Steel shot production data analysis system and method based on deep learning

Abstract

The invention discloses a steel shot production data analysis system and method based on deep learning, and belongs to the field of deep learning. A fault equipment searching model is constructed, and obtained steel shot production evaluation data and equipment operation data are imported into a fault reason searching model for fault reason searching; according to the method, the shape and the flaw production characteristics of the steel shot are comprehensively analyzed, so that the production quality of the steel shot is accurately obtained, then the fault reason in the operation process is quickly and accurately found according to the production quality characteristics of the steel shot, the operation data of the equipment and the historical data, and the operation efficiency is improved. And the accuracy and efficiency of fault reason discovery are improved.

Description

Technical Field

[0001] The present application belongs to the field of deep learning, and specifically to a steel shot production data analysis system and method based on deep learning. Background Art

[0002] Steel shot is a metal material widely used in metal surface treatment, casting, steel structure, metal rust removal and other fields. The production process of steel shot usually includes raw material preparation, smelting, casting, crushing, screening and other links. The main raw material of steel shot is scrap steel or pig iron. After the production of steel shot, the produced steel shot needs to be inspected for quality. The quality inspection includes appearance inspection, size measurement, chemical composition analysis and other aspects to ensure that the quality of steel shot meets the standard requirements. The production of steel shot requires strict control of the process parameters and quality standards of each link to ensure that the produced steel shot meets the use requirements and has good performance; When analyzing the production data of steel shots, the prior art cannot comprehensively analyze the shape and defect production characteristics of the steel shots, and thus cannot accurately obtain the production quality of the steel shots, resulting in the inaccurate discovery of the causes of failures during operation based on the production quality characteristics of the steel shots, the operation data of the equipment, and historical data. Most of the prior art has the above problems. In order to solve the problems raised by this background technology, the present application designs a steel shot production data analysis system and method based on deep learning. Summary of the invention

[0003] In response to the shortcomings of the existing technology, this application proposes a steel shot production data analysis system and method based on deep learning.

[0004] To achieve the above objectives, the present application provides the following technical solutions: In the first aspect, the present application provides a steel shot production data analysis method based on deep learning, which includes the following specific steps: S1. Obtaining the production data of steel shot and the operation data of the production equipment during operation; S2, obtaining steel shot production data and importing it into the steel shot production evaluation model to perform steel shot production evaluation; S3. Determine whether to search for faulty equipment based on the steel shot production assessment result. If yes, proceed to S4. If no, end. S4, construct a fault equipment search model, import the obtained steel shot production evaluation data and equipment operation data into the fault cause search model to search for the fault cause; S5. Feedback the found fault cause to the management end in real time.

[0005] As a preferred technical solution for the steel shot production data analysis method based on deep learning, the specific content of obtaining the steel shot production data and the operation data during the operation of the production equipment is: S11. Acquire production data corresponding to the steel shot production process, wherein the steel shot production data includes surface defect data of the steel shot and production shape data of the steel shot, wherein the surface defect data of the steel shot includes pit area and pit depth data on the surface of the steel shot, and the production shape data of the steel shot includes image data and diameter data of each parallel section of the steel shot, and the roundness data and diameter data of each corresponding parallel section are acquired by acquiring a three-dimensional image of the steel shot, cutting the steel shot image with a number of parallel sections at the same distance, acquiring images of all the cut surfaces as parallel sections, and acquiring roundness data and diameter data corresponding to each parallel section; S12, obtaining the production equipment operation data in the corresponding steel shot production process, wherein the production equipment operation data includes the operation equipment voltage, current, temperature and operation pressure data; S13, storing the obtained production data of the steel shot production process and the corresponding production equipment operation data of the steel shot production process in the storage component.

