Remote control system and method for exhaust gas on-line detection equipment

Abstract

This invention provides a remote control system and method for an online exhaust gas monitoring device, belonging to the field of remote control technology. The system includes: an exhaust gas monitoring network deployment module for deploying a gridded exhaust gas monitoring network in the monitoring area; an exhaust gas diffusion characteristic analysis module for performing dynamic diffusion characteristic analysis of exhaust gas; an anomaly verification node deployment module; a local consistency verification module for performing local consistency verification based on K diffusion verification topologies to obtain K device verification results; and a fatigue detection device positioning module for performing multi-source verification result fusion analysis, locating the fatigue detection device, and initiating remote device parameter adaptive refresh. This invention solves the technical problem in the prior art where the monitoring of online exhaust gas monitoring devices relies on manual or simple periodic inspections, and the equipment gradually experiences performance degradation after long-term operation, resulting in affected accuracy and reliability of exhaust gas data, thus impacting the overall performance of the equipment.

Description

Technical Field

[0001] This invention relates to the field of remote control technology, and specifically to a remote control system and method for an online exhaust gas detection device. Background Technology

[0002] Online exhaust gas monitoring equipment continuously collects the concentration of harmful gases in exhaust gases and transmits the data to a remote cloud platform for analysis in real time. This data allows for real-time monitoring of industrial enterprises' exhaust emissions, ensuring compliance with environmental protection requirements. However, during long-term operation, these online exhaust gas monitoring devices experience fatigue, meaning their performance gradually declines, leading to decreased data accuracy. Most online exhaust gas monitoring equipment relies on manual labor or simple periodic inspections. While some systems can send alarms, these monitoring methods are not intelligent enough to detect equipment fatigue in real time. When the equipment is fatigued, the accuracy and reliability of the exhaust gas data are affected, leading to data distortion or errors, thus impacting the overall performance of the equipment. Summary of the Invention

[0003] This application provides a remote control system and method for an online exhaust gas monitoring device, aiming to solve the technical problem that the monitoring of online exhaust gas monitoring devices in the prior art relies on manual or simple periodic inspections, and the device gradually experiences performance degradation after long-term operation, which affects the accuracy and reliability of exhaust gas data and thus affects the overall performance of the device.

[0004] The first aspect disclosed in this application provides a remote control system for an online exhaust gas monitoring device. The system includes: an exhaust gas monitoring network deployment module for deploying a gridded exhaust gas monitoring network in a monitoring area; an exhaust gas diffusion characteristic analysis module for dividing the exhaust gas monitoring network into K diffusion-related subnets by analyzing the dynamic diffusion characteristics of the exhaust gas in the monitoring area; an anomaly verification node deployment module for deploying K edge anomaly verification nodes locally in the K diffusion-related subnets; a local consistency verification module for, after P online exhaust gas monitoring devices in the exhaust gas monitoring network collect exhaust gas data in the monitoring area, group P real-time monitoring data into K groups of exhaust gas monitoring streams based on the K diffusion-related subnets, upload them to the K edge anomaly verification nodes, perform local consistency verification based on the K diffusion verification topologies, and obtain K device verification results; and a fatigue detection device positioning module for the K edge anomaly verification nodes to upload the K device verification results to a remote control cloud via an industrial IoT gateway, perform multi-source verification result fusion analysis, locate the fatigue detection device, and initiate remote device parameter adaptive refresh.

[0005] The second aspect of this application discloses a remote control method for an online exhaust gas monitoring device. The method is implemented through a remote control system for the aforementioned online exhaust gas monitoring device. The method includes: deploying a gridded exhaust gas monitoring network in a monitoring area; dividing the exhaust gas monitoring network into K diffusion-related subnets by analyzing the dynamic diffusion characteristics of the exhaust gas in the monitoring area; deploying K edge anomaly verification nodes locally in each of the K diffusion-related subnets; after P online exhaust gas monitoring devices in the monitoring network collect exhaust gas data from the monitoring area, grouping P real-time monitoring data into K groups of exhaust gas monitoring streams based on the K diffusion-related subnets, uploading them to the K edge anomaly verification nodes, performing local consistency verification based on the K diffusion verification topologies, and obtaining K device verification results; the K edge anomaly verification nodes upload the K device verification results to a remote control cloud via an industrial IoT gateway, performing multi-source verification result fusion analysis, locating fatigue detection devices, and initiating remote device parameter adaptive refresh.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects: By deploying a gridded exhaust gas monitoring network within the monitoring area, comprehensive monitoring of exhaust gas concentration and diffusion can be achieved. The gridded layout ensures the even distribution of monitoring points, thereby better capturing changes in exhaust gas within the area. This deployment method improves the coverage of exhaust gas monitoring, ensures monitoring without blind spots, and enhances the comprehensiveness of monitoring data. By analyzing the dynamic diffusion characteristics of exhaust gas within the monitoring area and dividing the exhaust gas monitoring network into K diffusion-related subnetworks based on the region's geographical and meteorological conditions, errors can be effectively reduced, the accuracy of exhaust gas data can be improved, and inconsistencies in monitoring results caused by differences in terrain and weather can be avoided. Deploying edge anomaly verification nodes locally within each diffusion-related subnetwork allows for preliminary anomaly detection and resolution of monitoring data locally. Verification reduces the computational burden on the central cloud, improving the system's real-time response capability and accuracy. Real-time exhaust gas data from the monitoring area is collected by online exhaust gas detection equipment. This data is then grouped according to the diffusion-related subnet and uploaded to edge verification nodes. Local consistency verification based on the diffusion verification topology is performed, ensuring the reliability of the monitoring data and reducing decision-making errors caused by data quality issues. After the edge verification nodes upload the equipment verification results to the remote control cloud, multi-source verification result fusion analysis is conducted. By analyzing the fusion results from multiple sources, the fatigue state of the equipment can be more accurately identified, and the remote equipment parameter adaptive refresh function can be activated. This allows for automatic adjustment and maintenance of the equipment through remote control, reducing the need for manual intervention and improving the system's automation level.

