Wetland microbial activity monitoring method and system

Abstract

The invention relates to the technical field of microorganism monitoring, and particularly discloses a wetland microorganism activity monitoring method and system.The method comprises the steps that a plurality of first time sequence data sets of different layers and related to different wetland state parameters are obtained; analyzing and acquiring a parameter change mode of each layer according to the first time sequence data set; similarity matching is carried out based on the parameter change mode and a pre-constructed parameter correlation model library, so that microbial metabolism modes of different layers are determined; determining an activity evaluation mode of the microorganisms in the corresponding layer based on a microorganism metabolism mode; generating an activity evaluation result according to the activity evaluation mode and the first time sequence data set; according to the method, the biological action signal and the non-biological process signal are distinguished, the evaluation accuracy is improved, the activity evaluation mode corresponding to the metabolism mode is determined based on the identified metabolism mode, and the difference of evaluating the activity of microorganisms in different metabolism pathways by adopting one index is avoided.

Description

Technical Field

[0001] The present application relates to the field of microbial monitoring technology, and in particular to a method and system for monitoring wetland microbial activity. Background Art

[0002] With the deepening of research on wetland ecosystem function, monitoring microbial activity has become an important tool for assessing wetland health. Current conventional monitoring methods utilize multi-parameter sensor networks to collect time-series data on key indicators such as redox potential, dissolved oxygen concentration, and gas composition, combined with chemometric analysis to reveal microbial metabolic characteristics. Monitoring technology is evolving towards multi-parameter integration and high-frequency data collection, providing fundamental data support for analyzing wetland microbial functional diversity.

[0003] However, existing methods face significant limitations in complex wetland environments: first, the dissolved oxygen gradient formed on the vertical profile of the wetland leads to the coexistence of multiple microenvironments such as aerobic zones, facultative zones, and anaerobic zones, and single-point or single-layer monitoring is difficult to capture the spatial differentiation characteristics of microbial communities; second, processes such as mineral redox and abiotic degradation of organic matter and microbial metabolism jointly affect monitoring parameters, and traditional data analysis is difficult to effectively separate biological and abiotic action signals; third, conventional static models are difficult to accurately identify real changes in metabolic patterns.

[0004] There is currently no effective technical solution to the above problems. Summary of the Invention

[0005] The purpose of this application is to provide a wetland microbial activity monitoring method and system to dynamically identify microbial metabolic modes at different layers, and determine appropriate evaluation methods based on the microbial metabolic modes to monitor and evaluate the activity of microorganisms at different layers, and reduce the impact of irrelevant data.

[0006] In a first aspect, the present application provides a method for monitoring wetland microbial activity for evaluating wetland microbial activity, the method comprising the following steps: S1. Acquire a first time series data set of different wetland state parameters at multiple different layers based on the monitoring component; S2. Analyze and obtain parameter change patterns of each layer according to the first time series data set; S3. performing similarity matching based on the parameter change pattern with a pre-built parameter association model library to determine the microbial metabolic modes of different layers, wherein the parameter association model library includes a plurality of metabolic models that match the microbial metabolic modes and are associated with wetland state parameters; S4. Determining an activity evaluation method for microorganisms in corresponding layers based on the microbial metabolism mode; S5. Generate an activity evaluation result according to the activity evaluation method and the first time series data set.

[0007] This method distinguishes biological action signals from non-biological process signals, improving the accuracy of the assessment. It also determines the activity evaluation method corresponding to the metabolic mode based on the identified metabolic mode, avoiding the use of a single indicator to evaluate the differences in microbial activity in different metabolic pathways.

[0008] The wetland microbial activity monitoring method, wherein the wetland state parameters include dissolved oxygen, redox potential, pH value, carbon dioxide content, methane content, hydrogen sulfide content, and nitrate content.

[0009] The method of the present application uses the above-mentioned wetland status parameters to focus on key indicators closely related to wetland microbial activity, thereby reducing data redundancy and processing burden and improving the targeted nature of monitoring.

[0010] The wetland microbial activity monitoring method, wherein the parameter change pattern includes the correlation and proportional relationship between different wetland state parameters in the same layer and the change rate of each wetland state parameter, and the metabolic model is represented as a theoretical multidimensional feature vector composed of a theoretical correlation matrix, a theoretical change rate vector, and a theoretical proportional relationship set between the wetland state parameters; Step S3 includes: S31, extracting the correlation matrix, change rate vector, and proportional relationship set in the parameter change pattern to form a multidimensional feature vector; S32. Calculate the cosine similarity between the multidimensional feature vector and the theoretical multidimensional feature vectors of each metabolic model, and select the microbial metabolic mode corresponding to the theoretical multidimensional feature vector with the highest cosine similarity as the microbial metabolic mode of the corresponding layer.

[0011] The wetland microbial activity monitoring method, wherein step S32 includes: S321, constructing a KD tree index structure, and storing the theoretical multidimensional feature vectors of each metabolic model in the KD tree; S322. Using the fast search capability of the KD tree, search for N theoretical multidimensional feature vectors that are closest to the multidimensional feature vector, where N is a positive integer; S323, calculating the cosine similarity between the multidimensional feature vector and N theoretical multidimensional feature vectors; S324. Select a theoretical multidimensional feature vector corresponding to the maximum value of the cosine similarity value, and use the microbial metabolic mode corresponding to the theoretical multidimensional feature vector as the microbial metabolic mode of the corresponding layer.

[0012] The wetland microbial activity monitoring method, wherein step S5 further comprises: S51, analyzing and filtering out parameter data of non-microbial metabolic patterns in the first time series data set to perform denoising on the first time series data set; S52 : Evaluate the first time series data set after denoising based on the activity evaluation method to generate an activity evaluation result.

