A Methodology for Multi-Objective Design Optimization (MDO) of Automotive Suspension Systems

Original Equipment Manufacturers (OEMs) should innovate ways to delight customers by creating affordable products with improved drive experience and occupant comfort. Vehicle refinement is an important initiative that is often take-up by the project teams to ensure that the product meets the customer’s expectations. A few important aspects of vehicle refinement include improving the Noise Vibration Harshness (NVH), ride and handling performance pertaining to the Functional Image (FI) of the product. Optimizing the suspension design variables to meet both ride and handling performance is often challenging as improving the ride will have a deteriorating effect on handling and vice-versa. The present work involves Multi-Objective Design Optimization (MDO) of the suspension system of an automotive Sports Utility Vehicle (SUV) platform considering both ride and handling requirements, simultaneously. Using advanced simulation tools (ADAMS) and a Design of Experiments (DoE) based approach, functional forms were derived for the objective function and constraint variables. The mathematical forms of the predictive models were obtained from various machine learning based algorithms. Trade studies were done to arrive at optimal design decisions. In the available literature, there are very few machine learning based optimization studies done considering both ride and handling attributes, simultaneously, covering a wide range of Design Verification Plans (DVPs).