A method and system for optimizing a feed pelleting process

Abstract

This invention belongs to the field of feed processing technology and provides a method and system for optimizing feed pelleting process. The method includes acquiring pelleting state characteristic values ​​and generating a pelleting state score to determine whether parameter adjustments are needed. If adjustments are required, an adjustable command analysis model is constructed based on the correlation between adjustable commands and pelleting state, as well as the difficulty of adjustment, to select the optimal adjustable command. Furthermore, historical data is used to establish a relationship curve between command parameters and pelleting state and production capacity. Combined with a preset parameter adjustment range, a comprehensive benefit analysis model is constructed to generate a comprehensive benefit score for parameter adjustments. An command parameter analysis model is established with the comprehensive benefit score as the target to solve for the optimal command parameters and execute the adjustment. This invention can automatically evaluate pelleting state and optimize process parameters, achieving a synergistic improvement in pelleting quality and production efficiency, reducing reliance on subjective experience and the lag in parameter adjustments, thereby improving the stability and overall efficiency of the feed pelleting process.

Description

Technical Field

[0001] This invention belongs to the field of feed processing technology, and in particular relates to a method and system for optimizing feed pelleting process. Background Technology

[0002] In the feed pelleting process, the setting of process parameters (such as conditioning temperature, ring die compression ratio, and feeding speed) directly affects pellet quality and production efficiency. Pelletization characteristics (such as particle size, moisture content, and degree of gelatinization) are key indicators for measuring the quality of intermediate products. How to dynamically adjust process parameters based on real-time pelleting status to achieve synergistic optimization of quality and production capacity is a significant technical challenge in the feed processing field.

[0003] Current feed pelleting process control mainly relies on manual experience-based adjustments or fixed parameter modes, with operators making coarse adjustments to individual parameters based on sensory judgment or offline testing results. Some systems introduce single-variable closed-loop control, such as adjusting steam addition based on temperature feedback, but these are mostly independent loop controls, failing to consider the coupling effects between multiple parameters. In terms of parameter optimization, existing methods are usually geared towards a single objective (such as maximizing output or minimizing energy consumption), lacking a comprehensive trade-off between pelleting quality and production capacity.

[0004] However, the aforementioned existing technologies have obvious drawbacks: manual experience-based adjustments are difficult to accurately quantify the correlation between pelleting status and parameters, resulting in delayed response and strong subjectivity; single-variable control cannot solve the problem of multi-parameter coupling, and parameter adjustments are prone to neglecting one aspect while focusing on another; optimization strategies oriented towards a single objective ignore the impact of pelleting quality on the nutritional value of feed and subsequent processing, which may lead to quality deterioration in pursuit of yield. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for optimizing feed pelleting processes, thus solving the aforementioned problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing a feed pelleting process, the method specifically comprising: Obtain pelleting state characteristic values ​​during the feed pelleting process, and generate pelleting state scores based on these characteristic values. Based on the pelleting status score, determine whether the current feed pelleting process requires adjustment of the command parameters; If the current feed pelleting process requires adjustment of instruction parameters, obtain the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting status, and the adjustment difficulty indicator value of the adjustable instructions; Based on the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between adjustable instructions and pelleting status, and the adjustment difficulty indicator value of adjustable instructions, an adjustable instruction analysis model is established to generate the optimal adjustable instructions. Establish the relationship curves between instruction parameters and pelleting status score changes, and between instruction parameters and production capacity changes, and preset the optimal adjustable instruction parameter adjustment range; wherein, the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes are established based on the corresponding pelleting status change values ​​and production capacity change values ​​when instruction parameters are adjusted in historical feed pelleting processes; Based on the parameter adjustment range of the optimal adjustable command, a comprehensive benefit analysis model is established according to the relationship curves of command parameters-granulation status score change and command parameters-capacity change, and a comprehensive benefit score for the adjustment of the optimal adjustable command parameters is generated. Based on the parameter adjustment range of the optimal adjustable command, with the parameter value of the optimal adjustable command as the independent variable and the comprehensive benefit score of the parameter adjustment of the optimal adjustable command as the dependent variable, an instruction parameter analysis model is established to generate the optimal instruction parameters of the optimal adjustable command. The feed pelleting process is optimized based on the optimal instruction parameters of the optimal adjustable instructions.

[0007] Based on the above technical solutions, the present invention also provides the following optional technical solutions: Further technical solution: The method for generating the granulation status score specifically includes: Through the formula: ; Generate granulation status score ; In the formula, This represents the i-th pelleting state characteristic value during the feed pelleting process. This represents the standard value of the i-th pelleting state characteristic during the feed pelleting process. This represents the deviation threshold of the i-th granulation state characteristic.

[0008] Further technical solution: The specific method for generating the optimal adjustable instruction includes: Based on the Pearson correlation coefficient between adjustable instructions and granulation status, an instruction correlation score is generated; An adjustable instruction analysis model is established based on the adjustment difficulty indicator value and instruction relevance score of adjustable instructions, and the optimal adjustable instructions are generated.

[0009] Further technical solution: The specific method for generating the instruction relevance score includes: Through the formula: ; Generate instruction relevance score ; In the formula, This represents the absolute value of the Pearson correlation coefficient between the adjustable instruction k and the i-th granulation state. This represents the absolute threshold of the Pearson correlation coefficient.

[0010] Further technical solution: The expression of the adjustable instruction analysis model is specifically as follows: ; In the expression, This represents the optimal adjustable instruction. This represents the instruction correlation score between the adjustable instruction k and the i-th granulation state. This represents the normalized value of the adjustment difficulty flag for the adjustable instruction k. , All are weighting coefficients, and .

[0011] Further technical solution: The method for generating the comprehensive benefit score of the optimal adjustable command parameter adjustment specifically includes: The optimal adjustable parameter adjustment amount is preset; Based on the relationship curve between instruction parameters and granulation status score, extract the change value of granulation status score and generate a granulation status benefit score for instruction parameter adjustment. Based on the relationship curve between command parameters and capacity changes, extract the capacity change value and generate a capacity benefit score for command parameter adjustments; A comprehensive benefit analysis model is established based on the granulation state benefit score and the capacity benefit score of the instruction parameter adjustment, and the comprehensive benefit score of the optimal adjustable instruction parameter adjustment is generated. The specific expression of the comprehensive benefit analysis model is as follows: ; In the expression, This represents the overall benefit score of the optimal adjustable instruction parameter adjustment when the optimal adjustable instruction parameter adjustment amount is j. This represents the pelletizing efficiency score when the parameter adjustment amount of the optimal adjustable command is j. This represents the capacity efficiency score when the parameter adjustment amount of the optimal adjustable instruction is j. This represents the instruction parameters of the optimal adjustable instruction when the parameter adjustment amount of the optimal adjustable instruction is j. This represents the minimum instruction parameters for the optimal adjustable instruction. This represents the maximum instruction parameters of the optimal adjustable instruction. , , All are weighting coefficients, and .

[0012] Further technical solution: The specific method for generating the granulation state benefit score of the instruction parameter adjustment includes: Through the formula: ; Granulation status benefit score based on generation instruction parameter adjustment ; In the formula, This represents the i-th pelleting status score during the feed pelleting process. This represents the change in granulation status score when the parameter adjustment amount of the optimal adjustable command is j. This represents the granulation status scoring threshold.

[0013] Further technical solution: The specific method for generating the capacity efficiency score of the instruction parameter adjustment includes: Through the formula: ; Production efficiency score based on generation instruction parameter adjustment ; In the formula, This indicates the current production capacity. This represents the change in production capacity when the parameter adjustment amount of the optimal adjustable command is j. This indicates the expected production capacity.

[0014] Further technical solution: The expression of the instruction parameter analysis model is specifically as follows: ; In the expression, This represents the optimal instruction parameters for the optimal adjustable instruction. This represents the comprehensive benefit score of the optimal adjustable instruction parameter adjustment when the parameter adjustment amount of the optimal adjustable instruction is j, and H represents the parameter adjustment range of the optimal adjustable instruction.

