Research On State Awareness Technology Of Marine Seawater Cooling System

The shipbuilding industry has been developed rapidly in recent years,unmanned ship and intelligent engine room are its development tendency.How to perceive and judge the state of the ship systems and equipment correctly is the premise of achieving intelligent ship.But with the development of complex ships,the ship power system is also becoming increasingly complex,the marine engineers have been difficult to judge the state of the ship systems and equipment according to their own experience,so it is of great significance to research the state awareness technology of the ship systems and equipment.Based on the analysis of the development of the state awareness technology both at home and abroad,this paper takes the Dalian Maritime University training ship MV"YUKUN" as research object,then builds the seawater cooling system state awareness model.First of all,the working principle and typical faults of MV "YUKUN" seawater cooling system and its influence on the system are analyzed in detail.Comparing with conventional state perception system framework,the state perception system framework for the ship's seawater cooling system is proposed.At the same time,based on thoroughly analysis of fault diagnosis technology and state parameter prognostics technology,and according to the characteristics of the ship sea water cooling system state parameters,then the SOM neural network and ARMA time series algorithm are chosen as the research method of this article.Secondly,the SOM neural network fault diagnosis model and state parameters of ARMA time series prognostics model are built.In the building process of fault diagnosis model,in order to improve the model,the fault samples self-updating operation is introduced.Then using the historical data of MV "YUKUN" cooling system which is collected under normal working condition to verify the validity of the fault diagnosis model and state parameter prognostics model,the verification results show that the two models are correct and effective.Finally,the SOM neural network fault diagnosis model and state parameters of the ARMA time series prognostics model are fused,building complete ship seawater cooling system state awareness model,and using the parameters of the normal and fault condition to verify this state awareness model and confirm it is feasible and effective.