Traceability management system for environmental monitoring data information

Abstract

The invention relates to the technical field of environmental monitoring data processing, in particular to a traceability management system for environmental monitoring data information, and solves the technical problem of low pollution abatement efficiency in the prior art. The system comprises a data acquisition unit used for acquiring pollution data of a plurality of environment monitoring points, the pollution data comprising pollutant types and pollutant concentrations at a plurality of moments; the pollution data processing unit is used for calculating the sudden similarity corresponding to each factory according to the pollution data of the first environment monitoring point which triggers the pollution alarm and the factory pollution discharge data of each factory in the candidate pollution source factory set under the condition that the pollution alarm is triggered; the sudden similarity is used for representing the matching degree between the factory pollution discharge data of the factory and the pollution data of the environment monitoring point; and the pollution source determining unit is used for screening a target pollution source factory from the candidate pollution source factory set according to the sudden similarity.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring data processing technology, and specifically to a traceability management system for environmental monitoring data information. Background Technology

[0002] With the advancement of industrialization, environmental problems have become increasingly prominent, making environmental monitoring a crucial means of detecting and stopping pollution. Environmental monitoring data covers various types of pollution, including river pollution, soil pollution, and air pollution. Collecting and analyzing this data to trace pollution sources is of great significance for environmental protection and sustainable development. Source tracing management systems can promptly detect new pollution and issue alerts based on environmental monitoring data, helping relevant personnel quickly reach the pollution source for intervention and preventing further spread of pollution.

[0003] In existing technologies, the main approach is to detect whether there are pollutant concentrations exceeding or abnormal in environmental monitoring data in real time. When such data is found, a pollution alarm is triggered, prompting staff to investigate potential pollution sources one by one. However, this approach is difficult to effectively locate pollution sources and has insufficient detection accuracy, resulting in low efficiency in pollution control. Summary of the Invention

[0004] To address the technical problem of low efficiency in existing pollution control technologies, the present invention aims to provide a source tracing and management system for environmental monitoring data. The specific technical solution adopted is as follows: This application provides a traceability management system for environmental monitoring data, including: The data acquisition unit is used to collect pollution data from multiple environmental monitoring points. The pollution data includes the types and concentrations of pollutants at multiple times. The pollution data processing unit is used to calculate the suddenness similarity of each factory based on the pollution data of the first environmental monitoring point that triggered the pollution alarm and the factory discharge data of each factory in the candidate pollution source factory set when a pollution alarm is triggered. The factory discharge data includes the types and concentrations of pollutants emitted by the factory at multiple times during operation. The suddenness similarity is used to characterize the degree of matching between the factory discharge data and the pollution data of the environmental monitoring point. The pollution source identification unit is used to screen target pollution source factories from the candidate pollution source factory set based on the similarity of suddenness.

[0005] In one possible implementation, the contaminated data processing unit is also used for: Based on the location of the first environmental monitoring point and the location relationship of multiple environmental monitoring points, an initial set of candidate pollution source factories is determined. Based on the types of pollutants with excessive concentrations in the pollution data, factories that emit pollutants with excessive concentrations are selected from the initial set of candidate pollution source factories to form a candidate pollution source factory set. The initial candidate pollution source factory set includes factories located between the first environmental monitoring point and the second environmental monitoring point. The first environmental monitoring point triggers a pollution alarm, while the second environmental monitoring point does not trigger a pollution alarm, and the first environmental monitoring point is located downstream of the second environmental monitoring point.

[0006] In one possible implementation, the contaminated data processing unit includes: The time series decomposition module is used to perform time series decomposition on the pollutant types at the first environmental monitoring point and the pollutant concentrations of each pollutant type emitted by each factory in the candidate pollution source factory cluster during operation at multiple time points, and obtain the time series decomposition results; the time series decomposition results are used to characterize the time-series variation characteristics of pollutant concentrations. The curve construction module is used to construct, based on the time series decomposition results, the actual monitoring concentration curves corresponding to the pollutant types at the first environmental monitoring point and the predicted emission concentration curves corresponding to the pollutant types emitted by each factory in the candidate pollution source factory cluster during operation. The similarity calculation module is used to calculate the sudden similarity of each factory based on the actual monitored concentration curve and the predicted discharge concentration curve.

[0007] In one possible implementation, the time series decomposition module includes: The sequence generation submodule is used to generate a pollutant concentration time series based on the pollutant types at multiple times and the pollutant concentrations at the first environmental monitoring point and the pollutant types emitted by each factory in the candidate pollution source factory cluster during operation. The sequence decomposition submodule is used to decompose the pollutant concentration time series into time series results. The time series decomposition results include the trend component, seasonal component, and residual component of the pollutant concentration time series.

[0008] In one possible implementation, the curve building module includes: The component proportion calculation submodule is used to calculate the trend component proportion and residual component proportion of each pollutant type in all pollutant types at each moment, based on the time series decomposition results corresponding to the pollutant types at the first environmental monitoring point and the pollutant types emitted by each factory in the candidate pollution source factory cluster during operation. The prediction submodule is used to predict the trend component proportion and residual component proportion of pollutant type in all pollutant types at each time point using a time series prediction algorithm. The curve construction submodule is used to construct the actual monitoring concentration curve and the predicted discharge concentration curve based on the trend component ratio and residual component ratio at future time and the most recent seasonal component at the current time.

