Multi-source sensor fused street canyon particulate matter distribution monitoring system

Abstract

The invention discloses a multi-source sensor fused street canyon particulate matter distribution monitoring system, which relates to the technical field of environment monitoring and comprises a particulate matter distribution monitoring platform. The particulate matter distribution monitoring platform is in communication connection with a sensor network module, a multi-source data fusion module, an air quality model module, a trend prediction module and an air quality early warning module, and all the modules are in electric signal connection; and the sensor network module is used for deploying a plurality of sensor nodes according to the spatial structure characteristics of the street canyon. By deploying a plurality of sensor nodes and comprehensively covering subdivided areas such as streets and roadways according to the spatial structure characteristics of the street canyons, the problem that traditional fixed monitoring stations are difficult to comprehensively monitor is effectively solved, particulate matter concentration data at different positions are collected, data quality is optimized through a data fusion algorithm, noise and errors are reduced, and the monitoring accuracy is improved. Therefore, the coverage range and precision of monitoring are obviously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and in particular to a street valley particulate matter distribution monitoring system using multi-source sensor fusion. Background Art

[0002] With the acceleration of urbanization and the continuous increase in industrial and transportation activities, particulate matter (such as PM2.5 and PM10) has become one of the main pollutants in the urban atmospheric environment. These particulate matter not only affects atmospheric visibility but also enters the human body through the respiratory system, causing respiratory and cardiovascular diseases, posing a serious threat to public health. Therefore, accurately monitoring the distribution of particulate matter is crucial for assessing urban air quality and formulating pollution prevention and control measures. Street canyons are narrow spaces formed by tall buildings on both sides and the ground. Their unique geometry restricts air circulation. Within street canyons, airflow is complex, easily forming vortices and stagnant areas, making it difficult for particulate matter to diffuse and dilute, resulting in high concentrations of pollution in localized areas. As areas of dense human and vehicle activity in cities, the particulate matter pollution in street canyons directly reflects one aspect of urban air quality.

[0003] In existing technology, street canyons have irregular spatial structures, with buildings of varying heights and densities on both sides, resulting in complex airflow. Traditional monitoring stations are mostly fixed points, making it difficult to comprehensively monitor subdivided areas such as streets and alleys, and to reflect changes in particulate matter in localized areas or over short periods of time. Therefore, the present invention aims to address the problem of deploying multiple sensor nodes to expand monitoring coverage, obtain particle concentration distributions in different areas, create accurate air quality maps, and ensure the accuracy of air quality warnings. To this end, a street canyon particulate matter distribution monitoring system using multi-source sensor fusion is proposed. Summary of the Invention

[0004] The present invention aims to provide a street valley particulate matter distribution monitoring system with multi-source sensor fusion to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A multi-source sensor fusion street valley particulate matter distribution monitoring system includes a particulate matter distribution monitoring platform, characterized in that: the particulate matter distribution monitoring platform is communicatively connected to a sensor network module, a multi-source data fusion module, an air quality model module, a trend prediction module, and an air quality warning module, wherein the modules are electrically connected;

[0007] The sensor network module is used to deploy multiple sensor nodes based on the spatial structure characteristics of the street valley and collect monitored particulate matter concentration data from each sensor node;

[0008] The multi-source data fusion module is used to fuse the particle concentration data from different locations using a data fusion algorithm;

[0009] The air quality model module is used to combine the fused particulate matter concentration data to construct a street valley particulate matter concentration distribution model and generate an accurate air quality map;

[0010] The trend prediction module is used to analyze the changing trend of street valley particulate matter concentration by using a time series analysis algorithm and combining historical and real-time monitoring particulate matter concentration data, so as to grasp the development trend of street valley particulate matter pollution in advance;

[0011] The air quality warning module is used to set corresponding warning thresholds based on the street valley particulate matter concentration distribution model and change trend analysis results, combined with air quality standards, to determine whether to issue air quality warning information.

[0012] A further improvement of the technical solution of the present invention is that: the sensor network module includes a node deployment unit and a data acquisition unit;

[0013] The node deployment unit is used to plan and deploy sensor nodes and install multiple particulate matter sensors based on the spatial structural characteristics of the street valley, including building height, density distribution, and street orientation;

[0014] The data collection unit is used to collect the monitored particulate matter concentration data from each sensor node, and perform pre-processing to match it with time and geographical information.

[0015] A further improvement of the technical solution of the present invention is that: the node deployment unit specifically includes:

[0016] Obtain building height and density distribution data through the urban geographic information system, use urban planning drawings to clarify street directions, widths, and surrounding topography, and integrate them into a detailed spatial structure database;

[0017] Based on the analysis of the street canyon's spatial structure, the layout of sensor nodes was planned. The installation locations and number of sensors were determined based on the characteristics of the street and laneway subdivisions. The spacing of sensors was also arranged based on the diffusion characteristics of particulate matter to ensure that the monitoring data effectively reflects the spatial distribution of particulate matter concentrations within the street canyon. Sensor nodes were deployed on both sides of the street, at laneway entrances, in areas with high traffic volume, and at locations with significant changes in building height.

