Rod and wire production line loop position abnormity detection method and device
Through the deep learning convolutional neural network model, the abnormal position of the rod wire production line is automatically detected, which solves the problems of low manual detection efficiency and high accident risk, and improves the operating efficiency and safety of the production line.
Patent Information
- Application Number
- CN202311743627.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-18
- Publication Date
- 2025-06-20
AI Technical Summary
In the prior art, the detection of position abnormality of the rod wire production line is mainly based on manual observation, resulting in low management efficiency and easy omissions, which in turn increases the risk of production accidents.
A detection method based on the deep learning convolutional neural network model is adopted. By obtaining the sensor data of the production line, drawing a line chart, and inputting it into the pre-trained detection model. The detection model is trained based on historical normal data and abnormal data, and automatically detects the abnormal position of the living sleeve and sends an alarm signal to the user.
It realizes automatic and accurate detection of abnormal position of the movable sleeve, improves the operating efficiency and operating quality of the production line, and reduces the risk of production accidents.
Smart Images
Figure CN120169840A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image processing and the field of bar and wire production technology, and particularly relates to a method and device for detecting abnormal loop positions in a bar and wire production line. Background Art
[0002] Rolled materials, that is, steel materials after rolling, include bars, wires and sheets. Among them, wires play a very important role in the current national economic development, ranking first in the world both in terms of the number of rolling mills and output, and the output is still increasing at a relatively fast rate. Currently, the output of bars and wires accounts for 48% - 50% of the total steel output, mainly used in the construction industry, and secondly in industries such as machinery, metallurgy, electric power, and coal.
[0003] In the transportation operation scenario of bars and wires, bars and wires are rolled by a continuous rolling mill. The principle that enables stable rolling is equal second flow, and tension-free rolling is achieved between adjacent stands. However, in the actual rolling process, due to factors such as steel temperature fluctuations, pass wear, and speed control, the tension between stands is actually dynamically changing. For the medium and finishing rolling areas, loops are set to eliminate tension to achieve tension-free rolling. The principle is to indirectly control the tension by controlling the loop to maintain it at a set height.
[0004] If there is a certain tension between stands, then the loop will not be able to reach the set height. At this time, the tension can be eliminated by adjusting the speed difference between stands to make the loop reach the set height. However, in actual production, when the tension between stands is too large, it will cause the loop to be unable to reach the set height or have a slow response to reaching the set height; when the steel storage between stands is too large, it will cause the loop to exceed the set height, and at this time, there will be a risk of steel piling up. Therefore, timely detecting abnormal situations of the loop position and timely adjusting it is the key to ensuring tension-free and stable rolling in the loop area.
[0005] Currently, the detection methods for whether the loop position is abnormal mainly rely on the on-site manual observation method, including observing the data of the loop position sensor and observing the production line status to understand the situation. However, no matter which of the above methods, it will make the on-site management efficiency low, and requires a dedicated person to track and observe, which is extremely prone to omissions, and thus leads to production accidents. Summary of the Invention
[0006] Aiming at the problems in the prior art, the present application provides a method and device for detecting abnormal loop positions in a bar and wire production line, which can automatically and accurately detect abnormal loop positions, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents.
[0007] To solve at least one of the above problems, the present application provides the following technical solutions:
[0008] According to the first aspect of the embodiments of the present application, the present application provides a method for detecting abnormal loop positions in a bar and wire production line, including:
[0009] In response to the sensor data of the production line loop starting to fluctuate, obtain the sensor data of the production line loop until the sensor data returns to zero;
[0010] Plot the obtained sensor data as a line chart and input the line chart into a pre-trained detection model, where the detection model is trained based on the historical normal data and abnormal data of the sensor;
[0011] In response to the output result of the detection model being abnormal data, send an alarm signal to the user.
[0012] According to any implementation manner of the present application, the training method of the detection model includes:
[0013] Obtain the historical sensor data of the production line loop, preprocess the historical sensor data, and intercept the interval of the historical sensor data during the period when the bar and wire pass through the production line;
[0014] Plot the interval as a line chart and perform binary classification annotation on the line chart to obtain the historical normal data and historical abnormal data of the sensor data;
[0015] Enhance the historical abnormal data until the sample quantity ratio of the historical abnormal data to the historical normal data meets a preset ratio;
[0016] Construct a deep convolutional model, input and train the historical normal data and the historical abnormal data until the model converges to obtain the detection model.
[0017] According to any implementation manner of the present application, the historical sensor data includes the standard value, actual measurement value, and compensation value of the loop position during the bar and wire production process.
[0018] According to any implementation manner of the present application, the preprocessing of the historical sensor data to intercept the interval of the historical sensor data during the period when the bar and wire pass through the production line includes:
[0019] In response to the historical sensor data starting to fluctuate, record the historical sensor data starting from a preset time before the fluctuation;
[0020] In response to the historical sensor data returning to zero, stop recording the historical sensor data after a preset time after the zeroing, to obtain multiple intervals of the historical sensor data, where each interval contains the standard value, actual measurement value, and compensation value of the historical sensor data during the period from the start of the fluctuation of the sensor data of the production line loop to the return of the sensor data to zero.
[0021] According to any embodiment of the present application, before inputting the historical abnormal data, it further includes:
[0022] Adding a perturbation amount to the actual measurement value in the historical abnormal data so that the actual measurement value has a deviation value within a preset range from the original value.
