Comparative Analysis of Renewable Energy Production Forecasting Algorithms Using Python

In this paper, forecasting algorithms were tested on renewable energy production data and their comparative analysis was carried outThe machine learning methods that were used in the work are DecisionTree, AdaBoost, RandomForest and LSTM networks.From the analysis of the results of the study, it can be seen that the accuracy of the forecast of the LSTM neural network is higher than all other methods considered, but at the cost of this was the speed of creating a forecast model.Of the considered methods, according to the combination of time and quality criteria, LSTM neural networks turned out to be the best for data analysis, showing the best accuracy. But for SMALL HYDRO and SOLAR data, RandomForest is the preferred method. It made it possible to build sufficiently accurate models in much less time than neural networks.