[0007] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the remote control system structure of an online exhaust gas detection device provided in an embodiment of this application.

[0009] Figure 2 This is a schematic flowchart of a remote control method for an online exhaust gas detection device provided in an embodiment of this application.

[0010] Figure labeling: 10 for exhaust gas monitoring network deployment module, 20 for exhaust gas diffusion characteristics analysis module, 30 for anomaly verification node deployment module, 40 for local consistency verification module, and 50 for fatigue detection equipment positioning module. Detailed Implementation

[0011] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0012] Example 1, as Figure 1 As shown in the figure, this application embodiment provides a remote control system for an online exhaust gas detection device, the system comprising: The system includes: a waste gas monitoring network deployment module 10 for deploying a gridded waste gas monitoring network in the monitoring area; a waste gas diffusion characteristic analysis module 20 for dividing the waste gas monitoring network into K diffusion-related subnets by analyzing the dynamic diffusion characteristics of waste gas in the monitoring area; an anomaly verification node deployment module 30 for deploying K edge anomaly verification nodes locally in the K diffusion-related subnets; a local consistency verification module 40 for collecting waste gas data from P online waste gas detection devices in the waste gas monitoring network, grouping P real-time monitoring data into K groups of waste gas monitoring streams based on the K diffusion-related subnets, uploading them to the K edge anomaly verification nodes, performing local consistency verification based on the K diffusion verification topologies, and obtaining K device verification results; and a fatigue detection device positioning module 50 for uploading the K device verification results to a remote control cloud via an industrial IoT gateway, performing multi-source verification result fusion analysis, locating fatigue detection devices, and initiating remote device parameter adaptive refresh.

[0013] Furthermore, the exhaust gas diffusion characteristic analysis module 20 is used to perform: After constructing a steady-state flow field model based on the geographical and meteorological information of the monitoring area, diffusion test events are introduced, and equipment-level concentration data are fitted to obtain P multivariate concentration time series of the P online exhaust gas detection devices; spatial diffusion correlation analysis is performed based on the P multivariate concentration time series to construct a diffusion correlation weight matrix; based on the dynamic propagation relationship analysis results of the diffusion correlation weight matrix, the exhaust gas monitoring network is divided into K diffusion correlation subnetworks.

[0014] Furthermore, the anomaly verification node deployment module 30 is used to perform: Based on the P multivariate concentration time series, the K diffusion-related subnets are quantified to form device associations, and the K diffusion verification topologies are constructed. The K diffusion-related subnets are connected to the K edge anomaly verification nodes through an industrial Ethernet ring network, wherein the K edge anomaly verification nodes are connected to the remote control cloud through a 5G private network. After the K diffusion verification topologies are deployed to the K edge anomaly verification nodes, the verification data path between the K diffusion-related subnets and the K diffusion verification topologies is constructed.

[0015] Furthermore, the exhaust gas diffusion characteristic analysis module 20 is used to perform: Based on the geographic fence boundary of the monitoring area, regional topographic data and short-term meteorological data are retrieved from the geographic information system and meteorological monitoring platform, respectively, to form the geographic meteorological information. After constructing a three-dimensional gridded spatial model based on the regional topographic data, the short-term meteorological data of the region is introduced based on the principle of computational fluid dynamics to perform flow field simulation calculations, resulting in the steady-state flow field model. After interactively obtaining the multi-condition production operation parameters of the monitoring area, the diffusion test event is constructed by sliding splicing conditions. The diffusion test event is introduced into the steady-state flow field model to simulate the pollutant diffusion trajectory and obtain the theoretical concentration distribution field. Based on the spatial deployment location of the P online exhaust gas detection devices in the monitoring area, P device measuring points are located in the theoretical concentration distribution field to extract the P multivariate concentration time series.

[0016] Furthermore, the exhaust gas diffusion characteristic analysis module 20 is used to perform: Perform combined enumeration based on exhaust gas consistency on the P multivariate concentration time series to obtain Group unit concentration time series; for the Asynchronous fluctuation similarity quantification and trend correlation quantification were performed on the time series of concentrations in group units to obtain... Group dynamic time warping distance and The Pearson correlation coefficient was calculated; based on the spatial topological relationship of the P device measurement points, the following was performed. Group dynamic time warping distance and The weighted fusion of the Pearson correlation coefficients yields One diffusion correlation strength coefficient; after constructing the symmetric sparse matrix framework of the P online exhaust gas detection devices, the following is used: The diffusion correlation strength coefficients are used to fill the matrix to obtain the diffusion correlation weight matrix.

