An advanced driver-assistance system is developed, in which the conventional steering wheel and brake pedal are replaced by a novel human-machine interface using surface electromyography (sEMG) electrodes attached to the driver’s muscles, and this paper is dedicated to braking control assistance. In the first part, how the sEMG signals are measured is presented, then the signals are analyzed in the frequency and time domain to determine the noise they contain, and signal processing using optimal linear Wiener filtering (finite impulse response) is proposed and compared with two other methods. In the second part, using deep learning for detecting whether the driver is braking the vehicle is discussed, where the authors present the steps of preparing the data sets, extracting time-series features, and training a long short-term memory network. Lastly, the potential of applying the proposed deep learning method online, i.e., while the vehicle is in use, is analyzed.
Surface Electromyography-Controlled Vehicle Braking Assistance System Using Deep Learning
Lect. Notes in Networks, Syst.
International Conference on Applied Human Factors and Ergonomics ; 2021 July 25, 2021 - July 29, 2021
2021-06-27
9 pages
Article/Chapter (Book)
Electronic Resource
English
Advanced driver-assistance systems (ADAS) , Human-machine interface (HMI) , Surface electromyography (sEMG) , Wiener filtering , Long short-term memory (LSTM) Engineering , Transportation Technology and Traffic Engineering , User Interfaces and Human Computer Interaction , Robotics and Automation , Industrial and Organizational Psychology
European Patent Office | 2020
|European Patent Office | 2020
|European Patent Office | 2020
|VEHICLE BRAKING ASSISTANCE DEVICE AND BRAKING ASSISTANCE METHOD
European Patent Office | 2018
|