Optimization of random forest for active power prediction based on three-phase voltage and current parameters

This study aims to apply random forest optimization in predicting total active power in a three-phase power system based on voltage and current parameters for each phase. The data used consists of measurement data collected using power quality meters, with a total of 2,343. The target output of the research focuses on the total active power value with a variable range of 0 - 4,915.05 kW, with a data division ratio of 80% for training and 20% for testing. The model scenario was configured using the Randomized Search CV optimization method, which produced regression evaluation metrics with an MAE of 5,911.61 watts, an RMSE of 74,308.10, and a determination coefficient R² of 0.8927. The model visualization is displayed in a scatter diagram, error histogram, and comparison graph of actual and predicted active power, showing that the model provides a stable error distribution and has a good level of capability. The results of the study indicate that the random forest model with hyperparameter optimization can be used to model the non-linear patterns and characteristics of data between voltage and total active power current parameters.