[0013] The wetland microbial activity monitoring method, wherein step S4 comprises: S41. For the determined microbial metabolic mode, query a preset mapping relationship table of microbial metabolic modes and activity evaluation indicators, and extract an evaluation mode corresponding to the microbial metabolic mode, wherein the evaluation mode includes at least one activity evaluation indicator, a data evaluation window, and a corresponding evaluation process.

[0014] The wetland microbial activity monitoring method, wherein step S52 includes: S521, extracting corresponding wetland state parameters from the first time series data set according to the evaluation process and the evaluation window to form a data set to be analyzed; S522: Processing the data set to be analyzed based on the evaluation process to generate an activity index value that matches the type of activity evaluation index; S523: Generate an activity evaluation result according to the activity index value and the corresponding activity evaluation index.

[0015] The wetland microbial activity monitoring method, wherein step S1 comprises: S11. Acquire a first time series data set about different wetland state parameters at multiple different layers based on the monitoring component, and simultaneously acquire a second time series data set about environmental state parameters at multiple different layers based on the monitoring component; Step S5 also includes the following steps performed between step S51 and step S52: S5A, aligning the first time series data set and the second time series data set based on the timestamp, and based on the preset correction rules, compensating and correcting the wetland state parameters in the first time series data set of the corresponding layer according to the environmental parameters in the second time series data set.

[0016] In the wetland microbial activity monitoring method, the environmental parameters are divided into temperature and conductivity.

[0017] In a second aspect, the present application further provides a wetland microbial activity monitoring system for evaluating wetland microbial activity, the system comprising: An acquisition module, configured to acquire a first time series data set of different wetland state parameters at multiple different layers based on the acquisition of the monitoring component; a change pattern analysis module, configured to analyze and obtain a parameter change pattern of each layer based on the first time series data set; a metabolic pattern analysis module for performing similarity matching based on the parameter change pattern with a pre-built parameter association model library to determine the microbial metabolic patterns at different layers, wherein the parameter association model library includes a plurality of metabolic models that match the microbial metabolic patterns and are associated with wetland state parameters; An evaluation method determination module, used to determine the activity evaluation method of the microorganisms in the corresponding layer based on the microbial metabolism method; An evaluation module is configured to generate an activity evaluation result according to the activity evaluation method and the first time series data set.

[0018] The system distinguishes biological action signals from non-biological process signals, improving the accuracy of the assessment. It also determines the activity evaluation method corresponding to the identified metabolic mode, avoiding the use of a single indicator to evaluate the differences in microbial activity in different metabolic pathways.

[0019] From the above, it can be seen that the present application provides a wetland microbial activity monitoring method and system, wherein the method of the present application combines the first time series data set of wetland status parameters of multiple different layers with parameter change pattern analysis and parameter association model library similarity matching, thereby realizing the identification of microbial metabolic modes in different layers, and based on this, determining the activity evaluation method corresponding to the metabolic mode, thereby distinguishing biological action signals from non-biological process signals, improving the accuracy of the evaluation, and determining the activity evaluation method corresponding to the metabolic mode based on the identified metabolic mode, avoiding the use of one indicator to evaluate the differences in microbial activity in different metabolic pathways. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of a method for monitoring wetland microbial activity provided in an embodiment of the present application.

[0021] Figure 2 A schematic structural diagram of a wetland microbial activity monitoring system provided in an embodiment of the present application.

[0022] Reference numerals: 201, acquisition module; 202, change pattern analysis module; 203, metabolic pattern analysis module; 204, evaluation method determination module; 205, evaluation module. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0024] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0025] First, please refer to Figure 1 Some embodiments of the present application provide a method for monitoring wetland microbial activity for evaluating wetland microbial activity, the method comprising the following steps: S1. Acquire a first time series data set of different wetland state parameters at multiple different layers based on the monitoring component; S2. Analyze and obtain parameter change patterns of each layer based on the first time series data set; S3. Based on the parameter change pattern, similarity matching is performed with a pre-built parameter association model library to determine the microbial metabolic mode of different layers. The parameter association model library includes multiple metabolic models that match the microbial metabolic mode and are associated with wetland state parameters; S4. Determine the activity evaluation method of the microorganisms in the corresponding layer based on the microbial metabolism mode; S5. Generate an activity evaluation result according to the activity evaluation method and the first time series data set.

[0026] Specifically, the monitoring component refers to the hardware equipment used to collect wetland environmental parameters. It is mainly used to obtain environmental data at different locations and depths of the wetland. It can be implemented using sensor arrays, probe networks, remote sensing equipment, etc., and can be integrated and designed according to the specific wetland status parameters that need to be collected. The types of structures such as sensor arrays are not described here.

[0027] More specifically, multiple different layers refer to multiple discrete or continuous depth areas on the vertical profile of the wetland, which can be achieved by layered sensor deployment, multi-depth probe integration, profile scanning, etc., and are mainly used to capture the environmental gradients and spatial differentiation of microbial communities in the vertical direction of the wetland.

[0028] More specifically, wetland status parameters refer to indicators that can reflect the physical, chemical or biochemical status of the wetland environment. They can be characterized by dissolved oxygen concentration, redox potential, pH value, gas composition, ion concentration, etc., and are mainly used to provide direct or indirect evidence of microbial metabolic activities.

[0029] More specifically, the first time series data set refers to a data set that is continuously or periodically collected over a period of time, for multiple different layers, and about different wetland status parameters. It can be stored in the form of a time series database, data matrix, etc., and is mainly used to record the changing dynamics of wetland environmental parameters over time.

