Internet-of-things human settlement ecological command management system

Abstract

The invention discloses an internet-of-things human settlement ecological command management system, particularly relates to the field of internet-of-things data modeling and analysis, and is used for solving the problem that an industrial park human settlement community lacks intelligent regulation and control management on air dust. According to the system, air dust data and meteorological sensing data of an industrial area are acquired through an acquisition module, and an air dust dynamic data flow is generated; the dust distribution prediction module is used for carrying out time sequence modeling on the dynamic data flow to form dust aggregation distribution prediction; the crowd distribution prediction module is used for constructing a crowd distribution prediction model based on the historical movement track and time distribution of the residents; the risk assessment module is used for spatially superposing an air dust prediction result and a crowd distribution prediction result and assessing an exposure risk; the internet-of-things management module obtains installation coordinates and action ranges of edge regulation and control equipment, maps the installation coordinates and action ranges to a community plane and accesses a unified command and dispatch center; and the regulation and control issuing module is used for issuing a regulation and control instruction by the dispatching center to realize spraying, purification or isolation intervention.

Description

Technical Field

[0001] This invention relates to the field of IoT data modeling and analysis technology, and more specifically, to an IoT-based human settlement ecological command and management system. Background Technology

[0002] With the advancement of smart city and smart park construction, the application of the Internet of Things (IoT) in the management of human settlements and environmental safety in industrial parks is gradually deepening. By deploying various types of sensors, intelligent control equipment, and a unified scheduling platform in the park, a real-time data perception and linkage system covering the entire area is formed. Building an IoT-based industrial park ecological command system has become a new trend in improving the livability of the park's working environment and ensuring the health and safety of employees.

[0003] Currently, air pollution problems such as particulate matter, dust, and exhaust fumes from large-scale production activities and logistics transportation in modern industrial parks are becoming increasingly prominent. Particularly during high-emission processes in industries such as smelting, chemicals, and building materials, air dust concentrations exhibit significant spatiotemporal dynamic fluctuations. Park workers are frequently exposed to high concentrations of air dust and harmful particulate matter during their daily activities such as commuting, production, and inspections, which can easily lead to respiratory diseases, occupational asthma, and even safety accidents. Traditional methods relying on single-point monitoring or post-incident medical and accident data lack real-time and spatial accuracy, making it difficult to provide timely and effective protection and early warning. Meanwhile, although many parks have installed air quality monitoring, meteorological sensors, dust suppression spraying, and exhaust purification control facilities, these devices are scattered and independent, lacking a unified perception and command system, and unable to achieve the fusion, correlation, and intelligent prediction of multi-source data. The lack of dynamic modeling of dust and harmful particle diffusion, accurate overlay of personnel operation and movement patterns, and intelligent assessment of exposure risks results in lagging control measures, low resource utilization efficiency, and significant shortcomings in public health and production safety assurance.

[0004] Therefore, there is an urgent need for an IoT-based ecological command and management system suitable for industrial parks that integrates dynamic monitoring, intelligent prediction, and coordinated control. This system should connect the entire chain of multi-source data collection, dynamic diffusion modeling, personnel risk superposition, and equipment coordinated scheduling to improve the safety level of public health and working environment in the park. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an Internet of Things human settlement ecological command and management system to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An IoT-based human settlement ecosystem command and management system includes: The data acquisition module is used to collect air dust detection data and meteorological sensor data from the industrial park and generate a dynamic air dust data stream. The dust distribution prediction module is used to perform time-series modeling of dynamic data streams of air dust, calculate the accumulation rate and diffusion gradient of air dust, and predict the aggregation and distribution of air dust. The population distribution prediction module is used to obtain the travel patterns of people in the industrial park's residential communities and to build a population distribution prediction model based on historical movement trajectories and time distribution. The risk assessment module is used to spatially overlay the airborne dust prediction results with the population distribution prediction results to generate a set of airborne dust exposure risk matrices. The IoT management module is used to obtain the installation coordinates and range of action of edge control devices, map them to the plane coordinate system of the industrial park and residential community, and connect them to the unified command and dispatch center. The control and control module is used to obtain the control location and control period when any element in the risk matrix is ​​identified as having a high risk of exposure, and then the control and control command is issued by the dispatch center.

[0007] In a preferred embodiment, the acquisition module acquires air dust detection data and meteorological sensor data from the industrial park to generate a dynamic air dust data stream, specifically including: Based on the actual geographical scope of the industrial park's residential communities and the location of industrial facilities, a unified planar coordinate system covering the entire community area is established and divided into spatial units according to a preset grid resolution. Based on the air dust detection equipment evenly deployed in the residential areas of the industrial park, data on the concentration of air dust in the air is collected. A meteorological sensor is installed in the vicinity of the air dust detection equipment, and the meteorological sensor collects wind speed, wind direction, humidity and temperature parameters in real time; The parameters collected by the air dust detection equipment and the meteorological sensor equipment are time-aligned and combined into a dynamic air dust data stream.

