Genetic algorithm-based optimization of a vehicle suspension system

This paper describes a method of optimizing the motion characteristics of a double wishbone front suspension system by using a genetic algorithm. The analysis considered only the kinematics of the system. The mechanism is mad e of two spatial closed loops - the first loop (the main double wishbone me chanism) analysed and results fed into the second loop, describing the stee ring mechanism. The mechanism is such that it allows hierarchical analysis. The analysis is carried out in the time domain. As the two loops are linke d together, overall motion influenced by the kinematics of these loops and the vertical oscillation of the suspension system would cause an oscillatio n of wheel orientation in space. The objective function is designed to repr esent the degree of coupling between the motion of the wishbone and steerin g mechanisms. A genetic algorithm is used to vary 20 selected design parame ters of the suspension-steering mechanism in order to minimize the amplitud e of wheel steer oscillation due to vertical suspension oscillation. Althou gh the study is not carried out on any 'real' suspension system, the softwa re could be used as a general design tool and it allows alternative objecti ve functions without substantial change to the software.