[0018] Install the particulate matter sensors according to the planned sensor node layout plan, and debug the sensors after installation to ensure that they can work properly.

[0019] A further improvement of the technical solution of the present invention is that: the data acquisition unit specifically includes:

[0020] Establishing a communication connection with each sensor node to receive real-time particle concentration data collected by the particle sensors of the sensor nodes, where the particle concentration data includes but is not limited to the concentration values of particulate matter such as PM2.5 and PM10;

[0021] A verification mechanism is used to perform preliminary verification on the received particulate matter concentration data, and then preprocessing operations are performed on the data that pass the preliminary verification, including data cleaning, normalization and code conversion steps;

[0022] Match the pre-processed particulate matter concentration data with time information, obtain the current time from the system clock, and add a timestamp to each data record. Through time matching, we can understand the change of particulate matter concentration over time;

[0023] According to the deployment location information of the sensor nodes, the particulate matter concentration data is matched with the corresponding geographic information, and the data records are associated with the corresponding geographic location coordinates to form a data set with spatial attributes.

[0024] A further improvement of the technical solution of the present invention is that the multi-source data fusion module specifically includes:

[0025] Receive particulate matter concentration data from different locations from the sensor network module and perform basic rechecks on the particulate matter concentration data, including data format, data volume, and data timestamp integrity;

[0026] After the particle concentration data passed the basic review, the Kalman filter method was selected as the data fusion algorithm for processing. Based on the laws and characteristics of the street canyon particle concentration changes, a state equation and an observation equation were established to reflect the sensor's observation process of the particle concentration. Through these two equations, the Kalman filter algorithm was used to dynamically estimate and update the particle concentration data. Continuous iterative calculations gradually approached the actual particle concentration value, realizing the fusion processing of multi-source data and optimizing data quality.

[0027] Verify and optimize the fused data results, collect known reference data for comparison with the fused data, evaluate the accuracy and reliability of the fused data, and adjust and optimize the fusion algorithm based on the verification results to further improve data quality.

[0028] A further improvement of the technical solution of the present invention is that the air quality model module specifically includes:

[0029] Obtain the fused particulate matter concentration data output by the multi-source data fusion module, and simultaneously collect geographic information data related to the street canyon, including building distribution, street orientation, and topography. The fused particulate matter concentration data and geographic information data are integrated to obtain the basic input data set.

[0030] Based on the characteristics of street canyons, a Gaussian diffusion model was selected to construct a street canyon particulate matter concentration distribution model. Based on the theoretical formula of the Gaussian diffusion model and the fused particulate matter concentration data and geographic information data, model parameters were set, including particulate matter emission source intensity, diffusion coefficient, deposition rate, wind speed and direction, to ensure that the model can accurately reflect the actual environmental conditions in the street canyon.

[0031] The fused particle concentration data and geographic information data are input into the constructed street canyon particle concentration distribution model for calculation. The model simulates the diffusion, migration, and transformation of particles within the street canyon based on the input initial particle concentration value, geographic information, and set model parameters. Through multiple iterative calculations, the model's internal variables are adjusted to gradually make the simulation results closer to the actual situation. After the calculation is completed, the particle concentration distribution data at each location in the street canyon is analyzed.

[0032] Using geographic information system (GIS) tools, the simulation results of particulate matter concentration distribution data at various locations in the street canyon were visualized to generate an accurate air quality map. The map shows the concentration distribution of particulate matter in different areas of the street canyon with different colors.

[0033] A further improvement of the technical solution of the present invention is that the analysis process of the particle concentration distribution data at each location in the street canyon is as follows:

[0034] The fused particulate matter concentration data and geographic information data were input into the street canyon particulate matter concentration distribution model to simulate the concentration distribution of particulate matter in the street canyon. The initial conditions of the model were set, including the initial value of the particulate matter concentration, geographic information, and model parameters. The model parameters included the emission source intensity, diffusion coefficient, deposition rate, wind speed, and wind direction of particulate matter.

[0035] The street canyon particle concentration distribution model simulates the diffusion, migration, and transformation of particles within the street canyon based on the input initial particle concentration value, geographic information, and model parameters. It also determines the basic formula for the Gaussian diffusion model. In the street canyon environment, the Gaussian diffusion model is modified based on the effect of particle settling velocity on concentration distribution, and the formula for the modified Gaussian diffusion model is then determined.

[0036] The street valley particulate matter concentration distribution model was used to perform multiple iterative calculations to gradually bring the simulation results closer to the actual situation. In each iteration, the model parameters and initial concentration distribution were initialized based on the current concentration distribution and geographic information. The concentration distribution at each location was calculated using the formula of the modified Gaussian diffusion model. The model parameters, including the diffusion coefficient, sedimentation rate, wind speed, and wind direction, were adjusted based on the geographic information. The concentration distribution calculation process at each location and the model parameter adjustment process were repeated until the error between the simulation results and the actual monitoring data was within an acceptable range.

