Automatic control system and method in livestock feed processing
An AI-driven automated control system for feed processing optimizes pelleting by adjusting ring die speed based on real-time monitoring of temperature, humidity, and flow rate, enhancing efficiency and quality while reducing energy use.
Patent Information
- Application Number
- CN202410566495.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-09
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-05-09
AI Technical Summary
The traditional pelleting process in feed processing relies on manual experience, leading to unstable quality, high energy consumption, and low production efficiency, which cannot meet the needs of modern production.
An automated control system using artificial intelligence based on deep learning to monitor and analyze material temperature, humidity, flow rate, and ring die speed in the ring die pelletizer, mining particle state and equipment operation characteristics to adaptively adjust the ring die speed for optimal pelleting.
Improves feed processing efficiency and pelletizing quality while reducing energy consumption through intelligent optimization and regulation of the pelleting process.
Smart Images

Figure CN118348865B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent control technology, and more specifically, to an automated control system and method in a livestock feed processing process. Background Art
[0002] The pelleting process of feed is one of the key steps in the feed processing process, which directly affects the nutritional value, digestibility, and convenience of storage and transportation of feed. The ring die pelletizer is a key equipment in the feed processing process, which squeezes the feed raw materials into pellets through mechanical action. When the raw materials enter the machine, the ring die rotates under the driving force of the main force, and the material layer is taken into the extrusion area. The pressure roller also rotates with the friction in the extrusion area. With the rotation of the ring die and the pressure roller, the extrusion force and material density gradually increase, and the material is squeezed into the die hole. The material in the die hole is continuously squeezed out of the die hole after being formed, and is cut into cylindrical pellets by the cutter. In this process, the material generates heat through mechanical friction, and a certain amount of heat is generated in the extrusion area between the ring die and the pressure roller, which helps to increase the temperature of the material and promote the conditioning and forming of the material.
[0003] With the rapid development of animal husbandry, the feed processing industry is facing multiple challenges of improving production efficiency, optimizing product quality and reducing energy consumption. The traditional pelleting process mostly relies on manual experience to adjust parameters, which is difficult to achieve precise control. There are problems such as unstable pelleting quality, high energy consumption, and low production efficiency, which cannot meet the needs of modern production. Therefore, an automated control system and method in the livestock feed processing process are expected. Summary of the invention
[0004] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an automated control system and method in the livestock feed processing process, which uses artificial intelligence technology based on deep learning to monitor and analyze the material temperature, material humidity, material flow rate and ring die speed in the ring die pelletizer, and mines the feed particle state characteristics from the material temperature and humidity data, and mines the equipment operation state characteristics from the material flow rate and ring die speed data, so as to adaptively adjust the ring die speed based on the timing response characteristics of the particle state relative to the equipment operation state to optimize the pelletizing effect. In this way, the intelligent optimization and regulation of the feed processing pelletizing process can be achieved to improve the feed processing efficiency and pelletizing quality, and reduce energy consumption.
[0005] Accordingly, according to one aspect of the present application, an automated control system for livestock feed processing is provided, comprising:
[0006] The ring die pelletizing monitoring module is used to collect material temperature values, material humidity values, material flow values and ring die speed values at multiple predetermined time points within a predetermined time period by a sensor group deployed in the ring die pelletizing machine;
[0007] A time series correlation analysis module is used to extract material temperature and humidity time series correlation characteristics and flow rate and speed time series correlation characteristics from the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points to obtain a material particle state comprehensive characterization feature vector and a device operation state comprehensive characterization feature vector;
[0008] A time series interactive response feature extraction module is used to calculate the responsiveness estimate of the comprehensive characterization feature vector of the material particle state relative to the comprehensive characterization feature vector of the equipment operation state to obtain a particle mass-equipment operation state time series interactive response feature matrix;
[0009] The ring die speed control module is used to determine the control instruction of the ring die speed based on the particle mass-equipment operation state timing interaction response characteristic matrix.
[0010] In the above-mentioned automated control system in the livestock feed processing process, the timing association analysis module includes: a timing arrangement unit, which is used to arrange the material temperature values, material humidity values, material flow values and ring die speed values at the multiple predetermined time points into material temperature timing input vectors, material humidity timing input vectors, material flow timing input vectors and ring die speed timing input vectors according to the time dimension; a particle state comprehensive characterization unit, which is used to perform timing association encoding on the material temperature timing input vector and the material humidity timing input vector to obtain the material particle state comprehensive characterization feature vector; an equipment operation state comprehensive characterization unit, which is used to perform timing association encoding on the material flow timing input vector and the ring die speed timing input vector to obtain the equipment operation state comprehensive characterization feature vector.
