Research Of A Soft Robotic Hand Controlled By Surface Electromyography

Compared to traditional rigid robots,soft robot has many advantages.The soft robot can change its shape arbitrarily to adapt to the complex environment of the outside world.The stability and safety of the soft robot driving are excellent.EMG as an important means of human-computer interaction,has broad application prospects and research significance.Based on the driving principle of soft robotic hand and the control method of EMG signal,this paper proposes a soft robotic hand that can be controlled by surface EMG signal.This article mainly did the following work:First of all,the related Chinese and foreign literature were consulted extensively,the status quo of soft robotic hand and EMG at home and abroad were investigated in depth,and the research background and significance of this topic were described.The technical route of this article and the main research content were described.Then,the three-dimensional model of the soft robotic hand is designed.The deformation characteristics of the soft robotic hand are analyzed based on the geometric relationship and the Kabelon equation.The relationship between the input air pressure and geometric parameters and the bending deformation angle is established.Based on the finite element analysis of Abaqus software platform,the effects of the number of cavities,the cross-sectional shape of cavities,the pressure in cavities and the pressure of single finger on the single-handed flexural behavior of soft robotic hand were obtained.Based on the results of structural design and finite element analysis,a soft robotic hand model was created.Then,based on the characteristics of surface EMG(sEMG)of the muscle groups such as the extensor muscles of the forearm based on the gestures of human hand gestures,a total of 24 eigenvalues including 21 time-domain features and 3 frequency domain features of surface EMG signals were extracted by signal pretreatment and active segment detection.After pattern recognition,a linear discriminant analysis classifier was used to establish a gesture recognition algorithm based on surface EMG signals.Finally,a soft robotic hand control platform is set up based on the functions of soft robotic hand and gesture reappearance,including hardware control platform and software control platform.Through the control platform,10 kinds of gesture recognition and control modes are established,and the control of the surface EMG signal to the soft robotic hand is realized.The software robotic hand can realize the task of capturing and reproducing human gestures.The offline recognition accuracy rate is 96.1%,Online recognition rate was 84.6%.The conclusions drawn can lay a theoretical and practical foundation for the further research and application of soft robotic hand.