A chicken manure drying strategy regulation method and device based on multi-source data fusion

Abstract

The application discloses a chicken manure drying strategy regulation method and device based on multi-source data fusion, and belongs to the technical field of livestock and poultry manure treatment. The method comprises the following steps: acquiring chicken manure material data, breeding end data, drying operation data, environment data and downstream resource demand data, and performing batch identification on each data to obtain chicken manure drying regulation data set; using the chicken manure drying regulation data set for time sequence alignment, feature extraction and fusion calculation to obtain the current drying state of chicken manure and nutrient form conversion prediction results; through comprehensive collection of multi-source data such as chemical properties of chicken manure, breeding end load, drying equipment operation parameters, environmental conditions and downstream resource demand, a full-chain transmission model covering drying parameters, nutrient form conversion and biological suitability of leaching liquid is established, and the chemical change trends such as uric acid conversion, ammonia nitrogen volatilization and organic acid chelation in the drying process can be predicted in real time.

Description

Technical Field

[0001] This invention relates to the field of livestock and poultry manure treatment technology, and in particular to a method and apparatus for regulating chicken manure drying strategy based on multi-source data fusion. Background Technology

[0002] In existing chicken manure treatment, processing chicken manure into slow-release nutrient solution precursors for fishpond fertilization or microalgae cultivation is a highly efficient utilization path that combines pollution control and resource regeneration value, and has received widespread attention in large-scale aquaculture and ecological agriculture in recent years. Drying is the core process in the preparation of this precursor. Its role is not only to reduce the moisture content for storage, transportation and extraction, but also to change the nitrogen speciation, organic acid composition, salt solubility and microbial community structure in chicken manure by controlling temperature, humidity and time. The organic extract made by soaking or rinsing the dried chicken manure needs to have an appropriate amount of available nitrogen to quickly start algae proliferation, a sufficiently low pathogen load to prevent putrefaction and disease, and a salinity and pH buffering capacity that matches the target fish or algae species. Therefore, the quality of the drying strategy directly determines whether the extract can be used as a safe and efficient biological nutrient solution, and thus affects the economic and ecological benefits of the entire chicken manure resource utilization chain.

[0003] Traditional chicken manure drying control generally relies on fixed-level operation or PID regulation based on single temperature and humidity feedback. Operators set hot air temperature, air volume, and turning frequency based on experience, and the equipment aims to maintain the set parameters. Some more advanced systems add a moisture content detector at the end of the drying section and use the moisture content threshold as the drying endpoint criterion to make limited adjustments to the hot air parameters. These technologies essentially simplify the drying process to the physical phenomenon of water evaporation, and the control logic revolves around moisture content and energy consumption, without considering the complex biochemical reactions inside the chicken manure and the relationship with downstream utilization needs, resulting in a significant disconnect between control decisions and the final product quality.

[0004] However, in the aforementioned existing technologies, if the drying temperature is too high or the drying time is too long, a large amount of readily available nitrogen in chicken manure, including uric acid, ammonia nitrogen, and amino acids, will volatilize or be converted into slow-release organic nitrogen. After the extract is made, the fertilizer effect will be greatly reduced, resulting in a delayed fertilization effect in fish ponds and the algae will easily die off due to nitrogen deficiency in the early stage of algae cultivation. Moreover, if the goal is simply to retain readily available nutrients by slow drying at low temperatures, the dehydration efficiency will be significantly reduced, and it will be impossible to effectively kill the miscellaneous bacteria and parasite eggs in chicken manure. This will cause the extract to deteriorate during storage and use, and the introduced miscellaneous bacteria will competitively inhibit the growth of beneficial algae and even induce fish diseases. These two problems make it impossible for traditional drying control, which only focuses on water content and energy consumption, to simultaneously take into account nutrient retention and microbial safety. The quality of the chicken manure precursor produced fluctuates greatly and it is difficult to consistently meet the requirements of downstream aquaculture. In addition, existing technologies lack the means to sense and control the deep chemical processes such as the chelation reaction between organic acids and mineral ions and the changes in the pH buffer system during the drying process, which further limits the adaptability of the dried products in downstream scenarios. Summary of the Invention

[0005] This invention overcomes the shortcomings of the prior art and provides a method and apparatus for regulating chicken manure drying strategy based on multi-source data fusion.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a method for regulating chicken manure drying strategy based on multi-source data fusion, comprising the following steps:

[0007] S1. Acquire chicken manure material data, breeding end data, drying operation data, environmental data, and downstream resource utilization demand data, and identify each data batch to obtain a chicken manure drying control dataset.

[0008] S2. Using the chicken manure drying control dataset, perform time-series alignment, feature extraction, and fusion calculation to obtain the current drying state and nutrient form transformation prediction results of chicken manure.

[0009] S3. Match the current dry state of the chicken manure and the nutrient form transformation prediction results with the downstream resource utilization requirements to obtain target control constraints including moisture content, available nitrogen retention, microbial inactivation and water-soluble nutrient retention;

[0010] S4. Using the target control constraint, candidate drying parameters are screened to obtain a nutrient-oriented variable temperature and humidity drying strategy;

[0011] S5. Implement drying control according to the aforementioned temperature and humidity drying strategy, obtain post-drying detection data and downstream user feedback data, and adjust the subsequent chicken manure drying strategy accordingly.

[0012] In a preferred embodiment of the present invention, in step S1, the chicken manure material data includes one or more of the following: initial moisture content of chicken manure, material temperature, material pH, uric acid content, ammonia nitrogen content, total nitrogen content, volatile fatty acid content, water-soluble phosphorus content, salinity, electrical conductivity, and microbial load.

[0013] In a preferred embodiment of the present invention, in step S1, the downstream resource utilization demand data includes one or more of the following: target fishpond type, target algae species type, target nitrogen speciation spectrum, salinity tolerance range, suitable pH range, required beneficial bacteria type, application time window, and target fertilizer effect activation speed.

[0014] In a preferred embodiment of the present invention, step S2, the fusion analysis includes: performing time-series alignment, missing value filling, normalization processing and feature weighting on the acquired data to obtain a fusion state vector for characterizing water load, equipment processing capacity, nitrogen loss risk, microbial residue risk and downstream demand urgency.

[0015] In a preferred embodiment of the present invention, in step S2, the nutrient form transformation prediction results include one or more of the following: uric acid to ammonia nitrogen conversion prediction results, ammonia nitrogen volatilization loss prediction results, organic nitrogen retention prediction results, volatile fatty acid retention prediction results, water-soluble phosphorus retention prediction results, and salt form change prediction results.

[0016] In a preferred embodiment of the present invention, in step S3, the target regulation constraints include one or more of the following: final moisture content constraint, available nitrogen retention constraint, pathogenic microorganism inactivation constraint, water-soluble nutrient retention constraint, salinity adaptation constraint, and pH adaptation constraint.

[0017] In a preferred embodiment of the present invention, in step S4, the variable temperature and humidity drying strategy includes: a nitrogen protection stage, a microbial control stage, and a dissolution adaptation stage; wherein, the nitrogen protection stage is used to reduce the loss of available nitrogen, the microbial control stage is used to reduce the pathogenic microbial load, and the dissolution adaptation stage is used to adjust the suitability of the extract of dried chicken manure.

