A method for regulating the expansion of the topsoil by integrating a carbon and nitrogen conversion model of organic materials

Abstract

This invention relates to the field of agricultural data analysis, and particularly to a method for expanding and regulating the topsoil volume by integrating an organic material carbon and nitrogen conversion model. The method includes: acquiring raw time-series data of carbon and nitrogen mineralization rates; obtaining a phase hysteresis correction factor by coupling and evaluating the temporal differences and morphological overlap characteristics of the mineralization energy centroid; obtaining a transient excitation smoothing factor by normalizing and attenuating the local oscillation energy functional of the sequence with the average oscillation energy of the entire sample; obtaining a corrected distance by performing multiplicative optimization of the Euclidean distance and performing iterative clustering calculations; and obtaining a topsoil volume expansion and regulation scheme by matching carbon and nitrogen conversion model parameters and inversely deriving the regulation strategy from the clustering results. This addresses the problems of distortion in the division of regulation type regions and decreased model matching accuracy caused by inoculation hysteresis translation and high-frequency oscillations of excitation effects in existing Euclidean distance-based mineralization rate time-series clustering.

Description

Technical Field

[0001] This invention relates to the field of agricultural data analysis technology, and in particular to a method for expanding and regulating the topsoil volume by integrating an organic material carbon and nitrogen conversion model. Background Technology

[0002] In agricultural topsoil expansion projects, the application of organic materials such as straw and livestock manure to the soil, along with fertilization, is a crucial means to improve topsoil thickness and soil fertility. To formulate precise material input and fertilization control strategies for different plots, it is typically necessary to sample the soil in the target area and conduct aerobic incubation experiments under constant temperature and humidity conditions in a laboratory. This involves continuously measuring carbon or nitrogen release per unit time to generate time-series data on the carbon and nitrogen mineralization rates over time. Existing control systems often employ clustering methods to classify the mineralization rate time series from different monitoring points, dividing them into control type zones. The system then uses representative curves from these zones to match carbon and nitrogen conversion model parameters to output agricultural operation suggestions. A commonly used similarity metric is the Euclidean distance calculated based on the point-to-point difference at sampling time. However, the decomposition of newly introduced organic materials by soil microorganisms is subject to inoculation lag differences and excitation effects, leading to time axis shifts and superimposed high-frequency transient sawtooth fluctuations in the mineralization rate curve. In this situation, Euclidean distance misinterprets phase shifts as large numerical differences and amplifies transient oscillation noise, causing clustering results to deviate from the true carbon and nitrogen transformation patterns in the soil. This reduces the reliability of the regulatory type region division and the accuracy of subsequent model parameter matching. Therefore, effectively removing the artificially inflated distance interference caused by inoculation lag and excitation effects during mineralization rate time series clustering and achieving stable classification of the true carbon and nitrogen transformation characteristics of the soil is a pressing problem that needs to be solved. Summary of the Invention

[0003] In view of this, the present invention aims to propose a method for expanding and regulating the topsoil volume by integrating a carbon and nitrogen conversion model of organic materials, in order to solve the problems of distortion in the division of regulation type regions and decrease in model matching accuracy caused by high-frequency oscillations of inoculation hysteresis translation and excitation effect in the existing mineralization rate time series clustering based on Euclidean distance.

[0004] To achieve the above objectives, the technical solution of the present invention is implemented as follows:

[0005] A method for regulating topsoil volume expansion by integrating an organic material carbon and nitrogen conversion model, the method comprising:

[0006] Step S1: Obtain raw time series data of carbon and nitrogen mineralization rate by collecting topsoil samples and measuring the carbon and nitrogen mineralization rate of organic materials;

[0007] Step S2: Obtain the phase hysteresis correction factor by coupling the time difference of the mineralization energy centroid with the morphological overlap characteristics;

[0008] Step S3: Obtain the transient excitation smoothing factor by normalizing and decaying the local oscillation energy functional of the sequence and the average oscillation energy of the whole sample;

[0009] Step S4: Obtain the corrected distance by multiplicatively correcting the Euclidean distance with the phase hysteresis correction factor and the transient excitation smoothing factor, and then perform clustering iterative calculation.

[0010] Step S5: Obtain the topsoil expansion control scheme by matching carbon and nitrogen conversion model parameters and back-deriving control strategies based on the clustering results.

[0011] Furthermore, the process of collecting topsoil samples and measuring the carbon and nitrogen mineralization rates of organic matter to obtain raw time-series data on carbon and nitrogen mineralization rates includes:

[0012] In the target area where the topsoil volume expansion and regulation is to be carried out, soil monitoring points are selected according to the spatial distribution uniformity; topsoil samples with a topsoil depth of 0 to 20 cm are collected at each soil monitoring point, and the topsoil samples are sieved after removing stones and plant residues.

[0013] Under constant temperature and humidity conditions in the laboratory, quantitative standard organic materials were added to soil samples from each topsoil layer and soil mineralization was carried out in aerobic culture at constant temperature. Sampling times were set within the culture period, and the carbon or nitrogen release per unit time of each topsoil sample was measured using a gas analyzer or chemical titration. The carbon or nitrogen release per unit time was used as the carbon and nitrogen mineralization rate value at the corresponding sampling time, thus obtaining the original time series data of carbon and nitrogen mineralization rate for each topsoil sample.

[0014] Data cleaning was performed on the raw time series data of carbon and nitrogen mineralization rates to remove outliers caused by instrument malfunctions and impute missing values, thus obtaining the cleaned raw time series data of carbon and nitrogen mineralization rates.

[0015] Furthermore, the step of obtaining a phase hysteresis correction factor by coupling the temporal difference of the mineralization energy centroid with morphological overlap characteristics includes:

[0016] The mineralization energy centroid time data is obtained by calculating the centroid time of mineralization energy from the raw time series data of carbon and nitrogen mineralization rates.

[0017] Phase hysteresis correction factor was obtained by performing time axis virtual alignment and morphological overlap feature evaluation on the mineralization energy centroid time data and the original carbon and nitrogen mineralization rate time series data.

[0018] Furthermore, the step of calculating the centroid time of mineralization energy from the raw carbon and nitrogen mineralization rate time series data to obtain the centroid time data of mineralization energy includes:

[0019] For any target topsoil sample, extract the carbon and nitrogen mineralization rate values ​​and time data corresponding to each sampling time from the original time series data of the carbon and nitrogen mineralization rate of the target topsoil sample.

