Optimization on handling stability using AdaBoost-CART with an improved evolution algorithm under uncertainty

Multi-objective optimization of handling stability for a tractor semitrailer during the designation process is a significant challenge in ensuring driving safety. In this paper, a data-driven modeling method that utilizes an uncertainty classification strategy to optimize handling stability is proposed. By considering the effect of yaw motion on the tractor semitrailer, a mathematical model was developed to identify the structure parameters that influence handling stability performance. The proposed method employs the AdaBoost-CART algorithm to simplify the modeling process by identifying key features in a dataset related to handling stability. The classification results were related to a population evolutionary strategy and a new recombination strategy improved the interaction of prediction results. The main objective of this optimization problem is to accurately predict the output results through the classification model and utilize this information to efficiently identify optimal parameters within the solution space under uncertain conditions. To validate the effectiveness of the proposed method, the optimized results were tested using both TruckSim software simulations and physical tests. The results show that the proposed method significantly improves the overall performance of handling stability in various driving scenarios.