Resilient Predictive Control of Connected Hybrid Vehicles Considering Denial-of-Service Attacks

The connected hybrid vehicles (CHVs) fleet, which consists of multiple inter-CHVs, serves as a significant driver of future intelligent transportation systems. Through vehicle-to-vehicle (V2V) communication, CHVs can greatly enhance driving safety and reduce fuel consumption. However, network exposure and frequent communication make CHVs highly vulnerable to information security attacks, particularly Denial of Service (DoS) attacks, which could potentially lead to communication interruptions and pose a threat to the safety of the fleet. Therefore, designing an advanced vehicular control strategy such that the control performance of CHVs is resilient to DoS attacks has become an urgent issue. To address this challenge, this article proposes an attack-resilient control strategy to ensure the security and cost-effectiveness of the CHVs in the event of a DoS attack. First, heterogeneous and uncertain vehicle dynamics models and DoS attack models are established. Building upon this foundation, a comprehensive resilient model predictive control (CRMPC) strategy is proposed, which ensures queue safety through defined safety functions and resilient MPC while enhancing economy by incorporating vehicle fuel consumption function. Then, a weight adjustment mechanism is employed to balance the relationship between security and cost-effectiveness. Furthermore, an adaptive equivalent consumption minimization strategy (A-ECMS) is adopted, with the equivalent factor being dynamically adjusted in real-time using a Proportional Integral algorithm. Finally, the results of two test scenarios demonstrate that the strategy improves fuel efficiency by 5.32% and 5.44%, respectively, while ensuring fleet safety.