A method and system for production quality control of fermentation fertilizer from agricultural and livestock wastes

Abstract

The present invention provides a production quality control method and system for fermentation fertilizer of agricultural and pastoral wastes, which relates to the technical field of fertilizer preparation, and includes: evenly distributing a monitoring sensor group to perform sequential monitoring and establish a real-time position data stream; locating interpolation points, performing trend fitting, and updating the interpolation data to the real-time position data stream; performing adaptive clustering at the same time node and performing adaptive clustering cluster correction using a preset time step; constructing a fermentation state partition, and each fermentation state partition is provided with an encoding identifier of a state evaluation index set; establishing a mapping difference control strategy using a multi-objective deviation control model; introducing a global control parameter optimization channel, performing global optimization, and performing production control. The present invention solves the technical problem that the prior art usually performs quality control based on a single or a small number of monitoring parameters, ignoring the mutual relationship and influence between different parameters, thereby resulting in low efficiency and unstable quality in the fermentation process.

Description

Technical Field

[0001] The present invention relates to the technical field of fertilizer preparation, and in particular to a production quality control method and system for fermented fertilizer from agricultural and animal husbandry wastes. Background Art

[0002] Agricultural and animal husbandry wastes, such as animal feces, plant residues, etc., as an important organic resource, have wide applications. Converting these wastes into fermented fertilizers can not only solve the problem of waste treatment, but also provide high-quality organic fertilizers for agriculture, improving the yield and quality of crops. However, the production process of fermented fertilizer is a complex biochemical reaction process involving multiple variables, such as temperature, humidity, oxygen concentration, pH value, etc. To ensure that the quality of the fermented fertilizer meets the standards, strict quality control must be carried out on the fermentation process. Existing technologies usually carry out regulation based on single or a small number of monitoring parameters, ignoring the mutual relationship and influence between different parameters, resulting in the inability to achieve global optimization of the fermentation process, and the multi-dimensional interactions not being reasonably regulated, thus leading to low efficiency and unstable quality in the fermentation process. Summary of the Invention

[0003] This application provides a production quality control method and system for fermented fertilizer from agricultural and animal husbandry wastes, aiming to solve the technical problems that existing technologies usually carry out quality control based on single or a small number of monitoring parameters, ignoring the mutual relationship and influence between different parameters, and being unable to achieve global optimization of the fermentation process, thus leading to low efficiency and unstable quality in the fermentation process.

[0004] In the first aspect disclosed in this application, a production quality control method for fermented fertilizer from agricultural and animal husbandry wastes is provided. The method includes: evenly distributing a monitoring sensor group in the fermentation heap body, using the monitoring sensor group to perform time-series monitoring on the fermented fertilizer in the fermentation heap body, and establishing a real-time position data stream; after reading the control accuracy of the production quality, positioning interpolation points with the control accuracy, and performing trend fitting on the real-time position data stream within a preset search area with the interpolation points as the interpolation centers, and updating the interpolation data to the real-time position data stream; performing adaptive clustering on the updated real-time position data stream at the same time node, and performing adaptive clustering cluster correction using a preset time step; constructing a fermentation state partition with the corrected adaptive clustering clusters, and each fermentation state partition is set with an encoded identifier of a state evaluation index set; establishing a mapping difference regulation strategy for each fermentation state partition based on the encoded identifier using a multi-objective deviation regulation model; introducing a global regulation parameter optimization channel, performing global optimization of the mapping difference regulation strategy, and performing production control with the global optimization result.

[0005] The second aspect disclosed in this application provides a production quality control system for agricultural and pastoral waste fermentation fertilizer. The system is used for the production quality control method of the above-mentioned agricultural and pastoral waste fermentation fertilizer, and the system includes: a timing monitoring module, which is used to evenly distribute a monitoring sensor group in the fermentation heap body, use the monitoring sensor group to perform timing monitoring on the fermentation fertilizer in the fermentation heap body, and establish a real-time position data stream; a trend fitting module, which is used to read the control accuracy of production quality, locate the interpolation points with the control accuracy, and perform trend fitting on the real-time position data stream within a preset search area with the interpolation points as the interpolation centers, and update the interpolation data to the real-time position data stream; an adaptive clustering module, which is used to perform adaptive clustering on the updated real-time position data stream at the same time node, and perform adaptive clustering cluster correction using a preset time step; a state partition construction module, which is used to construct a fermentation state partition with the corrected adaptive clustering clusters, and each fermentation state partition is set with an encoding identifier of a state evaluation index set; a regulation strategy establishment module, which is used to establish a mapping difference regulation strategy for each fermentation state partition based on the encoding identifier using a multi-objective deviation regulation model; a global optimization module, which is used to introduce a global regulation parameter optimization channel, perform global optimization of the mapping difference regulation strategy, and perform production control with the global optimization result.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] By evenly distributing a monitoring sensor group within the fermentation pile body, multiple important parameters during the fermentation process can be monitored in real time. These sensors continuously collect data and convert it into a real-time position data stream, providing the necessary basic data for subsequent fermentation state analysis and optimization. Through the frame interpolation technique, the gaps in the real-time data stream are filled to ensure the continuity and integrity of the data. By setting the control accuracy and performing trend fitting on the data within the preset search area, the data stream can be optimized and corrected, reducing data discontinuity and volatility, which makes subsequent analysis and decision-making more reliable. Through the adaptive clustering method, the data is intelligently classified according to time nodes, and areas with similar characteristics are clustered together, enabling more accurate analysis of different regions and states during the fermentation process. Through the correction of the clustering results, the stability and accuracy of the clustering clusters are ensured, contributing to further improving the accuracy of fermentation state analysis. Based on the correction results of the adaptive clustering clusters, the fermentation pile body is divided into multiple fermentation state partitions, each partition is equipped with a unique coding identifier, which not only improves the regional management ability of the fermentation process but also facilitates the state tracking and quality control of different regions. Through the coding identifier, the fermentation state and data of each region can be effectively monitored, facilitating subsequent analysis and optimization. Through the multi-objective deviation regulation model, according to the coding identifier of each fermentation state partition, a targeted regulation strategy is formulated. This strategy can handle the differences in each partition on the basis of multi-objective optimization, ensuring that each fermentation region can reach the predetermined fermentation state under different conditions, ultimately improving the quality of the overall fermented fertilizer. By introducing a global regulation parameter optimization channel, the global optimization of the mapping difference regulation strategy is executed, and the control parameters in the production process are adjusted according to the optimization results. Through this global optimization, the differences between different partitions can be coordinated, the demands of each partition can be balanced, ensuring the high efficiency and stability of the overall fermentation process, optimizing resource utilization, reducing production costs, and improving the overall quality of the fermented fertilizer.

