Wolf tail grass biomass fermentation carbon sequestration benefit evaluation method and system

Abstract

The application discloses a method and system for evaluating the carbon fixation benefit of Pennisetum biomass fermentation, and particularly relates to the field of agricultural biomass fermentation carbon fixation evaluation, which comprises the following steps: obtaining the fermentation carbon records of the Pennisetum agricultural fermentation site edge computing nodes formed according to a uniform sampling time sequence; dividing the material exchange position into a non-material exchange section; determining the temporary carbon amount according to the inorganic carbon amount of the fermentation liquid and the gas phase carbon amount in the tank; and combining the outlet carbon discharge amount to form a fermentation carbon sequence; for the non-material exchange section, constructing a forward block Hankel matrix and a backward block Hankel matrix according to the sampling position; and through generalized Schur decomposition, the matrix pencil method is used to solve the migration eigenvalue and the modal coefficient; the non-material exchange section is divided by the edge computing node, the matrix pencil method is used to dynamically identify the carbon migration time lag, the outlet carbon discharge amount is back matched to the actual formation position by combining the Schroedinger bridge, and accordingly, the dynamic carbon conversion sequence is formed and the agricultural fermentation carbon fixation performance analysis is completed.

Description

Technical Field

[0001] This invention relates to the field of agricultural biomass fermentation carbon sequestration assessment technology, and more specifically, to a method and system for assessing the carbon sequestration benefits of Napier grass biomass fermentation. Background Technology

[0002] Existing assessments of the carbon sequestration benefits of agricultural biomass fermentation, such as Napier grass, are mostly based on the amount of raw material input, the carbon content of raw materials, the carbon content of fermentation products, and the amount of carbon-containing gas emissions. They calculate the carbon input, carbon output, and carbon retention before and after fermentation, and analyze the carbon conversion at each fermentation stage by combining operational data such as temperature, liquid level, gas flow rate, and gas composition. During large-scale closed fermentation, biomass carbon is not immediately discharged from the gas outlet after being transformed by microorganisms. Some of it first enters the fermentation liquid to form dissolved inorganic carbon, and some enters the gas phase space inside the tank and is then released during the exhaust process. Furthermore, changes in stirring intensity, liquid level, temperature, and exhaust state can alter the time process of carbon migration from the liquid phase to the gas phase and its discharge from inside the tank. This leads to the following phenomenon: even after feeding, discharging and sampling have stopped at a certain time, carbon-containing gases are still continuously detected at the gas outlet, and some emission peaks will be later than the corresponding bioconversion stage. The existing method of directly collecting carbon emissions based on sampling time or exhaust time will include carbon that has already formed but is delayed in emission due to dissolution and temporary storage in the subsequent fermentation stage, resulting in a misalignment between the carbon conversion amount in each stage and the actual formation time. Even if the sampling frequency or gas detection points are increased, only more concentrated emission data can be obtained, and it is still impossible to determine which fermentation stage the delayed emission carbon should belong to. The technical problem to be solved by this application is: how to accurately determine the carbon conversion amount and carbon sequestration efficiency of each fermentation stage by corresponding the carbon content that is delayed due to gas-liquid migration and temporary storage in the tank to its actual formation stage based on the continuous operation data of the fermentation process of Napier grass, so as to accurately determine the carbon conversion amount and carbon sequestration efficiency of Napier grass biomass fermentation in each fermentation stage. Summary of the Invention

[0003] To overcome the aforementioned deficiencies in the prior art, embodiments of the present invention provide a method and system for evaluating the carbon sequestration benefits of Napier grass biomass fermentation. This method divides non-material exchange sections using edge computing nodes, dynamically identifies carbon migration lags using the matrix pencil method, and combines the Schödinger bridge to reallocate the outlet carbon emissions back to the actual formation location. Based on this, a dynamic carbon conversion sequence is formed, and the performance analysis of agricultural fermentation carbon sequestration is completed, thereby solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation, comprising: S1. Obtain the fermentation carbon record formed by the edge computing node of the agricultural fermentation site of Napier grass according to the unified sampling time sequence, divide the non-material exchange section according to the material exchange location, determine the temporary carbon amount based on the inorganic carbon amount of fermentation liquid and the gas phase carbon amount in the tank, and form the fermentation carbon sequence by combining the outlet carbon discharge amount. S2. For sections without material exchange, construct forward block Hankel matrix and backward block Hankel matrix according to sampling position. Use generalized Schul decomposition to perform matrix pencil method to obtain migration eigenvalues ​​and modal coefficients, reconstruct carbon migration amount at position intervals, complete dynamic carbon migration modeling and form carbon migration weights. S3. Based on carbon migration weights, a Scholdinger bridge prior matrix is ​​established. The initial temporary carbon amount and the newly formed carbon amount at each position are used as the source quality, and the outlet carbon emission amount and the temporary carbon amount at the end of the section are used as the target quality. The matrix elements are written in chronological order and row and column scaling is performed alternately until the row and column quality is conserved, thus forming a carbon migration matrix. S4. Based on the carbon migration matrix, the carbon emission at the outlet is allocated to the corresponding formation position, and the temporary carbon amount at the end of the section is written as the initial temporary carbon amount of the next section without material exchange. The matrix pencil method and Scholdinger bridge operation are repeated sequentially by the edge computing nodes to form a dynamic carbon conversion sequence. S5. Based on the dynamic carbon conversion sequence, the fermentation retention carbon amount is obtained by subtracting the cumulative export carbon emission amount assigned to the completed formation location from the initial carbon amount of the Napier grass agricultural batch. The ratio of the fermentation retention carbon amount to the initial carbon amount is used to perform agricultural fermentation carbon sequestration performance analysis, and the carbon sequestration benefit assessment results are output.

[0005] In a preferred embodiment, S1 includes: S11. Obtain the agricultural batch identifier of Napier grass and the material exchange time of fermentation container recorded by the edge computing node. Form a sampling interval with adjacent sampling times, write the material exchange that occurs in the interval to the end sampling position of the interval, and connect each sampling position according to the agricultural batch identifier to form a batch sampling sequence. The Napier grass agricultural batch identifier is a fixed identifier used to uniquely correspond to the source of Napier grass agricultural raw material input. It is generated by the source plot of Napier grass, the harvesting batch and the input batch in a fixed field order, and remains unchanged after the same batch of Napier grass enters the fermentation process. It is used to associate all sampling locations and carbon records corresponding to the batch of Napier grass. The material exchange moment in the fermentation container refers to the time and position at which the raw material of Napier grass, fermentation liquid or fermentation material actually crosses the boundary of the fermentation container and causes a change in the amount of material in the container. It is determined by the moment when the edge computing node receives the corresponding metering device to form an effective material increase or decrease record. S12. Based on the batch sampling sequence, the inorganic carbon content of the fermentation broth is obtained by multiplying the inorganic carbon concentration of the fermentation broth by the liquid phase volume, and the carbon content of the gas phase in the tank is obtained by multiplying the carbon concentration of the gas phase in the tank by the gas phase volume. The carbon discharge at the outlet is obtained by accumulating the product of the outlet gas flow rate and the carbon concentration in the sampling interval, thus forming a carbon record of the sampling location. The inorganic carbon concentration in the fermentation broth refers to the mass of inorganic carbon per unit volume of fermentation broth at the corresponding sampling location at the given time. It is output by the inorganic carbon detection equipment in a uniform carbon mass unit and is used to calculate the actual amount of inorganic carbon remaining in the fermentation broth at that sampling location. Liquid phase volume refers to the effective volume occupied by the fermentation liquid in the fermentation container at the corresponding sampling location and time. It is calculated from the liquid level detection value according to the pre-calibrated relationship between the liquid level and the effective liquid phase volume of the fermentation container. The carbon concentration in the gas phase inside the tank refers to the mass of carbon contained in a unit volume of gas in the gas phase space of the fermentation vessel at the corresponding sampling location at the time of sampling. It is obtained by converting the gas component detection results to a unified volume standard according to the temperature and pressure at the time of detection. The gas phase volume refers to the effective internal volume of the fermentation vessel that is not occupied by the liquid phase at the corresponding sampling location. It is obtained by subtracting the liquid phase volume at the same sampling location from the effective internal volume of the fermentation vessel. The outlet gas flow rate refers to the actual volumetric flow rate of gas discharged through the gas outlet of the fermentation vessel within the sampling interval. It is collected by the outlet flow meter and converted according to the temperature, pressure and volume benchmark consistent with the carbon concentration of the gas phase inside the tank. S13. Based on the carbon record of the sampling location, the sampling locations between adjacent material exchange locations form a non-material exchange section. The inorganic carbon content of the fermentation liquid is added to the gas phase carbon content in the tank to obtain the temporary carbon content. Then, the temporary carbon content is connected with the outlet carbon content according to the sampling position to form the fermentation carbon sequence of the corresponding Napier grass agricultural batch.

