Deep Learning–Based Prediction of Suspension Dynamics Performance in Multi-Axle Vehicles

This article is mainly to present a deep learning–based framework for predicting the dynamic performance of suspension systems for multi-axle vehicles, which emphasizes the integration of machine learning with traditional vehicle dynamics modeling. A multitask deep belief network deep neural network (MTL-DBN-DNN) was developed to capture the relationships between key vehicle parameters and suspension performance. Numerical simulation–generated data were utilized to train the model. This model also showed better prediction accuracy and computational speed compared to traditional deep neural network (DNN) models. Full sensitivity analysis has been performed in order to understand how different vehicle and suspension parameters may affect suspension dynamic performance. Furthermore, we introduce the suspension dynamic performance index (SDPI) in order to measure and quantify overall suspension performance and the effectiveness of multiple parameters. The findings highlight the effectiveness of multitask learning in improving predictive models for complex vehicle systems. By using SDPI and MTL-DBN-DNN model for optimization, the optimized parameters resulted in a significant improvement in suspension dynamic performance.