Indoor air pollutant tracing algorithm

Abstract

The invention relates to an indoor air pollutant traceability algorithm, and belongs to the technical field of pollutant traceability. The method comprises the steps that a plurality of pollutant sensors are arranged indoors, and a control center is arranged; acquiring sensor detection data, and performing data cleaning to obtain preprocessed detection data; constructing a building three-dimensional model, and setting a corresponding sensor model in the building three-dimensional model; and simulating the release of the pollutants on the sampling points, simulating the propagation process of the pollutants, and comparing the simulation result with the preprocessed detection data to determine the source of the pollutants. According to the invention, by arranging a plurality of pollutant sensors, concentration detection can be carried out on indoor pollutants, a three-dimensional model of a building is imported into simulation software, a virtual pollutant source is arranged in the three-dimensional model of the building, analogue simulation is carried out, and the source of the pollutants is judged according to an analogue simulation result; and the positioning effect on the pollution source is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of pollutant source tracing, and in particular relates to an indoor air pollutant source tracing algorithm. Background Art

[0002] Indoor air pollutant source tracing involves scientifically identifying the specific sources and contribution ratios of harmful substances in indoor air, thereby providing a precise basis for pollution prevention and control. This process typically combines pollutant testing (such as formaldehyde, TVOC, and radon), building material or furniture release rate analysis, environmental sampling, and model calculations (such as volume loading rate and release contribution percentage). For example, graphene electrochemical sensors, isotope traceability technology, or patented methods (CN202311363062.4) are used to quantify the impact of different building materials on pollutant concentrations. Source tracing can identify the primary cause of pollution (such as decoration materials, chemicals, or biological sources), allowing targeted measures such as ventilation, treatment, or source replacement to effectively improve indoor air quality and protect human health.

[0003] Existing technologies are unable to trace the source of indoor air pollutants, are unable to accurately treat them, and it is difficult to eliminate the source of pollutants. Summary of the Invention

[0004] The purpose of the present invention is to provide an indoor air pollutant tracing algorithm, aiming to solve the problems that the existing technology is unable to trace the source of indoor air pollutants, cannot accurately treat them, and is difficult to eliminate the source of pollutants.

[0005] To solve the above problems in the prior art, the present invention provides an indoor air pollutant source tracing algorithm, comprising: Multiple pollutant sensors are installed indoors, and a control center is set up, which is electrically connected to all pollutant sensors; Acquire sensor detection data of each pollutant sensor, perform data cleaning on the sensor detection data, and obtain pre-processed detection data; Constructing a three-dimensional building model, configuring the three-dimensional building model, and setting corresponding sensor models in the three-dimensional building model; Set up multiple sampling points in the three-dimensional model of the building, simulate the release of pollutants at the sampling points, and simulate the propagation process of pollutants. Compare the simulation results with the pre-processing detection data to determine the source of the pollutants.

[0006] Preferably, the step of acquiring sensor detection data from each pollutant sensor, performing data cleaning on the sensor detection data, and obtaining pre-processed detection data specifically includes: Traverse each pollutant sensor according to a preset number sequence, and obtain sensor detection data within a corresponding time period when accessing a pollutant sensor; Segment the sensor detection data according to the preset time intervals, dividing it into data corresponding to multiple time periods; The numerical values in the sensor detection data are cleaned, abnormal values are eliminated and missing values are filled to obtain preprocessed detection data.

[0007] Preferably, the steps of constructing a three-dimensional building model, configuring the three-dimensional building model, and setting a corresponding sensor model in the three-dimensional building model include: Obtain a floor plan of a building, identify the walls and windows in the floor plan, and construct a corresponding 3D model of the building in 3D software; Obtaining the decoration configuration in the building, and generating corresponding interior configuration in the three-dimensional model of the building based on the decoration configuration; The setting position of each pollutant sensor is obtained, and a corresponding sensor model is generated at the corresponding position in the three-dimensional model of the building.

[0008] Preferably, the step of setting a plurality of sampling points in the three-dimensional model of the building, simulating the release of pollutants at the sampling points, simulating the propagation process of the pollutants, and comparing the simulation results with the pre-processed detection data to determine the source of the pollutants includes: Load the 3D building model, grid the internal space of the 3D building model, and determine the position coordinates of each grid; The 3D building model is loaded into the simulation engine, and the center of each grid is used as the source of pollutants to simulate the flow path of pollutants; The simulation results are collected, compared with the pre-processed detection data actually collected, and the source of the pollutants is determined based on the comparison results.

