Hip exoskeleton walking aid method based on neural network and robust constraint following

The invention discloses a hip exoskeleton walking aid method based on a neural network and robust constraint following. The hip exoskeleton walking aid method comprises the steps that an encoder detects changes of joint angles; converting the change of the joint angle into an electric signal, and inputting the electric signal into a BP neural network; a BP neural network is designed, and input signals are classified and screened; training a neural network, and returning an output result to the robust constraint following controller; the robust constraint following controller performs trajectory tracking control on the hip exoskeleton robot, and the control method can quickly and correctly make a decision when the system judges the change of the joint angle, has the characteristics of self-learning property and high reliability, effectively improves the walking aiding effect, and provides great convenience for the robot industry. 本发明公开了一种基于神经网络和鲁棒约束跟随的髋部外骨骼助行方法，包括编码器检测关节角度的变化；将关节角度的变化转换为电信号，输入到BP神经网络；设计BP神经网络，