[0018] In a preferred embodiment of the present invention, in step S4, the variable temperature and humidity drying strategy includes one or more control parameters selected from hot air temperature, relative humidity, air volume, turning frequency, conveying speed, return air ratio, and stage switching time.

[0019] In a preferred embodiment of the present invention, in step S5, the downstream use feedback data includes one or more of the following: fishpond water color change data, transparency data, plankton density data, algal cell density data, extract bacterial rate data, extract pH data, extract conductivity data, and ammonia nitrogen release curve data.

[0020] Secondly, the present invention provides an apparatus for regulating chicken manure drying strategy based on multi-source data fusion, comprising:

[0021] The data acquisition module is used to acquire data on chicken manure materials, farming operations, drying operations, environmental conditions, and downstream resource utilization needs.

[0022] The fusion analysis module is used to determine the current drying state of chicken manure and the predicted results of nutrient form transformation.

[0023] The constraint determination module is used to determine the target control constraints;

[0024] The strategy generation module is used to generate nutrient-oriented temperature and humidity-controlled drying strategies.

[0025] The control output module is used to output the control parameters corresponding to the variable temperature and humidity drying strategy;

[0026] The feedback correction module is used to correct subsequent chicken manure drying strategies based on post-drying test data and downstream user feedback data.

[0027] This invention addresses the shortcomings of the prior art and has the following beneficial effects:

[0028] (1) By comprehensively collecting multi-source data such as the physicochemical characteristics of chicken manure, the load at the breeding end, the operating parameters of drying equipment, environmental conditions and downstream resource utilization needs, a full-chain transfer model covering drying parameters, nutrient form transformation and the biocompatibility of extracts is established. This model can predict the chemical change trends such as uric acid transformation, ammonia nitrogen volatilization and organic acid chelation during the drying process in real time. Starting from the end, the specific requirements of downstream fish ponds or algae for available nitrogen form, microbial safety, salinity and pH are transformed into a multi-dimensional constraint vector including available nitrogen retention rate, pathogenic bacteria inactivation rate and water-soluble nutrient retention rate, and a batch-customized drying strategy is dynamically generated. This solves the defects of traditional fixed-gear control that cannot respond to material batch differences and downstream diverse needs. Compared with feedback adjustment based solely on moisture content, this model elevates drying control from a simple physical dehydration task to a chemical form-oriented regulation of extract biocompatibility, laying the foundation for the stable production of high-quality slow-release nutrient solution precursors, thereby solving the core contradiction that available nitrogen retention and pathogenic bacteria inactivation cannot be simultaneously addressed.

[0029] (2) The drying strategy is specifically manifested as a nutrient-oriented variable temperature and humidity segmented process curve, which sequentially sets up a nitrogen protection stage, a microbial control stage, and a dissolution adaptation stage, and differentiates the hot air temperature, relative humidity, air volume, and turning parameters for each stage; the nitrogen protection stage inhibits the premature decomposition of uric acid and the escape of ammonia nitrogen through low temperature and high humidity conditions; the microbial control stage adopts short-term precise temperature rise to inactivate pathogenic bacteria while retaining beneficial bacteria such as Bacillus as much as possible; the dissolution adaptation stage regulates the organic acid and the microbial nitrogen through temperature-controlled slow drying. The degree of calcium and magnesium ion binding ensures that the target mineral salts can be fully dissolved during leaching and maintain suitable pH buffering performance; it achieves the triple goals of nitrogen retention, sterilization, and salt adjustment within the same drying batch; compared with the dilemma of traditional high-temperature long-time drying causing a large loss of readily available nitrogen and solidification of organic acids, and low-temperature slow drying leading to the growth of miscellaneous bacteria and excessively high alkalinity of the extract, through the precise control of staged variable parameters, the dried product can provide rapid fertilizer effect while possessing microbial safety and physicochemical compatibility, meeting the stringent requirements of specific algae species or fish ponds.

[0030] (3) A large closed-loop optimization mechanism spanning aquaculture, drying and aquaculture was constructed. Downstream actual usage data such as changes in fishpond water color, plankton density, algal cell density, bacterial rate of extract, pH of extract and ammonia nitrogen release curve were incorporated into the feedback loop as the ultimate basis for modifying the drying strategy generation model. The system adjusts the constraint weights and segmented process parameters of the multi-source data fusion model in reverse according to the real biological effects of extract in the integrated farming scenario, so that the drying control logic continues to evolve. It changed the traditional open-loop mode of drying control that takes moisture content or energy consumption as the endpoint regardless of downstream usage effect. Compared with the current situation where the drying stage and aquaculture application are separated, the terminal aquaculture effect is used as the test standard to ensure the applicability of each batch of dried chicken manure at the field scale, and effectively improve the closed-loop level and industrial value of chicken manure resource utilization. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart of a preferred embodiment of the present invention;

[0033] Figure 2 This is a flowchart of the data acquisition and fusion analysis sub-process of a preferred embodiment of the present invention;

[0034] Figure 3This is a flowchart of the strategy generation and feedback correction sub-process of a preferred embodiment of the present invention;

[0035] Figure 4 This is a device module diagram of a preferred embodiment of the present invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Application Overview:

[0038] During the high-temperature drying process of this invention, the organic acids in chicken manure combine with calcium and magnesium ions to form insoluble salts. These minerals cannot be effectively dissolved during subsequent extraction, resulting in a reduction of available mineral elements in the extract, which in turn affects the nutrient supply to microalgae or fish. On the other hand, under low-temperature and high-humidity drying conditions, chicken manure is prone to ammonification, causing the pH of the material to rise sharply. This results in an excessively alkaline extract, which is not conducive to the cultivation of certain microalgae that require an acidic environment, such as acidophilic Chlorella, and may also impact the acid-base balance of fishpond water. This contradiction involves profound changes in the salt form and pH buffering system during the drying process, which traditional drying methods that use moisture content as the sole control target are completely unaware of and cannot regulate.

[0039] The applicant discovered that drying temperature, humidity, duration, and turning method not only determine the dehydration rate but also directionally affect chemical processes such as uric acid hydrolysis, ammonia volatilization, organic acid chelation, and mineral salt solubility. These changes are precisely the key factors determining whether the extract is suitable for a specific algae species or fishpond. If we start from the biological needs of downstream aquaculture organisms and define multidimensional constraints on the drying process in reverse, including the retention rate of available nitrogen, the inactivation rate of pathogens, the dissolution rate of water-soluble minerals, and the pH range of the extract, we can transform drying control from a simple dehydration task into a targeted regulation process oriented towards the biocompatibility of the extract.

[0040] The applicant discovered that by comprehensively collecting initial physicochemical data of chicken manure materials, load at the breeding end, operating parameters of drying equipment, environmental temperature and humidity, and downstream resource utilization demand data, and through time-series alignment and feature fusion, the drying status can be diagnosed in real time and the trend of nutrient form transformation can be predicted. At the decision-making level, the multi-constraint optimization problem is transformed into solving nutrient-oriented variable temperature and humidity process curves, forming a segmented control strategy for the nitrogen protection stage, the microbial control stage, and the leaching adaptation stage. Finally, by incorporating feedback from downstream use such as fishpond water color, algae density, and extract bacterial rate, a large closed-loop correction mechanism spanning breeding, drying, and aquaculture is constructed, achieving a synergistic improvement in drying efficiency, energy consumption, and biocompatibility of the extract; thus providing a refined technical path for the resource utilization of chicken manure.

