Bio-organic fertilizer production process data system and extraction method

Abstract

The present application relates to the technical field of organic fertilizer production, and more particularly to a biological organic fertilizer production process data system and extraction method. The method comprises the following steps: obtaining a raw material information set and a process parameter set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on the raw material information set; calculating a reference microbial activity characteristic based on

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of organic fertilizer production, and in particular to a biological organic fertilizer production process data system and extraction method. BACKGROUND

[0002] In the existing biological organic fertilizer production process, the data extraction and monitoring means generally have problems such as single data dimension, rough process feature extraction, delayed fermentation state recognition, etc. Especially in the face of different raw material components and complex process environment, the traditional method is difficult to realize the accurate capture of the microbial activity state and the dynamic tracing of the abnormal production line. Most systems only rely on basic environmental parameters such as temperature and humidity, ignoring the fine-grained monitoring of microbial metabolism, heat production, gas production, and acid-base neutralization reaction data, resulting in that the fermentation behavior cannot be effectively modeled, and the process control lacks data support; at the same time, the existing method lacks correlation mining between the quality results and the process behavior in the fermentation process, lacks an efficient data structure for tracing and deviation identification, and is difficult to realize the fusion analysis and quality feedback of multi-source heterogeneous data. In addition, the detection of the output finished product is mostly limited to static index collection, lacking correlation mapping between historical trajectory and dynamic reaction rate, and it is difficult to build a systematic diagnosis mechanism for process quality, which limits the intelligent and fine development of the entire biological organic fertilizer production process. SUMMARY

[0003] Therefore, it is necessary to provide a biological organic fertilizer production process data system and extraction method to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, the biological organic fertilizer production process data extraction method comprises the following steps:

[0005] Step S1: obtaining a raw material information set and a process parameter set; based on the microbial activity characteristics of the raw material information set, calculating the reference microbial population rate;

[0006] Step S2: labeling the raw material information set and the process parameter set with time sequence timestamps, and constructing a fermentation behavior trajectory; analyzing the microbial activity characteristics of the fermentation behavior trajectory, and calculating the microbial population reaction speed;

[0007] Step S3: after the completion of the organic fertilizer production line, collecting the output quality feature set of the organic fertilizer finished product; comparing the corrosion degree index in the output quality feature set with the reference microbial population rate and the microbial population reaction speed, and outputting the fermentation process deviation data;

[0008] Step S4: tracing and correlating the fermentation process deviation data and the fermentation behavior trajectory to obtain an organic fertilizer process deviation mapping table; identifying the abnormal items of the fermentation production line according to the organic fertilizer process deviation mapping table, and constructing an organic fertilizer production report.

[0009] In the present specification, a bio-organic fertilizer production process data extraction system is provided for performing the bio-organic fertilizer production process data extraction method described above, the bio-organic fertilizer production process data extraction system comprising:

[0010] A raw material analysis module is configured to obtain a raw material information set and a process parameter set, and calculate a reference microbial population rate based on a microbial activity feature of the raw material information set;

[0011] A fermentation analysis module is configured to label a time sequence timestamp for the raw material information set and the process parameter set, and construct a fermentation behavior trajectory;

[0012] An analysis module is configured to analyze a microbial activity feature of the fermentation behavior trajectory, and calculate a microbial population reaction speed;

[0013] An analysis module is configured to analyze a microbial activity feature of the fermentation behavior trajectory, and calculate a microbial population reaction speed;

[0014] A quality deviation module is configured to collect an output quality feature set of the bio-organic fertilizer product after the bio-organic fertilizer production line is completed, and compare a corrosion degree index in the output quality feature set with the reference microbial population rate and the microbial population reaction speed, and output fermentation process deviation data;

[0015] A traceability diagnosis module is configured to trace and correlate the fermentation process deviation data and the fermentation behavior trajectory to obtain a bio-organic fertilizer process deviation mapping table, identify a control abnormal item of the fermentation production line according to the bio-organic fertilizer process deviation mapping table, and construct a bio-organic fertilizer production report.