[0009] In one possible implementation, the contaminated data processing unit further includes: The correlation calculation module is used to calculate the correlation between the types of pollutants at the first environmental monitoring point and the types of pollutants emitted by each factory in the candidate pollution source factory cluster.

[0010] In one possible implementation, the relevance calculation module includes: The first selection submodule is used to select multiple pollutant types from the pollutant types at the first environmental monitoring point to form a first pollutant type set; The second selection submodule is used to select multiple pollutant types from the pollutant types emitted by each factory in the candidate pollution source factory set to form a second pollutant type set; The correlation calculation submodule is used to calculate the correlation between the first pollutant type set and the second pollutant type set corresponding to each factory using a set similarity algorithm.

[0011] In one possible implementation, the similarity calculation module includes: The mutation point detection submodule is used to detect mutation points in the actual monitored concentration curve and each predicted discharge concentration curve through the mutation point detection algorithm, and generate the corresponding mutation point set. The distance calculation submodule is used to calculate the curve distance between the actual monitored concentration curve and each predicted discharge concentration curve, as well as the set distance between the sets of abrupt change points. The similarity calculation submodule is used to calculate the burst similarity of factories based on relevance, curve distance, and set distance for each factory.

[0012] In one possible implementation, the data acquisition unit is also used to preprocess the acquired pollution data; The preprocessing operations include: filling missing values ​​in contaminated data using a preset interpolation method, and removing outliers in contaminated data using a preset outlier detection method.

[0013] In one possible implementation, the pollution source determination unit is specifically used for: The factory with the highest degree of sudden similarity among the candidate pollution source factories is selected as the target pollution source factory.

[0014] The present invention has the following beneficial effects: In view of the technical problem of low efficiency in pollution control in existing technologies, this application provides a source tracing management system for environmental monitoring data. The data acquisition unit can comprehensively collect pollutant type and concentration data from multiple monitoring points and at multiple times, providing a rich data foundation for source tracing analysis. The pollution data processing unit quantifies the matching degree between factory discharge data and monitoring data by calculating the suddenness similarity, avoiding the limitations of single-dimensional analysis. The pollution source identification unit screens target pollution sources based on suddenness similarity, ensuring the relevance of the source tracing results. The above scheme forms a complete link from data acquisition to pollution source identification, effectively solving the problems of one-sided and inaccurate source tracing analysis in existing technologies, improving the efficiency and accuracy of pollution source tracing, further improving pollution control efficiency, and providing strong support for rapid pollution control and ecological environment protection. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a system architecture diagram of an environmental monitoring data traceability management system provided in one embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a contamination data processing unit provided in one embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a time series decomposition module provided in one embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a curve construction module provided in one embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of another contamination data processing unit provided in one embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a correlation calculation module provided in one embodiment of the present invention; Figure 7 This is a schematic diagram of the structure of a similarity calculation module provided in one embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an environmental monitoring data traceability management system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] In view of the technical problem of low efficiency in pollution control in existing technologies, this application provides a source tracing management system for environmental monitoring data. The data acquisition unit can comprehensively collect pollutant type and concentration data from multiple monitoring points and at multiple times, providing a rich data foundation for source tracing analysis. The pollution data processing unit quantifies the matching degree between factory discharge data and monitoring data by calculating the suddenness similarity, avoiding the limitations of single-dimensional analysis. The pollution source identification unit screens target pollution sources based on suddenness similarity, ensuring the relevance of the source tracing results. The above scheme forms a complete link from data acquisition to pollution source identification, effectively solving the problems of one-sided and inaccurate source tracing analysis in existing technologies, improving the efficiency and accuracy of pollution source tracing, further improving pollution control efficiency, and providing strong support for rapid pollution control and ecological environment protection.

[0020] The specific solution of the environmental monitoring data traceability management system provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0021] Please see Figure 1 The diagram illustrates a system architecture of an environmental monitoring data traceability management system 10 provided by an embodiment of the present invention. The environmental monitoring data traceability management system 10 includes a data acquisition unit 11, a pollution data processing unit 12, and a pollution source determination unit 13.

[0022] The data acquisition unit 11 is used to collect pollution data from multiple environmental monitoring points. The pollution data includes the types and concentrations of pollutants at multiple times.

[0023] In some embodiments, environmental monitoring points can be rationally deployed according to different monitoring scenarios. For example, for river pollution monitoring, a monitoring point can be set up at certain intervals (e.g., 1 kilometer) along the river from upstream to downstream. In addition to collecting pollution data such as pollutant types and concentrations, auxiliary data such as monitoring time, water flow velocity, and flow rate can also be collected simultaneously, as well as information on the location of the source factory and the concentration of pollutants emitted by the factory, providing a more comprehensive reference for subsequent source tracing analysis.

[0024] For example, the data acquisition unit 11 can collect data in real time through devices such as sensors and water quality analyzers, and upload the data to the system database for storage via wireless transmission devices. The data can also be stored as historical records.

[0025] In some embodiments, the data acquisition unit 11 is further configured to perform preprocessing operations on the acquired pollution data.

