Study On Trajectory Planning And Rehabilitation Evaluation Of Upper Limb Rehabilitation Robot Based On Improved Whale Optimization Algorithm

As the aging population increases,the number of elderly people gradually increases.Stroke is one of the major diseases that affect the elderly population,and stroke patients usually have hemiplegia as a sequela.The traditional rehabilitation method involves rehabilitation therapists assisting patients in rehabilitation training,which is not only time-consuming and laborious,but also the therapeutic effect depends on the therapist’s personal experience,and there are differences in multiple training sessions.With the development of rehabilitation medicine and mechatronics,rehabilitation robots have become a new method of assisting hemiparalysis patients in rehabilitation training.As a rehabilitation equipment that closely interacts with human body,trajectory planning of rehabilitation robot is one of the keys to assist patients in rehabilitation training.This paper focuses on the four-degree-of-freedom upper limb exoskeleton rehabilitation robot,and mainly studying the trajectory planning of rehabilitation robot and the rehabilitation assessment based on Surface Electromyography signal(s EMG)muscle force estimation.The main research work includes:(1)Robot kinematics is the basis for analyzing the workspace,motion planning and motion control.Firstly,the DH parameters of the rehabilitation robot were established,and the forward kinematics of the rehabilitation robot was calculated and analyzed.Then,the inverse kinematics of the rehabilitation robot was solved,and the forward kinematics and inverse kinematics simulation experiments were carried out to verify the correctness of the rehabilitation robot kinematic equations.Finally,the kinematics and dynamics models of the rehabilitation robot were established in Simscape Multibody environment.(2)To maintain a uniform or slowly changing motion and a smooth and nonsudden requirement during the patient’s rehabilitation training process,a three-segment polynomial rehabilitation trajectory containing a uniform speed segment was proposed.The acceleration and deceleration segments of the trajectory adopt fifth-order polynomial functions,and the middle segment is a linear function.In order to make the three-segment polynomial trajectory more in line with the natural upper limb movements,two optimization models of the trajectory were established,one is a single objective time-optimal trajectory optimization model,and the other is a two-objective time-jerk optimal trajectory optimization model established using the coefficient weighting method to meet the needs of patients in different rehabilitation stages.(3)In order to solve the problem of insufficient global search ability of the Whale Optimization Algorithm(WOA)under complex trajectory optimization constraints,three strategies were adopted to improve the Whale Optimization Algorithm,and a multi-strategy improved Whale Optimization Algorithm(MWOA)was proposed.To verify the performance of the MWOA algorithm,it was compared with four algorithms including the Grey Wolf Optimization algorithm in 14 standard test functions.The numerical experiments show that the MWOA algorithm has faster convergence speed,higher convergence accuracy and better robustness.The trajectory optimization experiment of the rehabilitation robot verified that MWOA algorithm has better global search ability and robustness.(4)To achieve good trajectory tracking performance of the rehabilitation robot,the parameters of the PID algorithm were optimized using the MWOA algorithm,and a trajectory tracking algorithm for the rehabilitation robot based on the MWOA-PID was proposed.The simulation experiments show that MWOA-PID control algorithm is fast response and small overshoot.Finally,the MWOA-PID control algorithm was used for position tracking experiment of the rehabilitation robot,and the experiment show that the MWOA-PID control algorithm has good position tracking accuracy with a maximum position tracking error of 0.033 rad.In order to overcome the impact of various disturbances from patients during rehabilitation activities,an improved reaching law sliding mode control was proposed.The simulation results show that the improved reaching law sliding mode control is insensitive to small disturbances in the system and has good robustness.(5)In response to the problem that the currently commonly used scale-based rehabilitation evaluation methods cannot quantitatively evaluate rehabilitation effects,a rehabilitation evaluation method based on s EMG signal muscle force estimation was proposed.Firstly,a muscle force acquisition device was designed,then,the s EMG and muscle force signals of the subjects were simultaneously collected,and the root mean square,mean absolute value,mean frequency and spectral moment ratio features of the s EMG signals were extracted after filtering.Finally,the Support Vector Machine was used for constructing the muscle force estimation model.The average error of muscle force estimation for six subjects is 9.5%MVC.The experiments show that the rehabilitation evaluation method proposed in this paper can provide quantitative rehabilitation evaluation for rehabilitation trainers and has a certain medical reference value.