Research On The Bio-Mechanical Model Of The Pathological Tremor Suppression Robotics

The pathogenesis of the pathological tremor has not been clear, so the tremor can not be proactively prevented and cured thoroughly. For the different etiology and characteristics of tremor, the common therapy including drugs and surgery, can only work on a kind of tremor, and drug therapy often brings varying degrees of side effects. After many years research, some new suppression measures emerge, such as bio-power load, and rehabilitative training. Biomechanical load increases the damping and inertia of the arm by controlling the external load in patients, for aiming at the objective of elimination of tremor, which is also divided into two kinds of active and passive suppression methods.Tremor suppression is mainly based on the mechanical model of upper limb, with real time tremor movement signals regulating the additional equipment, to inhibit non-intention movement. This paper is based on the wearable tremor suppression of two degrees of freedom in my laboratory. The robot is made from the suppression mechanism,damper and the characteristics of pathological tremor,in order to research the bio-mechanical mode of upper limb on the human body.In addition, this paper proposes a biomechanical model. The model analyses and identifies the EMG of human muscle, joint angle and joint torque signal, in order to get access to biomechanical model. Deciding method of collection and measuring signals of the experimenter by using of surface EMG acquisition system, six-axis force sensors and angle sensors.A mechanical model in line with the mechanical characteristics of tremor movement is achieved, for the intention of inhibiting the tremor movement and decreasing the affect on the intention movement. Basing on the characteristics of the movements concentrating in different frequency ranges, the frequency is considered as the objective function. And then the zero-phase digital filter and priori filtering algorithm are selected,which are respectively used to process the joint angle and EMG that are measured during intention and tremor movement, in order to acquire the signal reflecting the characteristics of the upper-limb tremor.According to the distinguished EMG of tremor movement, the profile curve and the root mean square value of the signal are extracted. And then the eigenvalues of the EMG combined with joint angle and joint torque, respectively , are used as input and output of the system , drawing the relation model between EMG and joint angles based on the auxiliary variable method with additional forgetting factor. Then the RBF neural network is employed for calibrating joint torque and EMG, and the mechanical models varying with the physical state of the limb movements is built.The traditional mechanical model between the joint angle and joint torque is established with the kinetic theory and the engineering parameters. And then the simplified rigid model of the upper limb is compared with the bio-mechanical model established in this paper, to analysis the validity of the bio-mechanical model having been built.