[0023] According to any embodiment of the present application, the step of inputting and training the historical normal data and the historical abnormal data until the model converges to obtain the detection model includes:
[0024] Inputting a preset proportion of the historical normal data and the historical abnormal data in all the historical sensor data into the deep convolutional model for training until the model converges;
[0025] Inputting the remaining proportion of the historical normal data and the historical abnormal data into the converged model to verify the converged model, and determining the detection model when the verification is passed.
[0026] According to the second aspect of the embodiments of the present application, the present application provides a loop position abnormal detection device for a bar and wire production line, including:
[0027] A data acquisition module, configured to: in response to the sensor data of the production line loop starting to fluctuate, acquire the sensor data of the production line loop until the sensor data returns to zero;
[0028] An abnormal detection module, configured to: plot the acquired sensor data as a line graph and input the line graph into a pre-trained detection model, where the detection model is obtained by training the historical normal data and abnormal data of the sensor according to the training device of the detection model;
[0029] An abnormal alarm module, configured to: in response to the output result of the detection model being abnormal data, send an alarm signal to the user.
[0030] According to any embodiment of the present application, the training device of the detection model includes:
[0031] A historical data acquisition module, configured to: acquire the historical sensor data of the production line loop and preprocess the historical sensor data, and intercept the interval of the historical sensor data during the passing of the bar and wire through the production line;
[0032] An image classification module, configured to: plot the interval as a line graph and perform binary classification annotation on the line graph to obtain the historical normal data and historical abnormal data of the sensor data;
[0033] An abnormal data enhancement module, configured to: enhance the historical abnormal data until the sample quantity ratio of the historical abnormal data to the historical normal data meets a preset ratio;
[0034] A repeated training module, configured to: construct a deep convolutional model, input and train the historical normal data and the historical abnormal data until the model converges, and obtain the detection model.
[0035] According to any embodiment of the present application, the historical sensor data includes the standard value, the actual measured value, and the compensation value of the loop position in the bar and wire production process.
[0036] According to any embodiment of the present application, the historical data acquisition module includes:
[0037] A recording start unit, configured to: in response to the start of fluctuation of the historical sensor data, record the historical sensor data starting from a preset time before the fluctuation;
[0038] A recording stop unit, configured to: in response to the historical sensor data returning to zero, stop recording the historical sensor data after a preset time after the return to zero, and obtain multiple intervals of the historical sensor data, where each interval includes the standard value, the actual measured value, and the compensation value of the historical sensor data during the period from the start of fluctuation of the sensor data of the production line loop to the return of the sensor data to zero.
[0039] According to any embodiment of the present application, before the repeated training module inputs the historical abnormal data, it further includes a deviation construction module, configured to:
[0040] Add a perturbation amount to the actual measured value in the historical abnormal data, so that the actual measured value has a deviation value within a preset range from the original value.
[0041] According to any embodiment of the present application, the repeated training module includes:
[0042] A partial training unit, configured to: input a preset proportion of the historical normal data and the historical abnormal data in all the historical sensor data into the deep convolutional model and perform training until the model converges;
[0043] A partial verification unit, configured to: input the remaining proportion of the historical normal data and the historical abnormal data into the converged model to verify the converged model, and determine the detection model when the verification is passed.
[0044] According to the third aspect of the embodiments of the present application, the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the method for detecting abnormal loop position in the bar and wire production line are implemented.
[0045] According to the fourth aspect of the embodiments of the present application, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for detecting abnormal loop position in the bar and wire production line are implemented.
[0046] According to the fifth aspect of the embodiments of the present application, the present application provides a computer program product, including a computer program / instructions. When the computer program / instructions are executed by a processor, the steps of the method for detecting abnormal loop position in the bar and wire production line are implemented.
[0047] As can be seen from the above technical solutions, the present application provides a method and device for detecting abnormal loop position in the bar and wire production line. In response to the sensor data of the production line loop starting to fluctuate, the sensor data of the production line loop is acquired until the sensor data returns to zero; the acquired sensor data is plotted as a line graph, and the line graph is input into a pre-trained detection model, which is trained based on the historical normal data and abnormal data of the sensor; in response to the output result of the detection model being abnormal data, an alarm signal is sent to the user, which can automatically and accurately detect abnormal loop position, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on the above drawings without creative efforts.
[0049] Figure 1 It is a schematic diagram of the loop position data in the bar and wire production line in the embodiments of the present application;
[0050] Figure 2 It is one of the flow schematic diagrams of the method for detecting abnormal loop position in the bar and wire production line in the embodiments of the present application;
[0051] Figure 3 It is another flow schematic diagram of the method for detecting abnormal loop position in the bar and wire production line in the embodiments of the present application;
[0052] Figure 4It is the third flowchart of the method for detecting abnormal loop positions in the bar and wire production line in the embodiments of the present application;
[0053] Figure 5 It is the visualization diagram of the loop data range in the embodiments of the present application;
[0054] Figure 6 It is the normal data diagram of the loop data range in the embodiments of the present application;
[0055] Figure 7 It is one of the abnormal data diagrams of the loop data range in the embodiments of the present application;
[0056] Figure 8 It is the second abnormal data diagram of the loop data range in the embodiments of the present application;
[0057] Figure 9 It is the third abnormal data diagram of the loop data range in the embodiments of the present application;
[0058] Figure 10 It is the fourth flowchart of the method for detecting abnormal loop positions in the bar and wire production line in the embodiments of the present application;
[0059] Figure 11 It is the first structural diagram of the device for detecting abnormal loop positions in the bar and wire production line in the embodiments of the present application;
[0060] Figure 12 It is the second structural diagram of the device for detecting abnormal loop positions in the bar and wire production line in the embodiments of the present application;
[0061] Figure 13 It is the third structural diagram of the device for detecting abnormal loop positions in the bar and wire production line in the embodiments of the present application;
[0062] Figure 14 It is the fourth structural diagram of the device for detecting abnormal loop positions in the bar and wire production line in the embodiments of the present application;
[0063] Figure 15 It is the structural diagram of the electronic device in the embodiments of the present application. Detailed implementation manners
[0064] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.
