Maotai-flavor liquor vinasse concentrated liquor fermentation treatment fertilizer production system

Abstract

The vinasse concentrated liquor is a stream of concentrated liquor generated in the fermentation process of the vinasse of the Maotai-flavor liquor, and is rich in organic substances such as alcohols and esters. The invention relates to the technical field of fertilizer fermentation, and discloses a fertilizer production system for fermentation treatment of Maotai-flavor liquor vinasse concentrated liquor. According to the system, a fermentation processing area is divided into a plurality of fermentation units, and Maotai-flavor liquor vinasse concentrated liquor component data, organic auxiliary material ratio data and environment monitoring data in each unit are collected to form a multi-source fermentation data set. Based on the data set, a vinasse concentrated liquid metabolic activity index and an auxiliary material decomposition efficiency index are constructed, and an initial fermentation potential value is generated through a dual-threshold response mechanism and a dynamic attenuation model. And carrying out collaborative correction on the initial fermentation potential value according to the material diffusion relevance of the adjacent fermentation units to obtain a corrected fermentation potential value. And performing descending sorting on the corrected fermentation potential values to generate a fermentation priority sequence, and further driving an execution strategy of a fermentation control unit based on the sequence. According to the system, accurate control over the fermentation process is achieved, and scientificity and efficiency of fertilizer production through fermentation treatment of the Maotai-flavor liquor vinasse concentrated liquor are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fertilizer fermentation, in particular to a fermented liquor of Jiang-flavor liquor stillage thick liquid treatment fertilizer production system. BACKGROUND

[0002] In the production process of Jiang-flavor liquor, a large amount of Jiang-flavor liquor stillage thick liquid will be produced. This kind of thick liquid contains rich organic substances such as alcohols and esters. If it is directly discharged, it will not only cause resource waste, but also may have adverse effects on the environment. Fermenting the Jiang-flavor liquor stillage thick liquid into fertilizer is an important way to realize resource recycling. However, the current fermentation treatment technology still has many deficiencies. Traditional fermentation treatment of Jiang-flavor liquor stillage thick liquid relies on empirical operation and lacks systematic monitoring and control means. In terms of fermentation area division, the extensive overall management mode is often used without fine unit division of the fermentation area, making it difficult to implement precise control according to the specific conditions of different areas. For key data in the fermentation process, such as component data of the stillage thick liquid, proportioning data of organic auxiliary materials, and environmental parameters, the collection is not comprehensive or the integration is low, which leads to the inability to form a complete fermentation data system and makes it difficult to support scientific fermentation decisions. In terms of fermentation potential evaluation, the existing technology often uses a single index or a static model for judgment without considering the correlation between the metabolic activity of the stillage thick liquid and the decomposition efficiency of the auxiliary materials. Since the fermentation process has dynamic change characteristics, environmental factors and material composition will change over time. Static evaluation models are difficult to reflect the changes in the fermentation state in real time, which may lead to misjudgment of the fermentation potential. At the same time, there is material diffusion phenomenon between adjacent fermentation units. The traditional technology does not consider this correlation, each unit is evaluated independently, and the mutual influence between units is ignored, which causes deviation between the evaluation results and the actual fermentation state. In terms of fermentation control execution level, there is a lack of priority ranking mechanism based on scientific evaluation, and the implementation of resource allocation and control measures lacks pertinence, often leading to insufficient fermentation in some areas and excess resources in some areas, affecting the overall fermentation efficiency and fertilizer quality. These problems make the stability and reliability of the fermentation treatment of Jiang-flavor liquor stillage thick liquid insufficient, which restricts the large-scale and efficient development of fertilizer production. SUMMARY

[0003] The purpose of the present application is to provide a fermented liquor of Jiang-flavor liquor stillage thick liquid treatment fertilizer production system to solve the problems raised in the background art.

[0004] To achieve the above purpose, the present application provides a fermented liquor of Jiang-flavor liquor stillage thick liquid treatment fertilizer production system, which comprises: The fermentation processing area is divided into a plurality of fermentation units, and component data of the Luzhou-flavor Luzhou-flavor liquor stillage thick liquid, organic auxiliary material proportioning data and environmental monitoring data in each fermentation unit are obtained to form a multi-source fermentation data set; A stillage thick liquid metabolic activity index and an auxiliary material decomposition efficiency index are constructed according to the multi-source fermentation data set, and an initial fermentation potential value is generated through a double-threshold response mechanism and a dynamic attenuation model; The initial fermentation potential value is cooperatively corrected according to material diffusion correlation of adjacent fermentation units to generate a corrected fermentation potential value; The corrected fermentation potential value is sorted in descending order to generate a fermentation priority sequence, and an execution strategy of a fermentation control unit is driven based on the fermentation priority sequence.

[0005] Preferably, the division rule of the fermentation unit specifically comprises: Based on the distribution topology of the fermentation tank group and the layout of the material conveying pipeline, the fermentation area is divided into independent fermentation units with a volume difference within a preset threshold range, and each independent fermentation unit is configured with a dedicated environmental monitoring sensor group.