[0037] After multiple iterative calculations, the street valley particle concentration distribution model outputs particle concentration distribution data for each location within the street valley, which is used to generate an accurate air quality map that intuitively displays the particle concentration distribution in different areas of the street valley.

[0038] A further improvement of the technical solution of the present invention is that the trend prediction module specifically includes:

[0039] Acquire historical monitored particulate matter concentration data from the database, covering the Street Valley's past particulate matter concentration information at different time periods, and record how particulate matter concentration changes over time. Simultaneously, receive real-time particulate matter concentration data collected and fused by the current sensor network, integrate historical data with real-time data, unify data formats and time accuracy, and ensure data continuity and comparability across the timeline.

[0040] Based on historical monitored particulate matter concentration data, trend characteristics related to particulate matter concentration changes are extracted, including the daily mean concentration change rate, the weekly mean concentration range, and the concentration seasonality index;

[0041] Based on the characteristics of street valley particulate matter concentration data, a trend prediction model was constructed using a time series analysis algorithm based on a long short-term memory network. The model took as input the trend characteristics of the daily mean concentration change rate, the weekly mean concentration range, and the concentration seasonality index, and output the future trend of street valley particulate matter concentration.

[0042] The real-time fused particle concentration data is input into the trend prediction model to obtain the predicted value of the particle concentration at a future time point. The future trend of the particle concentration in the street valley is then analyzed and the prediction results are output in the form of a chart, showing the change trend of the particle concentration over time.

[0043] A further improvement of the technical solution of the present invention is that the calculation process of the predicted value of the particle concentration at the future time point is:

[0044] Obtain the particle concentration value at the current time point from the real-time fused particle concentration data as the basic value for prediction;

[0045] Calculate the ratio of the daily average concentration change rate to the particle concentration value at the current time, add 1 to calculate the exponential growth or decay of the particle concentration within t days, analyze the exponential growth or decay portion of the particle concentration value at the current time, and then multiply it by the particle concentration value at the current time to obtain the adjusted value of the current concentration;

[0046] Calculate the ratio of the weekly average concentration range to the particle concentration value at the current time point, and multiply it by the sine function to obtain the intra-week periodic fluctuation function;

[0047] Calculate the ratio of the concentration seasonality index to the particle concentration value at the current time point, and calculate the exponential function based on the ratio of the predicted time step to the seasonal period. Multiply the two to obtain the seasonal variation function.

[0048] The current concentration adjustment value, the weekly periodic fluctuation function, and the seasonal variation function are added together to obtain the predicted value of the particulate matter concentration at a future time point.

[0049] A further improvement of the technical solution of the present invention is that the air quality warning module specifically includes:

[0050] Collect the particle concentration distribution data for each area in the street canyon, as output by the street canyon particle concentration distribution model, and the analysis results of future particle concentration trends obtained by the trend prediction module. At the same time, obtain national and local air quality standards to clarify the particle concentration threshold ranges corresponding to different air quality standards.

[0051] Based on the integrated street valley particulate matter concentration data, trend prediction results, and air quality standards, corresponding particulate matter concentration warning thresholds are set;

[0052] Compare real-time monitoring data and forecast results with the set particulate matter concentration warning threshold to determine whether it is necessary to issue air quality warning information. Once the warning is triggered, the air quality warning information will be released through multiple channels (SMS, APP push, social media, etc.).

[0053] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art:

[0054] 1. This invention provides a street valley particulate matter distribution monitoring system that uses multi-source sensor fusion. By deploying multiple sensor nodes, this system comprehensively covers subdivided areas such as streets and alleys based on the spatial structure of street valleys. This effectively solves the problem of traditional fixed monitoring stations being unable to provide comprehensive monitoring, ensuring that particulate matter concentration data at different locations can be collected. A data fusion algorithm optimizes data quality, reduces noise and errors, and significantly improves monitoring coverage and accuracy.