[0041] Example 1:

[0042] like Figure 1 As shown, a method for regulating chicken manure drying strategy based on multi-source data fusion includes the following steps:

[0043] S1. Acquire chicken manure material data, breeding end data, drying operation data, environmental data, and downstream resource utilization demand data, and identify each data batch to obtain a chicken manure drying control dataset.

[0044] S2. Using the chicken manure drying control dataset, perform time-series alignment, feature extraction, and fusion calculation to obtain the current drying state and nutrient form transformation prediction results of chicken manure.

[0045] S3. Match the current dry state of the chicken manure and the nutrient form transformation prediction results with the downstream resource utilization requirements to obtain target control constraints including moisture content, available nitrogen retention, microbial inactivation and water-soluble nutrient retention;

[0046] S4. Using the target control constraint, candidate drying parameters are screened to obtain a nutrient-oriented variable temperature and humidity drying strategy;

[0047] S5. Implement drying control according to the aforementioned temperature and humidity drying strategy, obtain post-drying detection data and downstream user feedback data, and adjust the subsequent chicken manure drying strategy accordingly.

[0048] This invention first gathers multi-source data on the same batch of chicken manure, including material collection, breeding, drying operations, environmental factors, and downstream resource utilization needs. A chicken manure drying control dataset is then created using batch identification, ensuring data traceability and consistency. Next, this dataset undergoes time-series alignment, feature extraction, and fusion calculations to determine the current drying status of the chicken manure and predict nutrient form transformation trends, providing quantitative input for subsequent decision-making. Subsequently, the aforementioned status and prediction results are matched with the fertilization or algae cultivation needs of fishponds to form target control constraints such as moisture content, available nitrogen retention, microbial inactivation, and water-soluble nutrient retention, clarifying the direction and boundaries of control. Finally, these target control constraints are used to screen drying parameters, generating a variable temperature and humidity drying strategy. The strategy is then modified based on post-drying detection and downstream usage feedback, achieving continuous optimization of the control strategy.

[0049] The main challenge in this invention, when employing multi-source data fusion and nutrient-oriented drying control, lies in the significant differences in moisture content, nitrogen composition, salinity, and microbial load among different batches of chicken manure. Furthermore, the data sources, sampling frequencies, and dimensions differ between the breeding, drying, environmental, and downstream user ends, making direct application to control the process prone to bias. Simultaneously, there are mutual constraints among dehydration efficiency, available nitrogen retention, microbial inactivation, and the applicability of the extract during the drying process. High-temperature rapid drying may cause nitrogen loss, while low-temperature slow drying may increase microbial risks. Therefore, it is necessary to transform the originally isolated drying control into dynamic regulation oriented towards resource utilization effectiveness through data fusion, nutrient conversion prediction, and downstream feedback correction to address the aforementioned complexities and uncertainties.

[0050] Example 2:

[0051] The technical solution in this embodiment is a further refinement based on the above embodiment 1.

[0052] like Figure 2 As shown, in step S1, data is collected from the material detection end, breeding management end, drying equipment end, environmental monitoring end and downstream demand end before and after the chicken manure enters the drying equipment, and a unique batch number is assigned to the same batch of chicken manure. Step S1 aims to clarify the initial state, breeding source, equipment operation status, environmental changes and final use of the batch of chicken manure, so as to provide complete basic information for subsequent data fusion and decision-making.

[0053] Among them, the data on chicken manure materials reflects the physical, chemical and biological state of chicken manure before and after entering the drying process. Specifically, it can include initial moisture content, material temperature, material pH, uric acid content, ammonia nitrogen content, total nitrogen content, volatile fatty acid content, water-soluble phosphorus content, salinity, electrical conductivity and microbial load, etc.

[0054] Data from the farming end is related to the chicken manure production process, such as the age of the flock, feed formulation, water consumption, manure removal intervals, bedding ratio, and manure collection time.

[0055] Drying operation data refers to the control and status parameters generated during the operation of the drying equipment, such as hot air temperature, relative humidity, air volume, turning frequency, conveying speed, return air ratio, moisture removal, material layer thickness, and energy consumption.

[0056] Environmental data mainly includes ambient temperature, ambient humidity, air pressure, inlet air temperature, and inlet air humidity. Changes in environmental conditions can affect the actual operating conditions and dehydration efficiency of the drying equipment. Incorporating environmental data helps improve the adaptability and robustness of the model.

[0057] Downstream resource utilization demand data is determined based on the target use, such as the target fishpond type, target algae species type, target nitrogen speciation spectrum, salinity tolerance range, suitable pH range, required beneficial bacteria type, application time window, and target fertilizer effect activation speed.

[0058] In this embodiment, the above data is acquired through online sensor acquisition, rapid laboratory testing, equipment PLC reading, synchronization with the aquaculture management system, manual entry, or import via downstream platform interfaces; the availability and diversity of data in actual production are fully considered.

[0059] The system assigns a unique batch number to each batch of chicken manure and binds the batch number with the sampling time, breeding house number, equipment number and target purpose to establish a correspondence between multi-source data.

[0060] After all the data is organized according to a unified field format, a chicken manure drying control dataset is formed for subsequent calculations.

[0061] This embodiment first uses a batch tracking mechanism to establish the correspondence between multi-source data. Chicken manure goes through multiple stages from generation, collection, transportation, drying to downstream use. Data from a single stage cannot accurately describe the resource applicability of a batch of chicken manure. Therefore, this embodiment links data from different sources through batch numbers and timestamps, so that material characteristics, equipment behavior and downstream demand can be calculated in the same data space, providing a structured data foundation for subsequent fusion analysis.

[0062] Specifically, step S1 includes;

[0063] S11. Assign batch numbers to the chicken manure entering the drying process, and record the source of the chicken manure, collection time, intended use, and planned drying equipment number;

[0064] S12. Collect or import chicken manure material data, breeding end data, drying operation data, environmental data and downstream resource utilization demand data corresponding to the batch number;

[0065] S13. Perform field unification, timestamp binding, outlier marking, and batch number association on the collected data to form a chicken manure drying control dataset.

[0066] In step S11, the batch number can be composed of the date, breeding house number, drying line number, and serial number to ensure that each batch has a globally unique identifier; the processing capacity of each batch of chicken manure can be 500-5000 kg, and the material layer thickness can be 5-40 cm; the above range is applicable to conventional continuous or intermittent drying equipment, and it is also convenient to match the test data with the downstream feedback data, which has good versatility.

[0067] In step S12, the moisture content can be obtained using a near-infrared moisture meter, a halogen moisture meter, or by drying and weighing; pH and conductivity can be obtained by mixing chicken manure with deionized water at a mass ratio of 1:5-20 and then detecting the results; uric acid content, ammonia nitrogen content, and total nitrogen content can be obtained using enzymatic methods, spectrophotometry, or the Kjeldahl method; and microbial load can be obtained using colony counting, ATP fluorescence detection, or rapid PCR detection. The detection methods for each indicator can be flexibly selected according to actual conditions and accuracy requirements.

