Analysis and optimization of energy efficiency of ships and decision management currently is the key development trend of intelligent ships,the study of ship condition monitoring as a research focus by high-tech company and scientific research.Through monitoring the performance status of ship's engine room to improve operational efficiency,energy saving and emission reduction,finally,the green ship and unmanned ship can be realized.The main work is as follows:In the view of data science,this paper studies the basic methods of feature engineering,including data acquisition and data processing.This work proposes a solution to ship integrated big data pool for ships multi-sensor data fusion.We constructed a binocular vision platform on the sea,do the work of the depth of data mining and information extraction of visual learning and monitoring model optimization by the use of big data technology operation and maintenance of ship,environmental perception system for ship operation is realized.Analysis of the processed main engine condition monitoring data.Based on Principal component analysis method,the key operating parameters such as speed,fuel consumption,smoke exhaust temperature,cylinder liner water temperature,roll value and speed are selected for correlation analysis and data fusion.K-means clustering and Gauss hybrid clustering are used to realize the automatic classification of the working conditions of the main engine.In order to verify the feasibility of the system,WPF-based human-machine interface is compiled on Visual Studio programming platform.Through the interface with MATLAB and MS SQL Server database,the functions of data reading,data cleaning,data classification and data visualization are realized.In this paper,the big data processing and analysis functions of ship engine room under the condition monitoring scenario are preliminarily realized.The content of this paper has reference value for the development of intelligent monitoring system of marine engine room based on data drive. |