A VOCs gas concentration distribution source tracing analysis method

Abstract

This invention relates to the field of VOCs source analysis and discloses a method for source analysis of VOCs gas concentration distribution. The method includes acquiring VOCs concentration data collected from multiple monitoring nodes within a monitoring area; identifying concentration overlap areas formed by multiple emission activities within the monitoring area; and forming an emission event influence chain based on the concentration transfer intensity corresponding to the continuity relationship. It also involves calculating the coverage of previous emission events in subsequent concentration changes to form historical contribution records; calculating the concentration difference and the correlation coefficient between the current concentration field and the historical contribution spatial distribution to obtain new concentration contribution records; determining the contribution competition relationship between multiple candidate emission sources when a new concentration contribution record can simultaneously correspond to multiple candidate emission sources, and forming a source contribution allocation result; and dynamically correcting the contribution values ​​of each emission source when a new concentration increase occurs again to form a source contribution change record. This invention has the advantage of improving the accuracy of source analysis.

Description

Technical Field

[0001] This invention relates to the field of VOCs source analysis, specifically to a method for source analysis of VOCs gas concentration distribution. Background Technology

[0002] Source analysis of volatile organic compounds (VOCs) typically relies on monitoring point concentration data, meteorological data, and diffusion models to determine pollution sources. However, in multi-source emission scenarios such as large chemical industrial parks and petrochemical storage and transportation bases, the emission behaviors of multiple sources are often intermittent, random, and temporally overlapping. Before a pollution cloud formed by one emission source has completely dispersed, other emission sources may have already begun new emission activities. Gas clouds formed by different emission events are prone to spatial convergence and concentration superposition during propagation, causing the concentration changes collected at monitoring points to simultaneously contain information on the impact of multiple emission events.

[0003] Most existing VOCs source analysis methods rely on wind field information, concentration peak characteristics, or reverse trajectory results at the monitoring time to identify pollution sources. They typically assume that concentration anomalies correspond to a single emission event, lacking the ability to dynamically analyze the temporal interrelationships between multiple emission events. When the emission intensities of different emission sources are similar and their diffusion paths partially overlap, it is easy to misjudge historical residual air masses as current emission source contributions, or to incorrectly merge multiple emission events into a single pollution source event, leading to distorted source contribution rate calculations and affecting the accuracy of pollution control decisions.

[0004] Therefore, it is necessary to design a source analysis method for VOCs gas concentration distribution that improves the accuracy of source analysis. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a source analysis method for VOCs gas concentration distribution, which has the advantage of improving the accuracy of source analysis and solves the problems mentioned in the background technology.

[0006] To achieve the aforementioned goal of improving the accuracy of source analysis, this invention provides the following technical solution: a method for source analysis of VOCs gas concentration distribution, comprising the following steps: VOCs concentration data collected by multiple monitoring nodes within the monitoring area are acquired, and overlapping concentration areas formed by multiple emission activities within the monitoring area are identified. The formation time, expansion direction and dissipation process of each overlapping concentration area are recorded. An influence transmission relationship is established based on the spatial overlap ratio between the later-formed overlapping concentration area and the earlier-formed overlapping concentration area, and an emission event influence chain is formed based on the concentration transmission intensity corresponding to the transmission relationship. The duration, concentration transfer intensity, and concentration contribution ratio of each impact relationship are traced along the impact chain of emission events. The coverage of previous emission events in subsequent concentration changes is calculated. Based on the coverage and corresponding concentration contribution values, the historical contribution retention of each emission event is calculated to form a historical contribution record. The historical contribution retention is compared with the current monitoring concentration according to the corresponding monitoring location. The concentration difference and the correlation coefficient between the current concentration field and the spatial distribution of historical contributions are calculated. The concentration components formed by the continuation of historical contributions are identified. The historical contribution retention is deducted step by step according to the preset stripping step size. The stripping termination condition is determined according to the correlation between the remaining concentration field and the spatial distribution of historical contributions to obtain the new concentration contribution record. When a newly added concentration contribution record can correspond to multiple candidate emission sources at the same time, the number of times each candidate emission source has received its impact, the degree of contribution retention, and the concentration response characteristics in the historical stage are statistically analyzed to determine the contribution competition relationship between each candidate emission source and to form the source contribution allocation result. The stability of the source contribution allocation results is continuously tracked during subsequent concentration changes. When a new concentration increase occurs again, the source attribution ratio of the new concentration is determined based on the current historical contribution record, the current new concentration contribution record, and the source contribution allocation result corresponding to the current moment. The contribution values ​​of each emission source are dynamically corrected to form a source contribution change record, and the corresponding VOCs gas concentration distribution source tracing analysis results are output.

[0007] Preferably, the process of recording the formation time, expansion direction, and dissipation process of each concentration overlap region is as follows: A multi-node gas concentration sensor array is deployed within the monitoring area to collect VOCs concentration data at each monitoring node at a preset sampling period, and a concentration distribution field is constructed using a spatial interpolation method. Based on the concentration change rate at each monitoring node at adjacent sampling times, the starting point of concentration rise is determined when the concentration change rate is higher than the preset rise threshold for multiple consecutive sampling periods, and the ending point of concentration fall is determined when the concentration change rate is lower than the preset fall threshold for multiple consecutive sampling periods. Identify independent concentration clouds caused by different emission activities based on the connectivity of concentration boundaries in the concentration distribution field; When two or more independent concentration clouds overlap in space and the peak concentration in the overlapping area exceeds the set ratio of the sum of the concentrations of each cloud when acting alone, which is extrapolated from the concentration distribution before the overlap, the overlapping area is defined as the concentration overlap area. The time when the concentration superposition effect first appears in the area, the direction vector of cloud expansion, and the dissipation process of the concentration superposition effect fading to below the level of a single cloud are recorded.

[0008] Preferably, the process of forming an emission event impact chain based on the concentration transfer intensity corresponding to the succession relationship is as follows: Based on the formation time of each concentration overlap region, multiple overlap regions appearing in the same spatial grid cell are sorted by time. The emission activities corresponding to the concentration overlap regions with earlier formation times are marked as earlier emission events, and the emission activities corresponding to the concentration overlap regions with later formation times are marked as later emission events. The cloud boundary tracking method is used to calculate the rate and arrival time of the expansion of the boundary of the rear overlapping region in the direction of the concentration centroid of the front overlapping region, and to determine the moment when the boundary of the rear overlapping region and the front overlapping region comes into contact. After the boundary of the rear overlapping region and the front overlapping region comes into contact, the spatial overlap ratio between the two is calculated. When the spatial overlap ratio exceeds the preset acceptance threshold, it is determined that it affects the establishment of the acceptance relationship. Multiple impact relationships established sequentially are linked together to form an emission event impact chain with emission events as nodes and impact relationships as edges. The establishment time, spatial span, and concentration transfer intensity of each impact relationship are recorded in the chain. The concentration transfer intensity is the ratio between the average concentration of the overlapping area at the time of the impact relationship establishment and the peak concentration of the previously overlapping area.

[0009] Preferably, the process for calculating the coverage of pre-emission events in subsequent concentration changes is as follows: Along the direction of the emission event impact chain, extract the previous emission event identifier and the subsequent concentration change period in each succession relationship. The subsequent concentration change period is the time interval between the establishment of the impact succession relationship and the dissipation of the corresponding subsequent emission event. During the subsequent concentration change period, the concentration component contributed by the previous emission event was continuously monitored, and the proportion of the previous emission event in the total concentration was separated by concentration tracer factor or correlation analysis method. Based on the curve of the proportion decaying over time, the duration for which the impact of the pre-emission event remains above the set contribution threshold during the subsequent concentration change process is calculated. The ratio of the duration to the total duration of the subsequent concentration change is taken as the coverage level, and the gradient of the coverage level with spatial location is recorded. The gradient is the ratio of the difference in coverage level between adjacent monitoring locations to the corresponding spatial distance.