[0008] The above description is only an overview of the technical solution of this application. In order to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the specific embodiments of this application are hereinafter specifically exemplified. Brief Description of the Drawings

[0009] Figure 1 It is a schematic flowchart of a method for controlling the production quality of fermented fertilizer from agricultural and animal husbandry waste provided by an embodiment of this application.

[0010] Figure 2 It is a schematic structural diagram of a system for controlling the production quality of fermented fertilizer from agricultural and animal husbandry waste provided by an embodiment of this application.

[0011] Description of the accompanying drawing reference numerals: the timing monitoring module 10, the trend fitting module 20, the adaptive clustering module 30, the state partition construction module 40, the regulation strategy establishment module 50, the global optimization module 60. Detailed implementation manners

[0012] In an embodiment of the present application, by providing a method and a system for controlling the production quality of fermented fertilizer from agricultural and animal husbandry wastes, the technical problem in the prior art that quality control is usually carried out based on a single or a small number of monitoring parameters, ignoring the mutual relationship and influence between different parameters and being unable to achieve global optimization of the fermentation process, thereby resulting in low efficiency and unstable quality in the fermentation process is solved.

[0013] After introducing the basic principle of the present application, various non-limiting implementation manners of the present application will be specifically introduced below with reference to the accompanying drawings of the specification. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0014] Embodiment 1, as Figure 1 shown, an embodiment of the present application provides a method for controlling the production quality of fermented fertilizer from agricultural and animal husbandry wastes, and the method includes:

[0015] Distribute a monitoring sensor group evenly in the fermentation heap body, and use the monitoring sensor group to perform timing monitoring on the fermented fertilizer in the fermentation heap body to establish a real-time position data stream.

[0016] In the fermentation heap body, a monitoring sensor group is evenly distributed. These sensor groups are used to monitor key parameters in the fermentation heap body, such as temperature, humidity, pH value, etc., and use, for example, an odor VOC sensor group to identify corruption and composting gases. The positions and quantities of these sensors are designed according to the size and shape of the heap body to ensure that the entire fermentation process of the heap body can be monitored comprehensively and accurately. Each sensor group continuously collects and uploads the real-time monitoring data, and records the data in chronological order to form a timing data stream. These timing data streams contain the data of each sensor at different time points and reflect the change process in the fermentation heap body. Among them, in the fermentation heap body, the specific position of each monitoring sensor is known. Therefore, after obtaining the sensor data, these data are matched with the position in combination with the actual position of the sensor to form a real-time position data stream. This real-time position data stream not only records the timing data of each sensor, but also includes the spatial position of the sensor, that is, the specific coordinates of the monitoring point. Through these data, the fermentation state and change trend at different positions in the fermentation heap body can be known at any time.

[0017] After reading the control accuracy of the production quality, locate the interpolation points with the control accuracy, and perform trend fitting on the real-time position data stream within a preset search area with the interpolation points as the interpolation centers, and update the interpolation data to the real-time position data stream.

[0018] In the actual production process, it is required to precisely control various parameters in the fermentation heap. To ensure that the quality control achieves the expected effect, a control accuracy is first set, which means that during the production process, the quality monitoring of the fermented fertilizer will determine the requirements for data accuracy according to pre-set standards (such as temperature and humidity fluctuation ranges).

[0019] Analyze the real-time data stream according to the control accuracy to find key interpolation points. An interpolation point refers to a data point in the time-series data stream that needs to be supplemented, updated, or analyzed in more detail based on the control accuracy requirements. The positioning of the interpolation point is determined according to the quality, change trend, and accuracy requirements of the current monitored data. For example, when there are intervals or drastic changes in the real-time monitored data, the corresponding data points are selected as interpolation points for more refined analysis and processing.

[0020] After determining the interpolation points, perform trend fitting around the interpolation points based on the surrounding real-time data stream. Trend fitting is to predict the change trend during the fermentation process through mathematical models such as curve fitting and regression analysis. By fitting the interpolation points and the data around them, the data at other time points can be estimated. Especially in the case of sparse or missing data, the interpolation points play a role in filling data gaps and complementing the time-series data. After calculating the interpolation data, insert these data into the real-time position data stream. In this way, all monitored data will be updated to ensure that the real-time data stream is more complete and accurate.

[0021] Perform adaptive clustering on the updated real-time position data stream at the same time node and use a preset time step for adaptive clustering cluster correction.

[0022] Perform adaptive clustering on the updated real-time position data stream. The basic idea of adaptive clustering is to group the data according to time nodes, that is, the same time point or time window. Each time node corresponds to a group of data, and these data reflect the change trends of different regions or different parameters in the fermentation heap. The purpose of adaptive clustering is to identify potential similarities or regularities in these data streams and aggregate the data with similar characteristics together. Adaptive clustering can adopt dynamically adjusted algorithms such as the K-means algorithm, DBSCAN, and hierarchical clustering. Adaptive clustering emphasizes dynamically adjusting the clustering parameters according to the actual distribution of the data, such as the number of clusters and distance metrics, in order to obtain more accurate results. According to different characteristics in the real-time data stream, such as temperature, humidity, pH value, etc., these data are divided into different clusters. The data points in the same cluster have certain similarities, and the differences between different clusters are relatively large, corresponding to different fermentation states or regions in the fermentation heap.

[0023] One challenge in the clustering process is how to ensure that the clustering results can accurately reflect the actual changes in the fermentation pile over time. To address this issue, a preset time step is used to correct the clustering clusters. The preset time step is a predefined time interval used to stabilize the clustering clusters and adjust the errors in the clustering process. Within each time step, the clustering clusters are re-evaluated and adjusted to ensure that their clustering results can accurately track the evolution of the fermentation pile state. For example, if the temperature and humidity in a certain area change rapidly, a smaller time step is required for correction; while for areas with slower changes, a longer time step is needed. These corrections ensure the accuracy of the data and the stability of the clustering results.

[0024] Construct a fermentation state partition based on the corrected adaptive clustering clusters, and each fermentation state partition is set with the coding identifier of the set of state evaluation indicators.