[0065] In the technical solution of this application, the acquisition, storage, use, processing, etc. of data all comply with the relevant provisions of national laws and regulations.
[0066] Currently, the detection method for whether the loop position is abnormal mainly relies on the way of on-site manual observation.
[0067] There are two kinds of information that can be observed manually. One is to indirectly understand the loop state through the sensor data of the loop position. Since the data is not conducive to intuitive display, the data is plotted into a Figure 1 chart as shown.
[0068] In Figure 1 , when the production line runs empty, the value of the vertical coordinate is 0, such as the flat area shown in the above figure; when there are bar and wire materials passing through the production line, the line in the image will fluctuate, corresponding to the periodic undulation and fluctuation areas in the figure. When the product completely passes through the production line, the vertical coordinate value of the line will quickly return to zero, and the production line enters the empty running stage.
[0069] The other way is to directly observe the production line state through human eyes to understand the situation. This method is relatively primitive. Usually, the abnormal signal is not significant. Only when obvious steel piling or obvious deviation of the loop position occurs can a relatively clear abnormal handling be made, but in most cases, it is too late.
[0070] No matter which method is used, it will make the on-site management efficiency low, and special personnel are required to track and observe, which is very easy to cause omissions, and then lead to production accidents.
[0071] Considering the above problems, this application provides a method and device for detecting abnormal loop position in a bar and wire production line. Through a detection algorithm based on a deep learning convolutional neural network model, it can automatically and accurately detect abnormal loop position without deploying image acquisition equipment, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents.
[0072] In order to be able to automatically and accurately detect abnormal loop position, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents, this application provides an embodiment of a method for detecting abnormal loop position in a bar and wire production line. Refer to Figure 2 , the method for detecting abnormal loop position in the bar and wire production line specifically includes the following content:
[0073] Step S101: In response to the sensor data of the production line loop starting to fluctuate, acquire the sensor data of the production line loop until the sensor data returns to zero.
[0074] Collect the real-time data of the loop sensor. When the loop data is 0, no record is made; when the data fluctuates, start recording until the data returns to zero again, and obtain the data in this section.
[0075] Step S102: Plot the acquired sensor data as a line chart and input the line chart into a pre-trained detection model, which is trained based on the historical normal data and abnormal data of the sensor.
[0076] The acquired sensor data will be processed and converted into the form of a line chart. The line chart is a visual representation that more comprehensively characterizes the loop position state at any time point during the entire rolling process. Exemplarily, it may include the standard value, actual measured value, and compensation value of the loop position. The line chart will be used as input and fed into a pre-trained detection model. The detection model is deployed at the production site after training to collect and graphically process the sensor data received in real time, and at the same time output the loop abnormal state.
[0077] Step S103: In response to the output result of the detection model being abnormal data, send an alarm signal to the user.
[0078] Once the detection model outputs abnormal data, the system will immediately respond and send an alarm signal to the user. This can be achieved in various ways, such as sound alarm, light alarm, or text message alarm, so as to promptly notify relevant personnel and take necessary measures.
[0079] As can be seen from the above description, the method for detecting abnormal loop position in the bar and wire rod production line provided by the embodiments of the present application can automatically and accurately detect abnormal loop position without deploying image acquisition equipment through a detection algorithm based on a deep learning convolutional neural network model, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents.
[0080] In an embodiment of the method for detecting abnormal loop position in the bar and wire rod production line of the present application, refer to Figure 3 , the training method of the detection model specifically includes the following content:
[0081] Step S201: Acquire the historical sensor data of the production line loop, and preprocess the historical sensor data to intercept the interval of the historical sensor data during the period when the bar and wire rod passes through the production line.
[0082] First, acquire the historical sensor data of the production line loop, and then preprocess the above historical sensor data to intercept the data interval during the period when the bar and wire rod passes through the production line. The interval contains the loop sensor data when there is a product passing through the production line.
[0083] In an optional embodiment, the historical sensor data includes the standard value, actual measured value, and compensation value of the loop position during the production process of the bar and wire rod.
[0084] Among them, the standard value of the loop position refers to the ideal position of the loop preset by the system for each time point or interval during the production of bar and wire products, which is used to evaluate the actual measured value and judge whether there is an abnormal loop position.
[0085] The actual measured value refers to the loop position data actually measured by the sensor. This is real-time data obtained during actual production and reflects the actual position of the bar and wire on the loop.
