An intelligent production management system based on data twins

Abstract

The present invention discloses an intelligent production management system based on data twin, which belongs to the field of production management system, including a crushing equipment data acquisition module, a drying equipment data acquisition module, a screening equipment data acquisition module, a data preprocessing and storage module, a digital twin modeling module, a simulation analysis module, an early warning classification module and a notification module; the crushing equipment data module is used to collect the feeding time, feeding image, discharging time and discharging image information of the crushing equipment, and send the collected information to the data preprocessing and storage module; the drying equipment data acquisition module is used to collect the drying time, manual participation time, waste gas emission and wastewater emission information of the drying equipment, and send the collected information to the data preprocessing and storage module. The present invention can effectively manage the early stage processing of raw coke, and then timely discover and warn when abnormalities occur.

Description

Technical Field

[0001] The present invention relates to the field of production management systems, and specifically to an intelligent production management system based on data twins. Background Art

[0002] Needle coke is a carbon material with a special shape and excellent performance. It is a silver-gray porous solid with a metallic luster. Its structure has obvious flow texture, large but few pores that are slightly elliptical, and the particles have a large aspect ratio, with a fibrous or needle-like texture and feel smooth to the touch.

[0003] The needle coke calcining process is an important part of the needle coke production process. Its main purpose is to remove moisture and volatiles from the raw coke through high-temperature treatment, and to improve the density, mechanical strength, electrical conductivity and chemical stability of the coke. Before calcination, the raw coke needs to be pre-processed (i.e. crushing, drying, and screening). More specifically, the raw needle coke in the coke pool is crushed by a crusher, and then transported to the dehydration bin by conveying equipment such as bucket elevators and belt conveyors for drainage. After drainage, it is transported to the drying equipment by belts and bucket elevators to dry the moisture, and then transported to the screening machine by belts. The dehydrated raw coke with a particle size smaller than the particle size enters the powder silo and is packaged for sale, while the dehydrated raw coke with a large particle size enters the pre-calcination bin.

[0004] The existing technology has the following problems: it is impossible to effectively manage the early processing of raw coke, and thus it is impossible to detect and warn in time when an abnormality occurs. Therefore, those skilled in the art provide an intelligent production management system based on data twins to solve the problems raised in the above background technology. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent production management system based on data twins, which can effectively manage the early processing of coke production, and then promptly discover and issue early warnings when abnormalities occur, so as to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An intelligent production management system based on data twins, including a crushing equipment data acquisition module, a drying equipment data acquisition module, a screening equipment data acquisition module, a data preprocessing and storage module, a digital twin modeling module, a simulation analysis module, an early warning classification module and a notification module;

[0008] The crushing equipment data module is used to collect the feeding time, feeding image, discharging time and discharging image information of the crushing equipment, and send the collected information to the data preprocessing and storage module; the drying equipment data acquisition module is used to collect the drying time, manual participation time, exhaust gas emission and wastewater emission information of the drying equipment, and send the collected information to the data preprocessing and storage module; the screening equipment data acquisition module is used to collect the total feed weight, screening product weight, screened material weight and screening time information of the screening equipment, and send the collected information to the data preprocessing and storage module; the data preprocessing and storage module is used to preprocess and store the received information, and send the preprocessed information to the digital twin modeling module; the digital twin modeling module constructs a digital twin model of needle coke calcination pretreatment based on the preprocessed information; the simulation analysis module analyzes the crushing equipment, drying equipment and screening equipment based on the digital twin model, and outputs the analysis results to the early warning classification module; the early warning classification module is used to perform graded early warning according to the analysis results, and send the classification results to the notification module; the notification module is used to issue early warning notifications of corresponding levels according to the classification results.

[0009] As a further solution of the present invention: the specific process of the simulation analysis module analyzing the crushing equipment based on the digital twin model is:

[0010] The feeding time of the material in this round of the crushing equipment is marked as T1, and the discharging time of the material in this round of the crushing equipment is marked as T2;

[0011] Calculate the crushing time of this round of materials: t1 = (T2-T1);

[0012] Compare the crushing time t1 with the preset time t2. If t1>t2, it means that the crushing time of the crushing equipment is abnormal. If t1≤t2, it means that the crushing time of the crushing equipment is normal.

[0013] Construct the image analysis sub-model and output the analysis results.