[0006] Preferably, the specific construction process of the stillage thick liquid metabolic activity index comprises: The concentration of organic matter in the Luzhou-flavor liquor stillage thick liquid is identified based on spectral analysis data; The content of alcohol and ester substances and the acidity value are extracted from the stillage thick liquid component data, and a metabolic activity parameter is calculated through a preset weight model; A metabolic stability index is generated according to the relationship between the metabolic activity parameter and a preset reference value; A metabolic state score is generated according to the microbial proliferation rate in a preset period in the fermentation log; The stillage thick liquid metabolic activity index value is calculated by comprehensively considering the metabolic stability index, the metabolic state score and a preset nutrient conversion coefficient.

[0007] Preferably, the specific construction process of the auxiliary material decomposition efficiency index comprises: The carbon-nitrogen ratio, fiber content and moisture content are extracted from the organic auxiliary material proportioning data and normalized; The normalized parameters are weighted and fused to calculate an auxiliary material decomposition efficiency initial value; When the moisture content is lower than a set threshold, the auxiliary material decomposition efficiency initial value is corrected in an incremental proportion according to the gradient interval divided by the moisture content deviation, and the correction upper limit is a preset maximum correction threshold.

[0008] Preferably, the generation process of the initial fermentation potential value comprises: The stillage thick liquid metabolic activity index alarm threshold and the auxiliary material decomposition efficiency index alarm threshold are set; If any index of the fermentation unit is lower than the corresponding alarm threshold, the initial fermentation potential value is assigned to a preset emergency response value; If all indexes are higher than the alarm threshold, the initial fermentation potential value is calculated according to the metabolic activity index value of the thick liquid and the auxiliary material decomposition efficiency index value through a dynamic attenuation function.

[0009] Preferably, the execution process of the synergistic correction includes: Selecting a target fermentation unit and obtaining the initial fermentation potential value of its adjacent fermentation unit; Generating a material diffusion efficiency value based on the length and flow rate data of the material conveying pipeline; Generating a nutrient complementarity index value by matching the material composition types of adjacent fermentation units and the target fermentation unit; Assigning a synergistic correction coefficient according to the material diffusion efficiency value and the nutrient complementarity index value; Weighting and correcting the initial fermentation potential value of the target fermentation unit based on the synergistic correction coefficient to generate a corrected fermentation potential value.

[0010] Preferably, the generation process of the material diffusion efficiency value includes: Calculating the diffusion timeliness degree according to the shortest conveying time length between the target fermentation unit and the adjacent fermentation unit; Calculating the diffusion path complexity according to the pipeline connection topological relationship; Generating a basic diffusion efficiency value by integrating the diffusion timeliness degree and the diffusion path complexity, and setting a compensation factor based on the pipeline connection type; Correcting the basic diffusion efficiency value through the compensation factor to obtain the material diffusion efficiency value.

[0011] Preferably, the generation process of the nutrient complementarity index value includes: Matching the main material component types of adjacent fermentation units with the material components of the target fermentation unit; If it belongs to a preset complementary combination type, calculating the nutrient complementarity index value according to the proportion of each component and the preset weight; If it does not belong to the complementary combination type, calculating a correlation index based on the material compatibility rules, and generating a nutrient complementarity index value through a preset mapping relationship.

[0012] Preferably, the calculation process of the correlation index includes: Obtaining a historical fermentation data set and counting the co-occurrence frequency of the material components of the target fermentation unit and the adjacent fermentation unit; Querying the dependency weight initial value between components according to the material function dependency table; Calculating a compatibility adjustment coefficient based on fermentation process constraint rules; The compatibility adjustment coefficient is used for modifying the initial value of the dependent weight, and the correlation degree index is generated in combination with the co-occurrence frequency.

[0013] Preferably, the distribution process of the synergistic correction coefficient comprises: An identification of a current fermentation stage is obtained, and a material diffusion weight and a nutrient complementarity weight are set according to the fermentation stage; The synergistic correction coefficient is generated according to the weight, the material diffusion efficiency value and the nutrient complementarity index value.

[0014] Compared with the prior art, the present application has the following advantages: The fermented liquor fermentation treatment fertilizer production system of the Maotai-flavor liquor brings multi-faceted optimization for the fermented liquor fermentation treatment of the Maotai-flavor liquor through multi-dimensional technical design. In the data acquisition link, the system divides the fermentation treatment area into a plurality of fermentation units, and obtains the Maotai-flavor liquor thick liquid component data, organic auxiliary material ratio data and environment monitoring data in each unit to form a multi-source fermentation data set. This fine data acquisition method breaks the limitations of scattered and incomplete data in the traditional fermentation process, and fully captures the key parameters in the fermentation process, providing rich basic information for subsequent fermentation evaluation and control.