[0020] Multiply the time data corresponding to each sampling time with the carbon and nitrogen mineralization rate value corresponding to each sampling time, and sum the product calculation results corresponding to all sampling times to obtain the time-weighted mineralization accumulation value; sum the carbon and nitrogen mineralization rate values ​​corresponding to all sampling times of the target topsoil sample to obtain the mineralization accumulation value; use the time-weighted mineralization accumulation value as the numerator and the mineralization accumulation value as the denominator, and use the resulting fraction as the mineralization energy centroid time data corresponding to the target topsoil sample.

[0021] Furthermore, the step of obtaining a phase hysteresis correction factor by performing time axis virtual alignment and morphological overlap feature evaluation on the mineralization energy centroid time data and the original carbon and nitrogen mineralization rate time series data includes:

[0022] For any two topsoil samples to be compared, the first target topsoil sample and the second target topsoil sample are used to extract the first mineralization energy centroid time data from the mineralization energy centroid time data corresponding to the first target topsoil sample, and the second mineralization energy centroid time data is extracted from the mineralization energy centroid time data corresponding to the second target topsoil sample. The difference between the first mineralization energy centroid time data and the second mineralization energy centroid time data is calculated and the difference is rounded to obtain the time axis offset data.

[0023] The carbon and nitrogen mineralization rate values ​​of the second target topsoil sample at each sampling time are extracted from the original time series data of the carbon and nitrogen mineralization rate of the second target topsoil sample. Based on the time axis offset data, the original time series data of the carbon and nitrogen mineralization rate of the second target topsoil sample is subjected to time axis virtual translation to obtain the virtual aligned second sequence data. The gaps generated by the time axis virtual translation are filled with zero values.

[0024] The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are extracted from the original time series data of the carbon and nitrogen mineralization rate of the first target topsoil sample. The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are then compared with the values ​​of the second sequence data at the corresponding sampling time by taking the minimum value time by time to obtain the time-by-time morphological intersection amplitude data. The time-by-time morphological intersection amplitude data is then summed to obtain the cumulative morphological intersection value.

[0025] The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are summed to obtain the first mineralization accumulation value; the values ​​of the virtual aligned second sequence data at each sampling time are summed to obtain the second mineralization accumulation value; the larger of the first mineralization accumulation value and the second mineralization accumulation value is taken as the maximum envelope mineralization assessment value.

[0026] The cumulative value of morphological intersection is used as the numerator, the maximum envelope mineralization assessment value is used as the denominator, and the corresponding fraction is used as the morphological overlap ratio assessment. The morphological overlap ratio assessment is processed by square root operation to obtain the morphological overlap mapping amount. The difference between constant 1 and the morphological overlap mapping amount is added to the non-zero small constant to obtain the phase hysteresis correction factor corresponding to the first target topsoil sample and the second target topsoil sample.

[0027] Furthermore, the step of obtaining the transient excitation smoothing factor by normalizing and decaying the local oscillation energy functional of the sequence with the average oscillation energy of the whole sample includes:

[0028] Local oscillation energy functional data were obtained by coupling normalized second-order difference and step amplitude logarithmic gating processing on the raw time series data of carbon and nitrogen mineralization rate.

[0029] The average oscillation energy of the whole sample is obtained by performing full-sample mean statistical processing on the local oscillation energy functional data;

[0030] The transient excitation smoothing factor is obtained by performing geometric mean normalization and reciprocal decay mapping on the local oscillation energy functional data and the full sample average oscillation energy.

[0031] Furthermore, the process of obtaining local oscillation energy functional data by coupling normalized second-order difference and step amplitude logarithm gating to the original carbon and nitrogen mineralization rate time series data includes:

[0032] For any target topsoil sample, extract the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time from the original carbon and nitrogen mineralization rate time series data corresponding to the target topsoil sample; for any target intermediate sampling time of the target topsoil sample, extract the carbon and nitrogen mineralization rate values ​​corresponding to the previous sampling time, the target intermediate sampling time, and the next sampling time of the target intermediate sampling time.

[0033] The difference between the carbon and nitrogen mineralization rate value corresponding to the next sampling time and the carbon and nitrogen mineralization rate value corresponding to the previous sampling time is used to obtain the step amplitude difference assessment; the calculation result of adding constant 1 to the absolute value of the step amplitude difference assessment is used as the step amplitude gate base value, and the step amplitude gate base value is logarithmically processed to obtain the step amplitude logarithmic gate quantity.

[0034] Add the carbon and nitrogen mineralization rate value corresponding to the next sampling time to the carbon and nitrogen mineralization rate value corresponding to the previous sampling time, and subtract twice the carbon and nitrogen mineralization rate value corresponding to the target intermediate sampling time to obtain the second-order difference evaluation; use the absolute value of the second-order difference evaluation as the numerator, and the calculation result of adding the carbon and nitrogen mineralization rate value corresponding to the target intermediate sampling time to a non-zero infinitesimal constant as the denominator, and use the corresponding fraction as the normalized second-order difference evaluation;

[0035] The normalized second-order difference evaluation is multiplied by the step amplitude logarithmic gating quantity to obtain the local oscillation energy evaluation corresponding to the intermediate sampling time. The local oscillation energy evaluations corresponding to the target topsoil sample at all intermediate sampling times are accumulated to obtain the local oscillation energy functional data corresponding to the target topsoil sample.

[0036] Furthermore, the step of obtaining the full-sample average oscillation energy by performing full-sample mean statistical processing on the local oscillation energy functional data includes:

[0037] Obtain the local oscillation energy functional data corresponding to all topsoil samples participating in clustering calculation, and construct the full sample local oscillation energy functional set from the local oscillation energy functional data corresponding to the topsoil samples.

[0038] The data of each local oscillation energy functional in the full sample local oscillation energy functional set are summed to obtain the total oscillation energy value of the full sample.

[0039] The average oscillation energy of the entire sample is obtained by dividing the cumulative value of the oscillation energy of the entire sample by the number of soil samples in the cultivated layer of the local oscillation energy functional set of the entire sample.

[0040] Furthermore, the transient excitation smoothing factor is obtained by performing geometric mean normalization and reciprocal decay mapping on the local oscillation energy functional data and the full-sample average oscillation energy, including:

[0041] For any two topsoil samples to be compared, the first target topsoil sample and the second target topsoil sample are used to extract the first local oscillation energy functional data from the local oscillation energy functional data corresponding to the first target topsoil sample, the second local oscillation energy functional data are extracted from the local oscillation energy functional data corresponding to the second target topsoil sample, and the average oscillation energy data is extracted from the average oscillation energy of the whole sample.

