A method and system for predicting the trend of the degree of humification of straw composting

Abstract

The application relates to the technical field of straw humification prediction, in particular to a trend prediction method and system for the humification degree of straw pile-fermentation, which comprises the following steps: determining the material consumption degree of each monitoring point and the average regional consumption degree of a target straw region based on the temperature time sequence data and the oxygen time sequence data of each monitoring point; determining the cemented region area and the cemented region height from the ground based on the material consumption degree of each monitoring point, the average regional consumption degree and three-dimensional space coordinates; determining the interference constraint characteristic value based on the cemented region area, the cemented region height from the ground, the blade clearance from the ground and the average regional consumption degree of the target straw region; and predicting the humification trend of the target straw region based on the interference constraint characteristic value and the monitoring data. The method can improve the accuracy of humification trend prediction.

Description

Technical Field

[0001] This application relates to the field of straw humification prediction technology, specifically to a trend prediction method and system for the degree of straw composting and humification. Background Technology

[0002] Straw composting is an important way to utilize agricultural waste resources. It involves piling up straw and using microorganisms for aerobic fermentation to convert it into organic fertilizer; this process is called humification. In actual production, straw windrow composting is widely used due to its simplicity and low cost. During composting, the fermenting material is continuously decomposed and consumed by microorganisms, while gradually settling and compacting under gravity. High-density anaerobic clumps easily form at the bottom of the pile, hindering oxygen entry and affecting fermentation efficiency. To maintain an aerobic environment, tracked turners are typically used on-site to periodically turn the pile, using rotary tillers to cut and break up the compacted material and replenish oxygen, thus ensuring the normal progress of the fermentation process.

[0003] In existing technologies, monitoring systems collect temperature and oxygen concentration data in real time by embedding temperature and oxygen probes inside the pile body, and then directly input these time-series data into time series prediction models (such as long short-term memory networks) to predict future humification trends.

[0004] However, there is a fixed safety gap between the rotary tiller blades and the ground. When the compacted material at the bottom of the pile settles below this safety gap due to heavy settling, the rotary tiller blades cannot reach and cut the bottom layer of compacted material, resulting in an oxygen supply disruption and hindering the fermentation process. Existing prediction methods rely solely on temperature and oxygen data for trend prediction, which leads to an excessively high humification index when the bottom layer is limited in cutting. This results in a large discrepancy between the prediction results and the actual fermentation progress, and the accuracy of the prediction is difficult to meet the actual production needs. Summary of the Invention

[0005] To address the technical problem of inaccurate humification index prediction based on temperature and oxygen concentration data, this application aims to provide a method and system for predicting the trend of straw composting humification degree. The specific technical solution adopted is as follows: This application provides a method for predicting the trend of straw composting humification, comprising: acquiring monitoring data from multiple monitoring points within a target straw area and the blade clearance of a turner, wherein the monitoring data includes the three-dimensional spatial coordinates of each monitoring point, temperature time-series data within a preset time period, and oxygen time-series data; determining the material consumption rate of each monitoring point and the average regional consumption rate of the target straw area based on the temperature time-series data and oxygen time-series data of each monitoring point; determining the area of ​​the compacted region and the height of the compacted region above the ground based on the material consumption rate, the average regional consumption rate, and the three-dimensional spatial coordinates of each monitoring point; determining an interference constraint feature value based on the area of ​​the compacted region, the height of the compacted region above the ground, the blade clearance, and the average regional consumption rate of the target straw area, wherein the interference constraint feature value is used to characterize the degree of comprehensive obstruction of the compacted region on the fermentation process; and predicting the humification trend of the target straw area based on the interference constraint feature value and the monitoring data.

[0006] Optionally, the temperature time-series data includes temperatures at multiple collection times, and the oxygen time-series data includes oxygen concentrations at multiple collection times. Based on the temperature and oxygen time-series data for each monitoring point, the material consumption rate and the average regional consumption rate of the target straw area are determined for each monitoring point. This includes: determining the effective fermentation temperature difference at each collection time based on the difference between each temperature in the temperature time-series data and the fermentation start-up temperature threshold; determining the effective oxygen supply concentration difference at each collection time based on the difference between each oxygen concentration in the oxygen time-series data and the aerobic baseline concentration threshold; determining the oxygen consumption degradation rate at each collection time based on the effective fermentation temperature difference and the effective oxygen supply concentration difference at the same collection time; determining the average oxygen consumption degradation rate at multiple collection times as the material consumption rate for each monitoring point; and determining the average material consumption rate at multiple monitoring points as the average regional consumption rate of the target straw area.

[0007] Optionally, the above-mentioned determination of the area of ​​the compacted region and the height of the compacted region above the ground based on the material consumption rate, the average regional consumption rate, and the three-dimensional spatial coordinates of each monitoring point includes: identifying monitoring points where the material consumption rate is greater than the average regional consumption rate of the target straw area as high-consumption monitoring points; projecting the three-dimensional spatial coordinates of each high-consumption monitoring point onto a vertical plane along the direction of travel of the turning machine to obtain the two-dimensional projected coordinates of each high-consumption monitoring point, which include horizontal coordinates and vertical coordinates; and determining the area of ​​the compacted region and the height of the compacted region above the ground based on the two-dimensional projected coordinates of the high-consumption monitoring points.