[0011] In the above-mentioned automated control system during livestock feed processing, the particle state comprehensive characterization unit is used to: arrange the material temperature time series input vector and the material humidity time series input vector in two dimensions into a material temperature and humidity time series input matrix, and then pass it through a time series association encoder based on a convolutional neural network model to obtain the material particle state comprehensive characterization feature vector.
[0012] In the above-mentioned automated control system in the livestock feed processing process, the equipment operation status comprehensive characterization unit is used to: arrange the material flow timing input vector and the ring die speed timing input vector in two dimensions into a flow speed timing input matrix and then pass it through the timing association encoder based on the convolutional neural network model to obtain the equipment operation status comprehensive characterization feature vector.
[0013] In the above-mentioned automated control system in the livestock feed processing process, the timing interaction response feature extraction module is used to: multiply the material particle state comprehensive characterization feature vector and the inverse vector of the equipment operation state comprehensive characterization feature vector to obtain the particle quality-equipment operation state timing interaction response feature matrix.
[0014] In the above-mentioned automated control system in the livestock feed processing process, the ring die speed control module includes: an optimization unit, which is used to perform label position sensitivity analysis on the particle quality-equipment operating status timing interaction response feature matrix to obtain an optimized particle quality-equipment operating status timing interaction response feature matrix; a control instruction generation unit, which is used to pass the optimized particle quality-equipment operating status timing interaction response feature matrix through a classifier-based ring die speed controller to obtain the control instruction, and the control instruction is used to indicate whether the ring die speed should be increased, decreased or maintained at the current time point.
[0015] In the above-mentioned automated control system for livestock feed processing, the optimization unit is used to: calculate the natural exponential function value raised to the power of the eigenvalue of the predetermined position of the particle quality-equipment operating status timing interaction response feature matrix and subtract one to obtain a characteristic index difference; calculate the ratio of the characteristic index difference to the eigenvalue of the predetermined position of the particle quality-equipment operating status timing interaction response feature matrix and then take the absolute value to obtain a first absolute value; activate the first absolute value through a softmax function to obtain an activation value, and multiply the activation value by the eigenvalue of the predetermined position of the particle quality-equipment operating status timing interaction response feature matrix to obtain a feature activation value; calculate the particle quality-equipment operating status The first natural exponential function value is obtained by raising the inverse of the eigenvalue of the predetermined position of the timing interaction response feature matrix to the power of the natural exponential function value; the quotient of the eigenvalue of the predetermined position of the particle mass-equipment operating state timing interaction response feature matrix divided by the first natural exponential function value is calculated and then the absolute value is taken to obtain a second absolute value; the logarithmic function value with base 2 of the second absolute value is calculated, and the logarithmic function value is multiplied by the eigenvalue of the predetermined position of the particle mass-equipment operating state timing interaction response feature matrix to obtain a characteristic logarithmic activation value; the weighted sum of the characteristic activation value and the characteristic logarithmic activation value is calculated to obtain the eigenvalue of the predetermined position of the optimized particle mass-equipment operating state timing interaction response feature matrix.
[0016] In the above-mentioned automated control system in the livestock feed processing process, the control instruction generation module is used to: expand the optimized particle quality-equipment operating status timing interaction response feature matrix along the row vector into a classification feature vector; use the fully connected layer of the classifier-based ring die speed controller to fully connect the classification feature vector to obtain a fully connected encoded feature vector; input the fully connected encoded feature vector into the Softmax classification function of the classifier-based ring die speed controller to obtain the probability value of the optimized particle quality-equipment operating status timing interaction response feature matrix belonging to each classification label, wherein the classification labels include should increase, should decrease and should maintain; and determine the classification label corresponding to the largest of the probability values as the control instruction.
[0017] According to another aspect of the present application, there is provided an automated control method in a livestock feed processing process, comprising:
[0018] Material temperature values, material humidity values, material flow values and ring die speed values at multiple predetermined time points within a predetermined time period collected by a sensor group deployed in the ring die pelletizer;
[0019] Extract material temperature and humidity time series correlation features and flow and speed time series correlation features from the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points to obtain a material particle state comprehensive characterization feature vector and an equipment operation state comprehensive characterization feature vector;
[0020] Calculating the responsiveness estimate of the comprehensive characterization feature vector of the material particle state relative to the comprehensive characterization feature vector of the equipment operation state to obtain a particle mass-equipment operation state time series interactive response feature matrix;
[0021] Based on the particle quality-equipment operation state timing interaction response characteristic matrix, a control instruction for the ring die rotation speed is determined.
