Feed production quality management method and system

Abstract

The invention is suitable for the technical field of feed processing intelligent control and quality monitoring, and provides a feed production quality management method and system.The method comprises the steps that when it is determined that a particle swarm migration track deviation phenomenon exists in the falling and cooling process of a current batch of feed in a target gravity type countercurrent cooling device, the current batch of feed is subjected to particle swarm migration; historical cooling data of the cooling device and an original accumulation risk score value generated for the current batch of feed are obtained. The invention provides a quality management method for a feed cooling process, which can dynamically correct an original accumulation risk score value in combination with a track deviation behavior in a specific particle shape. An image comparison and trend extraction mechanism is introduced, a historical offset amplitude sequence under the current particle specification is obtained, the average slope of the change trend of the historical offset amplitude sequence is calculated to serve as a correction factor, and the response relation between the score value and the particle group movement trend is constructed.

Description

Technical Field

[0001] The present invention belongs to the technical field of feed processing intelligent control and quality monitoring, and in particular relates to a feed production quality management method and system. Background Art

[0002] In the existing feed processing and production process, after high-temperature granulation or puffing, granular feed usually needs to be cooled through a gravity countercurrent cooling device to reduce the temperature, stabilize the shape, and ensure the quality of subsequent packaging and storage. During this process, the feed particles fall from top to bottom into the cooling chamber by gravity, and the cold air introduced from the bottom of the device forms a countercurrent contact with the particles, thereby completing the heat exchange. However, due to the variety of particle types and shapes, problems such as local poor ventilation, uneven particle distribution, or accumulation and blockage often occur during the actual cooling process, seriously affecting the cooling efficiency and consistency of the finished product. To prevent such risks, some equipment has been equipped with an automatic scoring module to attempt to predict and evaluate the accumulation risk before cooling or in the early stages of cooling.

[0003] Existing methods for scoring pile-up risks are mostly based on static parameters, such as average particle size, material layer thickness, fan setting value, or production experience model, etc., and a static scoring value is generated through rule matching or experience mapping to guide operational adjustments. However, such methods often lack dynamic response to the actual movement behavior of the particle group. In particular, when faced with non-standard particles of special shapes or large particle sizes, the scoring results cannot effectively reflect their actual trajectory migration or local offset in the cooler, resulting in some risks being underestimated or adjustments being delayed. In addition, existing systems rarely use image information or historical operating data from the cooling process to make structured corrections to the scoring values, and lack a scoring compensation mechanism based on the evolution trend of particle group behavior, which limits the accuracy and adaptability of risk assessment. Summary of the Invention

[0004] The purpose of the present invention is to provide a feed production quality management method and system, aiming to solve the problems raised in the background technology.

[0005] The present invention is achieved by providing a feed production quality management method, the method comprising:

[0006] When it is determined that the particle migration trajectory of the current batch of feed deviates during the cooling process of the current batch of feed in the target gravity countercurrent cooling device, historical cooling data of the cooling device and an original accumulation risk score value generated for the current batch of feed are obtained;

[0007] Analyze historical cooling data and extract reference local cooling data where the geometric characteristics of several feed particles meet the preset matching conditions and there is deviation in the particle group migration trajectory;

[0008] For each reference local cooling data, extract the corresponding reference trajectory offset image and the specification information of the processed feed particles;

[0009] Determine the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sort the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data;

[0010] Determine whether there is a trend that the deviation amplitude increases with the increase of particle volume. If so, modify the original accumulation risk score value based on this trend.

[0011] As a further limitation of the technical solution of the embodiment of the present invention, the preset matching condition refers to: the similarity between the geometric shape characteristics of the feed particles processed corresponding to the reference local cooling data and the geometric shape characteristics of the current batch of feed particles is within a preset similarity range.

[0012] As a further limitation of the technical solution of the embodiment of the present invention, for each reference local cooling data, the steps of extracting the corresponding reference trajectory offset image and processing the feed particle specification information include:

[0013] Analyze each reference local cooling data in turn, obtain the image frame recorded therein when the deviation of the corresponding particle group migration trajectory reaches a peak, and use the image frame as the corresponding reference trajectory deviation image;

[0014] The specification information of the feed particles processed by each target gravity countercurrent cooling device recorded in the reference local cooling data is obtained, wherein the specification information includes an average particle size, a particle length, a particle width and a particle shape code.

