A multi-source heterogeneous sensor collaborative industrial carbon pollution emission online monitoring and tracing system

Abstract

This invention relates to the field of carbon pollution emission source tracing technology, specifically disclosing a multi-source heterogeneous sensor collaborative online monitoring and source tracing system for industrial carbon pollution emissions. Based on a diffusion model, this invention analyzes the initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration at the monitoring site. This initial contribution ratio is input into a receptor model for iterative calculation to obtain the final contribution ratio, which is then used to generate a source tracing analysis report. This collaborative iterative method combining diffusion and receptor models achieves joint source tracing and accurate decomposition of contribution ratios for carbon emissions and pollutant emissions. Furthermore, this invention periodically calculates the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, and the average equipment availability coefficient, and weights them to obtain an operational quality index. This index is compared with a preset threshold to determine whether the system's operational quality meets expectations for the current period, thus achieving continuous monitoring and quantitative evaluation of the system's own operational quality.

Description

Technical Field

[0001] This invention relates to the field of carbon pollution emission tracing technology, and in particular to a multi-source heterogeneous sensing collaborative online monitoring and tracing system for industrial carbon pollution emissions. Background Technology

[0002] With the continuous expansion of industrial scale, key industries such as chemicals, steel, and building materials have become the core sources of carbon emissions and pollutant emissions, making precise control of carbon emissions crucial for environmental governance.

[0003] An existing method for online monitoring and source tracing of industrial carbon emissions involves deploying monitoring points around industrial parks or key pollution source areas, installing multi-parameter online monitoring equipment including CO2, and collecting data on pollutant concentrations, carbon emission concentrations, meteorological parameters (wind speed, wind direction, temperature, humidity, etc.), and geographic information. The collected data from multiple sources undergoes preliminary screening, noise reduction, and spatiotemporal matching. A dynamic carbon emission distribution model is established, integrating the WRF-STILT atmospheric transport and diffusion model with a Bayesian inversion algorithm to achieve dynamic inversion and prediction of carbon emissions. The actual monitoring data of each emission source within the region is compared with the model's prediction results. If the data falls within the predicted range, the emission activity is considered normal; significant differences indicate abnormal emissions. A wireless sensor network is used to analyze and simulate the trajectory and changes of pollutant plumes, identifying and tracing the source of abnormal emissions from the monitoring nodes. Combined with a GIS trajectory tracing system, a real-time pollution concentration cloud map and a dynamic model of the diffusion path are generated to locate the abnormal emission area.

[0004] Existing methods possess advantages such as strong multi-source data fusion capabilities, outstanding source tracing capabilities, and the ability to achieve real-time dynamic monitoring. However, existing methods still have some shortcomings: First, they suffer from high equipment costs and insufficient representativeness of monitoring points. Direct measurement methods have significant limitations such as high equipment costs and insufficient representativeness of monitoring points. The equipment is often expensive, and there are limitations in distinguishing carbon-based solid particles. Relying on a limited number of monitoring points for discrete data collection makes it difficult to meet the needs of comprehensive and accurate emission monitoring. Second, they lack the ability to trace the source of greenhouse gases and pollutants in a coordinated manner, and it is difficult to accurately decompose the joint contribution ratio of carbon emissions and characteristic pollutants. Third, they lack quantitative assessment of operational quality. Existing methods lack a systematic quantitative assessment of their own operational quality, which affects the long-term operation and continuous optimization of the system. Therefore, it is necessary to optimize existing methods. Summary of the Invention

[0005] This invention addresses the problems in the prior art, such as high equipment costs, insufficient representativeness of equipment deployment, inadequate collaborative tracing and contribution decomposition capabilities, and lack of quantitative assessment of operational quality. It proposes a multi-source heterogeneous sensor collaborative online monitoring and tracing system for industrial carbon pollution emissions.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a multi-source heterogeneous sensing collaborative online monitoring and tracing system for industrial carbon pollution emissions, comprising:

[0007] The real-time pollution emission data acquisition module collects characteristic pollutant concentrations, carbon emission concentrations, meteorological parameters, and source-receiver distance information at each monitoring site in real time through deployed sensors.

[0008] The multi-source heterogeneous data fusion module is used to unify time and space benchmarks, remove outliers, and generate synchronous data stream records.

[0009] The pollution emission exceedance determination module retrieves the effective characteristic pollutant concentration and carbon emission concentration from the synchronous data stream record, compares them with the corresponding matching emission standards, calculates the pollution emission exceedance coefficient, and generates and outputs the exceedance coefficient record.

[0010] The alarm module receives records of exceeding limits, generates alarm records of the corresponding level based on preset hierarchical alarm rules, and pushes the alarm records to the pollution source coarse location module.

[0011] The pollution source coarse location module receives alarm records, retrieves corresponding synchronous data stream records, filters candidate pollution sources from the static pollution source list, generates a candidate target pollution source list, and pushes the candidate target pollution source list to the carbon pollution collaborative tracing and contribution decomposition module.

