KUKA Real-Time Control through Angle Estimation of Wrist from sEMG with Support Vector Regression

摘要

In this paper, a wrist joint angle estimation model based on support vector regression(SVR) is established, which is optimized by combining the cuckoo algorithm with the steepest descent method. The sEMG signals were sampled from two forearm muscles. In order to shorten the model train time and improve the accuracy of the estimation model, a new technique combining cuckoo algorithm with the steepest descent optimization method is proposed to find the bestc and bestg of SVR in a short time. Experiments on a subject showed, the proposed method shows a good performance in both accuracy and timeliness. Then, in order to verify the estimation effectiveness and practicality of sEMG, the estimation angle was used to control a KUKA robot in real time, and obtaining a good experimental result. The method discussed in this paper can provide a valuable reference for the sEMG based control for rehabilitation robot systems.