An interval-based multi-objective robust design optimization for vehicle dynamics

Abstract

This study presents an Interval-based Multi-objective Robust Design Optimization (IB-MORDO) algorithm applied to a vehicle dynamic problem. The proposed algorithm optimizes a full 15 degrees-of-freedom (15-DOF) vehicle model, subjected to a double-lane change (DLC) maneuver under random road profiles, to attain driver comfort and safety. This study does not make assumptions about uncertain parameter statistics; instead, the uncertainties are quantified using a non-probabilistic α-cut level interval analysis. These uncertainties are applied to the system parameters and design variables to ensure robust results. After the optimization process, the root mean square (RMS) vertical acceleration at the driver’s seat resulted in a robust solution of 1.041 m/s² and a parameter interval radius (IR) equals to 0.631 m/s², whereas the RMS lateral acceleration at the driver’s seat resulted in a solution of 1.908 m/s² with an interval radius of 0.168 m/s². Unlike the Robust Optimization, the algorithm proposed herein considers uncertainties at system parameters and design variables without assuming any statistical data.

HIGHLIGHTS

• An Interval-based Robust Multi-objective Optimization procedure is proposed and tested on a 15-DOF vehicle model.

• Αn α-cut level methodology is used to deal with the uncertainty propagation.

• Resulted optimal suspension parameters minimize center and interval radius of driver’s vertical and lateral accelerations.

KEYWORDS:

• Multi-objective robust optimization
• uncertainty quantification
• α-cut level method
• multibody system dynamics
• Pareto frontier