[0030] More specifically, the parameter change pattern refers to the law of change of wetland status parameters over time in the first time series data set, the relationship between different parameters, and the change rate and other characteristics. It can be extracted and described using time series analysis methods, correlation analysis, principal component analysis and other techniques. It is mainly used to extract signals related to microbial metabolic activities from the original data.

[0031] More specifically, a parameter association model library refers to a collection of pre-established mathematical or statistical models that encompass the relationships between various known microbial metabolic pathways and changes in wetland state parameters. These models can take the form of models based on theoretical stoichiometry, models trained with experimental data, or models compiled from literature. They are primarily used as a reference standard for identifying the microbial metabolic pathways reflected in actual monitoring data. The parameter association model library can include models for parameter change patterns corresponding to wetland microbial metabolic pathways such as aerobic respiration, nitrate reduction, sulfate reduction, and methanogenesis. These models can be constructed based on stoichiometric principles or existing research data.

[0032] More specifically, similarity matching refers to the process of comparing the parameter change pattern obtained from actual monitoring with the metabolic model in the parameter association model library to determine the degree of similarity between the two. It can be achieved by using similarity measurement methods such as cosine similarity, Euclidean distance, and correlation coefficient. It is mainly used to identify the microbial metabolic mode that is closest to the actual wetland environment. For example, if the dissolved oxygen in a certain layer drops rapidly while the redox potential decreases and the nitrate concentration decreases, this may be similar to the activity pattern of nitrate-reducing bacteria.

[0033] More specifically, microbial metabolic mode refers to the specific pathway or type of material transformation and energy metabolism of wetland microbial communities under specific environmental conditions, such as aerobic respiration, anaerobic respiration, methanogenesis, fermentation, etc., which is mainly used to characterize the functional status of wetland microbial communities.

[0034] More specifically, the activity evaluation method refers to the specific indicators, data processing methods and evaluation processes used to evaluate the activity level of a specific microbial metabolism. It can use the production rate of specific metabolites, key enzyme activity, gene expression level, the rate or amplitude of change of specific parameters, etc. as evaluation indicators. It is mainly used to provide standards and methods for evaluating microbial activity.

[0035] More specifically, the activity evaluation result refers to the indicator or evaluation conclusion reflecting the activity level of wetland microorganisms calculated based on the determined activity evaluation method and the first time series data set. It can be presented in the form of activity index, rate value, grade classification, etc., and is mainly used to indicate the functional status of wetland microorganisms.

[0036] More specifically, the solution of the present application acquires environmental dynamic information on a vertical profile of a wetland by collecting first-order time-series data sets of wetland status parameters from multiple different layers. Based on this time-series data, the parameter change patterns of each layer are analyzed and extracted. These patterns reflect the temporal changes of wetland environmental parameters and the mutual influence between parameters. Precisely because microbial metabolic activity causes specific changes in wetland environmental parameters, by similarity matching the extracted parameter change patterns with a pre-built parameter association model library, it is possible to identify microbial metabolic modes that are highly consistent with actual monitoring data. Once the microbial metabolic modes of different layers are determined, corresponding activity evaluation methods can be determined based on these metabolic modes, as different metabolic pathways require different indicators and methods to evaluate their activity. Ultimately, based on the determined activity evaluation method and the original first-order time-series data set, evaluation results reflecting the microbial activity levels at different layers are generated. The entire process forms a closed loop from data collection, pattern recognition, metabolic mode determination, to targeted evaluation, thus achieving the assessment of wetland microbial activity.

[0037] The method of the present application combines the first time series data set of wetland status parameters of multiple different layers with parameter change pattern analysis and parameter association model library similarity matching, thereby realizing the identification of microbial metabolic modes in different layers, and based on this, determining the activity evaluation method corresponding to the metabolic mode, thereby distinguishing biological action signals from non-biological process signals, improving the accuracy of the evaluation, and determining the activity evaluation method corresponding to the metabolic mode based on the identified metabolic mode, avoiding the use of one indicator to evaluate the differences in microbial activity in different metabolic pathways.

[0038] In some preferred embodiments, the wetland state parameters include multiple ones of dissolved oxygen, redox potential, pH value, carbon dioxide content, methane content, hydrogen sulfide content, and nitrate content.

[0039] Specifically, the method of the present application uses the above-mentioned wetland status parameters to focus on key indicators closely related to wetland microbial activity, thereby reducing data redundancy and processing burden and improving the targeted nature of monitoring.

[0040] It should be noted that the carbon dioxide content, methane content, hydrogen sulfide content, and nitrate content mainly refer to the dissolved amounts of the corresponding substances in the corresponding layers of the wetland.

[0041] In some preferred embodiments, the parameter change pattern includes the correlation and proportional relationship between different wetland state parameters in the same layer and the change rate of each wetland state parameter, and the metabolic model is represented as a theoretical multidimensional feature vector composed of a theoretical correlation matrix, a theoretical change rate vector, and a theoretical proportional relationship set between the wetland state parameters; Step S3 includes: S31, extracting the correlation matrix, change rate vector, and proportional relationship set in the parameter change pattern to form a multidimensional feature vector; S32. Calculate the cosine similarity between the multidimensional feature vector and the theoretical multidimensional feature vector of each metabolic model, and select the microbial metabolic mode corresponding to the theoretical multidimensional feature vector with the highest cosine similarity as the microbial metabolic mode of the corresponding layer.

