Corrosion Prediction Model Of Aging Aircraft Structure Based On Grey Neural Network

With the gradual aging of domestic fleet aging problem,the adoption of scientific and effective maintenance mode can not only meet the airworthiness requirement of aging aircraft,but also increase the operation economic line during the Life span of aircraft.In the face of aging corrosion in aircraft structure is very popular,and increase the maintenance tasks and the installation of the sensor device can reduce the premise of economy,the intelligent algorithm is effective to predict the corrosion damage reasonably is very important,therefore become the main research focus of this paper.The domestic aging aircraft actual corrosion damage repair data,analysis the regional distribution characteristics of aging aircraft corrosion damage,corrosion depth and corrosion from a number of features to predict corrosion damage are analyzed in this paper.At the same time,combined with the characteristics of the aging aircraft maintenance task,the grey theory and neural network are used to predict the corrosion damage.In this paper,from the perspective of a single model,based on the grey theory in the unequal time interval model is optimized using the improved unequal interval prediction model of corrosion damage;the results show that the corrosion depth of one feature has a good predictive ability,but the number of predicted results of poor corrosion can not be done well.In order to overcome the shortcomings of the single model,the grey neural network combined model is used.The combination model of unequal time interval model and BP neural network is determined and realized.The two characteristics of the corrosion damage are predicted by the grey neural network mapping model and the coupling model respectively.The effect of prediction is greatly improved.Finally,the multi parameter grey neural network model is established to comprehensive predict the corrosion damage.Prediction results show,the corrosion depth has the characteristics of monotonicity,randomness and periodicity;the average number of corrosion with each structure contains a periodic variation of corrosion damage under fluctuated upward and related rules.It is explained in the article.