Intelligent evaluation method for automobile steel

Abstract

The invention relates to an intelligent evaluation method for automobile steel, and belongs to the technical field of automobile steel. Preprocessing the automobile steel product image to obtain a preprocessed automobile steel product image; performing feature extraction on the preprocessed automobile steel product image by using a feature extraction network to obtain a feature image; marking the feature image to obtain a training sample; training a deep learning model by using the training sample to obtain an automobile steel product quality evaluation model; and performing preliminary evaluation on a target automobile steel product by using the automobile steel product quality evaluation model to obtain an evaluation result. According to the method, evaluation of the automobile steel is completed through the deep learning model, subjective differences caused by manual visual inspection and experience judgment are avoided, and a stable quality judgment rule can be established conveniently.

Description

Technical Field

[0001] This invention relates to the field of automotive steel technology, and more specifically to an intelligent evaluation method for automotive steel. Background Technology

[0002] In existing technologies, the evaluation of automotive steel mainly relies on traditional mechanical property tests and process performance tests, such as tensile tests, bending tests, impact tests, cupping tests, and forming limit tests, combined with metallographic observation and finite element simulation analysis to determine the material's suitability. While these methods are relatively mature, they often suffer from drawbacks such as long testing cycles, high costs, reliance on a large number of physical samples, and sensitivity to experimental conditions, making it difficult to promptly feed the evaluation results into composition design and process optimization processes. Furthermore, many evaluation conclusions rely on the experience of materials experts for comprehensive judgment, which is highly subjective and hinders the achievement of a standardized, quantifiable, and traceable evaluation system. Summary of the Invention

[0003] To address the aforementioned problems, the purpose of this invention is to provide a method for intelligent evaluation of automotive steel.

[0004] A method for intelligent evaluation of automotive steel, comprising:

[0005] Step 1: Obtain images of automotive steel products;

[0006] Step 2: Preprocess the automotive steel product image to obtain the preprocessed automotive steel product image;

[0007] Step 3: Use a feature extraction network to extract features from the preprocessed automotive steel product image to obtain a feature image;

[0008] Step 4: Label the feature images to obtain training samples;

[0009] Step 5: Use the training samples to train a deep learning model to obtain an automotive steel product quality assessment model;

[0010] Step 6: Use the automotive steel product quality assessment model to conduct a preliminary assessment of the target automotive steel product, obtain the assessment results, and send the assessment results to the technical department via video conference.

[0011] Preferably, in step 2, the automotive steel product image is subjected to denoising, histogram equalization, and normalization to obtain a preprocessed automotive steel product image.

[0012] Preferably, the input image of the automotive steel product is fed into a convolutional neural network, and the initial features of the image are extracted through multiple convolutions to obtain a first feature map. The first feature map is then max-pooled to obtain a second feature map.

[0013] Preferably, in step 3, the image of the automotive steel product is input into the VGG16 model to obtain a detail feature map, and the detail feature map and the second feature map are stitched together to obtain a feature image.

[0014] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0015] This invention relates to an intelligent evaluation method for automotive steel. Compared with the prior art, this invention completes the evaluation of automotive steel through a deep learning model, avoiding subjective differences caused by manual visual inspection and experience judgment, and facilitating the establishment of stable quality judgment rules.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 The present invention provides a flowchart of an intelligent evaluation method for automotive steel. Detailed Implementation

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0020] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0022] Please see Figure 1 A method for intelligent evaluation of automotive steel, comprising:

[0023] Step 1: Obtain images of automotive steel products;

[0024] Step 2: Preprocess the automotive steel product image to obtain the preprocessed automotive steel product image;

[0025] In step 2, the automotive steel product image is subjected to denoising, histogram equalization, and normalization to obtain the preprocessed automotive steel product image.

[0026] Step 3: Use a feature extraction network to extract features from the preprocessed automotive steel product image to obtain a feature image;

[0027] In step 3, the input image of the automotive steel product is fed into a convolutional neural network. Initial features of the image are extracted through multiple convolutional layers to obtain a first feature map. The first feature map is then max-pooled to obtain a second feature map. The automotive steel product image is then input into a VGG16 model to obtain a detail feature map. The detail feature map and the second feature map are then concatenated to obtain a feature image.

[0028] Step 4: Label the feature images to obtain training samples;

[0029] Step 5: Use the training samples to train a deep learning model to obtain an automotive steel product quality assessment model;

[0030] Step 6: Use the automotive steel product quality assessment model to conduct a preliminary assessment of the target automotive steel product, obtain the assessment results, and send the assessment results to the technical department via video conference.

[0031] Product certification and on-site audits are conducted using video conferencing. During certification and factory audits, initial technical exchanges are moved online, with production, sales, R&D, and user departments participating simultaneously to jointly understand user needs. During product trial runs, rapid collaboration with user on-site engineers is established, using real-time video to understand product status and promptly connect with technical experts for quick solutions. This maximizes the efficiency of resolving product application issues, further reducing travel expenses incurred from repeated personnel trips and lowering user application technical support costs.

[0032] During the application technical support process, personnel from R&D, production, sales, and application will work together to provide technical support to users. Dedicated WeChat contact groups will be established for different users, and feedback from on-site representatives will be combined to understand user needs and product usage in real time, enhancing cooperation between different departments. Problem-solving progress will be shared in the WeChat groups in real time. The product application technology department will summarize the overall product application situation to establish user files, parts trial mold files, certification and testing files, stamping defect files, etc., summarizing product certification and trial mold patterns to improve product process performance.

[0033] This invention uses a deep learning model to evaluate automotive steel, avoiding subjective differences caused by manual visual inspection and experience-based judgment, and facilitating the establishment of stable quality judgment rules.

[0034] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent evaluation of automotive steel, characterized in that, include: Step 1: Obtain images of automotive steel products; Step 2: Preprocess the automotive steel product image to obtain the preprocessed automotive steel product image; Step 3: Use a feature extraction network to extract features from the preprocessed automotive steel product image to obtain a feature image; Step 4: Label the feature images to obtain training samples; Step 5: Use the training samples to train a deep learning model to obtain an automotive steel product quality assessment model; Step 6: Use the automotive steel product quality assessment model to conduct a preliminary assessment of the target automotive steel product, obtain the assessment results, and send the assessment results to the technical department via video conference.

2. The intelligent evaluation method for automotive steel according to claim 1, characterized in that, In step 2, the automotive steel product image is subjected to denoising, histogram equalization, and normalization to obtain the preprocessed automotive steel product image.

3. The intelligent evaluation method for automotive steel according to claim 2, characterized in that, In step 3, the input image of the automotive steel product is fed into the convolutional neural network. The initial features of the image are extracted through multiple convolutions to obtain the first feature map. The first feature map is then max-pooled to obtain the second feature map.

4. The intelligent evaluation method for automotive steel according to claim 3, characterized in that, In step 3, the image of the automotive steel product is input into the VGG16 model to obtain a detail feature map, and the detail feature map and the second feature map are stitched together to obtain a feature image.

Images