Research On Joint Servo Control Of The Lower Limb Exoskeleton For Walking Assistance

In recent years,the use of lower-limb exoskeletons(referred to as exoskeleton)to aid the normal walking has gradually become a research hotspot in the field of medical rehabilitation.In this paper,the self-tuning control method is used to improve the selfadaptive capacity of the exoskeleton joint load,and the hardware and software design of the servo driver and the implementation of the relevant algorithms are accomplished.The main research content is as follows:Based on the analysis of the exoskeleton joint load theory,the joint servo control scheme was determined.Based on the establishment of the mathematical model of permanent magnet synchronous motor,the basic principle of vector control and the digital realization method of Space Voltage Vector Pulse Width Modulation(SVPWM)are analyzed;the exoskeleton load is theoretically analyzed,a Simulink simulation platform is built.The influence by load changes of the servo performance of the joint are evaluated.Finally,a self-tuning servo control scheme for the exoskeleton joint is proposed.Aiming at the problem of time-varying inertia and load torque,an on-line identification of mechanical parameters of the exoskeleton joint servo system is proposed.For the time-varying inertia problem,the simulation compares the fixed inertia and variable inertia identification performances of forgetting factor recursive least squares(FFRLS)inertia identification and Landau discrete-time recursive algorithm,and the result shows that the latter has faster convergence speed and higher accuracy.For the time-varying problem of load torque,a reduced-order Luenberger disturbance observer is built which parameter design is simplified.Finally,the FFRLS and reduced-order Luenberger disturbance observer is combined to form a mechanical parameter identification scheme.Simulations verify the effectiveness of the scheme.A three-loop servo system controller was designed,the disturbance torque feedforward compensation transfer function was derived,and simulate the exoskeleton joint servo.The zero-pole cancellation strategy was used to derive the PI controller parameter analysis formula of the current controller.The simulation comparison between the minimum peak criterion auto-tuning and the speed loop auto-tuning method based on the damping ratio.The result shows that the latter was better and did not affect the current loop bandwidth;PVT interpolation was used to improve servo compliance.In order to compensate the load disturbance,the disturbance feedforward compensation transfer function was deduced.Finally,combined with exoskeleton working conditions,the simulation verified the effectiveness of self-tuning control.The software and hardware design of the exoskeleton joint servo system was implemented,and the experimental experiment was performed on the back-to-back experiment platform to verify the algorithm.In terms of hardware,the main control circuit,power drive circuit,cascade power supply circuit,sensing circuit,and protection circuit are mainly designed.The system software is layered and modularly designed,and the main program and interrupt program design are implemented based on a finite state machine to achieve decoupling.Finally,based on the self-made driver,a back-to-back experiment platform was built to simulate the actual working conditions of the exoskeleton,and the validity and feasibility of the exoskeleton joint self-tuning control algorithm was verified.The self-tuning control algorithm proposed in this paper improves the self-adaptive capacity of the exoskeleton's joints,ensures the followability and stability of the servo,and is suitable for the servo of the exoskeleton joints with variable load conditions.