[0042] The first local oscillation energy functional data and the second local oscillation energy functional data are multiplied to obtain the oscillation energy product evaluation; the oscillation energy product evaluation is then square-rooted to obtain the geometric mean oscillation energy evaluation.

[0043] The geometric mean oscillation energy assessment is used as the numerator, the average oscillation energy data is used as the denominator, and the corresponding fraction is used as the relative oscillation intensity assessment; the result of adding the relative oscillation intensity assessment to a constant is used as the attenuation denominator assessment.

[0044] Using constant 1 as the numerator and the attenuation denominator evaluation as the denominator, the resulting fraction is used as the transient excitation smoothing factor for the first target topsoil sample and the second target topsoil sample.

[0045] Furthermore, the step of obtaining the corrected distance and performing clustering iterative calculation by multiplicatively correcting the Euclidean distance using a phase hysteresis correction factor and a transient excitation smoothing factor includes:

[0046] Set the number of regulation type regions, and select multiple carbon and nitrogen mineralization rate time series original data corresponding to the number of regulation type regions from the carbon and nitrogen mineralization rate time series original data as the initial cluster center sequence;

[0047] For any target topsoil sample and any target cluster center sequence, the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time of the target topsoil sample are extracted from the original carbon and nitrogen mineralization rate time series data corresponding to the target topsoil sample. Similarly, the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time of the target cluster center sequence are extracted from the original carbon and nitrogen mineralization rate time series data corresponding to the target cluster center sequence. The difference between the carbon and nitrogen mineralization rate values ​​of the target topsoil sample at each sampling time and the carbon and nitrogen mineralization rate values ​​of the target cluster center sequence at the corresponding sampling time is calculated to obtain a time-by-time difference assessment. The time-by-time difference assessment is squared, and the squared results are summed to obtain the accumulated squared Euclidean distance. The accumulated squared Euclidean distance is then square-rooted to obtain the Euclidean distance between the target topsoil sample and the target cluster center sequence.

[0048] Obtain the phase hysteresis correction factor corresponding to the target topsoil sample and the target cluster center sequence, and obtain the transient excitation smoothing factor corresponding to the target topsoil sample and the target cluster center sequence; perform multiplicative correction processing on the phase hysteresis correction factor and the transient excitation smoothing factor and the Euclidean distance to obtain the corrected distance corresponding to the target topsoil sample and the target cluster center sequence.

[0049] The target topsoil samples are assigned to the cluster corresponding to the target cluster center sequence with the smallest corrected distance, and the cluster assignment is completed for all topsoil samples to obtain the sample assignment results. Based on the sample assignment results, the carbon and nitrogen mineralization rate values ​​of all topsoil samples in each cluster at each sampling time are arithmetically averaged to obtain the updated cluster center sequence for each cluster, and the updated cluster center sequence for each cluster is used as the new target cluster center sequence. The corrected distance calculation, sample cluster assignment, and cluster center sequence update are repeated until the change in the cluster center sequence is lower than the preset threshold or the preset number of iterations is reached to obtain the clustering results of the clustering iteration calculation.

[0050] Compared with the prior art, the present invention has the following advantages:

[0051] This invention discloses a method for expanding and regulating the topsoil volume by integrating an organic material carbon and nitrogen conversion model. By introducing a phase hysteresis correction mechanism based on mineralization energy centroid alignment and combining it with a nonlinear mapping of morphological overlap after alignment, phase misalignment is no longer misjudged as a fundamental difference. This eliminates the interference of inoculation hysteresis on curve similarity determination at the distance metric level, enabling soil samples with the same decomposition potential and similar fertilization trends to be stably merged into the same regulatory type region. This significantly reduces the misclassification rate of type regions and improves the biological consistency and interpretability of regulatory zoning. Furthermore, this technical solution constructs a local oscillation energy characterization highly sensitive to high-frequency oscillations and employs cross-sample adaptive normalization and reciprocal decay smoothing suppression to reduce the contribution weight of transient noise from excitation effects in distance calculation. This transforms the clustering results from being dominated by "oscillation amplitude differences" to being dominated by "steady-state transformation trends." Therefore, when matching carbon and nitrogen transformation model parameters based on this clustering result, a more stable template curve that better matches the intrinsic transformation kinetics of the soil can be obtained, thereby improving the robustness of model parameter fitting and the reliability of regulation inversion. Ultimately, more precise synergistic optimization can be achieved in decisions such as plot-level material input, nitrogen fertilizer application ratio, and tillage depth, reducing the failure of regulation schemes and waste of inputs due to misjudgment of soil type. Attached Figure Description

[0052] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0053] Figure 1 This is a flowchart of a method for regulating the expansion of the tillage layer by incorporating an organic material carbon and nitrogen conversion model, as described in an embodiment of the present invention. Detailed Implementation

[0054] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0055] See Figure 1 This is a flowchart of a method for regulating the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model, as provided in Embodiment 1 of the present invention. Figure 1 As shown, a method for regulating the expansion of the tillage layer by integrating an organic material carbon and nitrogen conversion model may include:

[0056] Step S1: Obtain raw time series data of carbon and nitrogen mineralization rate by collecting topsoil samples and measuring the carbon and nitrogen mineralization rate of organic materials.

[0057] First, in the target area where the topsoil volume expansion and regulation is to be carried out, soil monitoring points are selected based on the spatial uniformity of distribution. Soil samples with a topsoil depth of 0 to 20 cm are collected at each soil monitoring point. After removing stones and plant residues, the topsoil samples are sieved for later use.

[0058] Under constant temperature and humidity conditions in the laboratory, quantitative standard organic materials were added to soil samples from each topsoil layer and soil mineralization was carried out under constant temperature and aerobic conditions. Sampling times were set within the incubation period, and the carbon or nitrogen release per unit time of each topsoil sample was measured using a gas analyzer or chemical titration. The carbon or nitrogen release per unit time was used as the carbon and nitrogen mineralization rate value at the corresponding sampling time, thus obtaining the original time series data of carbon and nitrogen mineralization rate for each topsoil sample.

[0059] Data cleaning was performed on the raw time series data of carbon and nitrogen mineralization rates to remove outliers caused by instrument malfunctions and impute missing values, thus obtaining the cleaned raw time series data of carbon and nitrogen mineralization rates.