[0006] In a preferred embodiment, S2 includes: S21. For each section without material exchange, convert the temporary carbon amount and the outlet carbon emission amount into an integer sequence according to the carbon metering resolution unit of the edge computing node. Construct the forward block Hankel matrix and the backward block Hankel matrix according to the sampling position. Perform Bareiss elimination on the forward block Hankel matrix. In each round, read the first non-zero element in column order as the pivot. Use the current pivot and the pivot of the previous round to cross-multiply and divide to eliminate the remaining elements in the pivot column. Determine the matrix bundle order according to the number of non-zero pivots. Record the pivot row number and pivot column number that are consistent with the number of matrix bundle orders according to the pivot formation order to form the matrix extraction position number. The carbon measurement resolution unit refers to the fixed discrete unit used to represent carbon quality in the same batch of Napier grass. After converting the resolution of each metering device participating in the S2 calculation to the same carbon quality dimension, a quality unit that can divide all resolutions is selected so that the amount of temporary carbon and the amount of carbon emitted at the outlet are both integers after being divided by the carbon measurement resolution unit. The process of constructing the forward and backward block Hankel matrices is as follows: the first to the last sampling positions of the non-material exchange section are numbered sequentially, and the temporary carbon storage and outlet carbon emission of each sampling position are combined into a two-dimensional column vector; the number of sampling positions is denoted as N, the integer quotient of N divided by two is taken as the number of blocks R, and N minus R is taken as the number of blocks C. The i-th row and j-th column of the forward block Hankel matrix is ​​written into the two-dimensional column vector of the i-th plus j minus one sampling position, and the same position of the backward block Hankel matrix is ​​written into the two-dimensional column vector of the i-th plus j sampling position, where i increases from one to R and j increases from one to C. This makes the backward block Hankel matrix uniformly shifted forward by one sampling position relative to the forward block Hankel matrix. The Bareiss elimination process is as follows: Starting from the first column on the left of the matrix to be eliminated, read the first non-zero element in the current processing row and below it in ascending order of row number and swap it to the current processing row as the pivot; for each element to be updated in the lower right region of the pivot, multiply the element to be updated by the current pivot, and then subtract the product of the pivot column element of the row to be updated and the element in the same column of the row to be updated. The first round directly writes the subtraction result, and from the second round onwards, divide by the pivot of the previous round; after completing the row and column corresponding to the current pivot, move to the next column. When all elements in the current column below the current processing row are zero, jump to the next column, until the remaining columns cannot form non-zero pivots; Column order refers to the order in which the column numbers of a matrix are read from left to right in ascending order. When the pivot element needs to be found in the same column, it is read from top to bottom in ascending order of row numbers. Therefore, each elimination position has a unique pivot element selection order. An element refers to a single integer value that occupies a specific row and column number after the amount of temporary carbon and the amount of carbon emitted from exports are converted to integers according to the carbon measurement units and expanded into a block Hankel matrix. It does not refer to the entire two-dimensional column vector. S22. Extract the index according to the matrix. Read the elements with the same row number and the same column number from the forward block Hankel matrix and the backward block Hankel matrix to form forward and backward shift matrices with order equal to the order of the matrix bundle, respectively. Perform generalized Shur decomposition on the two matrices and transform them into upper triangular matrices simultaneously to form the migration eigenvalues ​​by corresponding to the ratio of the diagonal elements. The process of performing generalized Shure decomposition and synchronously transforming into upper triangular matrices is as follows: A matrix pair is formed by shifting the matrix forward and shifting it backward. A common left orthogonal transformation and a right orthogonal transformation are then performed on both matrices sequentially, ensuring that each transformation simultaneously affects the corresponding rows and columns of both matrices, until the forward-shifted upper triangular matrix and the backward-shifted upper triangular matrix are obtained. The diagonal elements of the two upper triangular matrices are read according to their diagonal indices. The diagonal elements of the backward-shifted upper triangular matrix are divided by the diagonal elements of the forward-shifted upper triangular matrix to form migration eigenvalues. Diagonal positions with a denominator of zero are not written into the finite set of migration eigenvalues.

[0007] In a preferred embodiment, S2 further includes: S23. Construct a modal basis matrix by writing the integer powers of the migration eigenvalues ​​in ascending order of the sampling interval. Perform Haushold orthogonal triangular decomposition on the modal basis matrix and substitute the temporary carbon amount and the outlet carbon emission amount into the triangular matrix in reverse order to obtain the modal coefficients. Then reconstruct the carbon migration amount by sampling interval according to the migration eigenvalues ​​and modal coefficients. The process of performing Haushold orthogonal triangular decomposition is as follows: starting from the first column of the modal basis matrix, read the column vector from the diagonal position downwards in the current column, construct the Haushold reflection vector according to the Euclidean length of the column vector, multiply the current column and the subsequent columns by the reflection matrix on the left, so that the elements below the diagonal position of the current column become zero, and then repeat the process according to the column number increment until the modal basis matrix is ​​transformed into an upper triangular matrix. At the same time, multiply the reflection matrices in the order of execution to form an orthogonal matrix. The process of obtaining the modal coefficients is as follows: the amount of temporary carbon and the amount of carbon emitted at the outlet of each sampling position in the non-material exchange section are arranged into an observation matrix according to the sampling position. The orthogonal matrix obtained by Haushold decomposition is transposed and multiplied by the observation matrix. Then, starting from the last row of the upper triangular matrix, the matrix is ​​substituted row by row upwards and backwards to obtain the modal coefficients of the amount of temporary carbon and the amount of carbon emitted at the outlet corresponding to each migration feature value, thus forming a modal coefficient matrix. The process of reconstructing carbon migration is as follows: taking the sampling interval zero as the starting interval, calculate the corresponding integer power of each migration feature value for each sampling interval, multiply it by the mode coefficient corresponding to the carbon emission at the outlet, and then sum them up; when the migration feature values ​​appear in conjugate pairs, the conjugate terms are combined into a real number result, and the real number results of each sampling interval are arranged in ascending order of the sampling interval to form a carbon migration sequence. S24. Divide the non-negative carbon migration amount of each sampling interval by the sum of the non-negative carbon migration amounts of all sampling intervals to obtain the carbon migration ratio. Use the sum of squared residuals between the carbon migration ratio and the carbon migration weight as the objective function. Constrain the carbon migration weight to be non-negative and the weight sum to be one. First, solve the linear equation involving all weights. When a negative weight appears, fix the first negative weight to zero according to the sampling interval order and solve it again. After all free weights are non-negative, release the zero weights one by one. When the objective function decreases, retain the release result and continue to solve until releasing any zero weight does not cause the objective function to decrease. Output the carbon migration weight. The carbon migration weight refers to the dimensionless distribution ratio of carbon at the formation location to the exit location after different sampling intervals. The carbon migration amount at each sampling interval is first divided by the sum of the carbon migration amounts at all sampling intervals to form the migration ratio. S24 then uses the migration ratio as the fitting object and obtains the carbon migration weight under the constraint that the carbon migration weight is non-negative and the sum of all carbon migration weights is one. The carbon migration weight is then written into the Scholdinger bridge prior matrix in S3 according to the same sampling interval.

[0008] In a preferred embodiment, S3 includes: S31. For each agricultural batch of Napier grass without material exchange, the source quality is established by the carbon amount temporarily stored at the beginning of the segment and the newly formed carbon amount at each formation location. The target quality is established by the carbon emission amount at the outlet of each emission location and the carbon amount temporarily stored at the end of the segment. When the source location is later than the emission location, write zero. For other locations, read the carbon migration weight according to the sampling interval between the two locations. Write the carbon migration weight remaining amount from the source location to the end of the segment at the temporary storage location at the end of the segment to form the Scholdinger bridge prior matrix. The formation location refers to the sampling location where carbon formation occurs after S2 dynamic modeling within the non-material exchange section. The writing is based on the fact that the newly formed carbon amount at the corresponding sampling location is greater than zero, and the sampling position is used as the unique tag number of the formation location. The newly formed carbon amount refers to the carbon mass that newly enters the temporary storage or exits the outlet at the current formation location relative to the previous sampling location, and is the remaining carbon mass after deducting the carbon mass that migrated from the previous formation location to the current location; the newly formed carbon amount is written to the corresponding formation location and is used as part of the source mass of S31. The emission location refers to the sampling location within the non-material exchange section where the carbon emission at the outlet is recorded. The end sampling position of the sampling interval to which the carbon emission at the outlet belongs is used as the emission location number, and the carbon emission at the outlet is used as the corresponding target mass. The source location refers to the time position in the Scholdinger bridge prior matrix that bears the source quality, including the segment start position and each formation position; the segment start position carries the initial temporary carbon amount of the segment, and each formation position carries the corresponding newly formed carbon amount, and the source location numbers are arranged in ascending order along the sampling position; The temporary storage location refers to the location at the end of a non-material exchange section that is used to receive carbon that has not yet been discharged within the section. The last sampling position of the section is used as the temporary storage location number, and the amount of carbon temporarily stored at the end of the section is used as the corresponding target mass. S32. Accumulate the source mass and outlet carbon emission along the sampling positions respectively. At each sampling position, check that the cumulative outlet carbon emission does not exceed the cumulative source mass, and check that the total source mass is equal to the total target mass. If any check relationship is not true, the edge computing node writes the current segment as a carbon record conflict and returns to S1 to re-form the fermentation carbon sequence. When all check relationships are true, retain the Scholdinger bridge prior matrix.

