Optimal partial filters of EEG signals for shared control of vehicle

Abstract

The development of equipment that measures EEG signals leads to the research that applies them to many domains. There are active research going on EEG signals for shared vehicle control system between human and car. An appropriate filtering method is also important because EEG signals normally have lots of noises. To reduce such noises, full matrix filter, sparse matrix reference filter, and common average reference (CAR) filter are presented and analyzed in this paper. In order to develop shared vehicle control system, we use controller, brain-computer interface (BCI), EEG signals, and car simulator program. By executing t-test, it was possible to find the optimal filter out of three filters mentioned above. With the analysis of t-test, it has revealed that full matrix filter is not appropriate for shared vehicle control system. In addition, it proves CAR filter has the best performance among these filters.