[0015] The thick liquid metabolic activity index and the auxiliary material decomposition efficiency index constructed based on the multi-source fermentation data set can reflect the core characteristics of the fermentation process from different dimensions. The metabolic activity index can reflect the conversion activity of organic matter in the thick liquid, and the auxiliary material decomposition efficiency index can reflect the utilization state of the organic auxiliary material in the fermentation environment. The combination of the two makes the description of the fermentation state more comprehensive. The initial fermentation potential value is generated through the double-threshold response mechanism and the dynamic attenuation model, which can dynamically adjust the evaluation standard according to the actual changes in the fermentation process, avoiding the problem of insufficient adaptability of the static evaluation model to the dynamic fermentation process, and making the initial potential value more consistent with the actual state of the fermentation. The initial fermentation potential value is synergistically corrected by considering the material diffusion correlation of adjacent fermentation units, further improving the accuracy of the potential value. In the actual fermentation process, the materials of adjacent units will diffuse and affect each other, and ignoring this correlation will cause deviations in the evaluation results. The synergistic correction mechanism can incorporate this mutual influence into the evaluation system, so that the corrected fermentation potential value can more accurately reflect the actual fermentation capacity of each unit. The modified fermentation potential values are sorted in descending order to generate a fermentation priority sequence, and the execution strategy of the fermentation control unit is driven based on the sequence, so that the fermentation regulation is more targeted and reasonable. Through priority sorting, the differences in resource demand and regulation urgency of each fermentation unit can be determined, so that the fermentation resources are reasonably allocated and differentiated regulation measures are developed. This regulation method based on data and scientific evaluation can avoid the blindness of traditional empirical regulation, optimize resource allocation in the fermentation process, reduce resource waste, improve overall fermentation efficiency, and improve the quality stability of fertilizer products. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A timing diagram of the Jiangxiang Baijiu (liquor) vinasse thick liquid fermentation treatment fertilizer production system according to the present application; Figure 2 A flowchart for constructing the metabolic activity index of the vinasse thick liquid; Figure 3 A flowchart for constructing the auxiliary material decomposition efficiency index; Figure 4 A flowchart for collaborative modification execution. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0018] Please refer to Figure 1 The present application provides a Jiangxiang Baijiu (liquor) vinasse thick liquid fermentation treatment fertilizer production system, which comprises: The fermentation treatment area is divided into several fermentation units, and the Jiangxiang Baijiu (liquor) vinasse thick liquid component data, organic auxiliary material ratio data and environmental monitoring data in each fermentation unit are obtained to form a multi-source fermentation data set. This step involves partitioning operation and data acquisition, ensuring that the data of each fermentation unit is independently traceable.

[0019] The vinasse thick liquid metabolic activity index and the auxiliary material decomposition efficiency index are constructed according to the multi-source fermentation data set, and the initial fermentation potential values are generated through a double-threshold response mechanism and a dynamic decay model. This step includes index calculation and model application, wherein the double-threshold response mechanism sets an alarm threshold to monitor the index change, and the dynamic decay model adjusts the potential value based on the time factor.

[0020] The initial fermentation potential value is cooperatively revised according to the material diffusion correlation of adjacent fermentation units, to generate a revised fermentation potential value. This step analyzes the interaction effects of adjacent units, including the calculation of material diffusion efficiency value and nutrient complementarity index value, to adjust the potential value in a weighted manner.

[0021] The revised fermentation potential value is sorted in descending order to generate a fermentation priority sequence, and the execution strategy of the fermentation control unit is driven based on the fermentation priority sequence. After sorting the potential value, the sequence is output to control the start-stop sequence of the fermentation equipment such as temperature regulator or stirring device.

[0022] Embodiment 1: refer to Figure 2 The division of fermentation units is strictly based on the physical distribution topology of the fermentation tank group and the layout of the material conveying pipeline. The fermentation area is divided into multiple independent fermentation units, and the boundaries of each unit are determined by the tank spacing threshold and the pipeline connection node. The volume difference of each unit is controlled within a preset floating range, which is dynamically adjusted by real-time calibration data of the volume sensor installed at the liquid level monitoring point of each fermentation tank. The volume data is collected by pressure sensing principle. Each independent fermentation unit is equipped with a dedicated multi-parameter environmental monitoring sensor group, including embedded temperature and humidity sensors and replaceable pH probes. The temperature and humidity sensors are installed in the middle of the side wall of the tank, and the pH probes are immersed in the liquid at a preset depth. Environmental monitoring data is uploaded to the central processor at a fixed sampling frequency, and the sampling interval is dynamically configured according to the fermentation stage.

[0023] The collection of Jiangxiang Baijiu (Luzhou-flavor liquor) vinasse thick liquid component data is realized through an online spectrum analysis system. The optical fiber probe of the near-infrared spectrometer is embedded in the observation window of the fermentation tank, and the spectral scanning wavelength range covers the characteristic absorption band of organic matter. After preprocessing, the spectral data are analyzed for alcohol content by a chemometrics model, which is established based on standard alcohol solution calibration. The acidity value is obtained by using an automatic titration analysis unit, and the titration end point is determined by the pH mutation point. The titration solution is quantitatively delivered by a precision peristaltic pump. The alcohol content extracted from the component data and the acidity value are input into a preset weight model, and the weight coefficient is periodically updated according to historical fermentation efficiency data. The model outputs the metabolic activity parameter.