[0068] The sampling period for drying operation data can be 1-60 seconds, the sampling period for environmental data can be 10-300 seconds, and material detection data should be collected at least once before feeding, during stage switching, and after discharging to ensure that key state changes can be captured.

[0069] In step S13, the outlier marking includes range outage marking, sampling missing marking, and mutation marking. For example, when the moisture content detection value is less than 0 or greater than 90%, it is marked as range outage; when the hot air temperature changes by more than 30°C between two consecutive sampling times and the equipment setting value does not change, it is marked as mutation outage. Through this processing, erroneous data is prevented from directly entering the subsequent fusion calculation, thus ensuring data quality.

[0070] Through step S1, this embodiment solves the problems of fragmented data from multiple sources, different sampling frequencies, and untraceable batches, enabling subsequent drying strategies to be generated based on the actual state and target use of the same batch of chicken manure, thus avoiding misjudgments and miscontrols caused by data chaos.

[0071] In step S2, the system performs time alignment, missing data completion, anomaly handling, normalization, and feature extraction on the data obtained in step S1, thereby determining the current drying state of the chicken manure and predicting possible changes in uric acid, ammonia nitrogen, organic nitrogen, volatile fatty acids, water-soluble phosphorus, and salt forms during the subsequent drying process.

[0072] The timing alignment here involves converting data obtained at different sampling frequencies and at different time points onto a unified time axis to ensure the time comparability of each data point.

[0073] Feature extraction involves calculating indicators from the raw data that characterize the significance of drying control, such as water load, nitrogen loss risk, microbial residue risk, salinity-pH mismatch, and downstream demand urgency.

[0074] Fusion computing normalizes and weights data from different sources and with different dimensions to form a fusion state vector, providing comprehensive input for subsequent decision-making.

[0075] Specifically, step S2 includes:

[0076] S21. Read the chicken manure drying control dataset using the batch number as the index, and perform time-series alignment of various types of data according to the preset time window;

[0077] S22. Preprocess missing values, outliers, and data with different dimensions to obtain standardized data;

[0078] S23. Extract water load characteristics, equipment processing capacity characteristics, nitrogen loss risk characteristics, microbial residue risk characteristics, salinity-pH compatibility characteristics, and downstream demand urgency characteristics from standardized data.

[0079] S24. The above features are fused to obtain a fused state vector;

[0080] S25. Predict the nutrient form transformation results based on the fusion state vector and output the current dry state of chicken manure.

[0081] In step S21, the preset time window can be 1-10 minutes. For data with a sampling period shorter than this time window, the mean, median, maximum or minimum value is aggregated.

[0082] For data with a sampling period longer than the time window, forward padding or linear interpolation is used to fill in the gaps; for key nodes such as feeding, stage switching, and discharging, event labels are set to avoid incorrect interpolation of cross-stage data.

[0083] In step S22, missing values ​​are handled as follows: sensor data with a continuous missing time of less than 3 time windows are interpolated linearly; data with a continuous missing time of more than or equal to 3 time windows are filled with the average value of the previous stable phase of the same batch or the average value of the same working condition of adjacent batches, and a missing value marker is added.

[0084] Outliers can be identified using equipment range rules, interquartile range rules, or... Rule identification; data with different dimensions are processed using min-max normalization, and the calculation formula is:

[0085] ;

[0086] in, This represents the original index value. This represents the normalized index value. and These represent the minimum and maximum values ​​of the indicator within the device's allowable range, in the training samples, or in the historical samples, respectively.

[0087] when At this time, the normalized value of the indicator is set to a preset neutral value of 0.5. Through the above normalization process, each indicator becomes a dimensionless value between 0 and 1, avoiding inconsistencies in dimensions caused by directly adding data from different units.

[0088] In step S23, the moisture load characteristics are determined by the initial moisture content, current moisture content, material layer thickness, and feed rate, reflecting the overall scale and current progress of the drying task; the equipment processing capacity characteristics are determined by the hot air temperature, air volume, return air ratio, conveying speed, and dehumidification rate, characterizing the drying capacity of the equipment under the current operating conditions; the nitrogen loss risk characteristics are determined by the material temperature, pH, ammonia nitrogen content, uric acid content, relative humidity, and ventilation intensity, reflecting the potential risks of nitrogen volatilization and conversion.

[0089] Microbial residue risk characteristics are determined by the material core temperature, maintenance time, initial microbial load, and water activity; salinity and pH compatibility characteristics are determined by salinity, conductivity, pH, water-soluble phosphorus, volatile fatty acid content, and the tolerance range of the target algae species or fishpond; downstream demand urgency characteristics are determined by the application time window and the target fertilizer effect activation speed. By extracting the above multi-dimensional characteristics, the drying status and control requirements of the current batch can be comprehensively described from different perspectives.

[0090] In step S24, the fused state vector is:

[0091] ;

[0092] in, Indicates the water content index, This indicates the equipment's processing capacity index. Indicates the nitrogen loss risk index, Indicates the risk index of microbial residues. Indicates the salinity-pH compatibility index. This represents the urgency index of downstream demand; each index is a normalized dimensionless value, ranging from 0 to 1.

[0093] Furthermore, the comprehensive risk index for regulation is:

[0094] ;

[0095] in, This indicates a comprehensive risk index for regulation; to These are weighting coefficients, all ranging from 0 to 1, and satisfying the following:

[0096] ;

[0097] When the intended use is microalgae cultivation, improve and Nitrogen retention and salinity / pH suitability have a greater impact on control results; when the target use is fishpond fertilization and pathogen control requirements are high, increasing... When downstream users have tight deadlines, improve... Since all the indicators involved in the calculation have been normalized, both sides of the formula are dimensionless risk values ​​with consistent dimensions.

[0098] In step S25, nutrient form transformation prediction can employ either an empirical kinetic model or a machine learning model. Taking an empirical kinetic model as an example, the ammonia nitrogen volatilization loss rate is:

[0099] ;

[0100] in, Indicates the ammonia nitrogen volatilization loss rate; This represents the normalized material temperature influence factor; This represents the normalized ventilation intensity influencing factor; This represents the normalized flipping frequency influence factor; This represents the normalized pH influence factor; This represents the empirical coefficient for ammonia nitrogen volatilization; It is not a fixed natural constant, but rather obtained by fitting historical drying experiments.

[0101] Specifically, no fewer than 30 batches, and preferably no fewer than 100 batches of chicken manure samples, can be selected. The changes in ammonia nitrogen content before and after drying can be detected under different temperature, humidity, air volume and turning frequency conditions, and the results can be obtained by least squares fitting. The value can range from 0.05 to 0.80, which corresponds to the range of change from low to high volatility risk under conventional hot air drying conditions for chicken manure.

[0102] The degree of microbial inactivation is as follows:

[0103] ;

[0104] in, Indicates the degree of microbial inactivation; Indicates the temperature-dependent inactivation coefficient; This indicates the cumulative time after the material reaches the effective inactivation temperature. The effective inactivation temperature can be set to 55-75℃; when the goal is to reduce the load of common pathogens, it is preferable to maintain the core temperature of the material at 60-70℃ for 5-30 minutes.