[0010] The preferred process for creating a historical contribution record is as follows: For each node in the emission event impact chain, the preceding emission events are weighted and accumulated based on their coverage in multiple subsequent connections and their corresponding concentration contributions. The weighting coefficient decreases exponentially with the increase of the connection distance. The connection distance is the number of directed edges traversed from the event node to the node belonging to the subsequent concentration change period in the impact chain. The weighted cumulative concentration contribution value is used as the historical contribution retention amount of the emission event within the monitoring area; Based on the original time sequence of emission events, the identifier, location, time of occurrence, and corresponding historical contribution retention of each emission event are associated and stored to form a traceable historical contribution record.

[0011] Preferably, the process for identifying concentration components formed by the continuation of historical contributions is as follows: Calculate the difference and ratio between the measured concentration and the historical contribution retention. When the absolute value of the difference between the monitored concentration and the historical contribution retention is less than the set deviation threshold, and the correlation coefficient between the current concentration field and the spatial distribution of historical contribution is higher than the set correlation threshold, it is determined that the concentration of the corresponding monitoring point is formed by the continuation of historical contribution. When the measured concentration is higher than the historical contribution retention and the difference exceeds the set deviation threshold, it is determined that there is a new concentration contribution. When the absolute value of the difference between the monitored concentration and the historical contribution retention is less than the set deviation threshold, but the correlation coefficient between the current concentration field and the spatial distribution of historical contributions is lower than the set correlation threshold, it is determined that the current concentration and the spatial distribution of historical contributions do not match, and the corresponding concentration is included in the scope of new concentration contribution analysis. Based on the determination results of each monitoring point, a historical contribution continuation concentration distribution map is generated, marking the proportion of historical contribution retention concentration in the current total concentration and its spatial distribution characteristics.

[0012] Preferably, the process for obtaining the newly added concentration contribution record is as follows: Based on the identified components formed by the continuation of historical contributions, the concentration values ​​formed by the continuation of historical contributions at the corresponding locations are subtracted sequentially from the total concentration field of the current monitored concentrations, according to the monitoring points. After deducting the concentration at each monitoring point, the remaining concentration value is calculated and used as the new concentration value for the corresponding monitoring point. When the remaining concentration value after deduction is less than zero, the new concentration value of the corresponding monitoring point is set to zero; when the remaining concentration value after deduction is greater than zero, the new concentration value of the corresponding monitoring point is retained. The newly added concentration values ​​at each monitoring point are summarized to form a spatial distribution field of the newly added concentration, which serves as a record of the contribution of the newly added concentration.

[0013] Preferably, the process of statistically analyzing the number of impacts, contribution retention levels, and concentration response characteristics of each candidate emission source in historical periods is as follows: The spatial distribution pattern and diffusion gradient of newly added concentration contribution records are extracted and matched with the historical emission location, emission height and emission intensity of each candidate emission source recorded in the emission source database. When the spatial distribution similarity and diffusion direction consistency both exceed the preset matching threshold, the corresponding emission source is determined as a candidate emission source. For each candidate emission source, the number of times the impact was absorbed within a set time window prior to the current monitoring time is counted from the historical contribution record, as well as the average degree to which the emission source contribution is retained in each absorption. The concentration response characteristics of each candidate emission source in historical emission events were analyzed, including the concentration rise slope, peak arrival time, and decay time constant.

[0014] Preferably, the process for forming the source contribution allocation result is as follows: The historical number of times each candidate emission source has received a contribution, the degree of contribution retention, and the concentration response characteristics are input into the Bayesian multi-source allocation model to calculate the prior contribution probability of each candidate emission source to the current newly added concentration contribution record. Based on the similarity between the newly added concentration contribution records and the individual concentration distributions of each candidate emission source obtained by the diffusion model simulation, the prior contribution probability is iteratively updated to obtain the posterior contribution probability. When the posterior contribution probabilities of two or more candidate emission sources all exceed a set threshold, it is determined that there is a contribution competition relationship between them, and the concentration share in the newly added concentration contribution record is allocated according to the ratio of the posterior contribution probabilities, thus forming the source contribution allocation result.

[0015] Preferably, the process of outputting the corresponding VOCs gas concentration distribution source analysis results is as follows: During the continuous monitoring period after the source contribution allocation results are formed, the contribution allocation ratio of each emission source is recalculated at each set time interval. The variance of each ratio value fluctuates over time. Periods with variance below the stability threshold are marked as allocation stability periods, and periods with variance exceeding the fluctuation threshold are marked as allocation adjustment periods. When a new concentration is detected to increase again, the starting time, peak concentration and spatial distribution of the growth peak are extracted. Combined with the historical contribution record at the current time and the source contribution allocation result at the current time, the proportion of the growth peak attributable to each historical emission source is calculated. The contribution values ​​of each emission source are dynamically corrected based on the calculated attribution ratio. The contribution values ​​before and after correction, as well as the correction time, are recorded as source contribution change records. These source contribution change records are then linked with the corresponding concentration overlap areas and emission event impact chains and output as the source tracing analysis results of VOCs gas concentration distribution.

[0016] Compared with existing technologies, this invention provides a source tracing and analysis method for VOCs gas concentration distribution, which has the following beneficial effects: This invention expands the traditional source tracing analysis method based solely on single-moment concentration distribution to a dynamic source tracing method encompassing the entire concentration evolution process by constructing concentration overlap regions, influence succession relationships, and emission event impact chains. It enables the correlation tracking of the transmission effects between different emission events in complex scenarios involving multiple emission sources coexisting, overlapping pollution plumes, and continuously changing emission processes. By calculating historical contribution retention and establishing historical contribution records, it achieves the separation analysis of historical emission residual effects from current new emission contributions, reducing the interference of historical pollution accumulation effects on current source apportionment results. Furthermore, by combining concentration differences and spatial distribution correlations, it identifies the continuation of historical contributions. The concentration components formed are extracted using a stepwise stripping mechanism to improve the accuracy of identifying new pollution sources. When multiple candidate emission sources simultaneously meet diffusion characteristics, contribution competition analysis is conducted by combining the number of impacts received, the degree of contribution retention, and concentration response characteristics to achieve quantitative allocation of source contributions under multi-source overlap conditions. By continuously tracking the changes in source contribution allocation results, the source attribution ratio during the increase of new concentrations is dynamically corrected, enabling the source apportionment results to be updated in real time with changes in emission activities and environmental conditions. This improves the accuracy, continuity, interpretability, and dynamic adaptability of source tracing results in complex VOCs emission scenarios. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the method of the present invention; Figure 2 This is a flowchart for recording the overlapping regions of each concentration in this invention. Detailed Implementation

[0018] 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.

[0019] Example 1: Please refer to Figure 1 A source tracing and analysis method for VOCs gas concentration distribution in an embodiment of the present invention includes the following steps: S1: Acquire VOCs concentration data collected by multiple monitoring nodes within the monitoring area, identify concentration overlap areas formed by multiple emission activities within the monitoring area, record the formation time, expansion direction and dissipation process of each concentration overlap area, establish the influence inheritance relationship based on the spatial overlap ratio between the later-formed concentration overlap area and the earlier-formed concentration overlap area, and form the emission event influence chain based on the concentration transfer intensity corresponding to the inheritance relationship.