[0060] Thus, the original time series data of carbon and nitrogen mineralization rates were obtained by collecting topsoil samples and measuring the carbon and nitrogen mineralization rates of organic materials.

[0061] Step S2: The phase hysteresis correction factor is obtained by coupling the time difference of the mineralization energy centroid with the morphological overlap characteristics.

[0062] When analyzing the carbon and nitrogen transformation processes of soil organic matter, the first challenge is addressing the spurious discrepancies in data caused by the heterogeneity of microbial communities. In real-world topsoil expansion scenarios, soil samples from different sampling points exhibit significant differences in response time to newly introduced organic matter due to variations in physical texture (e.g., porosity differences between clay and sand) and initial microbial biomass. This biological phenomenon is known as the inoculation lag effect. Specifically, in some soil samples, microorganisms require a longer time to synthesize specific extracellular enzymes to adapt to the substrate after inoculation, resulting in a delayed peak in the carbon and nitrogen mineralization rate curves. Conversely, other highly active soil samples rapidly enter the logarithmic growth phase, with an earlier peak. This leads to two data curves representing the same decomposition potential (i.e., similar curve envelope shapes and similar total release amounts) exhibiting a significant numerical distance when using traditional Euclidean distance for point-to-point calculations simply because of their different phases on the time axis. This artificially inflated mathematical distance caused by biological lag can mislead clustering algorithms into classifying soils that essentially belong to the same regulatory type into different categories. To address this issue, direct comparison of the original time series is insufficient; instead, a correction mechanism capable of sensing and compensating for this temporal phase difference must be introduced. This requires first locating the energy centroid of each curve on the time axis, quantifying the degree of hysteresis by calculating the centroid offset, and then assessing the morphological overlap between the two curves based on a virtual alignment of the centroids. This approach constructs a phase hysteresis correction factor, aiming to allow the algorithm to ignore the timing of startup and focus on identifying whether the core kinetic patterns of soil organic carbon and nitrogen transformation are consistent.

[0063] In summary, this invention first obtains the mineralization energy centroid time data by processing the raw carbon and nitrogen mineralization rate time series data through mineralization energy centroid time calculation. Specifically, for any target topsoil sample, the carbon and nitrogen mineralization rate values ​​and time data corresponding to each sampling time of the target topsoil sample are extracted from the raw carbon and nitrogen mineralization rate time series data of the target topsoil sample. The time data corresponding to each sampling time is multiplied by the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time, and the product calculation results corresponding to all sampling times are summed to obtain the time-weighted mineralization accumulation value. The carbon and nitrogen mineralization rate values ​​corresponding to all sampling times of the target topsoil sample are summed to obtain the mineralization accumulation value. The time-weighted mineralization accumulation value is used as the numerator, the mineralization accumulation value is used as the denominator, and the resulting fraction is used as the mineralization energy centroid time data of the target topsoil sample.

[0064] After obtaining the mineralization energy centroid time data, further processing is performed on the mineralization energy centroid time data and the original carbon and nitrogen mineralization rate time series data. This involves virtual alignment of the time axis and evaluation of morphological overlap features to obtain the phase hysteresis correction factor. Specifically, for any two topsoil samples to be compared, the first target topsoil sample and the second target topsoil sample are used to extract the first mineralization energy centroid time data from the first target topsoil sample, and the second mineralization energy centroid is extracted from the second target topsoil sample. The time-series data is processed as follows: The time-series data of the first and second mineralization energy centroids are subtracted and rounded to obtain the time axis offset data. The carbon and nitrogen mineralization rate values ​​of the second target topsoil sample at each sampling time are extracted from the original time-series data of the carbon and nitrogen mineralization rate of the second target topsoil sample. Based on the time axis offset data, the original time-series data of the carbon and nitrogen mineralization rate of the second target topsoil sample is virtually shifted along the time axis to obtain the virtually aligned second sequence data. Gaps created by the virtual shift are filled with zero values. The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are extracted from the original time-series data of the first target topsoil sample. The minimum value of the carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time is then taken from the values ​​of the virtually aligned second sequence data at the corresponding sampling time to obtain the time-by-time morphological intersection amplitude data. The time-by-time morphological intersection amplitude data is then summed to obtain the cumulative morphological intersection value. The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are summed to obtain the first accumulated mineralization value. The values ​​of the second sequence data after virtual alignment at each sampling time are summed to obtain the second accumulated mineralization value. The larger of the first and second accumulated mineralization values ​​is used as the maximum envelope mineralization assessment value. The morphological intersection accumulation value is used as the numerator, and the maximum envelope mineralization assessment value is used as the denominator. The resulting fraction is used as the morphological overlap ratio assessment. The morphological overlap ratio assessment is processed by square root operation to obtain the morphological overlap mapping amount. The difference between the constant and the morphological overlap mapping amount is added to the non-zero small constant to obtain the phase hysteresis correction factor corresponding to the first and second target topsoil samples.

[0065] In one implementation, assume the first The topsoil layer in the first The carbon and nitrogen mineralization rate at each sampling time is ;No. The topsoil layer in the first The carbon and nitrogen mineralization rate at each sampling time is After virtual alignment, the first The first sequence data The carbon and nitrogen mineralization rate at each sampling time is The non-zero small constant used to prevent computational anomalies when there is complete overlap and to ensure the numerical stability of the distance metric is: Then the first The topsoil sample and the first The formula for calculating the phase hysteresis correction factor for each topsoil sample is as follows:

[0066]

[0067] in, Indicates the first The topsoil sample and the first Phase hysteresis correction factor for each topsoil sample; Indicates the first The topsoil layer in the first The carbon and nitrogen mineralization rate values ​​at each sampling time; Indicates the first after virtual alignment The first sequence data The carbon and nitrogen mineralization rate values ​​at each sampling time; This represents a non-zero, small constant used to prevent computational anomalies during complete overlap and to ensure the numerical stability of the distance metric. In this embodiment of the invention, it is set... ; This indicates the length of the time series of carbon and nitrogen mineralization rates for topsoil samples.