[0008] Optionally, the above-mentioned determination of the area of ​​the hardened region and the height of the hardened region above the ground based on the two-dimensional projected coordinates of the high-consumption monitoring points includes: when the number of high-consumption monitoring points is greater than or equal to three and they are not collinear, calling the convex hull algorithm to enclose the two-dimensional projected coordinates of the high-consumption monitoring points into a two-dimensional convex polygon; determining the two-dimensional convex polygon as the hardened region and determining the area of ​​the two-dimensional convex polygon as the area of ​​the hardened region; and using the minimum vertical height coordinates of the vertices of the two-dimensional convex polygon as the height of the hardened region above the ground.

[0009] Optionally, the above-mentioned determination of the slab formation area and the ground clearance of the slab formation area based on the two-dimensional projection coordinates of the high-consumption monitoring points includes: when the number of high-consumption monitoring points is less than three, or when the two-dimensional projection coordinates of all high-consumption monitoring points are collinear, determining the minimum value of the vertical height coordinate among the two-dimensional projection coordinates of all high-consumption monitoring points as the ground clearance of the slab formation area, and determining the preset single-cut projection area of ​​the blade as the area of ​​the slab formation area.

[0010] Optionally, determining the interference constraint characteristic value based on the area of ​​the compacted region, the height of the compacted region above the ground, the blade clearance above the ground, and the average regional consumption rate of the target straw area includes: determining the height difference between the blade clearance above the ground and the height of the compacted region above the ground; determining the cutting interference amplification coefficient based on the height difference, which characterizes the extent to which the compacted region exceeds the cutting limit of the turner; and determining the interference constraint characteristic value based on the average regional consumption rate, the area of ​​the compacted region, and the cutting interference amplification coefficient.

[0011] Optionally, the above-mentioned determination of the cutting interference amplification factor based on the height difference includes: when the height difference is positive, determining the ratio of the height difference to the clearance between the cutting tool and the ground, and summing the ratio with 1 to determine the cutting interference amplification factor; when the height difference is not positive, setting the cutting interference amplification factor to 1.

[0012] Optionally, the above-mentioned prediction of the humification trend of the target straw area based on the interference constraint feature value and the monitoring data includes: averaging the temperature of all monitoring points at each collection time and averaging the oxygen concentration of all monitoring points at each collection time to obtain a temperature mean sequence and an oxygen concentration mean sequence; inputting the temperature mean sequence, the oxygen concentration mean sequence, and the interference constraint feature value into the humification trend prediction model to obtain a humification prediction curve, which is used to characterize the humification trend of the target straw area.

[0013] Optionally, after obtaining the humification prediction curve, the method further includes: obtaining an ideal fermentation progress curve; accumulating the difference between the humification prediction curve and the ideal fermentation progress curve within a preset time span to determine the predicted progress lag deviation area; and issuing a lag warning signal when the predicted progress lag deviation area is greater than or equal to a preset tolerance threshold.

[0014] This application also provides a trend prediction system for the degree of humification in straw composting, including a data acquisition unit, a data analysis unit, and a prediction unit. The data acquisition unit acquires monitoring data from multiple monitoring points within a target straw area and the blade clearance of a turner. The monitoring data includes the three-dimensional spatial coordinates of each monitoring point, temperature time-series data within a preset time period, and oxygen time-series data. The data analysis unit determines the material consumption rate of each monitoring point and the average regional consumption rate of the target straw area based on the temperature and oxygen time-series data. The data analysis unit also determines the area of ​​the compacted region and the height above the ground of the compacted region based on the material consumption rate, the average regional consumption rate, and the three-dimensional spatial coordinates of each monitoring point. Furthermore, the data analysis unit determines an interference constraint characteristic value based on the area of ​​the compacted region, the height above the ground of the compacted region, the blade clearance, and the average regional consumption rate of the target straw area. This interference constraint characteristic value characterizes the comprehensive obstruction effect of the compacted region on the fermentation process. The prediction unit predicts the humification trend of the target straw area based on the interference constraint characteristic value and the monitoring data.

[0015] This application has the following beneficial effects: By using time-series temperature and oxygen data from multiple monitoring points, the material consumption rate at each monitoring point is assessed, and the area and height of the compacted region above the ground are further determined. Discrete data are transformed into geometric boundaries characterizing the lateral scale and longitudinal settling depth of the compaction, enabling spatial perception of whether the compacted region has reached the mechanical cutting limit. Subsequently, by comparing the height of the compacted region above the ground with the blade clearance above the ground, and integrating the average regional consumption rate with the area of ​​the compacted region, an interference constraint feature value is determined to quantify the comprehensive obstruction effect of the compacted region on the fermentation process. Finally, this interference constraint feature value is introduced when predicting the humification trend of the target straw area, enabling the output humification prediction curve to truly reflect the fermentation lag caused by mechanical cutting limitations, significantly improving the accuracy of humification prediction. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of this application 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 of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating a method for predicting the trend of straw composting humification degree, provided as an embodiment of this application; Figure 2 This is a structural diagram of a trend prediction system for the degree of humification of straw composting, provided as an embodiment of this application. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method and system for predicting the degree of humification of straw composting proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0020] The following, in conjunction with the accompanying drawings, details a specific scheme for a method and system for predicting the trend of straw composting and humification provided in this application.

[0021] Please see Figure 1 The diagram illustrates a flowchart of a method for predicting the trend of straw composting humification degree according to an embodiment of this application.

[0022] like Figure 1 As shown, the trend prediction method for the degree of straw composting and humification includes S101-S105.

[0023] S101. Obtain monitoring data from multiple monitoring points within the target straw area and the blade clearance of the turner.

[0024] The monitoring data includes the three-dimensional spatial coordinates of each monitoring point, temperature time-series data within a preset time period, and oxygen time-series data.