[0020] The process of recording the formation time, expansion direction, and dissipation process of overlapping regions of each concentration in S1 is as follows: A multi-node gas concentration sensor array is deployed within the monitoring area to collect VOCs concentration data from each monitoring node at a preset sampling period. A spatial interpolation method is then used to construct the concentration distribution field. Within the target monitoring area, a VOCs gas concentration sensor array is deployed using a combination of gridding and intensified density in key areas, based on the emission source distribution characteristics and prevailing wind direction. Each sensor node has a built-in photoionization detector or gas chromatography detection module, capable of real-time detection of the total concentration and specific component concentrations of VOCs in the air. The system has a preset uniform sampling period, such as collecting data every two or five seconds. All sensor nodes maintain time consistency via a wireless synchronization clock. At each sampling moment, each node uploads its concentration data, along with its node number, timestamp, and location coordinates, to the central data processing unit. Upon receiving the concentration data from each discrete node, the central data processing unit uses spatial interpolation methods, such as inverse distance weighted interpolation or Kriging interpolation, to extend the concentration values ​​of the discrete nodes onto a continuous spatial grid across the entire monitoring area, generating the concentration distribution field for that sampling moment. The concentration distribution field is stored in the form of a two-dimensional planar grid or a three-dimensional spatial grid, with each grid cell assigned a concentration estimate, thus providing a continuous spatial data basis for subsequent identification of concentration clouds and overlapping areas; Based on the concentration change rate at each monitoring node at adjacent sampling times, the system determines the starting point of a concentration increase when the concentration change rate is above a preset increase threshold for multiple consecutive sampling periods, and the ending point of a concentration decrease when the concentration change rate is below a preset decrease threshold for multiple consecutive sampling periods. For each monitoring node, the system extracts the concentration time-series data for that node at multiple consecutive sampling times. The system calculates the concentration change between adjacent sampling times and divides it by the sampling time interval to obtain the concentration change rate of that node within that time period. A positive change rate indicates a concentration increase, and a negative change rate indicates a concentration decrease. The system presets an increase threshold and a decrease threshold, with the increase threshold being positive and the decrease threshold being negative. When the concentration change rate of a monitoring node is above the preset increase threshold for multiple consecutive sampling periods (e.g., three or five consecutive sampling periods), the system determines the start time of that consecutive period as the starting point of a concentration increase at that node, marking the beginning of the impact of an emission event reaching that node. Similarly, when the concentration change rate is below the preset decrease threshold for multiple consecutive sampling periods, the system determines the start time of that consecutive period as the end point of the concentration decrease, marking the substantial subsidence of the impact of the emission event at that node. The system performs the above detection synchronously on all monitoring nodes, and records the rising start point and falling end point of each node as the time boundary for that node's participation in the formation of the concentration cloud. When two or more independent concentration clouds spatially overlap and the peak concentration within the overlapping area exceeds a predetermined proportion of the sum of the concentrations of each cloud individually, extrapolated from the concentration distribution before overlap, the overlapping area is defined as a concentration overlap region. The system records the moment the concentration superposition effect first appears in the region, the direction vector of cloud expansion, and the dissipation process as the concentration superposition effect fades below the level of a single cloud. Based on the concentration rise and fall endpoints of each monitoring node, a spatiotemporal clustering algorithm identifies groups of nodes with similar rise times and spatial proximity as an independent concentration cloud. Each cloud is associated with a unique emission event identifier. The system continuously monitors the boundary changes of different cloud clouds. When the spatial boundaries of two or more cloud clouds first come into contact or overlap, the system delineates the overlap range within the overlapping area. For this overlap range, the system extracts the concentration distribution trend of each cloud cloud before the overlap occurs, extrapolates the concentration change patterns from several sampling times before the overlap, estimates the concentration value that each cloud cloud should have in the overlapping area if it existed alone, and adds the estimated values ​​of each cloud cloud to obtain the sum of the concentrations when acting alone. The system then detects the actual concentration peak within the overlapping region. When this peak exceeds a set percentage of the sum of the individual concentrations (e.g., 120%, meaning the cumulative effect is more than 1.2 times the sum of the individual concentrations), the system formally defines the overlapping region as a concentration overlap region. The system records the sampling time when this condition is first met as the formation time of the concentration superposition effect. Simultaneously, based on the expansion trajectory of each cloud before entering the overlapping region, the system calculates the direction vector of each cloud's expansion. Subsequently, the system continuously tracks the change in the ratio of the actual concentration to the sum of the individual concentrations within the overlapping region. When this ratio continuously decreases and first falls below the exit threshold (e.g., 105% of the sum of the individual concentrations), this time is recorded as the dissipation completion time, and the entire process from the formation time to the dissipation completion time is recorded as the dissipation process.

[0021] The process in S1 of forming an emission event impact chain based on the concentration transfer intensity corresponding to the succession relationship is as follows: Based on the formation time of each concentration overlap region, multiple overlapping regions appearing in the same spatial grid cell are sorted chronologically. The emission activities corresponding to the concentration overlap regions with earlier formation times are marked as "pre-emission events," and those corresponding to the concentration overlap regions with later formation times are marked as "post-emission events." All concentration overlap regions identified according to the method of claim 2 are obtained, each overlap region being associated with its specific formation time and the identifiers of the emission activities participating in the superposition within that region. The system divides the monitoring area into spatial grid cells of a uniform scale, each grid cell having a fixed spatial range. For each grid cell, the system retrieves all concentration overlap regions that have appeared within the spatial range of that grid cell and sorts them in ascending order according to the formation time of each overlap region. For any two overlap regions formed at different times within the same grid cell, the system marks the original emission activity corresponding to the overlap region with the earlier formation time as a "pre-emission event," and the original emission activity corresponding to the overlap region with the later formation time as a "post-emission event." If three or more overlap regions appear in the same grid cell, adjacent preceding and following events are paired sequentially according to their formation times. The system employs a cloud boundary tracking method to calculate the rate and arrival time of the expansion of the boundary of the subsequent overlapping region towards the concentration centroid of the preceding overlapping region. This determines the moment when the boundaries of the subsequent and preceding overlapping regions come into contact. Once contact occurs, the spatial overlap ratio between the two regions is calculated. If the spatial overlap ratio exceeds a preset threshold, an impact on the establishment of a connection relationship is determined. For each pair of overlapping regions marked with preceding and following emission events, the concentration centroid position of the preceding overlapping region at the moment its concentration superposition effect reaches its peak is extracted. This centroid position is obtained by weighted averaging of the concentration values ​​and coordinates of each grid cell within the calculation area. The trajectory of the subsequent overlapping region's outward expansion from its formation moment is extracted. The system uses a cloud boundary tracking method, identifying the point on the boundary of the subsequent overlapping region closest to the concentration centroid of the preceding overlapping region at each sampling period, and recording the change in distance between this point and the concentration centroid of the preceding overlapping region over time. By numerically differentiating the distance variation curve, the instantaneous rate of expansion of the boundary of the rear overlapping region towards the centroid of the concentration in the front overlapping region is obtained. The rates of multiple consecutive sampling periods are smoothed to obtain stable expansion rate values. The system continuously monitors the minimum distance between the boundaries of the rear and front overlapping regions. When this minimum distance first reaches zero, the moment is recorded as the boundary contact moment. Starting from the boundary contact moment, the system calculates the spatial overlap ratio between the rear and front overlapping regions, i.e., the ratio of the number of grid cells in the overlapping portion to the total number of grid cells in the front overlapping region and the total number of grid cells in the rear overlapping region, respectively. The smaller of the two values ​​is taken as the spatial overlap ratio. When this spatial overlap ratio exceeds a preset threshold, the system formally determines that the influence relationship between the preceding and following emission events has been established, and records the moment of establishment as the first sampling moment after boundary contact that satisfies the overlap ratio condition. Multiple impact relationships established sequentially are linked together to form an emission event impact chain, with emission events as nodes and impact relationships as edges. The chain records the establishment time, spatial span, and concentration transfer intensity of each impact relationship. The concentration transfer intensity is the ratio between the average concentration of the overlapping area at the time of establishment and the peak concentration of the previously overlapping area. All established impact relationships within the monitoring area are traversed. Each relationship includes a preceding emission event identifier, a following emission event identifier, the establishment time, and the spatial overlapping area at the time of establishment. All impact relationships are globally sorted in ascending order based on their establishment time. A directed graph construction method is then used: each unique emission event is considered a node, and a directed edge is drawn from each preceding emission event node to its corresponding following emission event node, with the direction of the edge representing the direction of impact transfer. When multiple impact relationships form a sequential sequence, they are automatically linked together into an emission event impact chain. For each directed edge representing a connection, the system records the following three parameters in the chain: First, the connection establishment time, i.e., the specific moment when the boundary of the subsequent emission event expands to the concentration centroid of the preceding emission event and effectively overlaps; Second, the spatial span, i.e., the straight-line distance between the concentration centroids of the preceding and subsequent emission events, and the actual path length traversed during the influence transmission process; Third, the concentration transmission intensity, which is calculated as follows: at the connection establishment time, extract the concentration values ​​of all grid cells in the spatially overlapping area and calculate their arithmetic mean to obtain the average concentration of the overlapping area, then extract the peak concentration of the previously overlapping area at its peak time, divide the average concentration of the overlapping area by the peak concentration of the previously overlapping area, and the resulting ratio is the concentration transmission intensity. The closer this ratio is to a certain value, the more completely the influence of the preceding emission event has been transmitted to the subsequent emission event.