[0008] In a preferred embodiment, the air dust distribution prediction module performs time-series modeling of the air dust dynamic data stream, calculates the air dust accumulation rate and diffusion gradient, and predicts the air dust aggregation distribution, specifically including: Trend decomposition was performed on the time series of air dust concentration. Long-term trend term and periodic term were obtained based on moving average and periodic fitting, and sudden term was extracted by residual anomaly detection. A two-dimensional vector field is constructed based on wind speed and wind direction parameters. The wind direction and wind speed of each meteorological monitoring point are mapped to the community spatial unit using the grid interpolation method to form a wind field vector model. By combining humidity and temperature parameters, the air dust settling rate is estimated, and a three-dimensional prediction field for diffusion and settling is established. In the prediction field, the cumulative concentration curve and residual time curve of air dust in each spatial unit are estimated. The cumulative concentration and residual time are weighted and calculated as the prediction aggregation index to generate a predicted aggregation distribution map of air dust.

[0009] In a preferred embodiment, establishing the three-dimensional prediction field for diffusion and sedimentation specifically includes: In each spatial unit of the industrial park's residential community planar coordinate system, a height layer is superimposed, and the settling velocity corresponding to the air dust particle size is corrected according to humidity and temperature parameters to establish a settling profile in the vertical direction. The long-term trend term, periodic term, and sudden term obtained from the decomposition are used as time boundary conditions and coupled with the wind field vector model and settlement profile. The evolution of air dust concentration in the three-dimensional grid is calculated iteratively according to the time step to establish a three-dimensional prediction field covering the horizontal and vertical directions.

[0010] In a preferred embodiment, the population distribution prediction module, which acquires the travel patterns of community residents and constructs a population distribution prediction model based on historical movement trajectories and time distribution, specifically includes: Collect residents' historical movement trajectory data, extract the areas and durations of stay for residents at different times within a set monitoring period, and integrate them to form a dataset of residents' travel patterns; Based on the age structure of residents, sensitive groups are identified, and clustering operations are performed on the travel pattern dataset of the identified sensitive groups to generate a set of travel patterns of the sensitive groups. Based on travel pattern sets, the population density distribution in different spatial units of the community is predicted within each time window, and a distribution prediction map of sensitive populations is established.

[0011] In a preferred embodiment, the risk assessment module spatially overlays the airborne dust prediction results with the population distribution prediction results to generate an airborne dust exposure risk matrix set, specifically including: The predicted distribution map of airborne dust and the predicted distribution map of sensitive populations are overlaid in the plane coordinate system of the industrial park residential community according to the preset time segment windows; For each overlapping spatial unit after stacking, the product of the predicted aggregation index of airborne dust and the predicted value of population density is calculated to obtain the initial exposure risk index. Exposure risk values ​​are obtained by weighting the regional attributes of the overlapping spatial units. These exposure risk values ​​are then integrated according to spatial units and time order to form an exposure risk matrix set.

[0012] In a preferred embodiment, the IoT management module acquires the installation coordinates and effective range of the edge control device, maps them to the industrial park's residential community plane coordinate system, and connects them to the unified command and dispatch center. Specifically, this includes: Obtain the fixed installation coordinates, effective range, and operational period of the edge control device, and convert the installation coordinates and effective range into a spatial coverage polygon. The spatial coverage polygon is compared with each spatial unit of the industrial park's residential community planar coordinate system to screen candidate control devices for each spatial unit. The edge control device includes a spray device, air purification facilities, and temporary isolation facilities; The edge control devices are networked according to their corresponding spatial coverage units, and a mapping relationship between device identifiers and spatial unit coverage is established and connected to the unified command and dispatch center.

[0013] In a preferred embodiment, in the control and control issuance module, when any element in the risk matrix is ​​identified as having a high risk of exposure, the control location and control period are obtained, and the control instruction is issued by the dispatch center. Specifically, this includes: Filter the matrix elements in the exposure risk matrix set whose exposure risk values ​​reach a preset risk threshold and mark them as high-risk exposures; Obtain the spatial units and temporal prediction windows corresponding to all high-risk exposure elements in the matrix set, and mark them as control locations and control periods; Based on the spatial unit to which the control location belongs, candidate edge control devices corresponding to the control location are extracted; The operating time of the candidate edge control device is compared with the control location to select the target control device; Obtain the device identifier of the target control device and generate control commands, then send the control commands to the target device.

[0014] The technical effects and advantages of the IoT-based human settlement ecological command and management system of the present invention are as follows: This invention, through the construction of an IoT-based human settlement ecosystem command and management system, achieves closed-loop linkage of multi-source data fusion, intelligent prediction, and automatic control, demonstrating significant technical effects and application advantages. First, the system deploys air dust detection and meteorological sensors within the community to achieve high spatiotemporal resolution acquisition of dynamic air dust data. Second, by establishing a three-dimensional prediction field through trend decomposition, wind field modeling, and deposition and diffusion calculations, it comprehensively considers long-term trends, periodic fluctuations, and sudden anomalies, achieving accurate prediction of air dust aggregation and distribution, significantly outperforming traditional methods based on single average values ​​or historical statistics. Third, the system combines population travel pattern predictions with the spatiotemporal distribution of sensitive populations with air dust prediction results to generate an exposure risk matrix, enabling targeted risk assessment and avoiding ineffective or excessive protection. Finally, the system unifies the connection of edge control devices to the command center, issuing control commands in real time based on the risk matrix, automatically linking spraying, sprinkler systems, air purification, and isolation facilities, improving response speed and resource utilization efficiency, forming a closed loop of prediction-assessment-control. This system improves the protection level against industrial dust in residential communities, achieving intelligent and refined management of the human settlement ecosystem, and has significant application value. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the structure of an IoT-based human settlement ecological command and management system according to the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1 Figure 1 The present invention discloses an IoT-based human settlement ecological command and management system, comprising: The data acquisition module is used to collect air dust detection data and meteorological sensor data from the industrial park and generate a dynamic air dust data stream. The dust distribution prediction module is used to perform time-series modeling of dynamic data streams of air dust, calculate the accumulation rate and diffusion gradient of air dust, and predict the aggregation and distribution of air dust. The population distribution prediction module is used to obtain the travel patterns of people in the industrial park's residential communities and to build a population distribution prediction model based on historical movement trajectories and time distribution. The risk assessment module is used to spatially overlay the airborne dust prediction results with the population distribution prediction results to generate a set of airborne dust exposure risk matrices. The IoT management module is used to obtain the installation coordinates and range of action of edge control devices, map them to the plane coordinate system of the industrial park and residential community, and connect them to the unified command and dispatch center. The control and control module is used to obtain the control location and control period when any element in the risk matrix is ​​identified as having a high risk of exposure, and then the control and control command is issued by the dispatch center.