[0025] In this embodiment, the cross-section of the pile currently being turned by the turner is defined as the current working section, the current direction of travel of the turner is defined as the X-axis, and the cross-section to be turned that is located in front of the turner and at a distance of a preset safe travel distance (e.g., 1-2 meters) from the current working section is defined as the target section. Using the target section as the reference plane, the three-dimensional space extending in the X-axis direction (in front of the turner) with a preset depth slice thickness (the empirical value can be configured as 0.5 meters) is defined as the target straw area.

[0026] Monitoring points refer to IoT sensor nodes pre-embedded within the target straw area, used to collect data on the composting environment. The three-dimensional spatial coordinates of the monitoring points are used to locate their spatial position within the compost pile, including the depth coordinate along the direction of travel of the compost turner (i.e., the X-axis), the horizontal coordinate perpendicular to the direction of travel (i.e., the Y-axis), and the vertical coordinate (i.e., the Z-axis); the temperature time series data includes the temperature at multiple collection times within a preset time period; the oxygen time series data includes the oxygen concentration at multiple collection times within a preset time period (corresponding to the collection times of the temperature time series data).

[0027] The preset duration refers to the time between the end of the last turning operation of the compost turner on the target straw area and the current time.

[0028] Optionally, within the preset duration, temperature time-series data and oxygen time-series data can be collected according to a preset data sampling frequency (e.g., once every 10 minutes).

[0029] The blade clearance of a compost turner refers to the constant height (generally 15-20 cm) of the rotary tillage blades from the ground when they are at full power. This parameter serves as a rigid physical limit for mechanical cutting.

[0030] S102. Based on the temperature time series data and oxygen time series data of each monitoring point, determine the material consumption rate of each monitoring point and the average regional consumption rate of the target straw area.

[0031] It should be understood that the material consumption rate of a monitoring point is used to characterize the degree of local compaction of straw material at that monitoring point due to excessive consumption; the regional consumption average is used to characterize the average level of overall fermentation consumption in the target straw area, serving as a dynamic comparison benchmark for screening high-consumption monitoring points.

[0032] In one optional implementation, the effective fermentation temperature difference at each collection moment can be determined based on the difference between each temperature and the fermentation start-up temperature threshold in the temperature time series data of each monitoring point; and the effective oxygen supply concentration difference at each collection moment can be determined based on the difference between each oxygen concentration and the aerobic baseline concentration threshold in the oxygen time series data of each monitoring point; the oxygen consumption degradation rate at each collection moment can be determined based on the effective fermentation temperature difference and the effective oxygen supply concentration difference at the same collection moment; the average of the oxygen consumption degradation rates at multiple collection moments can be determined as the material consumption rate at each monitoring point; and the average of the material consumption rates at multiple monitoring points can be determined as the regional consumption rate average of the target straw area.

[0033] It should be understood that the fermentation initiation temperature threshold is used to characterize the bottom line of the biochemical reaction initiation temperature at which thermophilic microorganisms begin to become active and substantially consume materials. It is used to distinguish between the effective temperature for promoting fermentation and the ineffective environmental temperature fluctuations caused by the natural diurnal cycle. For example, an empirical value of 50 degrees Celsius can be taken. When the temperature inside the pile is lower than the fermentation initiation temperature threshold, it indicates that the metabolic activity of microorganisms inside the pile is weak. At this time, the temperature data only reflects the diurnal variation of the environment rather than the heat generated by fermentation. Only when the pile temperature exceeds the fermentation initiation temperature threshold does the temperature data represent the true heat effect of fermentation.

[0034] The aerobic baseline concentration threshold is used to characterize the minimum oxygen concentration limit required to maintain the aerobic fermentation process. It is used to distinguish between effective oxygen supply and ineffective oxygen concentration in the anaerobic critical state. An empirical value of 5% can be used as an example. When the oxygen concentration is lower than this threshold, the pile enters the anaerobic state, and the oxygen concentration data at this time does not represent the aerobic fermentation support capacity.

[0035] In this embodiment, after subtracting the fermentation start-up temperature threshold from each temperature in the temperature time series data to obtain the initial temperature difference, a one-dimensional linear rectification function can be used to retain the positive values ​​in the initial temperature difference as the effective fermentation temperature difference, and the negative values ​​can be reset to zero. Similarly, after subtracting the aerobic baseline concentration threshold from each oxygen concentration in the oxygen time series data to obtain the initial oxygen difference, a one-dimensional linear rectification function can be used to retain the positive values ​​in the initial oxygen difference as the effective oxygen supply concentration difference, and the negative values ​​can be reset to zero.

[0036] It should be understood that the function of the one-dimensional linear rectified unit (ReLU) is to eliminate the invalid environmental fluctuations caused by the natural alternation of day and night, and retain only the temperature difference exceeding the fermentation start-up temperature threshold as the effective fermentation temperature difference for actual fermentation promotion, and the oxygen concentration difference exceeding the aerobic baseline concentration threshold as the effective oxygen supply concentration difference.

[0037] It is understandable that the humification of materials is a process that requires both heat generation (temperature increase) and oxygen consumption (oxygen depletion) to occur simultaneously. Only the synergistic effect of both can truly reflect the decomposition intensity of microorganisms. Therefore, the maximum and minimum values ​​of the effective fermentation temperature difference and the effective oxygen concentration difference can be extracted from multiple monitoring points. These values ​​are then normalized to the maximum and minimum values ​​respectively, mapping them to the range of [0, 1] to eliminate the influence of dimensions. The normalized values ​​are then multiplied to obtain the oxygen consumption degradation rate at each sampling time. A higher oxygen consumption degradation rate indicates more intense humification and a more significant structural collapse, material compaction, and settling effect caused by excessive microbial consumption of straw.