[0015] As a further limitation of the technical solution of the embodiment of the present invention, the steps of determining the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sorting the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data include:

[0016] Based on the historical cooling data, a cooling process image is obtained that is consistent with the feed particle specifications processed in each reference local cooling data and does not deviate from the particle group migration trajectory during the cooling process, as the corresponding normal trajectory image;

[0017] Based on image registration and center of gravity shift analysis technology, the reference trajectory shift image and the corresponding normal trajectory image are compared to determine the deviation between the two in the center of gravity position of the particle group falling path;

[0018] The reference trajectory offset images are sorted according to the volume order of the feed particles corresponding to the reference local cooling data.

[0019] As a further limitation of the technical solution of the embodiment of the present invention, the step of determining whether there is a trend in which the deviation amplitude increases with the increase of particle volume, and if so, correcting the original accumulation risk score value according to the trend includes:

[0020] Determining whether, in the sorted reference trajectory offset images, there is a trend that as the volume of the feed particles gradually increases, the corresponding offset amplitude also gradually increases;

[0021] If it exists, then the average slope of each reference trajectory offset image is calculated based on the change trend of the offset amplitude in the sequence, and the average slope is used as the correction factor to correct the original accumulation risk score value.

[0022] If not present, the original stacking risk score value remains unchanged.

[0023] As a further limitation of the technical solution of the embodiment of the present invention, when correcting the original accumulation risk score value, a preset linear correction model is used;

[0024] The linear correction model is: ,in Refers to the modified stacking risk score value, Refers to the original stacked risk score value, Refers to the average slope of the change trend of the offset amplitude, is the adjustment factor, and Greater than 0.

[0025] A feed production quality management system, comprising: a data acquisition module, a data screening module, a data analysis module, an image sorting module, and a scoring correction module, wherein:

[0026] a data acquisition module for acquiring historical cooling data of the cooling device and an original accumulation risk score value generated for the current batch of feed when it is determined that there is a deviation in the migration trajectory of the particle group during the cooling process of the current batch of feed in the target gravity countercurrent cooling device;

[0027] A data screening module is configured to analyze historical cooling data and extract reference local cooling data for feed pellets whose geometric features meet a preset matching condition and for which deviations from the particle group migration trajectory are present; the preset matching condition being that the geometric features of the feed pellets processed by the reference local cooling data are similar to those of the feed pellets in the current batch within a preset similarity range;

[0028] A data analysis module is used to extract the corresponding reference trajectory offset image and process the feed particle specification information for each reference local cooling data;

[0029] An image sorting module is used to determine the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sort the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data;

[0030] The score correction module is used to determine whether there is a trend in which the deviation amplitude increases with the increase of particle volume. If so, the original accumulation risk score value is corrected according to the trend.

[0031] As a further limitation of the technical solution of the embodiment of the present invention, the data parsing module specifically includes:

[0032] A reference image extraction unit is used to analyze each reference local cooling data in turn, obtain an image frame recorded therein when the deviation of the corresponding particle group migration trajectory reaches a peak, and use the image frame as the corresponding reference trajectory deviation image;

[0033] The particle information acquisition unit is used to obtain the specification information of the feed particles processed by each target gravity countercurrent cooling device recorded in the reference local cooling data, wherein the specification information includes the average particle size, particle length, particle width and particle shape code.

[0034] As a further limitation of the technical solution of the embodiment of the present invention, the image sorting module specifically includes:

[0035] A normal image extraction unit is used to obtain, based on the historical cooling data, a cooling process image that is consistent with the feed particle specifications processed in each reference local cooling data and does not deviate from the particle group migration trajectory during the cooling process, as a corresponding normal trajectory image;

[0036] A deviation amplitude calculation unit is used to compare the reference trajectory deviation image with the corresponding normal trajectory image based on image registration and center of gravity deviation analysis technology to determine the deviation amplitude of the two at the center of gravity position of the particle group falling path;

[0037] The offset image sorting unit is used to sort the reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data.