[0009] Preferably, the number of grids in the three-dimensional building model is determined according to the total volume of the three-dimensional building model. The larger the total volume of the three-dimensional building model is, the more grids there are.

[0010] Preferably, during the operation of the pollutant sensor, the concentration of the pollutants is monitored in real time, and when the concentration of any pollutant reaches a preset value, it is determined that the pollutant source should be traced.

[0011] Preferably, the pollutant sensor has a built-in wireless communication module.

[0012] Specifically, the simulation engine uses a computational fluid dynamics algorithm to simulate the flow path of pollutants, which specifically includes the following steps: Based on the grid structure of the building's three-dimensional model, the governing equations of fluid motion are established at each grid node; Setting boundary conditions based on the interior configuration generated in the 3D model of the building; The finite volume method is used to discretize the governing equations on the grid cells. By coupling the pressure-velocity field, a turbulence model is introduced to simulate the unsteady flow effect. The spatiotemporal distribution of pollutant concentrations for each grid is output through iterative calculation until the residual converges. The simulated concentration at the sensor model position is extracted, and the root mean square error is calculated with the pre-processed detection data. Based on the root mean square error, the current position is determined to be a valid candidate source.

[0013] Specifically, the three-dimensional building model supports dynamic update configuration, specifically including: The opening and closing status sensors installed on each door and window transmit data to the control center in real time via wireless communication modules. A lightweight 3D engine is used to update the door and window models in real time based on sensor data. When the door and window opening angle changes, the ventilation boundary conditions are synchronously updated. Local re-simulation domain calculations are performed based on the door and window positions.

[0014] Specifically, the real-time data pipeline of the wireless communication module adopts an event-driven architecture instead of a polling mechanism, and realizes millisecond-level push of sensor status through the MQTT protocol; the particle system of pollutant diffusion is dynamically colored based on CFD simulation results through a three-dimensional model dynamic rendering engine.

[0015] The present invention provides an indoor air pollutant source tracing algorithm. By setting up multiple pollutant sensors, it is possible to detect the concentration of indoor pollutants. The three-dimensional model of the building is imported into simulation software, and a virtual pollutant source is set up in the three-dimensional building model for simulation. The source of the pollutants is determined based on the simulation results, thereby greatly improving the effect of locating the pollution source. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0017] Figure 1 A flowchart of an indoor air pollutant source tracing algorithm provided by an embodiment of the present invention; Figure 2 A flowchart of the steps of obtaining sensor detection data of each pollutant sensor, performing data cleaning on the sensor detection data, and obtaining pre-processed detection data provided by an embodiment of the present invention; Figure 3 A flowchart of the steps of constructing a three-dimensional building model, configuring the three-dimensional building model, and setting corresponding sensor models in the three-dimensional building model provided by an embodiment of the present invention; Figure 4 A flowchart of the steps provided in an embodiment of the present invention for setting multiple sampling points in a three-dimensional building model, simulating the release of pollutants at the sampling points, simulating the propagation process of the pollutants, comparing the simulation results with pre-processed detection data, and determining the source of the pollutants. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] like Figure 1 FIG. 1 is a flow chart of an indoor air pollutant source tracing algorithm provided by an embodiment of the present invention. The indoor air pollutant source tracing algorithm includes: S100 , multiple pollutant sensors are set up indoors, and a control center is set up. The control center is electrically connected to all the pollutant sensors.

[0020] In this step, multiple pollutant sensors are set up indoors. The pollutant sensors monitor the same type of pollutants. All pollutant sensors are fixed at various locations indoors through fixed installation. In order to improve the monitoring accuracy, the minimum distance between the pollutant sensors is limited. For example, there must be at least one other pollutant sensor within 3 meters of any pollutant sensor. A wireless communication module is installed inside the pollutant sensor, and a control center is installed indoors. The pollutant sensor can be connected to the control center by wired means or by wireless communication.

[0021] S200 , acquiring sensor detection data of each pollutant sensor, performing data cleaning on the sensor detection data, and obtaining pre-processed detection data.

[0022] In this step, the sensor detection data of each pollutant sensor is obtained, and data collection is performed at preset time intervals to extract the sensor detection data detected by each pollutant sensor. The sensor detection data is transmitted in encrypted form. In order to ensure data quality, the sensor detection data needs to be cleaned to remove erroneous data in the original data and supplement the missing data to obtain preprocessed detection data.

[0023] S300: construct a three-dimensional building model, configure the three-dimensional building model, and set a corresponding sensor model in the three-dimensional building model.