[0105] The value can be obtained by fitting the measured changes in microbial numbers at different temperatures, with a range of 0.03 to 0.50 min. -1 This range can cover the main operating conditions of short-term temperature-heating sterilization, which can not only improve the inactivation effect of pathogenic microorganisms, but also avoid excessive loss of fast-acting nitrogen caused by prolonged high temperature.

[0106] Through step S2, this embodiment solves the problems that multi-source data cannot be directly used for control, the drying state cannot be accurately identified, and the nutrient conversion trend cannot be predicted.

[0107] In step S3, the system matches the current drying state and nutrient form transformation prediction results of chicken manure obtained in step S2 with the needs of downstream fish ponds or algae, and converts the downstream needs into executable target control constraints for the drying process; step S3 is used to convert what kind of extract solution the downstream needs into what control target the drying process must achieve.

[0108] The target control constraints include final moisture content constraints, available nitrogen retention constraints, pathogenic microorganism inactivation constraints, water-soluble nutrient retention constraints, salinity adaptation constraints, and pH adaptation constraints.

[0109] Among them, the final moisture content constraint is the range of moisture content that chicken manure should reach after drying, which directly affects the storage and transportation stability and extraction effect of the product; the available nitrogen retention constraint requires that the readily available nitrogen forms such as uric acid, ammonia nitrogen and amino acid nitrogen be kept within the target range to ensure the rapid activation of fertilizer effect; the pathogenic microorganism inactivation constraint requires that the load of pathogenic bacteria or miscellaneous bacteria after drying be lower than the given limit to ensure the biosafety of the product.

[0110] Water-soluble nutrient maintenance constraints focus on whether extractable nutrients such as water-soluble phosphorus, soluble potassium, calcium and magnesium ions, and volatile fatty acids are within suitable ranges, affecting the nutritional integrity of the extract. Salinity and pH compatibility constraints ensure that the extract made from dried chicken manure is compatible with the target fishpond or target algae species in terms of salinity and pH, avoiding stress on the target organisms.

[0111] Specifically, step S3 includes:

[0112] S31. Read downstream resource utilization demand data to determine the target application type;

[0113] S32. Determine the target moisture content range, target available nitrogen retention rate, target microbial inactivation rate, target water-soluble nutrient retention rate, target salinity range, and target pH range according to the target application type.

[0114] S33. Compare the current drying state and nutrient form transformation prediction results output in step S2 with the target range to obtain the deviation vector;

[0115] S34. Generate target control constraints based on the deviation vector.

[0116] In step S31, the target application types include fishpond fertilization type, microalgae cultivation type, organic extract preparation type, and general storage and transportation type.

[0117] Fishpond fertilization focuses on the speed of fertilizer activation and microbial safety; microalgae cultivation focuses on nitrogen speciation, pH and salinity; organic extract preparation focuses on the retention of water-soluble nutrients and extraction stability; general storage and transportation focuses on final moisture content and storage stability. Different types of applications correspond to different constraints and weighting configurations.

[0118] In step S32, the final moisture content can be set to 10-35%; when long-term storage is required, the final moisture content is preferably 10-18%; when rapid subsequent extraction and short storage time are required, the final moisture content is preferably 18-30%; the available nitrogen retention rate can be set to 60-90%, preferably 70-85%.

[0119] The inactivation rate of pathogenic microorganisms can be set to 90-99.9%; the retention rate of water-soluble nutrients can be set to 60-95%; the pH range of the extract can be set according to the target application, which is 6.0-8.5 for microalgae culture and 6.5-9.0 for fishpond fertilization; the salinity or conductivity range is determined according to the target fish species, algae species or water conditions. The above ranges are set by comprehensively considering the technical feasibility in actual production and the quality requirements of downstream applications.

[0120] In step S33, the deviation vector is:

[0121] ;

[0122] in, This indicates the deviation between the current or predicted final moisture content and the target moisture content. This indicates the deviation between the predicted available nitrogen retention rate and the target available nitrogen retention rate. This indicates the deviation between the predicted microbial inactivation rate and the target inactivation rate. This indicates the deviation between the predicted water-soluble nutrient retention rate and the target retention rate. Indicates salinity deviation. This indicates pH deviation.

[0123] In step S34, When the nitrogen content is negative and the risk of nitrogen loss is high, the target control constraint increases the nitrogen protection weight and limits the high-temperature ventilation time; when When the value is negative and the risk of microbial residue is high, the target regulation constraint is to increase the sterilization temperature or prolong the effective inactivation time; when or When the target range is exceeded, the target control constraint adjusts the temperature, humidity and drying rate of the subsequent dissolution adaptation stage to reduce pH shift caused by excessive solidification or ammoniation of organic acids.

[0124] Through step S3, this embodiment solves the problem that traditional drying control cannot respond to differentiated downstream needs, enabling the drying endpoint and process parameters to be dynamically determined around the quality requirements of fishponds, algae species, or extracts.

[0125] like Figure 3 As shown, in step S4, the system selects a suitable combination of parameters for the current batch of chicken manure from the candidate hot air temperature, relative humidity, air volume, turning frequency, conveying speed, return air ratio and stage switching time based on the target control constraints obtained in step S3, and generates a nutrient-oriented variable temperature and humidity drying strategy.

[0126] The candidate drying parameters set includes the control parameters that the drying equipment can execute; the variable temperature and humidity drying strategy is a process curve that adjusts the temperature, humidity, air volume, turning frequency and stage duration according to different stage objectives during the same batch drying process; nutrient orientation means that the strategy is generated not only by considering dehydration efficiency, but also by considering nitrogen retention, microbial inactivation, water-soluble nutrient retention and extraction compatibility.

[0127] This embodiment divides the drying process into a nitrogen protection stage, a microbial control stage, and a dissolution adaptation stage. The nitrogen protection stage is used to reduce premature decomposition of uric acid and ammonia nitrogen escape. The microbial control stage is used to reduce the pathogen load in a short period of time. The dissolution adaptation stage is used to adjust the pH, salinity, and mineral nutrient dissolution capacity of the dried chicken manure during extraction. By controlling the process in stages, the contradiction between nitrogen preservation and sterilization caused by a single high-temperature or low-temperature process is avoided.

[0128] Specifically, step S4 includes:

[0129] S41. Generate a set of candidate parameters based on the target control constraints;

[0130] S42. Perform constraint screening on the candidate parameter set and remove parameter combinations that may lead to nitrogen loss, microbial residue, energy consumption exceeding limits or downstream adaptation failure.

[0131] S43. Evaluate the retained parameter combinations comprehensively and select the parameter combination with the highest score;

[0132] S44. Convert the highest-scoring parameter combination into a variable temperature and humidity drying strategy that includes a nitrogen protection phase, a microbial control phase, and a dissolution adaptation phase.

[0133] In step S41, the candidate range for hot air temperature can be 35-85℃, the candidate range for relative humidity can be 20-80%, the candidate range for air volume can be 30-100% of the rated air volume of the equipment, the turning frequency can be once every 5-60 minutes, the conveying speed can be 0.1-5m / min, and the return air ratio can be 0-70%. The above ranges cover the adjustable range of conventional hot air drying equipment and can meet the needs of most production scenarios.

