Optimal LQG Controller for Variable Speed Wind Turbine Based on Genetic Algorithms

Abstract

Linear Quadratic Gaussian (LQG) control methodology shows useful properties of good performance and robustness in controller design applied to wind turbine. Typically, in the design procedure LQG method is necessary to select weighting matrices in order to solve the Algebraic Riccati Equations and then get the matrices Kalman Filter gain and optimal state-feedback. In order to optimize a LQG control applied to Double-Fed Induction Generator in wind power system, a Genetic Algorithms (GA) adapted to get the best values of the element of weighting matrices is proposed in this paper. The performance indices ISE and ITSE are a good alternative to obtain the fitness function to design LQG controllers with GA. The simulation results show the high effectiveness of this optimal design method.

A novel fractional‐order nonlinear PID ‐based pitch angle control strategy for a PMSG ‐based wind energy system

“…To optimize the controller parameters, the cost function is considered a weighted sum of IAE and ISE. These measures are selected to meet the basic properties of the controller operation, which is minimization of instantaneous error between the desired and measured angular velocity of the rotor at the instance while wind gets fluctuated 38,39 . The applied cost function is given in Equation (17).…”

Automatic Generation Control using Genetic Algorithm Based PID Controller

“…The controller was used in the system for automatic generation control purpose. GA has been implemented by Rene Barrera-Cardenas [6] for optimization of Linear Quadratic Gaussian control used in double-fed induction generator in a wind power system. Thus GA serves the purpose of designing an intelligent and robust control technology, in spite of nonlinearity, load disturbances and system parameter variations.…”

A MIMO-Based Compatible Fuzzy Logic Controller for DFIG-Based Wind Turbine Generator

“…So, many revolutionary control techniques have been introduced to overcome the issues created by conventional PI controllers such as fuzzy logic controllers (FLC) for RSC [6][7][8][9], deadbeat fuzzy controllers for GSC [9], PI and fuzzy hybrid RSC controllers [10] and self-tuning neural fuzzy controllers [11]. There are many adaptive controllers such as LQR [12,13], LQG [14], particle swarm optimizing algorithm technique [15], etc. All existing control techniques proposed the separate control techniques at RSC and GSC.…”