CN-122016997-A - Nondestructive testing method and medium for steel belt

CN122016997ACN 122016997 ACN122016997 ACN 122016997ACN-122016997-A

Abstract

The application discloses a nondestructive testing method and medium for a steel belt, wherein the method comprises the steps of inputting identification information and attribute information of the steel belt to be tested, scanning and detecting the steel belt to be tested to obtain testing data corresponding to a plurality of testing positions, analyzing the testing data to correspondingly obtain characteristic data of the steel belt to be tested, and outputting the characteristic data of the steel belt to be tested. The method can detect a plurality of different positions of the steel belt without damaging the steel belt, improves the detection comprehensiveness, is beneficial to improving the detection accuracy and the detection efficiency by constructing an analysis model between the detection data and the output characteristic data, directly correlates and outputs the detection data with the identification information of the steel belt, and improves the automation degree of industrial production detection.

Inventors

LUO KAN
CHEN YU
XU GE
WANG PING
XU WEILEI
XU QUAN
ZHAO JIAJUN
Yao Tenglong
WANG PENGCHENG
Liu Wuyu

Assignees

上海实达精密不锈钢有限公司
南京派光高速载运智慧感知研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20251113

Claims (10)

1. The nondestructive testing method for the steel belt is characterized by comprising the following steps of: firstly, inputting identification information and attribute information of a steel strip to be detected; Secondly, scanning and detecting the steel belt to be detected to obtain detection data corresponding to a plurality of detection positions; Thirdly, analyzing the detection data to correspondingly obtain the characteristic data of the steel belt to be detected; And step four, outputting the characteristic data of the steel belt to be detected.
2. The method according to claim 1, wherein in the first step, the identification information of the steel strip to be detected includes a brand number, a coil number, supplier information, and/or a smelting number of the steel strip to be detected, and the attribute information of the steel strip to be detected includes model information, specification information, material information, and/or usage information.
3. The method according to claim 1, wherein in the second step, the steel strip to be inspected is scanned and inspected, which comprises locally exciting the steel strip with an inspection probe to obtain electromagnetic characteristic signals, amplifying and filtering the signals, and transmitting the signals to an electromagnetic inspection device through a cable.
4. A steel strip nondestructive testing method according to claim 3, wherein in the second step, the detection data corresponding to the plurality of detection positions is obtained, including determining a predetermined detection position of the steel strip to be tested in advance based on the identification information and the attribute information of the steel strip to be tested.
5. The steel strip nondestructive testing method according to claim 1, wherein in the second step, in the case where the test data is large, the test data is also screened, the screening method comprising: the priority of the test data performance of different positions of the steel belt is as follows, the priority of the test data in the middle of the steel belt is higher than the priority of the test data at the tail of the steel belt, and the priority of the test data at the head of the steel belt is higher than the priority of the test data at the tail of the steel belt; or/and, the priority of the detection data obtained at the latest test time is greater than the priority of the detection data obtained at the past test time.
6. The nondestructive testing method of steel strip of claim 1 wherein in the third step, the method of analyzing the test data comprises a stepwise regression method or an artificial neural network method.
7. The method of claim 6, wherein for the stepwise regression method, comprising: Step 301, firstly, acquiring a plurality of detection data for detecting the steel strip; step S302, obtaining actual mechanical characteristic data of the steel strip through detection and test; step S303, combining the detection data serving as an input independent variable and the mechanical characteristic data serving as an output dependent variable to construct one or more regression calculation formulas; step S304, checking the significance of the regression coefficient in each regression calculation formula, deleting the independent variable with insignificant coefficient significance, recombining the independent variable, and returning to step S303 for iterative check; step S305, if the significance of the regression coefficient in the regression calculation formula accords with the inspection standard, the corresponding independent variable is reserved, the step S304 is returned to, and the significance of the regression coefficient of other independent variables which are not subjected to significance inspection is continuously inspected until all the independent variables are inspected; And step S306, finally obtaining the required optimal regression calculation formula.
8. The method of nondestructive testing of steel strip of claim 6, wherein for the neural network method, comprising: step S401, firstly, acquiring a plurality of detection data for detecting the steel strip; step S402, obtaining actual mechanical characteristic data of the steel strip through detection and test; step S403, constructing a neural network model, taking detection data as an independent variable input into the neural network model, and taking mechanical characteristic data as an independent variable output by the neural network model; step S404, initializing weight and threshold corresponding to each layer of the neural network model; Step S405, inputting the existing detection data and mechanical characteristic data as sample data into the neural network model, calculating an output layer error and an hidden layer error in the neural network model, and correcting weights and thresholds of the output layer error and the hidden layer until all sample data are trained; And step S406, judging whether the trained neural network model meets the training ending condition, if so, finishing training, and utilizing the neural network model to carry out practical application, and if not, returning to step S404 to reconstruct and train until the neural network model capable of carrying out practical application is obtained.
9. The method according to claim 1, wherein the output of the characteristic data of the steel strip to be inspected includes longitudinal and transverse tensile strength, longitudinal and transverse yield strength, longitudinal and transverse elongation after break, vickers hardness and/or confidence interval.
10. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method of any of claims 1-9.