[0026] The preprocessing operations include: filling missing values ​​in contaminated data using a preset interpolation method, and removing outliers from contaminated data using a preset outlier detection method.

[0027] For example, the preset interpolation method can be linear interpolation, which is simple to operate, computationally efficient, and suitable for situations where there are few missing values ​​in the pollution data. When pollutant concentration data at a certain moment is missing, the missing value is calculated using the effective concentration data at adjacent moments before and after that moment through linear interpolation, ensuring the continuity of the time series.

[0028] The preset outlier detection method can adopt 3 The criterion, based on the normal distribution of data, identifies outliers and removes data points when their deviation from the mean exceeds three times the standard deviation. Outliers may be caused by factors such as equipment failure or sudden interference; removing them prevents them from adversely affecting subsequent time series decomposition, curve construction, and similarity calculation.

[0029] The pollution data processing unit 12 is used to calculate the suddenness similarity of each factory based on the pollution data of the first environmental monitoring point that triggered the pollution alarm and the factory discharge data of each factory in the candidate pollution source factory set when a pollution alarm is triggered.

[0030] Among them, the factory's wastewater discharge data includes the types and concentrations of pollutants emitted by the factory at multiple times during operation, and the suddenness similarity is used to characterize the degree of matching between the factory's wastewater discharge data and the pollution data of environmental monitoring points.

[0031] In some embodiments, the triggering condition for a pollution alarm can be determined based on a preset pollutant concentration threshold. When the concentration of one or more pollutants collected at an environmental monitoring point exceeds the corresponding pollutant concentration threshold, a pollution alarm is triggered. This environmental monitoring point is designated as the first environmental monitoring point. Factory discharge data can be collected and stored in the system through prior investigation, including the factory's production type, historical pollutant types, and corresponding concentration time-series data, ensuring that it can be quickly retrieved for analysis when needed.

[0032] It should be noted that different factories produce different products and use different processes, resulting in variations in the pollutant content of their wastewater. Even if multiple factories have wastewater containing similar types of pollutants, their composition and proportions are often different. Therefore, this application can accurately locate the source of pollution by analyzing the similarity between each factory and the currently detected pollution data to assess the corresponding suddenness of pollution.

[0033] The pollution source identification unit 13 is used to screen target pollution source factories from the candidate pollution source factory set based on the similarity of suddenness.

[0034] In some embodiments, the pollution source determination unit 13 is specifically used to: select the factory with the highest sudden similarity among candidate pollution source factories as the target pollution source factory.

[0035] For example, the pollution source identification unit 13 compares the suddenness similarity of each factory calculated by the pollution data processing unit 12, and selects the factory with the highest suddenness similarity as the target pollution source factory. This is because the factory with the highest suddenness similarity has the highest correlation between its discharge data and the pollution data of the first environmental monitoring point in terms of pollutant type matching, concentration change trend, and abrupt change characteristics, and therefore has the highest probability of becoming an actual pollution source. This application can push relevant information about the factory (such as address, person in charge, contact information, etc.) to relevant personnel for timely handling.

[0036] Based on the above technical solution, the data acquisition unit 11 in this application can comprehensively collect pollutant type and concentration data from multiple monitoring points and at multiple times, providing a rich data foundation for source tracing analysis. The pollution data processing unit 12 quantifies the matching degree between factory discharge data and monitoring data by calculating the suddenness similarity, avoiding the limitations of single-dimensional analysis. The pollution source determination unit 13 screens target pollution sources based on suddenness similarity, ensuring the pertinence of the source tracing results. The above solution forms a complete link from data acquisition to pollution source determination, effectively solving the problems of one-sided and inaccurate results in the source tracing analysis of existing technologies, improving the efficiency and accuracy of pollution source tracing, further improving the efficiency of pollution control, and providing strong support for rapid pollution control and ecological environment protection.

[0037] In addition, in order to further narrow down the scope of candidate pollution source factories, reduce the amount of subsequent calculations, and improve the efficiency of source tracing, this application can also conduct preliminary screening of candidate pollution source factories through environmental monitoring points.

[0038] As one possible embodiment of this application, the contamination data processing unit 12 is further configured to: Based on the location of the first environmental monitoring point and the location relationship of multiple environmental monitoring points, an initial set of candidate pollution source factories is determined. Then, based on the types of pollutants with excessive concentrations in the pollution data, factories that emit pollutants with excessive concentrations are selected from the initial set of candidate pollution source factories to form a candidate pollution source factory set.

[0039] The initial candidate pollution source factory set includes factories located between the first environmental monitoring point and the second environmental monitoring point. The first environmental monitoring point triggers a pollution alarm, while the second environmental monitoring point does not trigger a pollution alarm, and the first environmental monitoring point is located downstream of the second environmental monitoring point.

[0040] For example, the second environmental monitoring point is the closest environmental monitoring point located upstream of the first environmental monitoring point that has not triggered a pollution alarm. Since pollutants diffuse in a certain direction in environmental media such as water and air, when the downstream environmental monitoring point triggers an alarm while the adjacent upstream environmental monitoring point does not, it can be inferred that the pollution source is located in the area between these two environmental monitoring points, thereby determining the initial set of candidate pollution source factories. This avoids analyzing all factories one by one and significantly narrows the scope of the investigation.

