CN-122007981-A - Large-scale equipment cutter processing monitoring method and system based on artificial intelligence

CN122007981ACN 122007981 ACN122007981 ACN 122007981ACN-122007981-A

Abstract

The invention provides a large-scale equipment cutter processing monitoring method and system based on artificial intelligence, and relates to the technical field of cold processing monitoring; processing and identifying the sequential image data, determining the position area information of the cutter in the image, determining the corresponding cutter area image according to the position area information, obtaining quantized cutter abrasion loss data according to the cutter area image based on an abrasion analysis model, and generating a monitoring result for guiding machining operation based on the quantized cutter abrasion loss data. According to the invention, through automatic closed loop of image acquisition, positioning, analysis and decision, an operator can continuously grasp key information without interrupting processing, so that the shutdown checking time is obviously reduced, and the comprehensive utilization efficiency of equipment and the continuity of production rhythm are improved.

Inventors

LIU WUQIANG
LI HONG
LUO DONG
GUO XINBO
Lv jifeng
MENG FANXING

Assignees

中国第一重型机械股份公司
一重集团(黑龙江)重工有限公司

Dates

Publication Date: 20260512
Application Date: 20251224

Claims (10)

1. The method for monitoring the processing of the large-scale equipment cutter based on artificial intelligence is characterized by comprising the following steps of: acquiring time sequence image data containing a cutter and a workpiece processing area in real time; processing and identifying the time sequence image data, determining position area information of the cutter in an image, and determining a corresponding cutter area image according to the position area information; based on the abrasion analysis model, obtaining quantized cutter abrasion loss data according to the cutter area image; and generating a monitoring result for guiding the machining operation based on the quantized tool wear amount data.
2. The method for monitoring tool processing of large-scale equipment based on artificial intelligence according to claim 1, wherein the processing and identifying the time-series image data to determine the position area information of the tool in the image comprises: And processing the time sequence image data based on a preset image recognition algorithm to determine the position area information of the cutter in the image.
3. The artificial intelligence based large scale equipment tool process monitoring method of claim 1, wherein the wear analysis model is constructed based on a modified YOLOv model, the modified YOLOv model construction process comprising: Obtaining an original YOLOv model; and adding a cascade aggregation attention mechanism to the neg network in the original YOLOv model.
4. The method for monitoring tool processing of large-scale equipment based on artificial intelligence according to claim 3, wherein the training and optimizing process of the wear analysis model comprises: Cutting out a cutter region sample image from the original training image according to the real cutter position label, measuring and quantifying the abrasion degree in the cutter region sample image, and obtaining real abrasion amount data serving as a supervision label; And taking the cutter region sample image as an original model input, taking the corresponding real abrasion loss data as a training target, training and optimizing the original model, and taking the optimized original model as the abrasion analysis model.
5. The method of claim 4, wherein generating a monitoring result for guiding a machining operation based on the quantized tool wear amount data comprises: Analyzing the time sequence image data and extracting at least one piece of processing visual characteristic data; Comparing the quantized cutter wear amount data with a preset wear threshold value to obtain a comparison result; And based on the comparison result, performing joint analysis according to at least one piece of processing visual characteristic data and the corresponding working condition standard to generate the monitoring result.
6. The method of claim 5, wherein the machining process visual characteristic data comprises at least one of chip status data, coolant coverage data, and trajectory deviation data.
7. The method for monitoring tool processing of a large-scale equipment based on artificial intelligence according to claim 1, wherein the generating of the monitoring result for guiding the processing operation based on the quantized tool wear amount data further comprises: and carrying out association storage on the time sequence image data, the corresponding position area information, the cutter area image, the quantized cutter abrasion loss data and the monitoring result to form a traceable machining process file, and setting a storage period and access authority of the machining process file.
8. The method for monitoring tool processing of a large-scale device based on artificial intelligence according to claim 7, wherein the storing the time-series image data, the corresponding position area information, the tool area image, the quantized tool wear data and the monitoring result in association with each other, after forming a traceable processing procedure file, further comprises: based on the processing process file, analyzing the correlation data between the monitoring result and the final processing quality; And dynamically adjusting and optimizing the abrasion threshold and the working condition standard used in the process of generating the monitoring result according to the associated data.
9. The method for monitoring the processing of the large-scale equipment cutter based on the artificial intelligence according to claim 1, wherein the method for monitoring the processing of the large-scale equipment cutter based on the artificial intelligence further comprises: And installing the image acquisition equipment in a machining area of a machine tool through an adjustable bracket, and adjusting the pose and parameters of the image acquisition equipment so that the image acquisition equipment covers the cutting contact point of the cutter and the workpiece.
10. Large-scale equipment cutter processing monitored control system based on artificial intelligence, characterized by, include: the image acquisition unit is used for acquiring time sequence image data containing the cutter and the workpiece processing area in real time; The processing and identifying unit is used for processing and identifying the time sequence image data, determining the position area information of the cutter in the image and determining a corresponding cutter area image according to the position area information; The abrasion analysis unit is used for obtaining quantized cutter abrasion loss data according to the cutter area image based on an abrasion analysis model; And the monitoring decision unit is used for generating a monitoring result for guiding the machining operation based on the quantized tool wear amount data.