[0014] As a further solution of the present invention: the specific analysis process of the image analysis sub-model is:

[0015] The feeding image of the material in this round of the crushing equipment is marked as Ai, i=1···n, n is a positive integer, and nine images are randomly selected from Ai and marked as ai, i=1···9 respectively;

[0016] The discharge image of the material in this round of crushing equipment is marked as Bi, i=1···n, n is a positive integer, and nine images are randomly selected from Bi and marked as bi, i=1···9 respectively;

[0017] Compare ai and bi at pixel level, calculate the difference between ai and bi and mark it as Ci, i=1···9;

[0018] Compare Ci with the preset value c. If Ci>c, Ci is planned to the first sequence. If Ci≤c, Ci is planned to the second sequence.

[0019] The nine images bi are compared pairwise at pixel level, and the differences are calculated and marked as Di, i=1···36;

[0020] Compare Di with the preset value d. If Di < d, Di is planned to the third sequence. If Di ≥ d, Di is planned to the fourth sequence.

[0021] Count the number of Ci in the second sequence and mark it as v1, and count the number of Di in the fourth sequence and mark it as v2;

[0022] Calculate the image difference weight value S = (v1*3+v2), and compare S with the preset value s. If S>s, it means that the crushing effect of the crushing equipment is abnormal. If S≤s, it means that the crushing effect of the crushing equipment is normal.

[0023] As a further solution of the present invention: the specific process of the simulation analysis module analyzing the drying equipment based on the digital twin model is:

[0024] The total drying time of the drying equipment in this round of materials is marked as t3, and the manual participation time of the drying equipment in this round of materials is marked as t4;

[0025] Calculate the automation degree of the drying equipment in this round of materials X = (t3-t4) / t3;

[0026] Compare X with the preset value x. If X>x, it means that the automation level of the drying equipment is normal. If X≤x, it means that the automation level of the drying equipment is abnormal.

[0027] The exhaust gas emission per minute of the drying equipment in this round of materials is marked as Wi, i=1···n, and the wastewater emission per minute of the drying equipment in this round of materials is marked as Mi, i=1···n;

[0028] Each Wi and Mi is represented by a line graph, wherein each Wi forms a line 1 and each Mi forms a line 2;

[0029] Based on broken line 1 and broken line 2, figure P is obtained, and the similarity between P and the preset figure p is compared. If the similarity between P and p is greater than a first preset value, it means that the pollutant emission of the drying equipment is normal. If the similarity between P and p is less than or equal to the first preset value, it means that the pollutant emission of the drying equipment is abnormal.

[0030] As a further solution of the present invention: the specific process of the simulation analysis module analyzing the screening equipment based on the digital twin model is:

[0031] The total weight of the feed of the current round of the screening equipment is marked as G real time, the weight of the screening product is marked as Y real time, the weight of the screened material is marked as Z real time, the screening time is marked as U real time, and the total weight of the feed of each round of the screening equipment in the past is marked as Gi, i=1···n;

[0032] Match Gi with G in real time. If the real-time matching between Gi and G is successful, Gi is planned to the fifth sequence. If the real-time matching between Gi and G is unsuccessful, Gi is planned to the sixth sequence.

[0033] The Gi in the fifth sequence is marked as gi, i = 1···n, and the sieving time corresponding to gi is marked as ui, the sieving product weight is marked as yi, and the weight of the oversize is marked as zi;

[0034] Match Ureal time with ui. If any one of ui matches Ureal time, it means that the screening time of the screening device is normal. If any one of ui does not match Ureal time, it means that the screening time of the screening device is abnormal.

[0035] Construct a screening effect analysis sub-model, analyze the difference between each round of materials corresponding to gi and the current round of materials, and output the analysis results.

[0036] As a further solution of the present invention: the specific analysis process of the screening effect analysis sub-model is:

[0037] Calculate the weight difference of the sieved product Ri=(Y real time / yi);

[0038] Calculate the weight difference of the material on the sieve: RRi = (Z real time / zi);

[0039] Calculate the material loss difference value RRRi=(G real-time-Y real-time-Z real-time) / (gi-yi-zi);

[0040] Obtain radar chart Fi based on Ri, RRi, and RRRi, and mark the preset radar chart as f;

[0041] Compare Fi with f for similarity. If the similarity between Fi and f is less than a second preset value, plan Fi to the seventh sequence. If the similarity between Fi and f is greater than or equal to the second preset value, plan Fi to the eighth sequence.