[0022] S2: Track the duration, concentration transfer intensity, and concentration contribution ratio of each impact relationship along the emission event impact chain, calculate the coverage of previous emission events in subsequent concentration changes, and calculate the historical contribution retention of each emission event based on the coverage and corresponding concentration contribution value to form a historical contribution record.

[0023] The process for calculating the coverage of preceding emission events in subsequent concentration changes in S2 is as follows: The process involves sequentially extracting the preceding emission event identifier and the subsequent concentration change period from each connection relationship along the direction of the emission event influence chain. The subsequent concentration change period is the time interval between the establishment of the influence connection relationship and the dissipation of the corresponding subsequent emission event. The emission event influence chain is then read. Following the direction of the influence chain, i.e., from earlier emission events to later emission events, each connection relationship in the chain is traversed sequentially. For the currently traversed connection relationship, the stored "previous emission event identifier" is extracted. This identifier corresponds to a previously occurring original emission activity. The "subsequent concentration change period" corresponding to this connection relationship is determined. The starting point of this period is the moment the influence connection relationship is established, i.e., the sampling moment when the boundary of the subsequent emission event and the concentration centroid of the preceding emission event effectively overlap and the spatial overlap ratio exceeds the connection threshold; the ending point of this period is the moment when the subsequent emission event dissipates, i.e., the moment when the concentration superposition effect in the concentration overlap area corresponding to the subsequent emission event fades to below the level of a single cloud cluster. The entire time interval from the starting point to the ending point is extracted as the subsequent concentration change period under this connection relationship. During the subsequent concentration change period, the concentration component contributed by the preceding emission event is continuously monitored. Concentration tracer factors or correlation analysis methods are used to separate the proportion of the preceding emission event in the total concentration. To separate the concentration component solely contributed by the preceding emission event from the total concentration, one or a combination of the following two methods is selected based on the compositional characteristics of the VOCs emitted by the preceding emission event. The first method is the concentration tracer factor method: In the initial stage of the preceding emission event, one or more tracer components unique to the emitted substances of the event are identified. These tracer components occupy a fixed and known proportion in the total VOCs concentration emitted by the event. During the subsequent concentration change period, the concentration change of this tracer component at the same monitoring node is measured, and then the total VOCs concentration component contributed by the preceding emission event is deduced based on the known proportion. The second method is the correlation analysis method: Concentration change curves of the preceding emission event acting alone in historical periods without interference from other events are collected as a standard template. The concentration time-series curves of each monitoring node during the subsequent concentration change period are compared with this template, and the correlation degree between the two at each time point is calculated. The correlation degree is used as a weight to estimate the contribution proportion of the preceding emission event. At each sampling time, a contribution percentage value is calculated for each monitoring node. This value is between zero and one, representing the proportion of the total concentration at that node that comes from pre-emission events at that time. Based on the curve of the percentage decaying over time, the duration for which the impact of the pre-emission event remains above a set contribution threshold during the subsequent concentration change is calculated. The ratio of this duration to the total duration of the subsequent concentration change is used as the coverage level, and the gradient of coverage with spatial location is recorded. The gradient is the ratio of the difference in coverage between adjacent monitoring locations to the corresponding spatial distance. For each monitoring node, the sequence of the node's contribution percentage changing over time during the entire subsequent concentration change period is extracted, and a curve showing the contribution percentage gradually decaying from high to low is plotted. A contribution threshold is pre-set, for example, 10%, indicating that the impact of the pre-emission event is considered to be still significant when its contribution percentage is still not less than 10% of the total concentration. Starting from the beginning of the subsequent concentration change period, the search proceeds along the decay curve to find the moment when the contribution percentage last drops below the threshold. The time between this moment and the beginning is recorded as the duration for which the impact of the pre-emission event remains above the contribution threshold. The total duration of the subsequent concentration change period is recorded, i.e., the time from the beginning to the end of dissipation. The ratio of this duration to the total duration of the subsequent concentration change is defined as the coverage level of the pre-emission event on the subsequent concentration change at that monitoring node. The closer the coverage level is to one, the more significant the impact of the preceding emission event remains throughout the subsequent concentration change. The system repeats the above calculation for all monitoring nodes within the spatial range covered by the system during the subsequent concentration change period to obtain the specific value of the coverage level at each node. The gradient of coverage level change with spatial location is calculated: for any two adjacent monitoring nodes, the difference between their coverage levels and the straight-line distance between them is calculated. The rate of change obtained by dividing the difference by the distance is the gradient in that direction. The gradients in multiple directions around each monitoring node are recorded, and areas with large gradients are identified. These areas typically correspond to boundary locations where the impact of the preceding emission event rapidly decays or is strongly disturbed by other emission events.

[0024] The process of creating historical contribution records in S2 is as follows: For each node in the emission event impact chain, the preceding emission events are weighted and accumulated based on their coverage in subsequent succession relationships and their corresponding concentration contributions. The weighting coefficient decreases exponentially with the succession distance, which is the number of directed edges traversed from the event node to the node belonging to the subsequent concentration change period in the impact chain. The generated emission event impact chain and the coverage in each succession relationship are obtained. For a node in the emission event impact chain that is a preceding emission event (denoted as event E), starting from that node, all succession relationships directly or indirectly involved with event E as a preceding event are retrieved downstream along the impact chain. Direct involvement means that event E is a preceding emission event in a certain succession relationship; indirect involvement means that the influence of event E is transmitted to the subsequent concentration change period downstream through intermediate events, i.e., event E → event A → event B, so event E also has an indirect influence on the subsequent concentration change period of event B. For each retrieved succession relationship, the succession distance from event E to the corresponding subsequent concentration change period is calculated. The transmission distance is defined as the number of directed edges traversed on the influence chain from node E of event E to the node belonging to the subsequent concentration change period (e.g., distance 1 for direct transmission, distance 2 for transmission with an intermediate event, and so on). A pre-set attenuation coefficient λ (usually between 0.5 and 0.8, adjusted according to the meteorological diffusion conditions and spatial scale of the monitoring area) is used. For transmission relationships with a transmission distance of d, the weight coefficient is calculated according to the formula w(d) = exp(-λ·d), meaning the weight coefficient decreases exponentially with the increase of transmission distance d. For a direct transmission relationship with a distance of 1, the weight coefficient is exp(-λ); for an indirect transmission relationship with a distance of 2, the weight coefficient is exp(-2λ), and so on. The system multiplies the calculated coverage degree of each retrieved transmission relationship by the corresponding weight coefficient to obtain the weighted contribution value of this influence transmission. The weighted contribution values ​​of all retrieved transmission relationships are summed to obtain the total weighted contribution value of event E in all subsequent transmission relationships. The weighted cumulative concentration contribution value is used as the historical contribution retention amount of the emission event within the monitoring area. The weighted cumulative concentration contribution value of event E is directly defined as the historical contribution retention amount of the emission event within the monitoring area at the current monitoring time. This historical contribution retention amount is a dimensionless relative value, and its physical meaning is: after the occurrence of event E, the VOCs emitted by it undergo a series of processes such as diffusion, dilution, and overlap with subsequent emission events, and still remain in the monitoring area in an identifiable manner. Since the weighting coefficient decays exponentially with the distance of connection, the emission event farther away from the current monitoring time and with more intermediate connection links has a smaller historical contribution retention amount, which is consistent with the objective law of the exponential decay of pollutant concentration with time in atmospheric diffusion. The above weighted cumulative calculation is performed on all emission events in the emission event influence chain (including the first, middle, and last events in the chain) to obtain the historical contribution retention amount corresponding to each emission event. For chain tail events (i.e. events that have not yet participated in any subsequent succession relationships as preceding events), their historical contribution retention only includes their own weighted contribution as preceding events in the existing direct succession relationships. If there are no subsequent succession relationships, their historical contribution retention is temporarily recorded as zero, and will be updated after a new succession relationship is established. Following the original chronological order of emission events, the system associates and stores the identifier, location, time of occurrence, and corresponding historical contribution retention of each emission event, forming a traceable historical contribution record. From the emission event impact chain and the initial emission event database, the system extracts the following four basic pieces of information for each emission event: a unique event identifier (an automatically generated code, such as "EVT_year_month_day_serial number"), location (including latitude and longitude coordinates and emission altitude, accurate to the meter level), and time of occurrence (the start time when the emission activity was first detected, accurate to the second level). The system then associates and binds this basic information with the historical contribution retention calculated in the second step for that emission event. Next, the system sorts all emission events according to their chronological order of occurrence (from earliest to latest). After sorting, the system writes all four pieces of information for each event (event identifier, location, time of occurrence, and historical contribution retention) into a persistent storage medium, such as a relational database table or a time-series database, in the form of structured data records. This storage structure supports fast retrieval by event identifier, time interval, or spatial location range. For the same emission event, as new continuation relationships emerge (i.e., the impact of the event continues to propagate downstream), the system periodically recalculates its historical contribution retention and updates the corresponding records in the storage medium, while retaining the timestamp of the update, forming a traceable change history. The resulting historical contribution record not only includes the retained contribution of each emission event at the current moment, but also fully records the entire lifecycle impact trajectory of the event from its occurrence to the current moment.