[0009] In a preferred embodiment, S3 further includes: S33. Based on the Scholdinger bridge prior matrix, perform iterative proportional fitting of the Scholdinger bridge. First, multiply each row by the ratio of the source mass to the current row sum, and then multiply each column by the ratio of the target mass to the current column sum. After each row and column scaling, convert the matrix elements according to the carbon measurement resolution unit until each row sum is consistent with the corresponding source mass and each column sum is consistent with the corresponding target mass, forming a real carbon migration matrix. S34. Divide the real carbon migration matrix by the carbon measurement resolution unit and round down each element. Subtract the corresponding row integer from the source mass integer value to form the row surplus. Subtract the corresponding column integer from the target mass integer value to form the column surplus. Establish residual transport edges at non-zero prior matrix positions and use the fractional part of one minus the corresponding real carbon migration amount as the edge cost. Use successive shortest paths to push the row surplus to zero column surplus, and then multiply by the carbon measurement resolution unit to form the carbon migration matrix. The process of establishing residual transport edges at non-zero prior matrix positions is as follows: a supply node is established at the source position with row surplus, and a receiving node is established at the target position with column surplus. The supply node and the receiving node are connected only when the corresponding Scholdinger bridge prior matrix element is non-zero. The capacity of the forward edge is the lower value between the corresponding row surplus and column surplus, and the edge cost is the value obtained by subtracting the decimal part of the corresponding real carbon migration amount from one. For each forward edge formed, a reverse edge with an initial capacity of zero and a cost that is the opposite of the cost of the forward edge is established simultaneously for subsequent withdrawal of the integer carbon amount that has been pushed. The successive shortest path process is as follows: Establish a source node and connect it to all supply nodes with remaining row capacity, with edge capacity set to the corresponding row capacity and cost set to zero; establish a sink node and connect it to all receiving nodes with remaining column capacity, with edge capacity set to the corresponding column capacity and cost set to zero; in the current residual network, execute the shortest path index in ascending order of node position, selecting the path with the earlier predecessor node position when costs are equal, obtaining a path from the source node to the sink node, using the lower of the remaining capacity of each edge of the path as the push amount, reducing the remaining capacity of the forward edge and increasing the remaining capacity of the corresponding reverse edge, while deducting the corresponding row and column capacity; repeat the shortest path search and push until all row and column capacity are zero, then write the cumulative net push amount of each forward edge back to the corresponding matrix position.

[0010] In a preferred embodiment, S4 includes: S41. Read the exit carbon emission column of the carbon migration matrix according to the sampling position, write the carbon amount of each matrix element into the formation position of the corresponding source row and accumulate them row by row to obtain the carbon emission amount belonging to the formation position, and then read the matrix elements of the segment end temporary storage column to form the temporary carbon composition saved according to the source row. The export carbon emission column refers to a matrix column in the carbon migration matrix that uses the export carbon emission amount of each emission location as the column quality. Each column corresponds to a specific emission location, and each element in the column represents the amount of carbon allocated to that emission location from different formation locations. The sum of all elements in the column is equal to the export carbon emission amount of the corresponding emission location. S42. For the next section without material exchange, the temporary carbon composition is split into the initial temporary carbon amount of the section according to the original formation position, and together with the newly formed carbon amount of the next section without material exchange, the source quality is established, so that the carbon amount retained across sections continues to maintain the original formation position, forming the source sequence of the next section without material exchange.

[0011] In a preferred embodiment, S4 further includes: S43. The carbon migration weight is obtained by re-executing the matrix pencil method on the next non-material exchange section by the edge computing node. Then, the Scholdinger bridge operation is performed with the source sequence to form the next carbon migration matrix. The carbon emission amount and temporary carbon composition of the location are recalculated according to S41. The non-material exchange sections of the Napier grass agricultural batch are recursively deduced. S44. After the last section without material exchange ends, the carbon emission amount of each section is accumulated according to the original formation position and the last temporary carbon composition is retained. The carbon quantity is checked by making the sum of the total source mass equal to the sum of the accumulated carbon emission amount and the last temporary carbon amount. When the equation is true, the dynamic carbon conversion sequence is obtained by arranging according to the formation position. When the equation is false, the edge computing node writes the corresponding agricultural batch as a carbon record conflict.

[0012] In a preferred embodiment, S5 includes: S51. Read the dynamic carbon conversion sequence according to the formation location, and associate the carbon emission amount of each formation location with the last temporary carbon composition to the same Napier grass agricultural batch. Subtract the carbon emission amount from the source carbon amount of each formation location to obtain the location-retained carbon amount, and arrange them according to the formation location to form a retained carbon sequence. S52. The amount of carbon retained at each position along the carbon retention sequence is accumulated to obtain the amount of carbon retained during fermentation. At the same time, the amount of carbon discharged at each formation position is accumulated to obtain the cumulative amount of carbon discharged at the export. The sum of the amount of carbon retained during fermentation and the cumulative amount of carbon discharged at the export are checked against the initial carbon amount of the Napier grass agricultural batch. If they match, the amount of carbon retained during fermentation of the corresponding Napier grass agricultural batch is retained. S53. Divide the amount of carbon retained during fermentation by the initial amount of carbon in the Napier grass agricultural batch to obtain the carbon sequestration performance value. Determine the carbon sequestration contribution of the corresponding formation location by the proportion of carbon retained at each location to the amount of carbon retained during fermentation. Write the data into the agricultural fermentation carbon sequestration performance record according to the Napier grass agricultural batch and output the carbon sequestration benefit assessment results.

[0013] A system for evaluating the carbon sequestration benefits of *Phragmites australis* biomass fermentation includes a carbon sequence construction module, a migration modeling module, a bridging allocation module, a dynamic relocation module, and a carbon sequestration evaluation module. The carbon sequence construction module is used to obtain the fermentation carbon records formed by the edge computing nodes of the Napier grass agricultural fermentation site according to the unified sampling time sequence, divide the non-material exchange section according to the material exchange location, determine the temporary carbon amount based on the inorganic carbon content of the fermentation liquid and the gas phase carbon content in the tank, and form the fermentation carbon sequence by combining the outlet carbon discharge amount. The migration modeling module is used to construct forward block Hankel matrices and backward block Hankel matrices according to the sampling position for sections without material exchange. It then uses the matrix pencil method to obtain migration eigenvalues ​​and modal coefficients through generalized Schur decomposition, reconstructs the carbon migration amount at the position interval, completes dynamic carbon migration modeling, and forms carbon migration weights. The bridging allocation module establishes a Scholdinger bridge prior matrix based on carbon migration weights. It uses the initial temporary carbon amount and the newly formed carbon amount at each position as the source quality, and the outlet carbon emission amount and the temporary carbon amount at the end of the segment as the target quality. The matrix elements are written in chronological order and row and column scaling is performed alternately until the row and column quality is conserved, thus forming a carbon migration matrix. The dynamic positioning module is used to allocate the carbon emission at the outlet to the corresponding formation position according to the carbon migration matrix, and write the temporary carbon amount at the end of the section as the initial temporary carbon amount of the next section without material exchange. The edge computing nodes repeat the matrix pencil method and Scholdinger bridge operation in sequence to form a dynamic carbon conversion sequence. The carbon sequestration assessment module, based on the dynamic carbon conversion sequence, obtains the fermentation-retained carbon content by subtracting the cumulative export carbon emissions from the completed formation location from the initial carbon content of the Napier grass agricultural batch. It then performs agricultural fermentation carbon sequestration performance analysis based on the ratio of fermentation-retained carbon content to initial carbon content, and outputs the carbon sequestration benefit assessment results.

[0014] The technical effects and advantages of this invention are as follows: 1. This scheme identifies carbon migration lags using the matrix pencil method and uses the Scholdinger bridge to rebalance the carbon emissions at the outlet, so that the delayed carbon emissions correspond to the actual formation stage, reducing the misattribution of stages. 2. Use the material exchange location as the section boundary to unify the sampling sequence of temporary carbon quantity and outlet carbon emission quantity, thereby reducing the interference of feeding, discharging and sampling on carbon migration identification. 3. Establish time-linked constraints on carbon migration weights and implement conservation allocation of source quality and target quality to make the correspondence between carbon content at formation location and emission location clearer; 4. During the cross-segment recursion process, the original formation location and temporary carbon composition are preserved, so that the delayed carbon continues to maintain its source attribution, reducing duplicate measurement and loss of attribution. 5. Update carbon migration relationships and verify carbon conservation by using edge computing nodes in sections without material exchange, so that dynamic modeling results are updated with fermentation status, improving the consistency of continuous evaluation. Attached Figure Description