[0068] It should be noted that, firstly, because soil mineralization data is often accompanied by noise, simply searching for the maximum value is prone to local spikes and misalignment. By calculating the weighted average time, the average occurrence time of microbial decomposition activity in the soil sample can be robustly characterized using the data integration over the entire time period. This provides an accurate physical benchmark for subsequent elimination of hysteresis effects. Secondly, by calculating... This step achieves a measure of morphological compatibility after alignment. After eliminating time axis misalignment through centroid translation, if two soil samples are essentially of the same type, their mineralization rate curves should be geometrically highly overlapping. In this case, the minimum overlap area (numerator) at each time step will be extremely close to their maximum envelope area (denominator). This design based on the intersection and union ratio directly maps the true similarity in decomposition ability between the two samples after removing the time factor. Finally, the formula uses... The final mapping is performed in the form of [formula missing]. The ratio term reflects the degree of overlap; when two curves are similar in shape and have a high degree of overlap after alignment, the ratio approaches [value missing]. This causes the value of the phase hysteresis correction factor to approach its minimum value (close to...). The phase hysteresis correction factor appears as a multiplicative factor in the final distance formula. This means that as long as the two curves have the same morphological pattern, regardless of how many days their start times differ, the phase hysteresis correction factor will force them to be closer together. Conversely, if the two curves still have significantly different morphologies even after alignment (e.g., one is a single peak and the other a double peak), the ratio will be smaller, and the phase hysteresis correction factor will be closer to the expected value. This preserves the original distances. It eliminates the interference of the biological artifact of inoculation lag on the clustering results, ensuring that the algorithm classifies based on the intrinsic transformation model characteristics of the soil rather than accidental time phenomena.

[0069] Thus, the phase hysteresis correction factor was obtained by coupling the evaluation of the temporal difference and morphological overlap characteristics of the mineralization energy centroid.

[0070] Step S3: Obtain the transient excitation smoothing factor by normalizing and decaying the local oscillation energy functional of the sequence and the average oscillation energy of the whole sample.

[0071] After addressing the misjudgment issue caused by different start-up times through a phase hysteresis correction factor, the clustering algorithm can focus more on the overall morphological trend of soil organic carbon and nitrogen transformation. However, in actual monitoring data, in addition to the overall trend characteristics, there are also significant local fluctuations. This is mainly because the introduction of fresh organic materials into the soil triggers a temporary change in the decomposition rate of existing organic matter by soil microorganisms, i.e., a stimulating effect. The stimulating effect is often accompanied by intense microbial community turnover and short-term competition for nitrogen sources. Coupled with the amplification effect of minor changes in the field monitoring environment (such as local water migration), this results in a large number of high-frequency, non-steady-state sawtooth oscillations on the collected mineralization rate curve. When using traditional algorithms for calculation, Euclidean distance captures and accumulates the vertical deviation of each oscillation point in numerical value, causing two soil samples that only differ in the degree of oscillation but have the same long-term fertilization trend to be calculated as having a large distance. In fact, this transient disturbance caused by the stimulating effect does not represent the long-term fertility characteristics of the soil and belongs to biological noise that should be suppressed in the long-term regulation model of topsoil expansion. If these oscillations are not weighted, the clustering results will be dominated by these high-frequency fluctuations rather than by the core transformation trend. Therefore, based on step S2, it is necessary to further construct a factor that can identify and smooth this non-steady-state characteristic. For curves that exhibit high roughness due to strong excitation effects, the algorithm should reduce its sensitivity to distance calculations, thereby uncovering the true steady-state trend hidden behind the oscillations.

[0072] In summary, this invention first obtains local oscillation energy functional data by coupling normalized second-order difference and step amplitude logarithmic gating processing on the raw carbon and nitrogen mineralization rate time series data. Specifically, for any target topsoil sample, the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time are extracted from the raw carbon and nitrogen mineralization rate time series data corresponding to the target topsoil sample. For any target intermediate sampling time of the target topsoil sample, the carbon and nitrogen mineralization rate values ​​corresponding to the previous sampling time, the target intermediate sampling time, and the next sampling time are extracted. The difference between the carbon and nitrogen mineralization rate value corresponding to the next sampling time and the carbon and nitrogen mineralization rate value corresponding to the previous sampling time is obtained to obtain the step amplitude difference assessment. The result of adding the absolute value of the step amplitude difference assessment to the constant is used as the step amplitude gating basis value, and the step amplitude gating basis value is logarithmically processed to obtain the step amplitude logarithmic gating quantity. The carbon and nitrogen mineralization rate value at the next sampling time is added to the carbon and nitrogen mineralization rate value at the previous sampling time, and the difference is taken from twice the carbon and nitrogen mineralization rate value at the target intermediate sampling time to obtain the second-order difference evaluation. The absolute value of the second-order difference evaluation is used as the numerator, and the result of adding the carbon and nitrogen mineralization rate value at the target intermediate sampling time to a non-zero infinitesimal constant is used as the denominator. The resulting fraction is used as the normalized second-order difference evaluation. The normalized second-order difference evaluation is multiplied by the logarithmic gating quantity of the step amplitude to obtain the local oscillation energy evaluation at the intermediate sampling time. The local oscillation energy evaluations of the target topsoil sample at all intermediate sampling times are accumulated to obtain the local oscillation energy functional data of the target topsoil sample.

[0073] After obtaining the local oscillation energy functional data corresponding to the topsoil samples, the average oscillation energy of the whole sample is obtained by performing full-sample mean statistical processing on the local oscillation energy functional data. Specifically, the local oscillation energy functional data corresponding to all topsoil samples participating in the cluster calculation are obtained, and the local oscillation energy functional data corresponding to the topsoil samples are used to form a full-sample local oscillation energy functional set. The local oscillation energy functional data of each local oscillation energy functional set in the full-sample local oscillation energy functional set are summed to obtain the cumulative value of the full-sample oscillation energy. The cumulative value of the full-sample oscillation energy is divided by the number of topsoil samples in the full-sample local oscillation energy functional set to obtain the average oscillation energy of the whole sample.

[0074] After obtaining the average oscillation energy of the entire sample, the transient excitation smoothing factor is obtained by performing geometric mean normalization and reciprocal decay mapping on the local oscillation energy functional data and the average oscillation energy of the entire sample. Specifically, for any two topsoil samples to be compared, the first target topsoil sample and the second target topsoil sample are used to extract the first local oscillation energy functional data from the local oscillation energy functional data corresponding to the first target topsoil sample, the second local oscillation energy functional data from the local oscillation energy functional data corresponding to the second target topsoil sample, and the average oscillation energy data from the average oscillation energy of the entire sample. The first local oscillation energy functional data and the second local oscillation energy functional data are multiplied to obtain the oscillation energy product assessment. The oscillation energy product assessment is then square-rooted to obtain the geometric mean oscillation energy assessment. The geometric mean oscillation energy assessment is used as the numerator, and the average oscillation energy data is used as the denominator. The resulting fraction is used as the relative oscillation intensity assessment. The result of adding the relative oscillation intensity assessment to a constant is used as the attenuation denominator assessment. The constant is used as the numerator, and the attenuation denominator assessment is used as the denominator. The resulting fraction is used as the transient excitation smoothing factor for the first and second target topsoil samples.