[0038] As a further limitation of the technical solution of the embodiment of the present invention, the score correction module specifically includes:

[0039] a change trend analysis unit for determining, in the sorted reference trajectory offset images, whether there is a change trend in which the corresponding offset amplitude gradually increases as the feed particle volume gradually increases;

[0040] A correction factor selection unit, if present, is used to calculate the average slope of each reference trajectory offset image based on the order of particle volume from small to large, based on the trend of the offset amplitude in the sequence, and use the average slope as a correction factor to correct the original accumulation risk score value;

[0041] A score value maintaining unit, used for keeping the original stacked risk score value unchanged if it does not exist;

[0042] When correcting the original accumulation risk score value, a preset linear correction model is used;

[0043] The linear correction model is: ,in Refers to the modified stacking risk score value, Refers to the original stacked risk score value, Refers to the average slope of the change trend of the offset amplitude, is the adjustment factor, and Greater than 0.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] The present invention proposes a quality management method for the feed cooling process, which can dynamically correct the original accumulation risk score value in combination with the trajectory deviation behavior under a specific particle shape. By introducing an image comparison and trend extraction mechanism, the historical deviation amplitude sequence under the current particle specification is obtained, and the average slope of its changing trend is calculated as a correction factor to construct a response relationship between the score value and the particle group movement trend. The score correction model of the present invention has good structural adaptability and trend sensitivity, and can achieve a more refined and dynamic expression of the degree of accumulation risk in the cooling process, which helps to improve the response accuracy of the cooling control strategy and the cooling consistency of the finished particle product. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A flowchart of a method provided by an embodiment of the present invention;

[0047] Figure 2 A flowchart of obtaining a reference trajectory offset image in the method provided in an embodiment of the present invention;

[0048] Figure 3 A flowchart of sorting a plurality of reference trajectory offset images in the method provided in an embodiment of the present invention;

[0049] Figure 4 A flowchart of correcting the original accumulation risk score value in the method provided in an embodiment of the present invention;

[0050] Figure 5 An application architecture diagram of the system provided by an embodiment of the present invention;

[0051] Figure 6 A structural block diagram of a data parsing module in a system provided by an embodiment of the present invention;

[0052] Figure 7 A structural block diagram of an image sorting module in a system provided by an embodiment of the present invention;

[0053] Figure 8 This is a structural block diagram of the score correction module in the system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0055] Figure 1 A flow chart of a method provided by an embodiment of the present invention is shown.

[0056] Specifically, a feed production quality management method comprises the following steps:

[0057] Step S100: When it is determined that the particle migration trajectory of the current batch of feed deviates during the cooling process of the current batch of feed in the target gravity countercurrent cooling device, historical cooling data of the cooling device and an original accumulation risk score value generated for the current batch of feed are obtained.

[0058] In the embodiments of the present invention, the gravity-type countercurrent cooling device generally refers to a cooling device used in feed pellet production lines. Its operating principle is as follows: After undergoing high-temperature processing such as pelleting or puffing, feed pellets flow downward by gravity into the cooler. Cool air is introduced from the bottom of the device for countercurrent cooling, which cools the pellets and stabilizes their shape through heat exchange. This device has a simple structure and low energy consumption, making it a widely used pellet cooling method in the feed processing industry.

[0059] The so-called "deviated particle migration trajectory during the drop cooling process" refers to the phenomenon in which particles in the cooling device, while supposed to follow a nearly symmetrical, vertically distributed flow path, experience spatial displacement, accumulation, skewness, or agglomeration. This causes particles to concentrate in specific areas within the cooler and no longer flow evenly across the cold air flow zone. This migration deviation can lead to localized ventilation obstruction, insufficient particle cooling, or accumulation and blockage, and is a significant risk factor affecting cooling uniformity and efficiency.

[0060] The original accumulation risk score value refers to the scoring parameter generated by the equipment control system or the upper data platform before or during the cooling of the current batch of feed, which is used to evaluate the risk of particle accumulation or abnormal flow inside the cooler. This score value is usually generated based on parameters such as particle size, raw material density, feeding rate, wind pressure setting, combined with empirical models or rule logic, to prompt operation and maintenance personnel or automatic control systems to take response measures such as ventilation adjustment and material guidance regulation. Some intelligent cooling equipment or industrial automation systems in the existing technology already have this type of scoring model, and often use rule-based static evaluation methods or expert system-based logical reasoning methods to calculate the initial risk value.

[0061] Historical cooling data primarily comes from the automatic recording systems used during cooling equipment operation, including but not limited to image acquisition modules, temperature and humidity sensors, wind pressure recorders, and pellet feed rate recorders. Data is typically automatically collected by the equipment's PLC or edge computing module and uploaded to a data platform, where it is stored in batches or time series. Historical cooling data should include at least the following information: cooling time period, particle image sequences (for trajectory analysis), particle physical parameters (such as volume and shape characteristics), ventilation parameters (such as wind speed and pressure), cooler configuration, ambient temperature and humidity, and a record of the current batch's accumulation risk score.