[0017] Furthermore, the exhaust gas diffusion characteristic analysis module 20 is used to perform: The diffusion correlation weight matrix is ​​coarsely divided based on spectral embedding clustering to obtain an initial subnet set. The initial subnet set is then subjected to connectivity verification based on the spatial Delaunay triangulation of the monitoring area. After separating isolated nodes that do not conform to spatial continuity, adjacent subnets are merged based on real-time wind direction data in the short-term meteorological data of the area to output the K diffusion correlation subnets.

[0018] Furthermore, the anomaly verification node deployment module 30 is used to perform: Based on the spatial topology of the P device measurement points, the K diffusion association subnets are structured and mapped into K initial verification topologies; based on the... Group dynamic time warping distance and Pearson correlation coefficient was constructed. Group time series fluctuation correlation layer and Group trend consistency association layer; using the above Group time series fluctuation correlation layer and The trend consistency association layer performs multi-dimensional topology optimization on the K initial verification topologies to obtain the K diffusion verification topologies.

[0019] Through the detailed description of a remote control method for an online exhaust gas detection device, those skilled in the art will clearly understand the remote control system of the online exhaust gas detection device in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section.

[0020] Example 2, based on the same inventive concept as the remote control system of the online exhaust gas detection device in the foregoing examples, such as... Figure 2 As shown in the figure, this application provides a remote control method for an online exhaust gas monitoring device, the method comprising: Deploy a grid-based exhaust gas monitoring network in the monitoring area.

[0021] A monitoring area typically refers to the region near the source of exhaust gas emissions. This area needs to monitor the concentration, composition, flow, and other characteristics of the exhaust gases. Within this area, the level of monitoring precision is determined based on actual needs, ultimately achieving comprehensive air quality monitoring. Multiple monitoring points are set up within the monitoring area to form a grid-like exhaust gas monitoring network. Each monitoring point has an online exhaust gas detection device to collect real-time exhaust gas data. The network structure can be regular, such as a uniform grid layout, or irregular, depending on the region's meteorological conditions, the location of pollution sources, and monitoring requirements.

[0022] By analyzing the dynamic diffusion characteristics of exhaust gas in the monitoring area, the exhaust gas monitoring network is divided into K diffusion-related subnetworks.

[0023] Based on topography and meteorological conditions, flow field simulations are performed using computational fluid dynamics and other methods to predict the diffusion trajectory of exhaust gases. Based on the flow field model, the diffusion path of exhaust gases within the monitoring area is simulated, yielding the exhaust gas concentration distribution. According to the diffusion patterns of the exhaust gases, their diffusion characteristics are further analyzed, identifying different diffusion behaviors. For example, some areas are pollution hotspots, while others are diffusion zones of pollution sources. Based on the diffusion characteristics, the exhaust gas monitoring network within the monitoring area is divided into multiple sub-networks. The sub-network division is based on the diffusion correlation of exhaust gases, which is quantified using a diffusion weight matrix. After division, each sub-network includes a certain number of monitoring points, and the exhaust gas concentration and diffusion characteristics of these monitoring points are highly correlated.

[0024] K edge anomaly verification nodes are deployed locally in the K diffusion association subnets.

[0025] Within each diffusion-related subnet, a local edge anomaly verification node is deployed. Each verification node is responsible for performing a local consistency check on the exhaust gas monitoring data within its subnet to detect any anomalies. Each verification node compares its data with that of neighboring monitoring devices to confirm the presence of anomalies, such as equipment errors. If a device exhibits data deviation, the verification node will report an anomaly and indicate the need for remote repair or adjustment.

[0026] After the P online exhaust gas detection devices in the exhaust gas monitoring network collect exhaust gas data in the monitoring area, they group P real-time monitoring data into K exhaust gas monitoring streams based on the K diffusion association subnets, upload them to the K edge anomaly verification nodes, perform local consistency verification based on the K diffusion verification topologies, and obtain the verification results of the K devices.

[0027] All P online exhaust gas monitoring devices deployed in the monitoring area collect exhaust gas data in real time. This data includes information such as pollutant concentration, temperature, humidity, and flow rate. Each device, according to its diffusion association subnet, divides the collected data into K groups of exhaust gas monitoring streams and uploads them to the edge anomaly verification node for subsequent verification processing. At the edge verification node, the real-time monitoring data undergoes consistency verification based on the diffusion verification topology. This process is performed according to the spatial and diffusion relationships between devices within each diffusion subnet to ensure the rationality and consistency of the data. If the data from a certain device is significantly inconsistent with the data from other devices, it will be marked as an anomaly, requiring further investigation into the reliability of the device or data.

[0028] The K edge anomaly verification nodes upload the verification results of the K devices to the remote control cloud through the industrial IoT gateway, perform multi-source verification result fusion analysis, locate the fatigue detection device, and initiate remote device parameter adaptive refresh.

[0029] Each edge anomaly verification node uploads the device verification results within each subnet to the remote control cloud. Data is transmitted via an industrial IoT gateway, ensuring reliable and real-time data transmission. Upon receiving verification results from multiple edge verification nodes, the remote control cloud performs multi-source data fusion analysis. The purpose of multi-source verification result fusion is to integrate data from different regions and devices, further improving the accuracy of the verification results. During the analysis, information such as the operating status of multiple devices and environmental factors is considered to identify abnormal devices and whether devices are fatigued. After the verification result fusion analysis, the fatigue status of the devices is automatically identified. Fatigued devices refer to those whose performance has deteriorated due to prolonged operation; through fusion analysis, these devices can be accurately located. When device fatigue or malfunction is detected, the remote device parameter adaptive refresh function is automatically activated. This means that device parameters can be automatically adjusted to restore optimal operating conditions without manual intervention. Device parameter adjustments include optimizing the operating status of exhaust gas monitoring sensors, adjusting operating frequencies, or adjusting other monitoring parameters according to environmental changes.