[0042] Specifically, the parameter change pattern refers to the comprehensive information obtained by analyzing the first time series data set of wetland status parameters, which can reflect the dynamic characteristics of the wetland microenvironment. It specifically includes the degree of correlation (correlation) between different wetland status parameters in the same layer over time, the relative content relationship (proportional relationship) between them, and the speed of their respective changes over time (rate of change). Its purpose is to comprehensively capture the complex dynamic characteristics of the wetland microenvironment.

[0043] More specifically, a metabolic model refers to a pre-constructed theoretical model used to characterize the specific microbial metabolic mode in wetland state parameters. It is specifically expressed as a theoretical multidimensional feature vector, which integrates the theoretical correlation matrix, theoretical change rate vector and theoretical proportional relationship set of wetland state parameters under this metabolic mode. Its purpose is to provide a theoretical reference for matching actual monitoring data.

[0044] More specifically, the correlation matrix refers to a matrix that describes the degree of correlation between different wetland status parameters in the same layer. For example, it can be calculated using the Pearson correlation coefficient, Spearman rank correlation coefficient, etc., and its purpose is to quantify the synergistic or antagonistic relationship between parameters; the change rate vector refers to a vector that describes the speed of change of each wetland status parameter in the same layer over time. For example, it can be calculated using the derivative, difference or linear regression slope of time series data, and its purpose is to reflect the trend and intensity of parameter changes; the proportional relationship set refers to a set that describes the relative content between different wetland status parameters in the same layer. For example, the average ratio or relative concentration of the parameters can be calculated, and its purpose is to characterize the relative abundance of the parameters.

[0045] More specifically, a multidimensional feature vector refers to a mathematical vector that integrates information such as the correlation matrix, change rate vector, and proportional relationship set in the parameter change pattern obtained from the analysis of actual monitoring data. Its purpose is to convert complex parameter change patterns into a computable numerical form; cosine similarity refers to the cosine value of the angle between two non-zero vectors, which is often used to judge the similarity of the directions of two vectors. Its purpose is to quantify the degree of match between the feature vector of actual monitoring data and the feature vector of the theoretical metabolic model.

[0046] More specifically, the solution of this application collects a first time-series data set of wetland state parameters from multiple different layers. Based on this data, the parameter variation patterns for each layer are analyzed. This parameter variation pattern not only includes the correlations and proportional relationships between different wetland state parameters within the same layer, but also further incorporates the rate of change of each wetland state parameter, thereby constructing a more comprehensive and detailed description of the parameter variation pattern. This multidimensional parameter variation pattern more fully reflects the complex dynamics of the wetland microenvironment, thereby more accurately capturing the unique parameter signatures of different microbial metabolic modes. On this basis, the pre-constructed metabolic model is represented as a theoretical multidimensional feature vector composed of a theoretical correlation matrix, a theoretical rate of change vector, and a set of theoretical proportional relationships between the wetland state parameters, thereby transforming the complex biological metabolic process into a quantifiable mathematical model. Subsequently, in step S31, the correlation matrix, rate of change vector, and proportional relationship set from the parameter variation pattern obtained through analysis of the actual monitoring data are extracted to form an actual multidimensional feature vector. This actual multidimensional feature vector is then subjected to cosine similarity calculations with the theoretical multidimensional feature vectors of each theoretical metabolic model. The similarity measurement method can accurately evaluate the degree of matching between the actual monitoring data characteristics and the theoretical metabolic model characteristics, and thus select the microbial metabolic mode corresponding to the theoretical multidimensional feature vector with the highest cosine similarity as the microbial metabolic mode of the corresponding layer.

[0047] The above processing method refines the change pattern of wetland state parameters into multi-dimensional characteristics including correlation, proportional relationship and change rate, and represents the metabolic model as the corresponding theoretical multi-dimensional feature vector. By calculating the cosine similarity between the multi-dimensional feature vector composed of actual monitoring data and the theoretical multi-dimensional feature vector for matching, it can capture the dynamic characteristics of the wetland microenvironment more comprehensively and accurately, effectively distinguish different microbial metabolic modes, reduce the interference of abiotic factors, thereby improving the accuracy of identifying microbial metabolic modes and providing a more reliable basis for the evaluation of wetland microbial activity.

[0048] In some preferred embodiments, step S32 includes: S321, constructing a KD tree index structure, and storing the theoretical multidimensional feature vectors of each metabolic model in the KD tree; S322. Using the fast search capability of the KD tree, search for N theoretical multidimensional feature vectors that are closest to the multidimensional feature vector, where N is a positive integer; S323, calculating the cosine similarity between the multidimensional feature vector and N theoretical multidimensional feature vectors; S324. Select a theoretical multidimensional feature vector corresponding to the maximum value of the cosine similarity value, and use the microbial metabolic mode corresponding to the theoretical multidimensional feature vector as the microbial metabolic mode of the corresponding layer.

[0049] Specifically, constructing a KD-tree index structure refers to establishing a multidimensional spatial data structure that recursively divides the space containing multiple data points along a certain coordinate axis to form a series of hyper-rectangular regions. Its purpose is to effectively organize and index high-dimensional data for fast neighbor search.

[0050] More specifically, storing the theoretical multidimensional feature vectors of each metabolic model in the KD tree means putting the theoretical multidimensional feature vector corresponding to each metabolic model in the parameter association model library as a data point into the constructed KD tree structure, which can store each vector as a node or leaf node of the tree and maintain its position information in the tree. The purpose is to use the indexing capability of the KD tree to manage these theoretical vectors and prepare for subsequent fast search.

