Building environment monitoring system based on internet of things

Abstract

This invention relates to the field of building environment, specifically disclosing an IoT-based building environment monitoring system. The system includes modules for data acquisition, constraint construction, data fusion, analysis model establishment, spatiotemporal variation verification and optimization, and environmental parameter evaluation. The data acquisition module establishes a three-dimensional coordinate system for the building and uses full-volume data. The constraint construction module preprocesses the data, calculates the spatial variation coefficient, temporal variation coefficient, and data fusion degree, and sets thresholds. The data fusion module extracts spatiotemporal features for multi-source fusion and outputs a standardized dataset. The analysis model establishment module outputs identification results through multi-dimensional analysis. The spatiotemporal variation verification and optimization module optimizes the identification results through multi-dimensional verification. The environmental parameter evaluation module calculates the deviation and comprehensive health index, and outputs tiered early warnings. This invention achieves three-dimensional dynamic monitoring, captures multi-parameter nonlinear coupling, and employs adaptive optimization strategies to improve evaluation accuracy.

Description

Technical Field

[0001] This invention relates to the field of building environment monitoring technology, and more specifically to a building environment monitoring system based on the Internet of Things. Background Technology

[0002] With the deepening application of IoT technology in the construction field, IoT-based building environment monitoring has become a key means to improve the quality of indoor and outdoor environments. It collects multi-dimensional parameters such as temperature, humidity, air quality, and sound environment by deploying sensor terminals, aiming to provide data support for the assessment of the health and comfort of the building environment.

[0003] Existing building environment monitoring systems typically follow a process of "data acquisition - feature analysis - quality assessment - feedback application," playing a certain role in the real-time monitoring and anomaly identification of basic environmental parameters. However, in practical applications, this method still has some areas for improvement. Existing monitoring methods often focus on a single environmental parameter independently, lacking collaborative analysis of multi-source parameters such as acoustic environment, air quality, and thermal environment, making it difficult to capture the nonlinear coupling relationships between different parameters. Meanwhile, the deployment of IoT monitoring terminals lacks standardized spatial adaptation rules, relying solely on area as a single criterion without considering building functional zoning, volume, ventilation characteristics, and national standards, resulting in insufficient spatial representativeness of the monitoring data. Furthermore, for multi-source parameters such as acoustic environment, air quality, and thermal environment, only independent statistical analysis is conducted, failing to quantify the spatiotemporal coupling characteristics and nonlinear correlations between parameters. This makes it impossible to capture the combined effects of multiple parameters on human health, resulting in monitoring results that cannot comprehensively reflect the actual condition of the building environment and cannot provide accurate scientific basis for environmental optimization. For example, focusing only on single indicators such as temperature or PM2.5 concentration makes it difficult to reveal the combined effects of simultaneous increases in noise and PM2.5 concentration on human health, resulting in monitoring results that cannot comprehensively reflect the actual condition of the building environment and cannot provide accurate scientific basis for environmental optimization.

[0004] Therefore, there is an urgent need for an Internet of Things-based building environment monitoring system that can achieve three-dimensional and dynamic monitoring of the building environment through multi-source data fusion and intelligent analysis. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, the present invention provides an Internet of Things-based building environment monitoring system, which solves the problems mentioned in the background art through the following solutions.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An Internet of Things (IoT) based building environment monitoring system includes:

[0008] Data acquisition module: A three-dimensional coordinate system is established based on the building structure. Monitoring points are selected within the coordinate system and IoT monitoring terminals are installed. The IoT monitoring terminals collect multidimensional indoor and outdoor geographic data and environmental parameters of the building to form a raw dataset of monitoring points with spatial location and time stamp.

[0009] Constraint Condition Construction Module: Receives the original dataset of monitoring points output by the data acquisition module, performs data preprocessing to obtain a valid dataset, completes quantitative calculation based on the valid dataset, sets constraint thresholds, and forms quantitative monitoring constraints containing the valid dataset, quantitative calculation results, and constraint thresholds, and outputs them.

[0010] Data fusion module: Receives the quantitative monitoring constraints output by the constraint construction module, and performs spatial feature extraction, temporal feature extraction and multi-source data spatiotemporal fusion processing on the effective dataset in sequence, using the constraint threshold in the quantitative monitoring constraints as the boundary, and outputs a standardized fusion dataset;

[0011] Analysis Model Building Module: Receives the standardized fusion dataset output by the data fusion module, performs multi-dimensional analysis on the standardized fusion dataset, and outputs the corresponding analysis results and risk identification results;

[0012] Spatiotemporal variation verification and optimization module: Receives the analysis results and risk identification results output by the analysis model building module, retrieves the effective dataset from the constraint construction module, performs multi-dimensional verification on the analysis results and risk identification results, dynamically adjusts the monitoring strategy based on the verification results, and sends the adjusted monitoring strategy back to the data acquisition module to complete the data iterative optimization process until the preset iteration termination condition is met, and outputs the optimal monitoring result;

[0013] Environmental parameter assessment module: Receives the optimal monitoring results output by the spatiotemporal variation verification and optimization module, calculates the deviation of environmental parameters and the comprehensive health index based on the optimal monitoring results, and outputs the building environment status assessment results and early warning signals.