[0041] In some embodiments, this application can pre-store the location information of all environmental monitoring points and the distribution of factories along the route. When the first environmental monitoring point triggers an alarm, it retrieves the nearest upstream environmental monitoring point that has not triggered an alarm (i.e., the second environmental monitoring point) and extracts all factories between these two environmental monitoring points as an initial candidate pollution source factory set. Subsequently, based on the types of pollutants exceeding the concentration standards in the pollution data of the first environmental monitoring point, it compares the historical discharge data of each factory in the initial candidate pollution source factory set, screens out factories that discharge such pollutants exceeding the standards, and forms the final candidate pollution source factory set, ensuring that subsequent analysis is only conducted on factories with potential pollution suspicion.

[0042] For example, the initial candidate pollution source factory set is ,include One factory, This application indicates that the first factory can be selected from the initial candidate pollution source factories based on the type of pollutant exceeding the concentration standard. The process involves screening factories that may be discharging this pollutant. Specifically, if a factory's wastewater contains pollutants at concentrations exceeding the standard, that factory is likely discharging that type of pollutant, thus initially identifying a cluster of candidate pollution source factories. , .

[0043] Based on the above technical solution, this application screens a set of candidate pollution source factories in two steps by combining the location relationship of monitoring points and the types of pollutants exceeding standards. First, the alarm status of upstream and downstream monitoring points is used to determine the initial range, eliminating obviously impossible areas. Then, further screening is conducted based on the types of pollutants exceeding standards to ensure that all candidate factories have the potential to emit such pollutants. This screening method effectively narrows the analysis scope, reduces the computational load of subsequent data processing, and improves the efficiency of the entire source tracing process. It also makes subsequent similarity calculations more targeted, laying the foundation for improving the accuracy of the source tracing results.

[0044] As one possible embodiment of this application, combined with Figure 1 ,like Figure 2 As shown, the pollution data processing unit 12 includes: a time series decomposition module 121, a curve construction module 122, and a similarity calculation module 123.

[0045] The time series decomposition module 121 is used to perform time series decomposition on the pollutant types at the first environmental monitoring point and the pollutant concentrations of each factory in the candidate pollution source factory cluster during operation at multiple times, so as to obtain the time series decomposition results.

[0046] The time series decomposition results are used to characterize the temporal variation of pollutant concentrations.

[0047] In some embodiments, the time series decomposition module 121 targets representative pollutant types (such as multiple pollutants with the highest ratio of monitored concentration to pollutant concentration threshold) and corresponding pollutant types emitted by candidate factories at the first environmental monitoring point. It decomposes the concentration data into different components through a time series decomposition algorithm, thereby extracting the trend, periodicity and other features of concentration changes, providing a basis for subsequent curve construction.

[0048] The curve construction module 122 is used to construct, based on the time series decomposition results, the actual monitoring concentration curves corresponding to the pollutant types at the first environmental monitoring point and the predicted emission concentration curves corresponding to the pollutant types emitted by each factory in the candidate pollution source factory cluster during operation.

[0049] The actual monitoring concentration curve is constructed based on the decomposition of real concentration data collected at the first environmental monitoring point, and can accurately reflect the temporal changes in pollutant concentrations at the monitoring point. The predicted discharge concentration curve is predicted based on the decomposition of historical discharge data from the factory and combined with the factory's discharge characteristics, and is used to simulate the changing patterns of pollutant concentrations after the factory discharges pollutants.

[0050] The similarity calculation module 123 is used to calculate the sudden similarity of each factory based on the actual monitoring concentration curve and the predicted discharge concentration curve.

[0051] In some embodiments, the similarity calculation module 123 quantifies the degree of matching between the two curves by comparing their overall trends, change nodes, and other features, thereby obtaining a sudden similarity. The higher the degree of curve matching, the stronger the correlation between the factory's pollution discharge behavior and the pollution situation at the monitoring point, and the greater the likelihood that it is a pollution source.

[0052] Based on the above technical solution, the time series decomposition module 121 in this application can mine the temporal variation characteristics of pollutant concentration, avoiding the limitations of analysis based solely on data from a single moment. The curve construction module 122 transforms the decomposition results into intuitive concentration curves for subsequent similarity comparison. The similarity calculation module 123 quantifies the degree of matching through curve comparison, ensuring the scientific rigor and accuracy of sudden similarity calculations. The above solution effectively improves the accuracy of sudden similarity calculations and further enhances the source tracing reliability of the entire system.

[0053] As one possible embodiment of this application, combined with Figure 2 ,like Figure 3 As shown, the time series decomposition module 121 includes a sequence generation submodule 1211 and a sequence decomposition submodule 1212.

[0054] The sequence generation submodule 1211 is used to generate a pollutant concentration time series based on the pollutant types at multiple times and the pollutant concentrations at multiple times, according to the pollutant types at the first environmental monitoring point and the pollutant types emitted by each factory in the candidate pollution source factory cluster during operation.

[0055] For ease of calculation, this application can sort pollutant concentrations from largest to smallest according to the ratio of each pollutant concentration to its corresponding pollutant concentration threshold, and select a preset number (e.g., 20) of pollutant types to generate a pollutant concentration time series. For example, the pollutant concentration time series generated for the first environmental monitoring point can be represented as follows: That is, the first Each type of pollutant at time The collected pollutant concentrations, the types of pollutants emitted by the factory during operation, and the time series of pollutant concentrations can be represented as follows: That is, the candidate pollution source factory group The first factory's emissions Each type of pollutant at time The concentration of pollutants collected.