[0016] The beneficial effects of the present application are: by constructing the raw material information set and the process parameter set based on time sequence labeling, a complete fermentation behavior trajectory is formed, so that the data flow in the fermentation process has clear time logic and structural hierarchy, thereby providing a high-precision data basis for subsequent microbial activity feature extraction and reaction speed calculation. By extracting multi-dimensional dynamic data such as temperature and humidity changes, pH fluctuations, gas release rate, heat generation rate and acid-base neutralization rate, a dynamic coupling model of microbial activity and metabolic response is established in the fermentation behavior trajectory, effectively revealing the stage change characteristics of microbial reaction. After the production line is completed, the corrosion degree index of the finished product is further collected and compared with the benchmark microbial group rate and microbial reaction speed, and the data correlation between the finished product quality and the process behavior is established, thereby realizing the quantitative analysis of process deviation. On this basis, the system traces the process deviation and behavior trajectory, constructs a deviation mapping table and identifies abnormal items of the production line, forms an end-to-end data closed-loop feedback path, and effectively realizes the whole process quality traceability and control optimization from raw material input to finished product output. The whole process takes data-driven as the core, emphasizes the time sequence fusion of multi-source data and the dynamic adjustment of index mapping relationship, improves the data explainability and analysis depth, and avoids the problems of coarse-grained processing of microbial state evaluation and lagging of process deviation identification in traditional methods. Therefore, by constructing the multi-source heterogeneous data-driven fermentation behavior trajectory and quality mapping system, the present application solves the problems of data isolation, state recognition lag and process deviation inaccurate positioning in traditional biological organic fertilizer production, and improves the intelligent diagnosis ability and quality controllability of the production process. BRIEF DESCRIPTION OF DRAWINGS

[0017] Fig. 1 It is a step flowchart of a biological organic fertilizer production process data extraction method;

[0018] Fig. 2 It is a biological organic fertilizer production line flowchart;

[0019] Fig. 3 It is a biological organic fertilizer production process gas concentration change monitoring schematic diagram;

[0020] The implementation, functional characteristics and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0021] The technical method of the present application will be described clearly and completely below in combination with the accompanying drawings. Obviously, the described embodiments are 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 skilled in the art without creative labor fall within the scope of protection of the present application.

[0022] In addition, the accompanying drawings are included to provide a thorough understanding of embodiments of the application and are not intended to be exhaustive or to limit the application to the precise outline described herein. Identical reference numerals in different drawings represent the same or similar elements, and repetitive descriptions can be omitted. Some of the blocks in the drawings are functional entities that do not necessarily have to correspond to physically or logically independent entities. The functional entities can be implemented in software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0023] It should be understood that, although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0024] To achieve the above object, please refer to Figs. 1 to 3 A bio-organic fertilizer production process data extraction method, the method comprises the following steps:

[0025] Step S1: Obtain a raw material information set and a process parameter set; based on the microbial activity characteristics of the raw material information set, calculate the reference bacterial population rate;

[0026] Step S2: Time-stamp the raw material information set and the process parameter set, and construct a fermentation behavior trajectory; analyze the microbial activity characteristics of the fermentation behavior trajectory, and calculate the bacterial population reaction speed;

[0027] Step S3: After the organic fertilizer production line is completed, collect the output quality characteristic set of the organic fertilizer product; compare the corrosion degree index in the output quality characteristic set with the reference bacterial population rate and the bacterial population reaction speed, and output the fermentation process deviation data;

[0028] Step S4: Trace and associate the fermentation process deviation data and the fermentation behavior trajectory to obtain an organic fertilizer process deviation mapping table; identify the abnormal items of the fermentation production line according to the organic fertilizer process deviation mapping table, and construct an organic fertilizer production report.

[0029] In the present embodiment, the sensors required to obtain the raw material information set and the process parameter set include but are not limited to sensors, NDIR sensors, temperature probes, dissolved / gas phase sensors, pH electrodes, weighing / gas flow meters, and camera / spectral sampling ends.

[0030] In the present embodiment, the referenceFig. 1 As shown, it is a step flow diagram of the bio-organic fertilizer production process data extraction method of the present application. In the present example, the bio-organic fertilizer production process data extraction method comprises the following steps:

[0031] Step S1: Obtain the raw material information set and the process parameter set; based on the microbial activity characteristics of the raw material information set, calculate the reference microbial population rate;

[0032] Especially important is that the microbial activity characteristic calculation formula of step S1 is as follows:

[0033] ;

[0034] Wherein, to is a weight factor, is the reference microbial population rate or the microbial population reaction speed.

[0035] In the present example, : indicates the reference microbial population rate or the microbial population reaction speed at time t, which is a dynamic index that comprehensively reflects the metabolic activity degree of the microbial population.

[0036] is a weight factor, reflecting the contribution of each monitoring variable to the calculation of the microbial population rate or reaction speed. These weights can be obtained by historical production data regression fitting or experimental calibration, and are used to balance the influence of different variables.

[0037] : the rate of change of ammonia gas concentration with time, usually in or reflects the decomposition rate of protein and nitrogen-containing organic matter.

[0038] : the rate of change of carbon dioxide concentration with time, in or , is the product rate of microbial respiratory metabolism in the organic matter decomposition process.

[0039] : the rate of change of pile temperature with time, in , embodies the dynamic change of heat release and dissipation in the fermentation process.

[0040] : the rate of change of oxygen concentration with time, in or , reflects the change trend of microbial oxygen consumption intensity and aeration condition.