[0038] Optionally, the material consumption rate at a monitoring point satisfies the following formula: in, Indicates the first Material consumption at each monitoring point This indicates the number of data collection moments within a preset time period. Indicates the index of the time of data collection; Indicates the first Each monitoring point at the time of data collection The normalized value of the effective fermentation temperature difference; Indicates the first Each monitoring point at the time of data collection The normalized value of the effective oxygen concentration difference. Indicates the first Each monitoring point at the time of data collection The oxygen-consuming degradation rate.

[0039] This formula averages the oxygen consumption degradation rate at multiple times, eliminating the data scale differences caused by different turning and turning cycle lengths, making the material consumption rates of different batches and different preset durations comparable.

[0040] It should be understood that after obtaining the material consumption rate at each monitoring point, the average of the material consumption rates at multiple monitoring points can be used to obtain the average regional consumption rate of the target straw area.

[0041] S103. Based on the material consumption rate, regional consumption rate average, and three-dimensional spatial coordinates of each monitoring point, determine the area of ​​the slab compaction region and the height of the slab compaction region above the ground.

[0042] It should be understood that the compacted area refers to the area where the anaerobic hard clumps are located at the bottom of the target straw area. The area of ​​the compacted area refers to the projected area of ​​the compacted area on a two-dimensional plane perpendicular to the direction of travel of the turning machine, which is used to characterize the lateral scale of the anaerobic hard clumps. The height of the compacted area from the ground refers to the vertical distance between the lowest boundary of the compacted area in the vertical direction and the ground.

[0043] In one alternative implementation, monitoring points where the material consumption rate is greater than the average consumption rate of the target straw area can be identified as high-consumption monitoring points. The three-dimensional spatial coordinates of each high-consumption monitoring point are projected onto a vertical plane along the direction of travel of the turner to obtain the two-dimensional projected coordinates of each high-consumption monitoring point, which include horizontal coordinates and vertical coordinates. The area of ​​the compacted region and the height of the compacted region above the ground are determined based on the two-dimensional projected coordinates of the high-consumption monitoring points.

[0044] It should be understood that the basic moisture content and fermentation potential of straw materials vary significantly in different seasons and batches. Therefore, the lower limit of the characteristic of local material over-compaction and downward settling caused by biochemical fermentation is a dynamic process that evolves with the composting process. A fixed material consumption threshold cannot adapt to the above batch differences. It is inaccurate to determine high consumption monitoring points based on a fixed material consumption threshold. Therefore, in this embodiment, the overall material consumption level of the target straw area (i.e., the average regional consumption) is used as the dynamic threshold for identifying high consumption monitoring points. This dynamic threshold increases dynamically with the increase of the overall fermentation intensity.

[0045] It is understandable that if the material consumption at a monitoring point is greater than the average consumption in the area, it means that the material consumption at that monitoring point is greater than the overall average level, and the location of that monitoring point exhibits significant high-density settlement and compaction characteristics. In this case, the monitoring point can be identified as a high-consumption monitoring point, and its three-dimensional spatial coordinates can be extracted.

[0046] It should be understood that during the cutting operation of the turner, the blade completes the cutting action in a two-dimensional cross section (i.e., the YZ plane) perpendicular to the direction of travel, compressing the three-dimensional spatial coordinates into a two-dimensional cross section. This corresponds to the physical reality of the turner's operation of "processing according to cross section" and transforms the problem into "the geometric characteristics of the distribution of the slab-forming area in the horizontal and vertical directions on the cross section to be processed".

[0047] Meanwhile, calculating the volume of the hardened region based on three-dimensional spatial coordinates would lead to a dramatic increase in computational load, and it would also be impossible to directly extract the virtual ground clearance of the hardened region.

[0048] Optionally, the three-dimensional spatial coordinates of each high-consumption monitoring point can be projected onto a vertical plane along the direction of travel of the turning machine. Specifically, the depth component (i.e., X-axis coordinate) along the direction of travel of the turning machine in the three-dimensional spatial coordinates of the high-consumption monitoring point is ignored, and only the horizontal coordinate (i.e., Y-axis coordinate) and vertical height coordinate (i.e., Z-axis coordinate) perpendicular to the direction of travel are retained to generate two-dimensional projected coordinates (y, z).

[0049] It should be understood that the X-axis is the time-progressing axis of the turner. The turner advances step by step along the X-axis, processing different cross sections in sequence, while the vertical plane (i.e., the YZ plane) is the working section of the turner. Since multiple monitoring points are usually distributed within the thickness of the longitudinal slice, these monitoring points are distributed in a "strip" shape in three-dimensional space. The three-dimensional convex hull algorithm will regard the depth difference of these points on the X-axis as the "length" of the compacted area, thus constructing an irregular three-dimensional body that is elongated along the direction of travel, incorrectly representing that the compaction is continuously distributed in the direction of travel. After being projected onto the vertical plane, these points are clustered on the two-dimensional plane, and the point set density is significantly improved, ensuring that the subsequent two-dimensional polygon algorithm has sufficient input point set density, and thus can accurately reconstruct the true lateral boundary of the compacted area on the cross section.

