Prediction Of Barrel Life Based On Grey Theory And Neural Networks

The barrel is an important part of the gun and a key component in determining the performance of the gun,and the life of the barrel represents the life of the whole gun to a certain extent.By predicting the life of the gun barrel,it can not only reduce the probability of risk and effectively reduce the maintenance cost of the gun,but also scientifically guide the combatants to use the gun reasonably and improve the scientific management of the gun by the troops in peacetime and wartime,so it has the value and necessity of research.This paper takes a tank gun as the research object,combines internal ballistics,heat transfer,finite element method,gray theory,neural network and other multidisciplinary theories and technologies to analyze and study the body tube life problem caused by its use.The main research of this paper is as follows.(1)On the basis of the study of the factors influencing ablative wear and the law of bore wear,the internal ballistic equations of ablative wear are established and solved to analyze the influence caused by ablative wear on ballistic performance and body tube life.The thermal response curves of the inner and outer walls and the radial temperature distribution of the gun barrel under different firing conditions are obtained by applying the finite element method to numerically analyze the heat transfer of the barrel during the firing of the gun.(2)Analyze the types of artillery body tube life and its judgment criteria,and determine the use of radial wear at the beginning of the body tube rifling to evaluate and predict the body tube life and calculate the maximum number of projectile rounds.By studying the gray theory prediction model,an improved gray GM(1,1)model with higher prediction accuracy is established.Based on the actual measured body tube wear volume data is always non-equally spaced,a reasonable method to handle non-equally spaced wear volume data is proposed.(3)Compare the respective advantages and disadvantages of BP neural network model and gray theory prediction model,and establish a gray neural network prediction model that combines the advantages of both prediction models.It is proved by example application that the established gray neural network prediction model has higher accuracy and better adaptability,and can make accurate prediction of the change pattern of the bore wear amount and body tube life of artillery.(4)Based on the established gray neural network prediction model,the Matlab software is used to build the body tube wear prediction software and complete the visualization of the body tube wear prediction and analysis results.Through example application,it demonstrates how to use the software to complete the wear amount and life prediction to meet the requirements of engineering practice,which has certain practical application value.