A novel switched model predictive control of wind turbines using artificial neural network-Markov chains prediction with load mitigation

Abstract

The existing model predictive control algorithm based on continuous control using quadratic programming is currently one of the most used modern control strategies applied to wind turbines. However, heavy computational time involved and complexity in implementation are still obstructions in existing model predictive control algorithm. Owing to this, a new switched model predictive control technique is developed for the control of wind turbines with the ability to reduce complexity while maintaining better efficiency. The proposed technique combines model predictive control operating on finite control set and artificial intelligence with reinforcement techniques (Markov Chains, MC) to design a new effective control law which allows to achieve the control objectives in different wind speed zones with minimization of computational complexity. The proposed method is compared with the existing model predictive control algorithm, and it has been found that the proposed algorithm is better in terms of computational time, load mitigation, and dynamic response. The proposed research is a forward step towards refining modern control techniques to achieve optimization in nonlinear process control using novel hybrid structures based on conventional control laws and artificial intelligence.