[0056] For example, the pollutant concentration time series is formed by arranging concentration data collected at different times in chronological order, with time as the horizontal axis and pollutant concentration as the vertical axis. Time can be in hours, and the historical emission periods of each factory are aligned relative to the time scale of the monitoring point. For the first environmental monitoring point, the time range of the time series can cover a period after the alarm is triggered (e.g., from the start of the alarm to 24 hours after the trigger); for candidate factories, the time range of the time series can be selected from emission periods under similar production conditions in their history to ensure data comparability.

[0057] The sequence decomposition submodule 1212 is used to perform time series decomposition on the pollutant concentration time series to obtain the time series decomposition results.

[0058] The time series decomposition results include the trend component, seasonal component, and residual component of the pollutant concentration time series.

[0059] In some embodiments, the sequence decomposition submodule 1212 may employ a time series decomposition algorithm (seasonal and trend decomposition using loess, STL), which can effectively separate the trend component (reflecting the long-term trend of concentration changes), the seasonal component (reflecting the periodic characteristics of concentration changes), and the residual component (reflecting concentration fluctuations caused by random factors) in a time series. Through this decomposition, the inherent laws of pollutant concentration changes can be more clearly identified, providing more valuable feature information for subsequent curve construction and similarity calculation.

[0060] For example, the relationship between the pollutant concentration time series and its individual components can be expressed by the following formula: in, This is a pollutant concentration time series, specifically a pollutant concentration time series generated from the pollutant types at the first environmental monitoring point mentioned above. Or the time series of pollutant concentrations generated by the types of pollutants emitted during factory operations. , As a trend component, For seasonal portions, These are the residual components. For ease of distinction, the first... The first factory The trend components of each pollutant type can be represented as: The first environmental monitoring point The trend components of each pollutant type can be represented as: , No. The first factory The seasonal component of each pollutant type can be expressed as: The first environmental monitoring point The seasonal component of each pollutant type can be expressed as: , No. The first factory The residual components of each pollutant type can be expressed as: The first environmental monitoring point The residual components of each pollutant type can be expressed as: .

[0061] Based on the above technical solution, the sequence generation submodule 1211 in this application transforms discrete concentration data into structured time series, providing a standardized data format for decomposition operations. The sequence decomposition submodule 1212 decomposes the pollutant concentration time series to obtain three components: trend, seasonality, and residual, comprehensively exploring the temporal variation characteristics of pollutant concentrations and avoiding analytical biases caused by mutual interference of various factors in the original data. This solution enables subsequent curve construction based on more accurate feature components, helping to improve the matching degree between predicted discharge concentration curves and actual monitored concentration curves, thereby improving the accuracy of sudden similarity calculations and providing more reliable technical support for precise source tracing.

[0062] As one possible embodiment of this application, combined with Figure 3 ,like Figure 4 As shown, the curve construction module 122 includes: a component proportion calculation submodule 1221, a prediction submodule 1222, and a curve construction submodule 1223.

[0063] The component proportion calculation submodule 1221 is used to calculate the trend component proportion and residual component proportion of each pollutant type in all pollutant types at each moment, based on the time series decomposition results corresponding to the pollutant types at the first environmental monitoring point and the pollutant types emitted by each factory in the candidate pollution source factory cluster during operation.

[0064] Since the discharge time and concentration of industrial wastewater from factories typically exhibit periodic variations, and the composition of wastewater discharged from the same factory is generally consistent (i.e., the types and proportions of pollutants remain relatively constant), this application can extract the discharge characteristics of each factory by analyzing the proportions of trend components and residual components. The trend component proportion refers to the ratio of the trend component of a specific pollutant at a given moment to the sum of the trend components of all pollutants at that moment, while the residual component proportion is the ratio of the residual component of that pollutant to the sum of the residual components of all pollutants at that moment. Because the types and proportions of pollutants in wastewater discharged from the same factory are usually stable, calculating these proportions can eliminate the influence of differences in total discharge from different factories, highlighting the characteristics of pollutant composition proportions. This is a key feature for distinguishing the discharge characteristics of different factories.

[0065] For example, taking 20 types of pollutants as an example, the trend component proportion satisfies the following formula: in, For the first The trend component proportion of each pollutant type For the first Trend components for each pollutant type.

[0066] The prediction submodule 1222 is used to predict the trend component proportion and residual component proportion of future times based on the trend component proportion and residual component proportion of each pollutant type among all pollutant types at each time moment using a time series prediction algorithm.

[0067] In some embodiments, the time series forecasting algorithm may employ an autoregressive integrated moving average (ARIMA) model, which can accurately predict future data based on the trends and patterns of historical data. Model parameters (initial core parameters: autoregressive order p, differencing order d, and moving average order q) can be set separately for the trend component proportion and the residual component proportion. For example, the parameters for predicting the trend component proportion can be set to (1,1,1), and the parameters for predicting the residual component proportion can be set to (1,0,1), outputting the predicted values ​​of each component proportion at different times.

