Rehabilitation Robot Perceptual System Based On Multi-sensor Design And Application Research

Thanks to the combination of rehabilitation medicine and engineering technology,the rehabilitation robot has achieved some success.As one of the key role in the rehabilitation robot system,the perceptual system not only needs to obtain the physiological information of the human body,but also is the precondition of the many applications such as intention recognition,spasm detection and local muscle fatigue assessment.In this paper,a multi-sensor perceptual system is designed for a treadmill rehabilitation robot and applied to the assessment of exercise-induced muscle fatigue.Firstly,this paper expounds the mechanism of sampling various information of human body,emphatically analyzes the mechanism of surface electromyography signal(sEMG),signal feature and source of noise and designs sEMG,the plantar pressure is designed based on the force measuring cantilever beam,designing a joint angle sensor using Hall angle sensor,pulse signal acquisition using PulseSensor.The spectral characteristics and noise of the sEMG were measured,the output noise is 0.545 mV.The sensitivity and nonlinearity of the plantar pressure sensor and the joint angle sensor were calibrated by the experiment,and the results show that each sensor index satisfies the actual demand.Secondly,a multi-sensor sensing system is designed based on NI Compactrio embedded platform,including hardware module configuration,software system architecture configuration and software design,which realizes multi-sensor synchronous acquisition,real-time data display and large amount of data storage.Then,for the application of local muscle fatigue assesment,based on the characteristics of non-stationary nonlinearity of sEMG signals,a method of muscle fatigue perception was proposed.Through the fatigue sEMG signals of 10 volunteers in the radial wrist extensor muscle,3 groups of experiments were designed to compare the performance of HHT marginal spectral entropy,approximate entropy and median frequency from fatigue index to data length dependence,evaluating muscle fatigue reliability and noise resistance.The experimental results show that the HHT marginal spectral entropy has many advantages,such as low dependence on data length,reliable evaluation of muscle fatigue and anti-noise,and provides a new method for sensory muscle fatigue.Finally,in order to verify the effectiveness of the perceptual system,the physiological movement signals of a volunteer in a treadmill rehabilitation robot impedance model were tested,and the muscle fatigue of the anterior tibial muscle was evaluated by HHT marginal spectral entropy.The results show that the sensing system based on Multi-sensor can reflect the information of patients ' rehabilitation training comprehensively,and can perceive exercise muscle fatigue through HHT marginal spectral entropy.