[0062] Furthermore, the feed production quality management method further comprises the following steps:

[0063] Step S200 , analyzing historical cooling data, extracting reference local cooling data of several feed particles whose geometric features meet preset matching conditions and have deviations from the particle group migration trajectory.

[0064] The preset matching condition means that the similarity between the geometrical shape characteristics of the feed particles processed corresponding to the reference local cooling data and the geometrical shape characteristics of the feed particles of the current batch is within a preset similarity range.

[0065] In an embodiment of the present invention, reference local cooling data is extracted, which shows that the geometric features of several feed particles meet preset matching conditions and there is a deviation in the migration trajectory of the particle group. This is intended to provide a control data source with strong comparability and consistent structural characteristics for the trend analysis of the deviation amplitude of the current batch of feed. Since the geometric shape of the particles has a significant impact on their movement state in the cooling device, different particle shapes may show completely different trajectory distributions and deviation behaviors. Therefore, only when the geometric shape features are close enough, the historical trajectory image data has a reference significance for the current batch comparison and trend analysis. If historical samples with excessive structural differences are blindly introduced, the impact of particle volume changes on the deviation phenomenon will not be accurately reflected, and may even mislead subsequent trend judgments and risk score corrections.

[0066] The process for determining whether the geometric features of the feed pellets processed in the reference local cooling data are similar to those of the current batch of feed pellets is within a preset similarity range. This process utilizes a technique based on image shape feature matching. Specifically, the image acquisition system first acquires representative morphological images of the current batch of feed pellets and those of previous batches. A corresponding geometric feature vector is generated using a shape contour extraction algorithm. Standard shape descriptors, such as Hu moments, Zernike moments, and boundary curvature histograms, can be used. The similarity between the current and historical particle shape feature vectors is then calculated. Common similarity metrics include cosine similarity, inverse normalized Euclidean distance, and structural similarity indices. Based on a set similarity threshold (e.g., 0.85 to 1.0), the system selects historical data with similarities that meet the preset criteria. The system then determines whether these historical data exhibit deviations from the particle migration trajectory. Ultimately, a reference local cooling data set that meets these two criteria is constructed. This process ensures that the underlying samples used for the offset image sorting in subsequent analysis are comparable in terms of structural dimensions, thereby ensuring the accuracy and interpretability of trend analysis.

[0067] Furthermore, the feed production quality management method further comprises the following steps:

[0068] In step S300 , for each reference local cooling data, the corresponding reference trajectory offset image and the specification information of the processed feed particles are extracted.

[0069] Specifically, Figure 2 A flow chart for obtaining a reference trajectory offset image is shown.

[0070] For each reference local cooling data, extracting the corresponding reference trajectory offset image and processing feed particle specification information specifically includes the following steps:

[0071] Step S301, analyzing each reference local cooling data in sequence, obtaining an image frame recorded therein when the deviation of the corresponding particle group migration trajectory reaches a peak, and using the image frame as the corresponding reference trajectory deviation image;

[0072] Step S302: Obtain specification information of feed particles processed by each target gravity countercurrent cooling device recorded in the reference local cooling data, wherein the specification information includes average particle size, particle length, particle width, and particle shape code.

[0073] In an embodiment of the present invention, in an embodiment of the present invention, in order to determine the image frame when the offset of the migration trajectory of the particle group reaches a peak, the system first analyzes the image sequence recorded in the reference local cooling data frame by frame, and based on the image registration and center of gravity offset analysis method, extracts the trajectory distribution state of the feed particle group in each frame of the image in the cooler space. Specifically, by setting a reference symmetry axis or a theoretical falling trajectory axis, the spatial offset distance between the center of gravity of the particle distribution in the image and the theoretical axis is calculated to form an offset amplitude change curve under the frame sequence. The system further identifies the local extreme value of the curve, determines the time point corresponding to the frame where the offset amplitude reaches the maximum value, and uses the image frame as the peak image representing the offset of the migration trajectory of the particle group of the reference data. The image frame has the technical characteristics of strong representativeness, significant offset phenomenon, and can be used for trend analysis, so it is selected as a reference trajectory offset image to participate in subsequent comparison and sorting.

