Upper limb exoskeleton rehabilitation robot control method based on radial basis neural network

Abstract

The invention discloses an upper limb exoskeleton rehabilitation robot control method based on a radial basis neural network. The method includes the steps that a human body upper limb musculoskeletal model is established; upper limb muscle myoelectric signals and upper limb movement data are collected, the movement data are imported into the upper limb musculoskeletal model, upper limb joint torque is obtained, the radial basis neural network is established and a neural network model is given out; the patient movement intention is recognized, the joint angular speed is subjected to fusion analysis, the result is used for recognizing the training object joint stretching state, and the limb movement intention is determined; and myoelectric signals and joint angles in affected side rehabilitation training are collected in real time, affected side joint torque is obtained through the neural network, joint torque needing to be compensated by an exoskeleton mechanical arm is calculated, myoelectric signal fatigue characteristics are analyzed, the compensation torque magnitude can be adjusted by classifying the degree of fatigue, and a torque controller can be controlled to achieve the effect that an upper limb rehabilitation robot assists patients in rehabilitation training by combining the movement intention. By means of the upper limb exoskeleton rehabilitation robot control method, the rehabilitation training process is made to be more suitable for the patients, man-machine interaction is strengthened, and the rehabilitation effect is improved.

Description

technical field [0001] The invention relates to the field of exoskeleton-assisted rehabilitation robot control, in particular to a control method for an upper limb exoskeleton rehabilitation robot based on a radial basis function (RBF) neural network. Background technique [0002] The wearable human exoskeleton robot assists the human body to complete actions through precise mechanical devices. It combines exoskeleton bionic technology and information control technology, and involves interdisciplinary knowledge such as biokinematics, robotics, information science, and artificial intelligence. The variety of exoskeleton systems also leads to the diversity of control methods for exoskeleton robots. From the perspective of system structure, it can be divided into two types: lower extremity exoskeleton and upper extremity exoskeleton. At present, the control methods for exoskeleton rehabilitation robots mainly include myoelectric control, force feedback control, and sensitivity...

AI Technical Summary

Problems solved by technology

The myoelectrically controlled exoskeleton rehabilitation robot mainly relies on the movement signal sent by the central nervous system of the human brain to change the characteristics of the surface muscle electrical signal of the human body, thereby controlling the rehabilitation robot. One-to-one relationship; in order to determine the relationship between muscle force and joint torque, the muscle force arm must also be determined, and the force arm usually changes with the change of the joint angle, so relying solely on myoelectricity makes the control of rehabilitation robots complicated. The degree is greatly increased, and this control method can only simply enable the patient to control the rehabilitation equipment, and the rehabilitation effect is not particularly obvious; the force feedback control maintains the force of the machine and the environment at a pre-set value through the feedback information of the force sensor, but The number and installation position of force sensors are difficult to determine; sensitivity amplification control does not need to install a large number of sensors between the wearer and the exoskeleton machine. This method maximizes the sensitivity function of the force exerted by the person output to the exoskeleton through the controller, Realize using the smallest force to change the position state of the exoskeleton, but this method is too dependent on the dynamic model of the system, and the actual exoskeleton system is a complex nonlinear system, so it is very difficult to establish an accurate model

[0003] In summary, there is still a lack of a better control method for exoskeleton rehabilitation robots

Images