[0015] Figure 1 This is a flowchart of the evaluation process for carbon sequestration by fermentation of Napier grass according to the present invention; Figure 2 This is a structural diagram of the Napier grass fermentation carbon fixation evaluation system of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Refer to the instruction manual appendix Figure 1-2 The present invention provides a method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation, comprising: S1. Obtain the fermentation carbon record formed by the edge computing node of the agricultural fermentation site of Napier grass according to the unified sampling time sequence, divide the non-material exchange section according to the material exchange location, determine the temporary carbon amount based on the inorganic carbon amount of fermentation liquid and the gas phase carbon amount in the tank, and form the fermentation carbon sequence by combining the outlet carbon discharge amount. After the agricultural batches of Napier grass enter the fermentation process, it is necessary to first unify the data generated by different metering devices to the same time reference and carbon quality caliber, and then divide the sampling sections that can reflect the internal carbon migration process of fermentation according to the material exchange location. Before the data collection begins, the recording times of the liquid level detection equipment, inorganic carbon detection equipment, gas component detection equipment, and outlet flow meter are unified by the edge computing node clock, and the agricultural batch identifier of Napier grass, sampling cycle, effective internal volume calibration version of fermentation container, and carbon mass unit are fixed. During the same batch collection period, the above data version remains unchanged. When another batch of Napier grass agricultural products enters the fermentation container, the current batch record is terminated and a new agricultural batch record is established. The specific execution is as follows: The sampling time assignment is used to determine which sampling location should be recorded for the carbon content change caused by material exchange; the edge computing node establishes a sampling interval with left open and right closed according to adjacent sampling times, reads the metering record formed by solid or liquid crossing the boundary of the fermentation vessel, and retains the normal outlet exhaust in the outlet carbon emission record; When executing S11, the material exchanges that occur within the sampling interval and at the end of the interval are uniformly written into the interval termination sampling position. When multiple material exchanges occur in the same interval, the exchange direction and measurement value are saved respectively and connected to the same termination sampling position. Then, based on the agricultural batch identifier of Napier grass and the sampling position, each sampling position is connected to form a batch sampling sequence. When the same device generates duplicate records with identical content, one record is retained; when the content is inconsistent, a record conflict is written. Late records are only added before the batch sampling sequence enters subsequent calculations. Data processing for the corresponding time range is stopped when the metering record that determines the material exchange location is missing. Carbon data conversion is used to convert different detection results into the carbon mass corresponding to the sampling location; after entering S12, the edge computing node multiplies the inorganic carbon concentration of the fermentation broth by the liquid volume at the same sampling location to obtain the inorganic carbon content of the fermentation broth, where the liquid volume is converted from the liquid level detection value according to the fixed liquid level volume calibration relationship. The amount of carbon in the gas phase inside the tank is obtained by multiplying the carbon concentration in the gas phase inside the tank by the effective internal volume of the fermentation vessel minus the volume of the liquid phase. The carbon emission at the outlet is calculated time-by-time within the sampling interval. For each time-time segment, the outlet gas flow rate corresponding to the previous outlet flow rate record is multiplied by the carbon concentration of the outlet gas at the same time, and then multiplied by the time difference between two adjacent records to obtain the carbon emission of the time segment. The carbon emission of the time segment is accumulated until the end of the sampling interval to obtain the carbon emission of the outlet in the interval. The carbon concentration in the gas phase inside the tank and the carbon concentration in the outlet gas are both converted to the same volume basis according to the temperature and pressure at the time of detection. When the inorganic carbon concentration, liquid level, or gas phase carbon concentration in the fermentation broth is missing, the missing sampling location is recorded. When the outlet gas flow rate or outlet gas carbon concentration is missing, the outlet carbon discharge missing range is recorded. The converted data forms the carbon record of the sampling location. Segmentation is used to obtain fermentation carbon data that is not directly disturbed by solids or liquids across the container boundary; edge computing nodes read carbon records at sampling positions along the sampling positions. In S13, the first complete sampling position after the material exchange position is taken as the starting point of the no-material exchange segment, and the last complete sampling position before the next material exchange position is taken as the ending point of the no-material exchange segment. The material exchange position is not written into the no-material exchange segment. The same boundary rules are used from the start of fermentation to the first material exchange point and from the last material exchange point to the end of fermentation; At each sampling location, the inorganic carbon content of the fermentation broth is added to the gas phase carbon content in the tank to obtain the temporary carbon content. Then, the temporary carbon content and the carbon discharge volume at the outlet of the corresponding sampling interval are paired and written into the fermentation carbon sequence according to the sampling position. If there are missing sampling locations within a section, the section is disconnected at the missing location. The two sides of the disconnection are re-formed into sections without material exchange. Sections that cannot form a front-to-back sampling relationship only retain boundary records and do not enter S2. In actual operation, source plot A, the third harvest batch and the feed batch 01 form a Napier grass agricultural batch, and the edge computing node records the fermentation process at a sampling cycle of ten minutes; If the material feeding is completed at 09:20 and the next material exchange occurs at 10:40, the material feeding record at 09:20 is written into the sampling interval ending at 09:20. The complete sampling position from 09:30 to 10:30 constitutes the section without material exchange. After each sampling location is sequentially written with the temporary carbon quantity and the carbon emission quantity at the interval outlet, a fermentation carbon sequence is formed. This ensures that the data entering S2 maintains the same agricultural batch, the same time base, and the same carbon quality caliber, and eliminates the direct interference of material exchange location on the dynamic modeling of carbon migration.

[0018] S2. For sections without material exchange, construct forward block Hankel matrix and backward block Hankel matrix according to sampling position. Use generalized Schul decomposition to perform matrix pencil method to obtain migration eigenvalues ​​and modal coefficients, reconstruct carbon migration amount at position intervals, complete dynamic carbon migration modeling and form carbon migration weights. After the non-material exchange section enters the carbon migration dynamic modeling, the edge computing node reads the fermentation carbon sequence of the current batch of Napier grass agricultural frozen, and uses the sampling position, section boundary and carbon quality caliber determined by S1; during the calculation, the carbon measurement resolution unit and numerical calculation accuracy are fixed so that each non-material exchange section adopts the same calculation rules. Integerization is used to ensure the consistency of subsequent matrix elimination; the edge computing nodes uniformly convert the carbon mass resolutions participating in the S2 operation into rational numbers under the same mass unit, take the common multiple of the denominator to complete the integerization, and then take the common factor of each resolution after integerization and convert it back to the mass unit to obtain the carbon measurement resolution unit. The amount of carbon stored and the amount of carbon emitted from exports are divided by the carbon measurement unit and then rounded to the nearest integer. If the amount is between two integers, the even integer is taken. S21 then writes the N sampling positions of the non-material exchange section into two-dimensional column vectors composed of the temporary carbon quantity and the outlet carbon emission quantity. The integer quotient R of N divided by two is taken as the number of blocks and the number of blocks and columns C is obtained by subtracting R from N. The i-th block row and j-th column of the Hankel matrix are moved forward and written into the two-dimensional column vector of the i-th plus j minus one sampling position. The same position of the Hankel matrix is ​​moved backward and written into the two-dimensional column vector of the i-th plus j sampling position. After expansion, both matrices are two R-rows and C-columns. The order of the matrix bundle is determined by Bareiss elimination. For the expanded forward block Hankel matrix, the first non-zero integer is read from left to right, column by column, and from the current processing row downwards as the pivot. In the first round, the previous pivot is taken as 1. Each element in the lower right region is updated by multiplying the current pivot by the element to be updated minus the product of the pivot column element and the pivot row element. From the second round onwards, the previous pivot is divided again. When a newline occurs, the original row number is saved synchronously. After elimination, the number of non-zero pivots is used as the matrix bundle order, and the original pivot row number and pivot column number are recorded according to the pivot formation order to form the matrix extraction bit number. A no-transition mode flag is written when the matrix bundle order is zero and both the forward block Hankel matrix and the backward block Hankel matrix are zero; a conflict identification flag is written when the matrix bundle order is zero and the backward block Hankel matrix contains non-zero elements. After obtaining the matrix extraction index, S22 reads elements from the forward block Hankel matrix and the backward block Hankel matrix respectively according to the same original row number and pivot column number, forming forward and backward shift matrices of the same order; The two matrices are subjected to complex generalized Schul decomposition, and the forward-shifted upper triangular matrix and the backward-shifted upper triangular matrix are obtained by common left unitary transformation and right unitary transformation. The number of significant digits is taken as the last digit of the decimal digits of the current agricultural batch integer sequence plus twice the matrix bundle order, and rounded to this number after each round of transformation. Read the results along the diagonal position. When the diagonal elements of the forward-shifted upper triangular matrix are non-zero, divide the corresponding diagonal elements of the backward-shifted upper triangular matrix by the corresponding diagonal elements of the forward-shifted upper triangular matrix to obtain the migration eigenvalue. Positions where the forward-shifted diagonal elements are zero are not written into the finite set of migration eigenvalues. When the finite migration feature value is empty, stop the modal reconstruction of the current segment, and record the number of repetitions for repeated migration feature values ​​after rounding. After the migration eigenvalues ​​enter S23, a modal basis matrix is ​​constructed according to the sampling interval; a single migration eigenvalue corresponds to an integer power column, and repeated migration eigenvalues ​​increase according to the number of repetitions by a confluence column composed of a product of decreasing order of the sampling interval and a lower power; The modal basis matrix is ​​decomposed using complex Haushold orthogonal trigonometric decomposition. The Euclidean length and the phase of the first element are calculated from the vector below the diagonal position of the current column. When the first element is zero, the phase is taken as one. The reflection matrix is ​​constructed using the reflection vector and its conjugate transpose, and the elements below the diagonal position are eliminated column by column. Each reflection acts synchronously on the observation matrix composed of the temporary carbon amount and the outlet carbon amount. Then, the modal coefficients are obtained by substituting the last row of the upper triangular matrix into the first row in reverse order. Then, the migration characteristic value is calculated to the integer power of the sampling interval and multiplied by the modal coefficient corresponding to the outlet carbon emission. Repeated characteristic values ​​are combined with the same confluence term, and complex conjugate terms are combined in pairs to obtain a real number to obtain the carbon migration amount; negative values ​​are written as zero, and non-negative values ​​are retained. When all carbon migrations are zero and export carbon emissions are zero, only the temporary path at the end of the segment is retained. When all carbon migrations are zero but export carbon emissions are non-zero, a migration reconstruction conflict is written. After entering S24, the carbon migration ratio is obtained by dividing the non-negative carbon migration amount of each sampling interval by the sum. Then, the sum of the squared residuals of the carbon migration ratio and the carbon migration weight is used as the objective function, and the carbon migration weight is limited to be non-negative and the weight sum is one. First, solve the constraint linear equation involving all weights. When a negative weight appears, fix the first negative weight to zero in ascending order of sampling interval and solve it again. After all free weights are non-negative, zero weights are released one by one and the objective function is recalculated. The release result is retained when the objective function decreases; otherwise, zero weights are restored. After traversal, the carbon migration weights are output and written into the current Napier grass agricultural batch record according to the sampling interval, for S3 to read when establishing the Scholdinger bridge prior matrix. For example, when the non-material exchange section contains eight valid sampling positions, the number of blocks R is four and the number of blocks C is four. After the two Hankel matrices are expanded, they are both eight rows and four columns. When Bareiss elimination forms three non-zero principal elements, the matrix bundle order is three, and a third-order forward shift matrix and a third-order backward shift matrix are formed according to the corresponding principal element row number and column number. After generalized Shure decomposition, confluence mode basis matrix and complex Haushold decomposition, the carbon migration amount of each sampling interval is obtained, and then carbon migration weight is formed, so that the edge computing node can update the dynamic modeling results of carbon migration according to the non-material exchange section, and provide the S3 with the time migration basis under the unified sampling interval.

