Research on Neural Network Optimization PID Control Based on Climbing Wheelchair Seat

为了解决爬楼轮椅在爬楼时由于机构变动给乘坐者带来的不适感,设计了一套座椅机构姿态调节系统。根据爬楼轮椅的功能需求,制定了座椅调节机构的机械结构,搭建了控制系统的硬件框架,根据PID算法参数的本身特性,借助双曲函数调整参数,再利用训练好的BP神经网络,补偿输出姿态误差,并对整个调节系统进行了试验。试验结果表明,该控制算法能实现对座椅机构的精准控制。

In order to solve the discomfort brought to the occupants due to the change of mechanism when the stair-climbing wheelchair climbs the stairs,a set of seat mechanism attitude adjustment system is designed.According to the functional requirements of the stair-climbing wheelchair,the mechanical structure of the seat adjustment mechanism is formulated,and the hardware framework of the control system is built.According to the characteristics of the parameters of PID algorithm,the parameters are adjusted with the help of hyperbolic function,and the trained BP neural network is utilized to compensate the output attitude error,and the whole adjustment system is tested.The test results show that the control algorithm can realize the precise control of the seat mechanism.