[0041] It is particularly important to note that the weighting factors need to be obtained by designing control experiments based on the raw material type. For example, for high-nitrogen raw materials (total nitrogen content > 2.5%), the weighting factors should be increased. The weighting is increased to 0.4-0.5; for high-carbon raw materials (C / N>30), the weighting is increased. To 0.3~0.4;

[0042] Each category requires at least three batches of production validation, with a finished product humic acid compliance rate >90% as the weighted optimization benchmark, thereby determining... Weighting factors.

[0043] Preferably, step S1 further includes:

[0044] Send data acquisition request signals to the raw material input unit and the process control unit;

[0045] Receive the data acquisition request signal for the current batch of raw material information set and add the raw material sampling timestamp;

[0046] Obtain the fermentation start-up and shutdown times, aeration or temperature control operation execution sequences from the process parameter set, and establish a process event timeline.

[0047] In embodiments of the present invention, four types of dynamic change quantities are extracted from raw material information: ammonia ( ),carbon dioxide( ),temperature( ) and dissolved oxygen ( These data require continuous time-series changes, therefore the system needs to perform periodic sampling via sensors or sampling ports, with each data point including concentrations or values ​​at least multiple time points. Based on this, by performing time differentiation on each data point (i.e., calculating the rate of change between adjacent time points), the ammonia change rate, rate of change, rate of temperature rise and fall and The rate of change, multiplied by the corresponding weighting factor ( to Finally, a numerical value representing the current activity state of the bacterial community is obtained through weighted summation. This value is defined as the "baseline microbial community rate" or "microbial community reaction rate," and can be used to subsequently determine the stability and abnormal characteristics of the fermentation process. Before sampling, a data acquisition request is automatically sent to the raw material input unit and the process control unit, and the corresponding "raw material sampling timestamp" and "process event timeline" are recorded. The latter includes key nodes such as fermentation start time, aeration and temperature control operation sequence, which are used to align and match the microbial community rate data with the fermentation process behavior trajectory later.

[0048] In one implementation of the present invention, assuming a batch of high-temperature composting organic fertilizer is being processed, the raw material data are as follows:

[0049] The ammonia concentration is reduced from 25 ppm to 20 ppm in 1 hour (a rate of change of -5 ppm / h);

[0050] The CO2 concentration is increased from 3000 ppm to 3500 ppm (a rate of change of +500 ppm / h);

[0051] The temperature is increased from 52°C to 58°C (a rate of change of +6°C / h);

[0052] The dissolved oxygen is reduced from 6.0 mg / L to 5.4 mg / L (a rate of change of -0.6 mg / L-h).

[0053] The set weight factors are: n1=0.3, n2=0.2, n3=0.4, n4=0.1, which are substituted into the microbial activity characteristic calculation formula:

[0054]

[0055]

[0056] At this time, the recorded microbial community reaction speed of the system is 100.84, which can be used as the numerical identifier of the fermentation activity in this time period. By continuous sampling calculation, a complete microbial community reaction speed curve can be formed for subsequent fermentation behavior trajectory modeling and deviation analysis.

[0057] Step S2: labeling time sequence timestamps for the raw material information set and the process parameter set, and constructing a fermentation behavior trajectory; analyzing the microbial activity characteristics of the fermentation behavior trajectory, and calculating the microbial community reaction speed;

[0058] Preferably, step S2 includes the following steps:

[0059] Step S21: labeling raw material sampling timestamps for the raw material information set and the process parameter set, and segmenting sampling according to the unified time window of the process event time axis to generate a time window index group;

[0060] Step S22: performing linear interpolation on temperature and humidity according to the time window index group to obtain a fertilizer time sequence data stream;

[0061] Step S23: segmenting the fertilizer time sequence data stream according to the nodes of the time sequence timestamps and the fermentation stage labels to construct a fermentation behavior trajectory;

[0062] Step S24: extracting temperature and humidity change gradients and pH fluctuation frequencies as microbial community metabolic excitation indicators, and comparing the indicator change rates of the before and after time windows to generate microbial community excitation indicator data;

[0063] Step S25: Calculate the microbial activity conversion rate of the microbial stimulation index data, and combine the gas release rate, heat release rate and acid-base neutralization rate per unit time in the fermentation behavior track to generate the microbial reaction speed.

[0064] Preferably, step S25 further comprises the following:

[0065] The fermentation behavior track is divided into three stages of initial stage, medium temperature stage and gas production stage according to the time stamp, and the proportion distribution of acid-base neutralization rate, heat release rate and gas release rate is adjusted respectively, wherein:

[0066] The initial stage takes the acid-base neutralization rate as the index;

[0067] The medium temperature stage takes the heat release rate as the index;

[0068] The gas production stage takes the gas release rate as the index, wherein the index is 60% of the proportion distribution.