Description

Large-scale equipment cutter processing monitoring method and system based on artificial intelligence Technical Field The invention relates to the technical field of cold working monitoring, in particular to a large-scale equipment cutter working monitoring method and system based on artificial intelligence. Background Heavy equipment such as a large vertical lathe, a floor boring and milling machine and the like is core equipment for manufacturing large key components (such as a nuclear power seal head, a marine crankshaft and a wind power main shaft). The quality and efficiency of the finishing process performed on these devices, especially the processing of large complex curved surfaces and deep hole cavities, are directly related to the success or failure of major projects. However, current industry monitoring of machining processes, especially the status of the core consumable tool, is severely dependent on personal experience of the operator and close visual inspection. The conventional method has a plurality of inherent and difficult-to-overcome defects that firstly, when a large workpiece is machined, a huge tool rest, a workpiece body and a complex clamp often form serious visual shielding, and an operator can not directly and clearly observe the actual contact area of a tool tip and the workpiece, so that the machining process is in a 'blind machining' state to a great extent. Secondly, the judgment of the cutter abrasion state is completely dependent on untimely and short-distance visual inspection of operators, and the method is high in subjectivity, incapable of quantification, low in efficiency, required to be frequently stopped and affected in production beat. More serious, in order to observe the processing condition in the deep hole or at a specific angle, an operator often needs to climb to the high position of a machine tool or approach to a processing area, and the operator is placed in the potential movement risks of chips, cooling liquid and equipment which can splash, so that the potential safety hazard of the operator is outstanding. The problems jointly cause unstable processing quality, unexpected damage of cutters and even high risk of equipment collision accidents, and the problems become technical bottlenecks for restricting quality improvement and efficiency improvement and safe production of high-end heavy equipment manufacturing industry. Therefore, a method for remotely, real-time, objectively and quantitatively monitoring tool processing is needed to overcome the inherent defects of the traditional manual monitoring mode, ensure processing safety and improve product quality and production efficiency. Disclosure of Invention The present invention solves one or more of the above-mentioned problems of the related art. In order to solve the problems, the invention provides a large-scale equipment cutter processing monitoring method and device based on artificial intelligence. In a first aspect, the present invention provides a method for monitoring tool processing of a large-scale device based on artificial intelligence, including: acquiring time sequence image data containing a cutter and a workpiece processing area in real time; processing and identifying the time sequence image data, determining position area information of the cutter in an image, and determining a corresponding cutter area image according to the position area information; based on the abrasion analysis model, obtaining quantized cutter abrasion loss data according to the cutter area image; and generating a monitoring result for guiding the machining operation based on the quantized tool wear amount data. Optionally, the processing and identifying the time-series image data, determining the position area information of the cutter in the image, includes: And processing the time sequence image data based on a preset image recognition algorithm to determine the position area information of the cutter in the image. Optionally, the wear analysis model is built based on a modified YOLOv model, the modified YOLOv model building process includes: Obtaining an original YOLOv model; and adding a cascade aggregation attention mechanism to the neg network in the original YOLOv model. Optionally, the training and optimizing process of the wear analysis model includes: Cutting out a cutter region sample image from the original training image according to the real cutter position label, measuring and quantifying the abrasion degree in the cutter region sample image, and obtaining real abrasion amount data serving as a supervision label; And taking the cutter region sample image as an original model input, taking the corresponding real abrasion loss data as a training target, training and optimizing the original model, and taking the optimized original model as the abrasion analysis model. Optionally, the generating a monitoring result for guiding the machining operation based on the quantized tool wear amount data includes