[0025] S3: Compare the historical contribution retention with the current monitoring concentration according to the corresponding monitoring location, calculate the concentration difference and the correlation coefficient between the current concentration field and the spatial distribution of historical contributions, identify the concentration components formed by the continuation of historical contributions, and deduct the historical contribution retention step by step according to the preset stripping step size. Determine the stripping termination condition based on the correlation between the remaining concentration field and the spatial distribution of historical contributions to obtain the new concentration contribution record.

[0026] The process for identifying concentration components formed by the continuation of historical contributions in S3 is as follows: Calculate the difference and ratio between the measured concentration and the historical contribution retention. When the absolute value of the difference between the monitored concentration and the historical contribution retention is less than a set deviation threshold, and the correlation coefficient between the current concentration field and the spatial distribution of historical contributions is higher than a set correlation threshold, it is determined that the concentration at the corresponding monitoring point is formed by the continuation of historical contributions. Obtain the measured VOCs concentration value and the corresponding historical contribution retention for each monitoring point. Calculate the difference between the two, i.e., the measured concentration minus the historical contribution retention, and take the absolute value of the difference. A deviation threshold is preset, which is calibrated based on the measurement error range of the monitoring instrument and the background concentration fluctuation level of the monitoring area. Calculate the spatial correlation coefficient between the entire current concentration field and the historical contribution spatial distribution field. This correlation coefficient reflects whether the concentration value change trends of the two concentration fields are consistent at each point. The closer the correlation coefficient is to one, the more similar the spatial distribution patterns of the two. When the absolute value of the difference between the measured concentration at a certain monitoring point and the historical contribution retention is less than the preset deviation threshold, and the correlation coefficient between the current concentration field and the spatial distribution of historical contributions is higher than the preset correlation threshold, it is determined that the current measured concentration at that monitoring point is entirely or mostly formed by the continuation of historical contributions, that is, there is no significant new emission activity interference at that point. When the measured concentration is higher than the historical contribution retention and the difference exceeds a set deviation threshold, it is determined that there is a new concentration contribution. When the measured concentration at a monitoring point is significantly higher than the historical contribution retention for that point, and the difference exceeds a preset deviation threshold, it is determined that there is an additional concentration source at that monitoring point besides the concentration formed by the continuation of historical contributions. This excess concentration is determined as a new concentration contribution, that is, the VOCs concentration generated by recently occurring emission activities, and the excess part at the point is marked as the new concentration contribution value. For cases where the measured concentration is lower than the historical contribution retention, it is determined that part of the historical contribution at that point has dissipated or transformed, and the remaining historical contribution continuation concentration is equal to the measured concentration itself, and no new contribution is generated. When the absolute value of the difference between the monitored concentration and the historical contribution retention is less than a set deviation threshold, but the correlation coefficient between the current concentration field and the spatial distribution of historical contributions is lower than a set correlation threshold, it is determined that the current concentration and the spatial distribution of historical contributions do not match, and the corresponding concentration is included in the analysis scope of new concentration contributions. When the absolute value of the difference between the measured concentration and the historical contribution retention at a certain monitoring point is less than the deviation threshold, that is, the two are close in value, but the correlation coefficient between the entire current concentration field and the spatial distribution field of historical contributions is lower than a preset correlation threshold, it is determined that the spatial distribution pattern of the current concentration is inconsistent with the expected distribution pattern of historical contributions. This situation may mean that although the concentration values ​​are similar, the high-value and low-value areas of the concentration have shifted, or that a new emission event has changed the original concentration distribution pattern. Instead of simply attributing this situation to the continuation of historical contributions, the concentration value at this monitoring point is included in the analysis scope of new concentration contributions. Based on the judgment results of each monitoring point, a historical contribution continuation concentration distribution map is generated, marking the proportion and spatial distribution characteristics of the historical contribution retention concentration in the current total concentration. The judgment results of all monitoring points are summarized, and each point is categorized into one of the following states according to the three rules mentioned above: "historical contribution continuation formation," "existence of new contribution," or "mismatch pending analysis." The system extracts all points judged as having historical contribution continuation formation, and uses the measured concentration values ​​at these points as the historical contribution continuation concentration values. For points with new contribution, the historical contribution retention amount is used as the historical contribution continuation concentration value. Points that do not match and await analysis are not included in the historical contribution continuation concentration distribution. The historical contribution continuation concentration values ​​from discrete points are extended to a continuous grid across the entire monitoring area, generating a historical contribution continuation concentration distribution map. On the same distribution map, the proportion of historical contribution continuation concentration to the current measured total concentration within each grid cell is calculated, and the spatial distribution of this proportion is marked with different colors or fill patterns. For example, areas with a proportion higher than 80% are represented by dark red, proportions between 50% and 80% by orange, and proportions lower than 50% by yellow or light colors. The following spatial distribution characteristics are also marked: the central location and coverage of the high-value area of ​​historical contribution concentration, the width of the transition zone between the high-proportion area and the low-proportion area, and the attenuation trend of the historical contribution proportion along the prevailing wind direction.