[0074] The obtained feed pellet specification information, including average particle size, length, width and particle shape coding, is mainly used to estimate the volume characteristics of the particles in the current reference data. In the present invention, particle volume is an important physical parameter that affects the movement behavior of the particle group and is a key variable for judging the trend of the deviation amplitude change. Based on the particle geometric parameters and the corresponding shape type, the approximate volume value of the particle can be calculated using the volume estimation model. For example, for cylindrical or ellipsoidal particles, the following can be used respectively: or Volume estimation is performed using models such as these. This provides a reliable quantitative basis for subsequent sorting of trajectory offset images by particle size. This volume information is not only used for trend analysis but also serves as a unified basis for comparison in multiple processing steps, such as image matching and correction factor generation, ensuring accuracy and consistency throughout the entire scoring and correction process.

[0075] Furthermore, the feed production quality management method further comprises the following steps:

[0076] Step S400 : determining the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sorting the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data.

[0077] Specifically, Figure 3 A flow chart for sorting several reference trajectory offset images is shown.

[0078] The steps of determining the offset amplitude of each reference trajectory offset image compared to the normal trajectory image and sorting the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data specifically include the following steps:

[0079] Step S401: Based on the historical cooling data, a cooling process image is obtained, which has the same feed particle specifications as those processed in each reference local cooling data and does not show any deviation from the particle group migration trajectory during the cooling process, as the corresponding normal trajectory image;

[0080] Step S402 , based on image registration and gravity center offset analysis technology, compare the reference trajectory offset image with the corresponding normal trajectory image to determine the deviation between the two in the gravity center position of the particle group falling path;

[0081] Step S403 : sorting the reference trajectory offset images according to the volume order of the feed particles corresponding to the reference local cooling data.

[0082] In an embodiment of the present invention, the selection of the normal trajectory image is based on screening of historical cooling data that is consistent with the particle specifications in the current reference local cooling data, and the cooling process corresponding to the selected image does not show deviation in the particle group migration trajectory. Specifically, the system analyzes the distribution state of the particle trajectory in the historical image sequence, screens out those image data with obvious center of gravity offset, trajectory deviation or concentrated accumulation area, and only retains image frames with relatively symmetrical and uniform distribution of particle falling paths as representative images of stable and non-offset trajectories. Under the premise that the particle specifications match the reference data, the selected normal trajectory image can provide a reference trajectory morphology with strong reference significance, provide a comparison basis for the subsequent calculation of the offset amplitude, and ensure that the offset measurement has a unified reference coordinate system and comparison baseline.

[0083] In step S402, the system first uses image registration technology to spatially align the reference trajectory offset image with its corresponding normal trajectory image to eliminate interference caused by shooting angle, image size, or position errors. After registration, the system identifies the pixel density or edge distribution area of ​​the particle group in the image and calculates its center of gravity position in the image coordinate system. The center of gravity coordinates of the particle group in the reference image and the normal image are extracted respectively, and the deviation amplitude of the two on the falling path is calculated based on the Euclidean distance or vertical offset distance. This deviation amplitude is a key quantitative indicator to characterize the degree of particle group trajectory deviation and is used in the subsequent trend identification and scoring correction process.

[0084] In step S403, the system sorts the multiple reference trajectory offset images based on the particle volume information extracted in step S302. The specific process involves inputting the particle size parameters from each reference local cooling data set into the volume estimation model to generate corresponding volume values; binding the reference images to the corresponding volume values ​​to form an image-volume comparison table; and then rearranging the images in ascending order of volume to form an ordered image sequence. This sorted sequence serves as the basic input for subsequent identification of the trend of offset amplitude versus volume, ensuring strict comparability and continuity of the horizontal axis variables relied upon for trend analysis.

[0085] Furthermore, the feed production quality management method further comprises the following steps:

[0086] Step S500 , determining whether there is a trend that the deviation amplitude increases with the increase of particle volume; if so, revising the original accumulation risk score value according to the trend.

[0087] Specifically, Figure 4 A flow chart for modifying the original stacked risk score value is shown.

[0088] The determination of whether there is a trend that the deviation amplitude increases with the increase of particle volume, and if so, the correction of the original accumulation risk score value according to the trend specifically includes the following steps:

[0089] Step S501, determining whether there is a trend in which the corresponding offset amplitude gradually increases as the volume of the feed pellet gradually increases in the sorted reference trajectory offset images;

[0090] Step S502: If so, the reference trajectory offset images are arranged in ascending order of particle volume. Based on the trend of the offset amplitude in the sequence, the average slope is calculated and used as a correction factor to correct the original accumulation risk score.

[0091] Step S503: If it does not exist, the original accumulation risk score value remains unchanged.