[0022] Compared with the prior art, the automated control system and method for livestock feed processing provided by the present application uses artificial intelligence technology based on deep learning to monitor and analyze the material temperature, material humidity, material flow rate and ring die speed in the ring die pelletizer, mines the feed particle state characteristics from the material temperature and humidity data, and mines the equipment operation state characteristics from the material flow rate and ring die speed data, thereby adaptively adjusting the ring die speed based on the timing response characteristics of the particle state relative to the equipment operation state to optimize the pelletizing effect. In this way, intelligent optimization and regulation of the feed processing pelletizing process can be achieved to improve feed processing efficiency and pelletizing quality, and reduce energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.
[0024] Figure 1 It is a block diagram of an automated control system in a livestock feed processing process according to an embodiment of the present application.
[0025] Figure 2 Schematic diagram of the architecture of an automated control system in a livestock feed processing process according to an embodiment of the present application.
[0026] Figure 3 It is a block diagram of a timing correlation analysis module in an automated control system in a livestock feed processing process according to an embodiment of the present application.
[0027] Figure 4 It is a block diagram of a ring die speed control module in an automated control system in a livestock feed processing process according to an embodiment of the present application.
[0028] Figure 5 It is a flow chart of an automated control method in a livestock feed processing process according to an embodiment of the present application. DETAILED DESCRIPTION
[0029] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.
[0030] Figure 1 It is a block diagram of an automated control system in a livestock feed processing process according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of an automated control system in a livestock feed processing process according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the automated control system 100 in the livestock feed processing process of the embodiment of the present application, it includes: a ring die pelletizing monitoring module 110, which is used to collect material temperature values, material humidity values, material flow values and ring die speed values at multiple predetermined time points within a predetermined time period by a sensor group deployed in the ring die pelletizer; a timing correlation analysis module 120, which is used to extract the material temperature and humidity timing correlation characteristics and the flow speed timing correlation characteristics from the material temperature values, material humidity values, material flow values and ring die speed values at the multiple predetermined time points to obtain a comprehensive characterization feature vector of the material particle state and a comprehensive characterization feature vector of the equipment operation state; a timing interactive response feature extraction module 130, which is used to calculate the responsiveness estimate of the comprehensive characterization feature vector of the material particle state relative to the comprehensive characterization feature vector of the equipment operation state to obtain a particle quality-equipment operation state timing interactive response feature matrix; a ring die speed control module 140, which is used to determine the control instruction of the ring die speed based on the particle quality-equipment operation state timing interactive response feature matrix.
[0031] As mentioned in the above background technology, with the rapid development of animal husbandry, the feed processing industry faces multiple challenges of improving production efficiency, optimizing product quality and reducing energy consumption. The traditional pelleting process mostly relies on manual experience to adjust parameters, which is difficult to achieve precise control. There are problems such as unstable pelleting quality, high energy consumption, and low production efficiency, which cannot meet the needs of modern production. In view of the above technical problems, the technical concept of this application is to use artificial intelligence technology based on deep learning to monitor and analyze the material temperature, material humidity, material flow rate and ring die speed in the ring die pelletizer, and to mine the feed particle state characteristics from the material temperature and humidity data, and to mine the equipment operation state characteristics from the material flow rate and ring die speed data, so as to adaptively adjust the ring die speed based on the timing response characteristics of the particle state relative to the equipment operation state to optimize the pelleting effect. In this way, the intelligent optimization and regulation of the feed processing pelleting process can be achieved to improve the feed processing efficiency and pelleting quality, and reduce energy consumption.
[0032] In the above-mentioned automated control system 100 in the livestock feed processing process, the ring die pelleting monitoring module 110 is used for collecting the material temperature value, material humidity value, material flow value and ring die speed value at multiple predetermined time points within a predetermined time period by the sensor group deployed in the ring die pelletizer. It should be understood that in the livestock feed processing process, especially in the operation of the ring die pelletizer, it is crucial to ensure the quality and pelleting efficiency of the feed pellets. Among them, the temperature and humidity of the feed pellets are key factors affecting their quality and performance. For example, too high a temperature may cause the feed pellets to burn or lose nutrition, while too low a humidity may cause the pellets to be unformed or fragile. By monitoring the temperature and humidity of the material, the production status of the feed pellets can be understood in time, so that the changes in the state of the feed pellets can be quickly responded to, and the corresponding pelleting parameters can be adjusted in time to ensure the quality of the feed pellets. At the same time, the material flow rate and the ring die speed are important indicators reflecting the operating status of the ring die pelletizer. The stability of the material flow rate directly affects the continuity and uniformity of the pelletizing, while the adjustment of the ring die speed is directly related to the molding quality and production efficiency of the pellets. By monitoring the material flow and ring die speed, the operating status of the ring die pellet mill can be fully understood, and the ring die speed can be automatically adjusted to match the material flow, thereby ensuring that the pellet mill operates in the best condition and improving production efficiency.
