Centralized ramp confluence cooperative control method with special connected and automated vehicle priority

To ensure that the special vehicles pass quickly and smoothly in the confluence area of typical Y-type ramps in unmanned driving environment, a cooperative control method was studied by considering the priority to special vehicles for the centralized control scenario of fully connected and automated vehicle (CAV). The confluence sequence arrangement in the controlled area was determined using games. By considering the task priority attribute of special CAV and characteristics of vehicles, attributes related to the acceleration-associated lane priority of special CAV, the time-associated vehicle type priority, and the acceleration change rate-associated vehicle stability priority were designed separately, and a joint characterization of these attributes was performed using the cost function. The confluence sequence arrangement with the special CAV was transformed into an optimal sequence set solution, and the optimal confluence sequence was determined using the income matrix method of two-person cooperative game. Based on the sequencing results, the Pontryagin's maximum principle was applied to determine the vehicle trajectory control, and the optimal analytical solution for the longitudinal trajectory was obtained when the cost was at the minimum, to achieve the cooperative control while prioritizing the special CAV. Through the case calculation, Python was used to simulate and verify the cooperative control method giving priority to the special CAV, and the respective fuel consumptions and transit times with no control strategy and using the first-in-first-out strategy were compared. Research results show that 86% of vehicles can travel at the maximum speed to smoothly pass through the confluence area, and the cooperative control method can effectively guarantee the priority of special CAV. Compared with cases without control strategy and under the first-in-first-out strategy, the cumulative fuel consumptions under the cooperative control decrease by 11.8% and 16.1%, respectively, and the overall fleet passing time is shorter than those obtained with the two traditional confluence strategies. There are corresponding thresholds for the maximum speed limit, initial speed, and controlled area length to enable the special CAV to pass quickly. These values can serve as references in the confluence area design. 3 tabs, 7 figs, 31 refs.