[0006] As a preferred technical solution for the steel shot production data analysis method based on deep learning, the acquisition of steel shot production data and importing it into the steel shot production evaluation model for steel shot production evaluation includes the following specific steps: S21, obtaining production data of the production process of several steel shot samples, importing the obtained production shape data of the steel shot into a shape abnormality value calculation formula to calculate the shape abnormality value, wherein the shape abnormality value calculation formula is: , where n is the number of steel shot samples, mi is the number of parallel sections, and Lij is the shortest distance from the jth parallel section to the center of the ball; S22, obtaining the surface defect data of the steel shot, and importing the surface defect data of the steel shot into the defect abnormal value calculation formula to calculate the defect abnormal value, wherein the defect abnormal value calculation formula is: , where Di is the number of pits on the i-th steel shot sample, V is the volume of the standard steel shot, dci is the depth of the c-th pit of the i-th steel shot sample, and sci is the area of ​​the c-th pit of the i-th steel shot sample; S23, obtain the calculated shape abnormal value and defect abnormal value, and substitute them into the steel shot production abnormal value calculation formula to calculate the steel shot production abnormal value, wherein the steel shot production abnormal value calculation formula is: , where b is the defect anomaly ratio; in this formula, the steel shot anomaly is accurately analyzed through shape anomaly and defect anomaly.

[0007] As a preferred technical solution of the steel shot production data analysis method based on deep learning, the method of judging whether to search for faulty equipment based on the steel shot production evaluation results includes the following specific steps: The obtained steel shot production abnormality value is compared with the set steel shot production abnormality threshold. If the obtained steel shot production abnormality value is greater than or equal to the set steel shot production abnormality threshold, it is necessary to search for faulty equipment. If the obtained steel shot production abnormality value is less than the set steel shot production abnormality threshold, it is not necessary to search for faulty equipment.

[0008] As a preferred technical solution for the steel shot production data analysis method based on deep learning, the construction of the fault equipment search model and importing the obtained steel shot production evaluation data and equipment operation data into the fault cause search model for fault cause search include the following specific contents: S41, extracting historical equipment operation data, shape anomaly values ​​of historical production products, defect anomaly value data of historical production products, and searched fault cause data, and constructing a deep learning neural network model whose input is the historical equipment operation data, shape anomaly values ​​of historical production products, and defect anomaly value data of historical production products, and whose output is the fault cause data; S42, setting the extracted historical equipment operation data, shape anomaly values ​​of historical production products, defect anomaly value data of historical production products and searched fault cause data into a parameter training set and a parameter test set in a ratio of 9:1; inputting 90% of the parameter training set into the deep learning neural network model for training to obtain an initial deep learning neural network model; using 10% of the parameter test set to test the initial deep learning neural network model, and outputting the initial deep learning neural network model that meets the maximum fault cause search accuracy as the deep learning neural network model; S43. Import the calculated shape anomaly values, defect anomaly values ​​and equipment operation data into the deep learning neural network model to output corresponding fault cause data.

[0009] In the second aspect, the present application provides a steel shot production data analysis system based on deep learning, which is implemented based on the above-mentioned steel shot production data analysis method based on deep learning, and specifically includes a data acquisition module, a steel shot production evaluation module, a judgment module, a fault cause finding module and a cause feedback module; Wherein, the data acquisition module is used to acquire the production data of steel shots and the operation data of the production equipment during operation; The steel shot production evaluation module is used to obtain steel shot production data and import it into the steel shot production evaluation model to perform steel shot production evaluation; The judgment module is used to judge whether to search for faulty equipment based on the steel shot production evaluation result; The fault cause finding module is used to build a fault equipment finding model, and import the obtained steel shot production evaluation data and equipment operation data into the fault cause finding model to find the fault cause; The cause feedback module is used to feed back the found fault cause to the management end in real time; It may also include a control module, which is used to control the operation of the data acquisition module, the steel shot production evaluation module, the judgment module, the fault cause finding module and the cause feedback module.

[0010] In a third aspect, the present application provides an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the above-mentioned steel shot production data analysis method based on deep learning by calling the computer program stored in the memory.

[0011] In a fourth aspect, the present application provides a computer-readable storage medium storing instructions, which, when executed on a computer, enables the computer to execute the steel shot production data analysis method based on deep learning as described above.