[0015] A feed pelleting process optimization system, which is used to execute the above-mentioned feed pelleting process optimization method, specifically includes: The state analysis unit is used to obtain the pelleting state characteristic values ​​during the feed pelleting process and generate a pelleting state score based on the pelleting state characteristic values. The judgment unit is used to determine whether the current feed pelleting process needs to be adjusted according to the pelleting status score. The data acquisition unit is used to obtain the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting status, and the adjustment difficulty indicator value of the adjustable instructions if the current feed pelleting process requires adjustment of the instruction parameters. The optimal instruction analysis unit is used to establish an adjustable instruction analysis model based on the adjustable instruction set of the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions, and generate the optimal adjustable instructions. The relationship curve establishment unit is used to establish the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes, and to preset the optimal adjustable instruction parameter adjustment range; wherein, the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes are established based on the corresponding pelleting status change values ​​and production capacity change values ​​when instruction parameters are adjusted in historical feed pelleting processes; The comprehensive benefit analysis unit is used to establish a comprehensive benefit analysis model based on the parameter adjustment range of the optimal adjustable command, according to the relationship curve of command parameter-granulation status score change and command parameter-capacity change, and to generate a comprehensive benefit score for the adjustment of the optimal adjustable command parameter. The optimal instruction parameter analysis unit is used to establish an instruction parameter analysis model based on the parameter adjustment range of the optimal adjustable instruction, with the parameter value of the optimal adjustable instruction as the independent variable and the comprehensive benefit score of the optimal adjustable instruction parameter adjustment as the dependent variable, and to generate the optimal instruction parameters of the optimal adjustable instruction. The optimization unit is used to optimize the feed pelleting process based on the optimal instruction parameters of the optimal adjustable instruction.

[0016] This invention provides a method and system for optimizing feed pelleting processes, which has the following advantages compared with existing technologies: This invention solves the problems of relying on subjective experience, lagging parameter adjustment, and ignoring multi-parameter coupling in existing technologies by acquiring pelleting state characteristic values ​​to generate scores, judging parameter adjustment needs, establishing models to generate optimal instructions and parameters, and dynamically optimizing the process based on historical data. It can automatically evaluate pelleting state and optimize process parameters, achieve synergistic improvement in pelleting quality and production efficiency, reduce reliance on subjective experience and lag in parameter adjustment, thereby improving the stability and overall efficiency of feed pelleting process. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for optimizing the feed pelleting process provided by the present invention.

[0018] Figure 2 This is a flowchart illustrating step S40 of the present invention.

[0019] Figure 3 This is a flowchart illustrating step S60 of the present invention.

[0020] Figure 4This is a schematic diagram of a feed pelleting process optimization system provided by the present invention.

[0021] Figure 5 A schematic diagram of the structure of the optimal instruction analysis unit provided by the present invention.

[0022] Figure 6 A schematic diagram of the structure of the comprehensive benefit analysis unit provided by the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 merely illustrative and not intended to limit the invention.

[0024] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0025] Please see Figure 1 The present invention provides a method for optimizing a feed pelleting process, which specifically includes the following steps: Step S10: Obtain the pelleting state characteristic values ​​during the feed pelleting process, and generate a pelleting state score based on the pelleting state characteristic values; Step S20: Based on the pelleting status score, determine whether the current feed pelleting process requires adjustment of the instruction parameters; Step S30: If the current feed pelleting process requires adjustment of instruction parameters, obtain the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions. Step S40: Based on the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions, establish an adjustable instruction analysis model to generate the optimal adjustable instructions; Step S50: Establish the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes, and preset the optimal adjustable instruction parameter adjustment range; wherein, the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes are established based on the corresponding pelleting status change values ​​and production capacity change values ​​when instruction parameters are adjusted in historical feed pelleting processes; Step S60: Based on the parameter adjustment range of the optimal adjustable command, establish a comprehensive benefit analysis model according to the relationship curve between command parameters and granulation status score and the relationship curve between command parameters and production capacity, and generate a comprehensive benefit score for the adjustment of the optimal adjustable command parameters; Step S70: Based on the parameter adjustment range of the optimal adjustable command, with the parameter value of the optimal adjustable command as the independent variable and the comprehensive benefit score of the parameter adjustment of the optimal adjustable command as the dependent variable, establish a command parameter analysis model to generate the optimal command parameters of the optimal adjustable command. Step S80: Optimize the feed pelleting process based on the optimal instruction parameters of the optimal adjustable instruction; Among them, the pelleting state characteristic value in the feed pelleting process refers to the state data of intermediate products formed by feed raw materials during the production process; for example, the particle size of powder after raw materials are crushed, the moisture content, the degree of gelatinization after conditioning, the pellet mill current, steam pressure, etc. These values ​​reflect the quality or operating status of the current production process. Command parameters refer to process variables that can be manipulated or controlled during the feed pelleting process; these parameters directly affect the pelleting effect and production efficiency, such as conditioning temperature, ring die compression ratio, feeding speed, steam addition amount, etc. The adjustable instruction set refers to the set of all adjustable instruction parameters in the current feed pelleting process; this set contains all potential control variables that can be used to optimize the process. The Pearson correlation coefficient is a statistical indicator used to measure the strength and direction of the linear correlation between two variables. In this embodiment, it is used to quantify the degree of correlation between adjustable instructions (i.e., process parameters) and granulation state characteristic values ​​in order to identify instructions that have a significant impact on granulation state. The difficulty indicator is used to characterize the complexity or resources required to adjust a certain instruction parameter; for example, adjusting some parameters may require downtime, time consumption, or wear and tear on the equipment, while other parameters can be adjusted in real time and smoothly; this indicator can be a qualitative or quantitative metric. The command parameter-granulation status score change curve describes the functional relationship between the adjustment of a specific command parameter and the change in the granulation status score. This curve is usually built based on historical data and is used to predict how the granulation status score will respond when a specific adjustment of the command parameter occurs. The command parameter-capacity change curve describes the functional relationship between the adjustment of a specific command parameter and the change in production capacity. This curve is also based on historical data and is used to predict the impact of command parameter adjustments on production efficiency.

[0026] Specifically, in step S10, it is first necessary to acquire the pelleting state characteristic values ​​during the feed pelleting process. These characteristic values ​​are key data for measuring the state of intermediate products formed during the feed raw material production process. For example, manual sampling can be used to obtain data such as particle size and moisture content through laboratory analysis, and then these data can be manually input into the system. Alternatively, basic sensors, such as temperature sensors and current sensors, can be deployed on the production line to collect relevant process data in real time. After acquiring the pelleting state characteristic values, a pelleting state score needs to be generated based on these characteristic values. One method is to set basic threshold rules; for example, when a certain characteristic value exceeds a preset range, it is directly judged as "unqualified" and assigned a lower score. Another method is to perform a weighted average of multiple characteristic values.

[0027] In step S20, based on the generated pelleting status score, it is determined whether the current feed pelleting process requires parameter adjustments. This determination can be based on a fixed scoring threshold; when the pelleting status score falls below this threshold, adjustment is considered necessary. Alternatively, it can be achieved by comparing the current score with the historical best score; if the deviation exceeds a certain percentage, adjustment is triggered.

[0028] In step S30, if it is determined that the current feed pelleting process requires adjustment of command parameters, it is necessary to obtain the set of adjustable commands for the current feed pelleting process, the Pearson correlation coefficient between the adjustable commands and the pelleting state, and the adjustment difficulty flag value of the adjustable commands. The set of adjustable commands can be pre-configured in the system with a preset parameter list, such as conditioning temperature and feeding speed. The Pearson correlation coefficient can be calculated through offline historical data analysis, determining the correlation between each command parameter and the pelleting state characteristic value, and storing it as a fixed value. The adjustment difficulty flag value can be determined by engineers based on experience, classifying the adjustment difficulty of each command parameter into different levels such as "high," "medium," and "low," and then assigning a value accordingly.

