Mechanism Reliability Analysis of Ammunition Conveyer Based on Deep Learning Neural Network

Ammunition conveyer is the main part of automatic loading mechanism, the reliability of it would affect the accuracy of artillery directly. Due to the errors generated in manufacturing and assembling, there will be a certain impact on the accuracy and reliability of the ammunition conveyer. In order to improve the computational efficiency of mechanism reliability, a mechanism reliability estimation method based on deep learning neural network was proposed in this paper. Firstly, the dynamic simulation model of the ammunition conveyer at 0 degree firing angle is established in ADAMS. During establishment, the structural uncertainties were considered and parameterized. Then establishing a co-simulation between ADAMS and MATLAB to obtain the database, which was utilized to the BP neural network training. After network training, building a reliability estimation process based on BP neural network trained before. The results show that the proposed method has a much higher computational efficiency compared to the MC method with guaranteed accuracy. Thus, the method provides new ideas for reliability analysis of the ammunition conveyer and other complex mechanisms.