The Improvement Of A Multi-objective Genetic Algorithm And Its Application In Active Suspension H2/H∞ Control

In the process of vehicle active suspension design, problems such as actuator saturation,actuator time-delay and parameter uncertainties should be considered.In applications, this paper presents a fuzzy static output feedback controller design approach for vehicle electro-hydraulic active suspensions.Due to the highly nonlinear dynamics of electro-hydraulic actuators, using the electro-hydraulic actuators to track the desired forces is fundamentally limited in its ability when the environment is dynamical.Therefore,we directly take dynamics of electro-hydraulic actuators into account in process of quarter-car suspension T-S fuzzy modeling,so,both actuator time-delay and dynamics of electro-hydraulic actuator are considered,which are closer to the practical application.As for actuator saturation, we take it as one of the H∞performance ,Takagi-Sugeno(T-S) fuzzy modeling is applied for approximating the dynamic nonlinear active suspension model; and,it has three main performance requirements which should be considered in designing suspension, ride comfort, handing stability and suspension deflection. This paper takes the vertical acceleration of the car body as the H2 performance, suspension deflection and tyre deflection as the H∞performance; utilizes a new improved Elitist Non-dominated Sorting Genetic Algorithm for searching the feedback gain matrix, acquires the H2 norm and H∞norm via solving the Linear Matrix Inequalities (LMI).As we all know,the best solution of multi-objective genetic algorithm(GA) is acquiring evenly distributed and actual optimal Pareto set . Generally , the terminating condition of GA is setting the maximum evolution generation,But there are no effective standards and mandates for setting it;if it is too big,the calculated quantities will be increased ; and if too small , the best results can not be acquired.This paper presents a new improved elitist Non-dominated Sorting Genetic Algorithm (INSGA₂) based on Non-dominated Sorting Genetic Algorithm (NSGA₂),a new terminating condition is presented,the strategy of limited elitist and uniformity keeping is suggested, effective crossover operator and mutation operator are introduced.