A multi-source sensor fusion indoor gaseous pollutant positioning method and system

Abstract

The application relates to the technical field of gas detection, and discloses an indoor gaseous pollutant positioning method and system based on multi-source sensor fusion, which comprises the following steps: arranging a multi-source sensor array, performing environment adaptation and error correction, synchronously collecting concentration data and environment influence parameters of indoor gaseous pollutants, obtaining a time-sequenced fusion data set of the indoor gaseous pollutants, dynamically associating and decoupling the concentration data and the environment influence parameters in the time-sequenced fusion data set to obtain a pollutant concentration feature data set of the indoor gaseous pollutants, obtaining a concentration spatial distribution gradient field of the indoor gaseous pollutants, obtaining a pollutant concentration field atlas of the indoor gaseous pollutants, and obtaining a precise positioning result of the indoor gaseous pollutants. Through multi-source sensor fusion and dynamic correction, the indoor gaseous pollutants can be precisely positioned with high accuracy and high reliability.

Description

Technical Field

[0001] This invention relates to the field of gas detection technology, and in particular to a method and system for locating indoor gaseous pollutants using multi-source sensor fusion. Background Technology

[0002] The presence of gaseous pollutants in the indoor environment directly affects the health of people within the space. Real-time monitoring and source location of gaseous pollutants provide a reliable basis for pollution control and environmental management, and are an important component of indoor environmental safety. Current indoor gaseous pollutant monitoring methods mostly employ a single type of sensor deployed in a dispersed manner. This sensor deployment lacks adaptability to the characteristics of indoor spatial distribution. Furthermore, the sensors are not adapted and adjusted for the actual indoor environment before being put into use, and the collected pollutant concentration data are easily affected by environmental factors, making it difficult to meet the accuracy and stability requirements for continuous monitoring.

[0003] Current data processing methods cannot effectively distinguish and decouple concentration data from environmental impact parameters. Interference factors such as ambient temperature, humidity, and airflow directly superimpose onto pollutant concentration data, causing the data to fail to accurately reflect the distribution of pollutants. Existing technologies struggle to extract spatial gradients and enhance features from concentration data based on multi-sensor spatial layouts. They also fail to combine sensor detection accuracy with appropriate weighting coefficients for spatial location, resulting in poor continuity and localized distortion in the constructed pollutant concentration distribution information. Current pollution source location methods rely solely on single concentration values ​​for location determination, without verifying the location results by considering concentration distribution characteristics and diffusion properties. This leads to biases and misjudgments in the location results, failing to provide reliable support for the precise treatment of indoor gaseous pollutants. Summary of the Invention

[0004] This invention provides a method and system for locating indoor gaseous pollutants using multi-source sensor fusion, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for locating indoor gaseous pollutants using multi-source sensor fusion, comprising: S1. Deploy a multi-source sensor array and perform environmental adaptation and error correction to obtain a multi-source sensing detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters; S2. Perform synchronization preprocessing on the concentration data of the indoor gaseous pollutants and environmental impact parameters to obtain a time-series fusion dataset of the indoor gaseous pollutants; S3. Based on the preset gas diffusion characteristic association matching rules, the concentration data and environmental impact parameters in the time-series fusion dataset are dynamically associated and decoupled to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants. S4. Based on the spatial layout of the multi-source sensing detection unit, perform spatial gradient extraction on the pollutant concentration feature dataset to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration. S5. Perform feature enhancement processing on the concentration spatial distribution gradient field, and generate a pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit. S6. Based on the pollutant concentration field map, the spatial location of the pollutant diffusion source is inverted to obtain the location result of the indoor gaseous pollutants. The validity of the location result is verified to obtain the accurate location result of the indoor gaseous pollutants.

[0006] In a preferred embodiment, the deployment of a multi-source sensor array and the subsequent environmental adaptation and error correction to obtain a multi-source sensing detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters, includes: Based on the indoor spatial distribution characteristics, multiple types of sensing and detection elements are deployed to form a multi-source sensor array, and the indoor deployment points and detection coverage of each sensing and detection element are determined. Collect basic environmental parameters of the indoor area to be detected, and adjust the multi-source sensor array according to the basic environmental parameters; The multi-source sensor array that has completed environmental adaptation adjustment is subjected to overall error correction to obtain a multi-source sensing and detection unit, and the synchronous acquisition triggering conditions of the multi-source sensing and detection unit are set. According to the synchronous acquisition triggering conditions, the concentration data of indoor gaseous pollutants and environmental impact parameters are synchronously acquired through the multi-source sensing detection unit.

[0007] In a preferred embodiment, the step of synchronizing the concentration data of the indoor gaseous pollutants with environmental impact parameters to obtain a time-series fused dataset of the indoor gaseous pollutants includes: The concentration data of indoor gaseous pollutants and environmental impact parameters are matched with timestamps. The concentration data and environmental impact parameters that have completed timestamp matching are subjected to abnormal data identification and format unification processing to obtain the indoor gaseous pollutant data to be fused. Based on the collection time sequence of the concentration data and the environmental impact parameters, the data to be fused is arranged to obtain the time-series fusion dataset of the indoor gaseous pollutants.

[0008] In a preferred embodiment, the dynamic correlation and decoupling of concentration data and environmental impact parameters in the time-series fusion dataset based on preset gas diffusion characteristic association matching rules yields the pollutant concentration feature dataset of the indoor gaseous pollutants, including: Based on the concentration data and environmental impact parameters corresponding to the same collection time in the time-series fusion dataset, a parameter association sample set for the indoor gaseous pollutants is constructed. Based on the preset gas diffusion characteristic correlation matching principle, a gas diffusion characteristic benchmark weight is generated, and the correlation degree analysis is performed on the parameter correlation samples to obtain the dynamic correlation degree between the concentration data and the environmental impact parameters. The decoupling weights of the time-series fusion dataset, the dynamic correlation degree, and the gas diffusion characteristic benchmark weights are decoupled to obtain the dynamic correlation decoupling weight coefficients of the indoor gaseous pollutants. The parameter association sample set is weighted and corrected based on the dynamic association decoupling weight coefficient to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants.

[0009] In a preferred embodiment, the formula for calculating the dynamic correlation decoupling weight coefficient is as follows: ; In the formula, These are the dynamic correlation decoupling weighting coefficients for the indoor gaseous pollutants. For the dynamic correlation degree, The environmental impact parameters for the current data collection point. These are preset standard environmental reference parameters. K represents the normal fluctuation range of the environmental impact parameters, and K represents the baseline weight of the gas diffusion characteristics.

[0010] In a preferred embodiment, the step of extracting the spatial gradient of the pollutant concentration feature dataset based on the spatially deployed locations of the multi-source sensing detection units to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration includes: Match the data identifiers in the pollutant concentration feature dataset with the spatial deployment locations of the multi-source sensing detection units to obtain the location-related data set of the indoor gaseous pollutants; Based on the location-related data set, the data correspondence between adjacent spatially deployed locations is determined, and a spatial neighborhood data group of indoor gaseous pollutants is generated. The concentration change trend of the spatial neighborhood data group is identified, the feature information representing the spatial change of concentration is screened out, the feature information is integrated and the spatial dimension is normalized and organized to obtain the spatial distribution gradient field of the indoor gaseous pollutants.