[0018] The acquisition module collects air dust detection data and meteorological sensor data from the industrial park to generate a dynamic air dust data stream.

[0019] A unified spatial reference framework is established based on the actual geographical scope of the target industrial park's residential community. This framework is defined using a Cartesian coordinate system, selecting a corner point with a fixed geographic identifier within the community boundary as the origin. The directions of the X and Y axes are determined using actual geographic surveying data to ensure consistency with the actual geographic directions; typically, the X-axis points due east and the Y-axis points due north. After establishing the coordinate system, a high-precision survey of the entire community area is required to obtain two-dimensional vector boundaries for all major geographic entities, including roads, industrial areas, and building outlines. To facilitate subsequent calculations and spatial mapping of data acquisition, the Cartesian coordinate system needs to be divided into spatial units using a grid. The resolution of the grid is set based on the spatial accuracy requirements of the target application; by default, the community area is divided into square grid units with sides of 10 meters or 20 meters. When the community area is large, a larger unit side length can be selected to reduce computation; when the community area is small and more refined risk prediction is required, a higher resolution unit side length can be selected. The side length is set based on the spatial variation characteristics of airborne dust diffusion and accumulation within the coverage area, ensuring that each unit accurately reflects the spatial differences of the local environment. After the grid is divided, each grid cell should be assigned a unique number, and the number should be mapped and stored to the spatial range of its planar coordinates.

[0020] After establishing a unified plane coordinate system, air dust detection equipment and meteorological sensors are strategically deployed within the community. The air dust detection equipment is primarily deployed near industrial facilities, as these are the main sources of air dust release. The deployment density and distribution of the equipment must comprehensively cover these areas. During deployment, the equipment is evenly distributed according to a spatial grid, with fixed installations chosen at the center of each industrial facility or along the main air duct, ensuring that the collected data represents the air dust concentration within that unit. The equipment installation height should be set between 1.5 and 2 meters above the ground, matching the typical breathing zone height of residents. The equipment's detection accuracy must be able to identify changes in air dust concentration. The sampling frequency is set to once every 5 minutes by default, but can be modified according to actual needs. Meteorological sensors must be deployed simultaneously near each air dust detection device, with the distance between them controlled within 3 to 5 meters to avoid data inconsistencies caused by local interference. Meteorological sensing equipment needs to be able to monitor environmental parameters such as wind speed, wind direction, humidity, and temperature in real time. The sampling frequency of wind speed and wind direction should be consistent with that of air dust detection equipment, while the sampling frequency of humidity and temperature can be slightly lower, such as once every 10 minutes, to ensure the synchronization and usability of the data.

[0021] Airborne dust concentration is aligned with meteorological parameters to unify the time reference. Time alignment uses interpolation methods to map data from different sampling frequencies to a unified time axis, for example, constructing time-series data points at 5-minute time steps. After alignment, multi-dimensional parameters such as airborne dust concentration, wind speed, wind direction, humidity, and temperature are integrated into a multimodal dynamic data stream. The dynamic data stream has temporal sequence and spatial location identifiers, and each record includes the sampling time, spatial unit number, and values ​​of various detection parameters.

[0022] In the air dust distribution prediction module, time-series modeling is performed on the dynamic data stream of air dust to calculate the accumulation rate and diffusion gradient of air dust and predict the aggregation and distribution of air dust.

[0023] Time series data of continuously monitored air dust concentrations were obtained. The selected dynamic air dust data came from various monitoring devices within the community, covering at least 60 consecutive days of records to ensure the sample included complete diurnal and weekly seasonal characteristics. Data preprocessing was performed before trend decomposition to ensure the input series was continuous and free from significant noise distortion. Preprocessing included: limiting concentration values ​​to between 0 and 2000 units; discarding distorted data if values ​​exceeded the limit; identifying jumps if the difference between adjacent sampling points exceeded twice the 90th percentile of the daily difference distribution, and replacing the value with the three-point median; imputing gaps of less than 30 minutes using linear interpolation, and backfilling gaps of more than 30 minutes but less than 2 hours using historical medians from the same week and time period.

