University of Mohaghegh Ardabili Journal of Operation and Automation in Power Engineering 2322-4576 2025 10 24 Deep Learning- Model Predictive Control for Load Frequency Control of Microgrids with Electric Vehicles 4175 10.22098/joape.2025.16193.2250 EN Farhad Amiri Department of Electrical Engineering, Tafresh University, Tafresh 39518-79611, Iran. Sajad Sadr Department of Electrical Engineering, Tafresh University, Tafresh 39518-79611, Iran. Journal Article 2024 11 14 In an islanded microgrid, ensuring frequency stability is essential for reliable system operation. Distributed generation (DG) and electric vehicles (EVs) make frequency stability challenging in an islanded microgrid because they increase generation and load variability and reduce system inertia. Load frequency control (LFC) is mainly used to enhance the frequency response of these types of microgrids. In addition, uncertainty in parameters and perturbations strongly impact the application of LFC. To address these challenges, this paper presents an LFC method for islanded microgrids using model predictive control (MPC) based on deep learning. The deep learning technique is used to enhance MPC controller performance against uncertainties and disturbances. The proposed method is validated through experiments, especially in the presence of disturbances and parameter instability. It is then compared with other methods, including linear active disturbance rejection control (LADRC), fractional-order PID (FOPID), and several others. The results show that the MPC method based on deep learning outperforms these approaches in terms of disturbance rejection, frequency response improvement, and system inertia enhancement. https://joape.uma.ac.ir/article_4175_aa84f32129913c5763cdb34609cf355c.pdf