[0025] Divide the fermentation pile into different fermentation state partitions according to the corrected adaptive clustering clusters. Each partition represents an area or a set of areas in the fermentation pile with similar fermentation states. The division of the fermentation state partitions can be based on different fermentation parameters, such as temperature, humidity, pH value, oxygen content, etc., and their change trends at different time nodes. For example, one partition represents an area with a higher temperature and a lower humidity, and another partition represents an area with a higher humidity and a lower temperature. The state of each area can reflect its corresponding fermentation environment, which in turn affects the production quality of the fertilizer.

[0026] Each fermentation state partition is set with the coding identifier of the set of evaluation indicators for the state of that area. The set of state evaluation indicators includes the fermentation temperature and humidity gradient index, the oxygen-containing stability index, the acid-base buffering index, and the biological activity trend factor. These indicators are calculated based on the actual monitoring data of the fermentation pile to form specific values to help determine whether the fermentation state of each partition is within the ideal range.

[0027] Use the multi-objective deviation regulation model to establish a mapping difference regulation strategy for each fermentation state partition based on the coding identifier.

[0028] Use the multi-objective deviation regulation model to regulate each fermentation state partition. The multi-objective deviation regulation model analyzes multiple objectives, such as temperature, humidity, pH value, etc., and their interactions. The ultimate goal is to minimize the deviations of these objectives to achieve an ideal fermentation environment. Deviation regulation means taking corresponding regulation measures according to the gap, that is, the deviation, between the actual monitoring data of each fermentation state partition and the set target value.

[0029] Each fermentation state partition has a corresponding coding identifier. Based on these coding identifiers, each partition is mapped to a specific regulation strategy. These strategies adjust the parameters of the fermentation pile according to the fermentation state and corresponding quality indicators of each partition to ensure that the state of each partition is as close as possible to the preset ideal state. The mapping difference regulation strategy takes into account the different characteristics and requirements of each fermentation state partition, so the regulation scheme will be different. For example, a region with higher humidity needs to strengthen ventilation to reduce humidity, while a region with too low temperature needs to be heated to increase temperature.

[0030] Introduce a global regulation parameter optimization channel to perform global optimization of the mapping difference regulation strategy and conduct production control based on the global optimization results.

[0031] To improve the effect of overall production control, a global regulation parameter optimization channel is introduced. The core of this channel is to globally optimize the regulation parameters of the entire fermentation process, not just limited to the local regulation of a single partition. The main task of this optimization channel is to integrate the regulation strategies of each fermentation state partition and coordinate their interactions. Specifically, globally optimize the control parameters (such as temperature, humidity, gas flow, etc.) of all partitions to ensure that the overall environment of the fermentation pile reaches the optimal state. Global optimization can be achieved through intelligent algorithms such as genetic algorithms, particle swarm optimization, simulated annealing, etc. These algorithms can find the best combination of regulation parameters in a multi-dimensional space.

[0032] In the global regulation parameter optimization channel, not only is each fermentation state partition regulated, but also global strategy optimization is carried out to ensure the maximization of the quality of the entire fermentation pile. The goal of global optimization is to reduce the control deviation between partitions and optimize the overall stability of each target variable (such as temperature, humidity, gas concentration, pH value, etc.). During this process, the mapping difference regulation strategy is executed, and the control parameters of each partition are continuously optimized through an iterative method. For example, when the temperature in a certain region is too high, the temperature inside the fermentation pile is reduced as a whole by adjusting the ventilation, heating, etc. of other regions to achieve global control.

[0033] After global optimization is completed, final production control strategies are formulated according to the optimization results. These strategies are directly applied to the production process to guide the real-time regulation of each partition in the fermentation pile. The optimized production control strategies ensure that all links in the fermentation process are coordinated, minimize the deviation of each quality indicator, and thus improve the quality and yield of the final product.

[0034] Furthermore, the introduction of the global regulation parameter optimization channel to perform global optimization of the mapping difference regulation strategy includes:

[0035] After inputting the mapping difference regulation strategy into the global regulation parameter optimization channel, the global optimization objective function is called. The evaluation features of the global optimization objective function include the regional state target approximation feature, the regulation cost feature, and the collaborative influence feature among intervals. After evaluating the mapping difference regulation strategy using the global optimization objective function, the adjustment focus items of the current mapping difference regulation strategy are identified. Using the adjustment focus items as the iteration index, iterative optimization is performed, and the global optimization result is established based on the iterative optimization result.

[0036] Input the mapping difference regulation strategy into the global regulation parameter optimization channel. These regulation strategies have been refined according to the characteristics of each fermentation state partition. Now, these strategies need to be globally adjusted to ensure that the production process of the entire fermentation heap reaches the optimal state. After inputting the mapping difference regulation strategy, the global optimization objective function is called. The core task of this objective function is to evaluate the effect of the mapping difference regulation strategy and provide guidance for the global optimization process.

[0037] The global optimization objective function evaluates the input regulation strategy through three key evaluation features, including the regional state target approximation feature, the regulation cost feature, and the collaborative influence feature among intervals. Among them, the regional state target approximation feature evaluates the gap between the actual state and the target state of each fermentation state partition, that is, the target approximation degree. This indicator measures whether each partition has reached the expected temperature, humidity, pH value and other targets; the regulation cost feature evaluates the cost brought by the regulation measures, including energy consumption, material consumption, etc. If a large amount of energy or cost is required to adjust the state of some regions, the evaluation value of this feature will be relatively high; the collaborative influence feature among intervals evaluates the interaction between different partitions. There may be mutual influence among various partitions in the fermentation heap. The regulation of a certain partition may affect the state of other partitions. For example, too high temperature in a certain partition may affect the humidity of adjacent partitions. Therefore, the regulation strategy needs to consider these collaborative influences to avoid systematic errors.

[0038] Use the global optimization objective function to comprehensively evaluate the input mapping difference regulation strategy, evaluate the current regulation strategy according to the three evaluation features in the objective function, and comprehensively evaluate each score to judge whether the current strategy has reached the optimal optimization state. After the evaluation, the adjustment focus items in the current mapping difference regulation strategy are identified. The adjustment focus items refer to those parts that need special attention in the current state. For example, the state of some partitions deviates far from the target value, or the regulation cost is too high, or there are strong collaborative influence problems. These parts are the adjustment focus items.

