This paper presents the design of an active suspension controller for an automotive vehicle using a genetic algorithm as the optimization technique. A four-degree-of-freedom model is used to represent a vehicle with different front and rear axes characteristics. The suspension deflection, tire deflection, vertical and angular acceleration are the performance criteria optimized. Different filters are used to model the frequency sensitivity of these criteria and the weighting is based on a passive suspension reference system. Independent front and rear controller optimization is performed with a genetic algorithm. The controllers include a linear gain matrix and a single filter. Each controller is designed to work with a minimum number of sensors and a limited order filter. To adapt the passive suspension components to the active system, the stiffness and damping of the suspension are optimized with values limited to a realistic range. Results show the impact of the various filters used to specify the critical frequency range of the inputs and outputs. This is observable for ride and handling criteria that are known to be frequency dependant. There is 38% improvement in the global performance of the active system compared to the baseline passive system.
展开▼