[0068] The curve construction submodule 1223 is used to construct the actual monitoring concentration curve and the predicted discharge concentration curve based on the trend component ratio and residual component ratio at future times and the most recent seasonal component at the current time.

[0069] For example, taking the predicted wastewater concentration curve as an example, the predicted wastewater concentration curve satisfies the following formula: in, For the first The first factory Predicted emission concentration curves for each pollutant type. For the predicted first The first factory Each type of pollutant at time The trend component proportion, For the first The first factory Trend components of each pollutant type For the first The first factory The most recent seasonal component of each pollutant type at the current time. For the predicted first The first factory Each type of pollutant at time The proportion of residual components, For the first The first factory The residual components for each pollutant type, This represents the function that takes the maximum value. This indicates the period from the detection of abnormal pollutant concentrations in the river to the point after the anomaly. Time period within an hour (in terms of) For example, that is, from arrive The predicted pollutant concentration curves were constructed based on the prediction results.

[0070] Based on the above technical solution, the component proportion calculation submodule 1221 in this application extracts the compositional proportion characteristics of pollutants, capturing the core differences in emissions from different factories. The prediction submodule 1222 utilizes a time series prediction algorithm to ensure the accuracy of future proportion predictions. The curve construction submodule 1223 integrates trend, seasonality, and residual components to construct a curve that truly reflects changes in pollutant concentration. This solution, based on the temporal variation patterns of pollutant concentrations and the characteristic stability of factory emissions, ensures high comparability between the two constructed curves, providing a solid foundation for accurate calculation of subsequent sudden similarity and further improving the system's source tracing accuracy.

[0071] As one possible embodiment of this application, combined with Figure 2 ,like Figure 5 As shown, the pollution data processing unit 12 also includes a correlation calculation module 124.

[0072] The correlation calculation module 124 is used to calculate the correlation between the types of pollutants at the first environmental monitoring point and the types of pollutants emitted by each factory in the candidate pollution source factory cluster.

[0073] The correlation is used to characterize the degree of overlap between the types of pollutants detected at the first environmental monitoring point and the types of pollutants emitted by the factory. The higher the correlation, the stronger the match between the pollutants emitted by the factory and the pollutants detected at the monitoring point, and the greater the likelihood that the factory is a source of pollution.

[0074] In some embodiments, combined with Figure 5 ,like Figure 6 As shown, the relevance calculation module 124 includes: a first selection submodule 1241, a second selection submodule 1242, and a relevance calculation submodule 1243.

[0075] The first selection submodule 1241 is used to select multiple pollutant types from the pollutant types at the first environmental monitoring point to form a first pollutant type set.

[0076] In some embodiments, the first selection submodule 1241 can sort pollutant types from largest to smallest according to the ratio of pollutant concentration to the corresponding pollutant concentration threshold, and preferentially select multiple pollutant types with the highest concentration ranking (such as the top 20) to form a first pollutant type set. This is because the ratio of pollutant concentration to the corresponding pollutant concentration threshold can reflect the emission level of the pollutant type and the emission characteristics of the pollution source, and selecting a portion of representative pollutants can reduce the amount of computation, improve the efficiency of correlation calculation, and avoid computational redundancy caused by including too many low-concentration, unrepresentative pollutant types.

[0077] For example, the set of pollutant types at the first environmental monitoring point is as follows: Each pollutant type is sorted from largest to smallest according to the ratio of its concentration to the corresponding pollutant concentration threshold. Taking 20 types as an example, the resulting first set of pollutant types is... .

[0078] The second selection submodule 1242 is used to select multiple pollutant types from the pollutant types emitted by each factory in the candidate pollution source factory set to form a second pollutant type set.

[0079] Similarly, the second selection submodule 1242 can select multiple pollutant types (such as the top 20) with the highest ratio of pollutant concentration to corresponding pollutant concentration threshold based on the pollutant concentration data in the candidate factory's historical discharge data to form a second pollutant type set, ensuring that the selection criteria are consistent with the first pollutant type set and guaranteeing the fairness and comparability of the correlation calculation.

[0080] For example, taking 20 types as an example, the set of second pollutant types can be represented as follows: .

[0081] The correlation calculation submodule 1243 is used to calculate the correlation between the first pollutant type set and the second pollutant type set corresponding to each factory through a set similarity algorithm.

[0082] In some embodiments, the set similarity algorithm can employ the MinHash algorithm, which can efficiently calculate the similarity between two sets, and is particularly suitable for rapid comparison of large-scale data. The MinHash algorithm calculates the ratio of the intersection size to the union size of the first and second pollutant type sets to obtain a correlation value. The value ranges from 0 to 1; the closer it is to 1, the higher the similarity between the two sets and the stronger the pollutant type matching.

[0083] For example, relevance satisfies the following formula: in, The first set of pollutant types and the second The correlation between the sets of second pollutant types in each factory This is the set of the first pollutant categories. For the first The second set of pollutant types from each factory It is the minimum hash function.

[0084] Based on the above technical solution, this application selects representative pollutant types to form a set through the first selection submodule 1241 and the second selection submodule 1242. This ensures the relevance calculation is targeted and improves computational efficiency. The relevance calculation submodule 1243 uses a set similarity algorithm to ensure the accuracy and reliability of the relevance results. The above solution can objectively reflect the matching of pollutant types, providing unified and accurate input parameters for the subsequent comprehensive calculation of sudden similarity, further enhancing the scientific nature and operability of the entire source tracing system.