[0051] More specifically, utilizing the fast search capability of the KD tree means using the KD tree's unique search algorithm to efficiently find the data point closest to a given query point (i.e., the multidimensional feature vector of the current layer) in multidimensional space. This can be achieved using the nearest neighbor search algorithm or the K nearest neighbor search algorithm. Its purpose is to quickly locate a theoretical multidimensional feature vector that is highly similar to the current multidimensional feature vector, avoiding traversing the entire data set.

[0052] More specifically, searching for the N theoretical multidimensional feature vectors that are closest to the multidimensional feature vector refers to finding the N theoretical multidimensional feature vectors that are closest to the multidimensional feature vector of the current layer in the multidimensional space through the KD tree search algorithm. This can be achieved by executing a K-nearest neighbor search query of the KD tree, where K is set to N. The purpose is to obtain a smaller set containing potential best matches, significantly reducing the scope of subsequent similarity calculations. The value of N can be determined based on the number of metabolic models in the parameter association model library, such as being set to 1 / 10 of the total number of metabolic models. The purpose is to balance the efficiency and accuracy of the search and ensure that enough potential matches are found for accurate calculations within a reasonable time.

[0053] More specifically, the method of the present application uses a KD tree for pre-indexing and fast nearest neighbor search, which can significantly reduce the number of times precise cosine similarity calculations are required when the parameter association model library is large, thereby significantly improving the efficiency of determining microbial metabolic patterns. This method, combined with the step of extracting parameter change patterns to form multidimensional feature vectors, makes the entire process of similarity matching based on parameter change patterns and model libraries more efficient. It is particularly suitable for processing large-scale metabolic model libraries, thereby enabling faster and more accurate assessment of microbial activity in different layers of wetlands. It is especially suitable for situations where the parameter association model library contains a large number of metabolic models, effectively solving the problem of low computational efficiency caused by the large size of the model library.

[0054] In some preferred embodiments, step S5 further includes: S51, analyzing and filtering out parameter data of non-microbial metabolic patterns in the first time series data set to perform denoising on the first time series data set; S52 : Evaluate the first time series data set after denoising based on an activity evaluation method to generate an activity evaluation result.

[0055] Specifically, step S51 is preferably analyzed and filtered according to the parameter change pattern and preset rules, wherein the preset rules refer to the judgment criteria or logic set used to identify non-microbial metabolic pattern data, which can be implemented using physical and chemical process knowledge based on wetland environment, statistical anomaly detection methods or machine learning models.

[0056] More specifically, the solution of the present application effectively improves the accuracy of the evaluation by adding a data denoising step before generating the activity evaluation results. The specific process can be to analyze the change patterns of parameters over time and the relationship between parameters, and combine with preset judgment criteria to identify data points or data segments that do not conform to typical microbial metabolic characteristics, such as abnormal fluctuations caused by physical diffusion, chemical precipitation or sensor failure. These non-microbial metabolic pattern data are removed or corrected from the original data set to obtain a purer denoised data set that better reflects microbial activity. Then, based on the previously determined activity evaluation method, this denoised first time series data set is evaluated to generate the final activity evaluation result, which reduces the interference of non-biological factors on the wetland status parameter monitoring data, making the data used for activity evaluation more real and reliable.

[0057] In some preferred embodiments, step S51 includes: S511, extracting parameter correlation information from the parameter change pattern, and constructing a parameter correlation network, where nodes represent wetland state parameters and edges represent correlation strengths between parameters; S512, based on a preset microbial metabolic pattern parameter association rule library, identifying a sub-network structure in the parameter correlation network that matches a known microbial metabolic pattern, wherein the microbial metabolic pattern parameter association rule library includes multiple microbial metabolic patterns and their corresponding parameter correlation network structures; S513. Calculate the network centrality index for the parameter correlation network structure that does not match the known microbial metabolic pattern and identify the key parameter nodes; S514. If the change amplitude of the wetland state parameter corresponding to the key parameter node exceeds the preset threshold, and the change amplitude is opposite to the parameter change direction under the known microbial metabolism mode, then the parameter data is determined to be parameter data of a non-microbial metabolism mode and is filtered out from the first time series data set.

[0058] Specifically, the parameter correlation network refers to a graph structure in which nodes represent wetland status parameters, edges connecting nodes represent the correlations between parameters, and the weights or attributes of the edges can represent the strength or type of correlation.

[0059] More specifically, the preset microbial metabolic pattern parameter association rule base refers to a database or knowledge base that stores a variety of known wetland microbial metabolic patterns and their typical correlation network structures on wetland status parameters. These rules are constructed based on wetland microbiology, biogeochemistry theory and experimental observations.

[0060] More specifically, network centrality metrics refer to quantitative indicators used to measure the importance or influence of nodes in a network, including degree centrality (the number of edges connecting a node), betweenness centrality (the frequency with which a node is located on the shortest path), and closeness centrality (the average distance from a node to other nodes).

[0061] More specifically, key parameter nodes refer to nodes with higher network centrality indicators in the parameter correlation network. Changes in these nodes have a greater impact on the entire network structure or parameter relationship; the preset threshold can be determined based on historical data, expert experience or statistical methods.

[0062] More specifically, the direction of parameter change under a known microbial metabolic pattern refers to the changing trend that the relevant wetland state parameters should theoretically exhibit during a specific microbial metabolic process.