[0012] The carbon pollution co-source tracing and contribution decomposition module analyzes the initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration of the monitoring site based on the diffusion model. The initial contribution ratio is input into the receptor model for iterative calculation to obtain the final contribution ratio, and then a source tracing analysis report is generated.

[0013] The operation quality assessment module periodically summarizes the system operation data within a preset period, calculates the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, and the average equipment availability coefficient, and obtains the operation quality index by weighted summation. It then compares the index with a preset threshold to determine whether the system operation quality for this period meets expectations and generates an operation quality assessment report.

[0014] Compared with the prior art, the present invention includes at least one of the following beneficial technical effects:

[0015] First, addressing the issues of high equipment costs and insufficient representativeness of existing technologies, this invention utilizes the collaborative operation of a pollution source coarse location module and a carbon pollution co-tracing and contribution decomposition module. This enables the effective tracing and contribution decomposition of pollution sources over a larger area with fewer sensor locations, reducing sensor deployment density and system construction costs per unit area, and improving the representativeness of the deployment.

[0016] Second, addressing the shortcomings of existing technologies in collaborative source tracing and contribution decomposition capabilities, this invention establishes a carbon pollution collaborative source tracing and contribution decomposition module. Based on a diffusion model, it analyzes the initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration at the monitoring site. This initial contribution ratio is then input into the receptor model for iterative calculation to obtain the final contribution ratio, subsequently generating a source tracing analysis report. This collaborative iterative method, combining diffusion and receptor models, achieves joint source tracing and accurate decomposition of contribution ratios for both carbon emissions and pollutant emissions.

[0017] Third, addressing the issue of insufficient quantitative assessment of operational quality in existing technologies, this invention establishes an operational quality assessment module. This module periodically summarizes system operational data within a preset period, calculates the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, and the average equipment availability coefficient, and then weights and sums them to obtain an operational quality index. This index is compared with a preset threshold to determine whether the system's operational quality for the current period meets expectations, generating an operational quality assessment report. This execution process enables continuous monitoring and quantitative assessment of the system's own operational quality, providing a scientific basis for system optimization and maintenance. Attached Figure Description

[0018] Figure 1 This is a module connection diagram for a multi-source heterogeneous sensing collaborative online monitoring and tracing system for industrial carbon emissions. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, this embodiment provides a multi-source heterogeneous sensor collaborative online monitoring and tracing system for industrial carbon pollution emissions, including a real-time pollution emission data acquisition module, a multi-source heterogeneous data fusion module, a pollution emission exceedance determination module, an alarm module, a pollution source coarse location module, a carbon pollution collaborative tracing and contribution decomposition module, an operation quality assessment module, and a database. The real-time pollution emission data acquisition module, multi-source heterogeneous data fusion module, pollution emission exceedance determination module, alarm module, pollution source coarse location module, and carbon pollution collaborative tracing and contribution decomposition module are sequentially connected. The multi-source heterogeneous data fusion module, alarm module, and pollution source coarse location module are all connected to the operation quality assessment module. All modules in the system are connected to the database.

[0021] The real-time pollution emission data acquisition module collects characteristic pollutant concentrations, carbon emission concentrations, meteorological parameters, and source-receiver distance information at each monitoring site in real time through deployed sensors.

[0022] The multi-source heterogeneous data fusion module is used to unify time and space benchmarks, remove outliers, and generate synchronous data stream records.

[0023] The pollution emission exceedance determination module retrieves the effective characteristic pollutant concentration and carbon emission concentration from the synchronous data stream record, compares them with the corresponding matching emission standards, calculates the pollution emission exceedance coefficient, and generates and outputs the exceedance coefficient record.

[0024] The alarm module receives records of exceeding limits, generates alarm records of the corresponding level based on preset hierarchical alarm rules, and pushes the alarm records to the pollution source coarse location module.

[0025] The pollution source coarse location module receives alarm records, retrieves corresponding synchronous data stream records, filters candidate pollution sources from the static pollution source list, generates a candidate target pollution source list, and pushes the candidate target pollution source list to the carbon pollution collaborative tracing and contribution decomposition module.

[0026] The carbon pollution co-source tracing and contribution decomposition module analyzes the initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration of the monitoring site based on the diffusion model. The initial contribution ratio is input into the receptor model for iterative calculation to obtain the final contribution ratio, and then a source tracing analysis report is generated.

[0027] The operation quality assessment module periodically summarizes the system operation data within a preset period, calculates the monitoring data reliability coefficient, pollution source location accuracy coefficient, and average equipment availability coefficient, and obtains the operation quality index by weighted summation. It then compares the index with a preset threshold to determine whether the system operation quality for this period meets expectations and generates an operation quality assessment report.

[0028] A database is used to store data information for all modules in the system.