[0011] In a preferred embodiment, the step of performing feature enhancement processing on the concentration spatial distribution gradient field, and generating a pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit, includes: The concentration spatial distribution gradient field is subjected to feature enhancement processing to obtain the enhanced concentration spatial distribution gradient field of the indoor gaseous pollutants. Based on the detection accuracy attribute information of the multi-source sensing detection unit, a set of detection accuracy weights for the multi-source sensing detection unit is generated. Based on the spatial layout of the multi-source sensing detection units, the spatial influence range is divided, and a set of spatial position weights for the multi-source sensing detection units is generated. The detection accuracy weight set and the spatial location weight set are fused to obtain a fusion coefficient that adapts to the enhanced concentration spatial distribution gradient field; The enhanced concentration spatial distribution gradient field is weighted and integrated based on the fusion coefficient to obtain the pollutant concentration field spectrum of the indoor gaseous pollutants.

[0012] In a preferred embodiment, fusing the detection accuracy weight set and the spatial location weight set to obtain a fusion coefficient adapted to the enhanced concentration spatial distribution gradient field includes: The detection accuracy weight set is normalized and regularized to generate the standardized detection accuracy weights of the multi-source sensing detection unit. The spatial location weight set is adjusted for distribution consistency to generate the standardized spatial location weights of the multi-source sensing detection unit; The standardized detection accuracy weights are matched and associated with the standardized spatial position weights to obtain the weight correspondence of the multi-source sensing detection units; Based on the weight correspondence, the two types of standardized weights are combined and integrated to generate the initial fusion coefficients of the multi-source sensing detection unit; Based on the distribution characteristics of the enhanced concentration spatial distribution gradient field, the initial fusion coefficient is adapted and adjusted to obtain a fusion coefficient adapted to the enhanced concentration spatial distribution gradient field.

[0013] In a preferred embodiment, the step of inverting the spatial location of pollutant diffusion sources based on the pollutant concentration field map to obtain the location result of the indoor gaseous pollutants, and verifying the validity of the location result to obtain the accurate location result of the indoor gaseous pollutants, includes: Identify the extreme concentration distribution regions in the pollutant concentration field map to generate a spatial set of candidate pollution sources for the indoor gaseous pollutants; Based on the distribution characteristics of the spatial set of candidate pollution sources, the initial spatial location of the pollutant diffusion source is confirmed, and the initial location result of the indoor gaseous pollutant is obtained. Based on the concentration distribution characteristics and diffusion trend characteristics corresponding to the initial positioning results, a set of verification reference characteristics for the indoor gaseous pollutants is generated. Based on the comparison and analysis between the verification reference feature set and the preset pollutant diffusion characteristics, the feature matching results of the indoor gaseous pollutants are obtained. The initial positioning results are filtered and corrected based on the feature matching results to obtain the accurate positioning results of the indoor gaseous pollutants.

[0014] To address the aforementioned problems, this invention also provides an indoor gaseous pollutant localization system based on multi-source sensor fusion. The system includes a sensor array acquisition module, a data synchronization and preprocessing module, a multi-source data association and decoupling module, a spatial gradient field generation module, a concentration field map construction module, and a pollution source localization verification module, wherein: The sensor array acquisition module is used to deploy a multi-source sensor array and perform environmental adaptation and error correction to obtain a multi-source sensor detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters. The data synchronization preprocessing module is used to perform synchronization preprocessing on the concentration data of indoor gaseous pollutants and environmental impact parameters to obtain a time-series fusion dataset of indoor gaseous pollutants. The multi-source data association and decoupling module is used to dynamically associate and decouple the concentration data and environmental impact parameters in the time-series fusion dataset based on preset gas diffusion characteristic association matching rules, so as to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants. The spatial gradient field generation module is used to extract the spatial gradient of the pollutant concentration feature dataset based on the spatial layout of the multi-source sensing detection unit, so as to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration. The concentration field map construction module is used to perform feature enhancement processing on the concentration spatial distribution gradient field, and generate the pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit. The pollution source location verification module is used to perform spatial location inversion of pollutant diffusion sources based on the pollutant concentration field map, obtain the location result of the indoor gaseous pollutants, verify the validity of the location result, and obtain the accurate location result of the indoor gaseous pollutants.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention deploys a multi-source sensor array by combining indoor spatial distribution characteristics, and performs environmental adaptation adjustment and overall error correction on the array. The gaseous pollutant concentration data and environmental impact parameters collected synchronously have higher accuracy and stability. Through time-series synchronous preprocessing and dynamic correlation decoupling processing based on gas diffusion characteristics, data interference caused by environmental factors can be effectively removed, and a pollutant concentration feature dataset that can truly reflect the distribution state of pollutants can be obtained, thereby improving the basic reliability of subsequent processing and positioning from the data source.

[0016] 2. This invention relies on the spatial location of multi-source sensing units to complete the extraction of spatial gradients of concentration and feature enhancement. It combines detection accuracy weights and spatial location weights to generate a continuous and complete pollutant concentration field map. Based on the map, it performs spatial inversion of pollution sources and verifies and corrects the effectiveness of the location results. This can significantly improve the accuracy and reliability of locating indoor gaseous pollutant diffusion sources, forming a closed-loop technical solution from data acquisition, data processing, field map construction to location verification. It provides stable and reliable technical support for the accurate monitoring and efficient treatment of indoor gaseous pollutants. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for locating indoor gaseous pollutants using multi-source sensor fusion, provided in an embodiment of the present invention. Figure 2 This is a functional block diagram of an indoor gaseous pollutant localization system based on multi-source sensor fusion, provided in an embodiment of the present invention. The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a method for locating indoor gaseous pollutants using multi-source sensor fusion. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for locating indoor gaseous pollutants using multi-source sensor fusion can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a method for locating indoor gaseous pollutants using multi-source sensor fusion, according to an embodiment of the present invention. In this embodiment, the method for locating indoor gaseous pollutants using multi-source sensor fusion includes: S1. Deploy a multi-source sensor array and perform environmental adaptation and error correction to obtain a multi-source sensing detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters; The deployment of a multi-source sensor array, followed by environmental adaptation and error correction, results in a multi-source sensing detection unit that simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters, including: Based on the indoor spatial distribution characteristics, multiple types of sensing and detection elements are deployed to form a multi-source sensor array, and the indoor deployment points and detection coverage of each sensing and detection element are determined. Collect basic environmental parameters of the indoor area to be detected, and adjust the multi-source sensor array according to the basic environmental parameters; The multi-source sensor array that has completed environmental adaptation adjustment is subjected to overall error correction to obtain a multi-source sensing and detection unit, and the synchronous acquisition triggering conditions of the multi-source sensing and detection unit are set. According to the synchronous acquisition triggering conditions, the concentration data of indoor gaseous pollutants and environmental impact parameters are synchronously acquired through the multi-source sensing detection unit.