[0014] Preferably, the three-dimensional coordinate system has the main entrance of the building as the origin O (0, 0, 0), the east direction as the x-axis, the north direction as the y-axis, and the vertical upward direction as the z-axis; the geographic data specifically includes: obtaining the three-dimensional coordinates (x, y, z) of each monitoring point inside and outside the building through positioning devices; the environmental parameters include outdoor parameters and indoor parameters, the outdoor parameters include 10 items, namely outdoor temperature, humidity, wind speed, wind direction, solar radiation, noise, PM2.5, PM10, air pressure, and illuminance; the indoor parameters include 9 items, namely indoor wind speed, wind direction, illuminance, air temperature, air humidity, air pressure, PM2.5, CO2 concentration, and oxygen content; and several IoT monitoring terminals are deployed inside and outside the building structure according to the hierarchical deployment rules.

[0015] Preferably, the hierarchical deployment rules are as follows: For enclosed indoor spaces: 1 IoT monitoring terminal is deployed for every single space with a building area ≤ 50㎡ and a floor height ≤ 3m; for every additional 30㎡ of building area or every additional 2m of floor height, 1 additional IoT monitoring terminal is deployed; For densely populated indoor areas: [The text abruptly ends here, likely due to an incomplete sentence or missing information.] One IoT monitoring terminal is installed in the space, with a horizontal spacing of ≤8m between IoT monitoring terminals; Outdoor monitoring area: divided into background monitoring points, building influence area and pedestrian influence area. Background monitoring points are installed in unobstructed areas ≥30m away from the building body, upwind of the prevailing wind direction, with 2 units installed; Building influence area is installed in the range of 10m-20m away from the building exterior wall, downwind of the prevailing wind direction, with 2-3 units evenly installed; Pedestrian influence area is installed in the range of 10m around the main entrance of the building and the entrance / exit of the underground parking garage, with 1-2 units installed; The spacing between outdoor IoT monitoring terminals is ≤15m; The installation height of all IoT monitoring terminals is uniformly 1.2m-1.5m, avoiding ventilation openings, heat sources, pollution sources and obstructions.

[0016] Preferably, the data preprocessing includes denoising and normalization. The denoising uses a 5-point window, second-order polynomial Savitzky-Golay filtering method to remove random noise from the data. The normalization uses a min-max normalization method to uniformly map all types of environmental parameters to the [0,1] interval, eliminating the dimensional differences between different parameters.

[0017] The quantitative calculation specifically involves calculating the spatial coefficient of variation, temporal coefficient of variation, and data fusion degree based on the effective dataset. The spatial coefficient of variation is used to quantify the uniformity of the distribution of environmental parameters within the building space at the same time, identify areas of abrupt parameter changes, and obtain a quantitative result of spatial distribution uniformity, reflecting the degree of difference in the distribution of environmental parameters within the building space. The temporal coefficient of variation is used to quantify the fluctuation amplitude of environmental parameters at a single monitoring point in the time dimension, identify periods of abnormal parameter fluctuations, and obtain a quantitative result of temporal fluctuation amplitude, reflecting the drastic change of environmental parameters in the time dimension. The data fusion degree calculation quantifies the spatiotemporal correlation of similar indoor and outdoor environmental parameters, reflecting the synchronicity of changes in the same parameter indoors and outdoors.

[0018] After completing all quantitative calculations, quantitative constraint thresholds are set for subsequent analysis processes, taking into account current national building environment standards and the building's functions.

[0019] Preferably, the spatial variation coefficient ,in, This represents the standard deviation of the same parameter value at all monitoring points at the same time. This represents the average value of the same parameter at all monitoring points at the same time. When the value is 1.5 times lower than the corresponding sensor's detection limit, Set to 0;

[0020] The coefficient of time variation ,in, The standard deviation of a single point parameter time series. This represents the mean of a single point parameter time series, when When the value is 1.5 times lower than the corresponding sensor's detection limit, Set to 0;

[0021] The degree of data fusion ,in, B represents the natural constant, and B represents the total number of pairs of parameters for the same physical quantity in indoor and outdoor environments. This represents the min-max normalized difference of the k-th pair of parameters at time t. This represents the global maximum value of the normalized difference of the k-th pair of parameters within the monitoring period. , These represent the timestamps for the collection of the k-th parameter group indoors and outdoors, respectively. Represents the time decay constant;

[0022] The constraint thresholds include the data fusion degree threshold, the spatial coefficient of variation threshold, and the temporal coefficient of variation threshold. The data fusion degree threshold is... >0.7, the threshold for spatial variation coefficient is >0.6, the threshold for the coefficient of variation over time is >0.5.

