The Research Of Dynamic Lower Limb Prosthesis Based On Flexible Iterative Control Of Finite State Machine

Dynamic lower limb prosthesis can follow the motion of the healthy limb side, and provide power support for the movement which need to be provided high torque of the knee joint motion, such as rapidly walk, climb up stairs, upslope. Dynamic lower limb has become major research and design direction, has important the theoretical significance and application value.This thesis contains three sections: information acquisition and feature analysis, database building and the motion control. This thesis mainly finished includes:Firstly, dynamic lower limb prosthesis movement decision procedure and artificial limb actuator delay always bring limb movement delays, resulting in movement of legs is not natural, and asymmetry. In order to overcome this drawback, this thesis collected the hip angle, acceleration and plantar pressure signal of human motion, and apply the acceleration signal on the hidden Markov model for pre-recognition of road condition, and take some key points of the acceleration sensor signals and foot pressure sensor signals for the division of gait and calculation of pace.Secondly, taking a healthy or disabled people sound leg knee joint motion curve as the expected curve, build up the database of five kinds road condition and the there pace through iterative learning control(ILC) for the knee joint movement control.Finally, according to the sequence of each gait phase transformation put forward the method of finite state machine(FSM) based on iterative learning control, which according to the output of finite state machine, to control the selection of control knowledge, to realize the control in different road condition, space, gait of dynamic lower limb prosthesis, has greatly improves the dynamic type prosthesis safety and speed of execution.