[0012] Compared with the prior art, the beneficial effects of this application are: The present application obtains steel shot production data and operation data during the operation of production equipment, obtains steel shot production data and imports it into a steel shot production evaluation model to conduct steel shot production evaluation, determines whether to search for faulty equipment based on the steel shot production evaluation results, builds a faulty equipment search model, imports the obtained steel shot production evaluation data and equipment operation data into a fault cause search model to search for fault causes, and feeds back the found fault causes to the management end in real time. The present application conducts a comprehensive analysis of the shape of steel shots and defective production characteristics to accurately obtain the production quality of steel shots, and then quickly and accurately discovers the fault causes in the operation process based on the production quality characteristics of steel shots, the equipment's operation data and historical data, thereby improving the accuracy and efficiency of fault cause discovery. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings; Figure 1 This is a schematic diagram of the overall process of the steel shot production data analysis method based on deep learning in this application; Figure 2 This is a schematic diagram of step S2 of the steel shot production data analysis method based on deep learning in this application; Figure 3 This is a schematic diagram of step S4 of the steel shot production data analysis method based on deep learning in this application; Figure 4 This is a schematic diagram of the overall framework of the steel shot production data analysis system based on deep learning in this application; Figure 5A three-dimensional schematic diagram is obtained for the parallel section of this application; Figure 6 This is a schematic diagram of the parallel surface cutting process of this application. DETAILED DESCRIPTION

[0014] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present application and its application or use.

