基于小脑内模机理的自平衡机器人平衡控制

Balance control of self-balancing robot based on cerebellar internal model mechanism

根据人脑系统中的感觉运动系统和中枢神经系统控制结构,模拟小脑的内模机理,建立了基于反馈误差学习的自适应逆模型,作为自平衡机器人的控制中枢,为机器人构建了具有生物学特性的感觉运动系统。小脑模型部分采用BP神经网络进行模拟,网络权值作为小脑模型内部的突触,通过突触修饰完成系统的自适应学习,与传统的反馈控制方法进行了对比分析,结果显示:采用本文提出的控制方法能够有效增强系统控制的稳定性。

In accordance with the control structure of sensorimotor system and central nervous system in human brain system,by simulating the internal model mechanism of cerebellum,an adaptive inverse model based on feedback-error-learn-ing is established,which functions as the control center of self-balancing robots.Hence,a sensorimotor system with biological characteristics is constructed.BP neural network is used to simulate the cerebellar model.With the network weights as the synapses in the cerebellar model,synaptic modification is used to complete the adaptive learning of the system.In comparison with the traditional feedback control method,the results show that the proposed control method can effectively enhance the stability of the system control.