[0029] In step S40, an adjustable instruction analysis model is established based on the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions, and the optimal adjustable instruction is generated. This model can employ basic decision tree logic, for example, prioritizing the instruction with the highest correlation coefficient and the lowest adjustment difficulty. Alternatively, a basic linear combination of the correlation coefficient and adjustment difficulty can be performed, selecting the instruction with the highest score as the optimal adjustable instruction.

[0030] In step S50, it is necessary to establish the relationship curves between instruction parameters and granulation status scores, and between instruction parameters and production capacity, and to preset the optimal adjustable instruction parameter adjustment range. These relationship curves can be established based on manually recorded historical data by plotting scatter plots and manually fitting basic linear or polynomial curves. The parameter adjustment range can be set to a large adjustment interval based on the upper and lower limits of the parameters provided by the equipment manufacturer.

[0031] In step S60, based on the parameter adjustment range of the optimal adjustable command, a comprehensive benefit analysis model is established according to the relationship curves between command parameters and granulation status score changes and between command parameters and production capacity changes, generating a comprehensive benefit score for the adjustment of the optimal adjustable command parameters. This model can simply sum the weighted changes in granulation status score and production capacity changes as the comprehensive benefit score, where the weights may be set based on experience.

[0032] In step S70, based on the parameter adjustment range of the optimal adjustable command, a command parameter analysis model is established with the parameter values ​​of the optimal adjustable command as independent variables and the comprehensive benefit score of the optimal adjustable command parameter adjustment as the dependent variable, generating the optimal command parameters for the optimal adjustable command. This model can calculate the comprehensive benefit score corresponding to each discrete parameter value within a preset parameter adjustment range using an exhaustive method or step search, and select the parameter value corresponding to the highest score.

[0033] In step S80, the feed pelleting process is optimized based on the optimal instruction parameters of the generated optimal adjustable instructions. This optimization can be achieved by manually inputting the generated optimal instruction parameters into the control system of the pelleting equipment. Alternatively, the optimal instruction parameters can be directly sent to the PLC (Programmable Logic Controller) of the pelleting equipment for execution via an automation interface.

[0034] Through the above technical solution, this application obtains pelleting state characteristic values ​​to generate scores, judges parameter adjustment needs, establishes models to generate optimal instructions and parameters, and dynamically optimizes the process based on historical data. This solves the problems of relying on subjective experience, parameter adjustment lag, and ignoring multi-parameter coupling in the prior art. It can automatically evaluate pelleting state and optimize process parameters, achieve synergistic improvement in pelleting quality and production efficiency, reduce reliance on subjective experience and parameter adjustment lag, thereby improving the stability and overall efficiency of feed pelleting process.

[0035] Preferably, the present invention further proposes a method for generating the granulation status score, specifically including: Through the formula: ; Generate granulation status score ; In the formula, This represents the i-th pelleting state characteristic value during the feed pelleting process. This represents the standard value of the i-th pelleting state characteristic during the feed pelleting process. This represents the deviation threshold of the i-th granulation state characteristic; The purpose of the granulation state score is to unify the granulation state characteristic values ​​of different dimensions and numerical ranges onto a comparable scale, facilitating subsequent judgment and analysis. This score can be a dimensionless numerical value used to quantify the degree of deviation between the current granulation state and the standard state.

[0036] The i-th pelleting state characteristic value in the feed pelleting process represents the actual state data of the intermediate products formed by the feed raw materials during the production process. For example, it can be the fineness of the crushed material, the moisture content after mixing, the temperature or steam pressure after conditioning, etc. These characteristic values ​​can be collected in real time by sensors or obtained through offline detection.

[0037] The standard value of the i-th pelleting state characteristic in the feed pelleting process represents the expected value of that characteristic under ideal or target pelleting conditions. This standard value can be set based on historical production data, process expert experience, or preset process parameters. For example, for fineness, the standard value can be the center value of a certain particle size distribution range; for moisture content, the standard value can be the target moisture percentage.

[0038] The i-th granulation state characteristic deviation threshold defines the maximum allowable deviation range of this characteristic value. When the actual characteristic value exceeds this threshold, it is generally considered that there is a significant abnormality in the granulation state. This threshold can be determined based on process requirements, product quality standards, or historical data analysis results. For example, if the standard value for grinding fineness is 100 micrometers, the deviation threshold can be set to ±20 micrometers, indicating that the range of 80-120 micrometers is acceptable.

[0039] This application's solution addresses the accuracy issue arising from the diversity of pelleting state characteristic values ​​by introducing a standardized formula to generate pelleting state scores. After obtaining the pelleting state characteristic values ​​during the feed pelleting process, the method no longer directly uses the original characteristic values. Instead, for each characteristic value, it subtracts its corresponding standard value to obtain the deviation of that characteristic value relative to the standard value. This deviation is then divided by a preset deviation threshold for the i-th pelleting state characteristic, thus standardizing the deviations of different characteristic values ​​into a unified score. This standardization process allows for fair comparison and comprehensive evaluation of pelleting state characteristic values ​​of different types and dimensions, ensuring the objectivity and accuracy of the pelleting state score. The pelleting state score generated in this way more accurately reflects the deviation between the actual operating status of the current pelleting process and the ideal state, providing a more reliable basis for subsequent judgment on whether the current feed pelleting process requires parameter adjustments. This precise scoring mechanism enables the entire optimization method to more sensitively capture process anomalies and trigger more timely and effective parameter adjustments, thereby improving the overall efficiency and effectiveness of feed pelleting process optimization.

[0040] The above technical solution, employing a standardized formula to generate pelleting state scores, effectively addresses the problem of insufficient scoring accuracy caused by significant differences in the dimensions and numerical ranges of characteristic values ​​for different pelleting states. This formula compares each pelleting state characteristic value with its standard value and normalizes it according to a preset deviation threshold, ensuring that the generated pelleting state score objectively and accurately reflects the degree of deviation between the current pelleting state and the ideal state. This allows subsequent process adjustment decisions to be based on more accurate evaluation results, avoiding misjudgments or delayed adjustments caused by inaccurate scoring. Consequently, it improves the sensitivity and effectiveness of feed pelleting process optimization, contributing to maintaining product quality stability and production efficiency.

[0041] For preferred options, please refer to [link / reference]. Figure 2 The present invention further proposes a method for generating the optimal adjustable instruction, specifically including: Step S41: Generate an instruction correlation score based on the Pearson correlation coefficient between the adjustable instructions and the granulation state; Step S42: Establish an adjustable instruction analysis model based on the adjustment difficulty indicator value and instruction relevance score of the adjustable instructions, and generate the optimal adjustable instructions; The instruction relevance score is an indicator used to quantify the correlation between adjustable instructions and granulation state. Its function is to transform the original Pearson correlation coefficient into a more comparable and operable score value, facilitating subsequent comprehensive evaluation by the model. This score can reflect the significance or importance of the instruction's impact on the granulation state. For example, the score can be based on a linear or nonlinear mapping of the absolute value of the Pearson correlation coefficient, or it can be segmented by introducing a threshold. The purpose of generating the instruction relevance score is to transform the Pearson correlation coefficient between adjustable instructions and granulation state into a unified score value that can be used for model calculation. This helps to standardize the quantification of the impact of different instructions on different granulation states. One implementation method is to map the absolute value of the Pearson correlation coefficient to a score range of 0 to 100 using a preset transformation function. Another implementation method is to assign different discrete score levels, such as high, medium, and low, based on the comparison between the absolute value of the Pearson correlation coefficient and a preset threshold.

