کد مقاله کد نشریه سال انتشار مقاله انگلیسی نسخه تمام متن
10226622 1701282 2019 11 صفحه PDF دانلود رایگان
عنوان انگلیسی مقاله ISI
Pattern recognition of SEMG based on wavelet packet transform and improved SVM
موضوعات مرتبط
مهندسی و علوم پایه سایر رشته های مهندسی مهندسی (عمومی)
پیش نمایش صفحه اول مقاله
Pattern recognition of SEMG based on wavelet packet transform and improved SVM
چکیده انگلیسی
The purpose of this paper is to solve the problem of low recognition accuracy of three-degree-of-freedom myoelectric prosthesis and long training time.According to the nonstationarity of the EMG signal, the wavelet packet is used to decompose the EMG signal and the energy and variance of the wavelet packet coefficients of the four-channel EMG signal are extracted as feature vectors.Then Particle Swarm Optimization(PSO) was combined with improved support vector machine(ISVM) to construct a new model(PSO-ISVM). Under the premise of ensuring the sparseness of the SVM, the slack variables and the decision function was improved to reduce the constraint conditions for solving the optimal face in the quadratic programming. SVM is optimized by the PSO in order to improve the accuracy of the model.The experimental results show that the improved algorithm can effectively identify six kinds of commonly used upper limb movements compared with the traditional SVM. The average recognition rate reaches 90.66% and training time can be shortened 0.042 s.
ناشر
Database: Elsevier - ScienceDirect (ساینس دایرکت)
Journal: Optik - Volume 176, January 2019, Pages 228-235
نویسندگان
, , ,