[0024] In this embodiment, a three-dimensional model of the building of the same size is constructed in the three-dimensional software according to the actual size of the building. Similarly, the three-dimensional model of the building is configured, and a corresponding sensor model is constructed according to the actual size of the pollutant sensor. The sensor model is loaded at the corresponding position in the three-dimensional model of the building according to the actual position of the pollutant sensor.

[0025] S400, multiple sampling points are set in the three-dimensional model of the building, the release of pollutants is simulated at the sampling points, and the propagation process of the pollutants is simulated. The simulation results are compared with the pre-processed detection data to determine the source of the pollutants.

[0026] In this embodiment, multiple sampling points are set in the three-dimensional model of the building. When generating the sampling points, the three-dimensional model of the building is segmented by a grid and divided into multiple independent areas. A sampling point is set in each area. The sampling points are used as the source of pollutants. Simulation is performed in the simulation software. Pollution sources of different concentrations are set at the sampling points. Then, the concentration data of the pollutants at each sensor model are extracted to obtain simulation results. The simulation results are compared with the pre-processed detection data to determine the source of the pollutants.

[0027] like Figure 2 As shown, as a preferred embodiment of the present invention, the steps of obtaining sensor detection data from each pollutant sensor, cleaning the sensor detection data, and obtaining pre-processed detection data specifically include: S201 , traversing each pollutant sensor according to a preset numbering sequence, and acquiring sensor detection data within a corresponding time period from a pollutant sensor when accessing the sensor.

[0028] In this step, the pollutant sensors are numbered. When numbering, natural numbers are used to number the pollutant sensors to determine the specific number of each pollutant sensor, and the data detected by each pollutant sensor are extracted one by one according to the order of the numbers.

[0029] S202 , segmenting the sensor detection data according to preset time intervals, and dividing the data into data corresponding to multiple time periods.

[0030] In this step, the sensor detection data is segmented according to preset time intervals, such as each minute as a segment, and all detection data within one minute is taken as a whole, thereby obtaining data corresponding to multiple time periods.

[0031] S203 , performing data cleaning on the values in the sensor detection data, removing abnormal values and filling in missing values to obtain pre-processed detection data.

[0032] In this step, the values in the sensor detection data are cleaned. A systematic data cleaning process is performed on the sensor-collected detection data. First, outliers that fall outside a reasonable range or are clearly erroneous are identified and removed. Statistical methods (such as the 3σ principle and boxplots) or a sliding window approach are used to identify outliers. Missing values in the data are filled using linear interpolation, previous and next value filling, or model-based prediction methods based on the characteristics of the time series. Ultimately, complete, continuous, and physically consistent preprocessed detection data is obtained, providing a reliable data foundation for subsequent analysis and modeling.

[0033] like Figure 3 As shown, as a preferred embodiment of the present invention, the steps of constructing a three-dimensional building model, configuring the three-dimensional building model, and setting a corresponding sensor model in the three-dimensional building model include: S301, obtaining a floor plan of a building, identifying walls and windows in the floor plan, and constructing a corresponding three-dimensional model of the building in three-dimensional software.

[0034] In this step, the floor plan of the building is obtained, image recognition is performed on the floor plan of the building, the walls and the locations of the windows are extracted, the sizes of the windows are determined according to the window number information, a three-dimensional model is constructed based on the floor plan, and windows are added.

[0035] S302: Acquire the decoration configuration in the building, and generate a corresponding interior decoration configuration in the three-dimensional model of the building based on the decoration configuration.

[0036] In this step, the decoration configuration of the building is obtained. The decoration configuration includes models of indoor objects such as furniture and appliances, so as to generate the same or similar decoration models. Specifically, a model database is constructed, and the corresponding decoration models are retrieved from it according to the type and size of furniture and appliances, and loaded into the three-dimensional model of the building.

[0037] S303: Obtain the installation position of each pollutant sensor, and generate a corresponding sensor model at the corresponding position in the three-dimensional building model.

[0038] In this step, the location of each pollutant sensor is obtained. Similarly, the corresponding sensor model is retrieved from the model database based on the model of the pollutant sensor and loaded into the three-dimensional building model.

[0039] like Figure 4 As shown, as a preferred embodiment of the present invention, the steps of setting multiple sampling points in the three-dimensional building model, simulating the release of pollutants at the sampling points, simulating the propagation process of the pollutants, comparing the simulation results with the pre-processed detection data, and determining the source of the pollutants include: S401, loading a three-dimensional building model, performing grid processing on the interior space of the three-dimensional building model, and determining the position coordinates of each grid.

