Taguchi-based Optimal Dry Electrode Design

Surface Electromyography (sEMG) electrodes, including both wet and dry classes, play a significant role in the EMG signals analysis and recording of muscle activation area. Users of prosthesis hands must occasionally manipulate and move their limbs for several hours during daily tasks. Hence, high-quality and durable muscle activation is crucial. Although the designs of electrodes with different probe shapes, sizes, metals, and roughness are presented in previous studies, finding the optimal electrode design is an open problem. Hence, this paper proposes a different application of the Taguchi method to specify a reduced set of examinations for determining the optimal electrode design. This study considers the length, width, and distance of probes as controllable parameters as well as subject dominant hand, skin humidity, and muscle fatigue as data recording noises. The design of the experiment process (DOE) selects reasonable orthogonal arrays for the inner and outer arrays of the DOE table. The probes of the electrodes are made of silver due to their high conductivity, less oxidation, and fewer skin allergies in the presented study. In this investigation, the primary focus lies in electrode optimization through the decrease in the error of grasp force prediction. The Support Vector Machine (SVM) method is used to evaluate the error of grasp force. For this purpose, the signal-to-noise ratio (SNR) assessment is performed, identifying each parameter’s optimum level and significance. Results indicate that the electrodes with probes of 1 cm in length, 1 cm in width, and 2 cm apart lead to better sEMG signals. These results offer a design for significantly improved muscle activation signal based on the FUMCARE sEMG setup to use in hand prosthesis applications.