[0042] Count the number of Fi in the seventh sequence and mark it as h1, count the number of Fi in the eighth sequence and mark it as h2;

[0043] If (h1 / h2) is greater than the third preset value, it means that the screening effect of the screening device is abnormal; if (h1 / h2) is less than or equal to the third preset value, it means that the screening effect of the screening device is normal.

[0044] As a further solution of the present invention: the specific process of the early warning classification module for performing the graded early warning is:

[0045] If the analysis results show any of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result is a first-level warning;

[0046] If the analysis results show any two of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result is a level 2 warning;

[0047] If the analysis results show any three of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result will be a level 3 warning;

[0048] If the analysis results show any four of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result will be a level 4 warning;

[0049] If the analysis results show any five of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result will be a level 5 warning;

[0050] If the analysis results show that all of the following items are abnormal in the crushing time of the crushing equipment, abnormal in the crushing effect of the crushing equipment, abnormal in the degree of automation of the drying equipment, abnormal in the amount of pollutant discharged by the drying equipment, abnormal in the screening time of the screening equipment, and abnormal in the screening effect of the screening equipment, the output classification result will be a level six warning.

[0051] As a further solution of the present invention: the notification module sends the analysis results and the warning classification results to a preset person in charge in the form of SMS notification.

[0052] As a further solution of the present invention: the data preprocessing and storage module preprocesses information including cleaning, integrating, converting and compressing the collected raw data to ensure the quality and consistency of the data.

[0053] As a further solution of the present invention: it also includes an interactive module for providing an intuitive user interface to facilitate production management personnel to view the analysis results of the simulation analysis module and the classification results of the early warning classification module.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] 1. This application can effectively manage the early processing of raw coke, and then promptly discover and issue early warnings when abnormalities occur. Among them, this application analyzes the early processing of raw coke from multiple angles from the crushing time and crushing effect of the crushing equipment, the degree of automation and pollutant emissions of the drying equipment, the screening time and screening effect of the screening equipment, so as to improve the accuracy of the management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A structural diagram of an intelligent production management system based on data twins

[0057] Figure 2 It is a schematic diagram of a graph P in an intelligent production management system based on data twins;

[0058] Figure 3 It is a schematic diagram of a radar chart Fi in an intelligent production management system based on data twins. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] As mentioned in the background technology of this application, research has found that the existing needle coke calcination production management system is unable to effectively manage the early processing of raw coke, and is unable to promptly detect and issue warnings when abnormalities occur, which results in certain defects.

[0061] In order to solve the above-mentioned defects, the present application discloses an intelligent production management system based on data twins, which can effectively manage the early processing of coke production, and then promptly detect and issue early warnings when abnormalities occur.

[0062] The following will describe in detail how the solution of the present application solves the above technical problems in conjunction with the accompanying drawings.

[0063] See also Figure 1 In an embodiment of the present invention, an intelligent production management system based on data twins includes a crushing equipment data acquisition module, a drying equipment data acquisition module, a screening equipment data acquisition module, a data preprocessing and storage module, a digital twin modeling module, a simulation analysis module, an early warning classification module and a notification module; the crushing equipment data module is used to collect the feeding time, feeding image, discharging time and discharging image information of the crushing equipment, and send the collected information to the data preprocessing and storage module; the drying equipment data acquisition module is used to collect the drying time, manual participation time, exhaust gas emission and wastewater emission information of the drying equipment, and send the collected information to the data preprocessing and storage module; the screening equipment data acquisition module is used to collect Collect the total feed weight, screening product weight, screened material weight and screening time information of the screening equipment, and send the collected information to the data preprocessing and storage module; the data preprocessing and storage module is used to preprocess and store the received information, and send the preprocessed information to the digital twin modeling module; the digital twin modeling module builds a digital twin model of needle coke calcination pretreatment based on the preprocessed information; the simulation analysis module analyzes the crushing equipment, drying equipment and screening equipment based on the digital twin model, and outputs the analysis results to the early warning classification module; the early warning classification module is used to perform graded early warnings based on the analysis results, and send the classification results to the notification module; the notification module is used to issue early warning notifications of corresponding levels based on the classification results. This application can effectively manage the early processing of raw coke, and then promptly detect and warn when abnormalities occur.

