A Comparative Analysis of Hyperparameter Tuned Stochastic Short Term Load Forecasting for Power System Operator

Intermittency in the grid creates operational issues for power system operators (PSO). One such intermittent parameter is load. Accurate prediction of load key to proper planning system. This paper uses regression analyses short-term forecasting (STLF). Assumed data are first analyzed and outliers identified treated. The cleaned fed methods involving Linear Regression, Decision Trees (DT), Support Vector Machine (SVM), Ensemble, Gaussian Process Regression (GPR), Neural Networks. best method based on statistical using parameters as Root Mean Square Error (RMSE), Absolute (MAE), (MSE), R2, Prediction Speed. further optimized with objective reducing MSE by tuning hyperparameters Bayesian Optimization, Grid Search, Random Search. algorithms implemented Python Matlab Platforms. It observed that obtained analysis hyperparameter an assumed set respectively. also that, due tuning, reduced 12.98%.