葡萄京·手机官网（中国）官方网站

Quality Inspection and Process Optimization of Stamping Parts Based on Artificial Intelligence Techn

The collision of “physical world” (represented by manufacturing equipment) and “digital world” (represented by artificial intelligence, sensors, etc.) has spawned a huge transformation in manufacturing, and the integration of the two worlds will be the next round of economic development. Inject new kinetic energy. New technologies, represented by artificial intelligence, are having a huge impact on production processes such as production processes, production models and supply chain systems. The application value of artificial intelligence technology in the diagnosis of manufacturing process is gradually becoming more and more prominent, especially in the quality inspection and process optimization of stamping parts, which is playing an incomparable advantage. In short, artificial intelligence related technology can replace the human eye to complete the functions of identification, measurement, positioning and judgment of stamping parts. Not only does artificial intelligence have the ability of “learning”, but it can be tuned by sample accumulation and model training to accurately predict. The risk of cracking of stamping parts, so as to achieve precise control and optimization of the quality of stamping products. The following is an application case of artificial intelligence technology in the automobile manufacturing press shop.

Background of the project

In mechanical manufacturing, press forming is a very important plastic processing method and is widely used in the automotive, aerospace, electrical and other industrial fields. As we all know, most of the cover parts and structural parts of the automobile body are thin plate stamping parts, and the level of stamping technology and the quality of stamping are very important for automobile manufacturers.

The stamping workshop of the production base of an automobile manufacturing enterprise has three stamping production lines, which mainly produce passenger car body coverings with large contours and spatial curved shapes, such as side panels, fenders, doors and hoods. In the stamping production process, part of the side wall is prone to local cracking during the drawing process, and parameter adjustment and trial production are required repeatedly; at the end of the production line, a large number of quality inspectors are required to perform manual inspection of the surface defects of the stamping parts.

Problems and challenges

1. The existing inspection method of the end line of the stamping production line is manual manual inspection. It is necessary to quickly sort out the stamping parts with surface defects such as cracks, scratches, slip lines and embossed bags within the limited production cycle time. It is not uniform, the stability is not high, and the quality inspection data is difficult to be effectively quantified and stored, which is not conducive to enterprise data resource collection, quality problem analysis and traceability.

2. In the trial production process of stamping production, there are many factors affecting the side cracking in the stretching process, such as equipment parameters, mold state, sheet properties, etc. The adjustment parameters and repeated trial production methods have certain blindness and cost. Large and inefficient.

3. There are many influencing factors and large differences in data forms, and they are distributed in different business systems of the workshop. Both real-time data and unstructured image data have extremely high requirements for data collection, management and storage.

solution

Based on the above situation, Merrill Lynch Data builds a big data platform for enterprises to realize the integration, storage and unified control of equipment, molds, materials, manufacturing process data, quality inspection data, and data mining based on machine learning. The intelligent detection technology based on machine vision realizes the prediction of the side cracking cracking and the intelligent identification of the surface defects of the product parts.

◎According to the processing parameters of stamping equipment, sheet parameters, mold performance parameters and maintenance records, the intelligent predictive model of stamping process was established through the data excavator learning algorithm. Accurately predict the risk of stamping cracking through sample accumulation and model training tuning. Finally, determine the correlation between the influencing factors of the manufacturing process, and formulate the production process parameter combination control strategy to support the process optimization and quality control of the stamping manufacturing process.

◎ Intelligent recognition and detection of stamping parts defects based on machine vision, based on the existing conditions of the production line, design the image acquisition system, through real-time image acquisition and intelligent analysis, quickly identify whether the stamping parts have surface defects, and automatically process all the detected images and process data. Store to the big data platform. Through the correlation between quality inspection data, production process process parameters and product design parameters, and using big data analysis technology, the closed-loop connection of analysis and management of stamping product quality problems is formed, and the precision control and optimization of stamping product quality are realized.

Value

1. By predicting the risk of cracking of stamping parts, improve the design efficiency of stamping parts processing parameters of new models, reduce the number of trial production and trial production.

2. Quickly and intelligently detect the surface defects of stamping parts, improve the stability and reliability of production line inspection, and reduce the labor intensity and labor cost of quality inspection workers. At the same time, product quality inspection data is effectively stored, providing important data support for quality closed-loop analysis and traceability.

3. Exploring a practical and feasible demonstration road for the enterprise's intelligent manufacturing transformation, and accumulated valuable experience for the promotion and application of industrial big data and artificial intelligence in peer enterprises.

Applicable industry

Automobile manufacturing, aerospace, home appliance production, etc. have stamping and spraying processes, and have high requirements on the surface quality of products.