[0042] The adjustment difficulty indicator of an adjustable instruction is an metric that measures the complexity or resource requirements of adjusting parameters for a specific instruction. Its purpose is to consider not only its impact on pelleting conditions but also the practical feasibility of adjustment when selecting the optimal adjustable instruction. This indicator can be a qualitative level, such as "easy," "medium," or "difficult," or a quantitative value, such as a difficulty index from 0 to 10. The adjustable instruction analysis model is a mathematical or logical model used to comprehensively evaluate multiple adjustable instructions and select the optimal one. Its function is to weighted or unweightedly combine multiple factors, such as instruction relevance scores and adjustment difficulty indicators, to arrive at a comprehensive evaluation result that guides the selection of the optimal instruction. This model can be a weighted summation model or a classification or regression model based on decision trees or neural networks. The purpose of generating the optimal adjustable instruction is to identify the most suitable instruction for parameter adjustment from the current set of adjustable instructions in feed pelleting technology. This instruction should have high potential for improving pelleting conditions while having relatively low adjustment difficulty. One approach is to calculate a comprehensive score for each adjustable instruction using an adjustable instruction analysis model, and then select the instruction with the highest score as the optimal adjustable instruction. Another approach is to set multiple filtering criteria, such as a relevance score above a certain threshold and an adjustment difficulty indicator value below a certain threshold, and then randomly or by priority select instructions from those that meet the criteria.

[0043] Specifically, firstly, based on the Pearson correlation coefficient between adjustable instructions and granulation status, a command correlation score is generated in step S41. This score transforms the original Pearson correlation coefficient into a standardized, quantifiable indicator, more intuitively reflecting the strength and direction of the command's influence on granulation status. Subsequently, in step S42, this command correlation score is combined with the adjustment difficulty indicator value of the adjustable command to establish an adjustable command analysis model. This model comprehensively considers the potential impact of the command on granulation status (reflected by the command correlation score) and the adjustment cost or complexity in actual operation (reflected by the adjustment difficulty indicator value). Through this model, the system can comprehensively evaluate all adjustable commands, thereby generating an optimal adjustable command that achieves the best balance between improving granulation status and operational convenience. This step-by-step and comprehensive evaluation mechanism ensures that the selection of the optimal command no longer relies solely on a single correlation indicator but incorporates practical operational considerations, guaranteeing the effectiveness and feasibility of the selected command.

[0044] Through the above technical solution, this application can more comprehensively and accurately evaluate the optimization potential of each adjustable instruction. By introducing an instruction correlation score, the Pearson correlation coefficient is transformed into a more comparable quantitative indicator, clearly presenting the degree of influence of the instruction on the pelleting state. Simultaneously, by combining the adjustment difficulty indicator value of the adjustable instructions, the selection of the optimal instruction takes into account both the convenience and cost of actual operation, avoiding the selection of instructions that are difficult to implement or have excessively high implementation costs. This comprehensive consideration ensures that the generated optimal adjustable instruction not only has the theoretical potential to improve the pelleting state but is also more operable and efficient in practical applications, thereby improving the accuracy and efficiency of feed pelleting process optimization.

[0045] Preferably, the present invention further proposes a method for generating the instruction relevance score, specifically including: Through the formula: ; Generate instruction relevance score ; In the formula, This represents the absolute value of the Pearson correlation coefficient between the adjustable instruction k and the i-th granulation state. This represents the absolute threshold of the Pearson correlation coefficient; The instruction correlation score is a metric used to quantify the influence of adjustable instruction k on the i-th granulation state. This score aims to provide a standardized numerical value reflecting the strength of the correlation between the instruction and the granulation state, thus providing input for subsequent adjustable instruction analysis models. The score can be a dimensionless numerical value, with its magnitude reflecting the strength of the correlation; for example, a higher score indicates a stronger correlation.

[0046] The formula for calculating instruction relevance scores compares and normalizes the original absolute value of the Pearson correlation coefficient with a preset absolute threshold for the Pearson correlation coefficient. This method highlights instructions with relevance exceeding a certain threshold and suppresses or distinguishes those with relevance below the threshold, thus more effectively identifying instructions that significantly impact granulation status.

[0047] The Pearson correlation coefficient is a statistic that measures the strength of a linear correlation between two variables, ranging from -1 to 1. The absolute value is used to indicate the strength of the correlation, without distinguishing between positive and negative correlations. This value can be obtained through statistical analysis of historical granulation data; for example, by collecting data on changes in granulation conditions under different instruction parameter settings and then calculating the Pearson correlation coefficient between the two variables.

[0048] The Pearson correlation coefficient absolute threshold is a preset threshold used to determine whether a correlation is meaningful. Its function is to filter out instructions with weak correlations that have no significant impact on granulation status, or to serve as a benchmark for distinguishing between strong and weak correlations. This threshold can be determined based on industry experience, expert knowledge, or through statistical analysis of historical data (e.g., setting a confidence interval). For example, it can be set to 0.5, meaning that a strong correlation is considered to exist only when the absolute value of the correlation coefficient is greater than 0.5.

[0049] This application's solution generates an instruction relevance score by introducing a standardized formula. This formula is based on the absolute value of the Pearson correlation coefficient between the adjustable instruction k and the i-th pelleting state, and a preset absolute threshold for the Pearson correlation coefficient. Specifically, when it is necessary to adjust the instruction parameters of the current feed pelleting process, the absolute value of the Pearson correlation coefficient between the adjustable instruction and the pelleting state is first obtained. Subsequently, the instruction relevance score is calculated using this absolute value and the preset absolute threshold for the Pearson correlation coefficient. The generation of this score allows the influence of different instructions on different pelleting states to be quantified and standardized, thus providing a unified and comparable input for subsequent adjustable instruction analysis models. In this way, the solution can more accurately assess the correlation strength between each adjustable instruction and the pelleting state, avoiding the evaluation bias that may be caused by relying solely on the original Pearson correlation coefficient, thereby improving the accuracy and reliability of the optimal adjustable instruction selection.

[0050] Through the above technical solution, this application provides a standardized and quantifiable method for generating instruction relevance scores. This method can effectively distinguish the strength of the correlation between instructions and pelleting state, and highlight those instructions that have a significant impact on pelleting state. This enables the acquisition of more accurate and comparable input data when establishing an adjustable instruction analysis model, thereby significantly improving the accuracy and reliability of optimal adjustable instruction selection, and ultimately optimizing the adjustment effect of the feed pelleting process.

[0051] Preferably, the present invention further proposes that the expression of the adjustable instruction analysis model is specifically as follows: ; In the expression, This represents the optimal adjustable instruction. This represents the instruction correlation score between the adjustable instruction k and the i-th granulation state. This represents the normalized value of the adjustment difficulty flag for the adjustable instruction k. , All are weighting coefficients, and ; This expression defines a mathematical model for selecting the instruction that maximizes a certain comprehensive score from a set of candidate adjustable instructions A. This comprehensive score combines the instruction's influence on granulation state (instruction relevance score) and the ease of instruction adjustment (normalized value of the adjustment difficulty indicator). The model can be implemented by iterating through each adjustable instruction k in set A, calculating its corresponding comprehensive score, and then selecting the instruction with the highest score as the optimal adjustable instruction; alternatively, for large-scale instruction sets, heuristic or optimization algorithms (e.g., genetic algorithms, particle swarm optimization, etc.) can be used to search for the optimal solution to improve computational efficiency.

[0052] The optimal adjustable command refers to the command deemed most suitable for parameter adjustment after evaluation by the adjustable command analysis model when parameter adjustment is required. The selection of this command aims to achieve the greatest improvement in granulation condition with minimal adjustment difficulty. This optimal adjustable command can be directly recommended to the operator by the optimization system based on model calculations; alternatively, in highly automated systems, this command can be directly sent to the actuator for subsequent parameter adjustments.