[0033] In the above-mentioned automated control system 100 for livestock feed processing, the time series correlation analysis module 120 is used to extract the material temperature and humidity time series correlation characteristics and the flow rate and speed time series correlation characteristics from the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points to obtain the material particle state comprehensive characterization feature vector and the equipment operation state comprehensive characterization feature vector. Figure 3 FIG. 1 is a block diagram of a time series correlation analysis module in an automated control system in a livestock feed processing process according to an embodiment of the present application. Figure 3 As shown, the timing association analysis module 120 includes: a timing arrangement unit 121, which is used to arrange the material temperature values, material humidity values, material flow values and ring die speed values at the multiple predetermined time points into a material temperature timing input vector, a material humidity timing input vector, a material flow timing input vector and a ring die speed timing input vector according to the time dimension; a particle state comprehensive representation unit 122, which is used to perform timing association encoding on the material temperature timing input vector and the material humidity timing input vector to obtain the material particle state comprehensive representation feature vector; an equipment operation state comprehensive representation unit 123, which is used to perform timing association encoding on the material flow timing input vector and the ring die speed timing input vector to obtain the equipment operation state comprehensive representation feature vector.
[0034] Specifically, the time series arrangement unit 121 is used to arrange the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points into material temperature time series input vectors, material humidity time series input vectors, material flow time series input vectors and ring die speed time series input vectors according to the time dimension. It should be understood that the temperature, humidity, flow rate and ring die speed of the material are parameters that change with time. In the technical solution of the present application, in order to capture the changing trends and patterns of these parameters over time, the material temperature values, material humidity values, material flow rate values and ring die speed values at the plurality of predetermined time points are further arranged according to the time dimension, so as to integrate the time series distribution information of the data in the time dimension, retain the time series characteristics of the data, so as to mine the time series dynamic change characteristics of each data, so that the model can use this time series dynamic change characteristics to accurately control the ring die speed.
[0035] Specifically, the particle state comprehensive characterization unit 122 is used to perform time-series association coding on the material temperature time series input vector and the material humidity time series input vector to obtain the material particle state comprehensive characterization feature vector. In a specific example of the present application, the encoding method for performing time-series association coding on the material temperature time series input vector and the material humidity time series input vector is to arrange the material temperature time series input vector and the material humidity time series input vector in two dimensions into a material temperature and humidity time series input matrix and then pass through a time series association encoder based on a convolutional neural network model to obtain the material particle state comprehensive characterization feature vector. It should be understood that considering that the quality state of feed particles is jointly determined by multiple factors such as its temperature and humidity. Therefore, in order to more comprehensively and accurately characterize the state of feed particles, in the technical solution of the present application, the material temperature time series input vector and the material humidity time series input vector are further arranged in two dimensions into a material temperature and humidity time series input matrix to integrate the time series distribution information on the two parameter dimensions of temperature and humidity. Then, the material temperature and humidity time series input matrix is subjected to time series association coding using a convolutional neural network model to extract comprehensive material particle state characterization features. Specifically, the convolutional neural network model has a strong feature extraction capability. By applying a two-dimensional convolution kernel to perform a sliding convolution operation on the material temperature and humidity time series input matrix, it can learn the local correlation features in the material temperature and humidity time series input matrix, and effectively mine the deep-level time series correlation features between material temperature and humidity, thereby achieving accurate characterization of the feed particle state. This provides an important basis for the subsequent ring die speed adjustment.