[0029] Furthermore, the concentrations of the characteristic pollutants include SO2 concentration, nitrogen oxide concentration, CO concentration, ozone concentration, and PM2.5 concentration. 10 and PM 2.5 The concentrations and total VOCs concentrations; the carbon emission concentrations include CO2 and CH4 concentrations; the meteorological parameters include wind speed, wind direction, temperature, relative humidity, atmospheric pressure, and boundary layer height; the source-receiver distance information includes the straight-line distance between the monitoring site and the candidate pollution source, the distance between the monitoring site and the plant boundary and the center of the plant area, and the distance to the upwind pollution source.

[0030] In this embodiment, it should be specifically noted that the sensors include a gas sensor, a particulate matter sensor, a photoionization detector, a wind speed and direction sensor, and a temperature, humidity, and air pressure sensor.

[0031] Gas sensors are used to monitor SO2 concentration, nitrogen oxide concentration, CO concentration, ozone concentration, and CH4 concentration.

[0032] For PM 10 and PM 2.5 For monitoring, particulate matter sensors are used; for monitoring total VOCs concentration, photoionization detectors are used.

[0033] For wind speed and direction, wind speed and direction sensors are used for monitoring.

[0034] In this embodiment, it should be specifically noted that the boundary layer height is calculated based on the collected meteorological parameters.

[0035] The straight-line distance between the monitoring site and the candidate pollution source is calculated using spherical geometry formulas after obtaining the coordinates of the two points based on GPS.

[0036] For the distance between the monitoring point and the factory boundary and the center of the factory area, after obtaining the coordinates of the vertices of the factory boundary polygon, the coordinates of the monitoring point, and the coordinates of the center of the factory area based on GPS, the distance between the monitoring point and the center of the factory area is calculated using the spherical geometry formula. The shortest perpendicular distance from the monitoring point to each side of the factory boundary polygon is the distance between the monitoring point and the factory boundary.

[0037] The distance to the upwind pollution source is calculated based on wind direction data and geometric projection.

[0038] In this embodiment, it should be specifically noted that the generation of the boundary layer height includes the following steps:

[0039] Based on the solar altitude angle, total cloud cover, low cloud cover, and average wind speed at a height of 10m, and referring to the stability level calculation table in the national standard, atmospheric stability is divided into six levels: A, B, C, D, E, and F, namely, strongly unstable, unstable, weakly unstable, neutral, weakly stable, and stable.

[0040] Calculate the geostrophic parameters based on the geographical latitude of the monitoring point;

[0041] The boundary layer height is calculated based on the determined stability level and the average wind speed at a height of 10m.

[0042] In this embodiment, it should be specifically explained that the generation of the upwind pollution source distance includes the following steps:

[0043] Obtain the current wind direction angle and calculate the upwind direction vector;

[0044] For each candidate pollution source, calculate its direction vector and straight-line distance relative to the monitoring point;

[0045] Calculate the directional consistency coefficient, which is the dot product of the upwind direction vector and the direction vector of the candidate pollution source relative to the monitoring point. The closer the directional consistency coefficient is to 1, the more likely the pollution source is upwind of the monitoring point (the wind blows directly towards the monitoring point). The closer the directional consistency coefficient is to -1, the more likely the pollution source is downwind. The closer the directional consistency coefficient is to 0, the more likely it is in the crosswind direction.

[0046] Set an angle threshold to filter out all pollution sources that meet the upwind condition;

[0047] Among all pollution sources that meet the upwind condition, the pollution source with the smallest straight-line distance is selected, and its straight-line distance is the upwind pollution source distance.

[0048] In this embodiment, it should be specifically noted that the setting of monitoring sites follows national or local standards.

[0049] In this embodiment, the execution flow of the multi-source heterogeneous data fusion module is as follows:

[0050] The timestamps of data from various sensors are unified to the NTP reference, and time alignment is achieved through interpolation and delay compensation.

[0051] Transform all spatial coordinates to a unified UTM projected coordinate system;

[0052] Abnormal data is identified and processed layer by layer by physical range verification, sliding interquartile range statistical detection, and physical consistency verification.

[0053] The processed data is integrated into a unified data record with data quality tags attached. The data is pushed to the pollution emission exceedance judgment module at fixed time intervals and a real-time cache is maintained. The data quality tags include normal, interpolation replacement, suspicious, and missing.

[0054] Furthermore, the excess coefficient record includes single excess coefficient, comprehensive excess coefficient, excess flag, low confidence label, high priority event flag, and persistent excess event flag.

[0055] Furthermore, the processing flow of the pollution emission exceeding limit determination module is as follows:

[0056] Pull synchronous data stream records from the message queue, parse the data quality flags in each record. If the characteristic pollutant concentration or carbon emission concentration flag is missing, skip this judgment, record the missing data in the system log, skip the over-limit judgment, and send an invalid data notification to the alarm module. If the characteristic pollutant concentration or carbon emission concentration flag is suspicious, perform a downgrade judgment, and attach a low confidence label to the generated pollution emission over-limit coefficient for the alarm module to refer to. If the characteristic pollutant concentration or carbon emission concentration flag is normal or interpolated, proceed to the formal judgment.