[0030] Furthermore, by analyzing the dynamic diffusion characteristics of exhaust gas in the monitoring area, the exhaust gas monitoring network is divided into K diffusion-related subnetworks. The method includes: After constructing a steady-state flow field model based on the geographical and meteorological information of the monitoring area, diffusion test events are introduced, and equipment-level concentration data are fitted to obtain P multivariate concentration time series of the P online exhaust gas detection devices; spatial diffusion correlation analysis is performed based on the P multivariate concentration time series to construct a diffusion correlation weight matrix; based on the dynamic propagation relationship analysis results of the diffusion correlation weight matrix, the exhaust gas monitoring network is divided into K diffusion correlation subnetworks.

[0031] Based on the geographic and meteorological information of the monitoring area, relevant geographic and meteorological data are acquired. This information includes topography, meteorological conditions such as wind speed, wind direction, temperature, and humidity, as well as the boundaries and climate characteristics of the monitoring area. A steady-state flow field model is constructed using computational fluid dynamics (CFD) models combined with meteorological data. This model simulates the diffusion process of exhaust gas under specific meteorological conditions and can predict the flow trajectory and concentration changes of exhaust gas from its source to various monitoring points. Diffusion test events are introduced to simulate the diffusion behavior of exhaust gas under different meteorological conditions and operating conditions. These test events help to better understand the diffusion characteristics of exhaust gas in different environments. Through device-level concentration data fitting, the data from each online exhaust gas monitoring device is combined with the flow field model to obtain the concentration data of each device at a specific time point. By fitting the sampling data from P online exhaust gas monitoring devices, a multivariate concentration time series for each device is obtained, representing the concentration changes of the device over different time periods.

[0032] Each exhaust gas monitoring device records a multivariate concentration time series, which includes concentration changes at each monitoring point. Based on this time series data, spatial diffusion correlation analysis is performed to identify the correlations in concentration changes between monitoring points. This analytical method helps to understand the diffusion patterns of exhaust gas within the monitoring area. The purpose of spatial diffusion correlation analysis is to analyze the relationships between different exhaust gas monitoring devices based on their locations and data. Some devices located in the same diffusion area may show similar trends in pollutant concentration changes; however, devices in different areas may exhibit significantly different trends in concentration changes due to differences in diffusion characteristics. Through spatial diffusion correlation analysis, a diffusion correlation weight matrix is ​​constructed. This matrix is ​​used to quantify the diffusion correlation between different exhaust gas monitoring devices. The elements of the weight matrix represent the correlation strength between each pair of devices; the larger the value, the stronger the concentration change relationship between the two; the smaller the value, the weaker the correlation.

[0033] Based on the diffusion correlation weight matrix, the propagation relationship of exhaust gas among different monitoring devices is analyzed. By analyzing these propagation relationships, it is possible to identify which devices are in the same diffusion subnetwork and whose exhaust gas concentration change trends and diffusion characteristics are similar. Based on the analysis results of the diffusion correlation weight matrix, the exhaust gas monitoring network is divided into K diffusion correlation subnetworks. The concentration changes among the devices in these subnetworks have a high correlation, reflecting the diffusion pattern of exhaust gas in the monitoring area.

[0034] Furthermore, the method involves deploying K edge anomaly verification nodes locally in the K diffusion-related subnets, and includes: Based on the P multivariate concentration time series, the K diffusion-related subnets are quantified to form device associations, and the K diffusion verification topologies are constructed. The K diffusion-related subnets are connected to the K edge anomaly verification nodes through an industrial Ethernet ring network, wherein the K edge anomaly verification nodes are connected to the remote control cloud through a 5G private network. After the K diffusion verification topologies are deployed to the K edge anomaly verification nodes, the verification data path between the K diffusion-related subnets and the K diffusion verification topologies is constructed.

[0035] Based on the multivariate concentration time series of P online exhaust gas monitoring devices, the first step is to quantify the correlation between these devices. This is done by analyzing the concentration time series data of each device and calculating their similarity or correlation. The multivariate concentration time series represents the concentration data of multiple pollutants changing over time for each device. By comparing these time series data, the strength of the relationship between the devices can be identified. For example, devices located in the same diffusion area have similar concentration change trends within the same time period. The device correlation quantification process includes modeling the spatial relationships between devices, constructing a device correlation metric based on the relative positions of the devices within the monitoring area and their concentration data changes, and constructing a correlation matrix between devices based on the similarity of concentration changes. This matrix can be used to represent the strength of the correlation between the devices. Based on the correlation quantification results, the relationships and topology of these devices within the diffusion subnet are optimized to form a diffusion verification topology. The diffusion verification topology refers to the spatial relationships and data interaction methods of a group of devices within a specific diffusion subnet. This topology provides the structural basis for data flow and information transmission for subsequent verification.

[0036] Within each diffusion-related subnet, an industrial Ethernet ring network connects all devices to the edge anomaly verification node. The advantage of this ring network structure is that it ensures that data can still be transmitted through other paths in the event of device or node failure, providing higher reliability. Each subnet uploads data to its corresponding edge anomaly verification node via the ring network. Each edge anomaly verification node is connected to the remote control cloud via a 5G private network. The 5G private network provides higher bandwidth and lower latency for this remote connection, ensuring that real-time data can be transmitted to the cloud quickly and stably.