[0027] The process of obtaining new concentration contribution records in S3 is as follows: Based on the identified components of historical contribution continuation, the concentration values ​​of historical contribution continuation at each monitoring point are subtracted sequentially from the total concentration field of the current monitored concentration. A historical contribution continuation concentration distribution map is obtained, from which the concentration values ​​identified as historical contribution continuation at each monitoring point are extracted. The measured total VOCs concentration values ​​at the same monitoring point within the current monitoring period are obtained; the measured total concentrations at all points together constitute the total concentration field of the current monitored concentration. The subtraction operation is performed one by one for each monitoring point: for a given monitoring point, the measured total concentration value at that point is subtracted from the historical contribution continuation concentration value at that point, resulting in a subtracted value. Here, the subtracted value is the determined "historical contribution continuation concentration value," that is, the portion of the concentration identified as originating from the continuation of historical emission activities after difference comparison and spatial correlation judgment. For locations identified as having "new contributions," the concentration value resulting from the continuation of historical contributions equals the retained historical contribution amount; for locations identified as having "continuation of historical contributions," this value equals the measured concentration itself, and the result after deduction is zero; for locations identified as "mismatched and awaiting analysis," the system does not perform deductions temporarily and reserves them for later processing. The deduction result value for each location is temporarily stored for further judgment. After deducting concentrations at each monitoring point, the remaining concentration value is calculated and used as the new concentration value for the corresponding monitoring point. For each monitoring point that has undergone the deduction operation, the result after deduction is defined as the remaining concentration value for that point. This remaining concentration value reflects the concentration component remaining in the current total concentration after excluding the continuation of historical contributions. This remaining concentration value is initially identified as the new concentration value for that point, i.e., the concentration contributed by recently occurring emission activities. When the remaining concentration value after deduction is less than zero, the new concentration value for the corresponding monitoring point is set to zero; when the remaining concentration value after deduction is greater than zero, the new concentration value for the corresponding monitoring point is retained. The remaining concentration values ​​for each monitoring point are then numerically checked. When the remaining concentration value is less than zero, this is usually due to measurement errors, spatial interpolation errors, or estimation biases in historical contribution retention, resulting in a non-physical negative value. Because the actual VOCs concentration in the atmosphere cannot be negative, the remaining concentration value for such points is forcibly corrected to zero, and the new concentration value for that point is set to zero, indicating that no new contribution was detected. When the remaining concentration value is greater than zero, this value is retained as the new concentration value for that point, indicating that there is a quantifiable new emission contribution at that point. When the remaining concentration value is equal to zero, zero is also retained, indicating that the new contribution at that point is zero. After correction, the final new concentration value for each monitoring point is obtained, and all values ​​are non-negative. The newly added concentration values ​​at each monitoring point are summarized to form a spatial distribution field of the newly added concentration, serving as a record of its contribution. The final determined newly added concentration values ​​at all monitoring points are then compiled, with each point including its coordinates and corresponding newly added concentration value. The newly added concentration values ​​at discrete points are extended to a continuous spatial grid across the entire monitoring area, generating a spatial distribution field of the newly added concentration. This field is presented as a two-dimensional planar cloud map or a three-dimensional isosurface, which can intuitively display the location of high-value areas, diffusion range, concentration gradient changes, and extension patterns along the prevailing wind direction.

[0028] S4: When a newly added concentration contribution record can correspond to multiple candidate emission sources at the same time, the number of times each candidate emission source has received influence, the degree of contribution retention, and the concentration response characteristics are statistically analyzed in the historical stage to determine the contribution competition relationship between each candidate emission source and form the source contribution allocation result.

[0029] The process in S4 to statistically analyze the number of impacts, contribution retention levels, and concentration response characteristics of each candidate emission source during historical periods is as follows: The spatial distribution pattern and diffusion gradient of newly added concentration contribution records are extracted and matched with the historical emission locations, emission heights, and emission intensities of each candidate emission source to screen out multiple candidate emission sources that can generate newly added concentration distributions. The newly added concentration contribution records are read, and the spatial distribution pattern of the newly added concentration throughout the monitoring area is extracted, including the center coordinates of the high-value area, the shape and range of the high-value area, and the direction and density of the concentration contour lines. The diffusion gradient of the newly added concentration field is calculated, i.e., the rate of decrease in concentration value extending from the center of the high-value area outwards, and the difference in the rate of decrease along different directions. A pre-established database of all potential emission sources within the monitoring area is accessed to obtain basic information corresponding to each potential emission source within the monitoring area. The potential emission source database is established by integrating pollution source directory data filed with environmental protection authorities, emission facility information declared by enterprises, and emission source information obtained through inversion of historical monitoring data. Specifically, the system obtains the location coordinates, emission outlet types, and emission facility information of each emission source based on environmental protection filing documents; it obtains the emission outlet height, production conditions, and historical emission records based on enterprise declaration documents; and it uses concentration diffusion inversion analysis, combined with historical VOCs concentration monitoring data within the monitoring area, to identify potential emission points that are not included in the filing management but have long-term concentration response characteristics, and adds them to the database. For each candidate emission source in the database, the system records its historical emission location coordinates, emission outlet height above ground, historical average emission intensity range, and historical concentration response characteristic parameters. The historical average emission intensity range is obtained from historical emission monitoring data or enterprise declaration data, and the historical concentration response characteristic parameters are calculated based on the concentration rise slope, peak arrival time, and decay time constant caused by the corresponding emission source during the historical monitoring period. The database is updated according to a preset update cycle to ensure that the emission source information is consistent with the actual emission situation. The system compares the distance between the center coordinates of the newly added high-value area and the location coordinates of each candidate emission source, selecting emission sources located within a preset radius of the high-value area center as preliminary candidates. The diffusion gradient of the newly added concentration field is matched with the theoretical diffusion characteristics of each preliminary candidate emission source: based on the emission height and intensity of each candidate emission source, combined with the current meteorological conditions (wind direction, wind speed, atmospheric stability) of the monitoring area, the spatial distribution pattern and diffusion gradient of the concentration that should be generated when the source emits alone are simulated in a forward manner. The system calculates the spatial similarity between the simulated distribution and the actual newly added concentration distribution, retains emission sources with similarity exceeding a preset threshold, and removes emission sources with too low similarity. After the above two rounds of screening by location matching and diffusion characteristic matching, a list of multiple candidate emission sources that can generate the current newly added concentration distribution is output, and each candidate source is accompanied by its spatial matching degree score with the actual newly added concentration distribution; For each candidate emission source, the number of impact relays within a set time window prior to the current monitoring time is counted from historical contribution records, along with the average degree to which the emission source's contribution is retained in each relay. For each selected candidate emission source, historical contribution records are retrieved to extract all records of the emission source's participation in the emission event impact chain during the historical period. A time window is set, such as the past 24 or 48 hours prior to the current monitoring time, and only historical relay events occurring within this time window are counted. Within this time window, each relay relationship is retrieved one by one to determine whether the candidate emission source appears as a preceding emission event in the relay relationship. Each occurrence is recorded as one impact relay count. For each retrieved relay relationship, the coverage degree of the candidate emission source in that relay is read, reflecting the proportion of the emission source's contribution retained during subsequent concentration changes. The coverage degrees corresponding to all retrieved relay relationships are summed and divided by the total number of impact relays to obtain the average degree to which the candidate emission source's contribution is retained during the historical period. For example, if a candidate emission source has participated in five impact takeovers in the past 24 hours, with the coverage levels of the five takeovers being high, medium, high, low, and medium, the arithmetic mean of these five coverage levels is calculated as the average contribution level of the emission source. The concentration response characteristics of each candidate emission source during historical emission events were analyzed, including the concentration rise slope, peak arrival time, and decay time constant. For each candidate emission source, the concentration time series data of that source as an independent emission event (i.e., before overlapping with other emission events) were retrieved from the historical contribution records. The concentration change curve from the start of emission to the peak concentration was extracted, and the magnitude of the concentration increase per unit time, i.e., the concentration rise slope, was calculated. This slope reflects the rate of release intensity of the emission source. The time length from the start of emission to the peak concentration was recorded as the peak arrival time, which reflects the transport distance and diffusion conditions between the emission source and the monitoring node. After the concentration reaches the peak, the stage from the peak to the background level or to half of the peak concentration was extracted, and the rate of concentration decay during this stage was calculated, i.e., the decay time constant. This constant comprehensively reflects the diffusion and dilution rate and chemical decay characteristics of the VOCs emitted by the emission source in the atmosphere.

