Research On Health Assessment Method Of Main Engine Fuel Supply Unit Based On Cloud Barycenter Method

With the development of science and technology,intelligent ships and unmanned ships have become the inevitable trend of ship development,which also put forward higher requirements for the safety and reliability of marine system.In order to ensure that the marine system can still complete the required tasks under the condition of less crew or no one,avoid the occurrence of faults,and achieve pre-maintenance or condition-based maintenance,it is necessary to focus on health assessment research.In order to improve the intellectualization level of ship engine room,enable people to fully grasp the health status of marine system,provide basis for assistant decision-making,realize pre-maintenance or condition-based maintenance,avoid failure,save maintenance costs and achieve the purpose of ensuring navigation safety,this paper takes the Cloud Barycenter Evaluation Method as health status evaluation method,and takes the Main Engine Fuel Oil Supply Unit as an example,carries out the research on the health status assessment of marine system.Firstly,through the research and analysis of relevant literatures,according to the multi-equipment and multi-parameter characteristics of marine system,combined with the functional analysis system technology of modern TRIZ theory,the establishment method of health status assessment index system is determined,and the classification type and benchmark determination method of evaluation index are discussed.In order to determine the pre-processing method of monitoring data of evaluation index,the monitoring data are analyzed to determine that gross errors are extreme,independent and low-frequency,so a dynamic database method is proposed for gross errors identification.Secondly,taking the main engine fuel oil supply unit as an example,the specific process of establishing the evaluation index system,determining the evaluation index benchmark and classifying the types is given,which also verifies the validity of the dynamic database method in identifying gross errors.Thirdly,in order to establish the calculation model of the optimal comprehensive weight of the evaluation index,the subjective and objective weights of the evaluation index are determined by using the Fuzzy Centralization Statistical Method and the Entropy Weight Method respectively.Then,according to the principle of minimizing the deviation of the total evaluation value,a non-linear programming model of the optimal comprehensive weight is established.Then,according to the types of evaluation indicators,the Cloud Barycenter Evaluation Method is improved,and a dynamic health assessment model is established with Sliding Time Window Method.In order to ensure the reliability of the results,a linear relationship is proposed to calibrate the relationship between the evaluation results and the health value of the evaluation object,and a calculation and analysis method of the health value reliability is given.Finally,taking the Main Engine Fuel Oil Supply Unit as an example,the real ship monitoring data are extracted,and the optimal comprehensive weight model of the evaluation index and the dynamic health evaluation model are verified.The validation results show that the optimal comprehensive weights calculated by the weighting model of evaluation indicators not only accord with people's experience,but also can adjust the weights according to the fluctuation of monitoring data,which improves the rationality of the weights.The health values calculated by using the health state dynamic evaluation model can reflect the changes of the actual health status of the main engine fuel supply unit under long-term operation and abnormal conditions.