[0039] Taking the adjustment focus item as the iteration index, the iteration index indicates which parts are preferentially adjusted during the optimization process, and optimization will be carried out on these focus items. Iterative optimization is a process of repeatedly adjusting the control strategy through optimization algorithms such as the gradient descent method and the genetic algorithm. In each iteration, the relevant control parameters will be adjusted according to the evaluation result of the optimization objective function until a globally optimal solution is found. After multiple rounds of iterative optimization, a global optimization result is finally obtained. This result represents the optimal combination of control strategies, covering the optimal control parameters for each fermentation state partition and ensuring the minimization of control costs while considering the synergistic effects between partitions.

[0040] Furthermore, taking the adjustment focus item as the iteration index and performing iterative optimization includes:

[0041] Obtain the attention data of the adjustment focus item; configure the mapping random perturbation using the attention data, search and update the control parameters with the random perturbation and the adjustment focus item, generate the first-round search update result, and identify the actual search direction of the first-round search update result; perform the search evaluation of multiple rounds of search update results. If the search evaluation result is that the update state is an inactive state and all rounds of searches are random perturbation searches, generate a random perturbation suppression. After adjusting the mapping random perturbation using the random perturbation suppression, continue the iterative optimization.

[0042] Obtain the attention data of the adjustment focus item. The attention data refers to the degree to which each adjustment focus item needs to be key-controlled during the global optimization process. These data reflect the influence of each adjustment focus item on the fermentation state and the final result during the optimization process. The attention data can be obtained through historical data, model simulation, or real-time monitoring data. For example, if the temperature in a certain fermentation state partition deviates significantly from the target, it will cause a higher attention level because the temperature deviation has a greater impact on the fermentation process.

[0043] Configure the mapping random perturbation according to the attention data. This perturbation is a random search mechanism used to explore different combinations of control parameters. By introducing the perturbation, different control schemes can be widely tested in the search space to find the optimal combination of control parameters. The attention data affects the intensity and direction of the perturbation. If the attention level of a certain adjustment focus item is high, the corresponding perturbation will be stronger to more finely control the parameter of this item; if the attention level is low, the perturbation will be weaker to avoid unnecessary over-adjustment.

[0044] With the introduction of perturbations, search and update are carried out in the control parameter space. This process involves changing control parameters (such as temperature, humidity, oxygen flow rate, etc.) to find the optimal solution that can reduce deviations and improve fermentation quality. Based on the perturbation effects of each adjustment concern item, the current control strategy is updated to generate the first-round search and update results, which include a set of new control parameters.

[0045] After the first-round search and update, identify the actual search direction of the first-round search and update results. This step is to determine whether the current control parameters are being optimized in the correct direction. If the search results fail to effectively approach the optimal solution, the search direction needs to be adjusted.

[0046] Perform multiple rounds of search and update. After each round of search, evaluate the results to ensure that the optimization direction of the control parameters is correct. The purpose of this search and evaluation process is to assess whether the search process is effective and whether it can drive the system to gradually approach the global optimal solution. If the search and evaluation results show that the current update status is "inactive", that is, the search process fails to find an effective optimization direction and multiple search rounds are based on random perturbations, then the random perturbation suppression mechanism will be executed. The purpose of random perturbation suppression is to reduce ineffective perturbations and avoid over-exploring search directions that have been proven to be ineffective. The suppression mechanism can be achieved by adjusting the range, frequency, or intensity of the perturbations. In this way, resources can be concentrated on directions that can effectively improve the control results and avoid being trapped in unnecessary random changes.

[0047] After generating and applying the perturbation suppression mechanism, adjust the mapped random perturbations to make them more refined and effective, and continue to execute the next round of iterative optimization process. The adjusted perturbations will be more concentrated in the search area most likely to lead to optimization, thereby improving the optimization efficiency.

[0048] Furthermore, the method of positioning the interpolation point with the control accuracy and performing trend fitting on the real-time position data stream within a preset search area with the interpolation point as the interpolation center and updating the interpolation data to the real-time position data stream includes:

[0049] Conduct a matching evaluation based on the control accuracy and the distribution of the monitoring sensor groups, and use the matching evaluation results to locate the interpolation point; after determining the preset search area, combine any two monitoring sensor groups within the preset search area to establish a combination set; connect the sensor groups for each combination within the combination set, configure trend trust weights based on the point-line distance between the sensor group connection and the interpolation point, then perform trend fitting on all combination sets, and calculate the interpolation data of the interpolation point using the trend fitting results and trend trust weights.

[0050] Match evaluation is carried out according to the control accuracy and the distribution of the monitoring sensor group. The control accuracy refers to the accuracy requirements in monitoring and adjusting the fermentation process, and the sensor group distribution refers to the specific positions of each sensor in the fermentation heap body. The match evaluation process includes evaluating whether the current position of the sensor group can meet the accuracy requirements in the production process, evaluating whether each sensor can collect sufficient information within its coverage area to support precise process control. When the evaluation results confirm which positions have problems of accuracy or data gaps, the interpolation points are located through these evaluation results. The interpolation points refer to the time or space positions where new data needs to be inserted in the data stream. Through the interpolation points, the areas with missing or incomplete data can be supplemented to enhance the accuracy and continuity of the overall data stream.

[0051] Define a preset search area, which is set in the fermentation heap body based on the range of the monitoring sensor distribution and the actual requirements. The preset search area is a space for data analysis and optimization, and data analysis and interpolation operations will be carried out within this area. The selection of the preset search area is based on the existing monitoring data and the structural characteristics of the fermentation heap body to ensure that the search area includes the areas that have a greater impact on the fermentation process while avoiding redundant data analysis. Within the search area, any two monitoring sensor groups are combined to establish a combination set. These sensor group combinations are used to analyze and compare the data differences and relationships between different monitoring points.

[0052] Based on the combination set, sensor group connections are made for each combination. Sensor group connection means establishing a connection between each pair of sensors in the combination set. For example, calculating distances, establishing spatial relationships, etc., to link their monitoring data. These connections help analyze the relationships between different sensors, especially their physical positions, monitoring parameters, etc., and can help better understand the change trends in different areas.

[0053] After connecting the sensor groups, the relationship between the sensor group and the interpolation point is analyzed by calculating the distance of each point-line. The point-line distance refers to the spatial distance between the sensor group and the interpolation point. According to the point-line distance, a trend trust weight is configured for the relationship between each sensor group and the interpolation point. This weight measures the credibility of the data of each sensor group for the interpolation point data. For example, a sensor group closer to the interpolation point has a higher trust level and a larger weight; while a sensor group farther from the interpolation point has a lower trust level and a smaller weight.

