Decomposition based CNN-RVFLN for Load Prediction and Congestion Management in Power System

Load forecasting is commonly employed by electricity market participants to refine their trading strategies and by system operators to maintain a stable grid. Specifically, it helps system operators anticipate potential power imbalances and other critical grid conditions, allowing them to take necessary corrective measures. Forecasting critical grid conditions, such as congestion, is particularly crucial in this regard. This paper introduces a Variational Mode Decomposition based CNN-RVFLN model designed to predict hour-ahead time-series residual loads. Load forecasting is important for significant operation in power system, loads become very much uncertain because it is mostly affected by various factors. Disturbances in the load results in congestion in power system. The new age intermittent energy sources leads to various system imbalances and congestion. Therefore faster branch flow estimation is necessary for the security assessment of the power system as the transmission line is exposed to various contingencies. This paper proposes the hybrid model for load forecasting that will help in congestion management by forecasting any overloading condition in the power system.