[0086] The compensation value is a correction term introduced to correct or adjust the actual measured value. In actual production, due to various factors (such as machine wear, temperature changes, etc.), there will be some deviations in the actual measured value. The compensation value can make the data more accurately reflect the true position of the loop by adding or subtracting a certain correction value to the actual measured value.
[0087] In an embodiment of the method for detecting abnormal loop position in the bar and wire production line of the present application, refer to Figure 4 , it may specifically include the following content:
[0088] Step S201A: In response to the start of fluctuation of the historical sensor data, record the historical sensor data starting from a preset time before the fluctuation.
[0089] Step S201B: In response to the historical sensor data returning to zero, stop recording the historical sensor data after a preset time after the return to zero, and obtain multiple intervals of the historical sensor data, where each interval includes the position standard value, the actual measured value, and the compensation value of the historical sensor data during the period from the start of fluctuation of the sensor data of the production line loop to the return of the sensor data to zero.
[0090] Among them, the data part when a product passes through the production line is extracted from the historical data. The specific method is to start recording and intercepting data at the observed time t, that is, when the sensor data ob > 0, until the data returns to zero again, denoted as time t + k; the recorded data includes 3 columns, corresponding to the standard value, the actual measured value, and the compensation value respectively. It can be understood that since most intervals are 0, representing the empty section of the production line, which has no analytical significance, there is no need to pay attention.
[0091] Secondly, draw a line chart for the data in the interval [t - 2, t + k + 2], and an example is shown as Figure 5 shown.
[0092] In the figure, the horizontal line represents the standard value of the loop position; the section line (dark color) roughly flush with the standard line corresponds to the actual measured value of the loop position; the lowest line (light color) represents the compensation value.
[0093] Step S202: Plot the interval as a line chart and perform binary classification annotation on the line chart to obtain the historical normal data and historical abnormal data of the sensor data.
[0094] Perform pictorial processing on all historical data required this time, plot the intercepted data interval as a line chart, and then perform binary classification annotation on the line chart, classifying it into historical normal data and historical abnormal data to form a training dataset, which includes normal and abnormal loop position states.
[0095] For Figure 4 the data in, perform binary classification annotation. The result classified as abnormal is labeled as "1", and the result classified as normal is labeled as "0". The classification results are as shown in Figure 6 and Figure 7 shown.
[0096] Among them, Figure 5 represents the historical normal data of the sensor data, Figure 6 represents the historical abnormal data of the sensor data.
[0097] Step S203: Augment the historical abnormal data until the sample quantity ratio of the historical abnormal data to the historical normal data meets a preset ratio.
[0098] Augment the historical abnormal data to ensure that the quantity ratio of normal and abnormal samples meets the preset ratio. The augmentation process includes copying existing abnormal pictures or adding random perturbations to the actual measured values to generate more and diverse abnormal data.
[0099] Specifically, since abnormal data is relatively rare in actual production and the distribution of the sample training data volume is very unbalanced this time, it is necessary to perform data augmentation on the abnormal samples so that the quantity ratio of these two types of samples is close to 1:1 (exemplary).
[0100] Especially, there are many types of abnormal data, such as shown in Figure 8 and Figure 9 shown:
[0101] In Figure 7 , this type of abnormality is that the loop height does not reach the standard value and there is a rapid upward surge at the tail;
[0102] In Figure 8 , this type of abnormality is that the loop height fails to reach the standard value.
[0103] Step S204: Construct a deep convolutional model, input and train the historical normal data and the historical abnormal data until the model converges to obtain the detection model.
[0104] By constructing a deep convolutional neural network model, it is used to detect abnormalities in the loop position. This model is trained with historical normal data and enhanced historical abnormal data until the model converges. To obtain a detection model for detecting abnormalities in the loop position of the bar and wire production line for subsequent real-time anomaly detection.
[0105] Exemplarily, construct a deep convolutional neural network model Alexnet, and fully train and test the neural network based on the above image data until the model fully converges.
[0106] In an alternative embodiment, before inputting the historical abnormal data, it further includes:
[0107] Add a perturbation amount to the actual measurement value in the historical abnormal data so that the actual measurement value has a deviation value within a preset range from the original value.
[0108] Exemplarily, there are two ways to generate abnormal data. One is to copy existing abnormal pictures and repeat training; the other is to add a random perturbation amount to the actual measurement value (the part greater than 0) of the loop position, so that it has a certain deviation on the basis of the original data, but try to keep the original line form.
[0109] In an embodiment of the method for detecting abnormalities in the loop position of the bar and wire production line of the present application, refer to Figure 10 , inputting and training the historical normal data and the historical abnormal data until the model converges to obtain the detection model, which may specifically include the following content:
[0110] Step S204A: Input a preset proportion of the historical normal data and the historical abnormal data in all the historical sensor data into the deep convolutional model and train until the model converges;
[0111] Step S204B: Input the remaining proportion of the historical normal data and the historical abnormal data into the converged model to verify the converged model, and determine the detection model when the verification passes.
[0112] First, randomly divide all the obtained data sets according to a ratio of 7:3 (exemplarily), where 70% of the data is used for training and 30% of the data is used for verification. This division is to fully utilize the data for learning during model training and to evaluate the generalization ability of the model during the verification stage.