[0055] 2. This invention provides a street valley particulate matter distribution monitoring system that uses multi-source sensor fusion. By receiving and processing particulate matter concentration data from various sensor nodes in real time, combined with a multi-source data fusion module and a trend prediction module, it can dynamically analyze the changing trends of street valley particulate matter concentrations. Furthermore, by setting corresponding warning thresholds, it can issue air quality warning information, helping environmental management departments to quickly respond to air quality changes and take effective measures to reduce the risk of public exposure to high-concentration particulate matter environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0057] Figure 1 Schematic diagram of the analysis process of the changing trend of the street valley particulate matter concentration in the present invention;

[0058] Figure 2 This is a schematic diagram of the functional modules of the system of the present invention. DETAILED DESCRIPTION

[0059] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0060] Example 1, as Figure 1 、 Figure 2 As shown, the present invention provides a street valley particulate matter distribution monitoring system with multi-source sensor fusion, including a particulate matter distribution monitoring platform, which is communicatively connected to a sensor network module, a multi-source data fusion module, an air quality model module, a trend prediction module, and an air quality warning module, wherein the modules are electrically connected;

[0061] The sensor network module is used to deploy multiple sensor nodes based on the spatial structure characteristics of the street valley and collect monitored particulate matter concentration data from each sensor node. The sensor network module includes a node deployment unit and a data acquisition unit;

[0062] The node deployment unit is used to plan and deploy sensor nodes based on the spatial structure characteristics of the street canyon, including building height, density distribution, and street orientation. Multiple particulate matter sensors are installed to ensure coverage of subdivided areas such as streets and alleys, comprehensively monitoring particulate matter concentrations at different locations. Building height and density distribution data are obtained through the urban geographic information system. Street orientation, width, and surrounding topography are determined using urban planning drawings and integrated into a detailed spatial structure database. Based on the results of the street canyon spatial structure analysis, a sensor node layout plan is planned. The installation locations and number of sensors are determined based on the characteristics of the street and alley subdivisions. Sensor spacing is arranged based on the diffusion characteristics of particulate matter to ensure that the monitoring data effectively reflects the spatial distribution of particulate matter concentrations within the street canyon. Sensor nodes are deployed on both sides of streets, at alley entrances, in areas with high traffic volume, and at locations with significant changes in building height to ensure comprehensive coverage of different areas within the street canyon. Particulate matter sensors are installed according to the planned sensor node layout plan. During installation, the sensor installation locations are ensured to meet design requirements, while also considering the sensor's stability and safety. After installation, the sensors are debugged to ensure proper operation.

[0063] The data acquisition unit is used to collect the monitored particulate matter concentration data from each sensor node, and pre-process it to match it with time and geographical information, establish a communication connection with each sensor node, and receive the particulate matter concentration data collected by the particulate matter sensor of the sensor node in real time. The particulate matter concentration data includes but is not limited to the concentration values of particulate matter such as PM2.5 and PM10. A verification mechanism is used to perform a preliminary verification on the received particulate matter concentration data to check whether the data packet is complete and whether there is a transmission error. If the data is abnormal, a retransmission request is sent to the sensor node to ensure that complete and accurate data is obtained. The data that passes the preliminary verification is then pre-processed, including the steps of data cleaning, normalization and encoding conversion. Noise and abnormal values in the data are removed through data cleaning, and a filtering algorithm is used. The method is used to smooth the data to reduce the impact of random errors, normalize the data to unify data of different dimensions into the same range, encode the data to ensure the consistency and compatibility of the data format, match the preprocessed particulate matter concentration data with time information, obtain the current time from the system clock, and add a timestamp to each data record. Through time matching, it is possible to understand the change of particulate matter concentration over time. According to the deployment location information of the sensor node, the particulate matter concentration data is matched with the corresponding geographic information, and the data records are associated with the corresponding geographic location coordinates to form a data set with spatial attributes. Through geographic information matching, it is possible to intuitively understand the distribution of particulate matter concentration in different geographic locations, providing basic data for building air quality maps and conducting spatial analysis;

[0064] The multi-source data fusion module is used to adopt a data fusion algorithm to fuse the particle concentration data from different locations, optimize data quality, reduce noise and errors, and eliminate the limitations of a single sensor through the fusion of multi-source data, thereby improving the accuracy, stability and consistency of the monitoring data and making the prediction of the particle distribution more reliable. It receives the particle concentration data from different locations from the sensor network module and performs a basic review of the particle concentration data, including the data format, data volume and timestamp integrity of the data. It checks whether the data format conforms to the preset unified format standard, such as the order of data fields, the definition of data type, etc. It verifies the data volume to ensure that the data volume of each data packet is complete without truncation or redundant data. It checks the timestamp integrity of the data to confirm whether the timestamp exists, is in the correct format and is continuous. Through a comprehensive review, it is checked whether the data has obvious format errors or is missing. The Kalman filter method is selected as the data fusion algorithm for processing after the particle concentration data passes the basic review. According to the law and characteristics of the change of the particle concentration in the street valley, a state equation is established. This equation is used to describe the change state of the particle concentration over time, which contains elements such as state variables and state transfer matrix. At the same time, an observation equation is established to reflect the sensor's observation process of the particle concentration. Through the two equations, the Kalman filter algorithm is used to dynamically estimate and update the particle concentration data. The calculation is continuously iterated to gradually approach the actual particle concentration value, realizing the fusion processing of multi-source data, optimizing data quality, verifying and optimizing the fused data results, collecting known reference data for comparison with the fused data, evaluating the accuracy and reliability of the fused data, and adjusting and optimizing the fusion algorithm according to the verification results to further improve the data quality.

