Multi-objective Predictive Control for Intelligent Vehicles by Considering Stability Constraints in Complex Scenarios

Intelligent vehicles are currently facing the challenge of improving driving stability and active safety while dealing with multiple objectives in complex traffic scenarios with multiple vehicles and various road surfaces. To overcome this, a new multi-objective control structure is proposed, which includes an integrated predictive model guided by a switching mechanism, and corresponding objective functions and constraints under different modes. The structure also coordinates multiple actuator inputs to integrate path planning and path following. The predictive model takes into account vehicle models, actuator dynamics, and the combined-effect tire model. A newly-developed quantitative risk calculation method is used to design the switching mechanism that switches among different driving modes. This helps the vehicle to improve active crash avoidance with approaching vehicles from various directions while satisfying driving stability constraints. Lastly, a driver-in-the-loop platform has been developed to validate the real-time performance and effectiveness of the proposed method under complex scenarios.