Shot peening strengthening surface integrity prediction method based on BP neural network

Abstract

The invention discloses a shot peening strengthening surface integrity prediction method based on a BP neural network. The method comprises the following steps: 1, collecting and arranging part shot peening strengthening test data, and determining input and output parameters of a surface integrity prediction model by analyzing process parameters and surface integrity evaluation indexes influencingshot peening performance in a shot peening strengthening process; preprocessing the data; 2, determining a hidden layer activation function type and a hidden layer node number, and constructing a BPneural network structure; 3, optimizing the initial weight and bias of the BP neural network through a genetic algorithm, and establishing a shot peening strengthening residual stress and surface roughness prediction model; 4, determining the weight and bias in the BP neural network structure according to the precision evaluation parameter R2; 5, predicting the shot peening strengthening residualstress and surface roughness by using the trained model meeting the precision requirement. The test cost of the shot peening strengthening process can be reduced, and the efficiency of the shot peening process is improved.

Description

technical field [0001] The invention belongs to the technical field of intelligent manufacturing, and in particular relates to a method for predicting surface integrity of shot peening strengthening. Background technique [0002] Shot peening is a widely used surface strengthening process at present, and it is often used for surface strengthening of key components in high-speed trains, aerospace, ships and other fields. Shot peening can introduce residual compressive stress on the surface of the part, improve the surface structure of the part, and delay the fatigue failure of the part. Shot peening treatment of parts can improve fatigue resistance, which is of great significance to improve their service performance. [0003] At present, in the shot peening process of parts, the research on the law of process parameters such as projectile diameter, spray angle, spray velocity, coverage rate, spray pressure and shot flow, and surface integrity evaluation indicators such as re...

AI Technical Summary

Problems solved by technology

Its disadvantage is that the establishment process of the relationship model between the residual compressive stress field and the shot peening process parameters is complicated, and some factors in the model are determined by physical tests.

[0006] Chinese patent document CN 109508488A discloses a method for predicting shot peening process parameters based on genetic algorithm optimization of BP neural network. It uses BP neural network to predict shot peening process parameters. The disadvantage is that it does not consider the input of the neural network prediction model. Correlation between parameters and is only used to predict nozzle movement velocity, not residual stress

Images