[0015] Embodiment 1. To solve the technical problems raised in the background technology, the present application provides a preferred embodiment: Figure 1 - Figure 3 As shown, the steel shot production data analysis method based on deep learning includes the following specific steps: S1. Obtaining the production data of steel shot and the operation data of the production equipment during operation; In one specific embodiment, the specific contents of obtaining the steel shot production data and the operation data of the production equipment during operation are: S11, obtaining production data of the corresponding steel shot production process, wherein the steel shot production data includes steel shot surface defect data and steel shot production shape data, wherein the steel shot surface defect data includes pit area and pit depth data on the steel shot surface, and the steel shot production shape data includes image data and diameter data of each parallel section of the steel shot, such as Figure 5 and Figure 6 As shown, Figure 5 A three-dimensional schematic diagram is obtained by parallel sectioning for this application. Figure 6 This is a schematic diagram of the parallel surface cutting process of this application. The roundness data and diameter data of each corresponding parallel section are obtained by obtaining a three-dimensional image of the steel shot, cutting the steel shot image with several parallel surfaces of the same distance, obtaining images of all the cut surfaces as parallel sections, and obtaining the roundness data and diameter data corresponding to each parallel section; The production data of steel shot can be obtained from the stereoscopic image of steel shot. The stereoscopic image of steel shot can be obtained using a 3D scanner: a 3D scanner is a device that can quickly obtain the surface shape and geometric data of an object. By placing the steel shot on the platform of the scanner, the 3D point cloud data of the steel shot can be obtained, and then a stereoscopic image can be generated; S12, obtaining the production equipment operation data in the corresponding steel shot production process, wherein the production equipment operation data includes the operation equipment voltage, current, temperature and operation pressure data; S13, storing the obtained production data of the steel shot production process and the operation data of the production equipment in the corresponding steel shot production process in the storage component; S2, obtaining steel shot production data and importing it into the steel shot production evaluation model to perform steel shot production evaluation; In one specific embodiment, obtaining steel shot production data and importing it into a steel shot production evaluation model to perform steel shot production evaluation includes the following specific steps: S21, obtaining production data of the production process of several steel shot samples, importing the obtained production shape data of the steel shot into a shape abnormality value calculation formula to calculate the shape abnormality value, wherein the shape abnormality value calculation formula is: , where n is the number of steel shot samples, mi is the number of parallel sections, Lij is the shortest distance from the jth parallel section to the center of the sphere. It should be noted that the center of the sphere here is the center of the steel shot, a is the shape anomaly ratio, S() is the area of ​​the image in brackets, Tij is the image of the jth parallel section of the i-th steel shot sample, and Tijm is the image of the jth parallel section of the standard steel shot. is the intersection of the images, is the union of the images, xij is the average diameter of the jth parallel section of the ith steel shot sample, and xijm is the average diameter of the jth parallel section of the standard steel shot. It should be noted that since a steel shot may have accidental factors, multiple steel shot samples are selected here, and the number of parallel sections is selected to be greater than or equal to 5. In the formula In the above formula, the area of ​​the intersection image of the image of the jth parallel section of the i-th steel shot sample and the jth parallel section of the standard steel shot is given, and the area of ​​the union image of the image of the jth parallel section of the i-th steel shot sample and the jth parallel section of the standard steel shot is given, that is, the similarity of the two image areas. Since the formula outputs a shape anomaly value, it is inversely proportional to the similarity of the two image areas, so it is necessary to use ; and for is the phase difference of the diameter. Therefore, this formula analyzes the anomaly of parallel sections through the anomaly of the image and the anomaly of the diameter. For parallel sections at different positions, the section closest to the center of the steel shot is definitely the largest and most important. Therefore, use , that is, the inverse of the distance to the center of the steel shot is weighted by importance. The significance of adding 1 to the distance here is to prevent the numerator from being 0; S22, obtaining the surface defect data of the steel shot, and importing the surface defect data of the steel shot into the defect abnormal value calculation formula to calculate the defect abnormal value, wherein the defect abnormal value calculation formula is: , where Di is the number of pits on the i-th steel shot sample, V is the volume of the standard steel shot, dci is the depth of the c-th pit of the i-th steel shot sample, and sci is the area of ​​the c-th pit of the i-th steel shot sample; S23, obtain the calculated shape abnormal value and defect abnormal value, and substitute them into the steel shot production abnormal value calculation formula to calculate the steel shot production abnormal value, wherein the steel shot production abnormal value calculation formula is: , where b is the defect abnormality ratio; in this formula, the steel shot abnormality is accurately analyzed through shape abnormality and defect abnormality; S3. Determine whether to search for faulty equipment based on the steel shot production assessment result. If yes, proceed to S4. If no, end. In one of the specific embodiments, the obtained steel shot production abnormality value is compared with the set steel shot production abnormality threshold. If the obtained steel shot production abnormality value is greater than or equal to the set steel shot production abnormality threshold, it is necessary to search for faulty equipment. If the obtained steel shot production abnormality value is less than the set steel shot production abnormality threshold, it is not necessary to search for faulty equipment. Here, it is necessary to specifically explain the value-taking method of the shape abnormality ratio coefficient, defect abnormality ratio and steel shot production abnormality threshold: obtain 5000 sets of steel shot production data and import them into the steel shot production abnormality value calculation formula to calculate the steel shot production abnormality value, perform equipment maintenance to see if a fault occurs, import the fault judgment result and the calculated steel shot production abnormality value into the fitting software, and output the value of the shape abnormality ratio coefficient, defect abnormality ratio and steel shot production abnormality threshold that meet the maximum accuracy of fault judgment; S4, construct a fault equipment search model, import the obtained steel shot production evaluation data and equipment operation data into the fault cause search model to search for the fault cause; In one specific embodiment, a fault equipment search model is constructed, and the obtained steel shot production evaluation data and equipment operation data are imported into the fault cause search model to search for the fault cause, including the following specific contents: S41, extracting historical equipment operation data, shape anomaly values ​​of historical production products, defect anomaly value data of historical production products, and searched fault cause data, and constructing a deep learning neural network model whose input is equipment operation data, shape anomaly values ​​of production products, and defect anomaly value data of production products, and whose output is fault cause data; S42, setting the parameter training set and parameter test set according to the ratio of 9:1 of the extracted historical equipment operation data, shape anomaly values ​​of historical production products, defect anomaly value data of historical production products and fault cause data found; inputting 90% of the parameter training set into the deep learning neural network model for training to obtain an initial deep learning neural network model; using 10% of the parameter test set to test the initial deep learning neural network model, and outputting the initial deep learning neural network model that meets the maximum fault cause finding accuracy as the deep learning neural network model, wherein the output strategy formula of the s-th neuron of the M-th layer in the deep learning neural network model is: ,in is the output of the sth neuron in layer M, is the connection weight between the cth neuron in the M-1th layer and the sth neuron in the Mth layer, represents the input of the cth neuron in the M-1th layer, represents the bias of the linear relationship between the cth neuron in the M-1th layer and the sth neuron in the M layer, Sig() represents the Sigmoid activation function, and w is the number of neurons in the M-1th layer; it should be noted that in this neural network, the three inputs of the input layer are: equipment operation data, shape anomaly values ​​of production products, and defect anomaly value data of production products, and the output of the output layer is the fault cause found; S43, importing the calculated shape anomaly value, defect anomaly value and equipment operation data into the deep learning neural network model to output corresponding fault cause data; S5. Feedback the found fault cause to the management end in real time, and the management end selects appropriate maintenance personnel to perform maintenance on the fault according to the found fault cause.