[0050] In one alternative implementation, if the number of high-consumption monitoring points is greater than or equal to three and they are not collinear, the convex hull algorithm is invoked to enclose the two-dimensional projected coordinates of the high-consumption monitoring points into a two-dimensional convex polygon; the two-dimensional convex polygon is determined as the hardening region, and the area of ​​the two-dimensional convex polygon is determined as the area of ​​the hardening region; the minimum vertical height coordinate of the vertex of the two-dimensional convex polygon is taken as the height of the hardening region above the ground.

[0051] It is understandable that if the number of high-consumption monitoring points is greater than or equal to three and they are not collinear, it indicates that the point density is sufficient and the convex hull algorithm can be executed. In this case, the two-dimensional projected coordinates of the high-consumption monitoring points can be enveloped into a two-dimensional convex polygon based on the convex hull algorithm.

[0052] Specifically, the method of using the convex hull algorithm to enclose the two-dimensional projected coordinates of high-consumption monitoring points into a two-dimensional convex polygon is as follows: the coordinates of the high-consumption monitoring points are determined as coordinate points, and the coordinate point with the smallest vertical height coordinate value among all high-consumption monitoring points is established as the pole origin. The polar angle of each of the remaining coordinate points relative to the pole origin is calculated with the horizontal direction as the polar axis. The coordinate points are sorted in ascending order of polar angle values. The cross product of vectors on the two-dimensional plane is calculated using the sorted coordinate points. The turning attribute of the connecting path is determined according to the sign of the cross product. Coordinate points that cause the path to be concave are removed, and the minimum two-dimensional envelope convex polygon is constructed.

[0053] It should be understood that the convex polygon is the geometric boundary of the compaction region, and its area is the area of ​​the compaction region, representing the lateral scale of the anaerobic hard blocks; the minimum vertical height coordinate of its vertex is the height of the compaction region above the ground, representing the lowest boundary of the compaction region extending towards the bottom of the pile.

[0054] In another alternative implementation, if the number of high-consumption monitoring points is less than three, or if the two-dimensional projected coordinates of all high-consumption monitoring points are collinear, the minimum value of the vertical height coordinate among the two-dimensional projected coordinates of all high-consumption monitoring points can be determined as the height of the slab region above the ground, and the preset single-cut projection area of ​​the blade can be determined as the area of ​​the slab region.

[0055] It is understandable that if the number of high-consumption monitoring points is less than three, or if the two-dimensional projected coordinates of all high-consumption monitoring points are collinear, it indicates that the number of high-consumption monitoring points is insufficient or their distribution is special and cannot form a two-dimensional convex polygon. In this case, the downgrade default strategy is triggered, and the minimum vertical height in the coordinates of the existing high-consumption monitoring points is directly extracted as the height of the slab region above the ground, and the preset single-cut projection area of ​​the blade is used as the area of ​​the slab region.

[0056] Optionally, the single-cut projection area of ​​the blade in the historical turning record of the turning machine can be determined as the preset single-cut projection area of ​​the blade.

[0057] The method provided in S103 above identifies high-consumption monitoring points as monitoring points where the material consumption is greater than the regional average consumption, thus achieving adaptive dynamic threshold filtering. Then, the three-dimensional spatial coordinates are projected onto the vertical plane along the direction of travel of the turning machine to obtain two-dimensional projected coordinates containing horizontal and vertical coordinates. This retains the horizontal width and vertical height information required by the subsequent convex hull algorithm, eliminates redundant dimensions in the direction of travel, and reduces computational complexity.

[0058] Finally, when the number of high-consumption monitoring points is sufficient and they are not collinear, a convex hull algorithm is used to construct a two-dimensional convex polygon, transforming discrete and isolated high-consumption monitoring points into continuous geometric regions with actual areas and minimum boundaries. This allows the lateral scale (area) and longitudinal extension depth (height above ground) of the hardened region to be quantified. When the number of high-consumption monitoring points is insufficient or they are collinear, the minimum vertical height is extracted from the existing high-consumption monitoring points as the height above ground of the hardened region, and the preset single-cut projection area of ​​the blade is used as the area of ​​the hardened region. This ensures the robustness and integrity of the algorithm under various data conditions and avoids the failure of the entire prediction process due to local data sparsity.

[0059] S104. Based on the area of ​​the compacted region, the height of the compacted region above the ground, the clearance between the blade and the ground, and the average regional consumption of the target straw area, determine the interference constraint characteristic value.

[0060] The interference constraint eigenvalue is used to characterize the overall obstruction effect of the compacted region on the fermentation process.

[0061] It should be understood that when the height of the compacted area from the ground is lower than the clearance between the blade and the ground, it indicates that it is difficult to cut the compacted area, the fermentation process is hindered, and the interference constraint characteristic value should be larger. The larger the average consumption of the area, the more intense the material consumption of the target straw area, the more vigorous the fermentation is, and the more likely sedimentation and compaction will occur, so the interference constraint characteristic value should be larger. The larger the area of ​​the compacted area, the wider the distribution of high consumption monitoring points, the larger the area occupied by the compacted area on the cross-section, and the greater the mechanical blocking load encountered by the blade during cutting, so the interference constraint characteristic value should be larger.

[0062] In one alternative implementation, the height difference between the blade clearance and the ground clearance of the slag region can be determined; the cutting interference amplification factor can be determined based on the height difference; and then the interference constraint characteristic value can be determined based on the average consumption of the region, the area of ​​the slag region, and the cutting interference amplification factor.

[0063] The cutting interference amplification factor is used to characterize the extent to which the caking region exceeds the cutting limit of the turning machine.

[0064] It should be understood that this height difference is the difference between the blade clearance from the ground and the height of the slab-forming area from the ground.