[0069] In the embodiment of the present application, step S2 constitutes a reproducible data pipeline from data alignment, missing data filling, time sequence segmentation to feature extraction: all raw material information and process parameters are marked with a unified time stamp by a synchronous time source (such as NTP or gateway clock) and stored in the database, sampling uses fixed or adaptive sampling rate (for example, 5-15 minutes / time), and the sampling frequency and device ID are registered as metadata at the collection end for traceability; based on the process event time axis (including tank entering, starting and stopping, aeration, temperature rising point, etc.), time window index groups are generated by segmenting the time sequence data with a configured time window width (for example, 4-8 hours, or set as needed); for the breakpoints or sparse sampling points in the time window, linear interpolation is used to fill in the temperature and humidity channels, forming continuous fertilizer time sequence data flow and writing into the time window index; the time sequence flow is segmented and fermentation behavior track records are generated according to the time window and process event label (the track table contains time window index, stage label and summary statistics of each channel);

[0070] In each time window, the basic features are extracted by numerical methods: the temperature and humidity gradient is calculated by finite difference (central difference or forward difference) to calculate its average change rate and variance, the pH fluctuation is counted by threshold zero crossing or short-time spectrum (peak count in short-time window) to obtain the fluctuation frequency, and then these are quantified as standardized microbial metabolic stimulation indexes (first normalized or z-score standardized, based on historical baseline or batch initial value); the time window change rate of the stimulation index is calculated by dividing the difference value of adjacent windows by the time interval and recorded as "stimulation change rate", these indexes are written into the stimulation index table together with the time window index for subsequent conversion rate and reaction speed calculation.

[0071] Especially to be noted, the establishment of historical baseline is based on the fermentation data of nearly 30 batches of the same type of raw materials, the mean value of each index at the same fermentation stage is calculated, if the historical data is lacking, the mean value of the index within 2 hours after the start of this batch fermentation is taken as the experience value.

[0072] In the embodiment of the application, the whole fermentation track is divided into the initial stage, the medium temperature stage and the gas production stage by event-driven feature detection, and the division logic is as follows: the earliest inflection point of the track, which detects "temperature slope continuously rising and exceeding the threshold value", is taken as the starting point of the medium temperature stage, and the inflection point before the starting point of the medium temperature stage is taken as the starting point of the initial stage. The sequence of local significant peak value (peak value greater than window average value plus several times standard deviation) is taken as the starting point of the gas production stage, and the initial stage is before the starting point; in each stage, three types of dominant rate sequences are extracted, the acid-base neutralization rate uses the time derivative of pH ) and can be multiplied by the buffer coefficient or corrected by the conductivity / alkalinity measurement, the heat generation rate uses the time derivative of the heap temperature (the moving average smoothing is performed on the sensing noise first), and the gas release rate uses the time derivative of the concentration or exhaust flow and the interval cumulative release amount as the measurement;

[0073] In the stage, several operable indexes are calculated: average change rate, peak density (peak occurrence times per unit hour), continuous threshold exceeding time length (continuous exceeding of the set threshold value) and standard deviation; the weight distribution is set according to the proportion of the dominant index (in the example, the proportion of the dominant index is 60%), the stage activity conversion parameter is calculated = dominant index value x 0.6 + the remaining 40% (or the experience weight distribution) after the remaining two indexes are normalized; finally, the stage activity conversion parameters are weighted and summarized according to the time length or the preset weight to obtain the colony reaction speed sequence of the whole track and are stored in the database.

[0074] The specific steps required in operation include: the equipment end uploads the sensing value every 5-15 minutes, the data receiving end first performs time alignment and denoising (moving average window 3-5 times), linear interpolation for missing points, finite difference derivation to obtain the rate sequence, peak value detection adopts sliding window local maximum judgment and combines threshold filtering, and the calculation result is stored in time window units and generates an alarm / record.

[0075] In one implementation mode of the application, taking a batch of high-nitrogen raw materials as an example, the system sets the raw material sampling time at 9:00 on August 1, and the fermentation start time at 10:00 on August 1. It is set to divide a time window every 6 hours. In the first two time windows, the pH value slowly decreases from 6.5 to 6.1, and the temperature rises from 31°C to 38°C. The system calculates that the temperature and humidity change gradient is small, and the acid-base neutralization rate is dominant, so it is determined as the initial stage.

[0076] ​In the next three time windows, the temperature rapidly rises to 55°C, and the system identifies this as the medium temperature stage, with the heat generation rate as the dominant factor. In the subsequent time window, the CO2 concentration suddenly rises to 9000 ppm, and the gas release rate is significantly high, thus establishing the gas production stage. During this process, the system automatically adjusts the proportion of the three types of indicators to match the fermentation activity logic at different stages, thereby providing stable reference indicators for subsequent quality comparison and process deviation analysis.