[0037] K diffusion verification topologies are deployed to corresponding K edge anomaly verification nodes. Each topology includes data exchange paths and relationship networks for devices within its diffusion subnet, ensuring that each node can perform effective verification tasks. The topology deployment process involves configuring the data flow and verification logic between verification nodes and network devices, ensuring that each topology can perform effective anomaly detection based on the spatial and data relationships of the devices. After topology deployment, verification data paths are established. These paths refer to the data flow paths between the various topology nodes, including real-time data transmission paths from devices to edge nodes and remote data upload paths from edge nodes to the cloud. These paths ensure that data flows according to the topology structure, enabling real-time data verification and analysis, ensuring no data loss or delay, and enhancing network stability and reliability.

[0038] Furthermore, after constructing a steady-state flow field model based on the geographical and meteorological information of the monitoring area, diffusion test events are introduced, and equipment-level concentration data are fitted to obtain P multivariate concentration time series from the P online exhaust gas monitoring devices. The method includes: Based on the geographic fence boundary of the monitoring area, regional topographic data and short-term meteorological data are retrieved from the geographic information system and meteorological monitoring platform, respectively, to form the geographic meteorological information. After constructing a three-dimensional gridded spatial model based on the regional topographic data, the short-term meteorological data of the region is introduced based on the principle of computational fluid dynamics to perform flow field simulation calculations, resulting in the steady-state flow field model. After interactively obtaining the multi-condition production operation parameters of the monitoring area, the diffusion test event is constructed by sliding splicing conditions. The diffusion test event is introduced into the steady-state flow field model to simulate the pollutant diffusion trajectory and obtain the theoretical concentration distribution field. Based on the spatial deployment location of the P online exhaust gas detection devices in the monitoring area, P device measuring points are located in the theoretical concentration distribution field to extract the P multivariate concentration time series.

[0039] A geofence boundary refers to the spatial boundary of a monitoring area, determined through a geographic information system (GIS). Setting the geofence boundary helps define the physical coverage of the exhaust gas monitoring network, ensuring that monitoring data accurately reflects the pollution status within the area. The GIS acquires topographic data of the monitoring area, including changes in ground elevation, landform features, buildings, rivers, and other factors, which influence the dispersion behavior of exhaust gases. Short-term meteorological data, including wind speed, wind direction, humidity, and air pressure, is obtained from a meteorological monitoring platform. These meteorological factors significantly affect the dispersion, dispersion rate, and direction of exhaust gases. Combining the retrieved topographic and meteorological data forms a complete geographic meteorological information system.

[0040] Based on topographic data of the monitoring area, a three-dimensional gridded spatial model is constructed using a geographic information system (GIS). This model includes all geographical factors affecting exhaust gas dispersion, such as topographic features, buildings, roads, rivers, and mountains. Each grid represents a small portion of the area, and finer grid division allows for more accurate simulation results. Using computational fluid dynamics (CFD) principles, exhaust gas flow simulations are performed within this three-dimensional spatial model. The CFD simulation analyzes the flow characteristics of air and exhaust gas under different environments through mathematical models and numerical solutions. By simulating the dynamic behavior of airflow, it predicts how exhaust gas will disperse under different topographic conditions. Short-term meteorological data is introduced into the flow field simulation as boundary conditions to simulate the flow of exhaust gas under different meteorological conditions. This meteorological data exhibits temporal and short-term variation characteristics. After CFD simulation, a steady-state flow field model is obtained. This model demonstrates the flow state and diffusion trend of exhaust gas in the monitoring area. The steady-state flow field model can help predict how exhaust gas will spread from the pollution source to the surrounding area under different meteorological conditions.

[0041] Multi-condition production operation parameters refer to production activities or industrial process parameters within the waste gas monitoring area, such as emissions, factory operating status, and equipment load. These parameters are obtained through an interactive system, reflecting the emission characteristics of waste gas under different environments, such as different production periods or weather conditions. Sliding splicing operation conditions refer to connecting multiple operation conditions in chronological order based on different parameters to create a comprehensive diffusion test event. This sliding splicing method allows the model to simulate the impact of different production states on waste gas diffusion over continuous time periods.

[0042] By incorporating diffusion test events into the previously obtained steady-state flow field model, the model simulates the diffusion process of exhaust gas under different operating conditions, taking into account the specific impact of different production environment conditions on exhaust gas diffusion. Based on the steady-state flow field model and diffusion test events, pollutant diffusion trajectory simulation is performed. This process mainly involves tracing the propagation path of exhaust gas through numerical simulation, simulating the diffusion process from the emission source to the monitoring point. The simulation process incorporates various factors, such as airflow speed and direction, terrain undulations, and temperature differences, to accurately predict the propagation path of exhaust gas. Through the simulation of the diffusion trajectory, a theoretical concentration distribution field is obtained, which represents the concentration distribution of exhaust gas at different locations and time points.

[0043] Based on the theoretical concentration distribution field, and according to the diffusion characteristics of exhaust gas, the spatial layout of online exhaust gas monitoring equipment is determined. The equipment is placed in areas with significant concentration variations or potential pollution risks. Within the theoretical concentration distribution field, P monitoring points are located based on the diffusion characteristics of the exhaust gas. The locations of these points are optimized based on changes in pollutant concentrations to ensure that the monitoring data comprehensively represents the exhaust gas diffusion. At each monitoring point, P multivariate concentration time series are obtained by monitoring the concentrations of different pollutants in the exhaust gas. These time series record the concentration changes of each monitoring device at different time points. The concentration time series recorded by each device includes the concentration changes of multiple pollutants. Through this data, the composition and trends of the exhaust gas can be analyzed.