[0021] It needs to be specifically explained that the indoor spatial distribution characteristics are the overall spatial attributes constituted by the spatial dimensions, functional zoning, partition layout, and airflow direction of the indoor area to be detected. Based on these indoor spatial distribution characteristics, multiple types of sensing elements used to detect different gaseous pollutants and environmental conditions are combined and arranged to form a multi-source sensor array capable of multi-dimensional parallel acquisition. The fixed installation positions of each sensing element in the indoor space are determined according to the indoor spatial distribution characteristics, i.e., the indoor deployment points. The spatial area in which the sensing elements can effectively collect signals at each indoor deployment point is also defined, i.e., the detection coverage area, so that the multi-source sensor array can completely cover the indoor area to be detected where pollutant monitoring and location are required.

[0022] The multi-source sensor array contains various types of sensing elements used to simultaneously collect indoor gaseous pollutant concentration data and environmental impact parameters. These sensing elements are specifically divided into two categories: gaseous pollutant concentration sensing elements and environmental parameter sensing elements. Gaseous pollutant concentration sensing elements collect concentration information of corresponding pollutants in indoor air, specifically including volatile organic compound (VOC) sensors, formaldehyde sensors, benzene series compound (BCC) sensors, carbon monoxide sensors, and carbon dioxide sensors. Environmental parameter sensing elements collect environmental state information affecting the diffusion and detection accuracy of gaseous pollutants, specifically including temperature sensors, humidity sensors, air pressure sensors, indoor wind speed sensors, and air convection sensors. These sensing elements are arranged in combination according to the indoor spatial distribution characteristics. Different types of sensing elements share the same indoor deployment points, and the various sensing elements at the same deployment point form a parallel acquisition structure, collectively constituting a multi-source sensor array capable of simultaneously acquiring multi-dimensional detection information. This provides complete raw information support for subsequent data processing and pollution source location.

[0023] Environmental baseline parameters, or basic environmental parameters, are collected in the indoor area to be detected under conditions without target pollutants. Based on these basic environmental parameters, the multi-source sensor array is adjusted for environmental adaptation. Environmental adaptation adjustment includes operating status, response threshold, and acquisition mode, so that the multi-source sensor array can adapt to the current environmental operating conditions.

[0024] A unified calibration is performed on the multi-source sensor array that has completed environmental adaptation and adjustment to eliminate the overall error between each sensing element, including zero-point deviation and sensitivity deviation. This results in an integrated sensing structure, namely the multi-source sensing unit, which has stable and accurate acquisition capabilities after environmental adaptation and error correction. A unified command condition, namely the synchronous acquisition trigger condition, is set according to the data processing timing requirements to control all sensing elements to start acquisition simultaneously.

[0025] According to the synchronous acquisition trigger conditions, the multi-source sensor detection unit is controlled to start the acquisition work synchronously, to obtain concentration data that characterizes the content of gaseous pollutants, and to obtain environmental impact parameters that may affect the pollutant concentration detection results. The concentration data is the actual content information of indoor gaseous pollutants collected by the multi-source sensor detection unit, specifically including formaldehyde content, benzene series content, total volatile organic compound content, and other detection data that can directly reflect the level of pollutant concentration. The environmental impact parameters are environmental state information that is acquired synchronously during the acquisition process and may interfere with the pollutant concentration detection results, specifically including ambient temperature, ambient humidity, indoor airflow velocity, ambient air pressure, and other parameters that can affect gas diffusion and sensor response.

[0026] The beneficial effects are as follows: by combining the indoor spatial distribution characteristics to deploy multiple types of sensing elements to form a multi-source sensor array, the indoor deployment points and detection coverage can be completely matched with the spatial layout of the indoor area to be detected, achieving comprehensive monitoring without blind spots. Based on basic environmental parameters, environmental adaptation adjustment and overall error correction are performed on the multi-source sensor array, which can effectively reduce the impact of environmental interference and inherent component deviations on the acquisition results, significantly improving the acquisition accuracy and operational stability of the multi-source sensing unit. By setting synchronous acquisition trigger conditions, the synchronous acquisition of concentration data and environmental impact parameters can be achieved, ensuring that the multi-source data maintains a high degree of consistency in time sequence. This provides real, stable, and reliable raw data support for subsequent synchronization preprocessing, dynamic correlation decoupling, spatial gradient extraction, and accurate location of pollution sources, comprehensively improving the accuracy and reliability of the entire process of indoor gaseous pollutant monitoring and location.

[0027] S2. Perform synchronization preprocessing on the concentration data of the indoor gaseous pollutants and environmental impact parameters to obtain a time-series fusion dataset of the indoor gaseous pollutants; The process of synchronizing the concentration data of indoor gaseous pollutants with environmental impact parameters to obtain a time-series fused dataset of indoor gaseous pollutants includes: The concentration data of indoor gaseous pollutants and environmental impact parameters are matched with timestamps. The concentration data and environmental impact parameters that have completed timestamp matching are subjected to abnormal data identification and format unification processing to obtain the indoor gaseous pollutant data to be fused. Based on the collection time sequence of the concentration data and the environmental impact parameters, the data to be fused is arranged to obtain the time-series fusion dataset of the indoor gaseous pollutants.

[0028] Specifically, time-stamp matching is performed on indoor gaseous pollutant concentration data and environmental impact parameters to establish a correspondence between the two types of data at the same collection time. Time-stamp matching establishes a correspondence between concentration data and environmental impact parameters generated at the same collection time, ensuring consistency in the collection time of the two types of data. Anomaly identification involves checking the time-stamp matched concentration data and environmental impact parameters item by item, filtering out collected data that exceeds the reasonable range. Anomaly identification is performed on the time-stamp matched concentration data and environmental impact parameters to remove data that does not conform to the collection rules. Format unification processing adjusts the concentration data and environmental impact parameters to the same data storage format and display standard, eliminating differences in the output formats of different sensor detection elements. Format unification processing is performed on the concentration data and environmental impact parameters that have completed anomaly identification, unifying the storage format and display standard of the two types of data. The resulting data is the set of indoor gaseous pollutant data to be fused, which is the set of concentration data and environmental impact parameters with a correspondence and a unified format after time-stamp matching, anomaly identification, and format unification processing.

[0029] Based on the acquisition sequence of concentration data and environmental impact parameters, the acquisition sequence refers to the order in which the multi-source sensing detection units acquire data. This sequence is used to determine the order of data arrangement. The data to be fused are arranged in sequence to form a time-series fusion dataset of indoor gaseous pollutants. The time-series fusion dataset is a collection of fused data that is sequential, uniform in format, and complete after being arranged according to the acquisition sequence.

