| The maintenance of ship equipment is the key to ensure the safe and reliable operation of ships.In the traditional maintenance work,the engineer can make accurate judgments based on his own experience based on the operating status of the ship system and its equipment.However,with the development of ship automation and intelligence,ship types,equipment,and systems of ships are becoming more and more complicated.It is difficult to solve problems by relying on experience alone.In order to guarantee the navigation safety of the ship,increase the service life of the equipment,and reduce the workload of the engineer,it is necessary to formulate a reasonable maintenance plan.In this paper,the sea water cooling system of "Yu Kun" ship,a teaching practice ship of Dalian Maritime University,is taken as the research object.By using the methods of fault identification,parameter prediction and state evaluation,a complete set of state prediction model of sea water cooling system is established.The model is mainly composed of three sub-models:fault identification model,parameter prediction model and state assessment model.The fault identification model is mainly based on the actual analysis of the working principle and typical faults of the sea water cooling system of the "Yu Kun" ship.The specific RBF neural network fault identification process and network structure are designed.The network is trained by the state parameters of the sea water cooling system in various states,and the accuracy of the RBF neural network fault identification model is verified by the test set.The parameter prediction model uses the time series ARIMA prediction model to initially predict the state parameters of the seawater cooling system.In order to avoid the information loss caused by the difference and the instability of the single model,at the same time,a weighted Markov chain prediction model is introduced to optimize the residual error of the ARIMA model.The historical data of the collected seawater cooling system is predicted,and the results are compared with the real data to verify.The results show that the prediction accuracy of the optimized model is significantly improved,and the accuracy meets the prediction requirements.The state evaluation model adopts AHP-FUZZY comprehensive evaluation method to evaluate the system state.Among them,the analytic hierarchy process determines the weight between the evaluation factors of the seawater cooling system.After the hierarchical single sorting and consistency test and the hierarchical total sorting and consistency test are passed,it is used as the weight vector of the fuzzy comprehensive evaluation factor set to realize the ship Evaluation of the operating health of the main seawater system.An example analysis is carried out on the established state prediction model of marine seawater cooling system,the collected seawater cooling system state parameter sequence is input into the model,and the experiment is carried out according to the established marine seawater cooling system state prediction model process.In comparison,the results show that the state prediction model is practical,accurate and effective.The innovation of this paper lies in the reasonable combination of the three methods of fault identification,parameter prediction and state evaluation.On the basis of parameter prediction,the combination of fault identification and state evaluation is combined to avoid the absoluteness of the fault identification results and explore potential fault The influence of the system operating state has obtained a more intuitive and instructive result for the system state,and achieved an effective prediction of the system state. |