[0024] The generation process of the metabolic stability index is associated with a preset reference curve, which is stored in the fermentation process database and contains standard metabolic values at different fermentation periods. The system calculates the deviation of the metabolic activity parameter from the corresponding period reference value in real time, and converts the deviation into a stability index after smoothing filtering. The metabolic state score is derived from the microbial proliferation monitoring data, which is detected in real time by the optical density method, and the detection probe is integrated at the bottom of the tank. The system calculates the state score according to the growth rate of the microbial population in three consecutive detection periods, and the growth rate threshold is set according to the microbial metabolic kinetics model.

[0025] The comprehensive calculation process for the metabolic activity index of distillers' grains concentrate integrates three inputs: metabolic stability index, metabolic state score, and preset nutrient conversion coefficient. The nutrient conversion coefficient is dynamically matched based on the nitrogen, phosphorus, and potassium elemental detection results of the material, which are performed using X-ray fluorescence spectrometry. The calculation employs a weighted arithmetic mean algorithm, with the weight allocation scheme stored in a coefficient configuration table. The calculation results are mapped to the activity index value in real time, which is used to characterize the metabolic intensity of microorganisms within the fermentation system.

[0026] The processing of organic excipient ratio data begins with the component analysis stage. Carbon content is determined using combustion oxidation-infrared detection, with the detection unit connected to the fermenter's air inlet pipe. Nitrogen content is automatically analyzed using the Kjeldahl method, with digestion and distillation processes completed within a closed reaction chamber. Fiber content is determined according to the standard for neutral detergent fibers, and the solvent is maintained under constant temperature circulation conditions. Moisture content is determined using the loss-of-weight method, with excipient samples dehydrated to constant weight in a microwave drying oven. The raw data for each parameter are normalized, with the conversion formula based on historical extreme values.

[0027] The metabolic activity parameter calculation model employs a multiple linear regression framework. Input variables include the normalized carbon-to-nitrogen ratio, fiber content, and water content. The regression coefficients are determined through orthogonal experimental design optimization. The model output values ​​are used to generate metabolic activity parameters after threshold validation, with validation rules excluding data outliers exceeding physically reasonable ranges. Parameter values ​​are updated in real-time to a central database for use in subsequent indicator calculations.

[0028] The entire implementation process forms a closed-loop data chain: from the physical division of fermentation units to environmental data acquisition, from spectral analysis to component analysis, and from weight calculation to indicator generation. Data interfaces at each stage adopt standardized protocols, and the sensor network is interconnected via an industrial bus architecture. During system execution, any data anomaly triggers an automatic verification mechanism; if verification fails, a backup acquisition channel is activated.

[0029] Example 2: See Figure 3 The construction of the excipient decomposition efficiency index begins with the structured processing of organic excipient ratio data. The carbon-to-nitrogen ratio parameter is obtained using a combustion oxidation-infrared detection unit, integrated at the junction of the fermenter group's air inlet pipes. The combustion chamber maintains a constant-temperature catalytic environment; oxidation products, after drying and filtration, enter the non-dispersive infrared detection module, and carbon concentration is quantified by the area of ​​characteristic absorption peaks. Nitrogen analysis employs a fully automated Kjeldahl nitrogen determination device. The sample digestion process is carried out in a titanium alloy reactor, and the condensate circulation temperature in the steam distillation stage is precisely controlled by a semiconductor cooling chip. Fiber content determination follows a standard neutral washing fiber procedure, with a reflux device equipped with a solvent recovery system, and the glass fiber filter crucible pore size conforming to international standard sieve specifications. The moisture content detection unit includes a microwave drying oven and electronic balance system; the sample weighing chamber has a built-in moisture-proof coating, and the drying endpoint is determined by a mass change rate threshold.

[0030] The extracted raw parameters are processed by a data normalization converter. The conversion baseline value is derived from a historical data database, the carbon-nitrogen ratio is taken from the process's allowed extreme range, the fiber content baseline is associated with the auxiliary material type code, and the moisture content reference value is dynamically adjusted according to seasonal parameters. The normalization algorithm uses linear scaling, mapping the output value to a unified dimensional range. The converted parameters are input into a weighted fusion processor, and the weight coefficients are stored in the auxiliary material attribute matrix. The matrix index is automatically matched based on the auxiliary material source code, with differentiated weight configurations used for agricultural waste and food processing by-products. The fusion calculation uses a scalar product operation mode, outputting the initial value of the auxiliary material decomposition efficiency.

[0031] The moisture content deviation detection module monitors normalized moisture content data in real time. A preset threshold is stored in the process parameter memory, with the threshold setting referencing the material's water-holding capacity curve. When the detected moisture content is below the threshold, the deviation calculation unit starts operating. The deviation value is generated using a percentage difference algorithm, and the calculation result is input into a gradient interval classifier. The classifier has a built-in three-level interval division rule: the slight deviation interval is set to a fixed percentage range below the threshold, the moderate deviation interval expands to double the percentage, and the severe deviation interval covers the remaining low-value range. Each interval corresponds to a preset proportional correction factor, whose value increases linearly with the degree of deviation. The correction engine applies the proportional correction factor to the initial value of the auxiliary material decomposition efficiency based on the interval matching results. The correction process performs multiplication operations, and the output value is limited by a preset upper limit constraint derived from the material decomposition kinetics model.