Description

Nondestructive testing method and medium for steel belt Technical Field The application relates to the technical field of steel strip detection, in particular to a steel strip nondestructive detection method and medium. Background At present, the mechanical detection of the performance of thin plate steels such as strip steel, plate steel and the like by domestic iron and steel enterprises mainly comprises the step of carrying out tensile test in a physical and chemical laboratory to obtain performance parameters. The existing method has the following main defects: 1. The method comprises the steps of (1) obtaining physical and chemical measurement values after the production is basically completed, 2) obtaining incomplete data, namely only reflecting mechanical property data values of an initial part and an end part of a roll of strip steel, and not obtaining the quality condition of the whole roll of strip steel, 3, shearing waste, namely shearing a section of suspected unqualified strip steel when the production is stopped for a certain reason or the production is carried out at a low speed, wherein the shearing is unknown, the actual practice is that the shearing is as much as possible, and the waste is obviously caused, 4, the operation is carried out by people beside the machine for 24 hours, the working efficiency is low, and the method does not accord with the trend of unmanned and intelligent. Therefore, there is a need to improve the efficiency, comprehensiveness and accuracy of steel strip detection. Content of the application The application mainly solves the technical problems of providing a nondestructive testing method and medium for a steel belt, and solving the problems of improving the efficiency and the accuracy of the nondestructive testing of the steel belt in the prior art. In order to solve the technical problems, the technical scheme adopted by the application is to provide a nondestructive testing method for a steel belt, which comprises the steps of inputting identification information and attribute information of the steel belt to be tested, scanning and detecting the steel belt to be tested to obtain detection data corresponding to a plurality of detection positions, analyzing the detection data to correspondingly obtain characteristic data of the steel belt to be tested, and outputting the characteristic data of the steel belt to be tested. In some embodiments, in the first step, the identification information of the steel strip to be detected includes a brand number, a coil number, supplier information, and/or a smelting number of the steel strip to be detected, and the attribute information of the steel strip to be detected includes model information, specification information, material information, and/or usage information. In some embodiments, in the second step, the steel strip to be detected is scanned and detected, including locally exciting the steel strip with a detection probe to obtain an electromagnetic characteristic signal, amplifying and filtering the signal, and transmitting the signal to an electromagnetic detection device through a cable. In some embodiments, in the second step, obtaining detection data corresponding to a plurality of detection positions includes determining a predetermined detection position of the steel strip to be detected in advance according to the identification information and the attribute information of the steel strip to be detected. In some embodiments, in the second step, the detection data is also screened if the detection data is more, wherein the screening method comprises the steps of prioritizing the performance of the detection data at different positions of the steel strip as follows, namely, prioritizing the detection data in the middle of the steel strip, prioritizing the detection data at the tail of the steel strip, prioritizing the detection data at the head of the steel strip, or/and, prioritizing the detection data obtained at the latest test time, prioritizing the detection data obtained at the past test time In some embodiments, in the third step, the method of analyzing the detection data comprises a stepwise regression method or an artificial neural network method. In some embodiments, for the stepwise regression method, comprising: Step 301, firstly, acquiring a plurality of detection data for detecting the steel strip; step S302, obtaining actual mechanical characteristic data of the steel strip through detection and test; step S303, combining the detection data serving as an input independent variable and the mechanical characteristic data serving as an output dependent variable to construct one or more regression calculation formulas; step S304, checking the significance of the regression coefficient in each regression calculation formula, deleting the independent variable with insignificant coefficient significance, recombining the independent variable, and returning to step S303 for iterative check; step S305