[0053] The instruction correlation score between adjustable instruction k and the i-th granulation state is a quantitative indicator reflecting the degree of influence of a specific adjustable instruction k on the i-th granulation state. A higher score indicates a stronger correlation between the instruction and the granulation state, meaning that adjusting the instruction may have a more significant effect on improving the granulation state. This score can be calculated based on the Pearson correlation coefficient between adjustable instructions and granulation states; alternatively, in some cases, it can be combined with the experience of domain experts to subjectively score or weight the correlation between different instructions and granulation states.

[0054] The normalized value of the adjustment difficulty flag for adjustable instruction k is a standardized metric used to measure the ease or difficulty of adjusting a specific instruction k. Normalization allows for comparison of the adjustment difficulty of different instructions on a uniform scale. A higher value indicates that adjusting the instruction is easier, and the operational cost or risk is lower. This adjustment difficulty flag value is comprehensively assessed and assigned based on historical operational data, considering factors such as the time required to adjust the instruction, potential production risks, and required human resources, before being normalized.

[0055] Weighting coefficient , This is used to balance the relative importance of instruction relevance score and adjustment difficulty indicator value in the overall score. By adjusting these two weights, the selection preference for the optimal instruction can be adjusted according to actual production needs (e.g., whether to prioritize effectiveness or ease of operation). These weight coefficients can be preset based on production experience or expert knowledge; or, they can be increased based on current production goals (e.g., increasing the weight when there is an urgent need to improve the granulation status score). Weighting; increase when rapid adjustment is needed. The weights can be dynamically optimized (either by weighting) or through machine learning algorithms.

[0056] This application's solution achieves the scientific selection of optimal adjustable instructions by explicitly defining the expression of the adjustable instruction analysis model. Specifically, when the system determines that the current feed pelleting process requires adjustment of instruction parameters, it first obtains the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions. Based on this, an instruction relevance score is generated according to the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions is normalized to obtain its normalized value. Subsequently, these quantified indicators are substituted into a preset weighted summation model. This model balances the potential improvement effect of the instructions on the pelleting state and the actual operational difficulty of instruction adjustment through weight coefficients. The system traverses all candidate adjustable instructions, calculates the comprehensive score of each instruction, and selects the instruction with the highest comprehensive score as the optimal adjustable instruction. This mechanism ensures that the selected optimal adjustable instruction is not only theoretically highly correlated with the improvement of the pelleting state, but also has high feasibility and low implementation cost in actual operation, thus providing a clear and optimized direction for subsequent instruction parameter adjustments.

[0057] Through the above technical solution, this application clarifies the specific expression of the adjustable instruction analysis model, namely, the normalized value that comprehensively considers the instruction relevance score and the adjustment difficulty indicator value through a weighted summation method. This enables a quantitative balance between the potential improvement effect of the instruction on the pelleting state and the convenience of actual operation when selecting the optimal adjustable instruction. This method avoids suboptimal selections that may result from judging based on a single indicator or subjective experience, ensuring that the selected optimal adjustable instruction not only efficiently guides the pelleting process towards the desired state but also effectively reduces operational complexity and potential risks during implementation. Therefore, the solution of this application significantly improves the scientificity, accuracy, and practicality of feed pelleting process optimization decisions.

[0058] For preferred options, please refer to [link / reference]. Figure 3 The present invention further proposes a method for generating the comprehensive benefit score of the optimal adjustable command parameter adjustment, specifically including: Step S61: Preset the parameter adjustment amount of the optimal adjustable command; Step S62: Based on the relationship curve between instruction parameters and granulation status score, extract the change value of granulation status score and generate a granulation status benefit score for instruction parameter adjustment. Step S63: Based on the relationship curve between instruction parameters and capacity changes, extract the capacity change value and generate a capacity benefit score for instruction parameter adjustments; Step S64: Establish a comprehensive benefit analysis model based on the granulation state benefit score and the capacity benefit score of the instruction parameter adjustment, and generate the comprehensive benefit score of the optimal adjustable instruction parameter adjustment; The specific expression of the comprehensive benefit analysis model is as follows: ; In the expression, This represents the overall benefit score of the optimal adjustable instruction parameter adjustment when the optimal adjustable instruction parameter adjustment amount is j. This represents the pelletizing efficiency score when the parameter adjustment amount of the optimal adjustable command is j. This represents the capacity efficiency score when the parameter adjustment amount of the optimal adjustable instruction is j. This represents the instruction parameters of the optimal adjustable instruction when the parameter adjustment amount of the optimal adjustable instruction is j. This represents the minimum instruction parameters for the optimal adjustable instruction. This represents the maximum instruction parameters of the optimal adjustable instruction. , , All are weighting coefficients, and ; In the above generation method, the preset parameter adjustment amount of the optimal adjustable instruction aims to provide discrete or continuous parameter adjustment options for subsequent comprehensive benefit evaluation. These parameter adjustment amounts can be determined through expert experience, historical data analysis, or equal-interval sampling based on a preset adjustment range. For example, if the parameter adjustment range of the optimal adjustable instruction is [10, 20], the preset adjustment amount can be a series of discrete values ​​such as 10.5, 11, 11.5, ..., 19.5, 20, or these discrete values ​​can be generated by defining a step size.

[0059] The process of extracting granulation state score changes based on the command parameter-granulation state score change curve and generating a granulation state benefit score based on command parameter adjustments quantifies the contribution of specific parameter adjustments to granulation state improvement. The command parameter-granulation state score change curve is a model describing the functional relationship between command parameter changes and granulation state score changes; it can be learned from historical data through regression analysis, interpolation fitting, and other methods. Extracting granulation state score changes involves finding the expected change in granulation state score on this curve based on a preset parameter adjustment amount. The generation of the granulation state benefit score aims to transform this change into a comparable benefit indicator, for example, by comparing the granulation state score change with the current granulation state score or a preset threshold.

[0060] Similarly, the capacity change value is extracted from the command parameter-capacity change curve, and a capacity benefit score for command parameter adjustments is generated. This process aims to assess the impact of specific parameter adjustments on production efficiency. The command parameter-capacity change curve is a model describing the functional relationship between command parameter changes and capacity changes, and can also be learned from historical data through methods such as regression analysis and interpolation fitting. Extracting the capacity change value means finding the corresponding expected change in capacity on the curve based on the preset parameter adjustment amount. The generation of the capacity benefit score aims to transform this capacity change into a comparable benefit indicator, for example, by comparing the capacity change value with the current or expected capacity.

[0061] Finally, a comprehensive benefit analysis model is established based on the granulation state benefit score and the production capacity benefit score of the adjusted command parameters, generating the comprehensive benefit score for the optimal adjustable command parameter adjustment. This step is the core decision support mechanism, used to comprehensively evaluate the overall effect of different parameter adjustment amounts. The comprehensive benefit analysis model weights and combines the granulation state benefit score, production capacity benefit score, and the influence of the parameter adjustment amount itself to generate a single comprehensive benefit score.

[0062] Among them, the granulation status benefit score reflects the improvement of product quality or process stability caused by adjustments; the capacity benefit score reflects the improvement of production efficiency caused by adjustments; and The terms take into account parameter adjustment amounts. Relative position within its permissible range is often used to penalize or reward adjustments that deviate from the center or boundary. For example, there is a tendency to choose adjustments closer to the middle value to increase robustness, or to favor more aggressive adjustments for rapid improvement.

[0063] Weighting coefficient , , It allows for a flexible trade-off between conservative and aggressive approaches to improving pelletizing conditions, increasing production capacity, and adjusting parameters, based on actual production needs and priorities. These weighting coefficients can be set by expert experience or learned from historical decision data through machine learning methods.