[0016] The advantages of this embodiment over the prior art are as follows: obtaining steel shot production data and operation data during the operation of production equipment, obtaining steel shot production data and importing it into a steel shot production evaluation model to conduct steel shot production evaluation, judging whether to search for faulty equipment based on the steel shot production evaluation results, building a faulty equipment search model, importing the obtained steel shot production evaluation data and equipment operation data into a fault cause search model to search for fault causes, and feeding back the found fault causes to the management end in real time. This application conducts a comprehensive analysis of the shape of steel shots and defective production characteristics to accurately obtain the production quality of steel shots, and then quickly and accurately discovers the fault causes in the operation process based on the production quality characteristics of steel shots, the equipment's operation data and historical data, thereby improving the accuracy and efficiency of fault cause discovery.

[0017] Example 2. Figure 4As shown, a steel shot production data analysis system based on deep learning is implemented based on the above-mentioned steel shot production data analysis method based on deep learning, and specifically includes a data acquisition module, a steel shot production evaluation module, a judgment module, a fault cause finding module and a cause feedback module; wherein the data acquisition module is used to acquire steel shot production data and operation data during the operation of production equipment; the steel shot production evaluation module is used to acquire steel shot production data and import it into a steel shot production evaluation model to perform steel shot production evaluation; the judgment module is used to determine whether to search for faulty equipment based on the steel shot production evaluation results; the fault cause finding module is used to construct a faulty equipment search model, and import the obtained steel shot production evaluation data and equipment operation data into the fault cause finding model to search for fault causes; the cause feedback module is used to feed back the found fault causes to the management end in real time; it may also include a control module, which is used to control the operation of the data acquisition module, the steel shot production evaluation module, the judgment module, the fault cause finding module and the cause feedback module; The above-mentioned parameters and steps for each unit module to realize the corresponding functions in the steel shot production data analysis system based on deep learning of the present application can refer to the parameters and steps in the embodiment of the steel shot production data analysis method based on deep learning above, and will not be repeated here.

[0018] Embodiment 3. This embodiment provides an electronic device, including: a processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the above-mentioned steel shot production data analysis method based on deep learning by calling the computer program stored in the memory.

[0019] The electronic device may have relatively large differences due to different configurations or performances, and may include one or more processors and one or more memories, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the steel shot production data analysis method based on deep learning provided by the above method embodiment. The electronic device may also include other components for realizing the functions of the device, for example, the electronic device may also have components such as a wired or wireless network interface and an input and output interface to input and output data. This embodiment will not be described in detail here.

[0020] Embodiment 4. This embodiment provides a computer-readable storage medium having a rewritable computer program stored thereon; When the computer program runs on a computer device, the computer device executes the above-mentioned steel shot production data analysis method based on deep learning.

[0021] For example, the computer readable storage medium can be a read-only memory, a random access memory, a read-only CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0022] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When a computer instruction or computer program is loaded or executed on a computer, a process or function according to an embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. Computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or / and a wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0023] The terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article, or apparatus.

[0024] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of application involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the aforementioned application concept. For example, the above features are replaced with (but not limited to) technical features with similar functions applied in the present application.

Claims

1. A steel shot production data analysis method based on deep learning, characterized in that: It includes the following specific steps: S1. Obtaining the production data of steel shot and the operation data of the production equipment during operation; S2, obtaining steel shot production data and importing it into the steel shot production evaluation model to perform steel shot production evaluation; S3. Determine whether to search for faulty equipment based on the steel shot production assessment result. If yes, proceed to S4. If no, end. S4, construct a fault equipment search model, import the obtained steel shot production evaluation data and equipment operation data into the fault cause search model to search for the fault cause; S5. Feedback the found fault cause to the management end in real time.

2. The steel shot production data analysis method based on deep learning according to claim 1, characterized in that: The specific contents of obtaining the steel shot production data and the operation data of the production equipment during operation are as follows: S11, obtaining production data of the corresponding steel shot production process, wherein the steel shot production data includes surface defect data of the steel shot and production shape data of the steel shot, wherein the surface defect data of the steel shot includes pit area and pit depth data on the surface of the steel shot, and the production shape data of the steel shot includes image data and diameter data of each parallel section of the steel shot; S12, obtaining the production equipment operation data in the corresponding steel shot production process, wherein the production equipment operation data includes the operation equipment voltage, current, temperature and operation pressure data; S13, storing the obtained production data of the steel shot production process and the corresponding production equipment operation data of the steel shot production process in the storage component.