[0063] More specifically, the method of the present application extracts parameter variation patterns, particularly correlation information between parameters, from a first time-series data set and constructs a parameter correlation network based on this information, graphically presenting the complex relationships between wetland state parameters. Subsequently, the constructed network is compared with a preset microbial metabolic pattern parameter association rule library to identify subnetwork structures that conform to known microbial metabolic patterns, thereby preliminarily distinguishing data patterns that may be caused by microbial activity. For network structures that do not match known patterns, network centrality indices are calculated to identify key parameter nodes, which may be more sensitive to abiotic factors. Finally, by analyzing the magnitude of change in the parameters corresponding to these key parameter nodes and comparing their direction of change with the theoretical direction of change under known microbial metabolic patterns, it is possible to more accurately determine whether the parameter data is significantly affected by non-microbial metabolic processes. If the magnitude of change is large and the direction is opposite to the known microbial metabolic pattern, it is determined to be non-microbial metabolic pattern data and filtered out. This combination of network structure analysis, rule matching, and threshold-based judgment method can effectively identify and eliminate the interference of abiotic factors on wetland state parameters, avoid the misjudgment that may result from simple threshold filtering, and improve the accuracy of denoising.

[0064] More specifically, the above processing method can more accurately identify and filter out parameter data that is not related to the microbial metabolism mode, avoid the misjudgment that may be caused by simple filtering methods, and improve the quality of the first time series data set, thereby making the subsequent wetland microbial activity evaluation results based on the denoised data more accurate and reliable.

[0065] In some preferred embodiments, step S4 includes: S41. For the determined microbial metabolic mode, query a preset mapping relationship table of microbial metabolic modes and activity evaluation indicators, and extract an evaluation mode corresponding to the microbial metabolic mode. The evaluation mode includes at least one activity evaluation indicator, a data evaluation window, and a corresponding evaluation process.

[0066] Specifically, the preset mapping relationship table between microbial metabolic modes and activity evaluation indicators refers to a pre-established database or lookup table that stores the correspondence between different microbial metabolic modes and their respective applicable activity evaluation indicators, data evaluation windows and evaluation processes. It can be implemented in the form of a relational database, key-value storage or configuration file.

[0067] More specifically, activity evaluation indicators refer to specific parameters used to quantitatively assess the activity level of microorganisms, such as the production rate of a certain metabolite, the consumption rate of a certain substrate, or the activity of a specific enzyme. They can be expressed by specific numerical values ​​such as the concentration change rate, gas flux, or redox potential change amplitude.

[0068] More specifically, the data evaluation window refers to the range or duration of time series data used to calculate the activity evaluation index value, which can be set as a fixed duration or a dynamic duration based on a specific event trigger.

[0069] More specifically, the evaluation process refers to the process of subjecting the original first time series data set to a series of data processing, calculation and conversion steps to ultimately obtain a numerical value or result that matches the activity evaluation index. It can be defined in the form of mathematical formulas, algorithm models or data processing scripts.

[0070] More specifically, the method of the present application actively queries a preset mapping relationship table by targeting the metabolic mode of microorganisms in a specific layer determined in the previous step. This mapping relationship table associates different microbial metabolic modes with a set of tailored activity evaluation methods. The extracted evaluation method is not a single concept, but includes three key components: at least one activity evaluation index, a data evaluation window, and an evaluation process. The activity evaluation index clarifies what parameters need to be measured to evaluate activity; the data evaluation window specifies the time range within which data should be used for calculation, ensuring the timeliness and pertinence of the evaluation; and the evaluation process provides a standardized calculation method for converting raw monitoring data into specific activity index values. This method forms a logical closed loop with the previous step of determining the metabolic mode and the subsequent steps of evaluation and generating results, ensuring the effectiveness and practicality of the entire wetland microbial activity monitoring method.

[0071] The method of this application can query a preset mapping table for a specific microbial metabolic pathway and extract the corresponding evaluation method, which includes specific activity evaluation indicators, data evaluation windows, and corresponding evaluation processes. This solves the problem of simply determining an activity evaluation method without a specific evaluation method, providing clear indicators, data ranges, and calculation methods for subsequent activity evaluation, allowing the activity evaluation to proceed smoothly and generate meaningful results.

[0072] In some preferred embodiments, step S52 includes: S521. Extract corresponding wetland state parameters from the first time series data set according to the evaluation process and evaluation window to form a data set to be analyzed; S522: Process the data set to be analyzed based on the evaluation process to generate an activity index value that matches the type of the activity evaluation index; S523: Generate an activity evaluation result according to the activity index value and the corresponding activity evaluation index.

[0073] Specifically, the solution of the present application accurately extracts wetland state parameters related to the current microbial metabolism mode from the original first time-series data set according to a preset evaluation process and evaluation window to form a data set to be analyzed, thereby avoiding the introduction of irrelevant or interfering data. The data set to be analyzed refers to a subset of wetland state parameters extracted from the first time-series data set according to the evaluation process and evaluation window for activity evaluation calculations, and its purpose is to provide accurate data input for subsequent evaluation calculations. Based on the data set to be analyzed, data processing is performed using the set evaluation process to convert the original parameters into activity index values ​​that can directly reflect the activity of microorganisms, thereby achieving an effective conversion from original monitoring data to quantitative evaluation indicators. The activity index value refers to a numerical value obtained after processing the data set to be analyzed based on the evaluation process and matching the activity evaluation index type, and its purpose is to quantify the activity of the microorganism to be evaluated. Step S523 compares or judges the activity index value with the corresponding activity evaluation index to generate the final activity evaluation result. The activity evaluation result refers to the final evaluation conclusion drawn based on the activity index value and the corresponding activity evaluation index. It can specifically be a judgment on the microbial activity level (such as high, medium, low) or status (such as active, inactive). Its purpose is to provide intuitive microbial activity assessment information.

