The architecture and key technologies of a BMI based telerobot system over Internet

摘要

In the study of Internet-based telerobot system, the aim is at human-machine integration and a BMI-based telerobot system over Internet is proposed using a new method of direct human-machine integration interface: brain-machine interface (BMI). The difference between this system and the traditional Internet-based telerobot system lies in that operators can directly control remote robots just by their thoughts. In traditional Internet-based telerobot system, operators control robots by joysticks which cannot directly integrate their intelligent system into robot system and also cannot meet the special application requirements of hand unfit teleoperation. To realize such a system, the key problems such as the optimal design of cognitive tasks fitting both operators and operating robots, the adaptive real-time online parameters optimization algorithm for feature extraction and classification, as well as the methods for overcoming the adverse impacts brought by network time-delay on the stability, telepresence, and transparency of the system are discussed in the study, which provides both methods and algorithms for further study.