Modeling Car-Following Behavior of Brain-Control Driving with Queuing Network Architecture

Directly using brain signals to drive vehicles is called brain-control driving, which has the potential to help people with disabilities to acquire driving ability. Developing driver models of the brain-control driving can help understand and simulate the driver behaviors, which contributes to the development of the brain-control driving. In this paper, to simulate the car-following behavior of the brain-control driving, we propose two brain-control car-following models, i.e., CF-BCB and CF-BCI models. The two models are built by integrating a car-following decision model with a brain-control behavior model and a brain-computer interface (BCI) performance model in the queuing network (QN) cognitive architecture, respectively. The manual- and brain-control experiments are conducted to collect the car-following data of the ideal and real brain-control driving, respectively. Simulations with the proposed models are performed to mimic the ideal and real brain-control driving of different subjects. We compare the performance of the car-following decision model and two decision models extended from the intelligent driver model (IDM) and the full velocity difference model (FVDM), respectively. The comparison results show that the car-following decision model performs better than the two decision models in simulating the ideal brain-control driving. Besides, we prove the effectiveness of the proposed models in simulating the car-following behaviors of the brain-control drivers by comparing the simulation and experimental results.