[0030] The beneficial effects are that by matching the timestamps of concentration data and environmental impact parameters, identifying abnormal data, and unifying the format, it is possible to ensure that the two types of data remain consistent in terms of collection time, data content, and storage format. The time-series fusion dataset formed by arranging the data according to the collection time sequence has continuous and stable time-series characteristics and a unified and standardized data format. This provides a clear time-series, standardized format, and reliable data foundation for subsequent dynamic correlation decoupling, spatial gradient extraction, and concentration field map construction, improving the smoothness and accuracy of subsequent data processing steps and ensuring the credibility of indoor gaseous pollutant location results.

[0031] S3. Based on the preset gas diffusion characteristic association matching rules, the concentration data and environmental impact parameters in the time-series fusion dataset are dynamically associated and decoupled to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants. The method, based on preset gas diffusion characteristic association matching rules, dynamically associates and decouples the concentration data and environmental impact parameters in the time-series fusion dataset to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants, including: Based on the concentration data and environmental impact parameters corresponding to the same collection time in the time-series fusion dataset, a parameter association sample set for the indoor gaseous pollutants is constructed. Based on the preset gas diffusion characteristic correlation matching principle, a gas diffusion characteristic benchmark weight is generated, and the correlation degree analysis is performed on the parameter correlation samples to obtain the dynamic correlation degree between the concentration data and the environmental impact parameters. The decoupling weights of the time-series fusion dataset, the dynamic correlation degree, and the gas diffusion characteristic benchmark weights are decoupled to obtain the dynamic correlation decoupling weight coefficients of the indoor gaseous pollutants. The parameter association sample set is weighted and corrected based on the dynamic association decoupling weight coefficient to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants.

[0032] The formula for calculating the dynamic correlation decoupling weight coefficient is as follows: ; In the formula, These are the dynamic correlation decoupling weighting coefficients for the indoor gaseous pollutants. For the dynamic correlation degree, The environmental impact parameters for the current data collection point. These are preset standard environmental reference parameters. K represents the normal fluctuation range of the environmental impact parameters, and K represents the baseline weight of the gas diffusion characteristics.

[0033] It should be specifically explained that, based on the concentration data and environmental impact parameters corresponding to the same collection time in the time-series fusion dataset, a parameter association sample set for indoor gaseous pollutants is constructed by binding them one-to-one according to the collection timestamp and the spatial deployment points of the sensors. The parameter association sample set uses a single sample as the basic unit. Each sample contains gaseous pollutant concentration data at the same time and location, as well as environmental impact parameters such as temperature, humidity, air pressure, and indoor wind speed. The samples are arranged in an orderly manner according to the collection time sequence and spatial location, providing a structured data foundation for subsequent correlation analysis and weight calculation.

[0034] For example, in an indoor living room area, the sensor at a certain collection time collects formaldehyde concentration data, real-time temperature, real-time humidity, and real-time indoor wind speed data. The above data are bound together to form a single parameter association sample. The parameter association samples of all collection times and all sensor locations are integrated to form a complete parameter association sample set.

[0035] Based on the preset gas diffusion characteristic correlation matching principle, the gas diffusion characteristic benchmark weights are generated according to the diffusion physics of the gaseous pollutants to be measured in the indoor environment. The gas diffusion characteristic correlation matching principle is that the degree of influence of environmental impact parameters on the diffusion rate and concentration distribution of gaseous pollutants is from strong to weak, namely indoor wind speed, temperature, humidity, and air pressure. Based on this, fixed gas diffusion characteristic benchmark weights are assigned to various environmental impact parameters.

[0036] The correlation analysis algorithm is used to calculate the correlation degree between concentration data and environmental impact parameters in the parameter correlation sample set, and the dynamic correlation degree between concentration data and environmental impact parameters is obtained. The dynamic correlation degree is used to characterize the degree of correlation between concentration data and environmental impact parameters under real-time acquisition conditions. The larger the value, the higher the correlation degree.

[0037] Calculating the dynamic correlation between concentration data and environmental impact parameters requires retrieving concentration data and environmental impact parameters from the same collection time and sensor location within the parameter correlation sample set. Using the concentration data as reference data and the environmental impact parameters as comparison data, the correlation between the changing trends of the reference and comparison data is determined sequentially. The numerical change directions of the concentration data and environmental impact parameters during the continuous collection period are compared one by one to determine the degree of consistency between their trends. Parameter combinations with a high degree of consistency are identified as having a high correlation, while those with a low degree of consistency are identified as having a low correlation. All parameter correlation samples are compared group by group to obtain the degree of correlation between the concentration data and each environmental impact parameter. The obtained correlation degrees are then organized according to a unified rule to form a dynamic correlation that can be directly used for subsequent weight calculations.

[0038] For example, for indoor formaldehyde gas, the baseline weights of gas diffusion characteristics are set according to diffusion characteristics, with the baseline weight corresponding to indoor wind speed being 0.5, temperature being 0.3, and humidity being 0.2. Through the grey relational analysis algorithm, the dynamic correlation degree between formaldehyde concentration data and indoor wind speed is calculated to be 0.82, the dynamic correlation degree with temperature is 0.65, and the dynamic correlation degree with humidity is 0.51.

[0039] The dynamic correlation decoupling weight coefficient of indoor gaseous pollutants is obtained by decoupling the time-series fusion dataset, dynamic correlation degree, and gas diffusion characteristic benchmark weight. The dynamic correlation decoupling weight coefficient is a dynamic weight calculated by combining dynamic correlation degree, real-time environmental parameters, benchmark parameters, and benchmark weight. It is used to remove environmental interference and extract true concentration characteristics. It reflects the correlation characteristics of real-time data and conforms to the physical laws of gas diffusion. It accurately removes environmental interference and is the core weight value used to correct concentration data. The preset standard environmental benchmark parameter is the standard environmental parameter value for gaseous pollutant concentration detection under laboratory conditions. It is the reference benchmark for calculating environmental parameter deviation. The normal fluctuation range of environmental impact parameters is the fluctuation range of environmental parameters in the indoor environment that will not significantly interfere with the detection of pollutant concentration. Exceeding this range is regarded as a strong interference factor. The deviation between the measured value of environmental parameters and the standard benchmark parameter is normalized and corrected by an exponential term. Real-time adaptation is achieved by combining dynamic correlation degree. The gas diffusion characteristic benchmark weight is superimposed to ensure that the calculation results conform to the physical laws of pollutant diffusion. Finally, a weight coefficient that can accurately remove environmental interference is obtained.

[0040] The concentration data in the parameter-related sample set is weighted and corrected based on the dynamic correlation decoupling weight coefficient. The weight coefficient is used as an adjustment factor to eliminate the concentration fluctuation error caused by environmental impact parameters and retain the true concentration characteristics generated by the release of gaseous pollutants themselves. The concentration data that have been weighted and corrected are then organized according to their original time sequence and spatial location to obtain the pollutant concentration characteristic dataset of indoor gaseous pollutants.