[0064] In this embodiment, the specific process of the simulation analysis module analyzing the crushing equipment based on the digital twin model is as follows: mark the feeding time of the material of the current round of the crushing equipment as T1, and mark the discharging time of the material of the current round of the crushing equipment as T2; calculate the crushing time t1=(T2-T1) of the material of the current round; compare the crushing time t1 with the preset time t2, if t1>t2, it means that the crushing time of the crushing equipment is abnormal, if t1≤t2, it means that the crushing time of the crushing equipment is normal; construct an image analysis sub-model and output the analysis results. This setting can quickly analyze whether the crushing time of the crushing equipment is normal. When the crushing equipment is crushing materials, the crushing time may fluctuate to a certain extent due to the different properties of the materials themselves (such as hardness, humidity, particle size, etc.). If this fluctuation is within the normal range specified in the equipment design or operation manual and does not affect production efficiency and product quality, it is usually not considered abnormal. If the crushing time suddenly becomes significantly longer and exceeds the normal fluctuation range, then this may indicate that there is some abnormality. The causes of the abnormality may include but are not limited to: increased equipment wear, damage to cutters or crushing components, reduced motor performance, unstable feed quantity or feed speed, significant changes in material properties (such as mixing with impurities, increased humidity, etc.), equipment control system failure, etc.

[0065] In this embodiment, the specific analysis process of the image analysis sub-model is as follows: the feeding image of the material of the current round of the crushing equipment is marked as Ai, i=1···n, n is a positive integer, and nine images are randomly selected from Ai and marked as ai, i=1···9 respectively; the discharging image of the material of the current round of the crushing equipment is marked as Bi, i=1···n, n is a positive integer, and nine images are randomly selected from Bi and marked as bi, i=1···9 respectively; ai and bi are compared at the pixel level, the difference between ai and bi is calculated and marked as Ci, i=1···9; Ci is compared with a preset value c (the value interval of c is 0.8≤c≤0.95), if Ci>c, Ci is planned to the first sequence, if Ci≤c, Ci is planned to the second sequence; the nine images bi are compared at the pixel level, the difference is calculated and marked as Di, i=1· ··36; compare Di with the preset value d (the value range of d is 0.2≤d≤0.4), if Di<d, then plan Di to the third sequence, if Di≥d, then plan Di to the fourth sequence; count the number of Ci in the second sequence and mark it as v1 (it should be noted that there are nine Ci in total, corresponding to the difference between the nine ai and the nine bi, and the second sequence contains Ci less than or equal to c, this number can be directly counted and marked as v1, and the value range of v1 is 0≤v1≤9), count the number of Di in the fourth sequence and mark it as v2; calculate the image difference weight value S=(v1*3+v2), compare S with the preset value s (the value range of s is 10≤s≤50), if S>s, it means that the crushing effect of the crushing equipment is abnormal, if S≤s, it means that the crushing effect of the crushing equipment is normal. This setting can quickly analyze whether there is any abnormality in the crushing effect of the crushing equipment. For the crushing equipment, the greater the difference between the material before and after crushing, the better the crushing effect. For all materials after crushing, the smaller the difference between them, the better the consistency of the crushed materials and the better the crushing effect. Among them, the calculation method of the difference degree is the existing technology. The most common calculation method is the method based on pixel comparison, and the method based on pixel comparison is divided into mean square error (MSE) and structural similarity index (SSIM). The mean square error (MSE) calculates the average of the square difference between each pixel value of the two images. It is simple and easy to implement and is applicable to any type of picture. The structural similarity index (SSIM) compares the similarity of two images from three aspects: brightness, contrast and structure. It takes into account the structural information of the image and is more accurate in evaluating the image quality. In this application, the mean square error (MSE) is used to calculate the difference value.

[0066] In this embodiment, the specific process of the simulation analysis module analyzing the drying equipment based on the digital twin model is as follows: the total drying time of the drying equipment in this round of materials is marked as t3, and the manual participation time of the drying equipment in this round of materials is marked as t4; the automation degree of the drying equipment in this round of materials is calculated as X=(t3-t4) / t3; X is compared with a preset value x (the value interval of x is 0.7≤x≤0.9). If X>x, it means that the automation degree of the drying equipment is normal; if X≤x, it means that the automation degree of the drying equipment is abnormal; the exhaust gas emission per minute of the drying equipment in this round of materials is marked as Wi, i=1···n, and the wastewater emission per minute of the drying equipment in this round of materials is marked as Mi, i=1···n; each Wi and Mi is represented by a broken line graph, wherein each Wi forms a broken line one, and each Mi forms a broken line two; based on the broken line one and the broken line two, a graph P is obtained (the graph P refers to a closed graph formed by connecting the ends of the broken line one and the broken line two, such as Figure 2 As shown), P is compared with a preset graphic p (the preset graphic p is a graphic preset in advance by the staff as needed) for similarity. If the similarity between P and p is greater than a first preset value (the value range of the first preset value is greater than or equal to 85% and less than or equal to 95%), it means that the pollutant emissions of the drying equipment are normal. If the similarity between P and p is less than or equal to the first preset value, it means that the pollutant emissions of the drying equipment are abnormal. This setting can quickly analyze whether the degree of automation of the drying equipment is normal. It should be noted that during the normal operation of the drying equipment, there are very few or even no manual participation links. If the overall drying time includes too long a manual participation time, it means that the drying equipment has an abnormality and needs to be adjusted or remedied manually. In addition, this setting can also quickly analyze whether the pollutant emissions of the drying equipment are abnormal based on the exhaust gas emissions and wastewater emissions of the drying equipment. It should be noted that similarity comparison to calculate similarity is a prior art. Common methods for calculating the similarity of two graphics include:

