Research On The Optimization Of Civil Aviation Engine Shop Visit Decision-making

The service life of civil aero-engine is long.The civil aero-engine shop visit is one of the important contents of civil aircraft maintenance business within the life time.A reasonable engine shop visit decision is not only related to the security of engine's operation,but also has an effect on the utilization rate and the maintenance cost of engine.Therefore,it is of great significance to optimize all aspects of engine shop visit decision-making.In view of the existing problems in the engine shop visit of airlines,this paper carries out the following studies:Firstly,this paper studies the prediction method of civil aero-engine's remaining useful life(RUL).Adopt kernel principal component analysis to process the engine monitoring data and obtain the degradation feature information of engine.Then,merge it into health index(HI).Use relevance vector machine(RVM)to establish a regression model of health index by which the prediction of engine's remaining useful life comes true.Compared with the existing methods,the results show that the proposed method can improve the prediction accuracy of engine's remaining useful life.Secondly,this paper optimizes the scheme of engine shop visit from the perspective of life limit parts.Regard the multiple life limit parts replacement problem as a multi-objective optimization problem and analyze the constraint conditions.Then,establish an optimization model and utilize particle swarm optimization algorithm to solve the model.After evaluating each within the obtained scheme set,select the optimum scheme as the ultimate scheme of engine shop visit.In addition,the proposed method is verified through the actual data of life limit parts in an engine.Finally,this paper studies the method of engine maintenance level decisionmaking.Take the maintenance level of engine's module as the decision-making object and establish an optimization model based on condition monitoring.The model sets the historical data of engine shop visit as the training samples and optimizes the sample space by large margin nearest neighbor algorithm.To determine the maintenance level of module,match the current data from the condition monitoring of engine with the training samples.A case study shows that the proposed model can meet the demand for practical engineering and provide support for the decision-making of engine maintenance level.