[0075] In one implementation, assume the first The local oscillation energy functional data corresponding to each topsoil sample are: ;No. The local oscillation energy functional data corresponding to each topsoil sample are: The average oscillation energy of the entire sample is Then the first The topsoil sample and the first The formula for calculating the transient excitation smoothing factor for each topsoil sample is as follows:

[0076]

[0077] in, Indicates the first The topsoil sample and the first Transient excitation smoothing factor corresponding to each topsoil sample; Indicates the first Local oscillation energy functional data corresponding to each topsoil sample; Indicates the first Local oscillation energy functional data corresponding to each topsoil sample; This represents the average oscillation energy of the entire sample.

[0078] It should be noted that, firstly, in The calculation employed a second-order difference operator for discrete data. Second-order differences are insensitive to smooth trends but extremely sensitive to high-frequency sawtooth fluctuations. Normalization by dividing by the current value eliminates the influence of absolute numerical magnitude, allowing the algorithm to fairly evaluate the relative oscillation levels between high- and low-mineralized areas. This ensures that the extracted features purely reflect the instability caused by the excitation effect, rather than the level of soil fertility. Secondly, a logarithmic gating term is introduced to nonlinearly amplify the impact of significant oscillations. For minor instrument errors or background noise, this term is extremely small, thus avoiding misjudgment; however, for significant jumps caused by the excitation effect, this term significantly increases. Value. This design allows the factor to precisely pinpoint biologically significant non-steady-state perturbations. Finally, The decay function is in reciprocal form. When comparing the oscillation energy of two samples... When both are relatively large (i.e., both are subject to strong excitation effects), the denominator contains The item will increase significantly, leading to The value decreases rapidly. In the final distance calculation, As a weighting coefficient, it compresses the distance values ​​between these highly oscillating samples. The corresponding logic is that since both samples are in a highly unstable excitation state, the huge numerical differences between the two samples caused by oscillations are unreliable. The algorithm forcibly narrows the distance between the two samples so that they can be classified into the same type of highly active soil for unified steady-state regulation, thereby improving the robustness of the algorithm in complex field environments.

[0079] Thus, the transient excitation smoothing factor was obtained by normalizing and decaying the local oscillation energy functional of the sequence with the average oscillation energy of the whole sample.

[0080] Step S4: The corrected distance is obtained by multiplicatively correcting the Euclidean distance with the phase hysteresis correction factor and the transient excitation smoothing factor, and then performing clustering iterative calculation.

[0081] This step embeds the calculated phase hysteresis correction factor and transient excitation smoothing factor into a traditional clustering algorithm to construct a hybrid distance metric and perform clustering. Specifically, the number of control type regions is set, and multiple original carbon and nitrogen mineralization rate time series data corresponding to the number of control type regions are selected from the original carbon and nitrogen mineralization rate time series data as the initial cluster center sequence. In this embodiment of the invention, the number of control type regions is set to 5. This number can be adjusted according to the actual scenario and is not required.

[0082] For any target topsoil sample and any target cluster center sequence, extract the carbon and nitrogen mineralization rate values ​​of the target topsoil sample at each sampling time from the original carbon and nitrogen mineralization rate time series data corresponding to the target topsoil sample, and extract the carbon and nitrogen mineralization rate values ​​of the target cluster center sequence at each sampling time from the original carbon and nitrogen mineralization rate time series data corresponding to the target cluster center sequence. Subtract the carbon and nitrogen mineralization rate values ​​of the target topsoil sample at each sampling time from the carbon and nitrogen mineralization rate values ​​of the target cluster center sequence at the corresponding sampling time to obtain a time-by-time difference assessment. Square the time-by-time difference assessment and sum the squared results to obtain the accumulated squared Euclidean distance. Perform square root operations on the accumulated squared Euclidean distance to obtain the Euclidean distance between the target topsoil sample and the target cluster center sequence.

[0083] Obtain the phase hysteresis correction factor corresponding to the target topsoil sample and the target cluster center sequence, and obtain the transient excitation smoothing factor corresponding to the target topsoil sample and the target cluster center sequence; perform multiplicative correction processing on the phase hysteresis correction factor and the transient excitation smoothing factor and the Euclidean distance to obtain the corrected distance corresponding to the target topsoil sample and the target cluster center sequence.

[0084] The target topsoil samples are assigned to the cluster corresponding to the target cluster center sequence with the smallest corrected distance, and the cluster assignment is completed for all topsoil samples to obtain the sample assignment results. Based on the sample assignment results, the carbon and nitrogen mineralization rate values ​​of all topsoil samples in each cluster at each sampling time are arithmetically averaged to obtain the updated cluster center sequence for each cluster, and the updated cluster center sequence for each cluster is used as the new target cluster center sequence. The corrected distance calculation, sample cluster assignment, and cluster center sequence update are repeated until the change in the cluster center sequence is lower than the preset threshold or the preset number of iterations is reached to obtain the clustering results of the clustering iteration calculation.

[0085] Thus, the corrected distance was obtained by multiplicatively correcting the Euclidean distance using a phase hysteresis correction factor and a transient excitation smoothing factor, and then performing clustering iterative calculations.

[0086] Step S5: By matching the parameters of the carbon-nitrogen conversion model and back-deriving the control strategy based on the clustering results, a control scheme for expanding the topsoil is obtained.

[0087] Using the clustering results output in step S4, the cultivated land within the target area is divided into several regulation type zones. For each type zone, its cluster center curve is selected as the standard template for the decomposition of soil organic matter in that area, and the corresponding kinetic parameters are matched (in this embodiment, the decomposition constant in the first-order kinetic equation is selected). and potential mineralization potential Based on a well-matched carbon-nitrogen conversion model and considering the target soil fertility improvement needs of the region, the optimal input amount of organic matter, nitrogen fertilizer application ratio, and tillage depth are derived in reverse. Finally, the system outputs specific agricultural operation suggestions for each plot, thereby achieving precise and coordinated improvement of topsoil thickness and soil fertility while eliminating interference from delayed microbial inoculation and stimulating effects. This solves the problem of existing technologies failing to control schemes due to misjudgment of soil characteristics.

