Multi-empirical formula and BP neural network model fused concrete penetration depth prediction algorithm

Abstract

The invention relates to a multi-empirical formula and BP neural network model fused concrete penetration depth prediction algorithm, including the steps: firstly, preprocessing an original sample; carry out dimensionality reduction on original sample data by using a dimensionless method; selecting a certain number of empirical formulas as required; and setting and establishing an empirical formula model and a BP neural network model, fusing the empirical formula model and the BP neural network model into a multi-empirical formula and BP neural network fusion model, and finally predicting thepenetration depth of the high-speed projectile acting on the concrete by using the model. According to the multi-empirical formula and BP neural network model fused concrete penetration depth prediction algorithm, a data fusion thought is adopted, and a plurality of empirical algorithms and the BP neural network are fused to form a new fusion model, so that the optimal empirical formula can be automatically used for compensating the precision under the condition that the prediction precision of the neural network is relatively low, and the prediction precision of the whole fusion model is improved.

Description

technical field [0001] The present invention relates to a prediction algorithm for the penetration depth of high-speed projectiles acting on concrete. It is a prediction algorithm for predicting the penetration depth of high-speed projectiles acting on concrete by using a neural network. Specifically, it is a fusion of multi-empirical formulas and BP neural network models. Concrete penetration depth prediction algorithm. Background technique [0002] The study of the penetration, penetration and damage effects of the projectile after acting on the target plays a very important role in the development of new warheads and the evaluation of the strike effect. The process of projectile penetration and penetration of the target is extremely complicated and involves many mechanical behaviors. Moreover, during the penetration process, deformation, abrasion, and burning of the target body and the projectile will occur, making it very difficult to accurately predict the penetration a...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to propose a concrete penetration depth prediction algorithm for the fusion of multiple empirical formulas and BP neural network models. For the problems in the above background technology, the present invention adopts the idea of ​​data fusion, and combines multiple empirical algorithms and BP neural network Fusion to form a new fusion model, so that the best empirical formula can be automatically used to make up for the accuracy when the prediction accuracy of the neural network is relatively low, thereby improving the prediction accuracy of the entire fusion model

Images