[0030] The process of forming the source contribution allocation results in S4 is as follows: The historical number of times each candidate emission source has been affected, its contribution retention level, and its concentration response characteristics are input into the Bayesian multi-source allocation model to calculate the prior contribution probability of each candidate emission source to the current increase in concentration. The three characteristic values ​​(historical number of times affected, average contribution retention level, and concentration response characteristics) of each candidate emission source are normalized and then input into the pre-defined Bayesian multi-source allocation model. Based on the historical performance of each candidate source, the prior probability of it being the main source of the current increase in concentration is calculated. Candidate sources with more historical affected times, higher contribution retention levels, and more consistent concentration response characteristics with historical patterns have higher prior probability values. The model normalizes the sum of the prior probabilities of all candidate sources to one and outputs the prior contribution probability corresponding to each candidate source. Based on the similarity between the newly added concentration contribution records and the concentration distributions generated by each candidate emission source acting alone, the prior contribution probability is iteratively updated to obtain the posterior contribution probability. For each candidate emission source, based on its historical emission intensity, emission height, and location coordinates, combined with current meteorological conditions, the theoretical concentration distribution generated by the source emitting alone within the monitoring area is positively simulated. The spatial similarity between this theoretical concentration distribution and the actual newly added concentration contribution records is calculated, including the overlap of high-value areas, the matching degree of contour morphology, and the consistency of diffusion gradient. According to the similarity level, the system iteratively corrects the prior contribution probability: the probability of candidate sources with high similarity is increased, and the probability of candidate sources with low similarity is decreased. After each round of correction, the system is re-normalized, and the iteration is repeated until the probability value change is less than a preset stable threshold, thus obtaining the final posterior contribution probability. When the posterior contribution probabilities of two or more candidate emission sources all exceed a set threshold, a contribution competition relationship is determined to exist between them. The concentration share in the new concentration contribution record is allocated according to the ratio of their posterior contribution probabilities, forming the source contribution allocation result. The posterior contribution probabilities of all candidate emission sources are sorted. When the probability values ​​of two or more candidate sources simultaneously exceed a preset effective contribution threshold, the system determines that a contribution competition relationship exists between these sources, meaning that the current new concentration is formed by the combined effect of multiple emission sources, rather than a single source. The total concentration share in the new concentration contribution record is allocated to each competing source according to the proportional relationship between their posterior contribution probabilities. Sources with higher probabilities are allocated larger concentration shares, and sources with lower probabilities are allocated smaller concentration shares. The sum of the allocated shares for all sources equals the total new concentration record value. The identifier of each competing source, its allocated concentration share, and its corresponding posterior contribution probability value are recorded together to form the source contribution allocation result.

[0031] S5: Continuously track the stability of the source contribution allocation results in the subsequent concentration change process. When the increase in concentration occurs again, determine the source attribution ratio of the new concentration based on the current historical contribution record, the current new concentration contribution record and the source contribution allocation result corresponding to the current moment. Dynamically correct the contribution value of each emission source, form a source contribution change record, and output the corresponding VOCs gas concentration distribution source tracing analysis results.

[0032] The process of outputting the corresponding VOCs gas concentration distribution source analysis results in S5 is as follows: During the continuous monitoring period following the formation of source contribution allocation results, the contribution allocation ratio of each emission source is recalculated at set time intervals. The variance of each ratio value fluctuates over time. Periods with variance below a stability threshold are marked as allocation stabilization periods, and periods with variance exceeding a fluctuation threshold are marked as allocation adjustment periods. After the source contribution allocation results are obtained, a continuous tracking process is initiated. At fixed time intervals (e.g., every five or ten minutes), newly added concentration contribution records within the current monitoring period are reread, and the contribution allocation ratio of each emission source is recalculated. As time progresses, the ratio values ​​of each emission source form a time series. The system uses a sliding window method to calculate the degree of fluctuation of each ratio value in this series, i.e., variance. When the variance of the ratio values ​​of all emission sources within a continuous period is below the preset stability threshold, the period is marked as an allocation stabilization period, indicating that the contribution relationship of each emission source is relatively certain; when the variance exceeds the preset fluctuation threshold, it is marked as an allocation adjustment period, indicating that the contribution relationship is changing. When a new concentration increase is detected, the starting time, peak concentration, and spatial distribution of the growth peak are extracted. Combined with the historical contribution records and the corresponding source contribution allocation results at the current time, the proportion of the growth peak attributable to each historical emission source is calculated. The concentration change trend at each monitoring point is monitored in real time. When a significant increase in the new concentration is detected within a short period (e.g., the concentration increase rate exceeds three times the background change rate for three consecutive sampling periods), the starting time of the growth peak, the peak concentration value corresponding to the peak arrival time, and the spatial distribution of the peak within the entire monitoring area are recorded. The historical contribution records at the current time and the source contribution allocation results within the most recent stable allocation period are retrieved. A pattern matching method is used to compare the spatial distribution of the growth peak with the contribution spatial patterns of each historical emission source during the stable period. Combining the temporal correlation between the peak start time and the historical emission times of each emission source, the proportion of the growth peak attributable to each historical emission source is calculated. The contribution values ​​of each emission source are dynamically adjusted based on the calculated attribution ratio. The contribution values ​​before and after adjustment, along with the adjustment time, are recorded as source contribution change records. These records are then linked to the corresponding concentration overlap areas and emission event impact chains, serving as the source tracing analysis results for VOCs gas concentration distribution. Based on the calculated attribution ratio, the cumulative contribution value of each emission source in the historical contribution record is dynamically adjusted: the concentration share of the current peak increase is added to the historical cumulative contribution value of the corresponding emission source according to the attribution ratio, forming a new, adjusted cumulative contribution value. The contribution value before adjustment, the adjusted contribution value, and the execution time of this adjustment for each emission source are recorded, forming a source contribution change record. This source contribution change record is linked to the concentration overlap area that triggered the adjustment (i.e., the spatial area where the peak increase is located) and the corresponding emission event impact chain. All historical contribution records of emission events, source contribution allocation results, source contribution change records, lists of concentration overlap areas, and sets of emission event impact chains are packaged and output as complete VOCs gas concentration distribution source tracing analysis results, allowing environmental regulators to analyze and judge the real-time contribution status and dynamic trends of each emission source within the monitoring area.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A VOCs gas concentration distribution source tracing analysis method, characterized in that, Includes the following steps: VOCs concentration data collected by multiple monitoring nodes within the monitoring area are acquired, and overlapping concentration areas formed by multiple emission activities within the monitoring area are identified. The formation time, expansion direction and dissipation process of each overlapping concentration area are recorded. An influence transmission relationship is established based on the spatial overlap ratio between the later-formed overlapping concentration area and the earlier-formed overlapping concentration area, and an emission event influence chain is formed based on the concentration transmission intensity corresponding to the transmission relationship. The duration, concentration transfer intensity, and concentration contribution ratio of each impact relationship are traced along the impact chain of emission events. The coverage of previous emission events in subsequent concentration changes is calculated. Based on the coverage and corresponding concentration contribution values, the historical contribution retention of each emission event is calculated to form a historical contribution record. The historical contribution retention is compared with the current monitoring concentration according to the corresponding monitoring location. The concentration difference and the correlation coefficient between the current concentration field and the spatial distribution of historical contributions are calculated. The concentration components formed by the continuation of historical contributions are identified. The historical contribution retention is deducted step by step according to the preset stripping step size. The stripping termination condition is determined according to the correlation between the remaining concentration field and the spatial distribution of historical contributions to obtain the new concentration contribution record. When a newly added concentration contribution record can correspond to multiple candidate emission sources at the same time, the number of times each candidate emission source has received its impact, the degree of contribution retention, and the concentration response characteristics in the historical stage are statistically analyzed to determine the contribution competition relationship between each candidate emission source and to form the source contribution allocation result. The stability of the source contribution allocation results is continuously tracked during subsequent concentration changes. When a new concentration increase occurs again, the source attribution ratio of the new concentration is determined based on the current historical contribution record, the current new concentration contribution record, and the source contribution allocation result corresponding to the current moment. The contribution values ​​of each emission source are dynamically corrected to form a source contribution change record, and the corresponding VOCs gas concentration distribution source tracing analysis results are output. 2.The VOCs gas concentration distribution source tracing analysis method according to claim 1, wherein, The process of recording the formation time, expansion direction, and dissipation process of overlapping concentration regions is as follows: A multi-node gas concentration sensor array is deployed within the monitoring area to collect VOCs concentration data at each monitoring node at a preset sampling period, and a concentration distribution field is constructed using a spatial interpolation method. Based on the concentration change rate at each monitoring node at adjacent sampling times, the starting point of concentration rise is determined when the concentration change rate is higher than the preset rise threshold for multiple consecutive sampling periods, and the ending point of concentration fall is determined when the concentration change rate is lower than the preset fall threshold for multiple consecutive sampling periods. Identify independent concentration clouds caused by different emission activities based on the connectivity of concentration boundaries in the concentration distribution field; When two or more independent concentration clouds overlap in space and the peak concentration in the overlapping area exceeds the set ratio of the sum of the concentrations of each cloud when acting alone, which is extrapolated from the concentration distribution before the overlap, the overlapping area is defined as the concentration overlap area. The time when the concentration superposition effect first appears in the area, the direction vector of cloud expansion, and the dissipation process of the concentration superposition effect fading to below the level of a single cloud are recorded. 3.The VOCs gas concentration distribution source apportionment method according to claim 1, wherein, The process by which an emission event impact chain is formed based on the concentration transfer intensity corresponding to the succession relationship is as follows: Based on the formation time of each concentration overlap region, multiple overlap regions appearing in the same spatial grid cell are sorted by time. The emission activities corresponding to the concentration overlap regions with earlier formation times are marked as earlier emission events, and the emission activities corresponding to the concentration overlap regions with later formation times are marked as later emission events. The cloud boundary tracking method is used to calculate the rate and arrival time of the expansion of the boundary of the rear overlapping region in the direction of the concentration centroid of the front overlapping region, and to determine the moment when the boundary of the rear overlapping region and the front overlapping region comes into contact. After the boundary of the rear overlapping region and the front overlapping region comes into contact, the spatial overlap ratio between the two is calculated. When the spatial overlap ratio exceeds the preset acceptance threshold, it is determined that it affects the establishment of the acceptance relationship. Multiple impact relationships established sequentially are linked together to form an emission event impact chain with emission events as nodes and impact relationships as edges. The establishment time, spatial span, and concentration transfer intensity of each impact relationship are recorded in the chain. The concentration transfer intensity is the ratio between the average concentration of the overlapping area at the time of the impact relationship establishment and the peak concentration of the previously overlapping area. 4.The VOCs gas concentration distribution source tracing analysis method according to claim 1, wherein, The process for calculating the coverage of pre-emission events in subsequent concentration changes is as follows: Along the direction of the emission event impact chain, extract the previous emission event identifier and the subsequent concentration change period in each succession relationship. The subsequent concentration change period is the time interval between the establishment of the impact succession relationship and the dissipation of the corresponding subsequent emission event. During the subsequent concentration change period, the concentration component contributed by the previous emission event was continuously monitored, and the proportion of the previous emission event in the total concentration was separated by concentration tracer factor or correlation analysis method. Based on the curve of the proportion decaying over time, the duration for which the impact of the pre-emission event remains above the set contribution threshold during the subsequent concentration change process is calculated. The ratio of the duration to the total duration of the subsequent concentration change is taken as the coverage level, and the gradient of the coverage level with spatial location is recorded. The gradient is the ratio of the difference in coverage level between adjacent monitoring locations to the corresponding spatial distance.