[0024] After preprocessing, trend decomposition is performed. Long-term trend terms are extracted using a moving average method, with a window size covering at least 24 hours to ensure short-term fluctuations are smoothed; for example, a window of 288 sampling points (corresponding to 24 hours) is set. Based on this, a quadratic polynomial fitting is used to compensate for long-term slow changes, resulting in a smooth background change curve. Periodic terms are fitted based on daily and weekly cycles, using a combination of sine and cosine fitting methods. The period length is first fixed (24 hours and 7 days), and then the amplitude and phase parameters are fitted using least squares, ensuring the fitted curve closely matches the residual changes after removing the long-term trend. After fitting, the residual sequence is calculated and anomaly detection is performed using a sliding window standard deviation method: for each residual sequence grouped into 30 sampling points (2.5 hours), the mean and standard deviation are calculated. If the current point deviates from the window mean by more than three times the standard deviation, it is identified as a sudden event, and the location is marked, along with the time of occurrence and the corresponding concentration value. After the above steps, three components are finally obtained: the long-term trend component describes the overall background concentration change, the periodic component reflects the recurring fluctuations in daily or weekly periods, and the sudden component corresponds to local or short-term strong abnormal events. Finally, these three components are stored in the database.

[0025] Real-time wind speed and direction parameters are collected from meteorological monitoring points deployed throughout the community. The collection frequency is synchronized with airborne dust concentration, uniformly every 5 minutes. Wind speed is standardized to meters per second, and wind direction is represented from 0° to 360°, increasing clockwise from true north. After collection, wind direction and speed are converted into two components of a planar vector: the east-west component is obtained by multiplying wind speed by the sine of wind direction, and the north-south component is obtained by multiplying wind speed by the cosine of wind direction. After conversion, each monitoring point corresponds to a pair of velocity components, along with their spatial coordinates. A continuous vector field is established across the entire community space, and spatial interpolation is performed on the discrete monitoring point data. Specifically, a grid interpolation method is used, with the grid resolution set according to the industrial park's residential community planar coordinate system to ensure sufficient resolution to capture local wind speed differences. For each grid point, an inverse distance weighted interpolation method is used: the four nearest monitoring points are selected, and the reciprocal of their distances is calculated as weights. After normalization, these weights are multiplied by the velocity components corresponding to each monitoring point and summed to obtain the east-west and north-south components of the grid point. If there are few monitoring points, a search radius is set to prevent extrapolation errors; monitoring points are ignored when the distance exceeds this radius. During interpolation, wind speed attenuation near buildings or trees needs to be corrected. For example, within 20 meters of the edge of a building, the wind speed component is multiplied by a attenuation factor of 0.7, while the wind direction remains unchanged. After interpolation is completed for the entire community, a two-dimensional vector field is obtained, with each grid point having a clear velocity direction and magnitude.

[0026] Vertical height layers are superimposed on each horizontal unit to form a spatially discrete grid with a three-dimensional structure. The vertical height range is determined based on the characteristics of the living environment, ensuring coverage from the ground surface to the top of community buildings. The number of layers is set according to the simulation accuracy requirements, and the thickness between layers remains uniform. The entire community area is formed into a complete three-dimensional spatial grid composed of horizontal units and height layers. When establishing the vertical settlement profile, the settling velocity of airborne dust within each height layer is determined. The settling velocity is affected by the particle size of the dust particles themselves, as well as air humidity and temperature. In practice, airborne dust is classified by particle size, and particles of different size ranges are divided into several levels according to their diameter. Each level corresponds to a baseline settling velocity, which is determined based on aerodynamic settling experiments. Subsequently, the baseline settling velocity is corrected based on air humidity and temperature parameters. When humidity increases, particles are more likely to absorb moisture, leading to an increase in equivalent particle size and thus accelerating settling. Increased temperature changes air viscosity, thereby affecting the settling velocity. The correction process is achieved by establishing a mapping relationship between environmental parameters and settling velocity. For example, a table of correspondence between humidity percentage and particle hygroscopic coefficient is pre-established, and an air viscosity adjustment coefficient is calculated based on real-time temperature values. These two corrections are then superimposed on the baseline settling velocity to obtain the actual settling velocity at the current height level. To ensure the continuity of the vertical profile, interpolation or smoothing methods are used between different height levels to ensure a reasonable continuous gradient of settling velocity with height rather than abrupt changes. After the above process, the corrected settling velocity distribution in the vertical direction of each horizontal grid cell is finally obtained, forming a complete vertical settling profile. The key to this process lies in the logical chain of the method rather than a fixed value. Therefore, the above process can be directly applied according to monitoring parameters in different implementation scenarios, ensuring the feasibility and adaptability of profile construction.

