Key points are not available for this paper at this time.
The algorithm of a complex methodology for assessing the technical condition of the cylinder piston group of a marine propulsion system is being investigated. Wear is a continuous process characteristic of all working mechanisms. Studies aimed at identifying factors contributing to the degradation of system elements of devices provide the basis for the development of preventive measures to reduce their effects. Knowledge of the technical condition of marine engine components is important for the development of measures that increase the reliability of equipment and reduce the risks of emergency situations. Some of the main approaches to modeling and evaluating the state of the cylinder-piston system of marine diesel engines are presented. To solve the problems of assessing the technical condition of the cylinder piston group during operation, classical methods of statistical data analysis are considered, methods that artificially increase the size of the data sample are proposed, machine learning methods are analyzed and the most effective for use are determined. An integrated approach is being created to study the operation process of a cylinder-piston group of diesel marine engines based on a combination of statistical methods, machine learning methods and probabilistic forecasting. A diagram of the properties of the studied parameters is illustrated for constructing a model for analyzing a cylinder-piston group system. Machine learning algorithms used to study systems are presented. The proposed technique allows, using the results of indirect measurements (data from lubrication analyses), to determine the technical condition of the engine system, in particular the cylinder piston group.
Mazur et al. (Tue,) studied this question.