[0077] Step S3: After the organic fertilizer production line is completed, the output quality characteristic set of the organic fertilizer product is collected; the corrosion degree index in the output quality characteristic set is compared with the benchmark bacterial population rate and the bacterial population reaction speed, and fermentation process deviation data is output;

[0078] Preferably, step S3 includes the following:

[0079] After the organic fertilizer production line is completed, the humus scale index is collected by the fertilizer product detection unit, and a detection timestamp is marked, forming an output quality characteristic set;

[0080] Preferably, the humus scale index is composed of the absorption ratio of humic acid and fulvic acid, and is paired with the raw material information set corresponding to the benchmark bacterial population rate to establish a raw material-humus correspondence relationship.

[0081] The raw material-humus correspondence relationship is compared with the benchmark bacterial population rate and the bacterial population reaction speed, and fermentation process deviation data is output, wherein the comparison standard is to calculate the fermentation state difference degree in each time window.

[0082] In the embodiment of the present application, at the end of the production line, the humus scale index of the fertilizer is collected by the detection unit, and the humus scale is composed of the spectral absorption ratio of humic acid and fulvic acid, which can reflect the degree of organic matter conversion in the product. After the detection is completed, the system marks the timestamp for this index and includes it in the output quality characteristic set. Then, the system binds the humus scale index with the corresponding raw material batch according to the previous raw material sampling information, forms a group of "raw material-humus" pairing relationship, and compares it with the benchmark bacterial population rate and the bacterial population reaction speed calculated previously. The comparison method is to extract the bacterial population reaction data in each time window and the corresponding window data in the product index after dividing the fermentation behavior trajectory according to the unified time window, and to calculate the deviation degree between the windows through the set fermentation state difference degree function, so as to form the fermentation process deviation data set, which provides the basis for subsequent anomaly identification and process adjustment. The key of the whole method is that the data label alignment, cross-stage correspondence and difference quantification are stable and traceable.

[0083] Preferably, the comparison of the corrosion degree index in the output quality characteristic set with the benchmark bacterial population rate and the bacterial population reaction speed includes the following:

[0084] extracting a corrosion degree index in the output quality feature set;

[0085] constructing a two-dimensional structure array with the corrosion degree index, the reference microbial population rate and the microbial population reaction speed, and nesting key node information to obtain a fermentation phase state array;

[0086] grading the fermentation phase state array with a preset deviation threshold to obtain fermentation process deviation data.

[0087] Preferably, the nested key node information includes the following:

[0088] The nested key node information includes a first node, a second node and a third node;

[0089] The first node represents a microbial population active critical point, representing a local maximum value of the reference microbial population rate or the microbial population reaction speed at the microbial population active critical point;

[0090] The second node represents a temperature high plane stable point, i.e. a stable point of the temperature peak;

[0091] The third node represents an abnormal point of oxygen fluctuation in the heap, satisfying a continuous downward trend of oxygen content and accompanied by a sudden rise in ammonia concentration.

[0092] Preferably, constructing the two-dimensional structure array further includes:

[0093] constructing the two-dimensional structure array, wherein the horizontal direction is a key time window in the fermentation behavior track, and the vertical direction is fermentation process deviation data constituted by the corrosion degree index;

[0094] The key time window is based on the raw material tanking time to trace back the gas release abnormal window in the fermentation stage, and an equal-width time window sequence is developed around the window as the center;

[0095] The corrosion degree index forms a corresponding sequence in each time window, constituting a unit comparison unit;

[0096] The key time window is used as an index to embed the fermentation state difference degree in the unit comparison unit to form the two-dimensional structure array.

[0097] In the embodiment of the present application, by constructing a structured data matrix, the quality index of the final output is correspondingly analyzed with the microbial population reaction in different time windows in the fermentation process to identify process deviations. Specifically, the system first extracts the corrosion degree index in the organic fertilizer product, mainly represented by the absorption ratio of humic acid and fulvic acid, and matches it with the reference microbial population rate and the microbial population reaction speed in the fermentation process. After aligning the time label, a two-dimensional structure array is constructed, wherein the horizontal axis is the key time window defined according to the gas release abnormal window as the center, and the vertical axis is the corrosion degree index sequence collected in different time windows.

[0098] Each cell of this array represents the mapping relationship between the raw material reaction and the product quality within a specific time window. To enhance the interpretability of the data, the array also embeds three types of key node information: the first type of node is the critical point of microbial activity, i.e., the time when the microbial reaction rate appears a local extreme value; the second type is the stable point of high temperature plateau, corresponding to the time node when the temperature change enters the plateau period; the third type is the abnormal point of oxygen fluctuation, i.e., the abnormal stage of continuous oxygen decline and sudden rise of ammonia concentration. These nodes are embedded as markers in the array, which helps to highlight the characteristics of biological activity or environmental response in different stages. The constructed array is then compared with the set deviation threshold to identify possible abnormal areas in different fermentation stages. This process not only completes the process state mapping in time and space dimensions, but also provides an operable way to quantify the volatility and stability of the process.