[0067] 1. Template matching based method

[0068] TM standard square difference: This is one of the template matching methods in OpenCV. By using one of the graphics as a template, looking for similar parts in the other graphics, and calculating the square difference of the matching results. The smaller the matching result, the more similar the two graphics are.

[0069] TM correlation: It is also a template matching method in OpenCV. Unlike the standard squared difference, it calculates the correlation coefficient between the template and the original image. The larger the result value, the more similar the two images are.

[0070] TM correlation coefficient: This is also a method of template matching. Its calculation principle is similar to TM correlation, but it may involve more complex mathematical operations. The larger the result value, the more similar the two graphics are.

[0071] 2. Hash-based methods

[0072] Mean hashing (aHash): This method calculates the average pixel value of the image and converts it into a hash value for comparison. The smaller the Hamming distance of the hash value, the more similar the two images are. The mean hashing method is faster, but may not be very accurate.

[0073] Perceptual hashing (pHash): Perceptual hashing reduces the frequency of the graph through discrete cosine transform (DCT) and selects the matrix in the upper left corner to calculate the hash value. Compared with mean hashing, it has better robustness. Similarly, the smaller the Hamming distance of the hash value, the more similar the two graphs are.

[0074] 3. Histogram-based methods

[0075] Grayscale histogram: Grayscale histogram describes the general appearance and quality of a graphic by counting the number of pixels of different gray levels in the graphic. By comparing the overlap of the grayscale histograms of two graphics, their similarity can be evaluated. However, this method may ignore spatial information and is sensitive to graphic rotation and scaling.

[0076] Color histogram: Similar to the grayscale histogram, the color histogram compares each channel separately using the grayscale histogram method and then calculates the average. This method can take into account the color information of the image, but it may also be affected by factors such as image rotation and scaling.

[0077] 4. Other methods

[0078] MSE (mean square error): For a single-channel grayscale image, MSE expresses the difference between the predicted value and the true value by calculating the average square difference. The smaller the value, the more similar the two images are. However, MSE imposes a greater penalty on larger errors and is more sensitive to outliers.

[0079] SSIM (Structural Similarity): SSIM is a measure of the similarity between two images based on the comparison of brightness, contrast and structure. The closer the SSIM value is to 1, the more similar the two images are. The SSIM method has a high accuracy in evaluating image quality.

[0080] Cosine similarity: After representing the graphs as feature vectors, similarity is evaluated by calculating their cosine similarity. This method focuses on direction rather than size and is suitable for comparing the relative composition of graph content.

[0081] Euclidean distance: The pixel values ​​or feature vectors of the graphics are regarded as points in multidimensional space, and the difference between two points is measured by calculating the Euclidean distance between them. The smaller the distance, the more similar the graphics are. However, this method is sensitive to scale and rotation.

[0082] In this application, the TM standard square difference is used to calculate the similarity between P and p.

