Quantile regression analysis of target parameters and operational risk forecasting for marine technology

• Main Author: N. Kovtun
• Format: Article
• Language: English
• Published: Krylov State Research Centre 2020-06-01
• Series: Труды Крыловского государственного научного центра
• Subjects: marine technology; quantile regression; machine learning; risk matrix; reliability; diagnostics; forecast; statistical simulatuion
• Online Access: https://transactions-ksrc.ru/eng/archive/quantile-regression-analysis-of-target-parameters-and-operational-risk-forecasting-for-marine-techno/

Object and purpose of research. This paper discusses forecasting methods for technical deterioration risks of complex marine technology. The purpose of the study was to formulate the proposals on development and induction of a system for non-stop monitoring, data acquisition and statistical analysis of key performance parameters of ship equipment so as to highlight its current condition and trends, as required by Russian Maritime Register of Shipping. Materials and methods. Statistical simulation was used to study the method of quantile regression analysis for target performance parameters of ship equipment, as well as applicability of this method to assessment of technical deterioration risks based on risk matrices and standard reliability schemes. Main results. Quantile regression analysis, risk matrices of target parameters and structural schemes of ship equipment reliability were used to suggest and justify the method, as well as describe the software based on it and intended to perform technical diagnosis and forecasting of ship equipment with calculation of corresponding risks. Conclusion. The method suggested in this paper to forecast technical deterioration risks of sophisticated marine technology complies with the requirements and recommendations of ITTC, IMO and Russian Maritime Register of Shipping.