[0092] When correcting the original accumulation risk score value, a preset linear correction model is used;

[0093] The linear correction model is: ,in Refers to the modified stacking risk score value, Refers to the original stacked risk score value, Refers to the average slope of the change trend of the offset amplitude, is the adjustment factor, and Greater than 0.

[0094] In an embodiment of the present invention, step S501 is specifically implemented as follows: the system sequentially analyzes the reference trajectory offset images, which are arranged in order of particle volume, extracts the offset amplitude corresponding to each image, and establishes a one-to-one correspondence between particle volume and offset amplitude. This data set is mapped into a two-dimensional coordinate point sequence, with particle volume on the X-axis and offset amplitude on the Y-axis, forming a line graph with discrete point features. The system further determines whether the Y value in the line graph shows an overall upward trend with the X value. Specifically, methods such as local monotonic segment identification, determination of the proportion of positive slopes in the sliding window, or determination of the trend of first-order derivatives can be used to comprehensively determine whether there is a trend in which the offset amplitude increases with increasing particle volume. To ensure the stability and comparability of trend determination, this embodiment requires consistent X-axis volume intervals. That is, the particle size is graded and normalized during reference image selection to prevent irregular changes in particle volume from affecting trend identification results.

[0095] In step S502, if the above trend is confirmed, the system will calculate the average slope of the series based on the above line chart using methods such as linear regression or differential averaging. This average slope is then used as a correction factor to adjust the original accumulation risk score. The introduction of this correction factor reflects the increasing risk of actual cooling offset caused by increased particle volume, and the initial score should be enhanced to improve the foresight and sensitivity of the risk prediction.

[0096] Using the average slope of the deviation amplitude change trend as a correction factor has the following beneficial effects: first, it provides a mechanism for dynamically calibrating the original accumulation risk score value based on the historical trajectory behavior pattern corresponding to the geometric shape characteristics of the current batch of feed pellets; second, the slope factor reflects whether the trajectory deviation amplitude caused by the volume increase has an increasing trend under the current particle shape type, and has clear structural adaptability and interpretability; third, correction based on the trend slope can provide quantitative feedback on the existence of volume-sensitive deviation behavior under the current particle shape, thereby improving the applicability and accuracy of the original accumulation risk score under specific feed types, and avoiding the situation where the existing scoring model underestimates the risk due to ignoring the particle group migration trend when facing "non-standard shape particles" or "niche specification particles".

[0097] The linear correction model provided by the present invention is an intuitive, easy-to-deploy, and low-computational-cost scoring adjustment method, and is particularly suitable for cooling control systems that need to be run in real time in industrial sites. In addition to the linear correction model, other methods such as polynomial fitting models, exponential adjustment models, segmented interval correction models, and weighted historical sample regression models can also be used to correct the original score values. For example, a segmented risk enhancement coefficient table can be constructed based on volume classification, or a nonlinear risk mapping function can be trained using the actual accumulation results of historical samples to further improve the accuracy and dynamics of the score adjustment. These methods can all be used as equivalent implementations of the present invention.

[0098] Further, Figure 5 The application architecture diagram of the system provided by the embodiment of the present invention is shown.

[0099] Among them, in another preferred embodiment provided by the present invention, a feed production quality management system includes:

[0100] The data acquisition module 100 is used to obtain historical cooling data of the cooling device and the original accumulation risk score value generated for the current batch of feed when it is determined that there is a deviation in the particle migration trajectory of the current batch of feed during the falling cooling process in the target gravity countercurrent cooling device.

[0101] Furthermore, the feed production quality management system also includes:

[0102] The data screening module 200 is used to analyze historical cooling data and extract reference local cooling data whose geometric features of several feed particles meet preset matching conditions and have deviations from the particle group migration trajectory; the preset matching conditions are: the similarity between the geometric features of the feed particles processed by the reference local cooling data and the geometric features of the feed particles in the current batch is within a preset similarity range.

[0103] Furthermore, the feed production quality management system also includes:

[0104] The data analysis module 300 is used to extract the corresponding reference trajectory offset image and process the feed particle specification information for each reference local cooling data.

[0105] Specifically, Figure 6 FIG. 3 is a block diagram showing the structure of the data analysis module 300 in the system provided by an embodiment of the present invention.

[0106] In a preferred embodiment of the present invention, the data parsing module 300 specifically includes:

[0107] The reference image extraction unit 301 is used to analyze each reference local cooling data in turn, obtain the image frame recorded therein when the deviation of the corresponding particle group migration trajectory reaches a peak, and use the image frame as the corresponding reference trajectory deviation image;

[0108] The particle information acquisition unit 302 is used to acquire the specification information of the feed particles processed by each target gravity countercurrent cooling device recorded in the reference local cooling data, wherein the specification information includes the average particle size, particle length, particle width and particle shape code.