[0074] The above-mentioned processing method can accurately extract key data for activity evaluation from a large amount of time-series monitoring data based on the determined microbial metabolic mode and the corresponding evaluation process and evaluation window, and convert these data into quantitative activity index values ​​based on the preset processing process. Finally, targeted and reliable activity evaluation results are generated based on these index values, thereby improving the accuracy and effectiveness of wetland microbial activity monitoring.

[0075] In some preferred embodiments, step S1 includes: S11. Acquire a first time series data set about different wetland state parameters at multiple different layers based on the monitoring component, and simultaneously acquire a second time series data set about environmental state parameters at multiple different layers based on the monitoring component; Step S5 also includes the following steps performed between step S51 and step S52: S5A, aligning the first time series data set and the second time series data set based on the timestamp, and based on the preset correction rules, compensating and correcting the wetland state parameters in the first time series data set of the corresponding layer according to the environmental parameters in the second time series data set.

[0076] Specifically, environmental state parameters refer to external or internal factors that affect the internal environmental conditions of the wetland, which may include temperature, conductivity, humidity, light intensity, etc., and their purpose is to quantify the potential impact of environmental factors on wetland state parameters; the second time series data set refers to a set of data records of environmental state parameters changing over time, which can be obtained by synchronous collection with the first time series data set, and its purpose is to provide background information on the changes of environmental factors over time.

[0077] More specifically, the preset correction rule refers to a mathematical model or logical rule that describes how environmental parameters affect wetland state parameters. It can be implemented using a regression model, a lookup table method, or an expert system rule based on experimental data. Its purpose is to quantify the degree of interference of environmental factors on wetland state parameters. Compensation correction refers to adjusting wetland state parameters using environmental parameters according to preset correction rules to eliminate or reduce the impact of environmental factors. It can be implemented by mathematical calculation or data search. Its purpose is to obtain wetland state parameter values ​​that are closer to the actual reflection of microbial metabolic activities. In the embodiment of the present application, the compensation correction process is manifested as compensating correction of the linear or nonlinear relationship between wetland state parameters based on environmental parameters.

[0078] More specifically, the method of this application simultaneously collects wetland state parameters (a first time-series data set) and, using a monitoring component, acquires a second time-series data set of environmental state parameters from multiple different layers, thereby obtaining environmental context information synchronized with the wetland state parameters. In the subsequent data processing process, specifically before denoising the first time-series data set and generating activity assessment results based on the denoised data, a key correction step is added. This step first aligns the first and second time-series data sets based on timestamps, ensuring temporal consistency between the wetland state parameters and the corresponding environmental parameters. Then, using pre-set correction rules, the wetland state parameters in the corresponding layers of the first time-series data set are compensated based on the environmental parameters in the second time-series data set. For example, if a pre-set rule indicates that rising temperature leads to lower dissolved oxygen readings, the dissolved oxygen data collected during periods of higher temperature will be adjusted upward according to this rule. This correction process effectively removes the interference of environmental factors on the wetland state parameters, ensuring that the corrected wetland state parameters more accurately reflect the actual metabolic activity of microorganisms. The corrected first time series data set is then sent to the subsequent denoising processing module and activity evaluation module for processing, ultimately generating a more accurate wetland microbial activity evaluation result.

[0079] The method of the present application collects environmental status parameters simultaneously with the wetland status parameters, and compensates and corrects the wetland status parameters based on the environmental parameters in subsequent processing. It can effectively eliminate or reduce the interference of environmental factors on the wetland status parameter monitoring results, so that the wetland status parameters can more accurately reflect the real metabolic activities of microorganisms, thereby significantly improving the accuracy and reliability of wetland microbial activity assessment.

[0080] In some preferred embodiments, the environmental parameters are differentiated into temperature and conductivity.

[0081] Specifically, because temperature directly affects the metabolic rate and enzyme activity of microorganisms, and conductivity reflects the ion concentration of water bodies, affecting microbial osmotic pressure and nutrient uptake, when correcting wetland state parameters (such as dissolved oxygen and redox potential), specific correction models or rules can be applied based on real-time temperature and conductivity data. It is precisely because of this targeted correction that the interference of environmental factors on the monitored values ​​of wetland state parameters can be more effectively eliminated, allowing the corrected wetland state parameters to more accurately reflect the true metabolic activity of microorganisms. This targeted environmental parameter correction, combined with the data correction steps based on the first and second time series data sets, forms a more accurate data preprocessing mechanism, providing a more reliable data foundation for subsequent analysis of microbial metabolic patterns and evaluation of microbial activity based on wetland state parameters.

[0082] Second, please refer to Figure 2 Some embodiments of the present application further provide a wetland microbial activity monitoring system for evaluating wetland microbial activity, the system comprising: The acquisition module 201 is configured to acquire a first time series data set of different wetland state parameters at multiple different layers based on the acquisition of the monitoring component; A change pattern analysis module 202 is configured to analyze and obtain a parameter change pattern of each layer based on the first time series data set; The metabolic pattern analysis module 203 is used to perform similarity matching based on the parameter change pattern with a pre-built parameter association model library to determine the microbial metabolic mode of different layers. The parameter association model library includes multiple metabolic models that match the microbial metabolic mode and are associated with wetland state parameters. An evaluation method determination module 204 is used to determine an activity evaluation method of microorganisms in a corresponding layer based on the microorganism metabolism method; The evaluation module 205 is configured to generate an activity evaluation result according to the activity evaluation method and the first time series data set.