[0036] Specifically, the equipment operation state comprehensive characterization unit 123 is used to perform time-series correlation coding on the material flow timing input vector and the ring die speed timing input vector to obtain the equipment operation state comprehensive characterization feature vector. In a specific example of the present application, the encoding method for performing time-series correlation coding on the material flow timing input vector and the ring die speed timing input vector is to arrange the material flow timing input vector and the ring die speed timing input vector in two dimensions into a flow speed timing input matrix and then pass the time-series correlation encoder based on the convolutional neural network model to obtain the equipment operation state comprehensive characterization feature vector. It should be understood that the material flow represents the ability of the equipment to process materials, while the ring die speed is directly related to the pelletizing effect and the load of the equipment. During the production process, changes in equipment operating parameters directly affect the quality and production efficiency of feed pellets. Therefore, in order to fully capture the operating state characteristics of the pelletizer equipment. In the technical solution of the present application, the same encoding method is adopted, and the timing association encoder based on the convolutional neural network model is used to perform timing association encoding on the material flow timing input vector and the ring die speed timing input vector, so as to capture the correlation structure information of the material flow and the ring die speed changing with time, so as to fully understand the operating status of the equipment, realize a comprehensive and accurate characterization of the operating status of the equipment, and more accurately describe the overall performance of the equipment.
[0037] In the above-mentioned automated control system 100 in the livestock feed processing process, the time-series interactive response feature extraction module 130 is used to calculate the responsiveness estimation of the material particle state comprehensive characterization feature vector relative to the equipment operation state comprehensive characterization feature vector to obtain the particle quality-equipment operation state time-series interactive response feature matrix. It should be understood that in the actual production process, there is a close interactive relationship between the material particle state (such as temperature, humidity, etc.) and the equipment operation state (such as material flow, ring die speed, etc.). The temperature and humidity state of the material directly affect the molding quality and nutritional retention of the particles, while the operation state of the equipment determines the efficiency of material processing and the conditions for particle formation. By calculating the responsiveness estimation of the material particle state comprehensive characterization feature vector relative to the equipment operation state comprehensive characterization feature vector, the interaction and influence between the particle quality and the equipment operation state can be more deeply understood, and the sensitivity of the particle quality to the change of the equipment operation parameters can be revealed, so that the ring die speed of the pellet mill can be dynamically adjusted based on this responsive association relationship, so that it can better adapt to the change of the material state to achieve the best pelletizing effect.
[0038] In a specific example of the present application, the timing interaction response feature extraction module 130 is used to: multiply the material particle state comprehensive characterization feature vector and the inverse vector of the equipment operation state comprehensive characterization feature vector to obtain the particle mass-equipment operation state timing interaction response feature matrix.
[0039] In the above-mentioned automated control system 100 for livestock feed processing, the ring die speed control module 140 is used to determine the control instruction of the ring die speed based on the particle quality-equipment operation state time series interactive response characteristic matrix. Figure 4 FIG. 1 is a block diagram of a ring die speed control module in an automated control system in a livestock feed processing process according to an embodiment of the present application. Figure 4 As shown, the ring die speed control module 140 includes: an optimization unit 141, which is used to perform label position sensitivity analysis on the particle quality-equipment operation state timing interaction response feature matrix to obtain an optimized particle quality-equipment operation state timing interaction response feature matrix; a control instruction generation unit 142, which is used to pass the optimized particle quality-equipment operation state timing interaction response feature matrix through a classifier-based ring die speed controller to obtain the control instruction, and the control instruction is used to indicate whether the ring die speed should be increased, decreased or maintained at the current time point.
[0040] Specifically, the optimization unit 141 is used to perform label position sensitivity analysis on the particle quality-equipment operation state time series interaction response feature matrix to obtain an optimized particle quality-equipment operation state time series interaction response feature matrix. In particular, in the technical solution of the present application, considering that the acquisition of material temperature value, material humidity value, material flow value and ring die speed value may be affected by factors such as sensor position and accuracy, resulting in concentrated distribution of data at certain time points, while data at other time points are relatively sparse, thereby affecting the feature distribution in the feature matrix. The convolutional neural network model used in the time series association encoder may be more sensitive to certain features, resulting in these features being highly concentrated in the feature matrix, while other features are relatively sparsely distributed. When calculating the particle quality-equipment operation state time series interaction response feature matrix, fusing the responsiveness estimates of different features may result in local significant aggregation anomalies in the feature distribution in the particle quality-equipment operation state time series interaction response feature matrix, that is, some features are highly concentrated at specific locations. This will produce local deviations, causing the classifier to be overly sensitive to the features at these locations and insufficiently sensitive to the features at other locations. On the other hand, when the classification is performed by the classifier, the position information in the particle mass-equipment operation state timing interaction response feature matrix is usually lost. This will cause the classifier to be unable to consider the position information of the feature, thereby causing the inductive bias of the classification. In order to solve the above problem, in the technical solution of the present application, the label position sensitivity analysis is performed on the particle mass-equipment operation state timing interaction response feature matrix.