[0032] The initial fermentation potential value generation process incorporates a dual-threshold monitoring mechanism. Alarm thresholds for the metabolic activity index of the distillers' grains concentrate and the decomposition efficiency index of the auxiliary materials are stored independently in a threshold database. This database is linked to fermentation stage timers, and differentiated threshold parameters are invoked for different fermentation cycles. Threshold settings are based on historical anomaly data mining, and the lower control limit is calculated using statistical process control methods. An index comparison device continuously receives real-time data streams from both indicators and performs threshold comparisons using a parallel processing architecture.

[0033] When either indicator falls below its corresponding alarm threshold, an emergency response trigger is activated. This trigger outputs a preset emergency response value to the initial fermentation potential value register. The emergency response value uses a globally unified constant, set below the lower limit of the normal potential value range. The response status is synchronously written to the system log and triggers an alarm signal to the central console. When both indicators exceed the alarm threshold, the dynamic decay function processor starts running. This processor has a built-in time decay coefficient table, with the coefficients negatively correlated with the fermentation duration. The processor receives real-time data on the metabolic activity indicators of the distillers' grains concentrate and the decomposition efficiency indicators of the auxiliary materials, and calculates the initial potential value base using a two-dimensional interpolation algorithm. After the base is corrected by the time decay coefficient, the initial fermentation potential value is output. The decay calculation uses a product model, with the time parameter derived from the fermentation process timer.

[0034] During the system's operational cycle, all intermediate calculation values ​​are stored in a buffer memory. The buffer data is retained for three fermentation cycles for traceability analysis, and the memory employs a cyclic overwrite storage strategy. The final generated initial fermentation potential value is transmitted to the collaborative correction module, using the industrial real-time data bus standard as the transmission protocol. The entire implementation process forms a complete data chain from auxiliary material detection to potential value output, with each stage's data including a timestamp and quality check code.

[0035] Example 3: See Figure 4 The collaborative correction process begins with the selection of the target fermentation unit. Based on the spatial topology map of the fermenter group, the system sequentially identifies the target units according to a preset scanning order. The spatial topology map is stored in the equipment location database, containing the three-dimensional coordinates and connection relationships of each fermenter. After selecting the target unit, the adjacent unit identification module is activated based on the principle of direct pipe connection, automatically retrieving the set of fermentation units that have physical pipe connections with the target unit. The initial fermentation potential value of the adjacent unit set is synchronously acquired via the data bus; this data originates from the output result cache of Example 2.

[0036] The calculation of material diffusion efficiency relies on a pipeline physical parameter acquisition system. Pipeline length data is obtained from the factory's digital model, which includes pipeline node coordinate information. Real-time flow velocity is acquired via an electromagnetic flowmeter installed at the upstream end of the pipeline junction. The diffusion time efficiency calculation unit receives the shortest transport time parameters between the target unit and each adjacent unit. The transport time is determined by the quotient of the pipeline length divided by the flow velocity, and the calculation result is processed by a unified converter. The diffusion path complexity analysis module calls the pipeline connection topology diagram and automatically calculates the following elements: the number of pipeline branch nodes from the target unit to adjacent units, the total number of elbow connections, and the number of reducers. Each element is weighted and summed according to a preset weight coefficient allocation table to generate a basic path complexity value. The weight coefficients are set based on the degree of influence of pipeline components on flow resistance and stored in a fluid dynamics parameter library.

[0037] The synthesis of the basic diffusion efficiency value employs a specific algorithm model. This model takes as input the basic values ​​of diffusion time efficiency and path complexity, and performs a linear combination operation. The combination weights are dynamically adjusted according to the pipe material type: metal pipes receive a higher time efficiency weight coefficient, while non-metallic pipes receive a higher complexity weight coefficient. The compensation factor generator operates based on a pipe connection type database, which records geometric parameters such as elbow curvature radius and tee branching angle. The compensation factor calculation formula is:

[0038] in: This represents the material diffusion efficiency value; This indicates the numerical value of diffusion timeliness; This represents the basic value of path complexity. This represents the compensation factor. The assignment rules are as follows: right-angle bends are assigned values ​​in the range of 0.8-0.9, obtuse-angle bends are assigned values ​​in the range of 0.9-1.0, and straight pipe sections are assigned values ​​in the range of 1.05-1.10. The specific values ​​are calibrated based on the pipe wall roughness.

[0039] Nutrient complementarity index values ​​are obtained through cross-unit data interaction. The material component type code of the target unit is provided by the distillers' grains concentrate component analysis module, and the coding rules are based on the International Classification of Chemicals (ICC). Material component data of adjacent units are synchronously transmitted through distributed caching. The matching engine compares the component codes of adjacent units with the target unit code, and calls a preset complementarity type lookup table. The lookup table stores typical complementary combination patterns, such as the pairing codes of high-carbon materials and high-nitrogen materials. When a match is successful, the proportion analyzer starts running and extracts the mass fraction of each component in the mixture. The mass fraction data comes from the real-time detection results of the automatic sampling unit, with the sampling point located in the middle of the material conveying pipeline. The proportion data is input into the weight calculation matrix, and the matrix coefficients are preset according to the principle of element balance. The index value generator outputs the nutrient complementarity index to the correction coefficient allocation module.