[0083] The system of the present application combines the first time series data set of wetland status parameters of multiple different layers with parameter change pattern analysis and parameter association model library similarity matching, thereby realizing the identification of the metabolic mode of microorganisms in different layers, and based on this, determining the activity evaluation method corresponding to the metabolic mode, thereby distinguishing biological action signals from non-biological process signals, improving the accuracy of the evaluation, and determining the activity evaluation method corresponding to the metabolic mode based on the identified metabolic mode, avoiding the use of one indicator to evaluate the differences in microbial activity in different metabolic pathways.

[0084] In addition, the units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0085] Furthermore, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0086] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0087] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A wetland microbial activity monitoring method for evaluating wetland microbial activity, characterized in that: The method comprises the following steps: S1. Acquire a first time series data set of different wetland state parameters at multiple different layers based on the monitoring component; S2. Analyze and obtain parameter change patterns of each layer according to the first time series data set; S3. performing similarity matching based on the parameter change pattern with a pre-built parameter association model library to determine the microbial metabolic modes of different layers, wherein the parameter association model library includes a plurality of metabolic models that match the microbial metabolic modes and are associated with wetland state parameters; S4. Determining an activity evaluation method for microorganisms in corresponding layers based on the microbial metabolism mode; S5. Generate an activity evaluation result according to the activity evaluation method and the first time series data set.

2. The wetland microbial activity monitoring method according to claim 1, characterized in that: The wetland status parameters include dissolved oxygen, redox potential, pH value, carbon dioxide content, methane content, hydrogen sulfide content, and nitrate content.

3. The wetland microbial activity monitoring method according to claim 1, characterized in that: The parameter change pattern includes the correlation and proportional relationship between different wetland state parameters in the same layer and the change rate of each wetland state parameter. The metabolic model is represented by a theoretical multidimensional feature vector composed of a theoretical correlation matrix, a theoretical change rate vector, and a theoretical proportional relationship set between the wetland state parameters. Step S3 includes: S31, extracting the correlation matrix, change rate vector, and proportional relationship set in the parameter change pattern to form a multidimensional feature vector; S32. Calculate the cosine similarity between the multidimensional feature vector and the theoretical multidimensional feature vectors of each metabolic model, and select the microbial metabolic mode corresponding to the theoretical multidimensional feature vector with the highest cosine similarity as the microbial metabolic mode of the corresponding layer.

4. The wetland microbial activity monitoring method according to claim 3, characterized in that: Step S32 includes: S321, constructing a KD tree index structure, and storing the theoretical multidimensional feature vectors of each metabolic model in the KD tree; S322. Using the fast search capability of the KD tree, search for N theoretical multidimensional feature vectors that are closest to the multidimensional feature vector, where N is a positive integer; S323, calculating the cosine similarity between the multidimensional feature vector and N theoretical multidimensional feature vectors; S324. Select a theoretical multidimensional feature vector corresponding to the maximum value of the cosine similarity value, and use the microbial metabolic mode corresponding to the theoretical multidimensional feature vector as the microbial metabolic mode of the corresponding layer.

5. The wetland microbial activity monitoring method according to claim 1, characterized in that: Step S5 further includes: S51, analyzing and filtering out parameter data of non-microbial metabolic patterns in the first time series data set to perform denoising on the first time series data set; S52 : Evaluate the first time series data set after denoising based on the activity evaluation method to generate an activity evaluation result.

6. The wetland microbial activity monitoring method according to claim 5, characterized in that: Step S4 includes: S41. For the determined microbial metabolic mode, query a preset mapping relationship table of microbial metabolic modes and activity evaluation indicators, and extract an evaluation mode corresponding to the microbial metabolic mode, wherein the evaluation mode includes at least one activity evaluation indicator, a data evaluation window, and a corresponding evaluation process.

7. The wetland microbial activity monitoring method according to claim 6, characterized in that: Step S52 includes: S521, extracting corresponding wetland state parameters from the first time series data set according to the evaluation process and the evaluation window to form a data set to be analyzed; S522: Processing the data set to be analyzed based on the evaluation process to generate an activity index value that matches the type of activity evaluation index; S523: Generate an activity evaluation result according to the activity index value and the corresponding activity evaluation index.

8. The wetland microbial activity monitoring method according to claim 5, characterized in that: Step S1 includes: S11. Acquire a first time series data set about different wetland state parameters at multiple different layers based on the monitoring component, and simultaneously acquire a second time series data set about environmental state parameters at multiple different layers based on the monitoring component; Step S5 also includes the following steps performed between step S51 and step S52: S5A, aligning the first time series data set and the second time series data set based on the timestamp, and based on the preset correction rules, compensating and correcting the wetland state parameters in the first time series data set of the corresponding layer according to the environmental parameters in the second time series data set.

9. The wetland microbial activity monitoring method according to claim 8, characterized in that: The environmental parameters are divided into temperature and conductivity.

10. A wetland microbial activity monitoring system for evaluating wetland microbial activity, characterized in that: The system comprises: An acquisition module, configured to acquire a first time series data set of different wetland state parameters at multiple different layers based on the acquisition of the monitoring component; a change pattern analysis module, configured to analyze and obtain parameter change patterns of each layer based on the first time series data set; a metabolic pattern analysis module for performing similarity matching based on the parameter change pattern with a pre-built parameter association model library to determine the microbial metabolic patterns at different layers, wherein the parameter association model library includes a plurality of metabolic models that match the microbial metabolic patterns and are associated with wetland state parameters; An evaluation method determination module, used to determine the activity evaluation method of the microorganisms in the corresponding layer based on the microbial metabolism method; An evaluation module is configured to generate an activity evaluation result according to the activity evaluation method and the first time series data set.

Images