[0109] Furthermore, the feed production quality management system also includes:

[0110] The image sorting module 400 is used to determine the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sort the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data.

[0111] Specifically, Figure 7 FIG. 4 is a block diagram showing a structure of an image sorting module 400 in a system provided by an embodiment of the present invention.

[0112] In a preferred embodiment of the present invention, the image sorting module 400 specifically includes:

[0113] The normal image extraction unit 401 is used to obtain, based on the historical cooling data, a cooling process image that is consistent with the feed particle specifications processed in each reference local cooling data and does not deviate from the particle group migration trajectory during the cooling process, as a corresponding normal trajectory image;

[0114] The deviation amplitude calculation unit 402 is used to compare the reference trajectory deviation image with the corresponding normal trajectory image based on image registration and center of gravity deviation analysis technology to determine the deviation amplitude of the two at the center of gravity position of the particle group falling path;

[0115] The offset image sorting unit 403 is used to sort the reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data.

[0116] Furthermore, the feed production quality management system also includes:

[0117] The score correction module 500 is used to determine whether there is a trend that the deviation amplitude increases with the increase of particle volume. If so, the original accumulation risk score value is corrected according to the trend.

[0118] Specifically, Figure 8 FIG. 5 shows a structural block diagram of the score modification module 500 in the system provided by an embodiment of the present invention.

[0119] In a preferred embodiment of the present invention, the score correction module 500 specifically includes:

[0120] A change trend analysis unit 501 is used to determine whether, in the sorted reference trajectory offset images, there is a change trend that as the feed particle volume gradually increases, the corresponding offset amplitude also gradually increases;

[0121] A correction factor selection unit 502 is configured to, if present, calculate an average slope of each reference trajectory offset image based on the order of particle volume from smallest to largest, based on the trend of the offset amplitude in the sequence, and use the average slope as a correction factor to correct the original accumulation risk score;

[0122] The score value maintaining unit 503 is used to keep the original stacked risk score unchanged if it does not exist;

[0123] When correcting the original accumulation risk score value, a preset linear correction model is used;

[0124] The linear correction model is: ,in Refers to the modified stacking risk score value, Refers to the original stacked risk score value, Refers to the average slope of the change trend of the offset amplitude, is the adjustment factor, and Greater than 0.

[0125] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0126] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When executed, the program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0127] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0128] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0129] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A feed production quality management method, characterized in that: The method comprises: When it is determined that the particle migration trajectory of the current batch of feed deviates during the cooling process of the current batch of feed in the target gravity countercurrent cooling device, historical cooling data of the cooling device and an original accumulation risk score value generated for the current batch of feed are obtained; Analyze historical cooling data and extract reference local cooling data where the geometric characteristics of several feed particles meet the preset matching conditions and there is deviation in the particle group migration trajectory; For each reference local cooling data, extract the corresponding reference trajectory offset image and the specification information of the processed feed particles; Determine the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sort the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data; Determine whether there is a trend that the deviation amplitude increases with the increase of particle volume. If so, modify the original accumulation risk score value based on this trend.

2. The feed production quality management method according to claim 1, characterized in that: The preset matching condition means that the similarity between the geometrical shape characteristics of the feed particles processed corresponding to the reference local cooling data and the geometrical shape characteristics of the feed particles of the current batch is within a preset similarity range.

3. The feed production quality management method according to claim 2, characterized in that: For each reference local cooling data, the steps of extracting the corresponding reference trajectory offset image and processing the feed particle specification information include: Analyze each reference local cooling data in turn, obtain the image frame recorded therein when the deviation of the corresponding particle group migration trajectory reaches a peak, and use the image frame as the corresponding reference trajectory deviation image; The specification information of the feed particles processed by each target gravity countercurrent cooling device recorded in the reference local cooling data is obtained, wherein the specification information includes an average particle size, a particle length, a particle width and a particle shape code.