[0041] In a specific example of the present application, the optimization unit 141 is used to: calculate the natural exponential function value of the eigenvalue of the predetermined position of the particle mass-equipment operating state timing interaction response feature matrix raised to the power and subtract one to obtain a characteristic index difference; calculate the ratio of the characteristic index difference to the eigenvalue of the predetermined position of the particle mass-equipment operating state timing interaction response feature matrix and then take the absolute value to obtain a first absolute value; activate the first absolute value through a softmax function to obtain an activation value, and multiply the activation value by the eigenvalue of the predetermined position of the particle mass-equipment operating state timing interaction response feature matrix to obtain a feature activation value; calculate the ratio of the characteristic index difference to the eigenvalue of the predetermined position of the particle mass-equipment operating state timing interaction response feature matrix to obtain a feature activation value. The first natural exponential function value is obtained by raising the inverse of the eigenvalue of the predetermined position of the mutual response characteristic matrix to the power of the natural exponential function value; the quotient of the eigenvalue of the predetermined position of the particle mass-equipment operation state timing interaction response characteristic matrix divided by the first natural exponential function value is calculated and then the absolute value is taken to obtain a second absolute value; the logarithmic function value with base 2 of the second absolute value is calculated, and the logarithmic function value is multiplied by the eigenvalue of the predetermined position of the particle mass-equipment operation state timing interaction response characteristic matrix to obtain a characteristic logarithmic activation value; the weighted sum of the characteristic activation value and the characteristic logarithmic activation value is calculated to obtain the eigenvalue of the predetermined position of the optimized particle mass-equipment operation state timing interaction response characteristic matrix.
[0042] Specifically, the label position sensitivity analysis of the particle quality-equipment operation state timing interaction response characteristic matrix is performed using the following optimization formula to obtain the optimized particle quality-equipment operation state timing interaction response characteristic matrix, wherein the optimization formula is:
[0043]
[0044] Among them, m i,j is the eigenvalue of the (i, j)th position of the particle mass-equipment operation state time series interaction response characteristic matrix, α and β represent weighting coefficients, softmax represents the normalized exponential function, log2 represents the logarithmic function value with base 2, and m i,j ' is the eigenvalue of the (i, j)th position of the optimized particle mass-equipment operating state timing interaction response characteristic matrix.
[0045] That is, in the technical solution of the present application, a label position sensitivity analysis is performed on the particle mass-equipment operating state timing interaction response feature matrix, and the above problem is solved by scattering the features in the particle mass-equipment operating state timing interaction response feature matrix to multiple positions related to its label position. Specifically, performing a label position sensitivity analysis on the particle mass-equipment operating state timing interaction response feature matrix retains the position information, ensuring that the scattered feature matrix still retains the original position information, and at the same time, the scattered feature distribution is aligned with the label position, thereby enhancing the classifier's ability to capture label position sensitivity. Furthermore, through the scattered features, the classifier can more comprehensively consider the position information of the features, thereby reducing the inductive bias of the classification.
[0046] Specifically, the control instruction generating unit 142 is used to pass the optimized particle quality-equipment operation state timing interactive response characteristic matrix through the classifier-based ring die speed controller to obtain the control instruction, and the control instruction is used to indicate whether the ring die speed at the current time point should be increased, decreased, or maintained. It should be understood that the classifier-based ring die speed controller can identify the interactive response pattern between particle quality and equipment operation parameters through in-depth analysis of the optimized particle quality-equipment operation state timing interactive response characteristic matrix, and then judge the adjustment direction of the ring die speed at the current time point according to this interactive response pattern, and map the optimized particle quality-equipment operation state timing interactive response characteristic matrix to a preset category label, so as to obtain the control instruction of the ring die speed for guiding the adjustment of the ring die speed, such as increasing the ring die speed, decreasing the ring die speed, or maintaining the current speed.
[0047] In a specific example of the present application, the control instruction generation module 142 is used to: expand the optimized particle quality-equipment operating status timing interaction response feature matrix along the row vector into a classification feature vector; use the fully connected layer of the classifier-based ring die speed controller to fully connect the classification feature vector to obtain a fully connected encoded feature vector; input the fully connected encoded feature vector into the Softmax classification function of the classifier-based ring die speed controller to obtain the probability value of the optimized particle quality-equipment operating status timing interaction response feature matrix belonging to each classification label, wherein the classification labels include should increase, should decrease and should maintain; and determine the classification label corresponding to the largest of the probability values as the control instruction.