[0088] Thus, the process of matching carbon and nitrogen conversion model parameters and back-deriving control strategies based on clustering results has been completed, resulting in a scheme for expanding and controlling the topsoil.

[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for regulating the expansion of the tillage layer by integrating a carbon and nitrogen conversion model of organic materials, characterized in that, The method includes: Step S1: Obtain raw time series data of carbon and nitrogen mineralization rate by collecting topsoil samples and measuring the carbon and nitrogen mineralization rate of organic materials; Step S2: Obtain the phase hysteresis correction factor by coupling the time difference of the mineralization energy centroid with the morphological overlap characteristics; Step S3: Obtain the transient excitation smoothing factor by normalizing and decaying the local oscillation energy functional of the sequence and the average oscillation energy of the whole sample; Step S4: Obtain the corrected distance by multiplicatively correcting the Euclidean distance with the phase hysteresis correction factor and the transient excitation smoothing factor, and then perform clustering iterative calculation. Step S5: Obtain the topsoil expansion and control scheme by matching carbon and nitrogen conversion model parameters and back-deriving control strategies based on the clustering results; The process of collecting topsoil samples and measuring the carbon and nitrogen mineralization rates of organic materials to obtain raw time-series data of carbon and nitrogen mineralization rates includes: selecting soil monitoring points based on spatial distribution uniformity within the target area for topsoil expansion and regulation; collecting topsoil samples from a depth of 0 to 20 cm at each monitoring point, removing stones and plant residues from the topsoil samples, and sieving them for later use; adding quantitative standard organic materials to each topsoil sample under constant temperature and humidity conditions in a laboratory and conducting aerobic incubation for soil mineralization; setting sampling times within the incubation period, measuring the carbon or nitrogen release per unit time of each topsoil sample using a gas analyzer or chemical titration, and using the carbon or nitrogen release per unit time as the carbon and nitrogen mineralization rate value at the corresponding sampling time to obtain raw time-series data of carbon and nitrogen mineralization rates for each topsoil sample; cleaning the raw time-series data of carbon and nitrogen mineralization rates, removing outliers caused by instrument malfunctions, and imputing missing values ​​to obtain cleaned raw time-series data of carbon and nitrogen mineralization rates. The step of obtaining a phase hysteresis correction factor by coupling and evaluating the difference in the time of mineralization energy centroid and the morphological overlap characteristics includes: calculating the time of mineralization energy centroid from the original carbon and nitrogen mineralization rate time series data to obtain mineralization energy centroid time data; and evaluating the time axis virtual alignment and morphological overlap characteristics of the mineralization energy centroid time data and the original carbon and nitrogen mineralization rate time series data to obtain a phase hysteresis correction factor.

2. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 1, characterized in that, The process of calculating the centroid time of mineralization energy from the raw time series data of carbon and nitrogen mineralization rates to obtain the centroid time data of mineralization energy includes: For any target topsoil sample, extract the carbon and nitrogen mineralization rate values ​​and time data corresponding to each sampling time from the original time series data of the carbon and nitrogen mineralization rate of the target topsoil sample. Multiply the time data corresponding to each sampling time with the carbon and nitrogen mineralization rate value corresponding to each sampling time, and sum the product calculation results corresponding to all sampling times to obtain the time-weighted mineralization accumulation value; sum the carbon and nitrogen mineralization rate values ​​corresponding to all sampling times of the target topsoil sample to obtain the mineralization accumulation value; use the time-weighted mineralization accumulation value as the numerator and the mineralization accumulation value as the denominator, and use the resulting fraction as the mineralization energy centroid time data corresponding to the target topsoil sample.

3. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 1, characterized in that, The process involves performing time axis virtual alignment and morphological overlap feature evaluation on the mineralization energy centroid time data and the original carbon and nitrogen mineralization rate time series data to obtain a phase hysteresis correction factor, including: For any two topsoil samples to be compared, the first target topsoil sample and the second target topsoil sample are used to extract the first mineralization energy centroid time data from the mineralization energy centroid time data corresponding to the first target topsoil sample, and the second mineralization energy centroid time data is extracted from the mineralization energy centroid time data corresponding to the second target topsoil sample. The difference between the first mineralization energy centroid time data and the second mineralization energy centroid time data is calculated and the difference is rounded to obtain the time axis offset data. The carbon and nitrogen mineralization rate values ​​of the second target topsoil sample at each sampling time are extracted from the original time series data of the carbon and nitrogen mineralization rate of the second target topsoil sample. Based on the time axis offset data, the original time series data of the carbon and nitrogen mineralization rate of the second target topsoil sample is subjected to time axis virtual translation to obtain the virtual aligned second sequence data. The gaps generated by the time axis virtual translation are filled with zero values. The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are extracted from the original time series data of the carbon and nitrogen mineralization rate of the first target topsoil sample. The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are then compared with the values ​​of the second sequence data at the corresponding sampling time by taking the minimum value time by time to obtain the time-by-time morphological intersection amplitude data. The time-by-time morphological intersection amplitude data is then summed to obtain the cumulative morphological intersection value. The carbon and nitrogen mineralization rate values ​​of the first target topsoil sample at each sampling time are summed to obtain the first mineralization accumulation value; the values ​​of the virtual aligned second sequence data at each sampling time are summed to obtain the second mineralization accumulation value; the larger of the first mineralization accumulation value and the second mineralization accumulation value is taken as the maximum envelope mineralization assessment value. The cumulative value of morphological intersection is used as the numerator, the maximum envelope mineralization assessment value is used as the denominator, and the corresponding fraction is used as the morphological overlap ratio assessment. The morphological overlap ratio assessment is processed by square root operation to obtain the morphological overlap mapping amount. The difference between constant 1 and the morphological overlap mapping amount is added to the non-zero small constant to obtain the phase hysteresis correction factor corresponding to the first target topsoil sample and the second target topsoil sample.

4. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 1, characterized in that, The method of obtaining the transient excitation smoothing factor by normalizing and decaying the local oscillation energy functional of the sequence with the average oscillation energy of the whole sample includes: Local oscillation energy functional data were obtained by coupling normalized second-order difference and step amplitude logarithmic gating processing on the raw time series data of carbon and nitrogen mineralization rate. The average oscillation energy of the whole sample is obtained by performing full-sample mean statistical processing on the local oscillation energy functional data. The transient excitation smoothing factor is obtained by performing geometric mean normalization and reciprocal decay mapping on the local oscillation energy functional data and the full sample average oscillation energy.

5. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 4, characterized in that, The process involves normalizing the raw carbon and nitrogen mineralization rate time series data, performing second-order difference normalization, and coupled with logarithmic gating of the stride amplitude to obtain local oscillation energy functional data, including: For any target topsoil sample, extract the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time from the original carbon and nitrogen mineralization rate time series data corresponding to the target topsoil sample; for any target intermediate sampling time of the target topsoil sample, extract the carbon and nitrogen mineralization rate values ​​corresponding to the previous sampling time, the target intermediate sampling time, and the next sampling time of the target intermediate sampling time. The difference between the carbon and nitrogen mineralization rate value corresponding to the next sampling time and the carbon and nitrogen mineralization rate value corresponding to the previous sampling time is used to obtain the step amplitude difference assessment; the calculation result of adding constant 1 to the absolute value of the step amplitude difference assessment is used as the step amplitude gate base value, and the step amplitude gate base value is logarithmically processed to obtain the step amplitude logarithmic gate quantity. Add the carbon and nitrogen mineralization rate value corresponding to the next sampling time to the carbon and nitrogen mineralization rate value corresponding to the previous sampling time, and subtract twice the carbon and nitrogen mineralization rate value corresponding to the target intermediate sampling time to obtain the second-order difference evaluation; use the absolute value of the second-order difference evaluation as the numerator, and the calculation result of adding the carbon and nitrogen mineralization rate value corresponding to the target intermediate sampling time to a non-zero infinitesimal constant as the denominator, and use the corresponding fraction as the normalized second-order difference evaluation; The normalized second-order difference evaluation is multiplied by the step amplitude logarithmic gating quantity to obtain the local oscillation energy evaluation corresponding to the intermediate sampling time. The local oscillation energy evaluations corresponding to the target topsoil sample at all intermediate sampling times are accumulated to obtain the local oscillation energy functional data corresponding to the target topsoil sample.

6. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 4, characterized in that, The step of obtaining the average oscillation energy of the entire sample by performing full-sample mean statistical processing on the local oscillation energy functional data includes: Obtain the local oscillation energy functional data corresponding to all topsoil samples participating in clustering calculation, and construct the full sample local oscillation energy functional set from the local oscillation energy functional data corresponding to the topsoil samples. The data of each local oscillation energy functional in the full sample local oscillation energy functional set are summed to obtain the total sample oscillation energy value. The average oscillation energy of the entire sample is obtained by dividing the cumulative value of the oscillation energy of the entire sample by the number of soil samples in the cultivated layer of the local oscillation energy functional set of the entire sample.

7. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 4, characterized in that, The transient excitation smoothing factor is obtained by performing geometric mean normalization and reciprocal decay mapping on the local oscillation energy functional data and the full sample average oscillation energy, including: For any two topsoil samples to be compared, the first target topsoil sample and the second target topsoil sample are used to extract the first local oscillation energy functional data from the local oscillation energy functional data corresponding to the first target topsoil sample, the second local oscillation energy functional data are extracted from the local oscillation energy functional data corresponding to the second target topsoil sample, and the average oscillation energy data is extracted from the average oscillation energy of the whole sample. The first local oscillation energy functional data and the second local oscillation energy functional data are multiplied to obtain the oscillation energy product evaluation; the oscillation energy product evaluation is then square-rooted to obtain the geometric mean oscillation energy evaluation. The geometric mean oscillation energy assessment is used as the numerator, the average oscillation energy data is used as the denominator, and the corresponding fraction is used as the relative oscillation intensity assessment; the result of adding the relative oscillation intensity assessment to a constant is used as the attenuation denominator assessment. Using constant 1 as the numerator and the attenuation denominator evaluation as the denominator, the resulting fraction is used as the transient excitation smoothing factor for the first target topsoil sample and the second target topsoil sample.

8. The method for controlling the expansion of the tillage layer based on an organic material carbon and nitrogen conversion model according to claim 1, characterized in that, The process involves multiplicatively correcting the Euclidean distance using a phase hysteresis correction factor and a transient excitation smoothing factor, obtaining the corrected distance, and performing iterative clustering calculations, including: Set the number of regulation type regions, and select multiple carbon and nitrogen mineralization rate time series original data corresponding to the number of regulation type regions from the carbon and nitrogen mineralization rate time series original data as the initial cluster center sequence; For any target topsoil sample and any target cluster center sequence, the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time of the target topsoil sample are extracted from the original carbon and nitrogen mineralization rate time series data corresponding to the target topsoil sample. Similarly, the carbon and nitrogen mineralization rate values ​​corresponding to each sampling time of the target cluster center sequence are extracted from the original carbon and nitrogen mineralization rate time series data corresponding to the target cluster center sequence. The difference between the carbon and nitrogen mineralization rate values ​​of the target topsoil sample at each sampling time and the carbon and nitrogen mineralization rate values ​​of the target cluster center sequence at the corresponding sampling time is calculated to obtain a time-by-time difference assessment. The time-by-time difference assessment is squared, and the squared results are summed to obtain the accumulated squared Euclidean distance. The accumulated squared Euclidean distance is then square-rooted to obtain the Euclidean distance between the target topsoil sample and the target cluster center sequence. Obtain the phase hysteresis correction factor corresponding to the target topsoil sample and the target cluster center sequence, and obtain the transient excitation smoothing factor corresponding to the target topsoil sample and the target cluster center sequence; perform multiplicative correction processing on the phase hysteresis correction factor and the transient excitation smoothing factor and the Euclidean distance to obtain the corrected distance corresponding to the target topsoil sample and the target cluster center sequence. The target topsoil samples are assigned to the cluster corresponding to the target cluster center sequence with the smallest corrected distance, and the cluster assignment is completed for all topsoil samples to obtain the sample assignment results. Based on the sample assignment results, the carbon and nitrogen mineralization rate values ​​of all topsoil samples in each cluster at each sampling time are arithmetically averaged to obtain the updated cluster center sequence for each cluster, and the updated cluster center sequence for each cluster is used as the new target cluster center sequence. The corrected distance calculation, sample cluster assignment, and cluster center sequence update are repeated until the change in the cluster center sequence is lower than the preset threshold or the preset number of iterations is reached to obtain the clustering results of the clustering iteration calculation.

Images