Desenvolvimento de um método de otimização para sistemas de suspensão veicular

Abstract

This work proposes a multi-objective optimization method that uses genetic algorithms to find the best spring and damper characteristics for a vehicle suspension system represented by a simplified quarter car model. Based on spectral power density equations that allow the description of the surfaces that have a Gaussian distribution pattern of the irregularities, five types of road were used. The maximum velocities where the comfort or safety limits for the roads are reached were obtained by four different calculation methods, two numerical and two analytical. The results are compared, showing errors that remain very low over a wide range of speeds. The efficacy and validity of the analytical method chosen as the basis of the optimization can be demonstrated by means of the obtained results. The optimization algorithm with an analytical base of acceleration calculations felt by the occupants and the standard deviation of the tire load shows a very short processing time alternative when compared to the numerical basis algorithms. More complex optimization models, with many more variables and objectives than the one presented here, are common in vehicle suspension projects and can benefit from the advantages offered by an analytical-based multi-objective optimization approach, reaching reliable solutions with more speed.