Özet:

nd. Time evaluation features of characteristics (TDF-characteristics) of a surface electromyogram when performing the palm functional movements and the possibility of their implementation in the bionic prosthetic limb intuitive control system. Objective. The aim of the paper is to develop the analytical model for evaluating the TDF-characteristics of myographic signals for basic functional movements of the patient’s wrist and fingers, as well as studying the possibilities of its implementation as a basis for non-parametric method of classification. Methods. A one-channel microcontroller based information measuring system was created for the registration of a surface electromyogram. The analytical model for evaluating its TDF-characteristics based on the use of variance and trapezoidal integral features was developed. Results. The developed model was tested using the method of k-NN classification of the measured signal in the analysis of normalized and non-normalized data of recognition of the fingers’ functional movements. The accuracy of movement classification was 86.11 % and is acceptable for use in the development of methods for automatic control in the control systems of bionic prosthetics of the upper limbs or fingers. Conclusions. The analytical model for evaluating the TDF-characteristics of myographic signals for the wrist basic functional movements based on one-channel measuring system and a simple method of machine learning technique were proposed in the study. The proposed model is effective when working with a small set of signal characteristics and a limited amount of input data, and its classification accuracy can be increased by using a wider sample of data for training.

Anahtar Kelimeler: