Study On Dynamic Baseline Methods Of Marine Power Equipment Affected By Typical Factors

In recent years,the research of intelligent ship and unmanned ship has become a hot topic in the ship industry.As an important part of intelligent ship,intelligent engine room needs to have the function of data collection,health status assessment,assistant decision-making and condition-based maintenance.Among the health status assessment methods of marine power equipment,measuring the deviation between real-time monitoring parameters and baseline to obtain health status assessment values has the advantages of less calculation and strong applicability.However,the monitoring parameters and working condition of marine power equipment will be affected by many factors.If static baseline is used in baseline status assessment method,the accuracy of health status assessment will be reduced,and even the problem of false alarm will arise.In order to solve these problems,this paper mainly focuses on the studies of the following aspects:Firstly,analyzing the influencing factors that lead to the trend change of marine power equipment monitoring parameters.And selecting the variable load of marine diesel engine and performance degradation of marine power equipment as typical influencing factors.Then establishing different dynamic baseline calculation methods under different typical influencing factor,and the prediction method under the condition of performance degradation.Secondly,taking the variable load as the typical factor.Because there is no monitoring parameters or monitoring parameters couldn't cover the full load range during the sea trail,the reference-site method is used to increase the reference-sites and update the reference points.Then the dynamic baseline under variable load conditions can be obtained by dynamic fitting of reference points.In the process of dynamically updating reference points,using sliding probability neural network to reduce computational complexity and improve the updating efficiency of reference points.Using the cylinder exhaust temperature to verify the feasibility of the method,the results show that this method can calculate baseline in real time and update baseline dynamically,the change rule of dynamic is consistent with the trend of monitoring parameters.Thirdly,taking the degradation of equipment performance as the typical factor.Collecting monitoring parameters according to performance degradation time series.Then establishing sliding probability neural network to obtain the dynamic performance reliability of the target equipment.And establishing the conversion function between performance reliability and baseline value to obtain the degradation baseline.Using the sea water pump outlet pressure to calculate degradation baseline,then comparing with the baseline that obtained by MarQuart algorithm.The results show that the method can update the baseline dynamically according to the target equipment's performance degradation degree and improve the dynamic baseline's accuracy.Finally,combining the degradation baseline calculation method and hybrid parameter prediction model,establishing the degradation baseline prediction method to predict the change time of the baseline value.Using the sea water pump outlet pressure to verify the feasibility of the method,the results show that this method can accurately predict the change time of the dynamic baseline value under the premise of determining the trend of baseline value change.The dynamic baseline calculation and prediction methods are basic research to realize the health status assessment and prediction of marine power equipment.It is of great significance to realize the intelligent operation and maintenance of engine room and improve the function of marine intelligent engine room system.The dynamic baseline that considers the external factors is more scientific and reasonable than static baseline status method.And these methods can improve the accuracy of the health status assessment value.The baseline prediction method provides idea for condition-based maintenance of engine room and health status prediction.