The Study on Mesh Optimization and Computational Efficiency Balance in Magnetic Network Modeling of Magnetic-Saturated Controlled Reactor

Abstract

In response to the dynamic reactive power demand surge caused by new energy grid integration, Magnetic-Saturated Controlled Reactor (MSCR) has become an ideal compensation device by virtue of its fast response and continuous regulation capability, but its modelling complexity and computational efficiency need to be solved urgently. This paper takes MSCR as the research object and focuses on the optimisation strategy of mesh dissections for magnetic network models, seeking a balance between computational accuracy and solution efficiency, and verifying the validity of the model through an experimental platform. Firstly, the magnetic leakage effect is considered and the magnetic flux tube equivalence method is adopted to establish the MSCR magnetic network model, and the validity of the model is verified by the measured data. Through the differential grid dissection of the magnetic valve region, it is revealed that the computational efficiency of the magnetic network model is about 100 times higher than that of the finite element method, and the error converges with the grid encryption; under the full-load condition, the local grid encryption of the magnetic valve section can significantly reduce the error, while the sparse mesh error is already very small in the case of no-load, and the optimization method based on the load-grid correlation is further proposed to realize the optimization of no-load and full-load regions by dynamically adjusting the mesh density. By dynamically adjusting the mesh density to achieve sparsification in the no-load region and local encryption in the full-load region, the relative variation of the calculation results under full working conditions is controlled within 1%, which effectively reduces the impact of load fluctuations on the calculation accuracy. This method realizes the dynamic balance between accuracy and efficiency in the modeling of complex electromagnetic devices, and provides theoretical support for the optimization of MSCR function and the engineering promotion of magnetic network model. The research results not only provide theoretical support for the engineering design of MSCR, but also provide important technical support for the system-level simulation of other electromagnetic devices, as well as the real-time simulation and safe and stable operation of smart grid.