[0054] After calculating the point-line distance and trend trust weight of each sensor group from the interpolation points, trend fitting is performed on the data of the sensor groups within all combination sets. Trend fitting uses mathematical models such as regression analysis and curve fitting to fit the data of the sensor groups and generate a trend curve describing the data changes. After completing the trend fitting, the interpolation data of the interpolation points is calculated based on the fitting results and trend trust weight. This data is generated through weighted averaging and conforms to the trend and logic of the overall data stream. The calculated interpolation data will be inserted into the real-time position data stream to fill the data gaps and optimize the continuity and accuracy of the data.

[0055] Furthermore, the adaptive clustering of the updated real-time position data stream at the same time node and the adaptive clustering cluster correction using a preset time step include:

[0056] After taking the data at the same time node as data of the same dimension, perform adaptive clustering of the real-time position data stream under the same dimension data to establish an adaptive clustering cluster; use the preset time step to perform stability search on the origin and development of each dimension adaptive clustering cluster, and update the adaptive clustering cluster using the stability search results to complete the adaptive clustering cluster correction.

[0057] Regarding the monitoring data at the same time node as data of the same dimension for processing means that if multiple sensors collect data at the same moment, these data will be combined into a unified time dimension data set. These data can be measurement values of different sensors, such as temperature, humidity, pH value, etc., but all belong to the same time node and will be processed as a unified data level during the analysis.

[0058] After organizing the data into the same dimension, perform adaptive clustering based on this data. The goal of adaptive clustering is to aggregate the monitoring data with similar characteristics into the same cluster. For example, multiple data dimensions such as temperature and humidity can identify which areas have similar fermentation states through clustering analysis at the same time node. Adaptive clustering algorithms can use methods such as K-means, DBSCAN, hierarchical clustering, etc., and dynamically allocate the data to different clusters according to the similarity of the data. Each cluster represents a group of areas with similar fermentation characteristics at the time node. After the clustering process is completed, a set containing multiple clusters is obtained, and each cluster corresponds to a group of data that shows similar change trends or characteristics at the time node. The purpose of the adaptive clustering cluster is to aggregate the fermentation characteristics of the monitoring points and provide meaningful data grouping for subsequent analysis.

[0059] Search and evaluate the stability of each adaptive clustering cluster using a preset time step. The preset time step refers to the set time interval or time window during the clustering process, which is used to track the evolution and changes of the clustering clusters in the time dimension. During the stability search process, check whether each adaptive clustering cluster remains stable within different time steps. For example, if the characteristics of a certain cluster change drastically, then this cluster needs to be re-evaluated or adjusted. The stability search process helps to judge whether the change trend of the clustering clusters in the time dimension is reasonable and whether it conforms to the actual situation of the fermentation pile body.

[0060] Based on the results of the stability search, update each clustering cluster. This update aims to ensure that the fermentation state reflected by each cluster is more consistent with the actual situation. For example, if the stability of a certain cluster is poor, then it is necessary to increase or decrease the data points within the cluster, or adjust the division criteria of the cluster. Through this stability correction, the structure of the clustering clusters can be dynamically adjusted so that each cluster can better adapt to the changes in the environment and state in the fermentation pile body.

[0061] Furthermore, the set of state evaluation indicators includes a fermentation temperature and humidity gradient index, an oxygen stability index, an acid-base buffering index, and a biological activity trend factor.

[0062] The set of state evaluation indicators includes a fermentation temperature and humidity gradient index, an oxygen stability index, an acid-base buffering index, and a biological activity trend factor. They are used to measure different physical and chemical characteristics during the fermentation process. These indicators help to comprehensively evaluate the fermentation state of different regions in the fermentation pile body, ensure that the fermentation process proceeds under ideal conditions, and ultimately obtain a fermentation fertilizer product that meets the expectations.

[0063] Among them, the fermentation temperature and humidity gradient index is used to measure the change gradient of temperature and humidity in the fermentation pile body. Temperature and humidity are key factors affecting the fermentation process. Excessive or too low temperature and humidity will affect the activity of microorganisms and the fermentation effect. By calculating the gradient of temperature and humidity at different positions in the pile body, that is, the change rate of temperature and humidity, the uniformity of the fermentation pile body can be evaluated. Ideally, temperature and humidity should be evenly distributed to promote the balanced progress of microbial activities and avoid local overheating or over-wetting.

[0064] The oxygen stability index represents the oxygen content and its stability in the fermentation pile body. During the fermentation process, the oxygen content is crucial for the activities of aerobic microorganisms. Too high or too low oxygen concentration will affect the fermentation process. By monitoring the change and stability of oxygen in the fermentation pile body, the normal progress of the aerobic fermentation reaction can be ensured. For example, if this index is too low, it means that the oxygen supply is insufficient and ventilation needs to be increased; if the index is too high, air circulation needs to be restricted or the ventilation volume needs to be reduced to avoid unnecessary energy consumption.

[0065] The acid-base buffering index represents the stability of the acidity and alkalinity (pH value) within the fermentation pile. During the fermentation process, microbial activities produce acidic substances, leading to a decrease in the pH value. An overly low pH value will inhibit the activity of microorganisms and affect the fermentation effect. If the pH value within the pile remains stable and is maintained within an appropriate range (usually between 4 and 7), it indicates that the fermentation process is proceeding well. If the acid-base buffering capacity is insufficient, the pH value can be adjusted by adding alkaline substances (such as lime) to avoid excessive acidification.

[0066] The biological activity trend factor represents the trend of microbial activities within the fermentation pile. Biological activity is directly related to the efficiency of fermentation. Therefore, monitoring biological activity helps predict the progress of fermentation and the quality of the final product. By tracking the trend of microbial activities, it is possible to identify whether there are phenomena of microbial inhibition or death during the fermentation process, and thus take timely measures to adjust the fermentation conditions. If the value of this factor decreases, it indicates a weakening of microbial activities, and environmental conditions such as temperature, humidity, or oxygen content need to be adjusted to promote microbial activity.

