Study On Control System Of Tele-operated Humanoid Robot Based On Bio-electricity Recognition

Physiological computing (including cognition, emotion, expression, etc.) and Brain-Computer Interface is a new man-machine interactive mode on which rapid development and great progress in theoretical research and practical applications has been achieved over the past decade. In the meantime, humanoid robotics research has extended from a simple organization field to an integrated cross-disciplinary combining of context-aware, intelligent decision, intelligent control, human-computer interaction, computer networks, and affective computing. Physiological calculation is used in humanoid robot research in this article and a tele-operated humanoid robot control system is designed based on the identification of bio-electric.Firstly, the related concepts such as bio-electricity robot control system, Brain-Computer Interface, bio-electricity signal and related research status were introduced. The composition and application of the robot control system based on bio-electricity signal was described. The applications and relative key technologies were introduced then.Next, bio-electricity signal control-based humanoid robot system was set up in the thesis. The principle of the system architecture, data acquisition and process was elaborated then. On this basis, Offline training and online control experiments in which two operators use only bio-electricity signal to control the robot and move it to the specified destination were taken. Comparing the same task bio-electricity control and manual control executive ratio is about1.15. The effectiveness of the system and control accuracy was verified.Meanwhile, a new concept:the operation quality (quality of tele-operation, QoT) was proposed after a set of study on fatigue driving. Its composition and generation principle was explained afterwards. The EEG perception system monitor real-time to collect and identify physiological signal changes of the operator and analyze their expressions of the physiological state. The relational model between QoT and fatigue extent was established by the neural network. Expected output and actual output error is less than0.03which proved the validity of the model built.At last, tele-operated system was combined with physiological monitoring to adjust parameters of control system real-time to adapt to changes in the user’s physiological state, and thus to achieve optimal performance and stability of the whole tele-operated system. The composition of QoT-based tele-operated system was designed in the paper. Experiments on QoT-based system and non-QoT system were carried out to verify the reliability of the system. Finally, a summary of the work and instructions at the characteristics of the study were built and future work was prospected.