Research On Modeling And Intelligent Control Method For Prosthesis System With Gait Adaptive Bionic Knee

The only way to restore the walking ability of lower limb amputee is to install lower limb prosthesis.Prosthetic knee is the most important component of the transfemoral prosthesis.Conventional prosthetic knee is a purely mechanical system and cannot adapt to the change of speed,gait phase and motion mode automatically.The lack of motion perception and interaction control system usually causes the reduced motion ability,abnormal gait and increased energy consumption.With the development of modern prosthetics,intelligent knee prosthesis,which integrates mechanical bionics,multi-source sensing and active control,has become the most important method to improve walking ability and realize physiological gait for transfemoral amputee.The key of intelligent knee prosthesis research lies in how to realize gait adaption through bionic system modeling and appropriate intelligent control method.Therefore,this work aimed to research the system modeling and intelligent control methods of gait adaptive bionic knee prosthesis.The research focused on the system modeling of the hydraulic damping intelligent knee prosthesis and the establishment of the intelligent interactive control mechanism and method which adapted to different human body,speed and motion mode based on kinematics and force information.The main research contents are as follows:(1)Structure and human-machine coupling model establishment of intelligent bionic prosthetic knee.The damping bionic way was determined by comparative analysis of damping and active intelligent knee prosthesis and working mechanism of normal knee joint.The different types of damping and structure characteristics of intelligent knee prosthesis were analyzed.The mechanical structure design method of intelligent lower limb prosthesis based on the hydraulic damping knee joint was proposed.The electric hydraulic damping cylinder for damping torque control of the prosthetic knee was mainly studied.The damping cylinder structure of direct push plate type with single-motor,butterfly valve-plate type with single-motor,quadrature type with single-motor,fanmouth type with double-motor and needle valve type with double-motor were proposed.Considering the characteristics of electronic control hydraulic damping cylinder of different structures,the experimental prototype was processed based on needle valve type with double-motor.The damping adjustment mechanism of intelligent prosthetic knee was explained through modeling and simulation analysis of swing phase of lower limb prosthesis system.(2)Research on gait transformation mechanism and control of intelligent prosthetic knee in typical motion modes.The hardware sensor system based on angle sensor,pressure sensor and accelerometer was constructed by analyzing the sensor information source and designing the special ankle pylon.The gait phase control interval and target of intelligent prosthetic knee were determined based on the study of gait phase of normal people and prosthesis wearers.The gait phase recognition algorithm based on kinematics and mechanics information was established.The gait phase identification methods and hydraulic damping control scheme of intelligent knee prosthesis in flat ground,up/down slope and up/down stairs were proposed by analyzing the gait phase of lower limb prosthesis wearers in different motion modes.In addition,the simulation on the motion pattern recognition was conducted by collecting the angle data of hip and knee of norm people.The motion pattern classification simulation were carried out based on BP,gradient descent with momentum BP and L-M algorithm.The feasibility of neural network algorithm to motion pattern classification was verified based on lower limb joint angles.By comparing the learning rate and error rate of training,it was found that L-M neural network had the best training effect,which provided a reference for future research on motion pattern recognition of intelligent prosthetic knee.(3)Research on speed adaptive control algorithms of intelligent knee prosthesis.Firstly,the speed recognition methods based on the plantar pressure and the landing time of the sole foot were compared and analyzed.The method based on the landing time of the sole foot was determined.Two speed adaptive control algorithms with different targets were proposed based on the research of physiological gait.One target was the maximum swing flexion knee angle.The target range was between sixty degree and seventy degree.The fuzzy logic control method was established.The other was the swing phase trajectory.The valve opening predictive control algorithm based on NARX neural network and particle swarm optimization was proposed.The simulation analysis of the predictive control in the swinging period was conducted,which proved its good gait following effect.Finally,the fuzzy logic speed adaptive control algorithm based on the maximum swing flexion knee angle was determined as the control algorithm of the experimental prototype.(4)Research on performance evaluation of intelligent knee prosthesis.By summarizing the research status of the performance evaluation of microprocessorcontrolled prosthetic knee,the common spatiotemporal parameters,angular,force parameters and physiological parameters used in the performance evaluation of microprocessor-controlled prosthetic knee were analyzed.The functional simulation and evaluation platform of microprocessor-controlled prosthetic knee was established.Firstly,the platform verification test was carried out to prove that it could well simulate lower limb movements of human.Then,the maximum swing flexion control and gait symmetry were studied based on the platform.The law of the physiological gait damping adjustment of the intelligent prosthetic knee under different speeds were revealed.Finally,a comparative study on the wearing tests of knee prosthesis was conducted.The results showed that the proposed speed adaptive control algorithm could effectively achieve the target by adjusting the valve opening at different speeds.The effects in preferred walking speed and maximum swing flexion knee angle were better than the conventional passive prosthetic knee.This research provided theoretic basis for gait adaptive intelligent bionic knee prosthesis.The innovation of design method of high performance intelligent knee prosthesis was realized.The progress of intelligent lower limb prosthesis industry was promoted effectively.