[0048] In summary, the automated control system in the livestock feed processing process according to the embodiment of the present application is explained, which uses artificial intelligence technology based on deep learning to monitor and analyze the material temperature, material humidity, material flow rate and ring die speed in the ring die pelletizer, mines the feed particle state characteristics from the material temperature and humidity data, and mines the equipment operation state characteristics from the material flow rate and ring die speed data, thereby adaptively adjusting the ring die speed based on the timing response characteristics of the particle state relative to the equipment operation state to optimize the pelletizing effect. In this way, intelligent optimization and regulation of the feed processing pelletizing process can be achieved to improve feed processing efficiency and pelletizing quality, and reduce energy consumption.
[0049] Figure 5 FIG. 1 is a flow chart of an automated control method for livestock feed processing according to an embodiment of the present application. Figure 5 As shown, according to the automated control method in the livestock feed processing process of the embodiment of the present application, the steps include: S110, collecting material temperature values, material humidity values, material flow values and ring die speed values at multiple predetermined time points within a predetermined time period by a sensor group deployed in the ring die pelletizer; S120, respectively extracting the material temperature and humidity time series correlation characteristics and the flow speed time series correlation characteristics from the material temperature values, material humidity values, material flow values and ring die speed values at the multiple predetermined time points to obtain a material particle state comprehensive characterization feature vector and an equipment operation state comprehensive characterization feature vector; S130, calculating the responsiveness estimate of the material particle state comprehensive characterization feature vector relative to the equipment operation state comprehensive characterization feature vector to obtain a particle quality-equipment operation state time series interaction response feature matrix; S140, determining the control instruction of the ring die speed based on the particle quality-equipment operation state time series interaction response feature matrix.
[0050] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned automated control method for livestock feed processing have been described in detail above. Figures 1 to 4 The description of the automatic control system in the livestock feed processing process has been introduced in detail, and therefore, its repeated description will be omitted.
[0051] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the module division is only a logical function division, and there may be other division methods in actual implementation. The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.
[0052] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0053] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.
[0054] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.
[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.
Claims
1. An automated control system for livestock feed processing, characterized in that: include: The ring die pelletizing monitoring module is used to collect material temperature values, material humidity values, material flow values and ring die speed values at multiple predetermined time points within a predetermined time period by a sensor group deployed in the ring die pelletizing machine; A time series correlation analysis module is used to extract material temperature and humidity time series correlation characteristics and flow rate and speed time series correlation characteristics from the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points to obtain a material particle state comprehensive characterization feature vector and a device operation state comprehensive characterization feature vector; A time series interactive response feature extraction module is used to calculate the responsiveness estimate of the comprehensive characterization feature vector of the material particle state relative to the comprehensive characterization feature vector of the equipment operation state to obtain a particle mass-equipment operation state time series interactive response feature matrix; A ring die speed control module, used to determine a control instruction for the ring die speed based on the particle mass-equipment operation state timing interaction response characteristic matrix; Wherein, the time series correlation analysis module includes: A timing arrangement unit, for arranging the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points into a material temperature timing input vector, a material humidity timing input vector, a material flow timing input vector and a ring die speed timing input vector according to the time dimension; A particle state comprehensive characterization unit, used for performing time series correlation coding on the material temperature time series input vector and the material humidity time series input vector to obtain a comprehensive characterization feature vector of the material particle state; The equipment operation state comprehensive characterization unit is used for performing time-series correlation coding on the material flow time series input vector and the ring die speed time series input vector to obtain a comprehensive characterization feature vector of the equipment operation state.
2. The automatic control system in the livestock feed processing process according to claim 1, characterized in that: The particle state comprehensive characterization unit is used for: The material temperature time series input vector and the material humidity time series input vector are arranged in two dimensions into a material temperature and humidity time series input matrix, and then a time series association encoder based on a convolutional neural network model is used to obtain a comprehensive characteristic vector representing the material particle state.
3. The automatic control system in the livestock feed processing process according to claim 2, characterized in that: The equipment operation status comprehensive characterization unit is used for: The material flow timing input vector and the ring die speed timing input vector are arranged in two dimensions into a flow speed timing input matrix and then passed through the timing association encoder based on the convolutional neural network model to obtain a comprehensive characteristic vector representing the equipment operation status.