[0067] Furthermore, the global optimization objective function is as follows:

[0068] Wherein, characterizes the global optimization objective function, characterizes the set of control parameters for all fermentation state partitions. N is the total number of fermentation state partitions, and i is the fermentation state partition index. is the regional state target approximation weight for the i-th fermentation state partition, is the control parameter The actual state of the i-th fermentation state partition under, is the target state of the i-th fermentation state partition, is the weight coefficient of the regulation cost item, is the control cost function of the i-th fermentation state partition, V is the weight of the collaborative influence term, characterizes the set of state partitions that have a coupled influence with the i-th fermentation state partition. j represents the index of the state partition that has a coupled influence with the i-th fermentation state partition, characterizes the state coupling strength between the i-th fermentation state partition and the j-th fermentation state partition d, is the control parameter The actual state of the j-th fermentation state partition under.

[0069] Specifically, the global optimization objective function is used to optimize various control parameters during the fermentation process to achieve an ideal fermentation state. The formula is as follows: Among them, the first term is the error term, which measures the actual state of each fermentation state partition and the target state The gap between them, through weighted sum of squared errors, aims to minimize the differences between each partition and its target state.

[0070] The second term is the control cost item, which is used to balance the adjustment cost of control parameters , avoid excessive control input, and ensure the economical use of control parameters.

[0071] The third term is the synergy impact item, which measures the synergy impact between different fermentation state partitions. By considering the relationships and synergy effects between partitions, the goal is to reduce unnecessary cross-regional regulation differences, thereby making the state of the entire fermentation heap more stable.

[0072] The purpose of this global optimization objective function is to find a set of optimal control parameters on the basis of minimizing the target errors of each fermentation state partition, control costs, and the synergy impact between partitions , so as to make the overall fermentation process reach the optimal state.

[0073] Furthermore, after production control and management with the global optimization results, it includes:

[0074] Record the control parameters of the fermentation process, and use the control parameter recording results as production labels to be embedded in the origin code; construct a traceable fermentation resume with the origin code embedded with production labels for fermentation fertilizer management.

[0075] Record the control parameters during the fermentation process. The control parameters usually include temperature, humidity, pH value, oxygen concentration, etc. These parameters play a key role throughout the fermentation process and can also be used to trace any problems during the fermentation process. Use the control parameter recording results as production labels and embed them into the origin code. The origin code is the unique identifier for identifying the production location and production batch of the fermentation fertilizer. By embedding the control parameter record into the origin code, the digital storage of production information is realized, enabling each batch of fermentation fertilizer to be traced back to the specific production process. This approach not only ensures the integrity of production data but also enhances production transparency, facilitating subsequent quality inspection and production auditing.

[0076] Use the origin code embedded with production labels to construct a traceable fermentation resume. This resume contains detailed information on the entire production process from raw material input to final product out-of-warehouse, including all key production parameters, production time, production batch, raw materials used, recording of control parameters, and measures related to quality control, etc.

[0077] Building a traceable fermentation resume not only improves the transparency of the production process but also helps with the management of fermented fertilizers. In this way, each batch of fermented fertilizers produced can be accurately traced, and every step in the production process can be audited and verified. For example, if there are quality problems with a certain batch of fermented fertilizers, the traceability system can quickly identify the root cause of the problem and find out the changes in control parameters that led to quality fluctuations, thereby improving the production process.

[0078] Furthermore, after the production control is carried out based on the global optimization results, it also includes:

[0079] Recording and storing the clustering results of each round of fermentation, the global optimization results, and the output effects in the production database; using the production database to update and optimize the strategies for subsequent fermented fertilizer production management.

[0080] Recording the clustering results of each round of fermentation, which reflect the state classification of different regions in the fermentation pile body. Through clustering, the fermentation characteristics of different regions at a certain time node can be identified and grouped. Recording these clustering results helps to understand the dynamic changes in each region of the fermentation pile body; recording the global optimization results, which are the results of globally adjusting the control parameters in the entire fermentation process based on the optimization algorithm. These results represent the optimal control strategies optimized according to the current fermentation state. After each round of fermentation, record the corresponding output effects, that is, the final product quality and yield of the fermented fertilizers. These data are the actual effects of the fermentation process and reflect whether the production meets the expected goals. These data provide important bases for subsequent quality control and production optimization. Storing all the above-recorded data in a unified production database, these data will serve as historical records and provide reference bases for future production management.

[0081] Based on the data stored in the production database, update and optimize the strategies for subsequent fermented fertilizer production management. By analyzing the historical data, patterns, trends, and potential problems in the production process can be identified, thereby improving the production management strategies. For example, by analyzing the clustering results and global optimization results of each round of fermentation, it can be determined which control measures have a positive effect on the fermentation quality and which measures have poor effects. This data-driven management model realizes the continuous improvement and optimization of the production process, thereby enhancing production efficiency, product quality, and management accuracy.

[0082] In summary, the production quality control method for fermented fertilizers from agricultural and animal husbandry waste provided by the embodiments of the present application has the following technical effects:

[0083] By evenly distributing a monitoring sensor group within the fermentation pile body, multiple important parameters during the fermentation process can be monitored in real time. These sensors continuously collect data and convert it into a real-time position data stream, providing the necessary basic data for subsequent fermentation state analysis and optimization. Through the frame interpolation technique, the gaps in the real-time data stream are filled to ensure the continuity and integrity of the data. By setting the control accuracy and performing trend fitting on the data within the preset search area, the data stream can be optimized and corrected, reducing data discontinuity and volatility, which makes subsequent analysis and decision-making more reliable. Through the adaptive clustering method, the data is intelligently classified according to time nodes, and regions with similar characteristics are clustered together, enabling a more precise analysis of different regions and states during the fermentation process. Through the correction of the clustering results, the stability and accuracy of the clustering clusters are ensured, contributing to further improving the accuracy of fermentation state analysis. Based on the correction results of the adaptive clustering clusters, the fermentation pile body is divided into multiple fermentation state partitions, each partition is equipped with a unique coding identifier, which not only improves the regional management ability of the fermentation process but also facilitates the state tracking and quality control of different regions. Through the coding identifier, the fermentation state and data of each region can be effectively monitored, facilitating subsequent analysis and optimization. Through the multi-objective deviation regulation model, according to the coding identifier of each fermentation state partition, a targeted regulation strategy is formulated. This strategy can handle the differences in each partition based on multi-objective optimization, ensuring that each fermentation region can reach the predetermined fermentation state under different conditions, ultimately improving the quality of the overall fermented fertilizer. By introducing a global regulation parameter optimization channel, the global optimization of the mapping difference regulation strategy is executed, and the control parameters in the production process are adjusted according to the optimization results. Through this global optimization, the differences between different partitions can be coordinated, the demands of each partition can be balanced, thereby ensuring the high efficiency and stability of the overall fermentation process, optimizing resource utilization, reducing production costs, and improving the overall quality of the fermented fertilizer.