5. The VOCs gas concentration distribution source apportionment method according to claim 1, wherein, The process of creating a record of historical contributions is as follows: For each node in the emission event impact chain, the preceding emission events are weighted and accumulated based on their coverage in multiple subsequent connections and their corresponding concentration contributions. The weighting coefficient decreases exponentially with the increase of the connection distance. The connection distance is the number of directed edges traversed from the event node to the node belonging to the subsequent concentration change period in the impact chain. The weighted cumulative concentration contribution value is used as the historical contribution retention amount of the emission event within the monitoring area; Based on the original time sequence of emission events, the identifier, location, time of occurrence, and corresponding historical contribution retention of each emission event are associated and stored to form a traceable historical contribution record.

6. The VOCs gas concentration distribution source apportionment method according to claim 1, wherein, The process of identifying concentration components formed by the continuation of historical contributions is as follows: Calculate the difference and ratio between the measured concentration and the historical contribution retention. When the absolute value of the difference between the monitored concentration and the historical contribution retention is less than the set deviation threshold, and the correlation coefficient between the current concentration field and the spatial distribution of historical contribution is higher than the set correlation threshold, it is determined that the concentration of the corresponding monitoring point is formed by the continuation of historical contribution. When the measured concentration is higher than the historical contribution retention and the difference exceeds the set deviation threshold, it is determined that there is a new concentration contribution. When the absolute value of the difference between the monitored concentration and the historical contribution retention is less than the set deviation threshold, but the correlation coefficient between the current concentration field and the spatial distribution of historical contributions is lower than the set correlation threshold, it is determined that the current concentration and the spatial distribution of historical contributions do not match, and the corresponding concentration is included in the scope of new concentration contribution analysis. Based on the determination results of each monitoring point, a historical contribution continuation concentration distribution map is generated, marking the proportion of historical contribution retention concentration in the current total concentration and its spatial distribution characteristics.

7. The VOCs gas concentration distribution source apportionment method according to claim 1, wherein, The process of obtaining new concentration contribution records is as follows: Based on the identified components formed by the continuation of historical contributions, the concentration values ​​formed by the continuation of historical contributions at the corresponding locations are subtracted sequentially from the total concentration field of the current monitored concentrations, according to the monitoring points. After deducting the concentration at each monitoring point, the remaining concentration value is calculated and used as the new concentration value for the corresponding monitoring point. When the remaining concentration value after deduction is less than zero, the new concentration value of the corresponding monitoring point is set to zero; when the remaining concentration value after deduction is greater than zero, the new concentration value of the corresponding monitoring point is retained. The newly added concentration values ​​at each monitoring point are summarized to form a spatial distribution field of the newly added concentration, which serves as a record of the contribution of the newly added concentration.

8. The method for tracing and analyzing the concentration distribution of VOCs gas according to claim 1, characterized in that, The process of statistically analyzing the number of impacts, contribution retention levels, and concentration response characteristics of each candidate emission source over historical periods is as follows: The spatial distribution pattern and diffusion gradient of newly added concentration contribution records are extracted and matched with the historical emission location, emission height and emission intensity of each candidate emission source recorded in the emission source database. When the spatial distribution similarity and diffusion direction consistency both exceed the preset matching threshold, the corresponding emission source is determined as a candidate emission source. For each candidate emission source, the number of times the impact was absorbed within a set time window prior to the current monitoring time is counted from the historical contribution record, as well as the average degree to which the emission source contribution is retained in each absorption. The concentration response characteristics of each candidate emission source in historical emission events were analyzed, including the concentration rise slope, peak arrival time, and decay time constant.

9. The method for tracing and analyzing the concentration distribution of VOCs gas according to claim 1, characterized in that, The process of forming the source contribution allocation results is as follows: The historical number of times each candidate emission source has received a contribution, the degree of contribution retention, and the concentration response characteristics are input into the Bayesian multi-source allocation model to calculate the prior contribution probability of each candidate emission source to the current newly added concentration contribution record. Based on the similarity between the newly added concentration contribution records and the individual concentration distributions of each candidate emission source obtained by the diffusion model simulation, the prior contribution probability is iteratively updated to obtain the posterior contribution probability. When the posterior contribution probabilities of two or more candidate emission sources all exceed a set threshold, it is determined that there is a contribution competition relationship between them, and the concentration share in the newly added concentration contribution record is allocated according to the ratio of the posterior contribution probabilities, thus forming the source contribution allocation result.

10. The VOCs gas concentration distribution source apportionment method of claim 1, wherein, The process of outputting the corresponding VOCs gas concentration distribution source analysis results is as follows: During the continuous monitoring period after the source contribution allocation results are formed, the contribution allocation ratio of each emission source is recalculated at each set time interval. The variance of each ratio value fluctuates over time. Periods with variance below the stability threshold are marked as allocation stability periods, and periods with variance exceeding the fluctuation threshold are marked as allocation adjustment periods. When a new concentration is detected to increase again, the starting time, peak concentration and spatial distribution of the growth peak are extracted. Combined with the historical contribution record at the current time and the source contribution allocation result at the current time, the proportion of the growth peak attributable to each historical emission source is calculated. The contribution values ​​of each emission source are dynamically corrected based on the calculated attribution ratio. The contribution values ​​before and after correction, as well as the correction time, are recorded as source contribution change records. These source contribution change records are then linked with the corresponding concentration overlap areas and emission event impact chains and output as the source tracing analysis results of VOCs gas concentration distribution.

Images