[0040] The generation process of the synergistic correction coefficient integrates the material diffusion efficiency value and the nutrient complementarity index value. A dual-parameter input dynamic weight allocator is used, with the weight ratio determined based on the current fermentation stage identifier. The stage identifier is derived from the central time controller, dividing fermentation into three states: early stage, middle stage, and late stage. The weight allocation rules are as follows: in the early stage, material diffusion accounts for 70% and nutrient complementarity accounts for 30%; in the middle stage, both account for 50%; and in the late stage, material diffusion accounts for 30% and nutrient complementarity accounts for 70%. The weighted calculation results are mapped to the 0.5-1.5 range by a normalization processor to generate the synergistic correction coefficient.

[0041] The final calculation of the corrected fermentation potential value applies to the initial value of the target unit. The initial fermentation potential value is read from a register, and the co-correction coefficient is obtained from the buffer. The correction operation uses a product model: the initial value is multiplied by the co-correction coefficient, and the result is written to the correction value storage array. This process is repeated until all fermentation units have been corrected. The data pipeline uses a double-check mechanism: data type verification is performed before transmission, and value range checks are performed after reception. The maximum time delay of the entire co-correction process is controlled within twice the system clock cycle.

[0042] Example 4: The generation process of nutrient complementarity index values ​​is illustrated through a specific fermentation unit example. Assume that the material components of the target fermentation unit T, detected by an online analyzer, have the main component codes A-7 (concentrated liquor from fermented baijiu of the soy sauce flavor) and B-3 (wheat straw powder), with mass percentages of 65% and 35%, respectively. Adjacent unit N1 detects component codes A-7 (concentrated liquor from fermented baijiu) and C-9 (soybean meal), while N2 detects B-3 (straw powder) and D-5 (rice husk powder). The system calls a preset complementarity type lookup table for matching and judgment. See Table 1.

[0043] Table 1: Preset Complementary Type Comparison Table.

[0044] Component type code Component name Matching rule classification A-7 Maotai-flavor liquor dreg concentrated liquor Core matrix B-3 Wheat straw powder High-fiber auxiliary material C-9 Soybean meal High-nitrogen auxiliary material D-5 Rice husk powder Inert filler Component matching detection between target unit T and adjacent unit N1: Core matrix A-7 exists in both, and high-fiber auxiliary material B-3 and high-nitrogen auxiliary material C-9 form a preset complementary combination (carbon-nitrogen complementary type). System execution proportion analysis: B-3 accounts for 35% in unit T, and C-9 accounts for 48% in unit N1. The weight allocator calls the element balance table, and the weight coefficient of the carbon-nitrogen complementary combination is set to 0.7. The nutrient complementarity index value is calculated as: 35%×0.7+48%×0.7=58.1%.

[0045] Component matching detection between target unit T and adjacent unit N2: Both contain B-3 (same type of fiber auxiliary material), and D-5 is not in the preset complementarity table. The system initiates correlation index calculation: accessing the historical fermentation database, it statistically analyzes the co-occurrence records of material type T (A-7+B-3) and material type N2 (B-3+D-5) over the past three months. Co-occurrence frequency analysis shows that co-treatment occurred 41 times in 217 fermentation records, with a frequency of 18.9%. Querying the material function dependency table, the initial dependency weight of B-3 and D-5 is 0.3 (both belonging to the fiber category). Fermentation process constraint rule detection: current pH=6.8, temperature=52℃, within the activity range of fiber-decomposing bacteria, the compatibility adjustment coefficient is set to 1.0. Correlation index = 18.9% × 0.3 × 1.0 = 5.67%. Through the mapping relationship table, this index corresponds to a nutrient complementarity index value of 12.3%.

[0046] For scenarios with multiple adjacent units, such as unit N3 containing E-2 (distillers' grains protein), the component matching detection between target units T and N3 shows that the combination of A-7 and E-2 is a pre-defined metabolic activation complementary pair. In unit T, A-7 accounts for 65%, and in unit N3, E-2 accounts for 82%. The activation weighting coefficient is 0.8, and the index value = 65% × 0.8 + 82% × 0.8 = 117.6%. The system automatically truncates values ​​exceeding 100%, ultimately taking 100%.

[0047] Component composition data were acquired using an automated sampling system. The sampler for target unit T was activated after the stirring cycle, extracting 500 mL of the mixture, which was then centrifuged. Solid components were analyzed using near-infrared diffuse reflectance spectroscopy, while liquid components were detected using ion chromatography. Data from adjacent units were synchronized via a distributed acquisition network, employing a timestamp alignment mechanism with a maximum time difference controlled within 200 milliseconds.