[0027] When calculating the three-dimensional prediction field, the long-term trend term, periodic term, and sudden term obtained from the previous decomposition are used as time boundary conditions and organically coupled with the aforementioned wind field vector model and settlement profile. Iterative calculations of concentration evolution are then performed according to a preset time step. Specifically, the iteration time interval is determined based on the data update frequency, dividing the entire prediction period into continuous time steps, such as every five or ten minutes, to ensure that the calculation process reflects the temporal dynamic changes in concentration. Initially, the air dust concentration of each three-dimensional grid cell is initialized; this value can be provided by historical monitoring data or the final value of the previous prediction period. In each time step, the long-term trend term is loaded as a baseline change factor for the overall concentration. This baseline change factor refers to the explicit numerical correction amount in the aforementioned fitted background change curve, which is added to or subtracted from the initial value of each spatial grid. The periodic term calculates the fluctuation amplitude corresponding to the current moment based on its fitted time function (a combination of sine and cosine functions) and superimposes it on the baseline to form regular time fluctuations. The sudden term only acts within its corresponding time interval. When a time step is detected to fall within the sudden term interval, the increment of the sudden term is superimposed on the current predicted value to simulate short-term anomalies. After loading the time boundary conditions, spatial transport calculations are performed. The horizontal transport part calls the wind speed and direction information corresponding to each grid cell in the wind field vector model to calculate the amount of concentration transported along the wind direction between grids, implemented based on the windward difference method, to ensure that the concentration change in the horizontal direction is consistent with the wind field. The vertical transport part calls the previously established subsidence profile to transfer the particle concentration downward to the adjacent lower layers according to the corresponding subsidence velocity in each height layer, until the surface layer. After completing the spatial transport, it is corrected again according to the trend term of the current time step to ensure that the time boundary conditions continue to act throughout the entire evolution process. In this way, the time trend and the spatial transport process are coupled, rather than calculated separately. After each time step is calculated, the result will be used as the initial value for the next time step to continue iterating until the end of the entire prediction period, and finally a three-dimensional air dust concentration prediction field covering the horizontal and vertical directions will be obtained.

[0028] The three-dimensional prediction field is iteratively calculated at time steps to obtain a complete sequence of concentration evolution over time for each spatial unit throughout the entire prediction period. Based on this sequence, the concentration accumulation curve and residual time curve for each spatial unit are derived, which are then used to calculate the predicted aggregation index. Specifically, for each spatial unit, all time steps within the prediction period are traversed, and the predicted concentration value at each time step is recorded progressively. These concentration values ​​are then accumulated in chronological order to generate a concentration accumulation sequence. By obtaining the previously uniformly set time step, the interval corresponding to this time step is used as the accumulation weight (if the step step remains unchanged, the weights are equal). The concentration values ​​at each moment are multiplied by the corresponding time interval and then summed to form an accumulation curve reflecting the total exposure. After obtaining the concentration accumulation curve, the residual time curve is calculated. Residual time refers to the duration during which the concentration of a spatial unit exceeds a specific reference value or persists within the prediction period. The calculation process for the residual time curve involves scanning the predicted concentration value at each time step in chronological order and determining whether it is higher than a preset reference value. This reference value is set based on the safe concentration value corresponding to the type of airborne particulate matter. When it is higher, it is included in the effective residual time period and the accumulated time length forms a sequence of residual time evolution with the prediction time. When the concentration is below the reference value, the residual time is not increased. Through this stepwise accumulation method, a residual time curve is generated for each spatial unit throughout the entire prediction period, which reflects the change in the cumulative exposure time of that unit during the prediction period.

[0029] After obtaining the concentration accumulation curve and residual time curve, the two are combined to form a comprehensive index describing the risk of airborne dust accumulation. In implementation, a weighted calculation is used, combining the cumulative concentration and residual time according to preset weights. The default weight for concentration accumulation is 0.6, and the weight for residual time is 0.4; the specific weight values ​​are determined based on the hazard of particulate matter and preset safe concentration reference values. The total cumulative amount and total residual time for each spatial unit within the prediction period are normalized to ensure they are of the same order of magnitude, and then added together according to their weights to obtain the predicted accumulation index value for that spatial unit. This calculation is performed separately for each spatial unit, thereby obtaining the accumulation index distribution covering the entire community's three-dimensional prediction field and mapping it into a visual distribution map.

[0030] The population distribution prediction module obtains the travel patterns of community residents and constructs a population distribution prediction model based on historical movement trajectories and time distribution.

[0031] Track data is collected from residents using positioning devices deployed within the community, ensuring the collection period covers at least four consecutive weeks of 24-hour coverage. Track data collection uses minutes as the basic time resolution, recording the coordinates of residents at each time point as a single data point. After collection, the raw data is cleaned and standardized, removing points with abnormal positioning accuracy, such as drift points exceeding the community's boundaries within the same time period, or points where the speed between two consecutive points exceeds the upper limit of normal human walking or cycling speed. Subsequently, latitude and longitude are converted to planar coordinates within the community, and each coordinate point is mapped to a specific spatial unit according to a predefined spatial grid. Continuity analysis is performed on the track points in chronological order. If multiple consecutive track points are within a 20-meter radius and remain for more than 5 minutes (the specific range and time can be flexibly adjusted according to the community area), it is determined as a dwelling event. Simultaneously with identifying dwelling events, the start and end times, the spatial unit where the event occurred, and the total dwelling time are recorded. Next, all stay events were divided into 48 half-hour time windows per day. The stay area and duration for each resident within each time window were statistically analyzed, generating daily spatiotemporal stay characteristics. To ensure the reliability of subsequent data mining, records of residents with a total daily stay of less than 15 minutes were excluded from the analysis. The final resident travel pattern dataset includes fields such as date, time window number, spatial unit number, and total stay duration, and is stored using a discrete encoding method.