[0064] This application's solution addresses the problem of accurately selecting specific parameter adjustments to achieve overall optimization after determining the optimal adjustable command and its adjustment range by introducing a multi-dimensional comprehensive benefit evaluation mechanism. Specifically, the solution first pre-defines a series of possible optimal adjustable command parameter adjustment amounts, providing discrete candidate values ​​for evaluation. Then, for each pre-defined parameter adjustment amount, based on pre-established command parameter-granulation state score change curves and command parameter-capacity change curves, it extracts the corresponding granulation state score change value and capacity change value. These change values ​​are then converted into a granulation state benefit score and a capacity benefit score for command parameter adjustment, thereby quantifying the potential benefits of the adjustment amount in improving granulation state and increasing capacity. Finally, by constructing a comprehensive benefit analysis model, a comprehensive benefit score is generated by weightedly combining the granulation state benefit score, capacity benefit score, and the relative position of the parameter adjustment amount within the adjustment range. This comprehensive benefit score fully reflects the overall impact of different parameter adjustments on the process, enabling the system to quantitatively weigh improvements in pelleting conditions, capacity increases, and the conservatism or aggressiveness of adjustments. This provides a solid quantitative basis for selecting the optimal command parameters. This method allows for refined decision-making based on data and models in complex feed pelleting processes, avoiding the negative impacts of optimizing a single indicator and ensuring the comprehensiveness and effectiveness of process optimization.

[0065] Through the above technical solution, this application enables a multi-dimensional and quantitative comprehensive benefit assessment of the parameter adjustment amount of the optimal adjustable command. Based on determining the optimal adjustable command and its parameter adjustment range, this solution further refines the decision-making process, enabling the system to no longer simply select a command, but to accurately weigh the impact of different parameter adjustment amounts on pelleting state improvement, capacity enhancement, and the adjustment itself. This comprehensive evaluation mechanism avoids suboptimal solutions that may result from single-index optimization, ensuring that the selected parameter adjustment amount maximizes the overall process benefit, thereby achieving more refined and intelligent optimization of the feed pelleting process.

[0066] Preferably, the present invention further proposes a method for generating the granulation state benefit score of the instruction parameter adjustment, specifically including: Through the formula: ; Granulation status benefit score based on generation instruction parameter adjustment ; In the formula, This represents the i-th pelleting status score during the feed pelleting process. This represents the change in granulation status score when the parameter adjustment amount of the optimal adjustable command is j. This represents the granulation status scoring threshold; This formula is used to calculate the granulation state benefit score of command parameter adjustments. Its purpose is to quantify the impact of command parameter adjustments on granulation state into a benefit score for subsequent comprehensive benefit analysis. The formula comprehensively evaluates the benefit of the adjustment by considering the current granulation state score, the change in the granulation state score due to parameter adjustments, and a preset granulation state score threshold.

[0067] The granulation state benefit score of the instruction parameter adjustment indicates the benefit of the adjustment to the granulation state when the parameter adjustment amount of the optimal adjustable instruction is j. The larger the value, the more significant the benefit of the parameter adjustment to improve the granulation state, and thus the greater the contribution to the overall comprehensive benefit.

[0068] The quantitative assessment value of the pelleting state in the current feed pelleting process is generated based on the pelleting state characteristic value in the feed pelleting process (such as the powder state data of raw materials after crushing), reflecting the quality or stability of the current pelleting process.

[0069] When the parameters of the optimal adjustable command are adjusted by a specific amount j, the expected or actual change in the granulation condition score is determined. This change can be positive (indicating improvement) or negative (indicating deterioration). One possible implementation is to obtain this change value by consulting a pre-established command parameter-granulation condition score change curve, which describes the relationship between different parameter adjustments and changes in the granulation condition score.

[0070] A pelleting condition score threshold is a preset benchmark or reference value used to measure the relative magnitude of the pelleting condition benefit score. It can represent an ideal pelleting condition score or an acceptable minimum pelleting condition score. One possible implementation is that the pelleting condition score threshold is a fixed value set based on industry standards, best practices, or historical best production data.

[0071] This application addresses the problem of inaccurate evaluation of pelleting state benefits in feed pelleting process optimization by introducing a calculation method that quantifies the impact of command parameter adjustments on pelleting state. Specifically, after determining the optimal adjustable command and its parameter adjustment range, to evaluate the pelleting state benefits brought about by different parameter adjustment amounts j, this solution uses the pelleting state score in the current feed pelleting process, the expected change in the pelleting state score, and a preset pelleting state score threshold to calculate the pelleting state benefit score of command parameter adjustments using a formula. This calculation method not only considers the direct changes brought about by parameter adjustments... It also incorporates the current granulation status and a baseline threshold. When When the value is positive (indicating improved granulation condition), The smaller the value, the closer the granulation state is to the ideal state, and the better the granulation state benefit score of the command parameter adjustment. The larger the value, the more significant the benefit. Conversely, the lower the value, the less effective the pelleting state benefit score. The smaller the value, the worse the benefit. By incorporating this benefit score into the comprehensive benefit analysis model and weighting it with the production capacity benefit score and the adjustment cost of the command parameters themselves, the final comprehensive benefit score can more comprehensively and accurately reflect the overall benefit brought about by different parameter adjustments. This precise quantification of benefits allows for more effective selection of the best parameters that significantly improve pelleting conditions while balancing production capacity and adjustment costs when establishing the command parameter analysis model and generating optimal command parameters, thereby achieving precise optimization of the feed pelleting process.

[0072] Through the above technical solution, this application provides a more accurate and comprehensive method for evaluating the benefits of pelleting status. This method not only considers the direct impact of instruction parameter adjustments on pelleting status scores but also combines the current pelleting status score with a preset pelleting status score threshold, making the calculation of the pelleting status benefit score more reasonable. This quantitative approach avoids the bias that may result from evaluating solely based on score changes, ensuring that the improvement in pelleting status is accurately reflected in the comprehensive benefit analysis. Therefore, when generating optimal instruction parameters for the optimal adjustable instructions, it is possible to more effectively balance the optimization of pelleting status with factors such as production capacity and adjustment costs, thereby achieving more accurate and effective optimization of the feed pelleting process and improving overall production efficiency and product quality.

[0073] Preferably, the present invention further proposes a method for generating the capacity efficiency score of the instruction parameter adjustment, specifically including: Through the formula: ; Production efficiency score based on generation instruction parameter adjustment ; In the formula, This indicates the current production capacity. This represents the change in production capacity when the parameter adjustment amount of the optimal adjustable command is j. This indicates the expected production capacity; The capacity benefit score for instruction parameter adjustment quantifies the impact of adjusting the parameters of the optimal adjustable instruction on the capacity of the feed pelleting process. It is an indicator that measures the influence of parameter adjustment on production efficiency; a higher value generally indicates a better effect of the parameter adjustment on capacity improvement or maintenance. This score can be generated based on a pre-set mathematical model or by training historical data using machine learning algorithms.

[0074] Current capacity refers to the actual output achieved by feed pelleting equipment or production line within a specific time period before any adjustments to the command parameters. It can be output per unit time (e.g., tons / hour) or batch output. Current capacity can be obtained through real-time sensor data acquisition, such as direct measurement using weighing sensors, flow meters, etc., or through statistical analysis of historical production data recorded by a production management system (MES).

[0075] The change in production capacity when the parameter adjustment amount of the optimal adjustable command is j. This represents the expected or actual change in production capacity after the parameters of the optimal adjustable command are adjusted by the adjustment amount j. This change can be positive (increased capacity) or negative (decreased capacity). The change in production capacity can be obtained by looking up a table or interpolating from a pre-established command parameter-production capacity change curve, or by predicting it using a simulation model, or by actual measurement after small-scale trial production.

[0076] Expected capacity refers to the output that feed pelleting equipment or production line should achieve under ideal or target production conditions. It is usually a target value set based on production plans, equipment design capacity, industry standards, or historical best production records. Expected capacity can be a fixed value or a dynamically adjusted range, and its setting is intended to provide a benchmark for capacity efficiency assessment.

