Design of Fault Tolerant Control on Wind Turbine Speed Control Based on Bias Fault Estimation Method with Optimization l_0 Norm Constraint

Abstract

The availability of onshore wind power plant systems (PLTB) reaches 98 percent, but the maintenance costs required are still very high for the wind turbine generator system. Meanwhile, the availability of offshore PLTB is decreased by 60 percent due to the main cause of damage to some components in wind turbine systems. This study proposes the use of fault estimation methods of wind turbine system components in fault tolerant control (FTC) strategy. The error estimation method is built using the l0 norm constraint optimization. The optimization formulation with l0 norm constraint is derived by applying the compressed sensing technique so that the estimation of the bias error can be used to estimate the error of several components by a single observer. This answers the observability issues encountered in single observer use cases. The proposed implementation of observers with the FTC results in better response characteristics when compared to systems without FTC. Response characteristics on actuator errors of 0.3 to 1.3 per unit, system with observers resulting in a maximum undershoot value of 0.4 to 1 percent while systems without observers resulting in a maximum undershoot value of 6.2 to 26.4 percent. The characteristics of the response with the observer on sensor errors resulting 0.3 to 1.3 per unit resulting in value of 1.6 to 4 percent, 0 percent and 63.9 to 70.7 seconds. System without observers, with sensor errors of 0.3 to 1.3 per unit resulting in maximum undershoot, steady state error and settling time of 6.2 to 26.2 percent, 6.2 to 26.4 percent, and 0 seconds.