[0083] In this embodiment, the specific process of the simulation analysis module analyzing the screening equipment based on the digital twin model is: mark the total feed weight of the current round of materials of the screening equipment as G real-time, the weight of the screening product as Y real-time, the weight of the screened material as Z real-time, and the screening time as U real-time, and mark the total feed weight of each previous round of materials of the screening equipment as Gi, i=1···n; match Gi with G real-time, if Gi and G real-time match successfully (successful matching means that the total feed weight of Gi and G real-time at this time is the same), then Gi is planned to the fifth sequence, if Gi and G real-time match unsuccessfully (unsuccessful matching means that the total feed weight of Gi and G real-time at this time is the same), then Gi is planned to the fifth sequence, At this time, Gi and G real-time have different total feed weights), then Gi is planned to the sixth sequence; Gi in the fifth sequence is marked as gi, i=1···n, and the screening time corresponding to gi is marked as ui, the weight of the screening product is marked as yi, and the weight of the screened material is marked as zi; U real-time is matched with ui, if any ui matches U real-time, it means that the screening time of the screening equipment is normal, if any ui does not match U real-time, it means that the screening time of the screening equipment is abnormal; a screening effect analysis sub-model is constructed to analyze the difference between each round of materials corresponding to gi and the current round of materials, and output the analysis results. This setting can quickly analyze whether the screening time of the screening equipment is normal. The screening time will fluctuate due to differences in material properties (such as particle size distribution, humidity, viscosity, etc.). If this fluctuation is within the normal range allowed by the equipment design or operating specifications and does not affect the screening efficiency and product quality, it is usually not considered abnormal. If the screening time suddenly becomes significantly longer and exceeds the normal fluctuation range, this indicates that there is some abnormality. The abnormal causes may include screen blockage, screen wear, vibration motor failure, unstable feeding amount or feeding speed, significant changes in material properties (such as mixing of large pieces of material, increased humidity causing increased material viscosity, etc.), etc.

[0084] In this embodiment, the specific analysis process of the screening effect analysis sub-model is as follows: calculate the weight difference value Ri of the screening product = (Y real-time / yi); calculate the weight difference value RRi = (Z real-time / zi) of the screened material; calculate the material loss difference value RRRi = (G real-time - Y real-time - Z real-time) / (gi-yi-zi); obtain the radar chart Fi based on Ri, RRi, and RRRi (a radar chart is a graphical method for displaying multivariate data in the form of a two-dimensional chart of three or more quantitative variables represented on an axis starting from the same point. It starts from a central point and extends multiple rays outward from the central point, each ray representing a specific variable or indicator. The point or line segment on each ray represents the value or score of the variable in different dimensions. Radar charts are often used to compare the performance of multiple variables in different dimensions and to show the relative relationship between the variables. In the present application, Ri, RRi, and RRRi correspond to three rays respectively to form a radar chart Fi, such as Figure 3 As shown), and mark the preset radar chart as f (the preset radar chart f is a graphic preset in advance by the staff as needed); compare the similarity between Fi and f. If the similarity between Fi and f is less than the second preset value (the value range of the second preset value is greater than or equal to 80% and less than or equal to 90%), Fi is planned to the seventh sequence. If the similarity between Fi and f is greater than or equal to the second preset value, Fi is planned to the eighth sequence; count the number of Fi in the seventh sequence and mark it as h1, and count the number of Fi in the eighth sequence and mark it as h2; if (h1 / h2) is greater than the third preset value (the value range of the third preset value is greater than or equal to 0.15 and less than or equal to 0.45), it means that the screening effect of the screening equipment is abnormal; if (h1 / h2) is less than or equal to the third preset value, it means that the screening effect of the screening equipment is normal. This setting can quickly analyze whether the screening effect of the screening equipment is normal.

[0085] In the present embodiment, the specific process of the early warning classification module performing graded early warning is as follows: if the analysis result shows any one of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, then the output classification result is a first-level early warning; if the analysis result shows any two of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, then the output classification result is a second-level early warning; if the analysis result shows any three of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment , the output classification result is a three-level warning; if the analysis results show any four of the abnormal crushing time of the crushing equipment, the abnormal crushing effect of the crushing equipment, the abnormal degree of automation of the drying equipment, the abnormal pollutant emission of the drying equipment, the abnormal screening time of the screening equipment, and the abnormal screening effect of the screening equipment, the output classification result is a four-level warning; if the analysis results show any five of the abnormal crushing time of the crushing equipment, the abnormal crushing effect of the crushing equipment, the abnormal degree of automation of the drying equipment, the abnormal pollutant emission of the drying equipment, the abnormal screening time of the screening equipment, and the abnormal screening effect of the screening equipment, the output classification result is a five-level warning; if the analysis results show all of the abnormal crushing time of the crushing equipment, the abnormal crushing effect of the crushing equipment, the abnormal degree of automation of the drying equipment, the abnormal pollutant emission of the drying equipment, the abnormal screening time of the screening equipment, and the abnormal screening effect of the screening equipment, the output classification result is a six-level warning. This setting can help staff members clearly and intuitively understand the urgency of the situation.