[0023] The preferred process for extracting a standardized fused dataset from a valid dataset is as follows:

[0024] The first step involves defining the criteria for determining adjacent monitoring points based on indoor and outdoor geographic data of the building. For indoor monitoring points of the same floor and type, those with a 3D spatial distance within a preset range are considered adjacent. Similarly, for outdoor monitoring points in the same area, those with a 3D spatial distance within a preset range are also considered adjacent. Then, for each monitoring point, the normalized difference value of its parameters with those of all adjacent points of the same type is calculated to form a spatial difference vector. This vector is then used to determine the spatial difference vector based on a spatial variation coefficient threshold. >0.6, filter out high-variability regions from the effective dataset where the spatial distribution difference exceeds the threshold, and remove low-value data with uniform spatial distribution and no significant changes;

[0025] The second step involves using a 30-minute fixed-duration sliding window to calculate the mean, standard deviation, and coefficient of variation of the single-parameter time series, based on the time coefficient of variation threshold. >0.5, in the high-variability area data selected in the first step, further filter out the high-variability period data with time fluctuation exceeding the threshold, and remove low-value data with stable time dimension;

[0026] The third step is to obtain the data integration degree. In the high spatiotemporal variability data selected in the second step, identify Strongly correlated parameter pairs with a value greater than 0.7 are used to form a standardized fusion dataset.

[0027] Preferably, the multi-dimensional analysis includes a spatiotemporal feature analysis model, a dynamic response model, and a chaotic correlation model.

[0028] The spatiotemporal feature analysis model is used to calculate spatial difference vectors and time series statistical characteristics, quantify and standardize the spatial heterogeneity and temporal fluctuation patterns of all parameters in the fusion dataset, and output a spatial heterogeneity distribution map and a temporal fluctuation trend map of the building-wide environmental parameters, along with a preliminary list of highly variable areas / time periods; the spatial difference vector... ,in, Indicates monitoring point The m-th parameter, The parameters represent adjacent points j; the statistical characteristics of the time series are calculated using the time variation coefficient.

[0029] The dynamic response model is used to calculate the dynamic response index. ,in, This represents the number of time series data points, which is determined by the monitoring duration. To fix the calculation step size; This represents the normalized rate of change of the outdoor drive parameters; Indicates the number of monitoring cycles. Outdoor wind speed after normalization at each sampling time; Indicates the number of monitoring cycles. Solar radiation after normalization at each sampling time; This is expressed as the normalized rate of change of the indoor response parameters; The table represents the number of monitoring periods. Indoor wind speed after normalization at each sampling time; Indicates the number of monitoring cycles. Indoor illuminance after normalization at each sampling time;

[0030] Sensitivity of indoor thermal environment to outdoor weather disturbances used to quantify the response of indoor thermal environment: Normal range: 0.3≤ ≤2.0 indicates that the indoor thermal environment exhibits reasonable response attenuation to outdoor meteorological disturbances, which is within the normal operating condition; low response is abnormal. <0.3 indicates that the indoor thermal environment does not change with outdoor weather conditions, which may be due to excessive airtightness of the building envelope, malfunction of the fresh air system, or over-adjustment of the air conditioning system; Highly sensitive anomaly: >2.0, Indoor thermal environment is excessively sensitive to changes in outdoor weather, corresponding to insufficient thermal insulation performance of the building envelope, air leakage through doors and windows, or malfunction of the air conditioning system; Reverse anomaly: =3, the indoor and outdoor environmental trends are completely opposite, which corresponds to abnormal heating of indoor equipment or failure of cold / heat source;

[0031] The weighted average of the normalized rates of change of the three core driving parameters—outdoor temperature, solar radiation, and wind speed—is used to obtain the comprehensive rate of change of the outdoor driving parameters. When the comprehensive rate of change of the outdoor driving parameters is <5% / 10min, and... When the denominator of the calculation formula is 0, the environment is considered to be in steady state, and the dynamic response index is determined. Marked as 1; when the comprehensive change rate of outdoor drive parameters is ≥5% / 10min, and When the denominator of the calculation formula is 0, it is determined that there is no response indoors, and the dynamic response index is... The value is marked as 0; when When the denominator of the calculation formula is negative, it is determined that the indoor and outdoor change trends are reversed, and the dynamic response index is affected. Marked as 3, included in the strongly abnormal operating condition;

[0032] The chaotic correlation model is used to quantify the nonlinear coupling strength among multiple parameters. It calculates the maximum cross-Lyapunov exponent through multivariable phase space reconstruction. ,in The time series step size is T, and the time series length is T. To determine the embedding dimension, a pseudo nearest neighbor method is used. ; The mutual information method is used to determine the delay time. t represents time; The distance between adjacent orbits of different parameter sequences in phase space; when When the value is greater than 0 and the dynamic response index is within the normal range, it is determined that there is a nonlinear coupling relationship between the corresponding parameters. The larger the value, the stronger the coupling; when When the dynamic response index is >0 and exceeds the normal range, it is determined to be a correlation driven by outdoor disturbances, not an inherent indoor risk, and is only recorded; when When ≤0, it is determined that there is no significant nonlinear coupling between the parameters.