[0134] The hot air temperature during the nitrogen protection stage can be 35-55℃, the relative humidity can be 50-80%, the air volume can be 30-60% of the rated air volume, and the stage duration can be 20-50% of the total drying time. This stage reduces the driving force of ammonia nitrogen volatilization by using lower temperature and higher humidity, while avoiding rapid crusting of chicken manure surface which affects internal moisture migration, thus retaining more readily available nitrogen for subsequent stages.

[0135] The hot air temperature in the microbial control stage can be 60-80℃, the relative humidity can be 25-55%, the air volume can be 50-90% of the rated air volume, and the stage duration can be 5-60 minutes. It is preferable to ensure that the core temperature of the material reaches 55-70℃ and is maintained for 5-30 minutes. This stage achieves the inactivation of pathogens through short-term precise heating, avoiding the loss of a large amount of fast-acting nitrogen caused by prolonged high temperature, and achieving the effect of rapid killing and minimal loss.

[0136] The hot air temperature during the dissolution adaptation stage can be 40-65℃, the relative humidity can be 30-65%, the air volume can be 40-80% of the rated air volume, and the stage duration can be 20-50% of the total drying time. This stage regulates the degree of binding between organic acids and calcium and magnesium ions through medium-low temperature slow drying, reduces the risk of excessive insoluble mineral salts, and controls the pH increase caused by ammoniation reaction, thus laying the foundation for the physicochemical compatibility of the extract.

[0137] In step S42, the constraint screening adopts a combination of hard and soft constraints. The hard constraints include the upper limit of the equipment's safe temperature, the target final moisture content, the minimum microbial inactivation rate, and the maximum allowable energy consumption, which are rigid conditions that must be met. The soft constraints include the available nitrogen retention rate, the water-soluble nutrient retention rate, pH compatibility, and salinity compatibility, which are flexible targets that are optimized as much as possible under the premise of meeting the hard constraints. If the combination of candidate parameters causes the predicted final moisture content to be higher than the target upper limit, or the predicted pathogenic bacteria inactivation rate to be lower than the target lower limit, they are directly eliminated to ensure the basic usability of the screening results.

[0138] In step S43, the overall score is:

[0139] ;

[0140] in, This indicates the overall score; Indicates the retention score of readily available nitrogen; Indicates the microbial inactivation score; This indicates the score for retaining water-soluble nutrients; This indicates the salinity and pH compatibility score; Indicates energy efficiency rating; to is the weighting coefficient, with a value ranging from 0 to 1, and satisfying the following:

[0141] ;

[0142] For microalgae culture applications, improve and For use as a fertilizer in fishponds, it can improve... and For short-term, rapid use scenarios, improve and The weights; each score is calculated based on the normalized closeness between the predicted value and the target value, for example... All are dimensionless values ​​between 0 and 1, so the dimensions on both sides of the formula are consistent.

[0143] Through step S4, this embodiment solves the problem that traditional fixed-gear drying cannot simultaneously achieve nitrogen retention, sterilization, salt adjustment and pH adaptation; through segmented control of variable temperature and humidity, the same batch of chicken manure can achieve nutrient protection, microbial control and extraction adaptation at different stages.

[0144] In step S5, the system controls the operation of the drying equipment according to the variable temperature and humidity drying strategy generated in step S4, and collects detection data after drying is completed. Feedback data is collected after downstream use, and these data are then used to correct the drying strategy for subsequent batches.

[0145] The data obtained after drying include moisture content, pH, ammonia nitrogen content, uric acid content, total nitrogen content, water-soluble phosphorus content, conductivity, salinity, volatile fatty acid content, and microbial load, which are used to verify whether the drying strategy has achieved the expected results.

[0146] Downstream user feedback data includes data on changes in fishpond water color, transparency, plankton density, algal cell density, bacterial content in the extract, pH of the extract, conductivity of the extract, and ammonia nitrogen release curves.

[0147] Specifically, step S5 includes:

[0148] S51. Send the variable temperature and humidity drying strategy to the drying equipment controller so that the drying equipment can perform drying according to the stage parameters;

[0149] S52. During the drying process, monitor the actual hot air temperature, relative humidity, air volume, material temperature, and material moisture content, and compare them with the target strategy curve.

[0150] S53. After drying is completed, collect the post-drying detection data to determine whether the target control constraints are met.

[0151] S54. After the dried chicken manure is made into an extract or used in fish ponds or algae cultivation, downstream usage feedback data is collected.

[0152] S55. Adjust the subsequent chicken manure drying strategy based on the test data after drying and the feedback data from downstream users.

[0153] In step S51, the controller can be a PLC, an industrial computer, an edge controller, or a cloud-edge collaborative control system. The specific selection can be determined according to the production scale and automation level. The control parameters are issued in the form of a stage table, including stage number, target temperature, target humidity, target air volume, turning frequency, conveying speed, return air ratio, and stage switching conditions, to ensure that the equipment can execute accurately according to the preset strategy.

[0154] In step S52, when the actual hot air temperature deviates from the target temperature by more than ±3℃ and lasts for more than 2 minutes, the system adjusts the heating power or issues an alarm; when the actual relative humidity deviates from the target humidity by more than ±10% and lasts for more than 5 minutes, the system adjusts the exhaust valve, return air ratio, or air intake volume; when the material center temperature does not meet the requirements of the microbial control stage, the system extends the duration of this stage, but the extension time is preferably no more than 50% of the original stage duration to avoid excessive nitrogen loss. Through the above real-time monitoring and dynamic adjustment, the robustness and adaptability of strategy execution are improved.

[0155] In step S53, if the moisture content after drying is higher than the target upper limit, the air volume in the dissolution adaptation stage of subsequent batches will be increased or the final drying time will be extended; if the available nitrogen retention rate is lower than the target lower limit, the temperature in the microbial control stage of subsequent batches will be reduced, the high temperature maintenance time will be shortened, or the relative humidity in the nitrogen protection stage will be increased; if the microbial load is higher than the target limit, the material center temperature in the microbial control stage of subsequent batches will be increased or the effective inactivation time will be extended.

[0156] In step S54, downstream feedback can be collected at 1h, 6h, 24h, 48h, or 72h after use to capture changes in effectiveness at different time scales; changes in pond water color are obtained through image acquisition devices or manual level recording; transparency is measured using a transparency disc; plankton density and algal cell density are obtained through microscopic counting, flow cytometry, or optical density detection; and the pH, conductivity, and ammonia nitrogen release curves of the extract are obtained through online detection or laboratory testing. This diverse feedback data provides multi-dimensional information support for a comprehensive evaluation of the drying strategy's effectiveness.

[0157] Through step S5, this embodiment solves the problem of the disconnect between traditional drying control and downstream application effects, enabling the drying strategy to be transformed from a one-time setting to a closed-loop optimization based on actual usage effects.

[0158] Example 3:

[0159] The technical solution in this embodiment is a further refinement based on the above embodiments.