[0040] In this step, the three-dimensional model of the building is loaded and opened through the three-dimensional modeling software. The total volume of the three-dimensional model of the building is extracted. The preset volume-grid number mapping table is queried based on the total volume to determine the total number of grids. The three-dimensional model of the building is segmented to obtain multiple grid units, and each grid is characterized by the center coordinates of the grid unit.

[0041] S402, loading the three-dimensional building model into a simulation engine, taking the center of each grid as a pollutant source, and simulating the flow path of the pollutants.

[0042] In this step, the three-dimensional model of the building is loaded into the simulation engine. In order to determine the location of the pollution source, each grid is regarded as a possible location of a pollution source. Therefore, a pollution source is set at this location. Through simulation, the pollutants will flow in the three-dimensional model of the building, that is, the pollutant concentration at each location in the three-dimensional model of the building will change.

[0043] S403: Collect simulation results, compare the simulation results with pre-processed detection data actually collected, and determine the source of the pollutants based on the comparison results.

[0044] In this step, based on the position of the sensor model, the pollutant concentration at the sensor model at different times is extracted, and the simulation results are compared with the actual data detected by each sensor (preprocessed detection data). If the two match, it means that the corresponding sampling point is the source of the pollutant, achieving accurate positioning of the pollutant source.

[0045] In this embodiment, during the operation of the pollutant sensor, the concentration of the pollutants is monitored in real time. When the concentration of any pollutant reaches a preset value, it is determined that the pollutant source should be traced.

[0046] Specifically, the simulation engine uses a computational fluid dynamics algorithm to simulate the flow path of pollutants, which specifically includes the following steps: Based on the grid structure of the building's three-dimensional model, the governing equations of fluid motion are established at each grid node; Setting boundary conditions based on the interior configuration generated in the 3D model of the building; The finite volume method is used to discretize the governing equations on the grid cells. By coupling the pressure-velocity field, a turbulence model is introduced to simulate the unsteady flow effect. The spatiotemporal distribution of pollutant concentrations for each grid is output through iterative calculation until the residual converges. The simulated concentration at the sensor model position is extracted, and the root mean square error is calculated with the pre-processed detection data. Based on the root mean square error, the current position is determined to be a valid candidate source.

[0047] In this embodiment, the governing equations for fluid motion include the continuity equation for conservation of mass, and the momentum equation is implemented based on the Navier-Stokes equations: , Concentration diffusion is calculated based on the pollutant transport equation: , Where ρ is the air density, u is the velocity vector, p is the pressure, μ is the dynamic viscosity, g is the acceleration of gravity, C is the pollutant concentration, D is the diffusion coefficient, and S is the pollution source term; The governing equations are discretized on the grid cells using the finite volume method (FVM). The pressure-velocity field is coupled using the SIMPLE algorithm. The k−ϵ turbulence model is introduced to simulate unsteady flow effects. The spatiotemporal distribution of pollutant concentration C(x, y, z, t) for each grid cell is output through iterative calculations until the residual converges. Extract the simulated concentration Csim(t) at the sensor model location and calculate the root mean square error RMSE with the preprocessed detection data Creal(t): , When the RMSE is less than the threshold, the current pollution source location is determined to be a valid candidate source.

[0048] Specifically, the three-dimensional building model supports dynamic update configuration, specifically including: The opening and closing status sensors installed on each door and window transmit data to the control center in real time via wireless communication modules. A lightweight 3D engine is used to update the door and window models in real time based on sensor data. When the door and window opening angle changes, the ventilation boundary conditions are synchronously updated. Local re-simulation domain calculations are performed based on the door and window positions.

[0049] In this embodiment, the data format transmitted by the wireless communication module is {"sensor_id":"W001","angle": 45.2,"timestamp":1719504000} under json / / Angle unit: degree; Update the door and window model in real time based on THREE.js. The pseudo code is as follows: Door and window model rotation update: function updateWindowModel(window_id, angle) { const window_mesh = model.getMeshById(window_id); / / Get the target mesh from the model generated by weight 3 const hinge_axis = window_mesh.userData.hinge_axis; / / Hinge axis vector (predefined) window_mesh.setRotationFromAxisAngle(hinge_axis,THREE.MathUtils.degToRad(angle)); } When the door and window opening angles change, the ventilation boundary conditions are updated synchronously (such as the opening area Aopen = W × H × sinθ, W: width, H: height, θ: angle).

[0050] According to the position of doors and windows, an influence sphere is automatically generated, and the CFD simulation described in right 4 is performed only on the grids within this spherical area.