[0033] Preferably, the multi-dimensional verification includes spatial heterogeneity verification, temporal periodicity verification, and comprehensive variation verification;

[0034] The spatial heterogeneity verification is based on the valid dataset and quantization results output by the constraint condition construction module. If... >0.6, marked as a region of high spatial variability;

[0035] The time-periodic verification: based on the valid dataset output by the constraint condition construction module and the quantization calculation results, if... >0.5 indicates a period of high temporal variability;

[0036] The comprehensive variation verification involves extracting the spatial and temporal variation coefficients from the data fusion module and calculating the spatiotemporal variation comprehensive index. ,in This indicates the weighting coefficient adjusted by the dynamic response index. When within the normal range, =1, the variation weight is fully included, and it is judged as indoor endogenous variation; when the dynamic response index When it exceeds the normal range, Table = 0.5, the variation weight is halved to reduce the impact of variation caused by outdoor disturbances and avoid false triggering; when the dynamic response index When marked as environmental steady state, Table = 0.2; α is the spatial variation weighting coefficient, and β is the time variation weighting coefficient; based on the natural variation characteristics of environmental parameters, they are divided into two categories: fast-changing parameters and slow-changing parameters. In steady state, α + β = 1, fast-changing parameters β = 0.7 and α = 0.3, and slow-changing parameters α = 0.6 and β = 0.4. In densely populated areas, α is increased by 0.2, and β is decreased by 0.2 accordingly. The densely populated areas refer to indoor public areas with a design personnel density ≥ 0.5 people / ㎡.

[0037] Preferably, the monitoring strategy is as follows:

[0038] When 0.4≤ When the value is ≤0.7, it is determined to be a medium variation region / time period. The fast variation parameter and the slow variation parameter adopt the conventional acquisition interval. The conventional acquisition interval is 10s for fast variation parameter and 60s for slow variation parameter.

[0039] when When the value is greater than 0.7, it is identified as a high-variability area / time period, triggering encrypted monitoring: the acquisition interval for fast-changing parameters is shortened to 5 seconds, and the acquisition interval for slow-changing parameters is shortened to 30 seconds;

[0040] when When the value is less than 0.4, it is determined to be a stable region / time period, and frequency reduction monitoring is implemented: the acquisition interval for fast-changing parameters is extended to 20s, and the acquisition interval for slow-changing parameters is extended to 120s.

[0041] Furthermore, when the dynamic response index is <0.3 or >2.0, regardless of the spatiotemporal variation comprehensive index... To determine whether the criteria are met, encrypted monitoring is triggered on the drive-response parameters of the paired area, the collection interval is shortened to half of the normal interval, and the area is simultaneously marked as a key traceability area;

[0042] When the dynamic response index is 3, the highest level of encrypted monitoring is triggered, and the collection interval of all parameters is shortened to 1 / 3 of the normal collection interval;

[0043] The iteration termination condition is: within 30 consecutive iteration cycles. and When the overall rate of change is less than 0.5%, the iteration is terminated and the optimal monitoring result is output.

[0044] Preferably, the deviation of the environmental parameters ,in, This represents the preprocessed real-time monitoring environmental parameter values. The values ​​represent the average parameters over a stable period of 7 consecutive days under normal operating conditions. , These are the upper and lower limits of the corresponding national standard values, respectively. For parameters without a lower limit requirement according to national standards, these are the upper and lower limits. Set to 0;

[0045] The comprehensive health index ,

[0046] in, The data integration degree is represented by ω1, ω2, ω3, ω4, and ω5, which represent weights. The weights are allocated as follows: ω1 + ω2 + ω3 + ω4 + ω5 = 1. The weights for indoor and outdoor correlation are ω1 = 0.15, spatiotemporal variation are ω2 = 0.25, nonlinear coupling risk is ω3 = 0.25, dynamic response performance is ω4 = 0.2, and parameter compliance is ω5 = 0.15.

[0047] The warning signal:

[0048] Level I: EHI ≥ 0.85, and It is within the normal range, with no coupling risk, and the environmental condition is stable, outputting a green normal signal;

[0049] Level II: 0.7 ≤ EHI < 0.85, and The system is within the normal range, with no significant coupling risk, and environmental fluctuations are controllable, resulting in a blue warning signal being output.

[0050] Level III: 0.5 ≤ EHI < 0.7, or If the value is outside the normal range, there is a risk of weak nonlinear coupling, or the deviation of a single parameter exceeds the standard, a yellow warning signal will be output, triggering an environmental operation and maintenance prompt.

[0051] Level IV: EHI < 0.5, or Reverse anomalies indicate a risk of strong nonlinear coupling or simultaneous exceedance of multiple parameters. A red warning signal is output, and the parameters exceeding the standard, coupling pairs, and the location and time period of the exceedance are pushed simultaneously. Based on three-dimensional coordinates and spatial difference vectors, the spatial area where the risk occurs is located, and the core coupling parameter pairs are identified.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] 1) This invention collects indoor and outdoor environmental parameters of buildings in multiple dimensions through IoT monitoring terminals and establishes a three-dimensional coordinate system for precise positioning, thereby obtaining a high-resolution dataset with spatiotemporal labels. This solves the problems of single parameters and ambiguous spatial positioning in traditional monitoring methods, and achieves the effect of capturing the spatiotemporal dynamic characteristics of the building environment more comprehensively.

