Automatic Generation Control of Isolated Two-area Microgrid with Multiple Resources Based on Learning Optimization

This paper studies the automatic generation control (AGC) problem of isolated two-area multiple source microgrid energies under the influence of stochasticity and volatility of renewable energies and loads. A Load Frequency Control (LFC) model of isolated two-area microgrid with multiple distributed energy resources and energy storage device is established. The wind power, the photovoltaic output and load fluctuation are formulated as discrete-time Markov processes based on their stochastic dynamic characteristics. The performance of the proposed learning optimization algorithm combined with the AGC principle of the power grid is demonstrated on microgrid system. At last, comparing SAQ-learning with traditional Q-learning and PI controller, simulation results illustrated the rationality of the proposed microgrid model and the expected performance of the controller.