[0057] The system retrieves the emission area type to which a monitoring site belongs based on the monitoring site number in the data record, determines the applicable time period standard based on the timestamp of the current data, and dynamically retrieves the corresponding emission limit from the built-in standard library based on the type of pollutant.

[0058] For each valid monitoring data record, the individual exceedance coefficient and the comprehensive exceedance coefficient are calculated separately. The individual exceedance coefficient is the ratio of the measured concentration of the pollutant to the standard emission limit. If the individual exceedance coefficient is greater than the upper limit of the normal range by 1, it means that the pollutant exceeds the standard. If the individual exceedance coefficient is less than or equal to the upper limit of the normal range by 1, it means that the pollutant does not exceed the standard. Based on the weighted exceedance contribution method, the comprehensive exceedance coefficient is calculated by applying the weights of each pollutant preset in the configuration file. Then, all individual coefficients and comprehensive coefficients are limited to the expected range. If the value exceeds the expected upper limit, it is uniformly marked as the expected upper limit value. If the value is lower than the expected lower limit value, it is uniformly marked as the expected lower limit value.

[0059] If the single exceedance coefficient of a pollutant is greater than or equal to 1.5 times the expected upper limit value, it is determined to be a serious exceedance. A high-priority event is immediately generated, and the pollution source coarse location module is notified to start the rapid location scanning process. If the comprehensive exceedance coefficient continues to exceed the upper limit value of the normal range for more than 30 minutes, a continuous exceedance event is sent to the alarm module to start a higher level alarm.

[0060] Generate and output records of the excess coefficients.

[0061] Furthermore, the tiered alarm system includes Level 1, Level 2, and Level 3 alarms. Level 1 alarms are instantaneous severe exceedance alarms, triggered when a high-priority event flag is true, meaning the single exceedance coefficient of a pollutant is greater than or equal to 1.5 times the expected upper limit. Level 2 alarms are persistent exceedance alarms, triggered when a persistent exceedance event flag is true, meaning the comprehensive exceedance coefficient continuously exceeds the upper limit of the normal range for more than 30 minutes. Level 3 alarms are ordinary exceedance reminder alarms, triggered when a single exceedance coefficient is greater than the upper limit of the normal range and the monitoring site has continuously recorded exceedances. If the quantity exceeds the data volume alarm value for continuous exceedance, a Level 1 alarm record is immediately generated upon triggering the alarm. This record includes the monitoring site number, location, type of pollutant exceeding the limit, measured concentration value, single exceedance coefficient, and timestamp. If a Level 2 alarm is triggered, a Level 2 alarm record is generated, including the monitoring site number, location, comprehensive exceedance coefficient, duration, and duration range. If a Level 3 alarm is triggered, a Level 3 alarm record is generated, including the monitoring site number, location, pollutant exceeding the limit, single exceedance coefficient, and number of consecutive exceedances.

[0062] Specifically, in this embodiment, when a Level 1 alarm is triggered, a Level 1 alarm record is immediately generated, pushed to the pollution source coarse location module within 1 second, marked as urgent, and a red highlighted prompt box pops up on the system monitoring interface, accompanied by a continuous beeping sound. The alarm event is recorded in the system log. When a Level 2 alarm is triggered, a Level 2 alarm record is generated and pushed to the pollution source coarse location module, the regular source tracing process is initiated, an orange continuous prompt bar is displayed on the system monitoring interface, and a preset SMS notification list is triggered. When a Level 3 alarm is triggered, a Level 3 alarm record is generated. If neither a Level 1 nor a Level 2 alarm is triggered, the Level 3 alarm record is pushed to the pollution source coarse location module, and a yellow prompt message is displayed on the system monitoring interface.

[0063] In this embodiment, it should be specifically noted that when the record of the excess coefficient contains a low confidence label, the alarm module performs the following special processing:

[0064] Active push notifications that do not trigger Level 1, 2, or 3 alarms;

[0065] Only display a suspicious data icon next to the data point on the system monitoring interface, and explain the reason in the floating details;

[0066] Records exceeding limits with low confidence labels are stored separately in the suspicious data log table, which is used as a basis for deduction when the operation quality assessment module calculates the system operation quality index.

[0067] If more than 10 records of low-confidence tags appear consecutively at the same monitoring point, the alarm module generates an equipment inspection suggestion to notify maintenance personnel to check the status of the corresponding sensors.

[0068] Furthermore, the execution flow of the pollution source coarse location module is as follows:

[0069] Upon receiving an alarm record, the system parses and extracts the monitoring site number, monitoring site coordinates, timestamp, alarm level, type of pollutant exceeding the standard, individual exceedance coefficient, and comprehensive exceedance coefficient. If it is a Level 1 alarm, the system automatically enters the rapid positioning mode; if it is a Level 2 or Level 3 alarm, the system enters the normal positioning mode.

[0070] Based on the monitoring site number and timestamp in the alarm record, a query request is sent to the multi-source heterogeneous data fusion module to obtain the synchronous data stream record corresponding to the alarm, retrieve meteorological parameters, source-receptor distance information and pollutant concentration. If the corresponding synchronous data stream record cannot be found, the coarse positioning process is terminated and the reason for the coarse positioning failure is recorded in the system log as missing synchronous data.