[0065] The air quality model module is used to combine the fused particulate matter concentration data to construct a street valley particulate matter concentration distribution model, generate an accurate air quality map, and intuitively display the concentration distribution of particulate matter in different areas of the street valley, so that environmental management personnel can quickly understand the spatial differences in street valley air quality. The fused particulate matter concentration data output by the multi-source data fusion module is obtained, and at the same time, geographic information data related to the street valley is collected, including building distribution, street direction, and topography. The fused particulate matter concentration data and geographic information data are integrated to obtain the basic input data set. According to the characteristics of the street valley, the Gaussian diffusion model is selected to construct the street valley particulate matter concentration distribution model. Based on the theoretical formula of the Gaussian diffusion model, the fused particulate matter concentration data and geographic information data are combined to set the model parameters, including the emission source intensity, diffusion, and so on. The model uses the particle concentration coefficient, deposition rate, wind speed, and wind direction to ensure that the model can accurately reflect the actual environmental conditions in the street canyon. The fused particle concentration data and geographic information data are input into the constructed street canyon particle concentration distribution model for calculation. The model simulates the diffusion, migration, and transformation of particles in the street canyon based on the input initial particle concentration value, geographic information, and set model parameters. Through multiple iterative calculations, the internal variables of the model are adjusted to gradually make the simulation results closer to the actual situation. After the calculation is completed, the particle concentration distribution data of each location in the street canyon is analyzed and visualized using geographic information system (GIS) tools to generate an accurate air quality map. The map shows the particle concentration distribution distribution of different areas in the street canyon with different colors.

[0066] In addition, the analysis process of the particle concentration distribution data at various locations in the street canyon is as follows:

[0067] The fused particulate matter concentration data and geographic information data are input into the street valley particulate matter concentration distribution model, where the particulate matter concentration data includes the concentration values of PM2.5, PM10, etc. monitored by each sensor node, and the geographic information data includes the distribution of buildings, street directions, topography, etc. in the street valley. The concentration distribution of particulate matter in the street valley is simulated, and the initial conditions of the model are set, including the initial value of the particulate matter concentration, geographic information, and model parameters. The model parameters include the emission source intensity, diffusion coefficient, sedimentation rate, wind speed, and wind direction of the particulate matter. The emission source intensity of the particulate matter represents the emission amount of the particulate matter, the diffusion coefficient represents the diffusion ability in the x, y, and z directions, the sedimentation rate represents the speed at which the particulate matter sinks due to factors such as gravity, the wind speed represents the speed of the wind in the x direction, and the wind direction represents the direction of the wind. The street valley particulate matter concentration distribution model simulates the diffusion, migration, and transformation process of particulate matter in the street valley based on the input initial value of the particulate matter concentration, geographic information, and model parameters, and determines The basic formula of the Gaussian diffusion model is used to modify the Gaussian diffusion model in a street canyon environment, taking into account the impact of particulate matter settling velocity on concentration distribution. The modified Gaussian diffusion model formula is then determined. Multiple iterative calculations are performed using the street canyon particle concentration distribution model to gradually approximate the simulation results to the actual situation. In each iteration, the model parameters and initial concentration distribution are initialized based on the current concentration distribution and geographic information. The concentration distribution at each location is calculated using the modified Gaussian diffusion model formula. Model parameters, including diffusion coefficient, settling velocity, wind speed, and wind direction, are adjusted based on geographic information. The concentration distribution calculation and model parameter adjustment process are repeated until the error between the simulation results and the actual monitoring data is within an acceptable range. After multiple iterative calculations, the street canyon particle concentration distribution model outputs particle concentration distribution data for each location within the street canyon, which is used to generate accurate air quality maps that intuitively display the particle concentration distribution in different areas of the street canyon.

[0068] The basic formula of the Gaussian diffusion model is expressed as:

[0069] ;

[0070] The formula of the modified Gaussian diffusion model is expressed as follows:

[0071] ;

[0072] Where, For the location The concentration of particulate matter at is the emission source intensity of particulate matter, is the wind speed, Respectively expressed in The diffusion coefficient in the direction, , Respectively in Direction and Distance in direction, is the sedimentation rate, For Distance in direction;

[0073] The trend prediction module uses a time series analysis algorithm to combine historical and real-time monitoring of particle concentration data to analyze the changing trends of street valley particle concentrations and identify the development trend of street valley particle pollution in advance.

[0074] The air quality warning module is used to set corresponding warning thresholds based on the street valley particulate matter concentration distribution model and change trend analysis results, combined with air quality standards, to determine whether to issue air quality warning information. This improves the accuracy and timeliness of air quality warnings and promptly reminds the public to take protective measures to reduce the risk of exposure to high concentrations of particulate matter and protect public health.