[0099] In an implementation manner of the present application, assuming that the gas release rate suddenly increases within 4 hours after the raw material is put into the tank, the system sets a sequence of equal-width time windows from the 2nd hour to the 6th hour centered on this point.

[0100] In these time windows, the collected humus scale indicators are 2.1, 2.5, 3.0, 2.7 and 2.4, respectively; at the same time, the microbial reaction rates are 0.8, 1.2, 1.5, 1.3 and 1.0, respectively; combined with the high-temperature plateau (e.g., the temperature remains above 65°C at the 3rd hour), the ammonia concentration sudden rise point (the 5th hour) and the microbial rate peak value (the 4th hour), these three key nodes are embedded in the array of this time period. After comparing these values with the set deviation threshold, it is found that the fermentation state difference of the 5th hour time window is higher than the standard value, thus recording it as an abnormal window of fermentation, outputting the process deviation segment for subsequent judgment of whether it is an operation error or insufficient raw material adaptability, etc.

[0101] This example illustrates that the core logic of the entire technical solution is to sequentially perform the three actions of alignment, nesting and quantization, and to complete process tracing and deviation detection with the help of structured data.

[0102] As shown in FIG. 1, it is a schematic diagram of a bio-organic fertilizer production line; Fig. 2

[0103] 101 is a raw material input unit - responsible for obtaining a set of raw material information, including microbial activity characteristic data collection;

[0104] 102 is a fermentation reactor - to construct a fermentation behavior trajectory and monitor the microbial metabolic process;

[0105] 103 is a process control unit - to perform ventilation and temperature control operations and establish a process event timeline;

[0106] ​104 is a data acquisition sensor - monitor temperature and humidity, pH, gas release and other key parameters;

[0107] 105 is a fertilizer product detection unit - collect humus scale indicators to form the output quality feature set.

[0108] Step S4: Trace and correlate the fermentation process deviation data and the fermentation behavior trajectory to obtain the organic fertilizer process deviation mapping table; identify the abnormal control items of the fermentation production line according to the organic fertilizer process deviation mapping table, and construct the organic fertilizer production report.

[0109] Preferably, step S4 comprises the following steps:

[0110] Step S41: Trace and correlate the fermentation process deviation data and the fermentation behavior trajectory to obtain the production line node abnormal data; construct the organic fertilizer process deviation mapping table based on the production line node abnormal data;

[0111] Step S42: Identify the abnormal control items of the fermentation production line according to the organic fertilizer process deviation mapping table to generate the full-process node abnormal items;

[0112] Step S43: If the full-process node abnormal items are within the preset range, it is determined that the organic fertilizer production line is in a normal production state, and the full-process monitoring is maintained until the end of the production cycle; if not, it is determined that the organic fertilizer production line is in an abnormal production state, and an audible and visual alarm is generated and an organic fertilizer production report with abnormal nodes is generated.

[0113] In the embodiment of the application, by comparing the output quality features with the fermentation stage data to obtain the fermentation process deviation data, and combining the time sequence records in the fermentation behavior trajectory, the abnormal time window and abnormal points on the fermentation time axis are located and corresponded, which are defined as the production line node abnormal data. Based on these abnormal data, the system constructs an "organic fertilizer process deviation mapping table" in the data table structure, which records the time of abnormal occurrence, the corresponding process parameters (such as temperature, air volume, pH, etc.) and the offset amount of the corresponding indicators such as the reaction speed of the microbial population or the gas release rate in the fermentation stage. Then, the system traverses the deviation mapping table to extract all identifiable abnormal control items, such as delayed heating response, excessive pH fluctuation, abnormal ventilation jitter, etc., and aggregates them into "full-process node abnormal items", which reflect the point set that is judged to deviate from the normal control trajectory in the entire production cycle. Finally, the system compares the abnormal item set with the preset threshold model (such as the maximum allowed deviation times, the maximum allowed offset amplitude, the maximum continuous abnormal duration, etc.), if the result is within the tolerance range, it enters the normal monitoring state; if it exceeds the range, it automatically triggers an audible and visual alarm, records the abnormal point and generates a production report with an abnormal marker. This process is based on time sequence data, process state data and threshold logic, and combines structured mapping relationship and multi-dimensional deviation indicators to construct an automatic closed-loop production line diagnosis method.