[0044] Furthermore, based on the P multivariate concentration time series, spatial diffusion correlation analysis is performed to construct a diffusion correlation weight matrix. The method includes: Perform combined enumeration based on exhaust gas consistency on the P multivariate concentration time series to obtain Group unit concentration time series; for the Asynchronous fluctuation similarity quantification and trend correlation quantification were performed on the time series of concentrations in group units to obtain... Group dynamic time warping distance and The Pearson correlation coefficient was calculated; based on the spatial topological relationship of the P device measurement points, the following was performed. Group dynamic time warping distance and The weighted fusion of the Pearson correlation coefficients yields One diffusion correlation strength coefficient; after constructing the symmetric sparse matrix framework of the P online exhaust gas detection devices, the following is used: The diffusion correlation strength coefficients are used to fill the matrix to obtain the diffusion correlation weight matrix.

[0045] Each online exhaust gas monitoring device records the concentration data of multiple pollutants. These data form a multivariate concentration time series. Exhaust gas consistency refers to the similarity of pollutant concentrations among devices within the same area, especially temporal consistency. This is because the concentration changes of exhaust gas at different locations and times exhibit a certain degree of spatial consistency. Through combination enumeration, the multivariate concentration time series of all devices are combined to generate concentration time series combinations among different devices. For example, if P devices have different concentration time series, then by combining these series, we can obtain... The purpose of grouping unit concentration time series is to find consistency and potential correlations between devices. The combined data provides rich information for subsequent analysis and helps to identify the correlation between different devices.

[0046] In the concentration time series of exhaust gas monitoring equipment, the concentration changes between devices may not be completely synchronized. Even under the same meteorological conditions, the concentration fluctuations of devices may differ due to location or other factors. Asynchronous fluctuation similarity quantification aims to measure whether the fluctuation patterns between these device concentration time series are similar, that is, whether they have similar fluctuation trends despite their different time points. Dynamic time warping (RTW) is an algorithm that measures the similarity of time series, matching the similarity between different series even if their time axes are different.

[0047] In addition to fluctuation similarity, the trend correlation between concentration time series is also analyzed. Trend correlation refers to whether the long-term trends of concentration changes between devices are consistent. For example, when the concentration of one device increases, does the concentration of another device also show a similar upward trend? The Pearson correlation coefficient is used to measure the linear correlation between two time series. A Pearson correlation coefficient value close to +1 indicates a high positive correlation, close to -1 indicates a negative correlation, and close to 0 indicates no correlation.

[0048] The spatial topology of monitoring equipment refers to the relative positions of monitoring equipment in geographic space and their connection methods. In exhaust gas diffusion monitoring, the spatial location of equipment directly affects the correlation between them. For example, adjacent equipment is affected by similar airflow and meteorological conditions, thus their concentration time series have a higher correlation, while equipment located relatively far apart has less correlation in concentration changes. The calculation results of dynamic time-warped distance and Pearson correlation coefficient are weighted and fused. Weighted fusion adjusts the weight of the similarity measure between each pair of equipment based on the spatial topology. The weighted fusion process means that for equipment with closer spatial relationships, such as those that are closer together, the correlation measures (dynamic time-warped distance and Pearson correlation coefficient) between them are given higher weights. This allows the correlation between adjacent equipment to be more fully reflected in the final analysis. Through weighted fusion, the diffusion correlation strength coefficient between each pair of equipment is finally obtained. These coefficients quantify the intensity of the exhaust gas diffusion correlation between equipment; higher values ​​indicate a stronger correlation in concentration changes between the equipment, suggesting they may be located in the same or similar diffusion areas.

[0049] In a symmetric sparse matrix, the elements represent the correlation strength between devices. In an exhaust gas monitoring system, to describe the diffusion correlation strength between each device—that is, whether their concentration changes are similar—the element values ​​of the matrix are determined by the previously calculated diffusion correlation strength coefficients. The matrix is ​​symmetric, meaning that the (i,j) element is the same as the (j,i) element. This is because the relationship between devices is bidirectional; the influence of device i on device j, or vice versa, should be equal. Since not all devices have strong diffusion correlations, most elements in the matrix will be zero, with only a few non-zero values. This structure is called a sparse matrix. Sparse matrices not only reduce storage space but also improve computational efficiency. The diffusion correlation strength coefficients are filled into the matrix to obtain the diffusion correlation weight matrix, which describes the diffusion correlation between all devices.

[0050] Furthermore, based on the dynamic propagation relationship analysis results of the diffusion correlation weight matrix, the exhaust gas monitoring network is divided into K diffusion correlation subnetworks. The method includes: The diffusion correlation weight matrix is ​​coarsely divided based on spectral embedding clustering to obtain an initial subnet set. The initial subnet set is then subjected to connectivity verification based on the spatial Delaunay triangulation of the monitoring area. After separating isolated nodes that do not conform to spatial continuity, adjacent subnets are merged based on real-time wind direction data in the short-term meteorological data of the area to output the K diffusion correlation subnets.