[0041] The weighted correction using weight coefficients as adjustment factors specifically requires extracting the original concentration data corresponding to a single sampling point and a single sampling time from the parameter-related sample set. The dynamic correlation decoupling weight coefficient corresponding to this sample is used as the basis for data adjustment. The dynamic correlation decoupling weight coefficient is numerically matched with the original concentration data, so that the weight coefficient acts on the corresponding concentration data. The original concentration data is adjusted in the same direction according to the magnitude of the weight coefficient. When the weight coefficient is greater than the set benchmark, the environmental interference component in the original concentration data is reduced. When the weight coefficient is less than the set benchmark, the effective concentration component in the original concentration data is increased. During the adjustment process, the trend of the concentration data itself is preserved, and only the fluctuations caused by temperature, humidity, air pressure, and indoor wind speed are removed. After the adjustment of a single data point is completed, the collection timestamp and spatial layout point information corresponding to the data are retained. All concentration data in the parameter-related sample set are adjusted item by item in turn. All the adjusted concentration data are rearranged and combined according to the original collection time sequence and spatial point distribution order, keeping the data structure and arrangement rules unchanged, forming a pollutant concentration feature dataset after removing environmental interference.

[0042] The pollutant concentration feature dataset includes the concentration feature values ​​of various target gaseous pollutants after environmental interference removal, the corresponding data acquisition time markers, the corresponding data acquisition spatial location markers, and decoupling correction records matching the concentration feature values. It also includes the acquisition timestamp information corresponding to each concentration feature value, sensor spatial deployment location information, and records of dynamic association decoupling weight coefficients. All data are arranged in an orderly manner according to the acquisition time sequence and spatial location distribution order, maintaining a consistent structural form with the parameter-related sample set. It can be directly used for subsequent spatial gradient extraction operations, providing real, stable, and environmentally interference-free concentration feature data support for the generation of the concentration spatial distribution gradient field.

[0043] For example, the original formaldehyde concentration data may be biased upwards due to indoor wind speed interference. After weighting correction by the calculated dynamic correlation decoupling weight coefficient, the interference component caused by wind speed is removed to obtain the concentration feature value that reflects the true release level of pollutants. After all samples are corrected, they are integrated to form a pollutant concentration feature dataset for subsequent spatial gradient extraction operations.

[0044] The beneficial effects are that by constructing a parameter-correlated sample set, the concentration data and environmental impact parameters are accurately matched in time and space. The baseline weight is set by combining the gas diffusion characteristics and the real-time dynamic correlation degree is calculated. The dynamic correlation decoupling weight coefficient is calculated by using an adaptive exponential formula to complete the weighted correction of the original concentration data. This can effectively remove the interference of environmental factors such as temperature, humidity and wind speed on the detection of gaseous pollutant concentration, accurately extract the true concentration characteristics of pollutants, and improve the accuracy and reliability of subsequent spatial gradient field construction and spatial positioning of pollution sources.

[0045] S4. Based on the spatial layout of the multi-source sensing detection unit, perform spatial gradient extraction on the pollutant concentration feature dataset to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration. The spatial gradient extraction of the pollutant concentration feature dataset based on the spatially deployed points of the multi-source sensing detection unit, to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration, includes: Match the data identifiers in the pollutant concentration feature dataset with the spatial deployment locations of the multi-source sensing detection units to obtain the location-related data set of the indoor gaseous pollutants; Based on the location-related data set, the data correspondence between adjacent spatially deployed locations is determined, and a spatial neighborhood data group of indoor gaseous pollutants is generated. The concentration change trend of the spatial neighborhood data group is identified, the feature information representing the spatial change of concentration is screened out, the feature information is integrated and the spatial dimension is normalized and organized to obtain the spatial distribution gradient field of the indoor gaseous pollutants.

[0046] It is important to explain that the data identifier carried by each data set in the pollutant concentration feature dataset is uniquely matched with the spatial deployment point corresponding to the multi-source sensing detection unit. This ensures that each pollutant concentration feature data can be assigned to its corresponding spatial acquisition location, forming a point-related data set of indoor gaseous pollutants. This point-related data set can establish a fixed binding relationship between pollutant concentration feature data and spatial location, ensuring the accuracy of the correspondence between data and location in the subsequent spatial gradient extraction process.

[0047] Based on the spatial distribution information of the points recorded in the point-related data set, the spatially adjacent points are classified and divided to clarify the attribution and correspondence of pollutant concentration characteristic data between adjacent spatial points, and to generate spatial neighborhood data groups of indoor gaseous pollutants. The spatial neighborhood data groups are combined with adjacent collection points as the unit, providing a continuous spatial data foundation for the identification of spatial variation trends of concentration.

[0048] The concentration characteristics of each pollutant in the spatial neighborhood data group are compared and judged group by group to determine the direction and magnitude of the increase or decrease of the concentration value with spatial location within the same spatial neighborhood, thus completing the concentration change trend identification. Information that can directly reflect the distribution difference and transmission direction of concentration in the indoor space is extracted from the concentration change trend identification results. All the extracted feature information representing the spatial change of concentration is integrated according to the planar and vertical distribution order of the indoor space. The integrated feature information is then coordinate-normalized and spatially aligned according to the actual layout of the indoor space, so that the feature information can completely reflect the continuous distribution state of pollutant concentration in the indoor space, thus obtaining the spatial distribution gradient field of indoor gaseous pollutant concentration.

[0049] For example, the concentration feature data corresponding to three adjacent sensor detection units in the living room area are divided into the same spatial neighborhood data group. By comparing the data within the group, the trend of concentration gradually increasing from south to north is identified. This trend is extracted as feature information, and combined with the actual spatial layout of the living room, the spatial dimension is regularized and organized to form the spatial distribution gradient field of concentration in the corresponding area.

[0050] The beneficial effects are as follows: by accurately matching pollutant concentration characteristic data with the spatial deployment points of multi-source sensing detection units, a point-related data set with one-to-one correspondence between data and spatial location is established, ensuring the uniqueness and accuracy of the spatial attribution of concentration data. Based on the point-related data set, a spatial neighborhood data group of adjacent points is constructed, which completely restores the continuous distribution state of pollutant concentration in indoor space and avoids the distortion of distribution characteristics caused by spatial data fragmentation. By identifying the concentration change trend of the spatial neighborhood data group, the direction and magnitude of concentration change at different spatial locations are effectively extracted. After spatial dimension normalization, a concentration spatial distribution gradient field that can completely reflect the spatial diffusion path and distribution differences of pollutants is formed. This process can accurately characterize the spatial distribution characteristics of indoor gaseous pollutants, providing stable and reliable spatial feature support for subsequent concentration field map generation and pollution source location inversion, and significantly improving the spatial rationality and reliability of pollutant location.