[0113] In the verification stage, the remaining 30% of the data is used. The above data can be arranged in ascending order, that is, the samples ranked in the top 30% of the data set are selected for verification to ensure that some relatively difficult and near-boundary samples are included in the verification set to more comprehensively evaluate the performance of the model.
[0114] To further illustrate this solution, the present application also provides a specific application example of a method for detecting abnormal loop positions in a bar and wire rod production line by using the above-mentioned abnormal loop position detection device for a bar and wire rod production line, which specifically includes the following contents:
[0115] An abnormal loop position detection algorithm for a bar and wire rod production line based on a deep convolutional neural network model, including: obtaining the loop position sensor data in an actual factory, intercepting the obtained data in intervals and plotting a line graph to achieve the purpose of visualizing the observed data. This picture more completely represents the loop position state at any time point during the entire rolling process.
[0116] The establishment of the detection model includes: based on the line graph picture drawn from the loop position data, annotating and expanding the data set; randomly dividing the obtained data set into a training set and a validation set according to a ratio of 7:3, and sending the training set into the detection model Alexnet for iterative training; and validating with 30% of the data set in ascending order.
[0117] The technical application proposed in the present application has the following beneficial effects:
[0118] This technical application is different from traditional abnormal detection applications. It is a detection algorithm based on a deep learning convolutional neural network model, and it can obtain intuitive visual picture data without deploying image acquisition devices such as cameras.
[0119] An abnormal loop position detection application for a bar and wire rod production line based on a deep convolutional neural network provided by the present application solves the technical problems in actual production such as low efficiency, low accuracy, processing lag, and high detection cost of the traditional manual detection method in production.
[0120] To further illustrate this solution, the present application also provides a specific application example of a method for detecting abnormal loop positions in a bar and wire rod production line by using the above-mentioned abnormal loop position detection device for a bar and wire rod production line, which specifically includes the following contents:
[0121] An abnormal loop position detection method for a bar and wire rod in a steel mill based on a deep convolutional neural network, the main detection steps of which are:
[0122] (1) Collect the loop position sensor data during the production of bars and wire rods in a steel mill, and perform picture processing on it by plotting a line graph. The information represented by this picture is more comprehensive and intuitive, and is perfectly suitable for the current deep convolutional neural network algorithm;
[0123] (2) Mark the abnormal values of the collected picture data to form an original training data set;
[0124] (3) Process the abnormal data by adding random perturbation amounts to increase the abnormal data samples and achieve the effect of data augmentation;
[0125] (4) Based on the labeled image dataset, input it into the Alexnet convolutional neural network for training to obtain the final abnormal diagnosis model;
[0126] (5) After being verified by the test set and on-site actual situation, the recognition accuracy (mAP) of this model reaches more than 99%.
[0127] (6) For the data processing process described in step (1), the specifically collected data includes three categories, namely the loop live position, the loop position standard value, and the adjustment compensation value;
[0128] (7) For what is described in step (1), the content of the line chart includes the three categories of data listed in (6), namely the loop live position, the loop position standard value, and the adjustment compensation value. Finally, the above three line segments are included in the figure respectively;
[0129] (8) For what is described in step (2), the specific abnormal classifications include multiple types, specifically including: the position height not meeting the standard, the position height exceeding the standard, the abnormal high rise at the tail, the position control disorder, and so on.
[0130] In order to be able to automatically and accurately detect the abnormal loop position, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents, this application provides an embodiment of a bar and wire product line loop position abnormal detection device for implementing all or part of the content of the bar and wire product line loop position abnormal detection method. See Figure 11 , the bar and wire product line loop position abnormal detection device specifically includes the following content:
[0131] The data acquisition module 1101 is used for: in response to the sensor data of the production line loop starting to fluctuate, acquiring the sensor data of the production line loop until the sensor data returns to zero;
[0132] The abnormal detection module 1102 is used for: plotting the acquired sensor data as a line chart and inputting the line chart into a pre-trained detection model, and the detection model is trained based on the historical normal data and abnormal data of the sensor by the training device of the detection model;
[0133] The abnormal alarm module 1103 is used for: in response to the output result of the detection model being abnormal data, sending an alarm signal to the user.
[0134] According to any implementation manner of this application, see Figure 12 , the training device of the detection model includes:
[0135] The historical data acquisition module 2201 is configured to: obtain the historical sensor data of the production line loop, preprocess the historical sensor data, and intercept the interval of the historical sensor data during the period when the bar and wire pass through the production line;
[0136] The image classification module 2202 is configured to: plot the interval as a line chart, and perform binary classification annotation on the line chart to obtain the historical normal data and historical abnormal data of the sensor data;
[0137] The abnormal data enhancement module 2203 is configured to: enhance the historical abnormal data until the sample quantity ratio of the historical abnormal data to the historical normal data meets a preset ratio;
[0138] The repeated training module 2204 is configured to: construct a deep convolutional model, input and train the historical normal data and the historical abnormal data until the model converges to obtain the detection model.
[0139] According to any embodiment of the present application, the historical sensor data includes the standard value, actual measurement value, and compensation value of the loop position during the production process of the bar and wire.