[0086] In this embodiment, the notification module sends the analysis results and the warning classification results to the preset person in charge in the form of SMS notification.

[0087] In this embodiment, the data preprocessing and storage module performs information preprocessing including cleaning, integrating, converting and compressing the collected raw data to ensure the quality and consistency of the data.

[0088] In this embodiment, an interactive module is also included to provide an intuitive user interface to facilitate production management personnel to view the analysis results of the simulation analysis module and the classification results of the early warning classification module.

[0089] The present invention can effectively manage the early stage processing of raw coke, and then timely discover and issue early warnings when abnormalities occur. Specifically, the present application analyzes the early stage processing of raw coke from multiple angles from the crushing time and crushing effect of the crushing equipment, the degree of automation and pollutant emissions of the drying equipment, the screening time and screening effect of the screening equipment, so as to improve the accuracy of the management system.

[0090] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0091] What is described above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent production management system based on data twins, characterized in that: It includes crushing equipment data acquisition module, drying equipment data acquisition module, screening equipment data acquisition module, data preprocessing and storage module, digital twin modeling module, simulation analysis module, early warning classification module and notification module; The crushing equipment data acquisition module is used to collect the feeding time, feeding image, discharging time and discharging image information of the crushing equipment, and send the collected information to the data preprocessing and storage module; the drying equipment data acquisition module is used to collect the drying time, manual participation time, exhaust gas emission and wastewater emission information of the drying equipment, and send the collected information to the data preprocessing and storage module; the screening equipment data acquisition module is used to collect the total feed weight, screening product weight, screened material weight and screening time information of the screening equipment, and send the collected information to the data preprocessing and storage module; the data preprocessing and storage module is used to preprocess and store the received information, and send the preprocessed information to the digital twin modeling module; The digital twin modeling module constructs a digital twin model of the pre-treatment of needle coke calcination based on the pre-processed information; the simulation analysis module analyzes the crushing equipment, drying equipment, and screening equipment based on the digital twin model, and outputs the analysis results to the early warning classification module; the early warning classification module is used to perform graded early warning according to the analysis results, and send the classification results to the notification module; the notification module is used to issue early warning notifications of corresponding levels according to the classification results; The specific process of the simulation analysis module analyzing the crushing equipment based on the digital twin model is as follows: The feeding time of the material in this round of the crushing equipment is marked as T1, and the discharging time of the material in this round of the crushing equipment is marked as T2; Calculate the crushing time of this round of materials: t1 = (T2-T1); Compare the crushing time t1 with the preset time t2. If t1>t2, it means that the crushing time of the crushing equipment is abnormal. If t1≤t2, it means that the crushing time of the crushing equipment is normal. Construct image analysis sub-model and output analysis results; The specific analysis process of the image analysis sub-model is as follows: The feeding image of the material in this round of the crushing equipment is marked as Ai, i=1···n, n is a positive integer, and nine images are randomly selected from Ai and marked as ai, i=1···9 respectively; The discharge image of the material in this round of the crushing equipment is marked as Bi, i=1···n, n is a positive integer, and nine images are randomly selected from Bi and marked as bi, i=1···9 respectively; Compare ai and bi at pixel level, calculate the difference between ai and bi and mark it as Ci, i=1···9; Compare Ci with the preset value c. If Ci>c, Ci is planned to the first sequence. If Ci≤c, Ci is planned to the second sequence. The nine images bi are compared pairwise at pixel level, and the differences are calculated and marked as Di, i=1···36; Compare Di with the preset value d. If Di < d, Di is planned to the third sequence. If Di ≥ d, Di is planned to the fourth sequence. Count the number of Ci in the second sequence and mark it as v1, and count the number of Di in the fourth sequence and mark it as v2; Calculate the image difference weight value S = (v1*3+v2), compare S with the preset value s, if S>s, it means that the crushing effect of the crushing equipment is abnormal, if S≤s, it means that the crushing effect of the crushing equipment is normal; The specific process of the simulation analysis module analyzing the drying equipment based on the digital twin model is as follows: The total drying time of the drying equipment in this round of materials is marked as t3, and the manual participation time of the drying equipment in this round of materials is marked as t4; Calculate the automation degree of the drying equipment in this round of materials X = (t3-t4) / t3; Compare X with the preset value x. If X>x, it means that the automation level of the drying equipment is normal. If X≤x, it means that the automation level of the drying equipment is abnormal. The exhaust gas emission per minute of the drying equipment in this round of materials is marked as Wi, i=1···n, and the wastewater emission per minute of the drying equipment in this round of materials is marked as Mi, i=1···n; Each Wi and Mi is represented by a line graph, wherein each Wi forms a line 1 and each Mi forms a line 2; Based on broken line 1 and broken line 2, figure P is obtained, and the similarity between P and the preset figure p is compared. If the similarity between P and p is greater than a first preset value, it means that the pollutant emission of the drying equipment is normal. If the similarity between P and p is less than or equal to the first preset value, it means that the pollutant emission of the drying equipment is abnormal.