[0019] S3. Based on carbon migration weights, a Scholdinger bridge prior matrix is ​​established. The initial temporary carbon amount and the newly formed carbon amount at each position are used as the source quality, and the outlet carbon emission amount and the temporary carbon amount at the end of the section are used as the target quality. The matrix elements are written in chronological order and row and column scaling is performed alternately until the row and column quality is conserved, thus forming a carbon migration matrix. After entering the Scholdinger bridge solution stage, the edge computing nodes use the carbon migration weight formed by S2 as the basis for time migration, and use the sampling position and carbon measurement resolution unit of the current Napier grass agricultural batch to uniformly map the formation, discharge and end-of-section retention of carbon in the non-material exchange section to the same mass transportation process. Before entering the calculation, the source mass and target mass are converted into the integer number of carbon measurement units. The time connection within the segment is only allowed to point from the source position to the same or subsequent target position, thereby ensuring that the carbon conservation and the time sequence relationship participate in the subsequent calculations simultaneously; the specific processing is as follows. The source and target ends are first established based on the carbon balance of the section, so that the Scholdinger Bridge does not depend on the migration and distribution results that have not yet been determined; the newly formed carbon amount at the current sampling location is the current temporary carbon amount minus the temporary carbon amount at the previous sampling location, plus the carbon emission amount at the outlet of the current sampling interval. When the result is positive, the current sampling position is registered as the formation position and the corresponding source mass is written. When the result is zero, no new source row is added. When the result is negative, the current section without material exchange is written into the carbon balance conflict and the bridge operation is stopped. When executing S31, the first source location is established with the carbon amount temporarily stored at the beginning of the segment, and the subsequent source locations are established with the newly formed carbon amount at each formation location; on the target side, the emission location index is retained for each sampling location, the zero target mass is written at the location where the carbon emission at the outlet is zero, and a temporary storage location is set at the end and the carbon amount temporarily stored at the end of the segment is written. For prior matrix elements where the source location is later than the emission location, write zero. For the remaining elements, read the carbon migration weights output by S2 according to the sampling interval between the two locations. Take one from the temporary column at the end of the segment corresponding to each source location and subtract the sum of the carbon migration weights of all emission locations from the source location to the end of the segment. If the remaining amount is less than zero, write the migration weight conflict. The resulting source quality satisfies the condition that the sum of the initial temporary carbon amount and the newly formed carbon amount in the segment equals the sum of the export carbon amount and the temporary carbon amount at the end of the segment, thus forming the Scholdinger bridge prior matrix. After the prior matrix is ​​written, it is necessary to confirm that the time support can support the entire source quality; the edge computing node accumulates the source quality and the outlet carbon emission along the sampling position, and S32 compares the two cumulative values ​​at each sampling position. When the cumulative outlet carbon emission exceeds the cumulative source quality, a causal conflict is recorded; then the total source quality is compared with the total target quality. If the two are different, a total conflict is recorded. After the two checks are confirmed, a source node is established for each positive source quality and a target node is established for each positive target quality. The source node and the target node are connected only at the non-zero position of the Scholdinger bridge prior matrix. Then, the source quality is used as the amount to be pushed by the source node and the target quality is used as the amount to be received by the target node. Starting from the first source node, search for connection positions with remaining receiving capacity by incrementing the target position number. Each time, push the carbon quantity corresponding to the first item in the remaining amount of the source node and the receiving capacity of the target node that reaches zero, and update the remaining amounts at both ends until all source nodes have finished pushing. When there is a source surplus but no connected target location, the section without material exchange is written into the support conflict; after all source surplus and target surplus are reduced to zero, the Scholdinger bridge prior matrix enters S33. Rereading S1 data is only allowed when new late records, supplementary collection records, or metering correction records have already been written; if the newly formed fermentation carbon sequence is consistent with the one before the conflict, the calculation of the current non-material exchange section will be terminated to avoid repeating the same operation. Quality scaling begins with the Scholdinger bridge prior matrix supporting the verification; S33 first uses the prior matrix as the initial real matrix, keeps all zeros for rows with zero source quality, and calculates the source quality divided by the current row and the resulting scaling factor for rows with positive source quality, and multiplies it into the whole row; After row scaling is completed, columns with a target quality of zero are kept all zero, and columns with a target quality of positive are calculated by dividing the target quality by the current column and the resulting scaling factor and multiplying it into the whole column. Positions where the prior matrix is ​​zero are kept zero throughout all rounds. If the source quality is positive and the current row sum is zero, or if the target quality is positive and the current column sum is zero, return to S32 to register a support conflict. After each row and column scaling is completed, the edge computing node divides the sum of each row and column by the carbon measurement resolution unit and converts it into the number of measurement units using the same integer value rule as S21; scaling stops when the sum of all row integers is equal to the source mass integer value and the sum of all column integers is equal to the target mass integer value. If the quantization matrix state obtained after a complete row and column scaling is the same as the quantization matrix state in a previous round, and the integer sums of the rows and columns are still not all consistent, then the scaling is stopped and the current real matrix is ​​retained for integer margin correction, thus forming a real carbon migration matrix. After the real number allocation is completed, the migration results with decimals need to be restored to the carbon mass that is consistent with the on-site measurement resolution; In S34, the edge computing node first converts the real carbon migration matrix into the number of carbon units, rounds down each element, and obtains the row margin by subtracting the corresponding row integer sum from the source mass integer value, and obtains the column margin by subtracting the corresponding column integer sum from the target mass integer value. Only at the position where the prior matrix of the Scholdinger Bridge is non-zero and the elements of the real matrix have a fractional part, establish a positive residual transport edge with a capacity of one carbon unit, and take the edge cost as one minus the corresponding fractional part. At the same time, establish a reverse edge with an initial capacity of zero and the opposite sign of the cost. The source node connects to the source node with row margin, and the receiving node connects to the sink node with column margin. When the residual network contains negative cost reverse edges, Bellman-Ford is used to perform relaxation in ascending order of node number. When the costs are the same, the path with the earlier predecessor node number is retained. For each path from the source node to the sink node obtained, one carbon metering unit is pushed. The capacity of the forward edge is reduced by one, the capacity of the reverse edge is increased by one, and the corresponding row margin and column margin are deducted simultaneously. Repeat the path search until all row and column reserves are zero. Then add the integer value of each matrix position to the net push unit number and multiply it back to the carbon measurement resolution unit to form the carbon migration matrix. When there are still reserves but the residual network has no source-sink path, write the current non-material exchange segment into an integer conflict and do not pass the result to S4. The above treatment is illustrated using a non-material exchange section with a total source mass of 170 carbon units. The initial temporary carbon quantity in the section is 120 units. The two formation sites within the section add 30 units and 20 units respectively, and the two emission sites discharge 40 units and 50 units respectively. The temporary carbon quantity at the end of the section is 80 units. Thus, both the source mass and the target mass are 170 units. The edge computing node first writes the carbon migration weights obtained by S2 into the Scholdinger bridge prior matrix according to the sampling interval, and then completes the cumulative causal verification and support transport verification. After obtaining the real carbon migration matrix by row and column scaling, the decimal part is corrected by single-unit residual transport so that the sum of each source row and the sum of the target column strictly correspond to the integer value of the unit of measurement. The final carbon migration matrix simultaneously stores the actual carbon distribution relationship between the formation location, emission location, and the temporary storage location at the end of the segment, which can be directly read by S4 to determine the carbon emission amount at the formation location and the temporary carbon composition.