[0054] 2) This invention constructs quantitative monitoring constraints based on spatial variation coefficient, temporal variation coefficient, and data fusion degree, and combines them with a multi-dimensional analysis model to obtain quantitative results on the uniformity of parameter spatial distribution, temporal fluctuation patterns, and synergy between indoor and outdoor parameters. This solves the problem that traditional methods are difficult to capture the nonlinear coupling relationship of multi-source parameters, and achieves a three-dimensional interpretation of the intelligent monitoring effect that combines the building environment with the human thermal comfort system.

[0055] 3) This invention integrates Savitzky-Golay filtering for noise reduction, global normalization, and chaotic correlation analysis by designing a data fusion and intelligent analysis process. This results in the chaotic characteristics of nonlinear coupling between parameters, which solves the problem that traditional linear analysis is difficult to reveal complex environmental mechanisms. It achieves a deep analytical effect from microscopic physiological responses to the coordinated changes of macroscopic environmental elements.

[0056] 4) This invention obtains adaptive and optimized monitoring results by dynamically adjusting the monitoring strategy through spatiotemporal variation verification and environmental parameter deviation assessment, combined with iteration termination conditions. This solves the problems of traditional monitoring being slow to respond to environmental variations and having a crude adjustment strategy, thereby improving monitoring efficiency and reducing equipment wear and material waste. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] As attached Figure 1 As shown, an Internet of Things (IoT) based building environment monitoring system includes:

[0060] Data acquisition module: A three-dimensional coordinate system is established based on the building structure. Monitoring points are selected within the coordinate system and IoT monitoring terminals are installed. The IoT monitoring terminals collect multi-dimensional indoor and outdoor geographic data and environmental parameters of the building to form the original dataset of the monitoring points.

[0061] Specifically, the core actions of this module are to establish a three-dimensional coordinate system for the entire building, standardize the deployment of IoT monitoring terminals, collect all indoor and outdoor geographic data and environmental parameters of the building, and output monitoring data with a unified time stamp. The specific implementation process is as follows:

[0062] First, a three-dimensional coordinate system covering the entire building area is established, with the center point of the main entrance of the target building as the reference origin, the east direction as the horizontal reference axis, the north direction as the vertical reference axis, and the vertical upward direction as the height reference axis.

[0063] Then, according to the functional attributes, space size, and human activity characteristics of different areas of the building, standardized terminal deployment is implemented: For independent and enclosed indoor spaces, the number of IoT monitoring terminals is matched according to the space area and floor height, and installed at the breathing zone height of people's daily activities, avoiding ventilation openings, heat-generating equipment, pollution sources, and obstructions; for densely populated indoor public areas, the number of IoT monitoring terminals is matched according to the space volume, ensuring that the horizontal spacing between IoT monitoring terminals meets the preset requirements and fully covers the range of human activities; for outdoor areas, IoT monitoring terminals are deployed according to the prevailing wind direction in unobstructed areas upwind of the building, in the surrounding areas of the building downwind, and in the surrounding areas of high-frequency entrances and exits, ensuring that the spacing of outdoor IoT monitoring terminals meets the requirements;

[0064] After the terminal is deployed, the three-dimensional coordinates of the indoor IoT monitoring terminal are obtained through the indoor high-precision positioning system, and the three-dimensional coordinates of the outdoor IoT monitoring terminal are obtained through the outdoor high-precision satellite positioning system, thus completing the spatial location calibration of all monitoring points;

[0065] Subsequently, the IoT monitoring terminal sets differentiated collection cycles according to the changing characteristics of environmental parameters, and cyclically collects all environmental parameters related to temperature, humidity, air quality, sound environment, light environment, and meteorology. It also adds a unified global standard time stamp to each set of collected raw data, forming a raw dataset of monitoring points with spatial location and time stamps. This dataset is then transmitted in real time to the basic data receiving and storage unit of the cloud control center via wired or wireless communication networks, and simultaneously output to subsequent modules.

[0066] In this embodiment, the following needs to be specifically explained:

[0067] The three-dimensional coordinate system has its origin O(0, 0, 0) at the main entrance of the building, with the x-axis pointing due east, the y-axis pointing due north, and the z-axis pointing vertically upwards. The geographic data specifically includes the three-dimensional coordinates (x, y, z) of various monitoring points inside and outside the building, obtained through positioning devices. The environmental parameters include outdoor and indoor parameters. The outdoor parameters include 10 items: outdoor temperature, humidity, wind speed, wind direction, solar radiation, noise, PM2.5, PM10, air pressure, and illuminance. The indoor parameters include 9 items: indoor wind speed, wind direction, illuminance, air temperature, air humidity, air pressure, PM2.5, CO2 concentration, and oxygen content.

[0068] Specifically, the deployment rules for monitoring points and IoT monitoring terminals are as follows:

[0069] For enclosed indoor spaces: if the building area of ​​a single space is ≤50㎡ and the floor height is ≤3m, one IoT monitoring terminal shall be installed; for every additional 30㎡ of building area or every additional 2m of floor height, one additional IoT monitoring terminal shall be installed; for densely populated indoor areas: one IoT monitoring terminal shall be installed for every 400㎡ of space volume, with a horizontal spacing of ≤8m between IoT monitoring terminals.

