Abstract—Reliable data-driven methods, utilizing validated predictive models of electricity consumption, offer substantial promise to effectively manage energy in densely populated electrical grids, especially in urban areas like Mosul City. This research presents a model for forecasting the hourly electrical load for Mosul City, taking into account meteorological variables such as temperature, humidity, wind speed, cloud cover, and the type of day (holiday or working day). It explores two distinct scenarios: the first one examines the influence of weather elements on predictions of electrical load, and the second one employs the Least Squares Support Machine (LSSVM) model to forecast electricity consumption in Mosul City using historical load data and meteorological information. Two optimization algorithms, the Particle Swarm Optimization algorithm (PSO) and the Whale Optimization Algorithm (WOA), are employed to improve model accuracy and adjust the parameters of the LSSVM. In addition, the performance of the models in this research is evaluated using the Mean Absolute Percentage Error (MAPE). The results demonstrate the superiority of the LSSVM+PSO model over the LSSVM+WOA model and the basic LSSVM model in terms of accuracy and error reduction, while according to execution time, the LSSVM+PSO model takes a little longer than the LSSVM+WOA model. Consequently, the LSSVM+PSO model is deemed suitable for forecasting hourly electricity consumption in the city of Mosul.