Power Predicting for Power Take-Off Shaft of a Disc Maize Silage Harvester Using Machine Learning

The relationship between the power consumed in the engine and the power take-off (P.T.O.) shaft of a maize silage harvester is critical to understanding the efficiency and performance of the harvester. The power consumed in the engine directly affects the power available for use on the P.T.O. shaft, which is the power source for the suspended silage harvesters. The research aimed to predict the power consumption of the P.T.O. shaft based on the power consumption of the tractor engine at different operating parameters, which are two applications of the P.T.O. shaft (540 and 540E rpm) and two forward speeds (1.8 and 2.5 km/h) using machine learning algorithms. The best results in terms of engine power consumption were achieved in the 540E P.T.O. application, and the forward speed was 1.8 km/h. The results also gave a correlation between the power consumed by the engine and the P.T.O shaft of 87%. Regarding prediction algorithms, the Tree algorithm gave the highest prediction accuracy of 98.8%, while the KNN, SVM, and ANN algorithms gave an accuracy of 98.1, 60, and 60%, respectively.