On optimal design of passive suspension using GA-based heuristic method

Abstract

In the automobile industry, vehicle suspension design is considered to be an important problem. In this paper, a real coded genetic algorithm (GA) with variable rate of crossover and mutation is applied to find the parameters of the passive suspension system of the half-car model with two passengers, which satisfy the performance as per ISO 2631 standards.To find the suspension parameters, we have formulated a non-linear constrained optimisation problem which is solved using GA. The objective function considered is the maximum bouncing transmissibility of the sprung mass at the centre of mass of the vibrating vehicle during its uniform motion over the road. Here, deterministic and probabilistic types of road conditions are taken into consideration. For deterministic type of road, we have developed a periodical waveform of multiple road bumps of saw-tooth type using Fourier series. To generate random road, we have used Gaussian distribution function. Since most of the random variables are assumed to follow Gaussian distribution, we have assumed probabilistic road as Gaussian. For validation, the results obtained by GA are compared to the initial suspension parameters in time domain and it is found that GA results show better performance.