[0051] S5. Perform feature enhancement processing on the concentration spatial distribution gradient field, and generate a pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit. The step of performing feature enhancement processing on the spatial distribution gradient field of the concentration, and generating a pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing unit, includes: The concentration spatial distribution gradient field is subjected to feature enhancement processing to obtain the enhanced concentration spatial distribution gradient field of the indoor gaseous pollutants. Based on the detection accuracy attribute information of the multi-source sensing detection unit, a set of detection accuracy weights for the multi-source sensing detection unit is generated. Based on the spatial layout of the multi-source sensing detection units, the spatial influence range is divided, and a set of spatial position weights for the multi-source sensing detection units is generated. The detection accuracy weight set and the spatial location weight set are fused to obtain a fusion coefficient that adapts to the enhanced concentration spatial distribution gradient field; The enhanced concentration spatial distribution gradient field is weighted and integrated based on the fusion coefficient to obtain the pollutant concentration field spectrum of the indoor gaseous pollutants.

[0052] The process of fusing the detection accuracy weight set and the spatial location weight set to obtain a fusion coefficient that adapts to the enhanced concentration spatial distribution gradient field includes: The detection accuracy weight set is normalized and regularized to generate the standardized detection accuracy weights of the multi-source sensing detection unit. The spatial location weight set is adjusted for distribution consistency to generate the standardized spatial location weights of the multi-source sensing detection unit; The standardized detection accuracy weights are matched and associated with the standardized spatial position weights to obtain the weight correspondence of the multi-source sensing detection units; Based on the weight correspondence, the two types of standardized weights are combined and integrated to generate the initial fusion coefficients of the multi-source sensing detection unit; Based on the distribution characteristics of the enhanced concentration spatial distribution gradient field, the initial fusion coefficient is adapted and adjusted to obtain a fusion coefficient adapted to the enhanced concentration spatial distribution gradient field.

[0053] It needs to be specifically explained that the characteristic information representing the spatial change of pollutant concentration in the concentration spatial distribution gradient field is amplified and highlighted, and irrelevant interference information in the concentration spatial distribution gradient field is weakened to obtain the enhanced concentration spatial distribution gradient field of indoor gaseous pollutants. The detection accuracy attribute information carried by the multi-source sensing detection unit itself is retrieved, and the corresponding weight level is divided according to the detection reliability of each sensing detection unit to generate the detection accuracy weight set of the multi-source sensing detection unit.

[0054] Based on the indoor area boundary where the multi-source sensor detection units are located, the spatial influence range that each sensor detection unit can effectively cover is divided. According to the size and importance of the spatial influence range, corresponding weight levels are assigned to generate a set of spatial location weights for the multi-source sensor detection units.

[0055] The weight values ​​in the detection accuracy weight set are standardized and transformed to a uniform scale, so that all weights are within the same numerical range, generating standardized detection accuracy weights for multi-source sensor detection units. The distribution of the weight values ​​in the spatial location weight set is balanced and adjusted so that all weights conform to the uniform rules of indoor spatial distribution, generating standardized spatial location weights for multi-source sensor detection units. The standardized detection accuracy weights and standardized spatial location weights corresponding to the same multi-source sensor detection unit are bound one-to-one to obtain the weight correspondence of multi-source sensor detection units.

[0056] Based on the weight correspondence, the standardized detection accuracy weights and standardized spatial location weights of each group are merged to generate the initial fusion coefficients of the multi-source sensing detection unit. Combining the density and intensity of the concentration distribution gradient field after enhancement, the initial fusion coefficients are adaptively increased or decreased to make the adjusted weights match the current concentration distribution state, thus obtaining the fusion coefficients adapted to the enhanced concentration spatial distribution gradient field. According to the fusion coefficients, each spatial feature information in the enhanced concentration spatial distribution gradient field is weighted and integrated, so that feature information with different spatial locations and different accuracies is fused and presented according to the weight ratio, thus obtaining the pollutant concentration field spectrum of indoor gaseous pollutants.

[0057] For example, after feature enhancement processing of the concentration spatial distribution gradient field in the living room area, a set of detection accuracy weights is generated based on the detection accuracy level of the sensing units in the area, and a set of spatial location weights is generated based on the spatial range covered by each unit. After normalization and distribution consistency adjustment, weight matching and integration are completed. The initial fusion coefficient is adapted and adjusted in combination with the concentrated distribution characteristics of the gradient field in the area. The final fusion coefficient is used to weight and integrate the enhanced gradient field to form a pollutant concentration field map that can clearly reflect the distribution of pollutant concentration.

[0058] The beneficial effects are that by enhancing the features of the spatial distribution gradient field of concentration, the core variation characteristics of pollutant concentration in spatial distribution can be effectively highlighted, irrelevant interference information can be weakened, and the recognizability and completeness of concentration distribution characteristics can be improved. A set of detection accuracy weights is generated based on the detection accuracy attribute information of multi-source sensing units. Combined with the spatial layout points, a set of spatial position weights is generated to divide the spatial influence range, which can comprehensively reflect the detection reliability and spatial contribution of the sensing units. By normalizing and adjusting the distribution consistency of the two types of weights, the uniformity of weight scale and the balance of distribution state can be achieved, ensuring the rationality and stability of weight fusion. The standardized detection accuracy weights and standardized spatial position weights are matched, associated, and combined, and adapted to the distribution characteristics of the enhanced spatial distribution gradient field of concentration. The resulting fusion coefficient can accurately adapt to the current spatial distribution state. Using the fusion coefficient to weight and integrate the enhanced spatial distribution gradient field of concentration can form an intuitive, accurate, and clearly characteristic pollutant concentration field map, providing stable, reliable, and highly recognizable data support for subsequent pollution source spatial location inversion and location result verification, improving the accuracy and consistency of the overall location process.

[0059] S6. Based on the pollutant concentration field map, the spatial location of the pollutant diffusion source is inverted to obtain the location result of the indoor gaseous pollutants. The validity of the location result is verified to obtain the accurate location result of the indoor gaseous pollutants.

[0060] The spatial location inversion of pollutant diffusion sources based on the pollutant concentration field map is used to obtain the location results of indoor gaseous pollutants. The validity of these location results is then verified to obtain precise location results of the indoor gaseous pollutants, including: Identify the extreme concentration distribution regions in the pollutant concentration field map to generate a spatial set of candidate pollution sources for the indoor gaseous pollutants; Based on the distribution characteristics of the spatial set of candidate pollution sources, the initial spatial location of the pollutant diffusion source is confirmed, and the initial location result of the indoor gaseous pollutant is obtained. Based on the concentration distribution characteristics and diffusion trend characteristics corresponding to the initial positioning results, a set of verification reference characteristics for the indoor gaseous pollutants is generated. Based on the comparison and analysis between the verification reference feature set and the preset pollutant diffusion characteristics, the feature matching results of the indoor gaseous pollutants are obtained. The initial positioning results are filtered and corrected based on the feature matching results to obtain the accurate positioning results of the indoor gaseous pollutants.