[0140] According to any embodiment of the present application, referring to Figure 13 , the historical data acquisition module includes:
[0141] The recording start unit 2201A is configured to: in response to the start of fluctuation of the historical sensor data, record the historical sensor data starting from a preset time before the fluctuation;
[0142] The recording stop unit 2201B is configured to: in response to the historical sensor data returning to zero, stop recording the historical sensor data after a preset time after the return to zero, and obtain multiple intervals of the historical sensor data, where each interval includes the standard value, actual measurement value, and compensation value of the historical sensor data during the period from the start of fluctuation of the sensor data of the production line loop to the return of the sensor data to zero.
[0143] According to any embodiment of the present application, before the repeated training module inputs the historical abnormal data, it further includes a deviation construction module configured to:
[0144] Add a perturbation amount to the actual measurement value in the historical abnormal data so that the actual measurement value has a deviation value within a preset range from the original value.
[0145] According to any embodiment of the present application, referring to Figure 14 , the repeated training module includes:
[0146] A partial training unit 2204A is configured to: input a preset proportion of the historical normal data and the historical abnormal data in all the historical sensor data into the deep convolutional model for training until the model converges;
[0147] A partial verification unit 2204B is configured to: input the remaining proportion of the historical normal data and the historical abnormal data into the converged model to verify the converged model, and determine the detection model when the verification is passed.
[0148] As can be seen from the above description, the loop position abnormal detection device for the bar and wire rod production line provided by the embodiment of the present application can automatically and accurately detect the loop position abnormality through the detection algorithm based on the deep learning convolutional neural network model without deploying an image acquisition device, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents.
[0149] At the hardware level, in order to automatically and accurately detect the loop position abnormality, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents, the embodiment of the present application provides an electronic device for implementing all or part of the content in the method for detecting the loop position abnormality of the bar and wire rod production line. The electronic device specifically includes the following content:
[0150] A processor, a memory, a communication interface, and a bus; wherein, the processor, the memory, and the communication interface complete communication with each other through the bus; the communication interface is used to implement information transmission between the loop position abnormal detection device for the bar and wire rod production line and related devices such as the core business system, the user terminal, and the relevant database. The logic controller can be a desktop computer, a tablet computer, a mobile terminal, etc., and this embodiment is not limited thereto. In this embodiment, the logic controller can be implemented with reference to the embodiment of the method for detecting the loop position abnormality of the bar and wire rod production line and the embodiment of the loop position abnormal detection device for the bar and wire rod production line, and the content thereof is incorporated herein, and the repeated parts will not be described again.
[0151] It can be understood that the user terminal may include a smart phone, a tablet electronic device, a network set-top box, a portable computer, a desktop computer, a personal digital assistant (PDA), a vehicle-mounted device, a smart wearable device, etc. Among them, the smart wearable device may include smart glasses, a smart watch, a smart bracelet, etc.
[0152] In practical applications, part of the method for detecting abnormal loop positions in bar and wire production lines can be executed on the side of the electronic device as described above, or all operations can be completed in the client device. Specifically, it can be selected according to the processing capabilities of the client device and the limitations of the user's usage scenario, etc. This application does not make any limitations in this regard. If all operations are completed in the client device, the client device may further include a processor.
[0153] The above-mentioned client device can have a communication module (i.e., communication unit), which can communicate with a remote server to achieve data transmission with the server. The server can include a server on the side of the task scheduling center. In other implementation scenarios, it can also include a server of an intermediate platform, such as a server of a third-party server platform that has a communication link with the task scheduling center server. The server can include a single computer device, or a server cluster composed of multiple servers, or a server structure of a distributed device.
[0154] Figure 15 It is a schematic block diagram of the system composition of the electronic device 9600 according to an embodiment of this application. As Figure 15 shown, the electronic device 9600 can include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It should be noted that this Figure 15 is exemplary; other types of structures can also be used to supplement or replace this structure to achieve telecommunication functions or other functions.
[0155] In one embodiment, the function of the method for detecting abnormal loop positions in bar and wire production lines can be integrated into the central processing unit 9100. Among them, the central processing unit 9100 can be configured to perform the following controls:
[0156] Step S101: In response to the sensor data of the production line loop starting to fluctuate, obtain the sensor data of the production line loop until the sensor data returns to zero;
[0157] Step S102: Plot the obtained sensor data as a line chart and input the line chart into a pre-trained detection model, which is trained based on the historical normal data and abnormal data of the sensor;
[0158] Step S103: In response to the output result of the detection model being abnormal data, send an alarm signal to the user.
[0159] As can be seen from the above description, the electronic device provided by the embodiments of the present application can automatically and accurately detect the abnormality of the loop position without deploying an image acquisition device through a detection algorithm based on a deep learning convolutional neural network model, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents.
[0160] In another embodiment, the loop position abnormality detection device of the bar and wire production line can be separately configured from the central processor 9100. For example, the loop position abnormality detection device of the bar and wire production line can be configured as a chip connected to the central processor 9100, and the function of the loop position abnormality detection method of the bar and wire production line can be realized through the control of the central processor.
[0161] As Figure 15 shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It should be noted that the electronic device 9600 does not necessarily have to include Figure 15 all the components shown in Figure 15 ; in addition, the electronic device 9600 may further include
[0162] components not shown in Figure 15 ; reference may be made to the prior art.
[0163] Among them, the memory 9140 may be, for example, one or more of a buffer, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory, or other suitable devices. The above information related to failures can be stored, and in addition, programs for executing relevant information can also be stored. And the central processor 9100 can execute the program stored in the memory 9140 to implement information storage or processing, etc.
