Optimization of Bagging Classifier for Diagnosis of Marine Electric Drives

Authors:

Abstract
The integration of Artificial Intelligence and its use to improve various aspects are one of the basic goals of modern industry. Such an industry is shipping. This article will examine the use of a machine learning classifier, namely Bagging classifier, to reduce, predict and diagnose defects in ship electric drives. For this purpose, a mathematical model of a ship power plant with limited energy production and a flexible model of an electric motor has been built, in which various defects can be simulated. The results of these simulations are used to train AI. An algorithm has been developed to optimize the performance of various popular classifiers, including the Bagging Classifier, and the achieved optimization results are analyzed from the point of view of accuracy and training time.