3. The steel shot production data analysis method based on deep learning according to claim 1, characterized in that: The method of obtaining steel shot production data and importing it into the steel shot production evaluation model to perform steel shot production evaluation includes the following specific steps: S21, obtaining production data of the production process of several steel shot samples, importing the obtained production shape data of the steel shot into a shape abnormality value calculation formula to calculate the shape abnormality value, wherein the shape abnormality value calculation formula is: , where n is the number of steel shot samples, mi is the number of parallel sections, Lij is the shortest distance from the jth parallel section to the center of the sphere, a is the shape anomaly ratio, S() is the area of ​​the image in brackets, Tij is the image of the jth parallel section of the i-th steel shot sample, and Tijm is the image of the jth parallel section of the standard steel shot. is the intersection of the images, is the union of the images, xij is the average diameter of the jth parallel section of the i-th steel shot sample, and xijm is the average diameter of the jth parallel section of the standard steel shot; S22, obtaining the surface defect data of the steel shot, and importing the surface defect data of the steel shot into the defect abnormal value calculation formula to calculate the defect abnormal value, wherein the defect abnormal value calculation formula is: , where Di is the number of pits on the i-th steel shot sample, V is the volume of the standard steel shot, dci is the depth of the c-th pit of the i-th steel shot sample, and sci is the area of ​​the c-th pit of the i-th steel shot sample; S23, obtain the calculated shape abnormal value and defect abnormal value, and substitute them into the steel shot production abnormal value calculation formula to calculate the steel shot production abnormal value, wherein the steel shot production abnormal value calculation formula is: , where b is the defect anomaly ratio.

4. The steel shot production data analysis method based on deep learning according to claim 3 is characterized in that: The method of judging whether to search for faulty equipment based on the steel shot production evaluation result includes the following specific steps: The obtained steel shot production abnormality value is compared with the set steel shot production abnormality threshold. If the obtained steel shot production abnormality value is greater than or equal to the set steel shot production abnormality threshold, it is necessary to search for faulty equipment. If the obtained steel shot production abnormality value is less than the set steel shot production abnormality threshold, it is not necessary to search for faulty equipment.

5. The steel shot production data analysis method based on deep learning according to claim 4, characterized in that: The construction of the fault equipment search model and importing the obtained steel shot production evaluation data and equipment operation data into the fault cause search model to search for the fault cause include the following specific contents: S41, extracting historical equipment operation data, shape anomaly values ​​of historical production products, defect anomaly value data of historical production products, and searched fault cause data, and constructing a deep learning neural network model whose input is the historical equipment operation data, shape anomaly values ​​of historical production products, and defect anomaly value data of historical production products, and whose output is the fault cause data; S42, setting the extracted historical equipment operation data, shape anomaly values ​​of historical production products, defect anomaly value data of historical production products and searched fault cause data into a parameter training set and a parameter test set in a ratio of 9:1; inputting 90% of the parameter training set into the deep learning neural network model for training to obtain an initial deep learning neural network model; using 10% of the parameter test set to test the initial deep learning neural network model, and outputting the initial deep learning neural network model that meets the maximum fault cause search accuracy as the deep learning neural network model; S43. Import the calculated shape anomaly values, defect anomaly values ​​and equipment operation data into the deep learning neural network model to output corresponding fault cause data.

6. A steel shot production data analysis system based on deep learning, which is implemented based on the steel shot production data analysis method based on deep learning as claimed in any one of claims 1 to 5, characterized in that: It specifically includes a data acquisition module, a steel shot production evaluation module, a judgment module, a fault cause finding module and a cause feedback module; Wherein, the data acquisition module is used to acquire the production data of steel shots and the operation data of the production equipment during operation; The steel shot production evaluation module is used to obtain steel shot production data and import it into the steel shot production evaluation model to perform steel shot production evaluation; The judgment module is used to judge whether to search for faulty equipment based on the steel shot production evaluation result; The fault cause finding module is used to build a fault equipment finding model, and import the obtained steel shot production evaluation data and equipment operation data into the fault cause finding model to find the fault cause; The cause feedback module is used to feed back the found fault cause to the management end in real time.

7. An electronic device comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; It is characterized in that the processor executes the steel shot production data analysis method based on deep learning as described in any one of claims 1 to 5 by calling the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer executes the steel shot production data analysis method based on deep learning as described in any one of claims 1 to 5.

Images