[0048] When a new type of uncoded material is detected (such as food waste coded X-0), the system initiates an offline analysis process. Manual sampling is conducted and sent to the laboratory for elemental analysis. The new material's code and attributes are updated to the database within three days. During the transition period, a default correlation index of 5.0% is used, corresponding to an indicator value of 10%. All indicator calculation results are written to a cache queue for sequential use by the collaborative correction module. The historical database employs a rolling update strategy, with new data overwriting the oldest record every ten fermentation cycles.

[0049] Example 5: The allocation of collaborative correction coefficients begins with the capture of fermentation stage identifiers. The system obtains the fermentation duration parameter through a central timing controller, which originates from a high-precision process timer. The stage classifier incorporates time threshold division points, dividing the fermentation cycle into three continuous intervals: the initial stage is defined as the period before the first time division point, the intermediate stage covers the middle time span, and the final stage occupies the last time interval. The timer data is updated once after each acquisition cycle, with the update frequency synchronized with the sensor network. The stage identifier encoding is written to a shared memory area for real-time access by the weight allocation module.

[0050] The weight configuration database stores the preset ratio between material diffusion weights and nutrient complementarity weights. The database uses a three-dimensional matrix structure: the first dimension indexes the fermentation stage code, the second dimension corresponds to the weight type identifier, and the third dimension stores the specific weight ratio values. The ratio values ​​are optimized based on long-term operational data: in the initial stage, a higher weight is used for material diffusion and a lower weight for nutrient complementarity; in the middle stage, the two weight ratios are balanced; and in the final stage, the weight for material diffusion decreases while the weight for nutrient complementarity increases. The database access interface uses direct address mapping, and the retrieval latency is controlled within the system clock cycle.

[0051] The material diffusion efficiency value input channel is connected to the output buffer of Example 3. A data format verification unit verifies whether the input value is within a valid range; invalid data triggers a retransmission mechanism. The nutrient complementarity index value is obtained through a cross-process communication pipeline, which is directly connected to the result register of Example 4. A dual-parameter receiver performs time synchronization alignment; when the difference between the timestamps of the two data streams exceeds a tolerance threshold, a data interpolation compensation program is initiated. The compensation program predicts the possible value at the current moment based on historical data trends.

[0052] The dynamic weight allocation engine is activated upon receiving the complete input dataset. The engine core comprises a proportional calculation unit and a weighted fusion processor. The proportional calculation unit extracts the corresponding proportional parameters from the weight configuration database based on the current fermentation stage identifier. The weighted fusion processor performs a two-parameter linear weighted calculation: the material diffusion efficiency value is multiplied by the corresponding weight ratio, and the nutrient complementarity index value is multiplied by the corresponding weight ratio; the two products are then added to obtain the original coefficient value. The fusion process is executed using a fixed-point arithmetic unit, ensuring computational accuracy that meets industrial control standards.

[0053] The coefficient post-processor performs normalization transformation on the original coefficient values. The transformation rule is based on preset statistical characteristics of the coefficient distribution, mapping the original coefficients to a standard numerical range. The mapping process uses a piecewise linear transformation algorithm, applying different scaling ratios to different value ranges. The normalization result is written to the collaborative correction coefficient register, which employs a dual-backup storage strategy to improve reliability. Before the coefficient values ​​are transmitted to the collaborative correction module, the range checker verifies whether they are within the allowed numerical space.

[0054] The system operation log fully records the key parameters of the coefficient allocation process, including stage identifier acquisition timestamps, database retrieval records, input parameter verification status, intermediate values ​​of fusion calculations, and final coefficient output values. Log entries have unique sequence numbers, and the storage period covers the entire fermentation process. When coefficient values ​​for multiple consecutive periods exceed the reasonable fluctuation range, the system automatically initiates a configuration parameter review process, generating a diagnostic report that is pushed to the maintenance terminal. The maximum execution time of the entire allocation process meets the response time requirements of the real-time control system.

[0055] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0056] 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 fertilizer production system for fermenting concentrated liquor lees from Maotai-flavor liquor, characterized in that, include: The fermentation treatment area was divided into several fermentation units. Data on the composition of concentrated liquor lees of Maotai-flavor liquor, the ratio of organic auxiliary materials, and environmental monitoring data were obtained in each fermentation unit to form a multi-source fermentation dataset. Based on the multi-source fermentation dataset, metabolic activity indexes of distillers' grains concentrate and decomposition efficiency indexes of excipients were constructed, and initial fermentation potential values ​​were generated through a dual-threshold response mechanism and a dynamic decay model. The initial fermentation potential value is collaboratively corrected based on the material diffusion correlation between adjacent fermentation units to generate a corrected fermentation potential value; The modified fermentation potential values ​​are sorted in descending order to generate a fermentation priority sequence, and the execution strategy of the fermentation control unit is driven based on the fermentation priority sequence.

2. The fertilizer production system for fermentation of concentrated liquor lees from Maotai-flavor liquor according to claim 1, characterized in that, The specific rules for dividing the fermentation units include: Based on the topology of the fermentation tank group distribution and the layout of the material conveying pipeline, the fermentation area is divided into independent fermentation units with a volume difference within a preset threshold range, and each independent fermentation unit is equipped with a dedicated set of environmental monitoring sensors.