[0070] Outdoor monitoring area: divided into background monitoring points, building influence area, and pedestrian influence area. Background monitoring points are deployed in unobstructed areas ≥30m away from the building body, upwind of the prevailing wind direction, with 2 units deployed. Building influence area is deployed in a range of 10m-20m away from the building exterior wall, downwind of the prevailing wind direction, with 2-3 units evenly deployed. Pedestrian influence area is deployed in a range of 10m around the main entrance of the building and the entrance / exit of the underground parking garage, with 1-2 units deployed. The spacing between outdoor IoT monitoring terminals is ≤15m. The installation height of all IoT monitoring terminals is uniformly 1.2m-1.5m, avoiding ventilation openings, heat sources, pollution sources, and obstructions.

[0071] Specifically, the IoT monitoring terminal integrates the following high-precision sensors: temperature sensor (accuracy ±0.5℃), humidity sensor (accuracy ±2%RH), ultrasonic anemometer (wind speed accuracy ±0.1m / s, wind direction accuracy ±5°), solar radiation sensor (accuracy ±5W / m²), noise monitoring module (accuracy ±1dB), PM2.5 / PM10 laser sensor (detection limit 0.3μg / m³), barometric pressure sensor (accuracy ±0.5hPa), illuminance sensor (accuracy ±2lux), miniature anemometer (wind speed accuracy ±0.2m / s, wind direction accuracy ±8°), illuminance sensor (accuracy ±3lux), integrated temperature, humidity and barometric pressure module (temperature ±0.3℃, humidity ±1.5%RH, barometric pressure ±0.3hPa), PM2.5 / CO2 multi-gas sensor (PM2.5 accuracy ±5%FS, CO2 accuracy ±50ppm), and electrochemical oxygen sensor (accuracy ±0.5%VOL).

[0072] In this embodiment, conventionally, the data acquisition module synchronously collects parameters from all monitoring points and adds a timestamp (t) to each set of data using the UTC time format. The positioning device uses the UWB indoor positioning system to obtain the coordinates of each IoT monitoring terminal.

[0073] Constraint Condition Construction Module: Receives the original dataset of monitoring points output by the data acquisition module, performs data preprocessing to obtain a valid dataset, completes quantitative calculation based on the valid dataset, sets constraint thresholds, and forms quantitative monitoring constraints containing the valid dataset, quantitative calculation results, and constraint thresholds, and outputs them.

[0074] Specifically, the core action of this module is to extract effective datasets from the raw monitoring data, perform quantitative calculations of the core features of environmental parameters based on the effective datasets, set threshold standards for data correlation determination, form quantitative constraints for subsequent data processing and analysis, and output the results. The specific implementation process is as follows:

[0075] First, after receiving the raw dataset from the monitoring points, data preprocessing is performed: noise is removed and the dimensional differences of different parameters are eliminated. At the same time, abnormal data is identified and removed, including fault data that remains unchanged for multiple consecutive acquisition cycles, invalid data that exceeds the sensor's range, and low-amplitude data with values ​​below the sensor's effective detection limit, so as to avoid invalid data interfering with subsequent calculations.

[0076] After preprocessing, the original datasets from the monitoring points are transformed into valid datasets. Core feature quantification calculations are then performed on these valid datasets. Specifically, spatial variation coefficients, temporal variation coefficients, and data fusion degree calculations are performed based on the valid datasets. The spatial variation coefficient quantifies the uniformity of environmental parameters within the building space at the same time, identifies areas of abrupt parameter changes, and yields a quantified result of spatial distribution uniformity, reflecting the degree of difference in the distribution of environmental parameters within the building space. The temporal variation coefficient quantifies the fluctuation amplitude of environmental parameters at a single monitoring point over time, identifies periods of abnormal parameter fluctuations, and yields a quantified result of temporal fluctuation amplitude, reflecting the drastic changes in environmental parameters over time. Both spatial and temporal variation coefficients provide a quantitative basis for the dynamic adjustment of the monitoring strategy. The data fusion degree calculation quantifies the spatiotemporal correlation between similar indoor and outdoor environmental parameters, reflecting the synchronicity of changes in the same parameter both indoors and outdoors. After completing all quantitative calculations, based on current national building environment standards and building functions, quantitative constraint thresholds are set for subsequent analysis processes, including anomaly thresholds for spatial distribution uniformity, anomaly thresholds for temporal fluctuation amplitude, and thresholds for determining the correlation between indoor and outdoor parameters. Finally, the preprocessed effective dataset, core feature quantification results, and quantitative constraint thresholds are integrated.

[0077] In this embodiment, it should be specifically noted that the data preprocessing includes denoising and normalization. The denoising uses a 5-point window, second-order polynomial Savitzky-Golay filtering method to remove random noise from the original data. The normalization uses a min-max normalization method to uniformly map all types of environmental parameters to the [0,1] interval, eliminating the dimensional differences between different parameters.