[0032] After obtaining the resident travel pattern dataset, residents are categorized into sensitive groups based on age structure. Sensitive groups are defined according to fixed rules, such as children under 12 years old and elderly residents under 65 years old, or flexibly based on job type and shift schedule. For the identified sensitive groups, their corresponding travel pattern datasets are extracted to construct behavioral pattern features. The proportion of each sensitive resident's stay in each half-hour time window and each spatial unit to their total stay for the day is used as a feature value, forming a vector of length equal to the number of time windows multiplied by the number of spatial units. To avoid excessive dimensionality affecting clustering results, spatial units are first clustered into activity clusters based on geographical proximity and functional similarity. The clustering criteria are fixed as adjacent units being less than a preset distance (default 30 meters) and having the same purpose. After dimensionality reduction, unsupervised clustering is used to cluster the feature vectors. The K-means clustering algorithm based on centroids is used, with the number of clusters selected between 3 and 6. The final value is selected through an internal scoring metric, which is a comprehensive evaluation of within-group squared error and between-group variance.

[0033] After clustering, the centroid vector of each cluster is obtained, reflecting the typical dwelling proportion of sensitive populations in each time window and activity cluster during the day under the corresponding pattern. Further statistical analysis of the occurrence probability of each pattern on weekdays, weekends, and holidays provides a reference for subsequent predictions. Each sensitive resident is labeled with the most frequently occurring pattern tag in their historical records, while the second most frequently occurring pattern tag is saved for pattern switching on abnormal days. Finally, a set of sensitive population travel patterns is formed, including fields such as time distribution and spatial distribution probability, all stored with discrete indices and proportion values ​​to ensure model reusability. Based on the generated sensitive population travel pattern set, the dwelling proportion of each pattern in each time window is extracted according to the predicted date, and then weighted and allocated to each spatial unit of the community in conjunction with the historical active population, forming the population density distribution for each time period, and establishing a sensitive population distribution prediction map.

[0034] In the risk assessment module, the air dust prediction results are spatially superimposed with the population distribution prediction results to generate a set of air dust exposure risk matrices.

[0035] The airborne dust prediction results and population distribution predictions are spatially overlaid in a unified industrial park residential community planar coordinate system. The setting method for each time segment window needs to be defined in advance, for example, using one hour as a time window, to ensure that it reflects the intraday variation trend. For the overlay operation, the airborne dust prediction aggregation distribution map is segmented and stored according to the preset time windows. Then, corresponding to the sensitive population distribution prediction map under the same time window, the corresponding values ​​are retrieved one by one using each spatial unit in the coordinate system as a matching benchmark to achieve complete spatial alignment. After alignment, numerical calculations are performed on each overlaid overlapping spatial unit, multiplying the corresponding airborne dust aggregation index and the predicted population density value. The purpose of this process is to quantify the coupling degree between the airborne dust aggregation intensity per unit space and unit time and the exposure probability of sensitive populations, forming an initial exposure risk index. For example, if the airborne dust aggregation index of a certain spatial unit in the current time window is 120 unit concentration and the predicted population density is 15 people / 100 square meters, then the calculated initial exposure risk index is 1800, which directly reflects the risk level of the area. Throughout the process, data consistency and strict adherence to time dimension correspondence must be ensured to avoid errors caused by time misalignment. After calculation, the initial exposure risk index of each spatial unit within the corresponding time window is stored in a matrix structure, forming an initial risk distribution map corresponding to spatial coordinates and time windows.

[0036] After obtaining the initial exposure risk index for each overlapping spatial unit, regional attributes are introduced to weight and correct the risk value. Regional attributes should be defined according to the community's functional zoning, typically including categories such as industrial facility proximity areas, public activity areas, and temporary stay areas. Each regional attribute should be assigned a fixed weight factor, set based on the exposure sensitivity and activity frequency of the sensitive population in that area. For example, the weight for industrial facility proximity areas can be set to 1.2; for temporary stay areas with shorter stay times, it can be set to 0.8; and for public activity areas, it can be set to 1.0 based on activity characteristics. In actual calculations, for each spatial unit, its regional attribute is first identified, and then the initial exposure risk index is multiplied by the corresponding weight factor to obtain the corrected exposure risk value. This correction process must be performed sequentially across all time windows to ensure consistency of the time series. After weighted correction, the exposure risk values ​​of each spatial unit in both the time and spatial dimensions are integrated according to coordinate index and time order to construct an exposure risk matrix set. The row dimension of this matrix set corresponds to the spatial unit index, the column dimension corresponds to the time window sequence, and the matrix elements are the weighted and corrected exposure risk values.

[0037] The IoT management module obtains the installation coordinates and operating range of the edge control device, maps them to the plane coordinate system of the industrial park and residential community, and connects them to the unified command and dispatch center.

[0038] Basic information was collected for all edge control devices within the community, including spray devices, air purification facilities, and temporary isolation facilities. Each device had clearly defined fixed installation coordinates, based on the community's unified planar coordinate system, obtained through on-site surveying or installation records. Coordinate accuracy was controlled within the meter range to ensure accurate spatial overlay. Each device also had a designed effective range, the specific value determined by the device's power and design standards. This effective range generally radiates outward from the device's air outlet or nozzle with a certain radius, or presents a regular or irregular coverage boundary. Furthermore, each device was equipped with operational time information, indicating the time interval within a day or week during which the device can be scheduled, recorded in the form of start and end time periods, such as 8:00 AM to 8:00 PM daily. After collecting information from these three dimensions, the device's installation coordinates and effective range were converted into spatial coverage polygons. Specifically, if the effective range is a regular circle, a circular buffer with a corresponding radius can be generated based on the installation coordinates, and then a covering polygon can be formed by using a polygon approximation algorithm (such as dividing the circumference into 360 equal division points); if the effective range is an irregular area, a polygon can be directly constructed based on the boundary point coordinates provided by the device design.