[0164] The input unit 9120 provides input to the central processor 9100. The input unit 9120 is, for example, a key or a touch input device. The power supply 9170 is used to supply power to the electronic device 9600. The display 9160 is used to display display objects such as images and texts. The display may be, for example, an LCD display, but is not limited thereto.
[0165] The memory 9140 can be a solid-state memory, such as a read-only memory (ROM), a random access memory (RAM), a SIM card, etc. It can also be a memory that stores information even when the power is off, can be selectively erased and has more data. Examples of such a memory are sometimes referred to as EPROMs, etc. The memory 9140 can also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 can include an application / function storage unit 9142 for storing application programs and function programs or the processes for operating the electronic device 9600 by the central processing unit 9100.
[0166] The memory 9140 can also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 can include various drivers of the electronic device for communication functions and / or for performing other functions of the electronic device (such as a messaging application, an address book application, etc.).
[0167] The communication module 9110 is a transmitter / receiver 9110 that transmits and receives signals via the antenna 9111. The communication module (transmitter / receiver) 9110 is coupled to the central processing unit 9100 to provide input signals and receive output signals, which can be the same as in the case of a conventional mobile communication terminal.
[0168] Based on different communication technologies, multiple communication modules 9110 can be provided in the same electronic device, such as a cellular network module, a Bluetooth module, and / or a wireless local area network module, etc. The communication module (transmitter / receiver) 9110 is also coupled to the speaker 9131 and the microphone 9132 via the audio processor 9130 to provide an audio output via the speaker 9131 and receive an audio input from the microphone 9132, so as to implement the usual telecommunication functions. The audio processor 9130 can include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor 9130 is also coupled to the central processing unit 9100, so that recording can be performed on the local machine through the microphone 9132 and the sound stored on the local machine can be played through the speaker 9131.
[0169] An embodiment of the present application also provides a computer-readable storage medium capable of implementing all steps of the method for detecting abnormal loop positions in a bar and wire production line where the execution entity in the above embodiment is a server or a client. A computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, all steps of the method for detecting abnormal loop positions in a bar and wire production line where the execution entity in the above embodiment is a server or a client are implemented. For example, when the processor executes the computer program, the following steps are implemented:
[0170] Step S101: In response to the sensor data of the production line loop starting to fluctuate, obtain the sensor data of the production line loop until the sensor data returns to zero;
[0171] Step S102: Plot the obtained sensor data as a line graph and input the line graph into a pre-trained detection model, where the detection model is trained based on the historical normal data and abnormal data of the sensor;
[0172] Step S103: In response to the output result of the detection model being abnormal data, send an alarm signal to the user.
[0173] As can be seen from the above description, the computer-readable storage medium provided by the embodiment of the present application automatically and accurately detects abnormal loop positions without deploying image acquisition devices through a detection algorithm based on a deep learning convolutional neural network model, thereby improving the operation efficiency and operation quality of the production line and reducing the risk of production accidents.
[0174] An embodiment of the present application also provides a computer program product capable of implementing all steps of the method for detecting abnormal loop positions in a bar and wire production line where the execution entity in the above embodiment is a server or a client. When the computer program / instructions are executed by a processor, the steps of the method for detecting abnormal loop positions in the bar and wire production line are implemented. For example, the computer program / instructions implement the following steps:
[0175] Step S101: In response to the sensor data of the production line loop starting to fluctuate, obtain the sensor data of the production line loop until the sensor data returns to zero;
[0176] Step S102: Plot the obtained sensor data as a line graph and input the line graph into a pre-trained detection model, where the detection model is trained based on the historical normal data and abnormal data of the sensor;
[0177] Step S103: In response to the output result of the detection model being abnormal data, send an alarm signal to the user.
[0178] As can be seen from the above description, the computer program product provided by the embodiments of the present application can automatically and accurately detect the abnormality of the loop position through a detection algorithm based on a deep learning convolutional neural network model, without the need to deploy image acquisition devices, thereby improving the operation efficiency and quality of the production line and reducing the risk of production accidents.
[0179] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, apparatus, or computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.
[0180] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or a device for implementing the functions specified in one block or multiple blocks.
[0181] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or a device for implementing the functions specified in one block or multiple blocks.
[0182] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or a device for implementing the functions specified in one block or multiple blocks.
[0183] In the present invention, specific embodiments are used to illustrate the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.
Claims
1. A method for detecting abnormal loop position in a bar and wire production line, characterized in that, The method includes: In response to the sensor data of the production line loop starting to fluctuate, obtaining the sensor data of the production line loop until the sensor data returns to zero; Plotting the obtained sensor data as a line chart and inputting the line chart into a pre-trained detection model, where the detection model is trained based on the historical normal data and abnormal data of the sensor; In response to the output result of the detection model being abnormal data, sending an alarm signal to the user.
2. The method for detecting abnormal loop position in a bar and wire production line according to claim 1, characterized in that, The training method of the detection model includes: Obtaining the historical sensor data of the production line loop and preprocessing the historical sensor data, and intercepting the interval of the historical sensor data during the period when the bar and wire pass through the production line; Plotting the interval as a line chart and performing binary classification annotation on the line chart to obtain the historical normal data and historical abnormal data of the sensor data; Enhancing the historical abnormal data until the sample quantity ratio of the historical abnormal data to the historical normal data meets a preset ratio; Constructing a deep convolutional model, inputting and training the historical normal data and the historical abnormal data until the model converges to obtain the detection model.