[0078] The quantification calculation is as follows:

[0079] The spatial variability coefficient ,in, This represents the standard deviation of the same parameter value at all monitoring points at the same time. This represents the average value of the same parameter at all monitoring points at the same time. When the value is 1.5 times lower than the corresponding sensor's detection limit, Set to 0;

[0080] The coefficient of time variation ,in, The standard deviation of a single point parameter time series. This represents the mean of a single point parameter time series, when When the value is 1.5 times lower than the corresponding sensor's detection limit, Set to 0;

[0081] The degree of data fusion Where B represents the total number of pairs of indoor and outdoor physical quantity parameters, and the paired parameters include: outdoor temperature-indoor temperature, outdoor humidity-indoor humidity, outdoor PM2.5-indoor PM2.5, outdoor PM10-indoor PM10, outdoor air pressure-indoor air pressure, outdoor illuminance-indoor illuminance, and outdoor noise-indoor noise, for a total of 7 pairs. This represents the min-max normalized difference of the k-th pair of parameters at time t. This represents the global maximum value of the normalized difference of the k-th pair of parameters within the monitoring period. Represents the natural constant. , These represent the timestamps for the collection of the k-th parameter group indoors and outdoors, respectively. Represents the time decay constant;

[0082] The constraint thresholds include the data fusion degree threshold, the spatial coefficient of variation threshold, and the temporal coefficient of variation threshold. The data fusion degree threshold is... >0.7, the threshold for spatial variation coefficient is >0.6, the threshold for the coefficient of variation over time is >0.5.

[0083] Data fusion module: Receives the quantitative monitoring constraints output by the constraint construction module, and uses the constraint thresholds in the quantitative monitoring constraints as boundaries to sequentially perform spatial feature extraction, temporal feature extraction, and multi-source data spatiotemporal fusion processing on the effective dataset, outputting a standardized fusion dataset.

[0084] The core function of this module is to extract spatial and temporal features of environmental parameters (i.e., spatiotemporal feature extraction), and to perform spatiotemporal fusion processing of multi-source data based on quantization constraints, resulting in a standardized fusion dataset output. Specifically:

[0085] First, spatial feature extraction is performed: the criteria for determining adjacent monitoring points are defined: for indoor monitoring points of the same type on the same floor, those with a three-dimensional spatial distance within a preset range are determined to be adjacent monitoring points; for outdoor monitoring points in the same area, those with a three-dimensional spatial distance within a preset range are determined to be adjacent monitoring points; then, for each monitoring point, the normalized difference value of the same type of parameters between the point and all adjacent monitoring points is calculated to form a spatial difference vector reflecting the degree of change in local environmental parameters, thus completing the spatial feature extraction.

[0086] Then, perform time feature extraction: using a fixed-duration sliding time window, with a single acquisition cycle as the sliding step, calculate the statistical characteristics of a single environmental parameter at a single monitoring point within the time window, including the average value, dispersion, and fluctuation amplitude within the window, to form complete time series statistical characteristics and complete time feature extraction.

[0087] Subsequently, based on the data fusion degree threshold, multi-source spatiotemporal fusion processing is performed on the datasets with completed spatial and temporal feature extraction. This identifies strongly correlated parameter pairs among similar indoor and outdoor parameters whose spatiotemporal correlation exceeds a preset threshold, while invalid data pairs with extremely low correlation are removed. The fusion results of spatial features, temporal features, and strongly correlated parameter pairs are integrated into a standardized fusion dataset, ensuring that the dataset retains only valid and highly correlated environmental feature data, which is then output to the subsequent analysis model building module.

[0088] In this embodiment, the following needs to be specifically explained:

[0089] Definition of adjacent monitoring points: The three-dimensional Euclidean distance between indoor monitoring points of the same type and on the same floor is ≤10m, and the three-dimensional Euclidean distance between outdoor monitoring points in the same area is ≤15m.

[0090] Specifically, the process of extracting a standardized fused dataset from a valid dataset:

[0091] The first step is to calculate the normalized difference value of each monitoring point with all neighboring points of the same type of parameter, forming a spatial difference vector, based on the spatial coefficient of variation threshold. >0.6, filter out high-variability regions from the effective dataset where spatial distribution differences exceed the threshold, and remove low-value data with uniform spatial distribution and no significant changes;

[0092] The second step involves using a 30-minute fixed-duration sliding window to calculate the mean, standard deviation, and coefficient of variation of the single-parameter time series, based on the time coefficient of variation threshold. >0.5, in the high-variability area data selected in the first step, further filter out the high-variability period data with time fluctuation exceeding the threshold, and remove low-value data with stable time dimension;

[0093] The third step is to obtain the data integration degree. In the high spatiotemporal variability data selected in the second step, identify Strongly correlated parameter pairs with a correlation coefficient greater than 0.7 are identified. Invalid data pairs with extremely low correlation coefficients are removed. After three layers of screening, spatial features, temporal features and strongly correlated parameters are finally integrated to form a standardized fusion dataset.

[0094] Analysis Model Building Module: Receives the standardized fusion dataset output by the data fusion module, performs multi-dimensional analysis on the standardized fusion dataset, and outputs the corresponding analysis results and risk identification results.