[0114] In an implementation of the present application, on a 48-hour organic fertilizer fermentation production line, the system detects that at the 12th hour, the 18th hour and the 36th hour, the reaction speed of the microbial flora decreases by 25%, 30% and 22% respectively, and the corresponding pH fluctuation is more than ±1.8, and the temperature at the 36th hour appears a high plateau for more than 2 hours, which are located as “production line node abnormal data”. After the system matches these abnormal points with their corresponding original behavior track, temperature and humidity, ventilation volume and other parameters, three high-risk nodes are marked in the deviation mapping table. After full-process scanning, the system judges that there are 3 high-risk abnormal nodes in this round of fermentation, of which 1 belongs to continuous deviation timeout (36th hour), which exceeds the preset maximum deviation duration limit of 2 hours, and finally the system determines that the batch production line state is “abnormal”, triggering an audible and light alarm, and marking the abnormal node information and corresponding time window, parameter fluctuation value in the generated organic fertilizer production report. In this way, the whole process of data acquisition, abnormal identification, deviation modeling and alarm output is completed.

[0115] In the overall implementation of the present application, a batch is set to have a sampling interval of 10 minutes and a window width of 6 hours (36 sample points). In the 18th hour to the 24th hour window, from 2500ppm to 5200ppm (regression slope in window ≈ +74 ppm / h), the temperature rises from 45 to 61 (slope ≈ +2.5 / h), and the pH fluctuates across the threshold once. The conversion of the window is calculated to be 0.12 (unit / window), the gas_rate is 74, the heat_rate is 2.5, and the neutralize_rate is -0.05;

[0116] The stage is determined to be the gas production stage (gas dominant, ),

[0117] The reaction_speed=0.10.12+0.674+0.22.5+0.1(-0.05)≈44.5 (example value is only used to illustrate dimensional synthesis logic).

[0118] If the final humus scale value corresponding to the window is 20% lower than expected (difference threshold), the window is marked as “deviation”, the node information (such as the 20th hour as the peak of microbial activity, and the 22nd hour as the temperature plateau) is embedded, and is recorded in the deviation mapping table. If two high-level deviation windows are found, an alarm is automatically triggered and the related time window, original sensor curve snapshot and suggested attention points (such as checking the ventilation system or raw material ratio) are listed in the report.

[0119] As Fig. 3 shown, an implementation of the gas concentration variation in the fermentation process of the present application is demonstrated, wherein

[0120] The concentration is represented by red triangles;

[0121] The concentration is represented by green circles;

[0122] The concentration is represented by blue diamonds.

[0123] Thus, from the standpoint of the prior art, the embodiments described are merely illustrative of the many possible specific embodiments which represent applications of the present application. Numerous and varied other embodiments are contemplated and can be readily devised by those skilled in the art without departing from the spirit and scope of the application. It is therefore intended that the scope of the application be limited only by the scope of the appended claims, and not by the foregoing description.

[0124] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and changes will readily occur to those skilled in the art, and it is intended to embrace all such modifications and changes that fall within the scope of the appended claims. Accordingly, the application is not to be restricted except in the spirit of the claims that follow.

Claims

1. A bio-organic fertilizer production process data extraction method, characterized by, Comprising the following steps: Step S1: Obtain the raw material information set and the process parameter set; Based on the microbial activity characteristics of the raw material information set, calculate the reference bacterial flora rate; Step S2: Time-stamp the raw material information set and the process parameter set, and construct the fermentation behavior trajectory; Analyze the microbial activity characteristics of the fermentation behavior trajectory, and calculate the bacterial flora reaction speed; Step S3: After the organic fertilizer production line is completed, collect the output quality characteristic set of the finished organic fertilizer; compare the corrosion degree index in the output quality characteristic set with the reference bacterial flora rate and the bacterial flora reaction speed, and output the fermentation process deviation data; Step S4: Trace the fermentation process deviation data and the fermentation behavior trajectory to obtain the organic fertilizer process deviation mapping table; identify the abnormal items of the fermentation production line according to the organic fertilizer process deviation mapping table, and construct the organic fertilizer production report; Step S4 is specifically: Step S41: Trace the fermentation process deviation data and the fermentation behavior trajectory to obtain the production line node abnormal data; construct the organic fertilizer process deviation mapping table based on the production line node abnormal data; Step S42: Identify the abnormal items of the fermentation production line according to the organic fertilizer process deviation mapping table, and generate the full-process node abnormal items; Step S43: If the full-process node abnormal items are within the preset range, it is determined that the organic fertilizer production line is in a normal production state, and the full-process monitoring is maintained until the end of the production cycle; If it is not within the preset range, it is determined that the organic fertilizer production line is in an abnormal production state, and a sound and light alarm is adopted to generate an organic fertilizer production report with abnormal nodes.

2. The bio-organic fertilizer production process data extraction method of claim 1, wherein, Step S1 further comprises: Send a data collection request signal to the raw material input unit and the process control unit; Receive the data collection request signal of the raw material information set of the current batch, and add the raw material sampling time stamp; Obtain the fermentation start-stop time, execution sequence of aeration or temperature control operation in the process parameter set, and establish a process event time axis.

