Controlling Megawatt Class WECS by ANFIS Network Trained with Modified Genetic Algorithm

TLDR: The generated mechanical power from Wind Energy Conversion System (WECS) is highly susceptible on wind energy absorbed by turbine blades whereas they can cause fluctuations in generated power which is addressed by applying a novel strategy for training ANFIS network as a controller to the WECS.

Abstract: The generated mechanical power from Wind Energy Conversion System (WECS) is highly susceptible on wind energy absorbed by turbine blades whereas they can cause fluctuations in generated power. Designing a controller for WECS that leads to smooth generated mechanical power as well as high efficiency even with the presence of low wind speeds can be a challenging problem. This paper addresses this challenge by applying a novel strategy for training ANFIS network as a controller to the WECS. We firstly introducing Genetic Algorithm and Modified Genetic Algorithm optimization methods for training the adaptive network and updating the parameters of ANFIS. Then we use this trained adaptive network as a pitch angle controller for wind turbine. Comparing the results of proposed method with standard parameter optimization methods of ANFIS shows less error and training time in desired results. Also the fluctuations of wind turbine mechanical power decrease by applying trained ANFIS as a control signal for pitch angle entrance. The simulation results by using actual detailed model for wind power system show the effectiveness of the proposed method. Keywords-component; wind energy conversion system (WECS), Wind turbine, Pitch angle, Adaptive-Network-Based Fuzzy Inference System (ANFIS) controller, Optimization, Genetic Algorithm, Modified Genetic algorithm