[0065] Optionally, if the height difference is positive, the ratio of the height difference to the clearance between the cutting tool and the ground can be determined, and the sum of the ratio and 1 can be determined as the cutting interference amplification factor; if the height difference is not positive, the cutting interference amplification factor is set to 1.

[0066] Understandably, the height difference represents the absolute depth difference between the boundary of the compacted area and the mechanical cutting baseline of the tiller. When the height difference is positive, the height of the compacted area above the ground is less than the blade clearance, indicating that the lower boundary of the compacted area has extended to the interference dead zone that the rotary tiller blades cannot reach. At this time, the cutting interference amplification factor should be large. The ratio of the height difference to the blade clearance is used to characterize the relative excess depth relative to the mechanical cutting baseline of the tiller. The larger this ratio is, the larger the cutting interference amplification factor should be.

[0067] When the height difference is negative or zero, the height of the compacted area above the ground is greater than or equal to the blade clearance above the ground, indicating that the rotary tiller blade can cut to the lower boundary of the compacted area. In this case, there is no need to amplify the cutting interference amplification factor, and the cutting interference amplification factor is kept at the reference value, i.e., 1.

[0068] It should be understood that the mean regional consumption, the area of ​​the compacted region, and the cutting interference amplification coefficient are all directly proportional to the interference constraint characteristic value. Therefore, the mean regional consumption, the area of ​​the compacted region, and the cutting interference amplification coefficient can be normalized to the maximum and minimum values ​​respectively, mapping their values ​​to the range of [0, 1]. The maximum and minimum values ​​required for normalization can be obtained from the historical monitoring data of the target straw area. Then, the three values ​​are multiplied together to obtain the interference constraint characteristic value, which is a dimensionless parameter.

[0069] The method provided in S104 above determines the cutting interference amplification factor based on the height difference between the blade's ground clearance and the ground clearance of the slag-forming region. Then, it multiplies the normalized values ​​of the average regional consumption, the area of ​​the slag-forming region, and the cutting interference amplification factor. This multiplication operation integrates three different dimensions of features (biochemical consumption intensity, spatial geometry, and mechanical over-limit amplitude) into a dimensionless composite feature value, achieving unified quantification of cross-domain information.

[0070] S105. Based on interference constraint eigenvalues ​​and monitoring data, predict the humification trend of the target straw area.

[0071] In one alternative implementation, the temperature and oxygen concentration at all monitoring points at each time point can be arithmetically averaged. The averages are then arranged in chronological order to obtain a temperature average sequence and an oxygen concentration average sequence. This temperature average sequence, the oxygen concentration average sequence, and the interference constraint feature value are then input into the humification trend prediction model to obtain the humification prediction curve.

[0072] The humification prediction curve is used to characterize the humification trend of the target straw area.

[0073] It should be understood that the average temperature series and the average oxygen concentration series can reflect the recent overall temperature and oxygen fluctuation trend of the entire target straw area.

[0074] Optionally, an array concatenation operation can be performed on the interference constraint feature values, the temperature mean sequence, and the oxygen concentration mean sequence, and the concatenated feature can be used as the input feature of the humification trend prediction model, so that the input feature simultaneously contains the fermentation basic trend features in the time dimension and the mechanical interference expectation features in the spatial dimension in the data structure.

[0075] In this embodiment, the humification trend prediction model is a pre-trained Long Short-Term Memory (LSTM) network. Specifically, it is a supervised pre-trained LSTM network based on real historical batch fermentation data. The input features of its training samples are historical environmental sequences (including historical average temperature sequences and historical average oxygen concentration sequences) and historical interference constraint feature values ​​generated using the same calculation steps described above. The label data is the daily humification index measured in the laboratory during the same period. Through pre-training, the humification trend prediction model learns the numerical mapping law between the magnitude of the interference constraint feature values ​​and the actual fermentation lag phenomenon. That is, a larger interference constraint feature value corresponds to a higher probability that the underlying hard blocks are not broken up and that oxygen supply is disrupted, and it has a strong negative correlation with the speed of humification process.

[0076] When applying the model, after receiving the aforementioned spliced ​​input features, the humification trend prediction model compresses the expected future increase based on the interference constraint feature values ​​in the current input features, thereby outputting a humification prediction curve constrained by mechanical cutting conditions. This humification prediction curve reflects the true trend of fermentation evolution within a set number of days in the future, after considering the risk of oxygen supply disruption caused by limited cutting at the bottom layer.

[0077] It should be understood that this humification prediction curve is a curve that predicts the change of the humification index over time.

[0078] In one implementation of this application, after obtaining the humification prediction curve, the method further includes: obtaining an ideal fermentation progress curve; accumulating the difference between the humification prediction curve and the ideal fermentation progress curve within a preset time span to determine the predicted progress lag deviation area; and issuing a lag warning signal when the predicted progress lag deviation area is greater than or equal to a preset tolerance threshold.

[0079] It should be understood that the ideal fermentation progress curve represents the fermentation evolution trend of a standard daily fermentation process under the premise that the material remains loose and there is no physical interference during the turning process. The humification prediction curve is a curve that predicts the change of the humification index over time.

[0080] Optionally, the ideal fermentation progress curve and the humification prediction curve can be aligned in the same time coordinate system. The value of each time node in the ideal fermentation progress curve within the preset time span is subtracted from the value of the corresponding time node in the humification prediction curve to obtain a difference sequence distributed along the time axis within the preset time span. Then, the difference sequence is subjected to discrete accumulation and summation calculation to obtain the area of ​​the predicted progress lag deviation.