[0051] Spectrograph embedding (SPE) is a technique for dimensionality reduction and clustering of data using graph theory. It identifies latent clustering structures in a graph by performing eigenvalue decomposition on the Laplacian matrix. SPE transforms the device relationships in the diffusion association weight matrix into a graph, where nodes represent devices and edges represent the diffusion association strength between devices. Based on the diffusion association weight matrix, SPE assigns devices to different clusters (subnets), initially defining the relationships between devices. This coarse partitioning involves approximate grouping to obtain an initial set of subnets.

[0052] Delaunay triangulation is a method of connecting a set of points in space to form triangles, ensuring that no point lies inside the circumcircle of any triangle. In exhaust gas monitoring systems, Delaunay triangulation is used to represent the spatial location and connectivity of monitoring equipment within a monitoring area. Delaunay triangulation is used to perform spatial connectivity checks on the initial set of subnets. By examining the spatial location and connectivity of the equipment, it verifies whether the equipment is located in the same diffusion area and whether a connected region has been formed. If some equipment is assigned to an isolated subnet due to geographical location or other factors, it is considered an isolated node. During connectivity verification, it identifies which devices lack actual connections or data associations. These isolated nodes, due to geographical distance or different diffusion characteristics, are incorrectly assigned to the same subnet and are separated from the current subnet.

[0053] Meteorological data from the monitoring area, especially real-time wind direction data, can help further optimize subnetting. Under the influence of wind direction, the diffusion paths between certain devices will change. If adjacent devices can influence each other under meteorological conditions, they should be merged into a single subnet. Merging adjacent subnets based on real-time wind direction data helps to more accurately reflect the diffusion behavior of exhaust gases. After completing the separation of isolated nodes and the merging of adjacent subnets, the final K diffusion-related subnets represent the device group within the exhaust gas monitoring area, with strong diffusion correlations between devices within each subnet.

[0054] Furthermore, based on the P multivariate concentration time series, device association quantization is performed on the K diffusion-related subnets to construct the K diffusion verification topologies. The method includes: Based on the spatial topology of the P device measurement points, the K diffusion association subnets are structured and mapped into K initial verification topologies; based on the... Group dynamic time warping distance and Pearson correlation coefficient was constructed. Group time series fluctuation correlation layer and Group trend consistency association layer; using the above Group time series fluctuation correlation layer and The trend consistency association layer performs multi-dimensional topology optimization on the K initial verification topologies to obtain the K diffusion verification topologies.

[0055] Spatial topology refers to the relative positions of exhaust gas monitoring devices and their distribution within the monitoring area. For example, are the devices adjacent to each other, located in the same diffusion area, or affected by similar meteorological conditions? Structured mapping refers to the rational organization and connection of devices in each diffusion-related subnet based on the spatial topology of the device measuring points. The goal here is to ensure that the devices in each diffusion-related subnet are combined according to certain spatial relationships to facilitate subsequent verification and anomaly detection. K diffusion-related subnets are mapped to K initial verification topologies, and these topologies reflect the physical relationships and data interaction paths between the devices.

[0056] The temporal fluctuation correlation layer is built based on dynamic time warping distance, representing the fluctuation patterns of concentration time series between devices. This layer allows analysis of whether the fluctuation patterns of devices are synchronized and detects short-term fluctuation relationships between devices. Each pair of devices' temporal fluctuation correlation layers measures the similarity of their concentration data fluctuations, i.e., whether they exhibit similar fluctuation patterns on the time axis. The trend consistency correlation layer is built based on Pearson correlation coefficient, representing the consistency of concentration change trends between devices. This layer is used to analyze whether the long-term trends between devices are similar, such as whether concentrations increase or decrease synchronously.

[0057] The initial verification topology was optimized based on a time-series fluctuation correlation layer and a trend consistency correlation layer. That is, the verification topology not only considers the spatial relationships between devices but also combines the fluctuation similarity and trend consistency between devices to optimize their connections. When optimizing the topology using the time-series fluctuation correlation layer, devices with strong similarity in concentration fluctuations are grouped together to form a denser subnet. The trend consistency correlation layer identifies devices with similar trends within the topology, ensuring long-term trend consistency among devices. Through optimization, the connection methods of devices are adjusted according to their similarity, making the verification topology of each device more consistent with actual exhaust gas diffusion and monitoring needs. After multi-dimensional topology optimization, K diffusion verification topologies are finally obtained, which are used for subsequent exhaust gas monitoring data verification and equipment anomaly detection.

[0058] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A remote control system for an exhaust gas on-line detection apparatus, characterized by, The system includes: The exhaust gas monitoring network deployment module is used to deploy a gridded exhaust gas monitoring network in the monitoring area; The exhaust gas diffusion characteristic analysis module is used to divide the exhaust gas monitoring network into K diffusion-related subnetworks by performing dynamic diffusion characteristic analysis of the monitored area. An anomaly verification node deployment module is used to deploy K edge anomaly verification nodes locally in the K diffusion-related subnets; The local consistency verification module is used to collect exhaust gas data in the monitoring area by P online exhaust gas detection devices in the exhaust gas monitoring network, group P real-time monitoring data into K groups of exhaust gas monitoring streams according to the K diffusion association subnets, upload them to the K edge anomaly verification nodes, perform local consistency verification based on the K diffusion verification topologies, and obtain the verification results of K devices. The fatigue detection equipment positioning module is used by the K edge anomaly verification nodes to upload the verification results of the K devices to the remote control cloud through the industrial IoT gateway, perform multi-source verification result fusion analysis, locate the fatigue detection equipment, and initiate remote device parameter adaptive refresh.

