The article discusses the features of the choice of an algorithm for creating an intelligent control system for an SEP in the operating conditions of autonomous marine vessels. On the basis of a comparative analysis of systems and methods of artificial intelligence, the expediency of using neural networks has been substantiated. The use of neural network modeling made it possible to develop an algorithm for the structural synthesis of a multi-mode NS-controller for controlling the operation of a diesel-generator set of an ESP during transient processes in the entire range, observing the principle of minimum complexity and taking into account the requirements for astatism, physical feasibility, stability and quality of transient processes.
autonomous ship, power plant, diesel, neural network control
1. Titov A.V., Barakat L., Lazovskaya O.Yu., Taktarov G.A., Kovalev O.P. Ocenka riskov ekspluatacii bezekipazhnyh sudov // Morskie intellektual'nye tehnologii- 2019,- № 1-4 (43).-S. 11-23.
2. Ambrosovskiy V.M., Kazunin D.V., Orlov A.P. Sistema upravleniya avtonomnyh i distancionno upravlyaemyh sudov // Morskoy vestnik-2019.-№ 3 (71).-S. 87-90.
3. Tang, Ruoli Optimal operation of hybrid energy system for intelligent ship: An ultrahigh-dimensional model and control method // Energy: the international journal. 2020. Volume211.
4. Machine learning paradigms: applications of learning and analytics in intelligent systems / George A. Tsihrintzis, Maria Virvou, Evangelos Sakkopoulos, Lakhmi C. Jain, editors. Cham: Springer, 2019. 287 p.
