An online detection system for the surface quality of aluminum alloy thin-walled parts during milling

Abstract

The invention discloses an online surface quality detection system in the milling process of aluminum alloy thin-walled parts. The module is connected with the surface quality online detection module, and the surface quality online detection module includes 4 well-trained Resnet18 network models in parallel, and a fully connected layer and a softmax layer that are connected successively to the output of the Resnet18 network model. The present invention uses the cutting force signal and noise signal collected in the milling process to perform wavelet packet decomposition and reconstruction and Fourier transform; then the processed signal is input into the trained neural network based on Resnet18 to obtain the surface residual stress , realize the online detection of surface quality during the milling process of aluminum alloy thin-walled parts; it has the advantages of low detection difficulty and good detection effect.

Description

technical field [0001] The invention relates to the field of manufacturing aluminum alloy thin-walled parts, in particular to an online detection system for the surface quality of aluminum alloy thin-walled parts during milling. Background technique [0002] In the field of high-end equipment manufacturing, aluminum alloy thin-walled parts are indispensable and important parts, and the flatness of their surfaces has an important impact on equipment. Therefore, the last link in the manufacture of aluminum alloy thin-walled parts is to test the flatness of the surface of aluminum alloy thin-walled parts. Usually, a batch of aluminum alloy thin-walled parts is tested by random inspection. This detection method exists Problems such as missed inspection of defective products, difficulty in inspection, and high inspection cost. Some authors also proposed to detect the surface quality of aluminum alloy based on image recognition defect method, but because the surface finish of alu...

AI Technical Summary

Problems solved by technology

Therefore, the last link in the manufacture of aluminum alloy thin-walled parts is to test the flatness of the surface of aluminum alloy thin-walled parts. Usually, a batch of aluminum alloy thin-walled parts is tested by random inspection. This detection method exists Problems such as missed inspection of defective products, difficult inspection, and high inspection cost

Some authors also proposed to detect the surface quality of aluminum alloy based on image recognition defect method, but because the surface finish of aluminum alloy is high and the reflection is serious, the details of the collected image will be covered by highlights, which makes the detection based on image recognition defect algorithm difficult and the detection effect is poor and other shortcomings

Images