3. The bio-organic fertilizer production process data extraction method of claim 1, wherein, Step S2 comprises the following steps: Step S21: Time-stamp the raw material information set and the process parameter set, and segment sampling according to the unified time window of the process event time axis to generate a time window index group; Step S22: According to the time window index group, linear interpolation is performed on the temperature and humidity to obtain a fertilizer time series data stream; Step S23: The fertilizer time series data stream is segmented according to the time series time stamp node and the fermentation stage label to construct the fermentation behavior trajectory; Step S24: Extract the temperature and humidity change gradient and the pH fluctuation frequency as the bacterial flora metabolic excitation index, and compare the index change rate of the previous and subsequent time windows to generate the bacterial flora excitation index data; Step S25: Calculate the bacterial flora activity conversion rate of the bacterial flora excitation index data, and combine the gas release rate, heat generation rate and acid-base neutralization rate per unit time in the fermentation behavior trajectory to generate the bacterial flora reaction speed.

4. The bio-organic fertilizer production process data extraction method of claim 1, wherein, Step S25 further comprises the following: Divide the fermentation behavior trajectory into three stages of initial stage, medium temperature stage and gas production stage according to the time stamp, and adjust the proportional distribution of the acid-base neutralization rate, heat generation rate and gas release rate respectively, wherein: The initial stage takes the acid-base neutralization rate as the index; The medium temperature stage takes the heat generation rate as the index; The gas production stage takes the gas release rate as the index, wherein the index is 60% of the proportional distribution.

5. The bio-organic fertilizer production process data extraction method of claim 1, wherein, Step S3 includes The following: When the organic fertilizer production line is completed, the humus scale index is collected through the fertilizer product detection unit, and a detection timestamp is marked to form an output quality feature set; The humus scale index is composed of the absorption ratio of humic acid and fulvic acid, and is paired with the raw material information set corresponding to the reference microbial population rate to establish a raw material-humus correspondence relationship; The raw material-humus correspondence relationship is compared with the reference microbial population rate and the microbial population reaction speed to output fermentation process deviation data, and the comparison standard is to calculate the fermentation state difference degree in each time window.

6. The bio-organic fertilizer production process data extraction method of claim 5, wherein, The comparison between the corrosion degree index in the output quality feature set and the reference microbial population rate and the microbial population reaction speed includes the following: Extract the corrosion degree index in the output quality feature set; Construct a two-dimensional structure array of the corrosion degree index, the reference microbial population rate, and the microbial population reaction speed, and embed key node information to obtain a fermentation phase array; Classify the fermentation phase array using a pre-set deviation threshold to obtain fermentation process deviation data.

7. The bio-organic fertilizer production process data extraction method of claim 6, wherein, The embedded key node information includes the following: The embedded key node information includes a first node, a second node, and a third node; The first node represents the microbial population active critical point, which represents the local maximum value of the reference microbial population rate or the microbial population reaction speed at the microbial population active critical point; The second node represents the temperature high plane stable point, which is the stable point of the temperature peak; The third node represents the abnormal point of oxygen fluctuation in the heap, which satisfies the continuous downward trend of oxygen content and is accompanied by a sudden rise in ammonia concentration.

8. The bio-organic fertilizer production process data extraction method of claim 6, wherein, The two-dimensional structure array also includes the following: Construct a two-dimensional structure array, where the horizontal direction is the key time window in the fermentation behavior track, and the vertical direction is the fermentation process deviation data composed of the corrosion degree index; The key time window is based on the raw material tanking time to trace back the gas release abnormal window in the fermentation stage, and expand the equal-width time window sequence centered on the window; The corrosion degree index forms a corresponding sequence in each time window to form a unit comparison unit; Embed the fermentation state difference degree into the unit comparison unit using the key time window as the index to form a two-dimensional structure array.

9. A bio-organic fertilizer production process data extraction system characterized by, The biological organic fertilizer production process data extraction system for executing the biological organic fertilizer production process data extraction method of claim 1 includes: A raw material analysis module for obtaining a raw material information set and a process parameter set; based on the microbial activity characteristics of the raw material information set, calculate the reference microbial population rate; A fermentation analysis module for labeling the time sequence timestamp for the raw material information set and the process parameter set, and constructing a fermentation behavior track; Analyze the microbial activity characteristics of the fermentation behavior track to calculate the microbial population reaction speed; Analyze the microbial activity characteristics of the fermentation behavior track to calculate the microbial population reaction speed; A quality deviation module for collecting the output quality feature set of the organic fertilizer product after the organic fertilizer production line is completed; comparing the corrosion degree index in the output quality feature set with the reference microbial population rate and the microbial population reaction speed to output fermentation process deviation data; A traceability diagnosis module for traceability correlation between fermentation process deviation data and fermentation behavior track to obtain an organic fertilizer process deviation mapping table; identify the abnormal items of the fermentation production line according to the organic fertilizer process deviation mapping table, and construct an organic fertilizer production report.

Images