[0160] The core invention of this invention is that it does not regard chicken manure drying as a simple dehydration process, but rather as a coupled and regulated process in which water migration, nitrogen form transformation, organic acid changes, mineral salt solubility changes, and microbial inactivation occur simultaneously. Furthermore, starting from the needs of downstream fish ponds or algae species, it defines multi-dimensional target constraints for the drying process in reverse, and then generates a staged temperature and humidity-changing drying strategy through multi-source data fusion.

[0161] This approach breaks away from the traditional control logic that drying is completed once the moisture content reaches the target. Instead, it moves the control objective forward to the drying process and the evaluation criteria backward to the downstream application effect, thereby establishing a closed-loop relationship between the aquaculture end, the drying end, and the aquatic application end.

[0162] This invention selects moisture content, available nitrogen retention, microbial inactivation, water-soluble nutrient retention, salinity, and pH as key indicators because these indicators correspond to the storage and transportation stability of chicken manure pre-dried products, fertilizer activation ability, biosafety, extraction effectiveness, and compatibility with target water bodies, respectively.

[0163] Excessive moisture content can lead to spoilage during storage and increased transportation costs; insufficient available nitrogen can result in slow fertilization of fishponds or slow start-up of algae cultivation; excessive microbial load can lead to spoilage of the extract or introduce disease risks; insufficient retention of water-soluble nutrients can result in insufficient effective nutrients in the extract; and inappropriate salinity and pH can inhibit the target algae species or affect the water quality of the fishpond. Therefore, there is a direct technical link between the above indicators and whether dried chicken manure can be used as a precursor for high-quality biological nutrient solutions.

[0164] Preferably, the present invention employs a machine learning model for predicting nutrient form transformation; the model includes an input layer, a temporal feature extraction layer, a multi-source feature fusion layer, a nutrient transformation prediction layer, and a constraint output layer.

[0165] The input layer receives data on chicken manure materials, livestock farming data, drying operation data, environmental data, and downstream demand data. The temporal feature extraction layer uses a two-layer LSTM with a hidden layer dimension of 64 to extract features of temperature, humidity, airflow, material temperature, and moisture content over time, capturing the dynamic evolution during the drying process. The multi-source feature fusion layer combines or weights material features, equipment features, environmental features, and demand features to integrate cross-source information. The nutrient conversion prediction layer outputs the available nitrogen retention rate, ammonia nitrogen volatilization loss rate, water-soluble nutrient retention rate, and microbial inactivation rate. The constraint output layer outputs candidate drying parameter scores or recommended stage parameters, transforming the prediction results into actionable control suggestions.

[0166] The data flow between each layer is as follows: the input layer outputs standardized time series data to the time series feature extraction layer; the time series feature extraction layer outputs time series features to the multi-source feature fusion layer; the multi-source feature fusion layer outputs fused features to the nutrient conversion prediction layer; the nutrient conversion prediction layer outputs prediction indicators to the constraint output layer; and the constraint output layer then forms candidate parameter scores or stage control suggestions. This architecture has a clear hierarchy, well-defined functions, and is easy to deploy and maintain.

[0167] The model's input data is tabular time series data. Each record includes batch number, timestamp, moisture content, pH, uric acid content, ammonia nitrogen content, total nitrogen content, volatile fatty acid content, water-soluble phosphorus content, salinity, conductivity, hot air temperature, relative humidity, air volume, turning frequency, conveying speed, ambient temperature and humidity, and target application code.

[0168] The model's output data includes predicted final moisture content, predicted available nitrogen retention rate, predicted microbial inactivation rate, predicted water-soluble nutrient retention rate, predicted pH deviation, and predicted salinity deviation. The above input data reflects the initial state of chicken manure, the intensity of drying, and downstream targets, while the output data reflects the actual utilization performance of dried chicken manure and its extract. There is a clear technical correlation and causal logic between the two.

[0169] The model training data consists of the enterprise's own drying production records, laboratory small-scale drying data, pilot-scale equipment operation data, and downstream user feedback data; the training sample is preferably no less than 100 batches, and each batch includes at least material detection data before drying, equipment operation data during drying, detection data after drying, and downstream user feedback data; before training, the input data is filled with missing values, removed outliers, normalized, and coded for the target use.

[0170] Missing values ​​are filled using linear interpolation, forward filling, or the average of adjacent batches under the same operating conditions; outlier removal is performed using equipment range rules, interquartile range rules, or... Rules; normalization uses min-max normalization; target application encoding uses one-hot encoding or integer encoding.

[0171] The model training employs supervised learning, using the dried detection data and downstream feedback data as labels; the loss function is a weighted mean squared error loss.

[0172] ;

[0173] in, This indicates the error in moisture content prediction. This indicates the prediction error of available nitrogen retention rate. This indicates the error in predicting the microbial inactivation rate. This indicates the error in predicting the retention rate of water-soluble nutrients. This indicates the fitness prediction error; to The loss weights range from 0 to 1 and satisfy the following conditions:

[0174] ;

[0175] Model training can employ the Adam optimizer with a learning rate of 0.0001-0.01, a batch size of 16-256, and 50-500 training epochs. An early stopping mechanism on the validation set is used to prevent overfitting. The aforementioned hyperparameter range has been experimentally verified, demonstrating good convergence and generalization performance on most datasets. During training, model parameters are updated through loss calculation and backpropagation. After training, the model is deployed on edge computing devices, industrial computers, or servers for nutrient form transformation prediction and candidate strategy scoring in chicken manure drying scenarios, enabling real-time inference applications.

[0176] The input data and training datasets used in this embodiment are all from compliant sources. The data are taken from legally authorized public datasets, enterprise-owned production data, laboratory test data, or business data authorized by user anonymization. No personal sensitive information or biometric data is illegally collected. For content involving farm entities, operators, or customer information in the data, anonymization or desensitization processing has been carried out to comply with relevant personal information protection requirements and ensure the legality and compliance of the data processing process.

[0177] In another optional embodiment, the present invention does not employ an AI model, but instead uses empirical formulas and constrained optimization models. The empirical coefficients are obtained through experimental fitting. The experimental conditions include: no less than 30 batches of chicken manure samples, preferably no less than 100 batches; hot air temperature covering 35-85℃, relative humidity covering 20-80%, air volume covering 30-100% of the rated air volume of the equipment, and turning frequency covering once every 5-60 minutes; and testing the moisture content, uric acid, ammonia nitrogen, total nitrogen, water-soluble phosphorus, pH, conductivity, and microbial load of each batch before and after drying.

[0178] By fitting empirical coefficients using least squares, ridge regression, or partial least squares regression, and since the input indicators are all normalized first, the calculation formulas use dimensionless or unidimensional data, thus avoiding inconsistencies in units.

[0179] Through the above model and parameter settings, this embodiment enables those skilled in the art to clearly understand the model architecture, input and output, training data source, training steps, parameter range, loss function, and empirical coefficient acquisition method, thereby enabling them to implement the solution of this invention as described in the specification.

[0180] Example 4:

[0181] like Figure 4 As shown, this embodiment provides a chicken manure drying strategy control device based on multi-source data fusion, including a data acquisition module, a fusion analysis module, a constraint determination module, a strategy generation module, a control output module, and a feedback correction module. The modules work together to achieve a complete closed-loop control from data acquisition to strategy execution and effect feedback.