[0039] After generating the spatial coverage polygons, each of them is compared with the gridded spatial units in the community one by one. The comparison is based on the spatial geometric overlay algorithm to determine the overlapping area between the spatial coverage polygon and the boundaries of each spatial unit. If the overlapping area is greater than the preset comparison threshold (default is 10% of the spatial unit area), it is considered that the spatial unit is covered by the device, and the device is recorded as a candidate control device for that spatial unit. By performing the above comparison on the spatial coverage polygons of all devices, one or more candidate devices are finally selected for each spatial unit.

[0040] Taking each spatial unit as an index, summarize all the edge control devices it covers and establish the corresponding relationship. For the case where multiple devices cover the same spatial unit, record the unique identifiers of all devices, and at the same time store their spatial coverage polygons and operable time period information. Secondly, logically network the devices covering the same area or adjacent areas to form a multi-device linkage cluster. The aggregation method based on spatial proximity is used during networking. After establishing the device network, further construct the mapping relationship between the device identifier and the spatial unit coverage. This mapping relationship uses the spatial unit as the primary key and the list of device identifiers as the value to form a mapping table, the content of which includes: spatial unit coordinate index, covered device identifier, device type (spray, sprinkler, purification, isolation), device coverage range boundary parameters, and device operable time period, etc. Information. The establishment of the mapping relationship needs to ensure the consistency and integrity of the data. For example, when a device is removed or its position is changed, the mapping record of the corresponding spatial unit should be updated in real time. During the process of constructing the mapping table, the information of the networked linkage cluster should also be incorporated. After completing the mapping relationship, import all the device mapping information into the unified command and dispatch center through the standardized data interface. The unified command and dispatch center is a platform for centralized management and coordination, which can flexibly allocate devices based on the mapping relationship in three dimensions: time, space, and type.

[0041] In the control instruction issuing module, when any element in the risk matrix is identified and determined to be at high exposure risk, obtain the control position and control time period, and issue a control instruction by the dispatch center.

[0042] For the constructed exposure risk matrix set, each element in the matrix is ​​read one by one along both the temporal and spatial dimensions. Each element corresponds to a specific spatial unit and a specific prediction time window. Simultaneously, a pre-set risk threshold is retrieved. This threshold should be dynamically and comprehensively set based on changes in the respiratory disease incidence rate. For example, it can be derived statistically from past clinical event records and community environmental statistics, using methods to derive the cutoff value for the likelihood of respiratory diseases at a certain concentration and population exposure level, ultimately forming a clear numerical definition. The exposure risk value of the matrix element is compared with this pre-set threshold. When the exposure risk value reaches or exceeds the threshold, the matrix element is identified as a high-risk exposure point, and the spatial unit coordinates and corresponding time window of the element are recorded and uniformly marked as the control location and control period.

[0043] For each marked control location, the established coverage mapping relationship between spatial units and edge control devices is used to quickly retrieve a list of candidate control devices covering that spatial unit. These candidate devices include spray devices, air purification facilities, and temporary isolation facilities. After obtaining the candidate device list, the control period is compared with the available operating time of each device. Devices whose control period falls entirely within their available operating time are marked as meeting the actual control time conditions. Through this screening process, a set of target control devices that meet both the coverage location and operating time conditions is obtained, and their unique device identification information is extracted. Finally, corresponding control commands are generated based on the device identification. The commands clearly include elements such as device start time, continuous operating time, and operating mode to ensure that the devices can execute as needed after issuance. After generating the commands, the control commands are accurately issued to the target devices through the unified command and dispatch center, enabling them to automatically perform operations such as spraying, rinsing, purification, or isolation within the predetermined time period, achieving targeted intervention in high-risk exposure areas.