[0061] It is necessary to explain in detail that the concentration distribution of all spatial locations in the pollutant concentration field map is traversed, and the spatial range with a concentration value higher than the surrounding area is marked as the concentration extreme distribution area. All concentration extreme distribution areas are summarized and integrated to generate a spatial set of candidate pollution sources for indoor gaseous pollutants. Based on the spatial concentration and continuous coverage of each area in the candidate pollution source spatial set, the area with the highest concentration value and the most concentrated spatial distribution is selected as the core location of the pollutant diffusion source, thus completing the initial spatial location confirmation of the pollutant diffusion source and obtaining the initial location result of indoor gaseous pollutants.

[0062] Extract the concentration distribution and diffusion trend from the core area to the periphery within the spatial range corresponding to the initial positioning results. Organize the above information into a set of verification reference features that can reflect the source and diffusion pattern of pollutants. Compare each feature in the set of verification reference features with the pre-set pollutant diffusion characteristics that conform to physical laws, and determine the degree of fit between the verification reference features and the preset diffusion characteristics to obtain the feature matching results of indoor gaseous pollutants.

[0063] Based on the feature matching results, initial positioning results that do not match the preset pollutant diffusion characteristics are eliminated, and positioning results with a high degree of fit are retained. The spatial position is then fine-tuned and calibrated to complete the screening and correction of the initial positioning results, thereby obtaining accurate positioning results for indoor gaseous pollutants.

[0064] For example, the kitchen area is identified as an area of ​​extreme concentration distribution in the indoor pollutant concentration field map and included in the candidate pollution source spatial set. Based on the spatial concentration distribution characteristics of this area, the core area of ​​the kitchen is determined as the initial location result. The concentration distribution and diffusion direction of this area, which gradually decrease from the inside to the outside, are extracted to form a set of verification reference features. This feature is compared with the characteristics of gaseous pollutants diffusing from the source to the outside to obtain a feature matching result with a high degree of fit. The initial location result is retained and the spatial position is finely adjusted to obtain the accurate location result of indoor gaseous pollutants.

[0065] The beneficial effects include: by accurately identifying the extreme concentration distribution areas in the pollutant concentration field map, the core spatial range of pollutant accumulation can be quickly located, forming a complete and reliable spatial set of candidate pollution sources, providing clear direction for the initial location of pollution sources. Initial spatial location confirmation based on the distribution characteristics of the candidate pollution source spatial set can efficiently obtain initial location results for indoor gaseous pollutants, ensuring the continuity and directionality of the location process. Extracting concentration distribution and diffusion trend characteristics from the initial location results and constructing a verification reference feature set can completely reconstruct the true state of pollutant release and diffusion. Comparing the verification reference feature set with preset pollutant diffusion characteristics can effectively determine the degree of fit between the location results and physical diffusion laws, eliminating location deviations caused by abnormal interference factors. Screening and correcting the initial location results based on feature matching results can eliminate unreasonable location information and optimize spatial location accuracy, ultimately obtaining stable and reliable accurate location results. This significantly improves the accuracy, rigor, and practicality of pollution source location, providing solid and reliable data support for the rapid treatment and precise management of indoor gaseous pollutants.

[0066] like Figure 2 The diagram shown is a functional block diagram of an indoor gaseous pollutant positioning system based on multi-source sensor fusion, provided in an embodiment of the present invention.

[0067] The multi-source sensor fusion indoor gaseous pollutant localization system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the multi-source sensor fusion indoor gaseous pollutant localization system 100 may include a sensor array acquisition module 101, a data synchronization and preprocessing module 102, a multi-source data association and decoupling module 103, a spatial gradient field generation module 104, a concentration field map construction module 105, and a pollution source localization verification module 106. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0068] In this embodiment, the functions of each module are as follows: The sensor array acquisition module 101 is used to deploy a multi-source sensor array and perform environmental adaptation and error correction to obtain a multi-source sensor detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters. The data synchronization preprocessing module 102 is used to perform synchronization preprocessing on the concentration data of indoor gaseous pollutants and environmental impact parameters to obtain a time-series fusion dataset of indoor gaseous pollutants. The multi-source data association and decoupling module 103 is used to dynamically associate and decouple the concentration data and environmental impact parameters in the time-series fusion dataset based on preset gas diffusion characteristic association matching rules, so as to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants. The spatial gradient field generation module 104 is used to extract the spatial gradient of the pollutant concentration feature dataset based on the spatial layout of the multi-source sensing detection unit, so as to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration. The concentration field map construction module 105 is used to perform feature enhancement processing on the concentration spatial distribution gradient field, and generate the pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit. The pollution source location verification module 106 is used to perform spatial location inversion of the pollutant diffusion source based on the pollutant concentration field map, obtain the location result of the indoor gaseous pollutants, verify the validity of the location result, and obtain the accurate location result of the indoor gaseous pollutants.

[0069] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0070] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0071] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0073] The embodiments of this application can acquire and process relevant data based on an artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for locating indoor gaseous pollutants using multi-source sensor fusion, characterized in that, The method includes: S1. Deploy a multi-source sensor array and perform environmental adaptation and error correction to obtain a multi-source sensing detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters; S2. Perform synchronization preprocessing on the concentration data of the indoor gaseous pollutants and environmental impact parameters to obtain a time-series fusion dataset of the indoor gaseous pollutants; S3. Based on the preset gas diffusion characteristic association matching rules, the concentration data and environmental impact parameters in the time-series fusion dataset are dynamically associated and decoupled to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants. S4. Based on the spatial layout of the multi-source sensing detection unit, perform spatial gradient extraction on the pollutant concentration feature dataset to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration. S5. Perform feature enhancement processing on the concentration spatial distribution gradient field, and generate a pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit. S6. Based on the pollutant concentration field map, the spatial location of the pollutant diffusion source is inverted to obtain the location result of the indoor gaseous pollutants. The validity of the location result is verified to obtain the accurate location result of the indoor gaseous pollutants.

2. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 1, characterized in that, The deployment of a multi-source sensor array, followed by environmental adaptation and error correction, results in a multi-source sensing detection unit that simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters, including: Based on the indoor spatial distribution characteristics, multiple types of sensing and detection elements are deployed to form a multi-source sensor array, and the indoor deployment points and detection coverage of each sensing and detection element are determined. Collect basic environmental parameters of the indoor area to be detected, and adjust the multi-source sensor array according to the basic environmental parameters; The multi-source sensor array that has completed environmental adaptation adjustment is subjected to overall error correction to obtain a multi-source sensing and detection unit, and the synchronous acquisition triggering conditions of the multi-source sensing and detection unit are set. According to the synchronous acquisition triggering conditions, the concentration data of indoor gaseous pollutants and environmental impact parameters are synchronously acquired through the multi-source sensing detection unit.

3. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 1, characterized in that, The process of synchronizing the concentration data of indoor gaseous pollutants with environmental impact parameters to obtain a time-series fused dataset of indoor gaseous pollutants includes: The concentration data of indoor gaseous pollutants and environmental impact parameters are matched with timestamps. The concentration data and environmental impact parameters that have completed timestamp matching are subjected to abnormal data identification and format unification processing to obtain the indoor gaseous pollutant data to be fused. Based on the collection time sequence of the concentration data and the environmental impact parameters, the data to be fused is arranged to obtain the time-series fusion dataset of the indoor gaseous pollutants.

4. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 3, characterized in that, The method, based on preset gas diffusion characteristic association matching rules, dynamically associates and decouples the concentration data and environmental impact parameters in the time-series fusion dataset to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants, including: Based on the concentration data and environmental impact parameters corresponding to the same collection time in the time-series fusion dataset, a parameter association sample set for the indoor gaseous pollutants is constructed. Based on the preset gas diffusion characteristic correlation matching principle, a gas diffusion characteristic benchmark weight is generated, and the correlation degree analysis is performed on the parameter correlation samples to obtain the dynamic correlation degree between the concentration data and the environmental impact parameters. The decoupling weights of the time-series fusion dataset, the dynamic correlation degree, and the gas diffusion characteristic benchmark weights are decoupled to obtain the dynamic correlation decoupling weight coefficients of the indoor gaseous pollutants. The parameter association sample set is weighted and corrected based on the dynamic association decoupling weight coefficient to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants.

5. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 4, characterized in that, The formula for calculating the dynamic correlation decoupling weight coefficient is as follows: ； In the formula, These are the dynamic correlation decoupling weighting coefficients for the indoor gaseous pollutants. For the dynamic correlation degree, The environmental impact parameters for the current data collection point. These are preset standard environmental reference parameters. K represents the normal fluctuation range of the environmental impact parameters, and K represents the baseline weight of the gas diffusion characteristics.

6. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 1, characterized in that, The spatial gradient extraction of the pollutant concentration feature dataset based on the spatially deployed points of the multi-source sensing detection unit, to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration, includes: Match the data identifiers in the pollutant concentration feature dataset with the spatial deployment locations of the multi-source sensing detection units to obtain the location-related data set of the indoor gaseous pollutants; Based on the location-related data set, the data correspondence between adjacent spatially deployed locations is determined, and a spatial neighborhood data group of indoor gaseous pollutants is generated. The concentration change trend of the spatial neighborhood data group is identified, the feature information representing the spatial change of concentration is screened out, the feature information is integrated and the spatial dimension is normalized and organized to obtain the spatial distribution gradient field of the indoor gaseous pollutants.

7. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 1, characterized in that, The step of performing feature enhancement processing on the spatial distribution gradient field of the concentration, and generating a pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing unit, includes: The concentration spatial distribution gradient field is subjected to feature enhancement processing to obtain the enhanced concentration spatial distribution gradient field of the indoor gaseous pollutants. Based on the detection accuracy attribute information of the multi-source sensing detection unit, a set of detection accuracy weights for the multi-source sensing detection unit is generated. Based on the spatial layout of the multi-source sensing detection units, the spatial influence range is divided, and a set of spatial position weights for the multi-source sensing detection units is generated. The detection accuracy weight set and the spatial location weight set are fused to obtain a fusion coefficient that adapts to the enhanced concentration spatial distribution gradient field; The enhanced concentration spatial distribution gradient field is weighted and integrated based on the fusion coefficient to obtain the pollutant concentration field spectrum of the indoor gaseous pollutants.

8. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 7, characterized in that, The process of fusing the detection accuracy weight set and the spatial location weight set to obtain a fusion coefficient that adapts to the enhanced concentration spatial distribution gradient field includes: The detection accuracy weight set is normalized and regularized to generate the standardized detection accuracy weights of the multi-source sensing detection unit. The spatial location weight set is adjusted for distribution consistency to generate the standardized spatial location weights of the multi-source sensing detection unit; The standardized detection accuracy weights are matched and associated with the standardized spatial position weights to obtain the weight correspondence of the multi-source sensing detection units; Based on the weight correspondence, the two types of standardized weights are combined and integrated to generate the initial fusion coefficients of the multi-source sensing detection unit; Based on the distribution characteristics of the enhanced concentration spatial distribution gradient field, the initial fusion coefficient is adapted and adjusted to obtain a fusion coefficient adapted to the enhanced concentration spatial distribution gradient field.

9. The method for locating indoor gaseous pollutants using multi-source sensor fusion as described in claim 1, characterized in that, The spatial location inversion of pollutant diffusion sources based on the pollutant concentration field map is used to obtain the location results of indoor gaseous pollutants. The validity of these location results is then verified to obtain precise location results of the indoor gaseous pollutants, including: Identify the extreme concentration distribution regions in the pollutant concentration field map to generate a spatial set of candidate pollution sources for the indoor gaseous pollutants; Based on the distribution characteristics of the spatial set of candidate pollution sources, the initial spatial location of the pollutant diffusion source is confirmed, and the initial location result of the indoor gaseous pollutant is obtained. Based on the concentration distribution characteristics and diffusion trend characteristics corresponding to the initial positioning results, a set of verification reference characteristics for the indoor gaseous pollutants is generated. Based on the comparison and analysis between the verification reference feature set and the preset pollutant diffusion characteristics, the feature matching results of the indoor gaseous pollutants are obtained. The initial positioning results are filtered and corrected based on the feature matching results to obtain the accurate positioning results of the indoor gaseous pollutants.

10. An indoor gaseous pollutant localization system based on multi-source sensor fusion, characterized in that, To implement the indoor gaseous pollutant localization method based on multi-source sensor fusion as described in claim 1, the system includes a sensor array acquisition module, a data synchronization and preprocessing module, a multi-source data association and decoupling module, a spatial gradient field generation module, a concentration field map construction module, and a pollution source localization verification module, wherein: The sensor array acquisition module is used to deploy a multi-source sensor array and perform environmental adaptation and error correction to obtain a multi-source sensor detection unit, which simultaneously collects indoor gaseous pollutant concentration data and environmental impact parameters. The data synchronization preprocessing module is used to perform synchronization preprocessing on the concentration data of indoor gaseous pollutants and environmental impact parameters to obtain a time-series fusion dataset of indoor gaseous pollutants. The multi-source data association and decoupling module is used to dynamically associate and decouple the concentration data and environmental impact parameters in the time-series fusion dataset based on preset gas diffusion characteristic association matching rules, so as to obtain the pollutant concentration feature dataset of the indoor gaseous pollutants. The spatial gradient field generation module is used to extract the spatial gradient of the pollutant concentration feature dataset based on the spatial layout of the multi-source sensing detection unit, so as to obtain the spatial distribution gradient field of the indoor gaseous pollutant concentration. The concentration field map construction module is used to perform feature enhancement processing on the concentration spatial distribution gradient field, and generate the pollutant concentration field map of the indoor gaseous pollutants based on the detection accuracy and spatial weight allocation fusion coefficient of the multi-source sensing detection unit. The pollution source location verification module is used to perform spatial location inversion of pollutant diffusion sources based on the pollutant concentration field map, obtain the location result of the indoor gaseous pollutants, verify the validity of the location result, and obtain the accurate location result of the indoor gaseous pollutants.

Images