2. The remote control system of an on-line exhaust gas detection apparatus according to claim 1, wherein The exhaust gas diffusion characteristic analysis module is used to perform: After constructing a steady-state flow field model based on the geographical and meteorological information of the monitoring area, diffusion test events are introduced, and equipment-level concentration data are fitted to obtain P multivariate concentration time series of the P online exhaust gas detection devices. Spatial diffusion correlation analysis was performed based on the P multivariate concentration time series to construct a diffusion correlation weight matrix; Based on the dynamic propagation relationship analysis results of the diffusion correlation weight matrix, the exhaust gas monitoring network is divided into K diffusion correlation subnetworks.

3. The remote control system for an exhaust gas on-line inspection apparatus according to claim 2, wherein The anomaly verification node deployment module is used to execute: Based on the P multivariate concentration time series, device association quantification is performed on the K diffusion-related subnets to construct the K diffusion verification topologies; The K diffused associated subnets are connected to the K edge anomaly verification nodes via an industrial Ethernet ring network, wherein the K edge anomaly verification nodes are connected to the remote control cloud via a 5G private network; After deploying the K diffusion verification topologies to the K edge anomaly verification nodes, the verification data paths of the K diffusion association subnets and the K diffusion verification topologies are constructed.

4. The remote control system for an online exhaust gas detection device as described in claim 2, characterized in that, The exhaust gas diffusion characteristic analysis module is used to perform: Based on the geographic fence boundary of the monitoring area, regional topographic data and regional short-term meteorological data are retrieved from the geographic information system and meteorological monitoring platform, respectively, to form the geographic meteorological information. After constructing a three-dimensional gridded spatial model based on the regional topographic data, flow field simulation calculations are performed using short-term meteorological data of the region based on computational fluid dynamics principles to obtain the steady-state flow field model. After interactively obtaining the multi-condition production operation parameters of the monitoring area, the diffusion test event is constructed by sliding splicing the conditions; By introducing the diffusion test event into the steady-state flow field model, the pollutant diffusion trajectory is simulated to obtain the theoretical concentration distribution field. Based on the spatial layout of the P online exhaust gas detection devices in the monitoring area, P device measuring points are located in the theoretical concentration distribution field to extract the P multivariate concentration time series.

5. The remote control system for an online exhaust gas detection device as described in claim 4, characterized in that, The exhaust gas diffusion characteristic analysis module is used to perform: Perform combined enumeration based on exhaust gas consistency on the P multivariate concentration time series to obtain Group unit concentration time series; Regarding the Asynchronous fluctuation similarity quantification and trend correlation quantification were performed on the time series of concentrations in group units to obtain... Group dynamic time warping distance and Group Pearson correlation coefficient; Based on the spatial topology relationship of the P device measurement points, the following is performed: Group dynamic time warping distance and The weighted fusion of the Pearson correlation coefficients yields Each diffusion correlation strength coefficient; After constructing the symmetric sparse matrix framework of the P online exhaust gas detection devices, the following is adopted: The diffusion correlation strength coefficients are used to fill the matrix to obtain the diffusion correlation weight matrix.

6. The remote control system for an online exhaust gas detection device as described in claim 5, characterized in that, The exhaust gas diffusion characteristic analysis module is used to perform: The diffusion correlation weight matrix is ​​coarsely divided based on spectral embedding clustering to obtain an initial set of subnets; Based on the spatial Delaunay triangulation of the monitoring area, the initial subnet set is subjected to connectivity verification. After separating isolated nodes that do not conform to spatial continuity, adjacent subnets are merged according to the real-time wind direction data in the short-term meteorological data of the area, and the K diffusion-related subnets are output.

7. The remote control system for an online exhaust gas detection device as described in claim 6, characterized in that, The anomaly verification node deployment module is used to execute: Based on the spatial topology relationship of the P device measurement points, the K diffusion association subnets are structured and mapped into K initial verification topologies; According to the above Group dynamic time warping distance and Pearson correlation coefficient was constructed. Group time series fluctuation correlation layer and Group trend consistency association layer; Using the above Group time series fluctuation correlation layer and The trend consistency association layer performs multi-dimensional topology optimization on the K initial verification topologies to obtain the K diffusion verification topologies.

8. A remote control method for an online exhaust gas detection device, characterized in that, The method, implemented based on the remote control system of the online exhaust gas detection device according to any one of claims 1-7, includes: Deploy a grid-based exhaust gas monitoring network in the monitoring area; By analyzing the dynamic diffusion characteristics of exhaust gas in the monitoring area, the exhaust gas monitoring network is divided into K diffusion-related subnetworks. K edge anomaly verification nodes are deployed locally in the K diffusion-related subnets; After the P online exhaust gas detection devices in the exhaust gas monitoring network collect exhaust gas data in the monitoring area, they group P real-time monitoring data into K exhaust gas monitoring streams based on the K diffusion association subnets, upload them to the K edge anomaly verification nodes, perform local consistency verification based on the K diffusion verification topologies, and obtain the verification results of the K devices. The K edge anomaly verification nodes upload the verification results of the K devices to the remote control cloud through the industrial IoT gateway, perform multi-source verification result fusion analysis, locate the fatigue detection device, and initiate remote device parameter adaptive refresh.

Images