[0081] For example, the preset time span can be set to the next five days, and the interval between two time points can be one day.

[0082] Optionally, the formula for calculating the predicted area of ​​schedule lag deviation is: in, This represents the area of ​​the projected progress lag. Indicates the number of time nodes within a preset time span; This indicates the ideal fermentation progress curve at time points. The ideal humification index; This indicates the humification prediction curve at the time node. The predicted humification index.

[0083] Optionally, the preset tolerance threshold can be determined using the benchmark proportional extraction method, that is, the sum of the ideal humification index at each time node of the ideal fermentation progress curve within the preset time span is calculated as the benchmark total area, and 10% to 15% of the benchmark total area is extracted as the preset tolerance threshold.

[0084] Understandably, when the predicted progress lag area is greater than or equal to the tolerance threshold, it indicates that the future fermentation process will be irreversibly delayed due to mechanical cutting limitations. At this time, a lag warning signal can be issued to prompt operators to take timely intervention measures.

[0085] It should be understood that by calculating the area of ​​deviation between the humification prediction curve and the ideal fermentation progress curve and setting a preset tolerance threshold, the abstract mechanical interference effect is transformed into an intuitive quantitative early warning indicator, realizing early warning of the problem of the bottom cutting blind zone, and providing a clear basis for timely human intervention measures.

[0086] The methods provided in S101-S105 above assess the material consumption at each monitoring point using time-series temperature and oxygen data from multiple monitoring points. This further determines the area and height of the compacted region above the ground, transforming discrete data into geometric boundaries characterizing the lateral scale and longitudinal settling depth of the compacted area. This enables spatial perception of whether the compacted region has reached the limits of mechanical cutting. Subsequently, by comparing the height of the compacted region above the ground with the blade clearance, and fusing the average regional consumption with the area of ​​the compacted region, an interference constraint feature value is determined to quantify the comprehensive obstruction effect of the compacted region on the fermentation process. Finally, this interference constraint feature value is introduced when predicting the humification trend of the target straw area, enabling the output humification prediction curve to accurately reflect the fermentation lag caused by limited mechanical cutting, significantly improving the accuracy of humification prediction.

[0087] like Figure 2As shown in the figure, this application embodiment also provides a trend prediction system 20 for the degree of straw composting and humification, including a data acquisition unit 201, a data analysis unit 202 and a prediction unit 203.

[0088] The data acquisition unit 201 is used to acquire monitoring data from multiple monitoring points within the target straw area and the blade clearance of the turner.

[0089] The monitoring data includes the three-dimensional spatial coordinates of each monitoring point, temperature time-series data within a preset time period, and oxygen time-series data.

[0090] The data analysis unit 202 is used to determine the material consumption rate of each monitoring point and the average regional consumption rate of the target straw area based on the temperature time series data and oxygen time series data of each monitoring point.

[0091] The data analysis unit 202 is also used to determine the area of ​​the slab-forming region and the height of the slab-forming region above the ground based on the material consumption rate, the average regional consumption rate, and the three-dimensional spatial coordinates of each monitoring point.

[0092] The data analysis unit 202 is also used to determine the interference constraint characteristic value based on the area of ​​the compacted region, the height of the compacted region from the ground, the blade clearance from the ground, and the average regional consumption of the target straw area.

[0093] The interference constraint eigenvalue is used to characterize the overall obstruction effect of the compacted region on the fermentation process.

[0094] The prediction unit 203 is used to predict the humification trend of the target straw area based on the interference constraint feature value and monitoring data.

[0095] It should be noted that the trend prediction system 20 for the degree of straw composting and humification can execute any of the above-mentioned optional trend prediction methods for the degree of straw composting and humification.

[0096] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0097] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for predicting the trend of straw composting and humification degree, characterized in that, The method includes: The monitoring data of multiple monitoring points within the target straw area and the blade clearance of the turner are obtained. The monitoring data includes the three-dimensional spatial coordinates of each monitoring point, temperature time series data within a preset time period, and oxygen time series data. Based on the time-series temperature and oxygen data of each monitoring point, the material consumption rate of each monitoring point and the average regional consumption rate of the target straw area are determined. Based on the material consumption rate, regional average consumption rate, and three-dimensional spatial coordinates of each monitoring point, the area of ​​the slab compaction zone and the height of the slab compaction zone above the ground are determined. Based on the area of ​​the compacted region, the height of the compacted region above the ground, the blade clearance from the ground, and the average regional consumption of the target straw area, interference constraint characteristic values ​​are determined. These interference constraint characteristic values ​​are used to characterize the degree of comprehensive obstruction of the compacted region on the fermentation process. Based on the interference constraint eigenvalues ​​and the monitoring data, the humification trend of the target straw area is predicted.

2. The method for predicting the trend of straw composting and humification degree according to claim 1, characterized in that, The temperature time-series data includes temperatures at multiple sampling times, and the oxygen time-series data includes oxygen concentrations at multiple sampling times. Based on the temperature and oxygen time-series data for each monitoring point, the determination of the material consumption rate and the average regional consumption rate of the target straw area for each monitoring point includes: Based on the difference between each temperature and the fermentation start-up temperature threshold in the temperature time series data of each monitoring point, the effective fermentation temperature difference at each collection time is determined; and based on the difference between each oxygen concentration and the aerobic baseline concentration threshold in the oxygen time series data of each monitoring point, the effective oxygen supply concentration difference at each collection time is determined. Based on the effective fermentation temperature difference and effective oxygen supply concentration difference at the same collection time, the oxygen consumption degradation rate at each collection time is determined; The average oxygen consumption degradation rate at multiple collection times is used to determine the material consumption rate at each monitoring point. The average material consumption rate at multiple monitoring points is determined as the regional average consumption rate of the target straw area.