[0051] Specifically, the real-time data pipeline of the wireless communication module adopts an event-driven architecture instead of a polling mechanism, and realizes millisecond-level push of sensor status through the MQTT protocol; the particle system of pollutant diffusion is dynamically colored based on CFD simulation results through a three-dimensional model dynamic rendering engine.

[0052] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An indoor air pollutant tracing algorithm, characterized in that: include: S100: Install pollutant sensors indoors and configure a control center, wherein the control center is electrically connected to all pollutant sensors; S200: Acquire sensor detection data of each pollutant sensor, perform data cleaning on the sensor detection data, and obtain pre-processed detection data; S300: Constructing a three-dimensional building model, configuring the three-dimensional building model according to the entity data, and setting a corresponding sensor model in the three-dimensional building model; S400: Sampling points are set in the three-dimensional model of the building, the release of pollutants is simulated at the sampling points, and the propagation process of the pollutants is simulated, and the simulation results are compared with the pre-processed detection data to determine the source of the pollutants.

2. The indoor air pollutant source tracing algorithm according to claim 1, characterized in that: The calculation method of the pre-processing detection data specifically includes: Traverse each pollutant sensor according to a preset number sequence, and obtain sensor detection data within a corresponding time period when accessing a pollutant sensor; Segmenting the sensor detection data according to a preset time interval into data associated with a time period; The numerical values in the segmented sensor detection data are cleaned to remove abnormal values and fill in missing values to obtain preprocessed detection data.

3. The indoor air pollutant source tracing algorithm according to claim 1, characterized in that: The configuration method of the sensor model includes: Obtain a floor plan of a building, identify the walls and windows in the floor plan, and construct a corresponding 3D model of the building in 3D software; Obtaining the decoration configuration in the building, and generating corresponding interior configuration in the three-dimensional model of the building based on the decoration configuration; The setting position of each pollutant sensor is obtained, and a corresponding sensor model is generated at the corresponding position in the three-dimensional model of the building.

4. The indoor air pollutant source tracing algorithm according to claim 1, characterized in that: The step of determining the source of the pollutants in S400 specifically includes: Load the 3D model of the building, perform grid processing on the internal space of the 3D model of the building, and determine the position coordinates of each grid; The 3D building model is loaded into the simulation engine, and the center of each grid is used as the source of pollutants to simulate the flow path of pollutants; The simulation results are collected, compared with the pre-processed detection data actually collected, and the source of the pollutants is determined based on the comparison results.

5. The indoor air pollutant source tracing algorithm according to claim 4, characterized in that: The number of grids in the three-dimensional building model is determined according to the total volume of the three-dimensional building model, and the total volume of the three-dimensional building model is in direct proportion to the number of grids.

6. The indoor air pollutant source tracing algorithm according to claim 1, characterized in that: During the operation of the pollutant sensor, the concentration of pollutants is monitored in real time. When the concentration of any pollutant reaches the preset value, the source of the pollutant is determined and traced.

7. The indoor air pollutant source tracing algorithm according to claim 1, characterized in that: The pollutant sensor has a built-in wireless communication module.

8. The indoor air pollutant source tracing algorithm according to claim 4, characterized in that: The simulation engine uses a computational fluid dynamics algorithm to simulate the flow path of pollutants, specifically including the following steps: Based on the grid structure of the building's three-dimensional model, the governing equations of fluid motion are established at each grid node; Setting boundary conditions based on the interior configuration generated in the 3D model of the building; The finite volume method is used to discretize the governing equations on the grid cells. By coupling the pressure-velocity field, a turbulence model is introduced to simulate the unsteady flow effect. The spatiotemporal distribution of pollutant concentrations for each grid is output through iterative calculation until the residual converges. The simulated concentration at the sensor model position is extracted, and the root mean square error is calculated with the pre-processed detection data. Based on the root mean square error, the current position is determined to be a valid candidate source.

9. The indoor air pollutant source tracing algorithm according to claim 3, characterized in that: The three-dimensional building model supports dynamic update configuration, specifically including: The opening and closing status sensors installed on each door and window transmit data to the control center in real time via wireless communication modules. A lightweight 3D engine is used to update the door and window models in real time based on sensor data. When the door and window opening angle changes, the ventilation boundary conditions are synchronously updated. Local re-simulation domain calculations are performed based on the door and window positions.

10. The indoor air pollutant source tracing algorithm according to claim 7, characterized in that: The real-time data pipeline construction of the wireless communication module adopts an event-driven architecture instead of a polling mechanism, and realizes millisecond-level push of sensor status through the MQTT protocol; the particle system of pollutant diffusion is dynamically colored based on CFD simulation results through a three-dimensional model dynamic rendering engine.

Images