[0084] Embodiment 2, based on the same inventive concept as the production quality control method for a fermented fertilizer from agricultural and animal husbandry waste in the foregoing embodiment, as Figure 2 shown, the embodiment of the present application provides a production quality control system for a fermented fertilizer from agricultural and animal husbandry waste, and the system includes:

[0085] A timing monitoring module 10, configured to evenly distribute a monitoring sensor group within the fermentation pile body, use the monitoring sensor group to perform timing monitoring on the fermented fertilizer within the fermentation pile body, and establish a real-time position data stream.

[0086] A trend fitting module 20, configured to, after reading the control accuracy of the production quality, locate the interpolation points with the control accuracy, and perform trend fitting on the real-time position data stream within a preset search area with the interpolation points as the interpolation centers, and update the interpolation data to the real-time position data stream.

[0087] The adaptive clustering module 30 is used to perform adaptive clustering on the updated real-time position data stream at the same time node, and perform adaptive clustering cluster correction using a preset time step.

[0088] The state partition construction module 40 is used to construct a fermentation state partition with the corrected adaptive clustering clusters, and each fermentation state partition is set with an encoding identifier of a state evaluation index set.

[0089] The regulation strategy establishment module 50 is used to establish a mapping difference regulation strategy for each fermentation state partition based on the encoding identifier using a multi-objective deviation regulation model.

[0090] The global optimization module 60 is used to introduce a global regulation parameter optimization channel, perform global optimization of the mapping difference regulation strategy, and perform production control with the global optimization result.

[0091] Furthermore, the global optimization module 60 is used to perform the following operation steps:

[0092] After inputting the mapping difference regulation strategy into the global regulation parameter optimization channel, call the global optimization objective function. The evaluation features of the global optimization objective function include regional state target approximation feature, regulation cost feature, and collaborative influence feature between intervals; after evaluating the mapping difference regulation strategy using the global optimization objective function, identify the adjustment attention items of the current mapping difference regulation strategy; use the adjustment attention items as the iteration index, perform iterative optimization, and establish a global optimization result with the iterative optimization result.

[0093] Furthermore, the global optimization module 60 is used to perform the following operation steps:

[0094] Obtain the attention degree data of the adjustment attention items; configure mapping random perturbations using the attention degree data, search and update the regulation parameters with the random perturbations and the adjustment attention items, generate a first-round search and update result, and identify the actual search direction of the first-round search and update result; perform search evaluation of multiple rounds of search and update results. If the search evaluation result is that the update state is inactive and multiple rounds of searches are all random perturbation searches, generate a random perturbation suppression, adjust the mapping random perturbation using the random perturbation suppression, and then continue iterative optimization.

[0095] Furthermore, the trend fitting module 20 is used to perform the following operation steps:

[0096] Perform matching evaluation according to the control accuracy and the distribution of the monitoring sensor groups, and use the matching evaluation results to locate the interpolation points; after determining the preset search area, combine any two monitoring sensor groups within the preset search area to establish a combination set; perform sensor group connection for each combination within the combination set, configure trend trust weights based on the point-line distance between the sensor group connection and the interpolation points, then perform trend fitting on all the combination sets, and calculate the interpolation data of the interpolation points using the trend fitting results and the trend trust weights.

[0097] Furthermore, the adaptive clustering module 30 is used to perform the following operation steps:

[0098] After taking the data at the same time node as the same-dimensional data, perform adaptive clustering of the real-time position data stream under the same-dimensional data to establish adaptive clustering clusters; use the preset time step to perform stability search on the traceability and development of each-dimensional adaptive clustering clusters, and update the adaptive clustering clusters using the stability search results to complete the correction of the adaptive clustering clusters.

[0099] Furthermore, the state evaluation index set includes a fermentation temperature and humidity gradient index, an oxygen-containing stability index, an acid-base buffering index, and a biological activity trend factor.

[0100] Furthermore, the global optimization objective function is as follows:

[0101] Among them, represents the global optimization objective function, represents the set of control parameters for all fermentation state partitions, N is the total number of fermentation state partitions, and i is the fermentation state partition index. is the regional state target approximation weight for the i-th fermentation state partition, is the control parameter The actual state of the i-th fermentation state partition under, is the target state of the i-th fermentation state partition, is the weight coefficient of the regulation cost item, is the control cost function of the i-th fermentation state partition, V is the weight of the collaborative influence item, represents the set of state partitions that have a coupling influence with the i-th fermentation state partition, j represents the index of the state partition that has a coupling influence with the i-th fermentation state partition, represents the state coupling strength between the i-th fermentation state partition and the j-th fermentation state partition d, is the control parameter The actual state of the j-th fermentation state partition under.

[0102] Furthermore, the global optimization module 60 is used to perform the following operation steps:

[0103] Record the control parameters during the fermentation process, and use the recorded results of the control parameters as production labels to be embedded in the origin code; construct a traceable fermentation resume with the origin code embedded with the production label for the management of fermented fertilizers.

[0104] Furthermore, the global optimization module 60 is further configured to perform the following operating steps:

[0105] Record and store the clustering results, global optimization results, and output effects of each round of fermentation in the production database; use the production database to update and optimize the strategies for subsequent fermented fertilizer production management.

[0106] Through the foregoing detailed description of a method for controlling the production quality of fermented fertilizers from agricultural and animal husbandry wastes in this specification, those skilled in the art can clearly know a system for controlling the production quality of fermented fertilizers from agricultural and animal husbandry wastes in this embodiment. Since it corresponds to the method disclosed in the embodiment, it is described relatively simply. For related parts, refer to the description in the method section.

