Design of an Adaptive Neural Network Controller for a Bidirectional Interlinking Converter in a Hybrid Ac/Dc Microgrid

• Published in: Conference record of the Industry Applications Conference pp. 1 - 8
• Main Authors: Wang, Li, Gao, Hou-Yu, Tzeng, Ching-Wen, Liao, Zi-Hao, Tseng, Ching-Chung, Mokhlis, Hazlie, Chua, Kein Huat, Tripathy, Manoj
• Format: Conference Proceeding
• Language: English
• Published: IEEE 20.10.2024
• Subjects: bidirectional interlinking converter; Control systems; Elman neural network controller; energy-storage system; Frequency control; hybrid AC/DC microgrid; Hybrid power systems; Microgrids; Neural networks; Power system dynamics; solar farm; Solar farms; Switches; Voltage control; wind farm; Wind farms
• ISSN: 2576-702X
• DOI: 10.1109/IAS55788.2024.11023763

This paper proposes an adaptive neural network controller for the bidirectional interlinking converter (BIC) connected between the DC microgrid (MG) and the AC MG of a hybrid AC/DC MG containing two wind farms, a solar farm, two energy storage systems (ESSs), a DC load, and an AC load. The Elman neural network (ENN) controller of the BIC is designed to effectively control the DC-link voltage of the DC MG as well as both voltage magnitude and frequency of the AC MG for the studied AC/DC MG switching between grid-connected mode (GCM) and islanding mode (ISM). The dynamic performance of the BIC using the designed ENN controller is also evaluated and compared with the one using a traditional proportional-integral (PI) controller when the studied hybrid AC/DC MG is switching from GCM to ISM. For the steady-state analysis, the BIC can maintain stable operation under the selected operating conditions. For the dynamic analysis, compared to the PI controller, the ENN controller can modulate the BIC faster to stabilize the active power, reactive power, voltage magnitude, and frequency of the studied system subject to switching conditions and reduce the amplitudes of system oscillations with shorter settling times.