2. According to claim 1, an intelligent production management system based on data twins is characterized in that: The specific process of the simulation analysis module analyzing the screening equipment based on the digital twin model is as follows: The total weight of the feed of the current round of the screening equipment is marked as G real time, the weight of the screening product is marked as Y real time, the weight of the screened material is marked as Z real time, the screening time is marked as U real time, and the total weight of the feed of each round of the screening equipment in the past is marked as Gi, i=1···n; Match Gi with G in real time. If the real-time matching between Gi and G is successful, Gi is planned to the fifth sequence. If the real-time matching between Gi and G is unsuccessful, Gi is planned to the sixth sequence. The Gi in the fifth sequence is marked as gi, i = 1···n, and the sieving time corresponding to gi is marked as ui, the sieving product weight is marked as yi, and the weight of the oversize is marked as zi; Match Ureal time with ui. If any one of ui matches Ureal time, it means that the screening time of the screening device is normal. If any one of ui does not match Ureal time, it means that the screening time of the screening device is abnormal. Construct a screening effect analysis sub-model, analyze the difference between each round of materials corresponding to gi and the current round of materials, and output the analysis results.

3. According to claim 2, an intelligent production management system based on data twins is characterized in that: The specific analysis process of the screening effect analysis sub-model is as follows: Calculate the weight difference of the sieved product Ri=(Y real time / yi); Calculate the weight difference of the material on the sieve: RRi = (Z real time / zi); Calculate the material loss difference value RRRi=(G real-time-Y real-time-Z real-time) / (gi-yi-zi); Obtain radar chart Fi based on Ri, RRi, and RRRi, and mark the preset radar chart as f; Compare Fi with f for similarity. If the similarity between Fi and f is less than a second preset value, plan Fi to the seventh sequence. If the similarity between Fi and f is greater than or equal to the second preset value, plan Fi to the eighth sequence. Count the number of Fi in the seventh sequence and mark it as h1, count the number of Fi in the eighth sequence and mark it as h2; If (h1 / h2) is greater than the third preset value, it means that the screening effect of the screening device is abnormal; if (h1 / h2) is less than or equal to the third preset value, it means that the screening effect of the screening device is normal.

4. According to claim 3, an intelligent production management system based on data twins is characterized in that: The specific process of the early warning classification module for performing graded early warning is as follows: If the analysis results show any of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result is a first-level warning; If the analysis results show any two of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result is a level 2 warning; If the analysis results show any three of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result will be a level 3 warning; If the analysis results show any four of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result will be a level 4 warning; If the analysis results show any five of the following: abnormal crushing time of the crushing equipment, abnormal crushing effect of the crushing equipment, abnormal automation degree of the drying equipment, abnormal pollutant emission of the drying equipment, abnormal screening time of the screening equipment, and abnormal screening effect of the screening equipment, the output classification result will be a level 5 warning; If the analysis results show that all of the following items are abnormal in the crushing time of the crushing equipment, abnormal in the crushing effect of the crushing equipment, abnormal in the degree of automation of the drying equipment, abnormal in the amount of pollutant discharged by the drying equipment, abnormal in the screening time of the screening equipment, and abnormal in the screening effect of the screening equipment, the output classification result will be a level six warning.

5. According to claim 4, an intelligent production management system based on data twins is characterized in that: The notification module sends the analysis results and warning classification results to the preset person in charge in the form of SMS notification.

6. According to claim 5, an intelligent production management system based on data twins is characterized in that: The data preprocessing and storage module preprocesses information including cleaning, integrating, converting and compressing the collected raw data to ensure the quality and consistency of the data.

7. The intelligent production management system based on data twin according to claim 6 is characterized in that: It also includes an interactive module for providing an intuitive user interface to facilitate production managers to view the analysis results of the simulation analysis module and the classification results of the early warning classification module.

Images