[0071] The list of all candidate pollution sources within the monitoring area is loaded from the static pollution source database. The list of pollution sources includes the pollution source number, name, type, UTM coordinates, and main emission pollutant spectrum. Candidate target pollution sources are obtained through spatial preliminary screening and feature fine screening.

[0072] Calculate the comprehensive suspicion score for the candidate target pollution sources, and sort them from high to low according to the comprehensive suspicion score to generate a list of candidate target pollution sources;

[0073] The alarm records, synchronized data stream records, candidate target pollution source list, and coarse location metadata are packaged and pushed to the carbon pollution collaborative tracing and contribution decomposition module.

[0074] Furthermore, the initial spatial screening includes: obtaining the current wind direction angle, calculating the upwind direction vector, for each candidate pollution source, calculating its direction vector and straight-line distance relative to the monitoring point, calculating the direction consistency coefficient, and if the direction consistency coefficient is greater than or equal to 0.5 and the straight-line distance is less than or equal to the maximum distance, the initial spatial screening is passed and the pollution source enters the feature fine screening.

[0075] The feature screening includes: obtaining the set of pollutant types exceeding the standard in this alarm, obtaining the main emission pollutant spectrum of the candidate pollution source, and calculating the pollutant category matching degree, that is, the proportion of the intersection of the emission spectrum of the pollutant exceeding the standard and the emission spectrum of the pollution source to the total number of pollutants exceeding the standard. If the pollutant category matching degree is greater than or equal to 0.5, the feature screening passes and the pollution source is marked as a candidate target pollution source. If it is a level 1 alarm and the screening result of the pollutant category matching degree greater than or equal to 0.5 is empty, the matching threshold is downgraded to 0.3 and the feature screening is re-executed.

[0076] The formula for calculating the overall suspect score is as follows: In the formula, Score i α, di k i Match i The following parameters are, in order: comprehensive score of pollution source i suspicion, distance attenuation coefficient, straight-line distance between monitoring site and pollution source i, directional consistency coefficient between pollution source i and monitoring site, and pollutant category matching degree, with w1, w2, and w3 being weighting coefficients, respectively.

[0077] Specifically, in this embodiment, the maximum distance is twice the boundary layer height for a Level 1 alarm, and the boundary layer height for a Level 2 or Level 3 alarm. When there is no boundary layer height data, the default value is 2KM. The default value of the distance attenuation coefficient is 1.0. The default values ​​of the weighting coefficients are 0.4, 0.4, and 0.2, respectively.

[0078] Furthermore, the candidate target pollution source list includes pollution source number, name, coordinates, suspicion comprehensive score, matching degree, and straight-line distance;

[0079] The coarse positioning metadata includes wind direction angle, wind direction fluctuation range before and after the alarm, coarse positioning mode, direction consistency threshold, angle corresponding to the direction threshold, maximum distance threshold, pollutant matching degree threshold, downgraded matching identifier, total number of static pollution sources, number of sources passing the initial spatial screening, number of sources passing the feature fine screening, number of final candidate target pollution sources, highest suspected comprehensive score, and lowest suspected comprehensive score.

[0080] In this embodiment, it should be specifically noted that the diffusion model includes a Gaussian plume model and a Lagrange particle diffusion model. Coarse location metadata is retrieved. If the wind direction is stable and the monitoring point is located downwind of the source, the Gaussian plume model is used. If the wind direction fluctuates greatly or complex terrain is involved, the model is switched to the Lagrange particle diffusion model. After the diffusion model is confirmed, the model parameters are configured.

[0081] Furthermore, the initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration at the monitoring site based on the diffusion model includes: for pollution source i in the candidate target pollution source list, calling the selected diffusion model to calculate the positive transfer coefficient K of pollutant j from pollution source i to monitoring point m. i，m，j Based on the forward transmission coefficient K i，m，j and the actual measured concentration C at the monitoring point 实测，m，j Inversely extrapolate the equivalent source strength Q of pollution source i at the monitoring time. i，j Equivalent source strength Q based on reverse inference i，j and forward transmission coefficient K i，m，j The contribution of pollution source i to monitoring point m with respect to pollutant j was recalculated, simulating the concentration C. 贡献，i，m，j For pollutant j, the simulated total concentration C is obtained by summing the simulated concentrations of all candidate target pollution sources at monitoring point m. 模拟总计，m，jCalculate the initial contribution ratio R of the i-th pollution source to pollutant j at monitoring point m. i，m，j C 贡献，i，m，j Divide by C 模拟总计，m，j The quotient is used to calculate the simulated total concentration C. 模拟总计，m，j Compared with the measured concentration C 实测，m，j The percentage deviation, wherein pollutant j is any one of carbon emission pollutants or characteristic pollutants.