3. The fertilizer production system for fermentation treatment of concentrated liquor lees from Maotai-flavor liquor according to claim 1, characterized in that, The specific construction process of the metabolic activity index of the distillers' lees concentrate includes: Identify the concentration of organic matter in concentrated liquor lees of Maotai-flavor liquor based on spectral analysis data; The content of alcohols and acidity values ​​were extracted from the composition data of the distillers' grains concentrate, and metabolic activity parameters were calculated using a pre-set weighted model. A metabolic stability index is generated based on the relationship between the metabolic activity parameter and a preset benchmark value; A metabolic status score is generated based on the microbial proliferation rate within a preset period in the fermentation log. The metabolic activity index value of the distillers' grains concentrate is calculated by combining the metabolic stability index, metabolic state score and preset nutrient conversion coefficient.

4. The fertilizer production system for fermentation treatment of concentrated liquor lees from sauce-flavored baijiu as described in claim 1, characterized in that, The specific construction process of the excipient decomposition efficiency index includes: The carbon-nitrogen ratio, fiber content, and moisture content were extracted from the organic auxiliary material ratio data and then normalized. The initial value of the decomposition efficiency of the auxiliary material was calculated by weighted fusion of the normalized parameters. When the moisture content is lower than the set threshold, the gradient interval is divided according to the moisture content deviation, and the initial value of the decomposition efficiency of the auxiliary material is corrected in an incremental manner, with the upper limit of correction being the preset maximum correction threshold.

5. The fertilizer production system for fermentation treatment of concentrated liquor lees from soy sauce-flavored baijiu as described in claim 1, characterized in that, The process of generating the initial fermentation potential value includes: Set alarm thresholds for the metabolic activity index of concentrated distiller's grains and the decomposition efficiency index of excipients; If any fermentation unit's index is lower than the corresponding alarm threshold, the initial fermentation potential value is assigned a preset emergency response value. If all indicators are higher than the alarm threshold, the initial fermentation potential value is calculated using a dynamic decay function based on the metabolic activity index value of the concentrated lees and the decomposition efficiency index value of the auxiliary materials.

6. The fertilizer production system for fermentation treatment of concentrated liquor lees from soy sauce-flavored baijiu as described in claim 1, characterized in that, The execution process of the collaborative correction includes: Select the target fermentation unit and obtain the initial fermentation potential values ​​of its adjacent fermentation units; The material diffusion efficiency value is generated based on the length and flow velocity data of the material conveying pipeline. Nutrient complementarity index values ​​are generated by matching the material composition types of adjacent fermentation units with those of the target fermentation unit. A synergistic correction coefficient is assigned based on the material diffusion efficiency value and the nutrient complementarity index value; The initial fermentation potential value of the target fermentation unit is weighted and corrected based on the aforementioned collaborative correction coefficient to generate a corrected fermentation potential value.

7. The fertilizer production system for fermentation treatment of concentrated liquor lees from Maotai-flavor liquor according to claim 6, characterized in that, The process of generating the material diffusion efficiency value includes: The diffusion efficiency is calculated based on the shortest transport time between the target fermentation unit and adjacent fermentation units. Calculate the diffusion path complexity based on the pipeline connection topology; A basic diffusion efficiency value is generated by combining the diffusion timeliness and diffusion path complexity, and a compensation factor is set based on the pipeline connection type; The material diffusion efficiency value is obtained by correcting the basic diffusion efficiency value using the compensation factor.

8. The fertilizer production system for fermentation treatment of concentrated liquor lees from soy sauce-flavored baijiu as described in claim 6, characterized in that, The process of generating the nutrient complementarity index value includes: Match the main material component types of adjacent fermentation units with the material components of the target fermentation unit; If it belongs to the preset complementary combination type, the nutrient complementarity index value is calculated according to the proportion of each component and the preset weight; If it does not belong to the complementary combination type, the correlation index is calculated based on the material compatibility rules, and the nutrient complementarity index value is generated through the preset mapping relationship.

9. The fertilizer production system for fermentation treatment of concentrated liquor lees from Maotai-flavor liquor according to claim 8, characterized in that, The calculation process of the correlation index includes: Obtain historical fermentation datasets and statistically analyze the co-occurrence frequency of material components between the target fermentation unit and adjacent fermentation units; Query the initial values ​​of the dependency weights between components based on the material function dependency table; Calculate the compatibility adjustment coefficient based on fermentation process constraint rules; The initial values ​​of the dependency weights are corrected based on the compatibility adjustment coefficients, and the correlation index is generated by combining the co-occurrence frequency.

10. The fertilizer production system for fermentation treatment of concentrated liquor lees from Maotai-flavor liquor according to claim 6, characterized in that, The allocation process of the collaborative correction coefficients includes: Obtain the current fermentation stage identifier, and set the material diffusion weight and nutrient complementarity weight according to the fermentation stage; A synergistic correction coefficient is generated based on the weights, material diffusion efficiency values, and nutrient complementarity index values.

Images