4. The automatic control system in the livestock feed processing process according to claim 3 is characterized in that: The temporal interaction response feature extraction module is used to: The material particle state comprehensive characterization feature vector and the inverse vector of the equipment operation state comprehensive characterization feature vector are multiplied to obtain the particle mass-equipment operation state time series interaction response feature matrix.
5. The automatic control system in the livestock feed processing process according to claim 4, characterized in that: The ring die speed control module comprises: An optimization unit, configured to perform a label position sensitivity analysis on the particle quality-equipment operation state timing interaction response characteristic matrix to obtain an optimized particle quality-equipment operation state timing interaction response characteristic matrix; A control instruction generating unit is used to pass the optimized particle quality-equipment operation state timing interaction response characteristic matrix through a classifier-based ring die speed controller to obtain the control instruction, and the control instruction is used to indicate whether the ring die speed should be increased, decreased or maintained at the current time point.
6. The automatic control system in the livestock feed processing process according to claim 5, characterized in that: The optimization unit is used to: Calculate a natural exponential function value raised to the power of the eigenvalue of a predetermined position of the particle mass-equipment operation state timing interaction response characteristic matrix and subtract one to obtain a characteristic exponent difference; Calculating a ratio of the characteristic index difference to a characteristic value of a predetermined position of the particle mass-equipment operation state timing interaction response characteristic matrix and then taking an absolute value to obtain a first absolute value; Activating the first absolute value through a softmax function to obtain an activation value, and multiplying the activation value by a feature value at a predetermined position of the particle mass-device operating state timing interaction response feature matrix to obtain a feature activation value; Calculating a natural exponential function value raised to a power of the inverse of the eigenvalue of a predetermined position of the particle mass-equipment operating state timing interaction response characteristic matrix to obtain a first natural exponential function value; Calculate the quotient of the eigenvalue of the predetermined position of the particle mass-equipment operation state time series interaction response characteristic matrix divided by the first natural exponential function value and then take the absolute value to obtain a second absolute value; Calculating a logarithmic function value with base 2 of the second absolute value, and multiplying the logarithmic function value by a characteristic value at a predetermined position of the particle mass-device operating state timing interaction response characteristic matrix to obtain a characteristic logarithmic activation value; A weighted sum of the feature activation value and the feature logarithmic activation value is calculated to obtain a feature value at a predetermined position of the optimized particle mass-equipment operating state timing interaction response feature matrix.
7. The automatic control system in the livestock feed processing process according to claim 6, characterized in that: The control instruction generating unit is used for: Expanding the optimized particle quality-equipment operation state time series interaction response characteristic matrix along the row vector into a classification characteristic vector; Using the fully connected layer of the classifier-based ring die speed controller to perform fully connected encoding on the classification feature vector to obtain a fully connected encoded feature vector; Inputting the fully connected encoded feature vector into the Softmax classification function of the classifier-based ring die speed controller to obtain the probability value of the optimized particle quality-equipment operation state time series interaction response feature matrix belonging to each classification label, wherein the classification labels include should increase, should decrease, and should maintain; The classification label corresponding to the largest probability value among the probability values is determined as the control instruction.
8. An automated control method in livestock feed processing, characterized in that: include: Material temperature values, material humidity values, material flow values and ring die speed values at multiple predetermined time points within a predetermined time period collected by a sensor group deployed in the ring die pelletizer; Extract material temperature and humidity time series correlation features and flow and speed time series correlation features from the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points to obtain a material particle state comprehensive characterization feature vector and an equipment operation state comprehensive characterization feature vector; Calculating the responsiveness estimate of the comprehensive characterization feature vector of the material particle state relative to the comprehensive characterization feature vector of the equipment operation state to obtain a particle mass-equipment operation state time series interactive response feature matrix; Determining a control instruction for the ring die speed based on the particle quality-equipment operation state timing interactive response characteristic matrix; Among them, the material temperature and humidity time series correlation characteristics and the flow rate and speed time series correlation characteristics are respectively extracted from the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points to obtain the material particle state comprehensive characterization feature vector and the equipment operation state comprehensive characterization feature vector, including: Arrange the material temperature values, material humidity values, material flow values and ring die speed values at the plurality of predetermined time points into a material temperature time series input vector, a material humidity time series input vector, a material flow time series input vector and a ring die speed time series input vector according to the time dimension; Performing time series correlation coding on the material temperature time series input vector and the material humidity time series input vector to obtain a comprehensive characterization feature vector of the material particle state; The material flow time series input vector and the ring die speed time series input vector are time series associated and encoded to obtain a comprehensive characteristic vector representing the equipment operation status.
