Gap metric–based model bank construction for wind turbine predictive control. (26th April 2018)

• Record Type: Journal Article
• Authors: Li, Dewen; Chen, Zhe; Li, Ning
• Is Part Of: Optimal control applications and methods. Volume 39:Number 5(2018)
• Page Start: 1610
• Page End: 1626
• Publication Date: 2018-04-26
• Subjects: gap metric -- model bank construction -- multiple model predictive control -- wind turbine system; Control theory -- Periodicals; Mathematical optimization -- Periodicals; 629.8312
• DOI: 10.1002/oca.2429
• Languages: English
• ISSNs: 0143-2087

Abstract

Summary: This paper presents a novel model bank construction method for the multiple model predictive control of wind turbine system. The gap metric is used to measure the dynamic difference between the linearized models of the wind turbine system at different wind speed. Two algorithms are then proposed to divide the wind speed range in different operating regions. Meanwhile, a complete and nonredundant linear model bank is established to approximate the wind turbine system in the whole operating region. We take the robust model predictive control algorithm to design the local controller and utilize the wind speed as the switching criterion to combine the submodels. The simulation study on a 5‐MW wind turbine verifies the efficiency of the proposed method.