[0020] S4. Based on the carbon migration matrix, the carbon emission at the outlet is allocated to the corresponding formation position, and the temporary carbon amount at the end of the section is written as the initial temporary carbon amount of the next section without material exchange. The matrix pencil method and Scholdinger bridge operation are repeated sequentially by the edge computing nodes to form a dynamic carbon conversion sequence. When crossing adjacent sections without material exchange, the carbon that has not yet been discharged from the previous section needs to retain both its original formation location and subsequent migration starting point. The amount of temporarily stored carbon should be corrected in conjunction with the material exchange that occurs at the section boundary to avoid the same amount of carbon being repeatedly used as a new source in multiple sections. The edge computing nodes therefore process each section without material exchange sequentially along the Napier grass agricultural batch, only transmitting the temporarily stored carbon composition, original formation location, and current transportation start location, while keeping the sampling position and carbon measurement resolution unit used by S2 unchanged; the specific execution is as follows; After the carbon migration matrix completes the integer mass allocation, the edge computing nodes read the outlet carbon emission column column by column according to the sampling position of the emission location, and determine the source location corresponding to each matrix element according to the matrix row number; When processing S41, the carbon amount allocated to each exit carbon emission column from the same source row is accumulated sequentially, and the resulting value is written into the carbon emission amount belonging to the formation position stored in the source row; the temporary storage column at the end of the segment is read row by row, and each matrix element is written into the temporary carbon composition as the temporary storage carbon amount that has not yet been discharged from the corresponding source position. For the first non-material exchange section of the current agricultural batch, the initial temporary carbon quantity of the section is taken as the initial formation position of the sampling position of the section and written into the initial temporary storage mark; for subsequent non-material exchange sections, the source line continues to read the previously saved original formation position; The sum of the carbon emission amount associated with the formation location of a source line and the temporary storage amount at the end of the segment should be equal to the source quality of the source line entering the current segment. If the values ​​are inconsistent, stop cross-segment recursion and write the carbon quantity conflict. When there is material exchange between adjacent non-material exchange sections, the initial temporary carbon quantity of the section needs to be corrected according to the boundary exchange record; before executing S42, the edge computing node reads the temporary carbon composition of the previous section and the carbon quantity carried by the boundary material exchange. When carry-out occurs, the amount of carbon carried out that is clearly attributed to the boundary record is deducted from the original formation position record. When the same batch of agricultural materials is returned, the corresponding temporary carbon composition is rewritten according to the original formation position saved in the return record to form the corrected starting temporary carbon amount of the segment. When the boundary exchange record only gives the total carbon amount but cannot determine the original formation location composition, the total carbon amount is not divided proportionally and is directly written into the boundary attribution conflict; when the carbon brought in cannot be attributed to the current Napier grass agricultural batch, the current agricultural batch recursion is terminated and a new batch record is established. Each modified temporary carbon composition retains its original formation position, while the starting sampling position of the next non-material exchange section is written as the current transportation starting position. It is then combined with the newly formed carbon amount obtained in the next non-material exchange section according to S31 to form the source sequence. The original formation position and the current transportation starting position of the newly formed carbon amount are both taken from their own formation positions. After entering the section without material exchange, the time migration calculation only uses the current transportation start position, and the final destination still uses the original formation position; In the processing of S43, the edge computing node first re-executes the matrix pencil method according to the temporary carbon amount and the outlet carbon emission amount of the next non-material exchange section to obtain the carbon migration weight applicable to the current section sampling interval. For the temporary carbon composition inherited from the previous segment, the carbon migration weight is read at the sampling interval from the current transportation start position to each emission position. For the newly formed carbon in the current segment, the carbon migration weight is read at the sampling interval from its own formation position to the emission position. Then, the Scholdinger bridge prior matrix is ​​established according to S3 and the current carbon migration matrix is ​​obtained. Then, the exit carbon emission column is read to update the cumulative carbon emission of each original formation position. At the same time, the segment end temporary storage column is read to generate a new temporary carbon composition. The current transport start position is updated once a non-material exchange segment is completed, while the original formation position remains unchanged. This allows the same amount of carbon spanning multiple segments to continue calculation using the migration timescale of the current segment and eventually write back to the initial formation position. After all material exchange-free sections are processed, the edge computing nodes perform the final conservation check based on the unique source carbon content. The unique source carbon content in S44 only includes the initial temporary carbon content of the first material exchange-free section, the newly formed carbon content of each material exchange-free section, and the boundary carbon content that can be attributed to the current agricultural batch, and deducts the carbon content that has been brought out by the boundary material exchange. The temporary carbon composition inherited from the previous section and used as the source quality of the next material exchange-free section is not counted repeatedly. Edge computing nodes accumulate the carbon emissions assigned to all formation locations based on their original formation locations, and read the last temporary carbon composition. When the carbon emissions from the unique source match the sum of the accumulated carbon emissions and the last temporary carbon composition, the agricultural batch identifier, source carbon emissions, accumulated carbon emissions, and last temporary carbon composition are written in ascending order based on the original formation location to form a dynamic carbon conversion sequence. If there is a discrepancy, an agricultural batch carbon record conflict is written, and transmission to S5 is stopped. To illustrate the cross-segment recursion process, at the end of the previous segment without material exchange, eighty carbon units are retained, of which fifty units correspond to formation position three and thirty units correspond to formation position five. Eight units are carried out from the segment boundary, and the boundary record clearly shows that five units come from formation position three and three units come from formation position five. Then, the initial temporary carbon quantity of the next segment without material exchange is corrected to seventy-two units. The two temporary carbon compositions retain the original formation positions three and five, respectively. At the same time, the initial sampling position of the next segment without material exchange is written as the current transportation start position common to both records. After the edge computing node completes the Scholdinger bridge operation using the carbon migration weights re-acquired in the next non-material exchange section, the carbon emission at the outlet is still written back to form positions three and five respectively, and the unemitted part continues to be transferred to the next non-material exchange section. With this approach, the carbon content across segments is only counted at the source when it is first formed. Segment inheritance only changes the starting point of transportation, without changing the original formation attribution, thus providing a consistent dynamic carbon conversion sequence for S5 to calculate the carbon sequestration performance of agricultural fermentation based on the formation location.

[0021] S5. Based on the dynamic carbon conversion sequence, the fermentation carbon content is obtained by subtracting the cumulative export carbon emission from the completed formation location from the initial carbon content of the Napier grass agricultural batch. The ratio of the fermentation carbon content to the initial carbon content is used to perform agricultural fermentation carbon sequestration performance analysis and output the carbon sequestration benefit assessment results. After the dynamic carbon conversion sequence is formed, the performance accounting uses the unique source carbon amount determined by S44 as the unified carbon domain. The initial carbon amount of the Napier grass agricultural batch is the sum of the initial source carbon amount of the first non-material exchange section, the newly formed carbon amount of each non-material exchange section, and the boundary carbon amount that can be attributed to the current agricultural batch, and then deducts the carbon amount that has been brought out by the boundary material exchange. Inoculum, reflux liquid or other carbon-containing inputs that cannot be attributed to the current Napier grass agricultural batch are separately accounted for and not included in the initial carbon content of the agricultural batch; before entering S5, the edge computing node freezes the dynamic carbon conversion sequence and the initial carbon content of the agricultural batch, so that the carbon content of the location source, the carbon emission amount attribution and the subsequent performance value adopt the same quality range; the processing order is as follows; The location-level accounting first determines the retention result of each source carbon at the end of fermentation; the edge computing node reads the dynamic carbon conversion sequence according to the original formation location. In the S51 processing, the source carbon amount of the same original formation location, the cumulative carbon emission amount of each non-material exchange section written back to the formation location, and the final temporary carbon composition are associated according to the agricultural batch identifier. The location-retained carbon amount is obtained by subtracting the carbon emission amount of the formation location from the source carbon amount of the formation location, and is checked against the carbon amount of the same original formation location in the final temporary carbon composition. When the two are consistent, the carbon sequence is written in increments according to the original formation position. When the two are inconsistent or the carbon retention at the position is negative, it is written as a conflict in the formation position accounting and will not be included in subsequent performance calculations. The carbon carried out by boundary material exchange that has already been deducted from the carbon quantity of the sole source in S44 will not be deducted again, thus ensuring that the same carbon quantity only participates in the source accounting once. Total amount conservation is used to confirm that the retained carbon sequence can represent the complete evaluation range of the current agricultural batch; after completing the location accounting, S52 accumulates the carbon amount retained at all locations along the retained carbon sequence to obtain the fermentation retained carbon amount, and at the same time accumulates the carbon emission amount assigned to the location along the dynamic carbon conversion sequence to obtain the cumulative export carbon emission amount, and then adds the two together and compares them with the initial carbon amount of the Napier grass agricultural batch. When the values ​​are consistent, the carbon content retained during fermentation is preserved. When the values ​​are inconsistent, the source carbon content, the carbon emission content, and the last temporary carbon composition are checked back according to the original formation location. The first position where the difference occurs is located and written into the agricultural batch carbon content to check for conflicts. When the initial carbon content of the Napier grass agricultural batch is zero, the performance calculation is terminated to avoid the subsequent ratio calculation losing its definition. The performance results retain both the total amount of agricultural batches and the location contribution formed by dynamic relocation; in S53, the carbon sequestration performance value is obtained by dividing the amount of carbon retained by fermentation by the initial amount of carbon in the Napier grass agricultural batch. When the amount of carbon retained during fermentation is greater than zero, the carbon retention amount at each location is divided by the amount of carbon retained during fermentation to obtain the carbon fixation contribution at the corresponding formation location. When the amount of carbon retained during fermentation is zero, the carbon fixation performance value and the carbon fixation contribution at each formation location are both written as zero. Edge computing nodes write carbon sequestration performance values, original formation location, carbon content from the location source, carbon emission attribution to the formation location, carbon content retained at the location, and carbon sequestration contribution into the agricultural fermentation carbon sequestration performance record of the same batch of Napier grass. This allows the formation location attribution relationship determined by the matrix pencil method and the Scholdinger bridge to be directly included in the carbon sequestration benefit assessment results, instead of only retaining the total cumulative export carbon emission. For example, an agricultural batch of Napier grass, after verification by S44, has an initial carbon content of 170 carbon units. The source carbon content of the three original formation sites is 120, 30, and 20 units respectively, and the cumulative carbon emissions attributable to the formation sites are 60, 20, and 10 units respectively. Therefore, the carbon retention at the three sites is 60, 10, and 10 units respectively, the carbon retention during fermentation is 80 units, and the cumulative carbon emissions from exports are 90 units. The sum of these two is consistent with the initial carbon content of the agricultural batch. The carbon sequestration performance value is written as 80 divided by 170, and the carbon sequestration contribution of each formation site is written as 60 divided by 80, 10 divided by 80, and 10 divided by 80 respectively. The resulting agricultural fermentation carbon sequestration performance record reflects both the overall carbon retention level of each batch of Napier grass and the contribution results of each formation location after dynamic relocation. It can be directly used to compare the fermentation carbon sequestration performance of different plots, harvesting batches, and feed batches under the same agricultural management caliber.