[0077] This application's solution constructs a quantitative capacity efficiency score by considering current capacity, expected capacity, and capacity changes under specific parameter adjustments. When adjusting the parameters of the optimal adjustable command, the current actual production capacity is first obtained, and the potential capacity change at a given parameter adjustment amount j is predicted based on a preset command parameter-capacity change relationship curve. Subsequently, the current capacity is added to the predicted capacity change to obtain the adjusted expected actual capacity, which is then compared with the preset expected capacity. The capacity efficiency score is calculated using this ratio. This calculation method allows the capacity efficiency score to intuitively reflect the contribution of parameter adjustments to achieving the expected capacity target. This capacity efficiency score is then integrated into a comprehensive benefit analysis model as an important component affecting the comprehensive benefit score. In this way, the solution not only focuses on improving granulation conditions but also fully considers improving production efficiency, enabling the final command parameter adjustment scheme to maximize production efficiency while ensuring product quality, thereby solving the problem that focusing solely on granulation conditions may lead to a decrease in capacity.

[0078] The above technical solution enables a quantitative assessment of the impact of parameter adjustments on production capacity when optimizing the feed pelleting process. Introducing a capacity benefit score into the comprehensive benefit analysis model ensures that the system considers not only improvements in pelleting conditions but also increases in production efficiency when selecting optimal parameters. This avoids one-sided optimization that might sacrifice capacity for product quality, thus ensuring that the final optimization scheme maximizes production efficiency while maintaining product quality, achieving a balance between quality and efficiency, and improving the economic benefits and production efficiency of the entire feed pelleting process.

[0079] Preferably, the present invention further proposes the following expression for the instruction parameter analysis model: ; In the expression, This represents the optimal instruction parameters for the optimal adjustable instruction. This represents the comprehensive benefit score of the optimal adjustable instruction parameter adjustment when the parameter adjustment amount of the optimal adjustable instruction is j, and H represents the parameter adjustment range of the optimal adjustable instruction. The instruction parameter analysis model aims to accurately identify the instruction parameters that bring the greatest overall benefit from a preset parameter adjustment range. This model can be implemented as a computational module, such as a software program integrated into an industrial control system, which receives the overall benefit score for different parameter adjustment amounts and the parameter adjustment range as input, and outputs the optimal instruction parameters. Alternatively, the model can also be implemented through an algorithm executed on a dedicated industrial control computer or a cloud optimization platform to achieve maximum operation. (Expression) This is the core logic of the instruction parameter analysis model. It represents finding the parameter adjustment amount j within the parameter adjustment range H of the optimal adjustable instruction that maximizes the comprehensive benefit score of the optimal adjustable instruction parameter adjustment, and determining this parameter adjustment amount as the optimal instruction parameter of the optimal adjustable instruction. This expression can be implemented by iterating through all discrete parameter adjustment amounts j within the parameter adjustment range H and comparing their corresponding values. The value is determined by selecting the j corresponding to the maximum value. When the parameter adjustment range H is a continuous interval, the optimal solution can be found using numerical optimization algorithms, such as gradient ascent.

[0080] The parameter adjustment range H of the optimal adjustable instruction limits the set of values ​​that the parameter j of the optimal adjustable instruction can take. These ranges are usually determined by factors such as equipment physical limitations, process requirements or safety regulations, and can be a discrete set of values ​​or a continuous range of values.

[0081] The proposed solution first generates a pelleting state score based on pelleting state characteristic values ​​and determines whether instruction parameter adjustments are necessary. If adjustments are required, an adjustable instruction analysis model is established based on the Pearson correlation coefficient between adjustable instructions and pelleting state, and the adjustment difficulty indicator value of the adjustable instructions, to generate the optimal adjustable instruction. Based on this, by establishing the relationship curves between instruction parameters and pelleting state score changes, and instruction parameters and production capacity changes, and by pre-setting the parameter adjustment range of the optimal adjustable instruction, a comprehensive benefit analysis model is established to generate a comprehensive benefit score for the optimal adjustable instruction parameter adjustment. Finally, through the instruction parameter analysis model, the parameter adjustment amount j that brings the maximum comprehensive benefit score is accurately identified from the pre-set parameter adjustment range H and determined as the optimal instruction parameter. This series of steps ensures that when process parameters need adjustment, not only can the most relevant instruction be selected for adjustment, but the optimal parameter value of that instruction can also be further accurately determined, thereby achieving refined and maximized optimization of the feed pelleting process. The entire process, from problem identification to final parameter determination, forms a closed-loop intelligent optimization system, avoiding the limitations of human experience-based judgment and improving the scientific nature and accuracy of optimization decisions.

[0082] Through the above technical solution, this application provides a clear and quantitative method to determine the optimal command parameters for the optimal adjustable command. This solves the problem of how to select the best value from many possible parameter values ​​after determining the optimal adjustable command and its parameter adjustment range. This solution establishes a command parameter analysis model and uses a comprehensive benefit score as the decision-making basis, ensuring that the selected parameters maximize overall benefits. This makes the optimization of the feed pelleting process more precise and efficient, avoiding suboptimal results or additional trial-and-error costs caused by improper parameter selection.

[0083] For preferred options, please refer to [link / reference]. Figure 4 In another embodiment of the present invention, a feed pelleting process optimization system is proposed. This system is used to execute the above-described feed pelleting process optimization method, specifically including: The state analysis unit 10 is used to acquire the pelleting state characteristic values ​​during the feed pelleting process and generate a pelleting state score based on the pelleting state characteristic values. Judgment unit 20 is used to determine whether the current feed pelleting process needs to be adjusted according to the pelleting status score; The data acquisition unit 30 is used to obtain the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions if the current feed pelleting process requires adjustment of the instruction parameters. The optimal instruction analysis unit 40 is used to establish an adjustable instruction analysis model based on the adjustable instruction set of the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions, and generate the optimal adjustable instructions. The relationship curve establishment unit 50 is used to establish the relationship curve between instruction parameters and pelleting status score changes and the relationship curve between instruction parameters and production capacity changes, and to preset the parameter adjustment range of the optimal adjustable instruction; wherein, the relationship curve between instruction parameters and pelleting status score changes and the relationship curve between instruction parameters and production capacity changes are established based on the corresponding pelleting status change values ​​and production capacity change values ​​when instruction parameters are adjusted in historical feed pelleting processes; The comprehensive benefit analysis unit 60 is used to establish a comprehensive benefit analysis model based on the parameter adjustment range of the optimal adjustable command, according to the relationship curve of command parameter-granulation status score change and the relationship curve of command parameter-capacity change, and to generate a comprehensive benefit score for the adjustment of the optimal adjustable command parameter. The optimal instruction parameter analysis unit 70 is used to establish an instruction parameter analysis model based on the parameter adjustment range of the optimal adjustable instruction, with the parameter value of the optimal adjustable instruction as the independent variable and the comprehensive benefit score of the optimal adjustable instruction parameter adjustment as the dependent variable, and to generate the optimal instruction parameters of the optimal adjustable instruction. The optimization unit 80 is used to optimize the feed pelleting process based on the optimal instruction parameters of the optimal adjustable instruction. For preferred options, please refer to [link / reference]. Figure 5 The present invention further proposes that the optimal instruction analysis unit 40 specifically includes: The correlation analysis module 41 is used to generate an instruction correlation score based on the Pearson correlation coefficient between the adjustable instructions and the granulation state. The result output module 42 is used to establish an adjustable instruction analysis model based on the adjustment difficulty flag value and instruction relevance score of the adjustable instruction, and generate the optimal adjustable instruction.

[0084] For preferred options, please refer to [link / reference]. Figure 6 The present invention further proposes that the comprehensive benefit analysis unit 60 specifically includes: The preset module 61 is used to preset the parameter adjustment amount of the optimal adjustable command; The state benefit analysis module 62 is used to extract the granulation state score change value based on the relationship curve between instruction parameters and granulation state score, and generate the granulation state benefit score after instruction parameter adjustment. The capacity efficiency analysis module 63 is used to extract the capacity change value based on the command parameter-capacity change relationship curve and generate a capacity efficiency score for command parameter adjustment. The comprehensive benefit output module 64 is used to establish a comprehensive benefit analysis model based on the granulation state benefit score and the production capacity benefit score adjusted by the instruction parameters, and to generate the comprehensive benefit score of the optimal adjustable instruction parameters.