3. The method for predicting the trend of straw composting and humification degree according to claim 1, characterized in that, The determination of the compaction area area and the height of the compaction area above the ground based on the material consumption rate, regional average consumption rate, and three-dimensional spatial coordinates at each monitoring point includes: Monitoring points where the material consumption rate is greater than the average consumption rate of the target straw area are identified as high consumption monitoring points. The three-dimensional spatial coordinates of each high-consumption monitoring point are projected onto a vertical plane along the direction of travel of the turning machine to obtain the two-dimensional projected coordinates of each high-consumption monitoring point. The two-dimensional projected coordinates include horizontal coordinates and vertical height coordinates. The area of ​​the sludge compaction region and its height above the ground are determined based on the two-dimensional projection coordinates of high-consumption monitoring points.

4. The method for predicting the trend of straw composting and humification degree according to claim 3, characterized in that, The determination of the area and height above ground of the compaction region based on the two-dimensional projected coordinates of high-consumption monitoring points includes: When the number of high-consumption monitoring points is greater than or equal to three and they are not collinear, the convex hull algorithm is invoked to enclose the two-dimensional projected coordinates of the high-consumption monitoring points into a two-dimensional convex polygon. The two-dimensional convex polygon is defined as the plated region, and the area of ​​the two-dimensional convex polygon is defined as the area of ​​the plated region. The minimum vertical height coordinates of the vertices of the two-dimensional convex polygon are used as the ground height of the slab-forming region.

5. The method for predicting the trend of straw composting humification degree according to claim 3, characterized in that, The determination of the area and height above ground of the compaction region based on the two-dimensional projected coordinates of high-consumption monitoring points includes: If the number of high-consumption monitoring points is less than three, or if the two-dimensional projection coordinates of all high-consumption monitoring points are collinear, the minimum value of the vertical height coordinate among the two-dimensional projection coordinates of all high-consumption monitoring points is determined as the height of the slab-forming area above the ground, and the preset single-cut projection area of ​​the blade is determined as the area of ​​the slab-forming area.

6. The method for predicting the trend of straw composting and humification degree according to claim 1, characterized in that, The determination of interference constraint characteristic values ​​based on the area of ​​the compacted region, the height of the compacted region above the ground, the blade clearance above the ground, and the average regional consumption of the target straw area includes: Determine the height difference between the blade clearance from the ground and the height of the slab-forming area from the ground; The cutting interference amplification factor is determined based on the height difference, and the cutting interference amplification factor is used to characterize the extent by which the slab-forming area exceeds the cutting limit of the turning machine; The interference constraint characteristic value is determined based on the average consumption of the region, the area of ​​the slab region, and the cutting interference amplification factor.

7. The method for predicting the trend of straw composting and humification degree according to claim 6, characterized in that, The determination of the cutting interference amplification factor based on the height difference includes: When the height difference is positive, the ratio of the height difference to the blade's ground clearance is determined, and the sum of the ratio and 1 is determined as the cutting interference amplification factor; If the height difference is not positive, the cutting interference amplification factor is set to 1.

8. The method for predicting the trend of straw composting and humification degree according to claim 1, characterized in that, The prediction of the humification trend of the target straw area based on the interference constraint eigenvalues ​​and the monitoring data includes: The temperature at all monitoring points at each acquisition time is arithmetically averaged, and the oxygen concentration at all monitoring points at each acquisition time is arithmetically averaged to obtain the temperature mean sequence and the oxygen concentration mean sequence. The average temperature sequence, the average oxygen concentration sequence, and the interference constraint feature value are input into the humification trend prediction model to obtain the humification prediction curve, which is used to characterize the humification trend of the target straw area.

9. The method for predicting the trend of straw composting and humification degree according to claim 8, characterized in that, After obtaining the humification prediction curve, the process also includes: Obtain the ideal fermentation progress curve; The difference between the predicted humification curve and the ideal fermentation progress curve within a preset time span is accumulated and determined as the predicted progress lag deviation area. If the area of ​​the predicted progress lag is greater than or equal to a preset tolerance threshold, a lag warning signal will be issued.

10. A trend prediction system for the degree of humification in straw composting, characterized in that, It includes a data acquisition unit, a data analysis unit, and a prediction unit; The data acquisition unit is used to acquire monitoring data from multiple monitoring points within the target straw area and the blade clearance of the turner. The monitoring data includes the three-dimensional spatial coordinates of each monitoring point, temperature time series data within a preset time period, and oxygen time series data. The data analysis unit is used to determine the material consumption rate of each monitoring point and the average regional consumption rate of the target straw area based on the temperature time series data and oxygen time series data of each monitoring point. The data analysis unit is also used to determine the area of ​​the slab formation region and the height of the slab formation region above the ground based on the material consumption rate, the average regional consumption rate, and the three-dimensional spatial coordinates of each monitoring point. The data analysis unit is also used to determine interference constraint feature values ​​based on the area of ​​the compacted region, the height of the compacted region above the ground, the blade clearance from the ground, and the average regional consumption of the target straw region. The interference constraint feature values ​​are used to characterize the degree of comprehensive obstruction of the compacted region on the fermentation process. The prediction unit is used to predict the humification trend of the target straw area based on the interference constraint feature value and the monitoring data.

Images