3. The method for detecting abnormal loop position in a bar and wire production line according to claim 2, characterized in that, The historical sensor data includes the standard value, actual measurement value, and compensation value of the loop position during the production process of the bar and wire.
4. The method for detecting abnormal loop position in a bar and wire production line according to claim 2, characterized in that, The preprocessing of the historical sensor data to intercept the interval of the historical sensor data during the period when the bar and wire pass through the production line includes: In response to the historical sensor data starting to fluctuate, recording the historical sensor data starting from a preset time before the fluctuation; In response to the historical sensor data returning to zero, stopping recording the historical sensor data after a preset time after the zeroing, to obtain multiple intervals of the historical sensor data, where each interval contains the standard value, actual measurement value, and compensation value of the historical sensor data during the time period from the start of the fluctuation of the sensor data of the production line loop to the return of the sensor data to zero.
5. The method for detecting abnormal loop position in a bar and wire production line according to claim 3, characterized in that, Before inputting the historical abnormal data, it further includes: Adding a perturbation amount to the actual measurement value in the historical abnormal data to make the actual measurement value deviate from the original value within a preset range.
6. The method for detecting abnormal loop position in a bar and wire production line according to claim 2, characterized in that, The inputting and training the historical normal data and the historical abnormal data until the model converges to obtain the detection model includes: Inputting a preset proportion of the historical normal data and the historical abnormal data in all the historical sensor data into the deep convolutional model and training until the model converges; Inputting the remaining proportion of the historical normal data and the historical abnormal data into the converged model to verify the converged model, and determining the detection model when the verification passes.
7. A device for detecting abnormal loop position in a bar and wire production line, characterized in that, The device includes: A data acquisition module for: in response to the sensor data of the production line loop starting to fluctuate, obtaining the sensor data of the production line loop until the sensor data returns to zero; An anomaly detection module for: plotting the obtained sensor data as a line chart and inputting the line chart into a pre-trained detection model, where the detection model is trained by the training device of the detection model based on the historical normal data and abnormal data of the sensor; An abnormal alarm module, configured to: send an alarm signal to a user in response to the output result of the detection model being abnormal data.
8. The device for detecting abnormal loop position in a bar and wire production line according to claim 7, characterized in that, The training device of the detection model includes: A historical data acquisition module, configured to: acquire historical sensor data of a live loop on a production line, preprocess the historical sensor data, and intercept an interval of the historical sensor data during which bar and wire materials pass through the production line; An image classification module, configured to: plot the interval as a line chart, and perform binary classification annotation on the line chart to obtain historical normal data and historical abnormal data of the sensor data; An abnormal data enhancement module, configured to: enhance the historical abnormal data until the sample quantity ratio of the historical abnormal data to the historical normal data meets a preset ratio; A repeated training module, configured to: construct a deep convolutional model, input and train the historical normal data and the historical abnormal data until the model converges, to obtain the detection model.
9. The device for detecting abnormal loop position in a bar and wire production line according to claim 8, characterized in that,The historical sensor data includes standard values, actual measured values, and compensation values of the live loop position during the production process of bar and wire materials.
10. The abnormal detection device for the loop position in the bar and wire production line according to claim 8, wherein, The historical data acquisition module includes: A recording start unit, configured to: record the historical sensor data starting from a preset time before the fluctuation in response to the historical sensor data starting to fluctuate; A recording stop unit, configured to: stop recording the historical sensor data after a preset time after the zeroing in response to the historical sensor data being zeroed, to obtain multiple intervals of the historical sensor data, where each interval includes the standard value, actual measured value, and compensation value of the historical sensor data during the time period from the start of the fluctuation of the sensor data of the live loop on the production line to the zeroing of the sensor data.
11. The abnormal detection device for the loop position in the bar and wire production line according to claim 9, wherein, Before the repeated training module inputs the historical abnormal data, it further includes a deviation construction module, configured to: Add a perturbation amount to the actual measured value in the historical abnormal data, so that the actual measured value has a deviation value within a preset range from the original value.
12. The abnormal detection device for the loop position in the bar and wire production line according to claim 8, wherein, The repeated training module includes: A partial training unit, configured to: input a preset proportion of the historical normal data and the historical abnormal data in all the historical sensor data into the deep convolutional model and perform training until the model converges; A partial verification unit, configured to: input the remaining proportion of the historical normal data and the historical abnormal data into the converged model to verify the converged model, and determine the detection model in the case of passing the verification.
13. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, When the processor executes the program, it implements the steps of the method for detecting abnormal position of the live loop of the bar and wire production line according to any one of claims 1 to 6.
14. A computer-readable storage medium, on which a computer program is stored, wherein, When the computer program is executed by a processor, it implements the steps of the method for detecting abnormal position of the live loop of the bar and wire production line according to any one of claims 1 to 6.
15. A computer program product, comprising a computer program / instructions, wherein, When the computer program / instructions are executed by a processor, it implements the steps of the method for detecting abnormal position of the live loop of the bar and wire production line according to any one of claims 1 to 6.