[0085] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing a feed pelleting process, characterized in that, The method specifically includes: Obtain pelleting state characteristic values ​​during the feed pelleting process, and generate pelleting state scores based on these characteristic values. Based on the pelleting status score, determine whether the current feed pelleting process requires adjustment of the command parameters; If the current feed pelleting process requires adjustment of instruction parameters, obtain the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting status, and the adjustment difficulty indicator value of the adjustable instructions; Based on the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between adjustable instructions and pelleting status, and the adjustment difficulty indicator value of adjustable instructions, an adjustable instruction analysis model is established to generate the optimal adjustable instructions. Establish the relationship curves between instruction parameters and pelleting status score changes, and between instruction parameters and production capacity changes, and preset the optimal adjustable instruction parameter adjustment range; wherein, the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes are established based on the corresponding pelleting status change values ​​and production capacity change values ​​when instruction parameters are adjusted in historical feed pelleting processes; Based on the parameter adjustment range of the optimal adjustable command, a comprehensive benefit analysis model is established according to the relationship curves of command parameters-granulation status score change and command parameters-capacity change, and a comprehensive benefit score for the adjustment of the optimal adjustable command parameters is generated. Based on the parameter adjustment range of the optimal adjustable command, with the parameter value of the optimal adjustable command as the independent variable and the comprehensive benefit score of the parameter adjustment of the optimal adjustable command as the dependent variable, an instruction parameter analysis model is established to generate the optimal instruction parameters of the optimal adjustable command. The feed pelleting process is optimized based on the optimal instruction parameters of the optimal adjustable instructions.

2. The method for optimizing the feed pelleting process according to claim 1, characterized in that, The specific methods for generating the granulation status score include: Through the formula: ; Generate granulation status score ; In the formula, This represents the i-th pelleting state characteristic value during the feed pelleting process. This represents the standard value of the i-th pelleting state characteristic during the feed pelleting process. This represents the deviation threshold of the i-th granulation state characteristic.

3. The method for optimizing the feed pelleting process according to claim 1, characterized in that, The specific methods for generating the optimal adjustable instructions include: Based on the Pearson correlation coefficient between adjustable instructions and granulation status, an instruction correlation score is generated; An adjustable instruction analysis model is established based on the adjustment difficulty indicator value and instruction relevance score of adjustable instructions to generate the optimal adjustable instructions.

4. The method for optimizing the feed pelleting process according to claim 3, characterized in that, The specific methods for generating the instruction relevance score include: Through the formula: ; Generate instruction relevance score ; In the formula, This represents the absolute value of the Pearson correlation coefficient between the adjustable instruction k and the i-th granulation state. This represents the absolute threshold of the Pearson correlation coefficient.

5. The method for optimizing the feed pelleting process according to claim 3, characterized in that, The expression for the adjustable instruction analysis model is specifically as follows: ; In the expression, This represents the optimal adjustable instruction. This represents the instruction correlation score between the adjustable instruction k and the i-th granulation state. This represents the normalized value of the adjustment difficulty flag for the adjustable instruction k. , All are weighting coefficients, and .

6. The method for optimizing the feed pelleting process according to claim 1, characterized in that, The specific methods for generating the comprehensive benefit score of the optimal adjustable command parameter adjustment include: The optimal adjustable parameter adjustment amount is preset; Based on the relationship curve between instruction parameters and granulation status score, extract the change value of granulation status score and generate a granulation status benefit score for instruction parameter adjustment. Based on the relationship curve between command parameters and capacity changes, extract the capacity change value and generate a capacity benefit score for command parameter adjustments; A comprehensive benefit analysis model is established based on the granulation state benefit score and the capacity benefit score of the instruction parameter adjustment, and the comprehensive benefit score of the optimal adjustable instruction parameter adjustment is generated. The specific expression of the comprehensive benefit analysis model is as follows: ; In the expression, This represents the overall benefit score of the optimal adjustable instruction parameter adjustment when the optimal adjustable instruction parameter adjustment amount is j. This represents the pelletizing efficiency score when the parameter adjustment amount of the optimal adjustable command is j. This represents the capacity efficiency score when the parameter adjustment amount of the optimal adjustable instruction is j. This represents the instruction parameters of the optimal adjustable instruction when the parameter adjustment amount of the optimal adjustable instruction is j. This represents the minimum instruction parameters for the optimal adjustable instruction. This represents the maximum instruction parameters of the optimal adjustable instruction. , , All are weighting coefficients, and .

7. The method for optimizing the feed pelleting process according to claim 6, characterized in that, The specific methods for generating the granulation state benefit score based on the adjustment of the command parameters include: Through the formula: ; Granulation status benefit score based on generation instruction parameter adjustment ; In the formula, This represents the i-th pelleting status score during the feed pelleting process. This represents the change in granulation status score when the parameter adjustment amount of the optimal adjustable command is j. This represents the granulation status scoring threshold.

8. The method for optimizing the feed pelleting process according to claim 6, characterized in that, The specific methods for generating the capacity efficiency score based on the adjustment of the instruction parameters include: Through the formula: ; Production efficiency score based on generation instruction parameter adjustment ; In the formula, This indicates the current production capacity. This represents the change in production capacity when the parameter adjustment amount of the optimal adjustable command is j. This indicates the expected production capacity.

9. The method for optimizing the feed pelleting process according to claim 1, characterized in that, The specific expression of the instruction parameter analysis model is as follows: ; In the expression, This represents the optimal instruction parameters for the optimal adjustable instruction. This represents the comprehensive benefit score of the optimal adjustable instruction parameter adjustment when the parameter adjustment amount of the optimal adjustable instruction is j, and H represents the parameter adjustment range of the optimal adjustable instruction.

10. A feed pelleting process optimization system, characterized in that, This system is used to execute the feed pelleting process optimization method according to any one of claims 1-9, specifically including: The state analysis unit is used to acquire the pelleting state characteristic values ​​during the feed pelleting process and generate a pelleting state score based on the pelleting state characteristic values. The judgment unit is used to determine whether the current feed pelleting process needs to be adjusted according to the pelleting status score. The data acquisition unit is used to obtain the set of adjustable instructions for the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting status, and the adjustment difficulty indicator value of the adjustable instructions if the current feed pelleting process requires adjustment of the instruction parameters. The optimal instruction analysis unit is used to establish an adjustable instruction analysis model based on the adjustable instruction set of the current feed pelleting process, the Pearson correlation coefficient between the adjustable instructions and the pelleting state, and the adjustment difficulty indicator value of the adjustable instructions, and generate the optimal adjustable instructions. The relationship curve establishment unit is used to establish the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes, and to preset the optimal adjustable instruction parameter adjustment range; wherein, the relationship curves between instruction parameters and pelleting status score changes and between instruction parameters and production capacity changes are established based on the corresponding pelleting status change values ​​and production capacity change values ​​when instruction parameters are adjusted in historical feed pelleting processes; The comprehensive benefit analysis unit is used to establish a comprehensive benefit analysis model based on the parameter adjustment range of the optimal adjustable command, according to the relationship curve of command parameter-granulation status score change and command parameter-capacity change, and to generate a comprehensive benefit score for the adjustment of the optimal adjustable command parameter. The optimal instruction parameter analysis unit is used to establish an instruction parameter analysis model based on the parameter adjustment range of the optimal adjustable instruction, with the parameter value of the optimal adjustable instruction as the independent variable and the comprehensive benefit score of the optimal adjustable instruction parameter adjustment as the dependent variable, and to generate the optimal instruction parameters of the optimal adjustable instruction. The optimization unit is used to optimize the feed pelleting process based on the optimal instruction parameters of the optimal adjustable instruction.

Images