[0085] As one possible embodiment of this application, combined with Figure 2 ,like Figure 7 As shown, the similarity calculation module 123 includes: a mutation point detection submodule 1231, a distance calculation submodule 1232, and a similarity calculation submodule 1233.

[0086] The mutation point detection submodule 1231 is used to detect mutation points in the actual monitored concentration curve and each predicted discharge concentration curve through the mutation point detection algorithm, and generate the corresponding mutation point set.

[0087] A mutation point refers to a point in time where the pollutant concentration in the curve changes significantly. For example, when a factory starts discharging pollutants, the pollutant concentration at a monitoring point will suddenly increase, forming a mutation point. If a factory suddenly starts discharging pollutants, the pollutant concentration in the river will suddenly increase, causing a mutation in the actual monitoring concentration curve of the river pollutants. As the discharge time increases, the pollutant concentration in the river gradually increases, and the concentration at the monitoring point may gradually dilute due to the influence of water flow. The pollutant concentration values ​​at the monitoring point change significantly throughout the discharge process. Since the mutation patterns of each pollutant concentration in the actual monitoring concentration curve are consistent with the mutation patterns in the predicted discharge concentration curves of each pollutant discharged by each factory, the most likely source of pollution can be traced based on the similarity of mutations between the actual monitoring concentration curve and the predicted discharge concentration curve.

[0088] For example, the mutation point detection algorithm can employ the Mann-Kendall (MK) mutation point detection algorithm. During the detection process, the significance level can be set to 0.05. The algorithm outputs all significant mutation points in the curve, forming a mutation point set. This set includes the corresponding time information for each mutation point, and the mutation points in the set are ordered according to the chronological order of their corresponding time information. The mutation point set corresponding to the actual monitored concentration curve can be represented as follows: , Indicates the first The first abrupt change point in the actual monitored concentration curve of each pollutant species. This represents the number of abrupt change points in the actual monitored concentration curve. The set of abrupt change points corresponding to the predicted wastewater concentration curve can be represented as... , Indicates the first The first factory's emissions The first abrupt change point in the predicted emission concentration curves for each pollutant type. To predict the number of abrupt change points in the wastewater concentration curve.

[0089] The distance calculation submodule 1232 is used to calculate the curve distance between the actual monitored concentration curve and each predicted discharge concentration curve, as well as the set distance between the sets of abrupt change points.

[0090] In some embodiments, the curve distance and set distance can be calculated using the dynamic time warping (DTW) algorithm. The smaller the DTW distance, the closer the overall trends of the two curves are and the more consistent their abrupt change characteristics are.

[0091] The similarity calculation submodule 1233 is used to calculate the burst similarity of factories based on relevance, curve distance and set distance for each factory.

[0092] Among them, relevance is positively correlated with sudden similarity, while curve distance and set distance are negatively correlated with sudden similarity.

[0093] For example, burst similarity satisfies the following formula: in, For the first The burst similarity corresponding to each factory For the first The relevance of each factory to the corresponding number of factories The number of pollutant types, For the first The first factory Predicted emission concentration curves for each pollutant type. For the first Actual monitoring concentration curves for each pollutant type For the first The first factory The set of abrupt change points corresponding to the predicted emission concentration curves of each pollutant type. For the first The set of abrupt change points corresponding to the actual monitored concentration curves of each pollutant species. This is the DTW distance function. Characterizing the difference between the actual monitored concentration curve and the predicted discharge concentration curve, Characterizing the difference between the abrupt change points in the actual monitored concentration curve and the abrupt change points in the predicted wastewater concentration curve, and Multiplication is used to combine these two differences to obtain a combined difference, which is then negatively correlated to obtain the burst similarity.

[0094] This is a normalization function used to normalize the product of the curve distance and the set distance to between 0 and 1. The more similar the predicted emission concentration curve and the actual monitored concentration curve are overall, and the more similar their abrupt change points, the smaller the curve distance. The larger the value, the more significant the effect. The higher the sudden similarity between the predicted emission concentration curve and the actual monitored concentration curve of a factory, the more likely it is to occur. The more likely a factory is to be a source of pollution, the greater the likelihood of it being one.

[0095] Based on the above technical solution, this application uses a mutation point detection submodule 1231 to capture the mutation characteristics of the curve, a distance calculation submodule 1232 to measure the matching degree of the overall trend of the curve and the mutation characteristics, and a similarity calculation submodule 1233 to integrate correlation and two distance parameters, comprehensively considering three key factors: matching of pollutant types, consistency of concentration change trends, and consistency of mutation characteristics. This similarity calculation method avoids the limitations of single-dimensional analysis, enabling sudden similarity to more comprehensively and accurately reflect the correlation between factory emissions and monitoring data, significantly improving the accuracy of pollution source identification and effectively reducing the risk of misjudgment.

