Research On Prediction Of Warship Spare Parts Based On Logistic Support Analysis And CBM

The structure of warship equipment is complex and there are many components.It has special task environment and limited storage space.The optimal allocation of ship support resources is an important research content of equipment support.Reasonable and efficient allocation of equipment support resources is a prerequisite for the completion of equipment support tasks.Whether the equipment support resources and train equipment are matched or not determines the quality of the equipment comprehensive support effect.In order to avoid equipment downtime or mission termination caused by the shortage of support resources,especially spare parts,it is necessary to have sufficient types and quantity of spare parts.Due to the limited storage space and cost of spare parts on board,there are some constraints on the configuration of spare parts.Therefore,the scientific allocation method of ship spare parts has always been a key research issue in the field of equipment support.This thesis presents a method for predicting the types and quantity of ship spare parts which are suitable for practical engineering through the research of supportability analysis and maintenance according to the situation.Firstly,by studying the basic concepts and methods of supportability analysis,the main steps and contents of determining initial spare parts are proposed.The structure tree decomposition of ship equipment is carried out to determine the maintenance work type of components and construct the maintenance task assignment table,which lays a foundation for the analysis of spare parts varieties in the support analysis work.In terms of spare parts allocation,the priority of spare parts allocation is considered.The spare parts are prioritized by value engineering theory.This thesis proposes a spare parts allocation priority model based on improved TOPSIS-VE,which optimizes the configuration scheme of initial spare parts of ship equipment.Secondly,this thesis proposes a kind prediction method of equipment follow-up spare parts based on condition-based maintenance,which reflects the difference and individuation of spare parts resource allocation.In order to find the weak points in the execution of the equipment task,the equipment state analysis model was built based on the maintenance of the equipment.The concept of social hierarchy in Grey Wolf Optimizer(GWO)was introduced to artificial fish swarms algorithm(AFSA).The experimental results show that the model has obvious advantages in the accuracy of equipment state recognition,which provides a strong theoretical support for determining the specific weak points of equipment.It can determine the type of spare parts according to the weak points and revise the maintenance support list.Finally,the thesis proposes a method to predict the quantity of spare parts on board based on the remaining life of equipment,which can solve the problem of the quantity prediction of the following spare parts.LSTM network is used to predict the remaining life of equipment.By constructing LSTM training network,the trend of degradation degree can be quantified.The degradation trend and specific degradation curve of bearings were obtained by polynomial fitting.Re-LSTM training model is proposed in this thesis to solve the lag problem of predicted degradation curve.This thesis obtained the residual life value and actual life cycle of equipment with high reliability.This thesis presents a prediction method for the quantity of spare parts on board based on the life cycle of the equipment.The actual life value of the equipment can be brought into the model to obtain the quantity of subsequent spare parts on board in line with the actual operation state of the equipment.