Reliability Assessment of Islanded Microgrids Considering Generation-Load Uncertainty

Due to the volatility and uncertainty of renewable energy, as well as the constantly changing load demand, the reliability assessment of islanded microgrids faces significant challenges. To address this issue, this paper proposes a reliability assessment method for islanded microgrids considering source-load uncertainties. Firstly, an uncertainty model of the islanded microgrid system is established: the “failure-repair” time series model is used to simulate the life process of the micro gas turbine unit; the “normal-deficiency-shutdown” time series model combined with wind and photovoltaic prediction errors is used to simulate the uncertainty of wind and solar power; a load uncertainty model considering load prediction errors is established; a storage charge-discharge model considering energy storage life is established. Secondly, the CNN-BILSTM-MATT (Convolutional Neural Network - Bi-directional Long Short-Term Memory Multi-head Attention) hybrid model optimized by the RIME algorithm is used to predict wind, solar, and load; combined with the Adaptive Bandwidth Kernel Density Estimation (ABKDE) method, interval prediction of wind and solar power is realized. Then, within the wind and solar prediction interval, Monte Carlo sampling is used to obtain the actual power of the simulated wind and solar output; finally, combined with the predicted power load data, conventional generators, and energy storage status, the reliability index of the microgrid is calculated.