Active Suspension Performance Improvements Using a Genetically Tuned Fuzzy Logic System

In this paper we present preliminary results of a program that explores the use of genetic algorithms (GA) in the development of a fuzzy logic control (FLC) scheme for an active vehicle suspension system. Simulations of the U. S. Army’s High Mobility Multi-Wheeled Vehicle (HMMWV), going over a variety of cross country terrains, were used to tune the final FLC rulebase. What is unique in this approach is that the final design is based on minimizing a nonlinear function of critical vehicle quantities (driver’s absorbed power and the HMMWV rms pitch rate). Results of simulations using the GA-Fuzzy design are compared to the performance of a passive HMMWV suspension and to five other active control strategies. Finally, shortcomings and planned enhancements to the design are presented.