[0075] Example 2, as Figure 1 、 Figure 2 As shown, based on Example 1, the present invention provides a technical solution: preferably, the trend prediction module specifically includes:

[0076] The historical monitored particle concentration data is obtained from the database, covering the particle concentration information of the street valley in different time periods in the past, and recording the changes of particle concentration over time. At the same time, the particle concentration data collected and fused by the current sensor network is received in real time, and the historical data is integrated with the real-time data to unify the data format and time accuracy to ensure that the data is continuous and comparable on the time axis. Based on the historical monitored particle concentration data, the trend characteristics related to the particle concentration change trend are extracted, including the daily mean concentration change rate, the weekly mean concentration range and the concentration seasonality index. The daily average concentration change rate is the ratio of the difference between the daily average particle concentrations of two consecutive days to the daily average of the previous day, and then multiplied by 100% to get the change rate. It is used to measure the amplitude and direction of the change in the daily average particle concentration between two consecutive days, and can reflect the fluctuation of the particle concentration in the short term (two consecutive days). The weekly average concentration range is the difference between the maximum and minimum daily average particle concentrations within a week, which is used to measure the fluctuation range of the particle concentration within the week. The larger the range, the more drastic the fluctuation of the particle concentration within the week, and the smaller the range, the relatively stable the particle concentration within the week. The concentration seasonality index is the ratio of the average particle concentration in a certain season to the average particle concentration throughout the year. It is used to measure the degree of change in particle concentration in a season relative to the whole year and reflect the seasonal characteristics of particle concentration. If the seasonality index is greater than 100%, it means that the particle concentration in this season is higher than the average level of the whole year. If it is less than 100%, it means that it is lower than the average level of the whole year. According to the characteristics of the street valley particle concentration data, a time series analysis algorithm based on the long short-term memory network is used to construct a trend prediction model. The trend characteristics including the daily mean concentration change rate, the weekly mean concentration range and the concentration seasonality index are used as input, and the future change trend of the street valley particle concentration is used as output. During the construction process, the model parameters are estimated and tested. By adjusting the parameters multiple times, the model can better fit the historical data and capture the inherent laws and trends in the data. To ensure the model has high accuracy and reliability, the trend prediction model is trained using most of the historical data to allow the model to learn the patterns and regularities in the data. During the training process, the model parameters are continuously optimized to improve the model's fit to the historical data. After training is completed, the model is verified using the remaining historical data to evaluate the model's prediction performance. The accuracy and stability of the model are judged by calculating the error index between the predicted value and the actual value. If the verification result is not ideal, the model is adjusted and optimized in the previous step until it meets the prediction requirements. The real-time fused particulate matter concentration data is input into the trend prediction model to obtain the predicted value of the particulate matter concentration at a future time point. The future trend of the street valley particulate matter concentration is then analyzed and the prediction results are output in the form of a chart to show the trend of the particulate matter concentration over time.

[0077] In addition, the calculation process of the predicted value of particulate matter concentration at a future time point is:

[0078] The particle concentration value at the current time point is obtained from the real-time fused particle concentration data as the basis for prediction. The ratio of the daily mean concentration change rate to the particle concentration value at the current time point is calculated, and 1 is added to calculate the exponential growth or decay of the particle concentration within t days. The exponential growth or decay portion of the particle concentration value at the current time point is analyzed and then multiplied by the particle concentration value at the current time point to obtain the adjusted value of the current concentration. The ratio of the weekly mean concentration range to the particle concentration value at the current time point is calculated and multiplied by the sine function to obtain the intra-weekly periodic fluctuation function. The ratio of the concentration seasonal index to the particle concentration value at the current time point is calculated, and an exponential function calculated based on the ratio of the predicted time step to the seasonal cycle is calculated. The two are multiplied to obtain the seasonal variation function. The adjusted value of the current concentration, the intra-weekly periodic fluctuation function, and the seasonal variation function are added together to obtain the predicted value of the particle concentration at the future time point.

[0079] The calculation expression for the predicted value of particulate matter concentration at a future time point is:

[0080] ;

[0081] Where, is the predicted value of the particulate matter concentration at a future time point, is the particle concentration value at the current time point, is the daily mean concentration change rate, The range of the weekly average concentration is is the concentration seasonality index, is the time step of prediction, For seasonal cycles, Indicates the length of a season and the exponential growth or decay trend of particulate matter concentration. The concentration may increase or decrease, Represents the intra-week periodic fluctuation of particulate matter concentration, according to The size and concentration of the represents the seasonal variation of particulate matter concentration, according to Size and The concentration may show a gradual increasing or decreasing trend within the season;

[0082] The air quality early warning module specifically includes:

[0083] The system collects the particle concentration distribution data for each area within the street canyon, as output by the street canyon particle concentration distribution model, and the analysis results of future particle concentration trends obtained by the trend prediction module. Furthermore, it obtains national and local air quality standards and defines the particle concentration threshold ranges corresponding to different air quality standards. Based on the integrated street canyon particle concentration data, trend prediction results, and air quality standards, it sets corresponding particle concentration warning thresholds to ensure that the warning thresholds accurately reflect the air quality status and can detect potential pollution risks in advance. Real-time monitoring data and prediction results are compared with the set particle concentration warning thresholds to determine whether air quality warning information should be issued. Once an alert is triggered, the air quality warning information is disseminated through various channels (such as text messages, app push notifications, and social media). The air quality warning information includes the current air quality status, future trends, health risk warnings, and recommended protective measures.