[0107] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for controlling the production quality of fermentation fertilizer from agricultural and livestock wastes, characterized in that The method includes: Uniformly distributing a monitoring sensor group in the fermentation heap body, using the monitoring sensor group to perform sequential monitoring on the fermented fertilizer in the fermentation heap body, and establishing a real-time position data stream; After reading the control accuracy of the production quality, positioning the interpolation point with the control accuracy, and performing trend fitting on the real-time position data stream within a preset search area with the interpolation point as the interpolation center, and updating the interpolation data to the real-time position data stream; Performing adaptive clustering on the updated real-time position data stream at the same time node, and performing adaptive clustering cluster correction using a preset time step; Constructing a fermentation state partition with the corrected adaptive clustering cluster, and each fermentation state partition is set with an encoded identifier of a set of state evaluation indicators; Establishing a mapping difference regulation strategy for each fermentation state partition based on the encoded identifier using a multi-objective deviation regulation model; Introducing a global regulation parameter optimization channel, performing global optimization of the mapping difference regulation strategy, and performing production control with the global optimization result; The introducing a global regulation parameter optimization channel and performing global optimization of the mapping difference regulation strategy includes: After inputting the mapping difference regulation strategy into the global regulation parameter optimization channel, calling a global optimization objective function, and the evaluation features of the global optimization objective function include regional state target approximation features, regulation cost features, and collaborative influence features between intervals; After evaluating the mapping difference regulation strategy using the global optimization objective function, identifying the adjustment attention items of the current mapping difference regulation strategy; Using the adjustment attention items as iteration indices, performing iterative optimization, and establishing a global optimization result with the iterative optimization result, including: Obtaining the attention data of the adjustment attention items; Configuring mapping random perturbations using the attention data, searching and updating the regulation parameters with the random perturbations and the adjustment attention items, generating a first-round search and update result, and marking the actual search direction of the first-round search and update result; Performing search evaluation on multiple rounds of search and update results. If the search evaluation result is that the update state is inactive and multiple rounds of searches are all random perturbation searches, generating a random perturbation suppression, adjusting the mapping random perturbation using the random perturbation suppression, and then continuing the iterative optimization.

2. The production quality control method of a fermentation fertilizer from agricultural and pastoral waste as claimed in claim 1, characterized in that, The positioning the interpolation point with the control accuracy, and performing trend fitting on the real-time position data stream within a preset search area with the interpolation point as the interpolation center, and updating the interpolation data to the real-time position data stream includes: Performing a matching evaluation according to the control accuracy and the distribution of the monitoring sensor group, and using the matching evaluation result to position the interpolation point; After determining the preset search area, combining any two monitoring sensor groups within the preset search area to establish a combination set; Connecting the sensor groups for each combination within the combination set, configuring a trend trust weight based on the point-line distance between the sensor group connection and the interpolation point, then performing trend fitting on all combination sets, and calculating the interpolation data of the interpolation point using the trend fitting result and the trend trust weight.

3. The production quality control method of a fermentation fertilizer from agricultural and animal husbandry wastes as described in claim 1, characterized in that, The performing adaptive clustering on the updated real-time position data stream at the same time node, and performing adaptive clustering cluster correction using a preset time step includes: After taking the data at the same time node as the data of the same dimension, perform adaptive clustering on the real-time position data stream under the data of the same dimension to establish an adaptive clustering cluster. Use the preset time step to perform stability search on the traceability and development of each dimension adaptive clustering cluster, and update the adaptive clustering cluster with the stability search result to complete the adaptive clustering cluster correction.

4. The production quality control method of a fermentation fertilizer from agricultural and pastoral wastes as claimed in claim 1, wherein The state evaluation index set includes a fermentation temperature and humidity gradient index, an oxygen-containing stability index, an acid-base buffering index, and a biological activity trend factor.

5. The production quality control method of a fermentation fertilizer from agricultural and animal husbandry wastes as claimed in claim 1, characterized in that, The global optimization objective function is as follows: Among them, J(Θ) represents the global optimization objective function, Θ = {Θ1, Θ2, …, Θ N} represents the set of control parameters for all fermentation state partitions, N is the total number of fermentation state partitions, i is the fermentation state partition index, ω i is the regional state target approximation weight for the i-th fermentation state partition, s i (θ i ) is the actual state of the i-th fermentation state partition under the control parameter θ i . is the target state of the i-th fermentation state partition, μ is the weight coefficient of the regulation cost term, c i (θ i ) is the control cost function of the i-th fermentation state partition, ν is the weight of the cooperative influence term, represents the set of state partitions that have a coupling influence with the i-th fermentation state partition, j represents the index of the state partition that has a coupling influence with the i-th fermentation state partition, κ ij represents the state coupling strength between the i-th fermentation state partition and the j-th fermentation state partition d, s j (θ j ) is the actual state of the j-th fermentation state partition under the control parameter θ j .

6. The production quality control method of a fermentation fertilizer from agricultural and animal husbandry waste as claimed in claim 1, characterized in that After performing production control with the global optimization result, it includes: Record the control parameters during the fermentation process, and use the control parameter recording result as a production label to embed in the origin code. Construct a traceable fermentation resume with the origin code embedded with the production label for fermentation fertilizer management.

7. The production quality control method of a fermentation fertilizer from agricultural and pastoral wastes according to claim 1, characterized in that, After performing production control with the global optimization result, it also includes: Record and store the clustering results of each round of fermentation, the global optimization result, and the output effect in the production database. Use the production database to update and optimize the strategy for subsequent fermentation fertilizer production management.

8. A production quality control system for fermented fertilizer from agricultural and animal husbandry waste, characterized in that, For implementing the production quality control method of a fermentation fertilizer from agricultural and animal husbandry waste according to any one of claims 1-7, the system includes: A time series monitoring module for evenly distributing a monitoring sensor group in the fermentation heap body, using the monitoring sensor group to perform time series monitoring on the fermentation fertilizer in the fermentation heap body, and establishing a real-time position data stream. A trend fitting module for reading the control accuracy of production quality, positioning the interpolation points with the control accuracy, and performing trend fitting on the real-time position data stream within a preset search area with the interpolation points as the interpolation centers, and updating the interpolation data to the real-time position data stream. An adaptive clustering module for performing adaptive clustering on the updated real-time position data stream at the same time node, and using a preset time step to perform adaptive clustering cluster correction. A state partition construction module for constructing a fermentation state partition with the corrected adaptive clustering cluster, and each fermentation state partition is set with an encoding identifier of the state evaluation index set. A regulation strategy establishment module for establishing a mapping difference regulation strategy for each fermentation state partition based on the encoding identifier using a multi-objective deviation regulation model. A global optimization module for introducing a global regulation parameter optimization channel, performing global optimization of the mapping difference regulation strategy, and performing production control with the global optimization result.

Images