[0022] exist Figure 1It should be noted that: This figure represents the overall process of the evaluation method for carbon sequestration benefits of Napier grass biomass fermentation, which sequentially goes through fermentation carbon sequence formation, dynamic carbon migration modeling, Schöringer bridge allocation, dynamic carbon conversion sequence formation, and carbon sequestration performance analysis, and finally outputs the carbon sequestration benefit evaluation results; solid arrows indicate the data transfer and processing order between each step, and partition boxes indicate different processing stages; exist Figure 2 It should be noted that: This figure represents the data processing structure of the biomass fermentation carbon sequestration benefit assessment system of Napier grass. The fermentation site data first enters the edge computing node, and then the carbon sequence construction, migration modeling, bridging allocation and dynamic relocation are completed in sequence to finally form the carbon sequestration performance result; the solid arrow indicates the data transmission direction, the ellipse represents the edge computing node, and the rectangle represents the input, processing and output units.

[0023] Furthermore, the present invention also includes a system for evaluating the carbon sequestration benefits of Napier grass biomass fermentation, the system comprising a carbon sequence construction module, a migration modeling module, a bridging allocation module, a dynamic relocation module, and a carbon sequestration evaluation module. The carbon sequence construction module is used to obtain the fermentation carbon records formed by the edge computing nodes of the Napier grass agricultural fermentation site according to the unified sampling time sequence, divide the non-material exchange section according to the material exchange location, determine the temporary carbon amount based on the inorganic carbon content of the fermentation liquid and the gas phase carbon content in the tank, and form the fermentation carbon sequence by combining the outlet carbon discharge amount. The migration modeling module is used to construct forward block Hankel matrices and backward block Hankel matrices according to the sampling position for sections without material exchange. It then uses the matrix pencil method to obtain migration eigenvalues ​​and modal coefficients through generalized Schur decomposition, reconstructs the carbon migration amount at the position interval, completes dynamic carbon migration modeling, and forms carbon migration weights. The bridging allocation module establishes a Scholdinger bridge prior matrix based on carbon migration weights. It uses the initial temporary carbon amount and the newly formed carbon amount at each position as the source quality, and the outlet carbon emission amount and the temporary carbon amount at the end of the segment as the target quality. The matrix elements are written in chronological order and row and column scaling is performed alternately until the row and column quality is conserved, thus forming a carbon migration matrix. The dynamic positioning module is used to allocate the carbon emission at the outlet to the corresponding formation position according to the carbon migration matrix, and write the temporary carbon amount at the end of the section as the initial temporary carbon amount of the next section without material exchange. The edge computing nodes repeat the matrix pencil method and Scholdinger bridge operation in sequence to form a dynamic carbon conversion sequence. The carbon sequestration assessment module, based on the dynamic carbon conversion sequence, obtains the fermentation-retained carbon content by subtracting the cumulative export carbon emissions from the completed formation location from the initial carbon content of the Napier grass agricultural batch. It then performs agricultural fermentation carbon sequestration performance analysis based on the ratio of fermentation-retained carbon content to initial carbon content, and outputs the carbon sequestration benefit assessment results.

[0024] Overall Implementation and Explanation: This plan takes the agricultural batch of Napier grass as the management object. Edge computing nodes collect data on inorganic carbon in fermentation liquid, gas phase carbon in tank and carbon emission at the outlet according to a unified sampling sequence. The plan also divides the non-material exchange zone according to the feeding, discharging and sampling locations. First, the migration relationship between temporary carbon and outlet carbon over time is analyzed using the matrix pencil method to obtain the carbon migration weights corresponding to different sampling intervals. Then, the carbon migration weights are written into the Scholdinger bridge prior matrix. Under the constraints of carbon conservation and time sequence, the carbon discharged from the outlet is redistributed to the actual formation location. After a section ends, the temporary carbon that has not yet been discharged remains at its original formation location and is transferred to the next section without material exchange. The edge computing node recalculates the migration relationship of the current section and continues to distribute it until fermentation ends. Finally, the source carbon, the carbon discharge amount, and the retained carbon amount of each formation location are obtained. Based on this, the carbon sequestration performance value of the Napier grass agricultural batch and the carbon sequestration contribution of each formation location are calculated. Taking a batch of Napier grass from the same plot and harvesting batch entering the fermentation container as an example, during the fermentation process, some carbon is formed in the morning and enters the fermentation liquid and the gas phase inside the tank. It is not discharged from the gas outlet until the afternoon. If the exhaust time is directly counted, the carbon formed in the morning will be recorded in the afternoon. After adopting this scheme, the edge computing node first uses the sampling data of the section with no material exchange between morning and afternoon to determine the time migration law of carbon from formation to discharge through the matrix pencil method. Then, the Scholdinger bridge is used to distribute the carbon discharge detected in the afternoon back to the morning and other actual formation locations according to the migration relationship. If some carbon is still left in the fermentation liquid or the gas phase in the tank at the end of the section, this part of carbon will be carried into the next fermentation section for continued tracking. After fermentation, the carbon released at each stage and the carbon remaining in the fermentation system can be mapped back to their original formation locations, thus obtaining carbon fixation performance results of Napier grass consistent with the actual fermentation process.

[0025] The above description is merely 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 evaluating the carbon sequestration benefits of Napier grass biomass fermentation, characterized in that, include: S1. Obtain the fermentation carbon record formed by the edge computing node of the agricultural fermentation site of Napier grass according to the unified sampling time sequence, divide the non-material exchange section according to the material exchange location, determine the temporary carbon amount based on the inorganic carbon amount of fermentation liquid and the gas phase carbon amount in the tank, and form the fermentation carbon sequence by combining the outlet carbon discharge amount. S2. For sections without material exchange, construct forward block Hankel matrix and backward block Hankel matrix according to sampling position. Use generalized Schul decomposition to perform matrix pencil method to obtain migration eigenvalues ​​and modal coefficients, reconstruct carbon migration amount at position intervals, complete dynamic carbon migration modeling and form carbon migration weights. S3. Based on carbon migration weights, a Scholdinger bridge prior matrix is ​​established. The initial temporary carbon amount and the newly formed carbon amount at each position are used as the source quality, and the outlet carbon emission amount and the temporary carbon amount at the end of the section are used as the target quality. The matrix elements are written in chronological order and row and column scaling is performed alternately until the row and column quality is conserved, thus forming a carbon migration matrix. S4. Based on the carbon migration matrix, the carbon emission at the outlet is allocated to the corresponding formation position, and the temporary carbon amount at the end of the section is written as the initial temporary carbon amount of the next section without material exchange. The matrix pencil method and Scholdinger bridge operation are repeated sequentially by the edge computing nodes to form a dynamic carbon conversion sequence. S5. Based on the dynamic carbon conversion sequence, the fermentation retention carbon amount is obtained by subtracting the cumulative export carbon emission amount assigned to the completed formation location from the initial carbon amount of the Napier grass agricultural batch. The ratio of the fermentation retention carbon amount to the initial carbon amount is used to perform agricultural fermentation carbon sequestration performance analysis, and the carbon sequestration benefit assessment results are output.

2. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 1, characterized in that: S1 includes: S11. Obtain the agricultural batch identifier of Napier grass and the material exchange time of fermentation container recorded by the edge computing node. Form a sampling interval with adjacent sampling times, write the material exchange that occurs in the interval to the end sampling position of the interval, and connect each sampling position according to the agricultural batch identifier to form a batch sampling sequence. S12. Based on the batch sampling sequence, the inorganic carbon content of the fermentation broth is obtained by multiplying the inorganic carbon concentration of the fermentation broth by the liquid phase volume, and the carbon content of the gas phase in the tank is obtained by multiplying the carbon concentration of the gas phase in the tank by the gas phase volume. The carbon discharge at the outlet is obtained by accumulating the product of the outlet gas flow rate and the carbon concentration in the sampling interval, thus forming a carbon record of the sampling location. S13. Based on the carbon record at the sampling location, a non-material exchange section is formed by the sampling locations between adjacent material exchange locations. The amount of inorganic carbon in the fermentation liquid is added to the amount of gas phase carbon in the tank to obtain the amount of temporary carbon. Then, the amount of temporary carbon is connected to the amount of carbon discharged at the outlet according to the sampling position to form the fermentation carbon sequence.

3. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 2, characterized in that: S2 includes: S21. For each section without material exchange, convert the temporary carbon amount and the outlet carbon emission amount into an integer sequence according to the carbon metering resolution unit of the edge computing node. Construct the forward block Hankel matrix and the backward block Hankel matrix according to the sampling position. Perform Bareiss elimination on the forward block Hankel matrix. In each round, read the first non-zero element in column order as the pivot. Use the current pivot and the pivot of the previous round to cross-multiply and divide to eliminate the remaining elements in the pivot column. Determine the matrix bundle order according to the number of non-zero pivots. Record the pivot row number and pivot column number that are consistent with the number of matrix bundle orders according to the pivot formation order to form the matrix extraction position number. S22. Extract the index according to the matrix. Read the elements with the same row number and the same column number from the forward block Hankel matrix and the backward block Hankel matrix to form forward and backward shift matrices with order equal to the order of the matrix bundle, respectively. Perform generalized Shur decomposition on the two matrices and transform them into upper triangular matrices simultaneously to form the migration eigenvalues ​​by corresponding to the ratio of the diagonal elements.

4. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 3, characterized in that: S2 also includes: S23. Construct a modal basis matrix by writing the integer powers of the migration eigenvalues ​​in ascending order of the sampling interval. Perform Haushold orthogonal triangular decomposition on the modal basis matrix and substitute the temporary carbon amount and the outlet carbon emission amount into the triangular matrix in reverse order to obtain the modal coefficients. Then reconstruct the carbon migration amount by sampling interval according to the migration eigenvalues ​​and modal coefficients. S24. Divide the non-negative carbon migration amount of each sampling interval by the sum of the non-negative carbon migration amounts of all sampling intervals to obtain the carbon migration ratio. Use the sum of the squared residuals between the carbon migration ratio and the carbon migration weight as the objective function. Constrain the carbon migration weight to be non-negative and the weight sum to be one. First, solve the linear equation involving all weights. When a negative weight appears, fix the first negative weight to zero according to the sampling interval order and solve it again. After all free weights are non-negative, release the zero weights one by one. When the objective function decreases, retain the release result and continue to solve until releasing any zero weight does not cause the objective function to decrease. Output the carbon migration weight.

5. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 4, characterized in that: S3 includes: S31. For each agricultural batch of Napier grass without material exchange, the source quality is established by the carbon amount temporarily stored at the beginning of the segment and the newly formed carbon amount at each formation location. The target quality is established by the carbon emission amount at the outlet of each emission location and the carbon amount temporarily stored at the end of the segment. When the source location is later than the emission location, write zero. For other locations, read the carbon migration weight according to the sampling interval between the two locations. Write the carbon migration weight remaining amount from the source location to the end of the segment at the temporary storage location at the end of the segment to form the Scholdinger bridge prior matrix. S32. Accumulate the source mass and outlet carbon emission along the sampling positions respectively. At each sampling position, check that the cumulative outlet carbon emission does not exceed the cumulative source mass, and check that the total source mass is equal to the total target mass. If any check relationship is not true, the edge computing node writes the current segment as a carbon record conflict and returns to S1 to re-form the fermentation carbon sequence. When all check relationships are true, retain the Scholdinger bridge prior matrix.

6. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 5, characterized in that: S3 also includes: S33. Based on the Scholdinger bridge prior matrix, perform iterative proportional fitting of the Scholdinger bridge. First, multiply each row by the ratio of the source mass to the current row sum, and then multiply each column by the ratio of the target mass to the current column sum. After each row and column scaling, convert the matrix elements according to the carbon measurement resolution unit until each row sum is consistent with the corresponding source mass and each column sum is consistent with the corresponding target mass, forming a real carbon migration matrix. S34. Divide the real carbon migration matrix by the carbon measurement resolution unit and round down each element. Subtract the corresponding row integer from the source mass integer value to form the row surplus, and subtract the corresponding column integer from the target mass integer value to form the column surplus. Establish residual transport edges at non-zero prior matrix positions and use the fractional part of the corresponding real carbon migration amount as the edge cost. Use successive shortest paths to push the row surplus to zero column surplus, and then multiply by the carbon measurement resolution unit to form the carbon migration matrix.

7. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 6, characterized in that: S4 includes: S41. Read the exit carbon emission column of the carbon migration matrix according to the sampling position, write the carbon amount of each matrix element into the formation position of the corresponding source row and accumulate them row by row to obtain the carbon emission amount belonging to the formation position, and then read the matrix elements of the segment end temporary storage column to form the temporary carbon composition. S42. For the next section without material exchange, the temporary carbon composition is split into the initial temporary carbon amount of the section according to the original formation position, and together with the newly formed carbon amount of the next section without material exchange, the source quality is established, so that the carbon amount retained across sections continues to maintain the original formation position, forming a source sequence.

8. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 7, characterized in that: S4 also includes: S43. The carbon migration weights are obtained by re-executing the matrix pencil method on the next non-material exchange section by the edge computing node. Then, the Scholdinger bridge operation is performed with the source sequence to form the next carbon migration matrix. The carbon emission amount and temporary carbon composition at the location are recalculated according to S41. S44. After the last section without material exchange ends, the carbon emission amount of each section is accumulated according to the original formation position and the last temporary carbon composition is retained. The carbon quantity is checked by making the sum of the total source mass equal to the sum of the accumulated carbon emission amount and the last temporary carbon amount. When the equation is true, the dynamic carbon conversion sequence is obtained by arranging according to the formation position. When the equation is false, the edge computing node writes the corresponding agricultural batch as a carbon record conflict.

9. The method for evaluating the carbon sequestration benefits of Napier grass biomass fermentation according to claim 8, characterized in that: S5 includes: S51. Read the dynamic carbon conversion sequence according to the formation location, and associate the carbon emission amount of each formation location with the last temporary carbon composition to the same Napier grass agricultural batch. Subtract the carbon emission amount from the source carbon amount of each formation location to obtain the location-retained carbon amount, and arrange them according to the formation location to form a retained carbon sequence. S52. The amount of carbon retained at each position along the carbon retention sequence is accumulated to obtain the amount of carbon retained during fermentation. At the same time, the amount of carbon discharged at each formation position is accumulated to obtain the cumulative amount of carbon discharged at the export. The sum of the amount of carbon retained during fermentation and the cumulative amount of carbon discharged at the export are checked against the initial carbon amount of the Napier grass agricultural batch. If they match, the amount of carbon retained during fermentation of the corresponding Napier grass agricultural batch is retained. S53. Divide the amount of carbon retained during fermentation by the initial amount of carbon in the Napier grass agricultural batch to obtain the carbon sequestration performance value. Determine the carbon sequestration contribution of the corresponding formation location by the proportion of carbon retained at each location to the amount of carbon retained during fermentation. Write the data into the agricultural fermentation carbon sequestration performance record according to the Napier grass agricultural batch and output the carbon sequestration benefit assessment results.

10. A system for evaluating the carbon sequestration benefits of *Phragmites australis* biomass fermentation, used to implement the method for evaluating the carbon sequestration benefits of *Phragmites australis* biomass fermentation as described in any one of claims 1-9, the system comprising a carbon sequence construction module, a migration modeling module, a bridging allocation module, a dynamic relocation module, and a carbon sequestration evaluation module, characterized in that: The carbon sequence construction module is used to obtain the fermentation carbon records formed by the edge computing nodes of the Napier grass agricultural fermentation site according to the unified sampling time sequence, divide the non-material exchange section according to the material exchange location, determine the temporary carbon amount based on the inorganic carbon content of the fermentation liquid and the gas phase carbon content in the tank, and form the fermentation carbon sequence by combining the outlet carbon discharge amount. The migration modeling module is used to construct forward block Hankel matrices and backward block Hankel matrices according to the sampling position for sections without material exchange. It then uses the matrix pencil method to obtain migration eigenvalues ​​and modal coefficients through generalized Schur decomposition, reconstructs the carbon migration amount at the position interval, completes dynamic carbon migration modeling, and forms carbon migration weights. The bridging allocation module establishes a Scholdinger bridge prior matrix based on carbon migration weights. It uses the initial temporary carbon amount and the newly formed carbon amount at each position as the source quality, and the outlet carbon emission amount and the temporary carbon amount at the end of the segment as the target quality. The matrix elements are written in chronological order and row and column scaling is performed alternately until the row and column quality is conserved, thus forming a carbon migration matrix. The dynamic positioning module is used to allocate the carbon emission at the outlet to the corresponding formation position according to the carbon migration matrix, and write the temporary carbon amount at the end of the section as the initial temporary carbon amount of the next section without material exchange. The edge computing nodes repeat the matrix pencil method and Scholdinger bridge operation in sequence to form a dynamic carbon conversion sequence. The carbon sequestration assessment module, based on the dynamic carbon conversion sequence, obtains the fermentation-retained carbon content by subtracting the cumulative export carbon emissions from the completed formation location from the initial carbon content of the Napier grass agricultural batch. It then performs agricultural fermentation carbon sequestration performance analysis based on the ratio of fermentation-retained carbon content to initial carbon content, and outputs the carbon sequestration benefit assessment results.

Images