Model Predictive Control For Brain-controlled Vehicles

A brain-controlled vehicle is directly controlled by the command acquired through brain-computer interface interpreting electroencephalograph(EEG) signals of human. The research about the brain-controlled vehicle can not only promote the development of brain-machine integration and rehabilitation devices based on biological and mechatronics integration, but it also can help the users with severe disabilities increase their mobility and improve their living quality. The research results also could be widely applied in civil and military areas. Thus, the study regarding brain-controlled vehicles has important scientific significance and practical values.But the current BCIs still have some problems, such as the low accuracy of command recognition, time delay in transmission and limited number of control command. These issues lead to a bad control performance for human when they use BCI to control a vehicle.In order to improve the control performance and safety of brain-control vehicle, this paper proposed a shared control method between the brain-controlled driver and a safety assistive controller designed by Model Predictive Control method(MPC). The work mainly includes:Modeling of brain control vehicle systems, designing of a safety assistive controller based on MPC, simulation analysis and experiment verification for designing controller.The dissertation achieved the following research accomplishments:1) Established the brain-controlled vehicle system model based on Steady-State Visual Evoked Potentials(SSVEPs) BCI. It includes QN-based decision-making brain-controlled driver model, BCI performance model and vehicle model. The system model was validated by experiment. The system model contribute to our understanding of brain-controlled vehicle system, and also provide a simulation platform for controller designing.2) Proposed a new predictive control method based on human-centered control and designed an assistive controller through MPC method. The controller improved the driving performance and safety of the brain-controlled vehicle.3) Analyzed the affection of accuracy of BCI to the performance of brain-controlledvehicle, and obtained the optimal relationship between the accuracy of BCI and parameterof assistive controller.These innovative accomplishments can not only promote the development of key technology of brain-controlled vehicle, but also benefit the research for human-centered control.