Design And Implementation Of Ship Main Engine Analysis System Based On Big Data

As an important part of the engine room,the main engine of a ship provides a source of power for the entire ship.Its safety and reliability directly affect the safety of navigation.Effective fault monitoring and diagnosis are very important.The development of the Internet and the Internet of Things technology has made ships move towards automation,informatization,and intelligence.Various collectors with excellent performance are deployed in different parts of the host to collect various parameter information of the host at all times.This process generates a huge amount of data..In order to further use "big data" to improve the monitoring and analysis of the host’s fault diagnosis,energy efficiency evaluation,performance and other issues,a big data platform was built to extract the multi-domain fault signals contained in the massive monitoring data,combined with intelligent algorithm identification,to realize the host status The role of management,thus breaking through the traditional manual inspection relying on expert experience.The work content of this paper is as follows:(1)By studying the application results of big data technology and shipbuilding industry big data technology,it is concluded that the big data analysis platform must have the functions of data signal acquisition,data mining,and model prediction.Against this background,a ship main engine analysis system based on a big data platform is proposed.(2)A statistical analysis of the current host state evaluation method was done,combined with system functions and feasibility requirements,and the overall system framework design was completed.(3)Since the realization of the functions of the entire system is based on the big data framework,it is necessary to build a Hadoop platform with big data processing capabilities.The platform combines Flume technology to achieve data collection,Storm technology to achieve data processing,Kafka messaging middleware to achieve decoupling between Flume and Storm,and data storage through Hbase.(4)Model prediction is the key link of the system.In this paper,through the combination of supervised learning K-means and unsupervised learning BP neural network,the establishment of the main engine energy efficiency analysis model is realized,and the main component analysis is designed to reduce the dimensionality of the data,and finally the model is realized.Verification and analysis.(5)Build the entire system in the cloud server,and the clustering parameter setting,weight distribution,database modification.