Optimization of Controllers for Gas Turbine Based on Probabilistic Robustness

An optimization method for controller parameters of a gas turbine based on probabilistic robustness was described in this paper. As is well known, gas turbines, like many other plants, are stochastic. The parameters of a plant model are often of some uncertainties because of errors in measurements, manufacturing tolerances and so on. According to model uncertainties, the probability of satisfaction for dynamic performance requirements was computed as the objective function of a genetic algorithm, which was used to optimize the parameters of controllers. A Monte Carlo experiment was applied to test the control system robustness. The advantage of the method is that the entire uncertainty parameter space can be considered for the controller design; the systems could satisfy the design requirements in maximal probability. Simulation results showed the effectiveness of the presented method in improving the robustness of the control systems for gas turbines.