Integrated Path Planning-Control Design for Autonomous Vehicles in Intelligent Transportation Systems: A Neural-Activation Approach
Xingyu Li, Xinle Gong, Ye‐Hwa Chen, Jin Huang, Zhihua Zhong
Abstract
Path tracking for autonomous vehicles is one of the most critical tasks in intelligent transportation systems (ITS). The ITS performance, including efficiency, safety, flexibility, and resilience, are all based on it. The two central issues for a successful path tracking are resilience and smoothness. We endeavor to adopt a neural-activation based constraint-following approach to resolve these two issues concurrently. First, an adaptive robust constraint-following control scheme is proposed. The control tracks a desired trajectory with guaranteed performance even in the presence of uncertainty. Second, a neural-activation mechanism is proposed, which generates desired trajectory effectively based on traffic pattern with sufficiently smoothness. Third, the trajectory is embedded into the control scheme to ensure that the control conforms to any changing traffic pattern while in motion. As a result, the control can rapidly adapt to the changing traffic condition with smoothness and resilience.