[0082] In this embodiment, it should be specifically noted that the source tracing analysis report includes a task identifier, details of each pollution source's contribution, and carbon-pollution synergistic characteristics. The task identifier includes the corresponding alarm record ID, monitoring site number, and timestamp. The details of each pollution source's contribution include the CO2 contribution ratio, CH4 contribution ratio, and contribution ratio of the main characteristic pollutants based on the alarm records for each candidate target pollution source. The carbon-pollution synergistic characteristics include the carbon-pollution homogeneity of each source and the ranking of the main contributing sources. The carbon-pollution homogeneity of each source is the ratio of the total carbon emission contribution ratio of each candidate target pollution source to the total contribution ratio of the main characteristic pollutants. The ranking of the main contributing sources is the descending order of each pollution source according to its contribution ratio.

[0083] Furthermore, the system operation data includes the total number of data records of each monitoring site of the multi-source heterogeneous data fusion module in this cycle, the number of records marked as normal data quality and the number of records marked as interpolation replacement data quality, the total number of suspicious data logs of the alarm module in this cycle, the total number of alarm records received by the pollution source coarse location module in this cycle and the number of times coarse location was successfully completed, the theoretical total operating hours of all equipment in this cycle and the cumulative number of hours that the equipment was unavailable due to faults or maintenance in this cycle;

[0084] The reliability coefficient of the monitoring data is the product of the proportion of valid data and the penalty factor for suspicious data. The proportion of valid data is calculated based on the total number of data records at each monitoring site in the current period, the number of records marked as normal data quality, and the number of records marked as interpolation replacement data quality. The penalty factor for suspicious data is calculated based on the total number of data records at each monitoring site in the current period and the total number of suspicious data logs.

[0085] The pollution source location accuracy coefficient is calculated based on the total number of alarm records received by the pollution source coarse location module within this cycle and the number of times coarse location was successfully completed.

[0086] The average equipment availability coefficient is calculated based on the total theoretical operating hours of all equipment in the current period and the cumulative number of hours that equipment is unavailable due to failure or maintenance in the current period.

[0087] The operational quality index is obtained by weighted summation of the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, and the average equipment availability coefficient, with the weighting coefficients being preset by the system.

[0088] In this embodiment, it should be specifically explained that determining whether the system operation quality of this cycle meets expectations includes: comparing the operation quality index with a preset threshold; if the operation quality index is greater than or equal to the preset threshold, it is determined that the system operation quality of this cycle meets expectations; if the operation quality index is less than the preset threshold, it is determined that the system operation quality of this cycle does not meet expectations.

[0089] The operational quality assessment report includes the start and end dates of the assessment period, the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, the average equipment availability coefficient, the operational quality index, and the judgment results.

[0090] In this embodiment, it should be noted that the preset values, thresholds, and weighting coefficients used are all selected based on actual needs, and no specific value restrictions are imposed here.

[0091] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-source heterogeneous sensing collaborative online monitoring and tracing system for industrial carbon pollution emissions, characterized in that, include: The real-time pollution emission data acquisition module collects characteristic pollutant concentrations, carbon emission concentrations, meteorological parameters, and source-receiver distance information at each monitoring site in real time through deployed sensors. The multi-source heterogeneous data fusion module is used to unify time and space benchmarks, remove outliers, and generate synchronous data stream records. The pollution emission exceedance determination module retrieves the effective characteristic pollutant concentration and carbon emission concentration from the synchronous data stream record, compares them with the corresponding matching emission standards, calculates the pollution emission exceedance coefficient, and generates and outputs the exceedance coefficient record. The alarm module receives records of exceeding limits, generates alarm records of the corresponding level based on preset hierarchical alarm rules, and pushes the alarm records to the pollution source coarse location module. The pollution source coarse location module receives alarm records, retrieves corresponding synchronous data stream records, filters candidate pollution sources from the static pollution source list, generates a candidate target pollution source list, and pushes the candidate target pollution source list to the carbon pollution collaborative tracing and contribution decomposition module. The carbon pollution co-source tracing and contribution decomposition module analyzes the initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration of the monitoring site based on the diffusion model. The initial contribution ratio is input into the receptor model for iterative calculation to obtain the final contribution ratio, and then a source tracing analysis report is generated. The operation quality assessment module periodically summarizes the system operation data within a preset period, calculates the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, and the average equipment availability coefficient, and obtains the operation quality index by weighted summation. It then compares the index with a preset threshold to determine whether the system operation quality for this period meets expectations and generates an operation quality assessment report.

2. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The concentrations of the characteristic pollutants include SO2 concentration, nitrogen oxide concentration, CO concentration, ozone concentration, and PM2.5 concentration. 10 and PM 2.5 The concentrations and total VOCs concentrations; the carbon emission concentrations include CO2 and CH4 concentrations; the meteorological parameters include wind speed, wind direction, temperature, relative humidity, atmospheric pressure, and boundary layer height; the source-receiver distance information includes the straight-line distance between the monitoring site and the candidate pollution source, the distance between the monitoring site and the plant boundary and the center of the plant area, and the distance to the upwind pollution source.

3. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The excess coefficient record includes single excess coefficient, comprehensive excess coefficient, excess flag, low confidence label, high priority event flag, and persistent excess event flag.

4. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The processing flow of the pollution emission exceeding limit determination module is as follows: Pull synchronous data stream records from the message queue, parse the data quality flags in each record. If the characteristic pollutant concentration or carbon emission concentration flag is missing, skip this judgment, record the missing data in the system log, skip the over-limit judgment, and send an invalid data notification to the alarm module. If the characteristic pollutant concentration or carbon emission concentration flag is suspicious, perform a downgrade judgment, and attach a low confidence label to the generated pollution emission over-limit coefficient for the alarm module to refer to. If the characteristic pollutant concentration or carbon emission concentration flag is normal or interpolated, proceed to the formal judgment. The system retrieves the emission area type to which a monitoring site belongs based on the monitoring site number in the data record, determines the applicable time period standard based on the timestamp of the current data, and dynamically retrieves the corresponding emission limit from the built-in standard library based on the type of pollutant. For each valid monitoring data record, the individual exceedance coefficient and the comprehensive exceedance coefficient are calculated separately. The individual exceedance coefficient is the ratio of the measured concentration of the pollutant to the standard emission limit. If the individual exceedance coefficient is greater than the upper limit of the normal range by 1, it means that the pollutant exceeds the standard. If the individual exceedance coefficient is less than or equal to the upper limit of the normal range by 1, it means that the pollutant does not exceed the standard. Based on the weighted exceedance contribution method, the comprehensive exceedance coefficient is calculated by applying the weights of each pollutant preset in the configuration file. Then, all individual coefficients and comprehensive coefficients are limited to the expected range. If the value exceeds the expected upper limit, it is uniformly marked as the expected upper limit value. If the value is lower than the expected lower limit value, it is uniformly marked as the expected lower limit value. If the single exceedance coefficient of a pollutant is greater than or equal to 1.5 times the expected upper limit value, it is determined to be a serious exceedance. A high-priority event is immediately generated, and the pollution source coarse location module is notified to start the rapid location scanning process. If the comprehensive exceedance coefficient continues to exceed the upper limit value of the normal range for more than 30 minutes, a continuous exceedance event is sent to the alarm module to start a higher level alarm. Generate and output records of the excess coefficients.

5. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The tiered alarm system includes Level 1, Level 2, and Level 3 alarms. Level 1 alarms are instantaneous severe exceedance alarms, triggered when a high-priority event flag is true, meaning the single exceedance coefficient of a pollutant is greater than or equal to 1.5 times the expected upper limit. Level 2 alarms are persistent exceedance alarms, triggered when a persistent exceedance event flag is true, meaning the comprehensive exceedance coefficient continuously exceeds the upper limit of the normal range for more than 30 minutes. Level 3 alarms are ordinary exceedance reminder alarms, triggered when a single exceedance coefficient is greater than the upper limit of the normal range and the number of consecutive exceedance records at that monitoring site is large. The alarm value is set for the amount of data that continuously exceeds the limit. When a Level 1 alarm is triggered, a Level 1 alarm record is immediately generated. The Level 1 alarm record includes the monitoring site number, location, type of pollutant exceeding the limit, measured concentration value, single exceedance coefficient, and timestamp. When a Level 2 alarm is triggered, a Level 2 alarm record is generated. The Level 2 alarm record includes the monitoring site number, location, comprehensive exceedance coefficient, duration, and duration range. When a Level 3 alarm is triggered, a Level 3 alarm record is generated. The Level 3 alarm record includes the monitoring site number, location, pollutant exceeding the limit, single exceedance coefficient, and number of consecutive exceedances.

6. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The execution flow of the pollution source coarse location module is as follows: Upon receiving an alarm record, the system parses and extracts the monitoring site number, monitoring site coordinates, timestamp, alarm level, type of pollutant exceeding the standard, individual exceedance coefficient, and comprehensive exceedance coefficient. If it is a Level 1 alarm, the system automatically enters the rapid positioning mode; if it is a Level 2 or Level 3 alarm, the system enters the normal positioning mode. Based on the monitoring site number and timestamp in the alarm record, a query request is sent to the multi-source heterogeneous data fusion module to obtain the synchronous data stream record corresponding to the alarm, retrieve meteorological parameters, source-receptor distance information and pollutant concentration. If the corresponding synchronous data stream record cannot be found, the coarse positioning process is terminated and the reason for the coarse positioning failure is recorded in the system log as missing synchronous data. The list of all candidate pollution sources within the monitoring area is loaded from the static pollution source database. The list of pollution sources includes the pollution source number, name, type, UTM coordinates, and main emission pollutant spectrum. Candidate target pollution sources are obtained through spatial preliminary screening and feature fine screening. Calculate the comprehensive suspicion score for the candidate target pollution sources, and sort them from high to low according to the comprehensive suspicion score to generate a list of candidate target pollution sources; The alarm records, synchronized data stream records, candidate target pollution source list, and coarse location metadata are packaged and pushed to the carbon pollution collaborative tracing and contribution decomposition module.

7. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 6, characterized in that, The initial spatial screening includes: obtaining the current wind direction angle, calculating the upwind direction vector, for each candidate pollution source, calculating its direction vector and straight-line distance relative to the monitoring point, calculating the direction consistency coefficient, and if the direction consistency coefficient is greater than or equal to 0.5 and the straight-line distance is less than or equal to the maximum distance, the initial spatial screening is passed and the pollution source enters the feature fine screening. The feature screening includes: obtaining the set of pollutant types exceeding the standard in this alarm, obtaining the main emission pollutant spectrum of the candidate pollution source, and calculating the pollutant category matching degree, that is, the proportion of the intersection of the emission spectrum of the pollutant exceeding the standard and the emission spectrum of the pollution source to the total number of pollutants exceeding the standard. If the pollutant category matching degree is greater than or equal to 0.5, the feature screening passes and the pollution source is marked as a candidate target pollution source. If it is a level 1 alarm and the screening result of the pollutant category matching degree greater than or equal to 0.5 is empty, the matching threshold is downgraded to 0.3 and the feature screening is re-executed. The formula for calculating the overall suspect score is as follows: In the formula, Score i α, d i k i Match i The following parameters are, in order: comprehensive score of pollution source i suspicion, distance attenuation coefficient, straight-line distance between monitoring site and pollution source i, directional consistency coefficient between pollution source i and monitoring site, and pollutant category matching degree, with w1, w2, and w3 being weighting coefficients, respectively.

8. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 7, characterized in that, The list of candidate target pollution sources includes pollution source number, name, coordinates, comprehensive suspicion score, matching degree, and straight-line distance; The coarse positioning metadata includes wind direction angle, wind direction fluctuation range before and after the alarm, coarse positioning mode, direction consistency threshold, angle corresponding to the direction threshold, maximum distance threshold, pollutant matching degree threshold, downgraded matching identifier, total number of static pollution sources, number of sources passing the initial spatial screening, number of sources passing the feature fine screening, number of final candidate target pollution sources, highest suspected comprehensive score, and lowest suspected comprehensive score.

9. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The initial contribution ratio of each candidate target pollution source to the carbon emission concentration and characteristic pollutant concentration at the monitoring site, analyzed based on the diffusion model, includes: for pollution source i in the candidate target pollution source list, using the selected diffusion model to calculate the positive transfer coefficient K of pollutant j from pollution source i to monitoring point m. i，m，j Based on the forward transmission coefficient K i，m，j and the actual measured concentration C at the monitoring point 实测，m，j Inversely extrapolate the equivalent source strength Q of pollution source i at the monitoring time. i，j Equivalent source strength Q based on reverse inference i，j and forward transmission coefficient K i，m，j The contribution of pollution source i to monitoring point m with respect to pollutant j was recalculated, simulating the concentration C. 贡献，i，m，j For pollutant j, the simulated total concentration C is obtained by summing the simulated concentrations of all candidate target pollution sources at monitoring point m. 模拟总计，m，j Calculate the initial contribution ratio R of the i-th pollution source to pollutant j at monitoring point m. i，m，j C 贡献，i，m，j Divide by C 模拟总计，m，j The quotient is used to calculate the simulated total concentration C. 模拟总计，m，j Compared with the measured concentration C 实测，m，j The percentage deviation, wherein pollutant j is any one of carbon emission pollutants or characteristic pollutants.

10. The multi-source heterogeneous sensing collaborative industrial carbon pollution emission online monitoring and tracing system according to claim 1, characterized in that, The system operation data includes the total number of data records of each monitoring site of the multi-source heterogeneous data fusion module in this cycle, the number of records marked as normal data quality and the number of records marked as interpolation replacement data quality, the total number of suspicious data logs of the alarm module in this cycle, the total number of alarm records received by the pollution source coarse location module in this cycle and the number of times coarse location was successfully completed, the theoretical total operating hours of all equipment in this cycle and the cumulative number of hours that the equipment was unavailable due to faults or maintenance in this cycle; The reliability coefficient of the monitoring data is the product of the proportion of valid data and the penalty factor for suspicious data. The proportion of valid data is calculated based on the total number of data records at each monitoring site in the current period, the number of records marked as normal data quality, and the number of records marked as interpolation replacement data quality. The penalty factor for suspicious data is calculated based on the total number of data records at each monitoring site in the current period and the total number of suspicious data logs. The pollution source location accuracy coefficient is calculated based on the total number of alarm records received by the pollution source coarse location module within this cycle and the number of times coarse location was successfully completed. The average equipment availability coefficient is calculated based on the total theoretical operating hours of all equipment in the current period and the cumulative number of hours that equipment is unavailable due to failure or maintenance in the current period. The operational quality index is obtained by weighted summation of the reliability coefficient of monitoring data, the accuracy coefficient of pollution source location, and the average equipment availability coefficient, with the weighting coefficients being preset by the system.

Images