[0182] Specifically, the data acquisition module is used to acquire data on chicken manure materials, farming data, drying operation data, environmental data, and downstream resource utilization needs.

[0183] Furthermore, this module includes a material detection unit, an aquaculture data interface, an equipment data interface, an environmental sensing unit, and a downstream demand input unit. The material detection unit connects to a moisture meter, pH meter, conductivity meter, nitrogen detection device, and rapid microbial detection device to achieve rapid detection of key material indicators. The equipment data interface communicates with the drying equipment PLC or frequency converter to obtain real-time operating parameters. The downstream demand input unit is provided by a human-machine interface, mobile terminal, or aquaculture management platform, facilitating users to input their target uses and quality requirements. Through diverse data acquisition methods, this module can adapt to chicken manure treatment scenarios of different scales and levels of automation.

[0184] Specifically, the fusion analysis module is used to perform time-series alignment, missing value imputation, outlier handling, normalization, and feature extraction on the data obtained by the data acquisition module, and output the current drying state of chicken manure and the prediction results of nutrient form transformation.

[0185] Furthermore, the fusion analysis module can be deployed in edge computing devices, industrial computers, or servers, and its output results include water load index, nitrogen loss risk index, microbial residue risk index, water-soluble nutrient retention index, salinity-pH compatibility index, and downstream demand urgency index.

[0186] Specifically, the constraint determination module is used to determine target control constraints based on the current drying state and nutrient form transformation prediction results of chicken manure output by the fusion analysis module, combined with downstream resource utilization needs.

[0187] Furthermore, the target regulation constraints include final moisture content constraints, available nitrogen retention constraints, pathogenic microorganism inactivation constraints, water-soluble nutrient retention constraints, salinity adaptation constraints, and pH adaptation constraints.

[0188] Specifically, the strategy generation module is used to screen candidate drying parameters based on target control constraints and generate nutrient-oriented variable temperature and humidity drying strategies.

[0189] Furthermore, the strategy generation module includes a candidate parameter library, a constraint screening unit, a comprehensive scoring unit, and a staged strategy generation unit.

[0190] Furthermore, the stage strategy generation unit is used to generate the temperature, humidity, air volume, turning frequency, conveying speed, return air ratio, and stage switching time corresponding to the nitrogen protection stage, microbial control stage, and dissolution adaptation stage.

[0191] Specifically, the control output module is used to output the temperature and humidity changing drying strategy generated by the strategy generation module to the drying equipment control system. The control output module is connected to the drying equipment PLC through industrial Ethernet, RS485, 4G / 5G communication or fieldbus, and converts the control parameters of each stage into control instructions that the equipment can execute.

[0192] Specifically, the feedback correction module is used to correct subsequent chicken manure drying strategies based on post-drying test data and downstream user feedback data.

[0193] Furthermore, the feedback correction module receives feedback data such as moisture content, ammonia nitrogen content, uric acid content, total nitrogen content, water-soluble phosphorus, pH, conductivity, microbial load, as well as changes in pond water color, transparency, plankton density, algal cell density, extract bacterial rate, and ammonia nitrogen release curve.

[0194] Furthermore, the feedback correction module adjusts the weights in the fusion analysis module, the range of stage parameters in the strategy generation module, or the model prediction parameters based on the feedback deviation.

[0195] Based on the preferred embodiments of the present invention described above, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A method for regulating chicken manure drying strategy based on multi-source data fusion, characterized in that, Includes the following steps: S1. Acquire chicken manure material data, breeding end data, drying operation data, environmental data, and downstream resource utilization demand data, and identify each data batch to obtain a chicken manure drying control dataset. S2. Using the chicken manure drying control dataset, perform time-series alignment, feature extraction, and fusion calculation to obtain the current drying state and nutrient form transformation prediction results of chicken manure. S3. Match the current dry state of the chicken manure and the nutrient form transformation prediction results with the downstream resource utilization requirements to obtain target control constraints including moisture content, available nitrogen retention, microbial inactivation and water-soluble nutrient retention; S4. Using the target control constraint, candidate drying parameters are screened to obtain a nutrient-oriented variable temperature and humidity drying strategy; S5. Implement drying control according to the aforementioned temperature and humidity drying strategy, obtain post-drying detection data and downstream user feedback data, and adjust the subsequent chicken manure drying strategy accordingly.

2. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S1, the chicken manure material data includes one or more of the following: initial moisture content of chicken manure, material temperature, material pH, uric acid content, ammonia nitrogen content, total nitrogen content, volatile fatty acid content, water-soluble phosphorus content, salinity, electrical conductivity, and microbial load.

3. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S1, the downstream resource utilization demand data includes one or more of the following: target fishpond type, target algae species type, target nitrogen speciation spectrum, salinity tolerance range, suitable pH range, required beneficial bacteria type, application time window, and target fertilizer effect activation speed.

4. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S2, the fusion analysis includes: performing time-series alignment, missing value imputation, normalization, and feature weighting on the acquired data to obtain a fusion state vector that characterizes water load, equipment processing capacity, nitrogen loss risk, microbial residue risk, and downstream demand urgency.

5. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S2, the nutrient form transformation prediction results include one or more of the following: uric acid to ammonia nitrogen conversion prediction results, ammonia nitrogen volatilization loss prediction results, organic nitrogen retention prediction results, volatile fatty acid retention prediction results, water-soluble phosphorus retention prediction results, and salt form change prediction results.

6. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S3, the target regulation constraints include one or more of the following: final moisture content constraint, available nitrogen retention constraint, pathogenic microorganism inactivation constraint, water-soluble nutrient retention constraint, salinity adaptation constraint, and pH adaptation constraint.

7. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S4, the variable temperature and humidity drying strategy includes: a nitrogen protection stage, a microbial control stage, and a dissolution adaptation stage; wherein, the nitrogen protection stage is used to reduce the loss of available nitrogen, the microbial control stage is used to reduce the pathogenic microbial load, and the dissolution adaptation stage is used to adjust the suitability of the extract of dried chicken manure.

8. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S4, the variable temperature and humidity drying strategy includes one or more control parameters among hot air temperature, relative humidity, air volume, turning frequency, conveying speed, return air ratio and stage switching time.

9. The method for regulating chicken manure drying strategy based on multi-source data fusion according to claim 1, characterized in that, In step S5, the downstream feedback data includes one or more of the following: fishpond water color change data, transparency data, plankton density data, algal cell density data, extract bacterial rate data, extract pH data, extract conductivity data, and ammonia nitrogen release curve data.

10. A device for regulating chicken manure drying strategy based on multi-source data fusion, characterized in that, include: The data acquisition module is used to acquire data on chicken manure materials, farming operations, drying operations, environmental conditions, and downstream resource utilization needs. The fusion analysis module is used to determine the current drying state of chicken manure and the predicted results of nutrient form transformation. The constraint determination module is used to determine the target control constraints; The strategy generation module is used to generate nutrient-oriented temperature and humidity-controlled drying strategies. The control output module is used to output the control parameters corresponding to the variable temperature and humidity drying strategy; The feedback correction module is used to correct subsequent chicken manure drying strategies based on post-drying test data and downstream user feedback data.

Images