An efficient Forecasting Tool using Machine Learning in Optimal Power Flow: A Case Study of Iraqi Weather

In the electricity sector, the hasty precise decisions must be taken by grids operators to guarantee a sufficient level of system`s resilience during various contingencies. In practice, the optimal power flow (OPF) is a substantial monitoring and assessment tool to meet invulnerable operation measures. This article examines an optimal power flow for the day ahead, considering the intermittent nature of renewable energy sources (RES) influenced by weather conditions. It incorporates machine learning into power system operations to accurately forecast the meteorological data (such as temperature, and solar irradiance) that directly impact the power output from solar photovoltaic generators. This enables power generation schedulers to make informed decisions for the next 24 hours. The research engages the conventional IEEE-30-bus system as a test system, applied to Iraqi capital as testing location. The Matlab-designed algorithm is employed to accomplish the day-ahead optimal power flow using whale optimization. The results demonstrate a close match between the actual and predicted meteorological data, validating the effectiveness of the forecast in optimizing the power flow. Economically, the results of Baghdad show that the predicted contribution of solar energy reduces the fuel cost to 568.79 USD/h, compared to 801.84 USD/h when renewables are not utilized. Environmentally, CO2 emissions are reduced to 269.24 kg/h from 424.81 kg/h in the conventional system. Furthermore; to evaluate the achievability of the whale optimization, the optimal power flow for the conventional system is compared with two other metaheuristic optimization techniques, providing statistical metrics for a comparative analysis.