4. The feed production quality management method according to claim 3, characterized in that: The steps of determining the offset amplitude of each reference trajectory offset image compared to the normal trajectory image and sorting the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data include: Based on the historical cooling data, a cooling process image is obtained that is consistent with the feed particle specifications processed in each reference local cooling data and does not deviate from the particle group migration trajectory during the cooling process, as the corresponding normal trajectory image; Based on image registration and center of gravity shift analysis technology, the reference trajectory shift image and the corresponding normal trajectory image are compared to determine the deviation between the two in the center of gravity position of the particle group falling path; The reference trajectory offset images are sorted according to the volume order of the feed particles corresponding to the reference local cooling data.

5. The feed production quality management method according to claim 4, characterized in that: Determining whether there is a trend that the deviation amplitude increases with the increase of particle volume, and if so, revising the original accumulation risk score value based on the trend includes the following steps: Determining whether, in the sorted reference trajectory offset images, there is a trend that as the volume of the feed particles gradually increases, the corresponding offset amplitude also gradually increases; If it exists, then the average slope of each reference trajectory offset image is calculated based on the change trend of the offset amplitude in the sequence, and the average slope is used as the correction factor to correct the original accumulation risk score value. If not present, the original stacking risk score value remains unchanged.

6. The feed production quality management method according to claim 5, characterized in that: When correcting the original accumulation risk score value, a preset linear correction model is used; The linear correction model is: ,in Refers to the modified stacking risk score value, Refers to the original stacked risk score value, Refers to the average slope of the change trend of the offset amplitude, is the adjustment factor, and Greater than 0.

7. A feed production quality management system, characterized in that: The system includes: a data acquisition module, a data screening module, a data analysis module, an image sorting module and a score correction module, wherein: a data acquisition module for acquiring historical cooling data of the cooling device and an original accumulation risk score value generated for the current batch of feed when it is determined that there is a deviation in the migration trajectory of the particle group during the cooling process of the current batch of feed in the target gravity countercurrent cooling device; A data screening module is configured to analyze historical cooling data and extract reference local cooling data for feed pellets whose geometric features meet a preset matching condition and for which deviations from the particle group migration trajectory are present; the preset matching condition being that the geometric features of the feed pellets processed by the reference local cooling data are similar to those of the feed pellets in the current batch within a preset similarity range; A data analysis module is used to extract the corresponding reference trajectory offset image and process the feed particle specification information for each reference local cooling data; An image sorting module is used to determine the offset amplitude of each reference trajectory offset image compared to the normal trajectory image, and sort the corresponding reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data; The score correction module is used to determine whether there is a trend in which the deviation amplitude increases with the increase of particle volume. If so, the original accumulation risk score value is corrected according to the trend.

8. The feed production quality management system according to claim 7, characterized in that: The data analysis module specifically includes: A reference image extraction unit is used to analyze each reference local cooling data in turn, obtain an image frame recorded therein when the deviation of the corresponding particle group migration trajectory reaches a peak, and use the image frame as the corresponding reference trajectory deviation image; The particle information acquisition unit is used to obtain the specification information of the feed particles processed by each target gravity countercurrent cooling device recorded in the reference local cooling data, wherein the specification information includes the average particle size, particle length, particle width and particle shape code.

9. The feed production quality management system according to claim 8, characterized in that: The image sorting module specifically includes: A normal image extraction unit is used to obtain, based on the historical cooling data, a cooling process image that is consistent with the feed particle specifications processed in each reference local cooling data and does not deviate from the particle group migration trajectory during the cooling process, as a corresponding normal trajectory image; A deviation amplitude calculation unit is used to compare the reference trajectory deviation image with the corresponding normal trajectory image based on image registration and center of gravity deviation analysis technology to determine the deviation amplitude of the two at the center of gravity position of the particle group falling path; The offset image sorting unit is used to sort the reference trajectory offset images according to the volume order of the feed particles corresponding to each reference local cooling data.

10. The feed production quality management system according to claim 9, characterized in that: The score correction module specifically includes: a change trend analysis unit for determining, in the sorted reference trajectory offset images, whether there is a change trend in which the corresponding offset amplitude gradually increases as the feed particle volume gradually increases; A correction factor selection unit, if present, is used to calculate the average slope of each reference trajectory offset image based on the order of particle volume from small to large, based on the trend of the offset amplitude in the sequence, and use the average slope as a correction factor to correct the original accumulation risk score value; A score value maintaining unit, used for keeping the original stacked risk score value unchanged if it does not exist; When correcting the original accumulation risk score value, a preset linear correction model is used; The linear correction model is: ,in Refers to the modified stacking risk score value, Refers to the original stacked risk score value, Refers to the average slope of the change trend of the offset amplitude, is the adjustment factor, and Greater than 0.

Images