[0084] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A multi-source sensor fusion street valley particulate matter distribution monitoring system, including a particulate matter distribution monitoring platform, characterized by: The particle distribution monitoring platform is communicatively connected to a sensor network module, a multi-source data fusion module, an air quality model module, a trend prediction module, and an air quality warning module, wherein electrical signals are connected between the modules; The sensor network module is used to deploy multiple sensor nodes based on the spatial structure characteristics of the street valley and collect monitored particulate matter concentration data from each sensor node; The multi-source data fusion module is used to fuse the particle concentration data from different locations using a data fusion algorithm; The air quality model module is used to combine the fused particulate matter concentration data to construct a street valley particulate matter concentration distribution model and generate an air quality map; The trend prediction module is used to analyze the changing trend of street valley particulate matter concentration by combining historical monitoring and real-time monitoring particulate matter concentration data using a time series analysis algorithm; The air quality warning module is used to set corresponding warning thresholds based on the street valley particulate matter concentration distribution model and change trend analysis results, combined with air quality standards, to determine whether to issue air quality warning information.

2. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 1, characterized in that: The sensor network module includes a node deployment unit and a data acquisition unit; The node deployment unit is used to plan and deploy sensor nodes and install multiple particulate matter sensors based on the spatial structural characteristics of the street valley, including building height, density distribution, and street orientation; The data collection unit is used to collect the monitored particulate matter concentration data from each sensor node, and perform pre-processing to match it with time and geographical information.

3. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 2, characterized in that: The node deployment unit specifically includes: Obtain building height and density distribution data through the urban geographic information system, use urban planning drawings to clarify street directions, widths, and surrounding topography, and integrate them into a spatial structure database; Based on the analysis of the street valley's spatial structure, the layout of sensor nodes was planned. The installation locations and number of sensors were determined based on the characteristics of the street and laneway subdivisions. The spacing of sensors was also arranged based on the diffusion characteristics of particulate matter. Sensor nodes were deployed on both sides of the street, at laneway entrances, in areas with high traffic flow, and in locations with large changes in building height. Install the particulate matter sensors according to the planned sensor node layout plan, and debug the sensors after installation.

4. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 2, characterized in that: The data acquisition unit specifically includes: Establish communication connections with each sensor node and receive real-time particle concentration data collected by the particle sensors of the sensor nodes; A verification mechanism is used to perform preliminary verification on the received particulate matter concentration data, and then preprocessing operations are performed on the data that pass the preliminary verification, including data cleaning, normalization and code conversion steps; Match the preprocessed particulate matter concentration data with the time information, obtain the current time from the system clock, and add a timestamp to each data record; According to the deployment location information of the sensor nodes, the particulate matter concentration data is matched with the corresponding geographic information, and the data records are associated with the corresponding geographic location coordinates to form a data set with spatial attributes.

5. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 1, characterized in that: The multi-source data fusion module specifically includes: Receive particulate matter concentration data from different locations from the sensor network module and perform basic rechecks on the particulate matter concentration data, including data format, data volume, and data timestamp integrity; After the particle concentration data passed the basic review, the Kalman filter method was selected as the data fusion algorithm for processing. Based on the laws and characteristics of the street canyon particle concentration changes, a state equation and an observation equation were established. Through these two equations, the Kalman filter algorithm was used to dynamically estimate and update the particle concentration data. Continuous iterative calculations gradually approached the actual particle concentration value. Verify and optimize the fused data results, collect known reference data for comparison with the fused data, and adjust and optimize the fusion algorithm based on the verification results.

6. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 1, characterized in that: The air quality model module specifically includes: Obtain the fused particulate matter concentration data output by the multi-source data fusion module, and simultaneously collect geographic information data related to the street canyon, including building distribution, street orientation, and topography. The fused particulate matter concentration data and geographic information data are integrated to obtain the basic input data set. According to the characteristics of street canyons, a Gaussian diffusion model was selected to construct a street canyon particulate matter concentration distribution model. Based on the theoretical formula of the Gaussian diffusion model and the fused particulate matter concentration data and geographic information data, the model parameters were set, including the emission source intensity, diffusion coefficient, deposition rate, wind speed and direction of particulate matter. The fused particle concentration data and geographic information data are input into the constructed street canyon particle concentration distribution model for calculation. The model simulates the diffusion, migration, and transformation of particles within the street canyon based on the input initial particle concentration value, geographic information, and set model parameters. Through multiple iterative calculations, the model's internal variables are adjusted to gradually make the simulation results closer to the actual situation. After the calculation is completed, the particle concentration distribution data at each location in the street canyon is analyzed. Using a geographic information system, the simulation results of the particle concentration distribution data at various locations in the street canyon were visualized to generate an air quality map. The map shows the concentration distribution of particle matter in different areas of the street canyon with different colors.

7. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 6, characterized in that: The analysis process of the particle concentration distribution data at each location in the street canyon is as follows: The fused particulate matter concentration data and geographic information data were input into the street canyon particulate matter concentration distribution model to simulate the concentration distribution of particulate matter in the street canyon. The initial conditions of the model were set, including the initial value of the particulate matter concentration, geographic information, and model parameters. The model parameters included the emission source intensity, diffusion coefficient, deposition rate, wind speed, and wind direction of particulate matter. The street canyon particle concentration distribution model simulates the diffusion, migration, and transformation of particles within the street canyon based on the input initial particle concentration value, geographic information, and model parameters. It also determines the basic formula for the Gaussian diffusion model. In the street canyon environment, the Gaussian diffusion model is modified based on the effect of particle settling velocity on concentration distribution, and the formula for the modified Gaussian diffusion model is then determined. The street valley particulate matter concentration distribution model was used to perform multiple iterative calculations to gradually bring the simulation results closer to the actual situation. In each iteration, the model parameters and initial concentration distribution were initialized based on the current concentration distribution and geographic information. The concentration distribution at each location was calculated using the formula of the modified Gaussian diffusion model. The model parameters, including the diffusion coefficient, sedimentation rate, wind speed, and wind direction, were adjusted based on the geographic information. The concentration distribution calculation process at each location and the model parameter adjustment process were repeated until the error between the simulation results and the actual monitoring data was within an acceptable range. After multiple iterative calculations, the street valley particle concentration distribution model outputs the particle concentration distribution data at each location in the street valley.

8. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 1, characterized in that: The trend prediction module specifically includes: Obtain historical monitored particulate matter concentration data from the database, covering the Street Valley's past particulate matter concentration information at different time periods, and recording changes in particulate matter concentration over time. Simultaneously, receive real-time particulate matter concentration data collected and fused by the current sensor network, integrating historical data with real-time data. Based on historical monitored particulate matter concentration data, trend characteristics related to particulate matter concentration changes are extracted, including the daily mean concentration change rate, the weekly mean concentration range, and the concentration seasonality index; Based on the characteristics of street valley particulate matter concentration data, a trend prediction model was constructed using a time series analysis algorithm based on a long short-term memory network. The model took as input the trend characteristics of the daily mean concentration change rate, the weekly mean concentration range, and the concentration seasonality index, and output the future trend of street valley particulate matter concentration. The real-time fused particle concentration data is input into the trend prediction model to obtain the predicted value of the particle concentration at a future time point. The future trend of the particle concentration in the street valley is then analyzed and the prediction results are output in the form of a chart, showing the change trend of the particle concentration over time.

9. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 8, characterized in that: The calculation process of the predicted value of the particulate matter concentration at the future time point is: Obtain the particle concentration value at the current time point from the real-time fused particle concentration data as the basic value for prediction; Calculate the ratio of the daily average concentration change rate to the particle concentration value at the current time, add 1 to calculate the exponential growth or decay of the particle concentration within t days, analyze the exponential growth or decay portion of the particle concentration value at the current time, and then multiply it by the particle concentration value at the current time to obtain the adjusted value of the current concentration; Calculate the ratio of the weekly average concentration range to the particle concentration value at the current time point, and multiply it by the sine function to obtain the intra-week periodic fluctuation function; Calculate the ratio of the concentration seasonality index to the particle concentration value at the current time point, and calculate the exponential function based on the ratio of the predicted time step to the seasonal period. Multiply the two to obtain the seasonal variation function. The current concentration adjustment value, the weekly periodic fluctuation function, and the seasonal variation function are added together to obtain the predicted value of the particulate matter concentration at a future time point.

10. The multi-source sensor fusion street valley particulate matter distribution monitoring system according to claim 1, characterized in that: The air quality early warning module specifically includes: Collect the particle concentration distribution data for each area in the street canyon, as output by the street canyon particle concentration distribution model, and the analysis results of future particle concentration trends obtained by the trend prediction module. At the same time, obtain air quality standards and clarify the particle concentration threshold ranges corresponding to different air quality standards. Based on the integrated street valley particulate matter concentration data, trend prediction results, and air quality standards, corresponding particulate matter concentration warning thresholds are set; Compare real-time monitoring data and prediction results with the set particulate matter concentration warning threshold to determine whether it is necessary to issue air quality warning information. Once the warning is triggered, the air quality warning information will be issued through multiple channels.

Images