[0044] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0045] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0046] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0047] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0048] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0049] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0050] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0051] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0052] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0053] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An Internet of Things (IoT) ecological command and management system for human settlements, characterized in that it comprises: The method comprises the following steps: An acquisition module is used to acquire air dust detection data and meteorological sensing data of the industrial park, and generate air dust dynamic data flow; A dust distribution prediction module is used to perform time series modeling on the air dust dynamic data flow, calculate the air dust accumulation rate and diffusion gradient, and predict the air dust aggregation distribution; A crowd distribution prediction module is used to obtain the crowd travel rule of the industrial park residential community, and construct a crowd distribution prediction model based on historical mobile trajectory and time distribution; A risk assessment module is used to perform spatial superposition on the air dust prediction result and the crowd distribution prediction result, and generate an exposure risk matrix set of air dust; An Internet of Things management module is used to obtain the installation coordinates and action range of edge regulation equipment, map to the industrial park residential community plane coordinate system, and access a unified command and dispatch center; A regulation issuing module is used to obtain the regulation position and regulation period when any element in the risk matrix is identified and determined as high exposure risk, and issue a regulation instruction from the dispatch center. 2.The Internet of Things (IoT) -based ecological command and management system for human settlements according to claim 1, wherein, In the acquisition module, the air dust detection data and meteorological sensing data of the industrial park are acquired, and the air dust dynamic data flow is generated, which specifically comprises: According to the actual geographical range of the industrial park residential community and the industrial facility laying position, a unified plane coordinate system covering the whole community is established, and the space is divided into units according to the preset grid resolution; Based on the air dust detection equipment uniformly arranged in the industrial park residential community area, the concentration data of air dust in the air is collected; Meteorological sensing equipment is arranged in the vicinity of the air dust detection equipment, which collects wind speed, wind direction, humidity and temperature parameters in real time; The collection parameters of the air dust detection equipment and the meteorological sensing equipment are time-aligned, and combined into air dust dynamic data flow. 3.The Internet of Things (IoT) -based ecological command and management system for human settlements according to claim 1, wherein, In the dust distribution prediction module, the air dust dynamic data flow is time series modeled, the air dust accumulation rate and diffusion gradient are calculated, and the air dust aggregation distribution is predicted, which specifically comprises: Perform trend decomposition on the time series of air dust concentration, obtain the long-term trend item and periodic item based on moving average and period fitting, and extract the burst item through residual anomaly detection; A two-dimensional vector field is constructed based on wind speed and wind direction parameters, and the wind direction and wind speed of each meteorological monitoring point are mapped to the community space unit by using grid interpolation method to form a wind field vector model; The air dust settling rate is estimated combined with humidity and temperature parameters, and a three-dimensional prediction field of diffusion and settling is established; In the prediction field, the air dust concentration accumulation curve and residual time curve of each space unit are estimated, the concentration accumulation and residual time are weighted and operated as a prediction aggregation index to generate a prediction aggregation distribution map of air dust.

4. The Internet of Things (IoT) ecological command and management system according to claim 3, characterized in that, The three-dimensional prediction field of diffusion and settling specifically comprises: Stack a height layer in each space unit of the industrial park residential community plane coordinate system, correct the settling velocity corresponding to the air dust particle size according to the humidity and temperature parameters, and establish a settling profile in the vertical direction; The long-term trend item, the periodic item and the burst item obtained by the decomposition are coupled with a wind field vector model and a deposition profile as time boundary conditions, and evolution of air dust concentration in a three-dimensional grid is iteratively calculated according to time steps to establish a three-dimensional prediction field covering horizontal and vertical directions.

5. The Internet of Things (IoT) ecological command and management system according to claim 1, characterized in that, In the population distribution prediction module, the population distribution prediction model is constructed based on the historical moving track and time distribution of the community population, and the population distribution prediction model specifically comprises: The historical moving track data of the residents is collected, the stay area and stay duration of the residents in different time periods within a set monitoring period are extracted, and a resident travel rule data set is integrated; The residents are divided into sensitive populations based on the age structure of the residents, clustering operation is performed on the travel rule data set of the sensitive populations after the division, and a sensitive population travel mode set is generated; The population density distribution in different spatial units in the community is predicted in each time window based on the travel mode set, and a sensitive population distribution prediction map is established.

6. The Internet of Things (IoT) ecological command and management system according to claim 1, wherein, In the risk assessment module, the air dust prediction result and the population distribution prediction result are spatially superimposed to generate an air dust exposure risk matrix set, specifically comprising: The predicted air dust aggregation distribution map and the sensitive population distribution prediction map are superimposed in the industrial park residential community plane coordinate system according to a preset time segmentation window; For each superimposed and coincided spatial unit, the product of the predicted air dust aggregation index and the population density prediction value is calculated to obtain an initial exposure risk index; According to the region attribute of the superimposed and coincided spatial unit, a weighted operation is performed to obtain an exposure risk value, and the exposure risk value is integrated according to the spatial unit and the time sequence to form an exposure risk matrix set.

7. The Internet of Things (IoT) ecological command and management system according to claim 1, wherein, In the Internet of Things management module, the installation coordinates and the action range of the edge regulation equipment are obtained, mapped to the industrial park residential community plane coordinate system, and connected to the unified command and dispatch center, specifically comprising: The fixed installation coordinates, the action range and the operable period of the edge regulation equipment are obtained, the installation coordinates and the action range are converted into a spatial coverage polygon, and the spatial coverage polygon is compared with each spatial unit of the industrial park residential community plane coordinate system to screen the candidate regulation equipment corresponding to each spatial unit. The edge regulation equipment includes a spraying device, an air purification facility and a temporary isolation facility. The edge regulation equipment is networked according to the corresponding covered spatial unit, a mapping relationship between the device identifier and the spatial unit coverage is established, and the unified command and dispatch center is connected. In the regulation issuing module, when any element in the risk matrix is identified and determined as high exposure risk, the regulation position and the regulation period are obtained, and the regulation instruction is issued by the dispatch center, specifically comprising: 8.The Internet of Things (IoT) -based ecological command and management system for human settlements according to claim 1, wherein, In the exposure risk matrix set, the matrix elements whose exposure risk values reach the preset risk threshold are screened and marked as high exposure risk; The spatial unit and the time prediction window corresponding to all high exposure risk elements in the matrix set are obtained and marked as the regulation position and the regulation period; Based on the spatial unit to which the regulation position belongs, the candidate edge regulation equipment corresponding to the regulation position is extracted; The regulation period and the operable period of the candidate edge regulation equipment corresponding to the regulation position are compared, and the target regulation equipment is screened. ​ Obtaining a device identifier of a target regulation device and generating a control instruction, and issuing the control instruction to the target device.

Images