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

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

Claims

1. A provenance management system for environmental monitoring data information, characterized in that, The method comprises: a data collection unit configured to collect pollution data of a plurality of environmental monitoring points, the pollution data comprising a plurality of pollution types and pollution concentrations at different times; a pollution data processing unit configured to, in the case of triggering a pollution alarm, calculate a burst similarity of each factory in a candidate pollution source factory set according to pollution data of a first environmental monitoring point triggering the pollution alarm and factory pollution data of each factory in the candidate pollution source factory set; wherein the factory pollution data comprises pollution types and pollution concentrations emitted by the factory at different times during operation, and the burst similarity is used to represent a matching degree between the factory pollution data and the pollution data of the environmental monitoring point; a pollution source determination unit configured to filter a target pollution source factory from the candidate pollution source factory set according to the burst similarity.

2. The provenance management system of environmental monitoring data information according to claim 1, characterized in that, The pollution data processing unit is further configured to: determine an initial candidate pollution source factory set according to a position of the first environmental monitoring point and a positional relationship of the plurality of environmental monitoring points, and filter factories emitting a pollution type with a concentration exceeding a standard from the initial candidate pollution source factory set to form the candidate pollution source factory set; wherein the initial candidate pollution source factory set comprises factories located between the first environmental monitoring point and a second environmental monitoring point, the first environmental monitoring point triggers the pollution alarm, the second environmental monitoring point does not trigger the pollution alarm, and the first environmental monitoring point is located downstream of the second environmental monitoring point.

3. The provenance management system of environmental monitoring data information according to claim 1, characterized in that, The pollution data processing unit comprises: a time series decomposition module configured to respectively perform time series decomposition on pollution concentrations of pollution types of the first environmental monitoring point and pollution types emitted by each factory in the candidate pollution source factory set during operation at different times to obtain time series decomposition results; the time series decomposition results are used to represent variation characteristics of the pollution concentrations in time sequence; a curve construction module configured to respectively construct actual monitoring concentration curves of the pollution types of the first environmental monitoring point and predicted factory emission concentration curves of the pollution types emitted by each factory in the candidate pollution source factory set during operation according to the time series decomposition results; a similarity calculation module configured to respectively calculate the burst similarity of each factory according to the actual monitoring concentration curves and the predicted factory emission concentration curves.

4. The provenance management system of environmental monitoring data information according to claim 3, characterized in that, The time series decomposition module comprises: a sequence generation sub-module configured to generate pollution concentration time sequences of pollution types of the first environmental monitoring point and pollution types emitted by each factory in the candidate pollution source factory set during operation according to pollution concentrations of the pollution types at different times; a sequence decomposition sub-module configured to perform time series decomposition on the pollution concentration time sequences to obtain the time series decomposition results; the time series decomposition results comprise trend components, seasonal components and residual components of the pollution concentration time sequences.

5. The provenance management system of environmental monitoring data information according to claim 4, characterized in that, The curve construction module comprises: The component proportion calculation submodule is configured to calculate, for each time point, a trend component proportion and a residual component proportion of the pollutant category in all pollutant categories based on a time series decomposition result corresponding to the pollutant category. The prediction submodule is configured to predict, by a time series prediction algorithm, the trend component proportion and the residual component proportion at a future time point according to the trend component proportion and the residual component proportion of the pollutant category in all pollutant categories at each time point. The curve construction submodule is configured to construct an actual monitoring concentration curve and a predicted emission concentration curve according to the trend component proportion and the residual component proportion at the future time point and the nearest seasonal component at the current time point.

6. The provenance management system of environmental monitoring data information according to claim 3, characterized in that, The pollution data processing unit further includes: The correlation calculation module is configured to calculate a correlation between the pollutant category of the first environmental monitoring point and the pollutant category emitted by each factory in the candidate pollution source factory set.

7. The provenance management system of environmental monitoring data information according to claim 6, characterized in that, The correlation calculation module includes: The first selection submodule is configured to select a plurality of pollutant categories from the pollutant categories of the first environmental monitoring point to form a first pollutant category set. The second selection submodule is configured to select a plurality of pollutant categories from the pollutant categories emitted by each factory in the candidate pollution source factory set to form a second pollutant category set. The correlation calculation submodule is configured to calculate, by a set similarity algorithm, a correlation between the first pollutant category set and the second pollutant category set corresponding to each factory.

8. The provenance management system of environmental monitoring data information according to claim 6, characterized in that, The similarity calculation module includes: The mutation point detection submodule is configured to detect, by a mutation point detection algorithm, a mutation point in the actual monitoring concentration curve and each predicted emission concentration curve to generate a corresponding mutation point set. The distance calculation submodule is configured to calculate a curve distance between the actual monitoring concentration curve and each predicted emission concentration curve and a set distance between the mutation point sets. The similarity calculation submodule is configured to calculate, for each factory, a burst similarity of the factory according to the correlation, the curve distance, and the set distance.

9. The provenance management system of environmental monitoring data information according to claim 1, characterized in that, The data acquisition unit is further configured to perform a preprocessing operation on the acquired pollution data. The preprocessing operation includes filling in missing values in the pollution data by using a preset interpolation method and eliminating abnormal values in the pollution data by using a preset abnormal value detection method.

10. The provenance management system of environmental monitoring data information according to claim 1, characterized in that, The pollution source determination unit is specifically configured to: Take the factory with the largest burst similarity in the candidate pollution source factory set as the target pollution source factory.

Images