Weight factor optimization of active power filter based on model predictive control for EAST harmonic current compensation

Jie Yang,Jing Lu,Yanan Wu,Jun Li,Bin Xu,Junmin Zhang,Liuwei Xu
DOI: https://doi.org/10.1016/j.fusengdes.2024.114263
IF: 1.905
2024-02-19
Fusion Engineering and Design
Abstract:With the EAST fusion device in experimental operation, its high-power device injects a large number of harmonic current into the power system. To effectively reduce the harmonic current generated during the operation of the EAST power system, this paper proposes an active power filter (APF), which uses model predictive control (MPC) with a genetic algorithm (GA) for finding the optimal weighting factors of the cost function. In the first step, the cost function comprising multi-step prediction and switching frequency control is established. The second phase uses the genetic algorithm strategy for the cost function's weighting factors. The optimized weight factors data are organized and analyzed, and the model predictive controller optimized by genetic algorithm is verified on the two-level active power filter system. The results demonstrate that the proposed optimistic algorithm can effectively tune the weight factors. The algorithm accomplishes the goal of optimizing the controller and enhances the APF's performance of harmonic current compensation.
nuclear science & technology
What problem does this paper attempt to address?