Data sample expansion method and system based on ship motion spectrum characteristics

• Main Authors: HUANG LIMIN, LI MAO, CHEN HANGYU, WANG XIANGLU
• Format: Patent
• Language: Chinese; English
• Published: 05.04.2024
• Subjects: CALCULATING; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention belongs to the technical field of ship and ocean engineering, and discloses a data sample expansion method and system based on ship motion spectrum characteristics. According to the method, Fourier transform is carried out on limited collection time calendar fragments to obtain frequency spectrum features, a random phase is added to discretize a motion frequency spectrum, and then motion time calendar data under the current working condition is expanded to achieve expansion of the number of samples. Through the method, the problem of insufficient actually measured ship motion data volume can be effectively improved, sufficient training samples are provided for training of a ship motion deep learning forecasting model, and effective improvement of forecasting precision is realized. According to the method, the distribution characteristics of the motion time calendar data in the frequency domain are effectively extracted, and the new ship motion time calendar with consistent characteristics is obt