[0095] Specifically, the implementation process of this module is as follows:

[0096] First, after receiving the standardized fusion dataset, the analysis model building module constructs a spatiotemporal feature analysis model, a dynamic response model, and a chaotic correlation model. These three computational models are then used to perform multi-dimensional analysis of the standardized dataset.

[0097] First, the spatiotemporal feature analysis model calculates spatial difference vectors and time series statistical characteristics based on standardized fusion datasets, quantifies the spatial heterogeneity and temporal fluctuation patterns of all parameters in the standardized fusion datasets, and outputs spatial heterogeneity distribution maps and temporal fluctuation trend maps of building-wide environmental parameters, as well as a preliminary list of highly variable areas / time periods, providing benchmark data for subsequent verification and tracing.

[0098] Second, the dynamic response model calculates the response sensitivity index of the indoor thermal environment to changes in outdoor meteorological conditions, clarifies whether the changes in the indoor environment are caused by outdoor meteorological disturbances or by factors inside the building, and divides the normal and abnormal ranges of response sensitivity. For special working conditions with a denominator of zero, steady-state and no-response classifications are made according to the magnitude of changes in outdoor meteorological conditions. For reverse change working conditions with negative calculation results, strong anomaly markings are made separately.

[0099] Third, the chaotic correlation model uses a multivariable phase space reconstruction method to reconstruct the nonlinear dynamic phase space of multiple parameters and calculate the maximum cross-track divergence rate index between different parameters to quantify the nonlinear coupling strength between multiple parameters.

[0100] After completing the calculation of all key parameters, dynamic monitoring and risk identification of the building environment are performed: First, based on the calculation results of the dynamic response model, the coupling strength calculation results are pre-screened. Only when the response sensitivity is within the normal range and the track divergence rate is greater than zero, it is determined that there is an inherent nonlinear coupling risk between the parameters. The larger the track divergence rate, the higher the coupling risk. When the response sensitivity exceeds the normal range and the track divergence rate is greater than zero, it is determined that the parameter association is driven by outdoor meteorological disturbances and does not belong to the inherent indoor risk. It is only recorded and no warning is triggered. When the track divergence rate is less than or equal to zero, it is determined that there is no significant nonlinear coupling relationship between the parameters. Finally, the complete environmental comprehensive evaluation model, the basic analysis data obtained from the model calculation, and the results of dynamic monitoring and risk identification of the building environment are simultaneously output to the subsequent spatiotemporal variation verification and optimization module.

[0101] In this embodiment, the following needs to be specifically explained:

[0102] The spatiotemporal feature analysis model is used to calculate the spatial difference vector and time series statistical features, wherein the spatial difference vector... ,in, This represents the m-th parameter of monitoring point i. The parameters represent adjacent point j; the time series statistical features are used to calculate the time variation coefficient. A spatial heterogeneity distribution map of the building-wide environmental parameters is obtained based on the spatial difference vector, clearly indicating the degree of parameter distribution differences in different areas; a time fluctuation trend map of the environmental parameters at each monitoring point is obtained based on the time series statistical features, clearly indicating the degree of parameter change drasticness in different time periods. Monitoring points with a value >0.6 were marked as areas of high spatial variability. A time marker >0.5 indicates a period of high temporal variability.

[0103] The dynamic response model is used to calculate the dynamic response index. Where K represents the number of time series data points, which is determined by the monitoring duration; To fix the calculation step size; This represents the normalized rate of change of the outdoor drive parameters; Indicates the number of monitoring cycles. Outdoor wind speed after normalization at each sampling time; Indicates the number of monitoring cycles. Solar radiation after normalization at each sampling time; This is expressed as the normalized rate of change of the indoor response parameters; Indicates the number of monitoring cycles. Indoor wind speed after normalization at each sampling time; Indicates the number of monitoring cycles. Indoor illuminance after normalization at each sampling time; Used to quantify the sensitivity of indoor thermal environment to outdoor weather disturbances:

[0104] Normal range: 0.3≤ ≤2.0 indicates that the indoor thermal environment exhibits a reasonable response attenuation to outdoor meteorological disturbances, which is within the normal operating condition.

[0105] Low response anomaly: <0.3 indicates that the indoor thermal environment does not change with the outdoor weather, which corresponds to the building envelope being too airtight, the fresh air system being faulty, or the air conditioning system being over-adjusted.

[0106] Highly sensitive abnormality: >2.0, the indoor thermal environment is excessively sensitive to changes in outdoor weather, which corresponds to insufficient thermal insulation performance of the building envelope, air leakage through doors and windows, or failure of the air conditioning system.

[0107] Reverse anomaly: =3, the indoor and outdoor environmental trends are completely opposite, which corresponds to abnormal heating of indoor equipment or failure of cold / heat source.

[0108] The weighted average of the normalized rates of change of the three core driving parameters—outdoor temperature, solar radiation, and wind speed—is used to quantify the overall drastic change in outdoor meteorological conditions. Specifically: Overall rate of change of outdoor driving parameters = (normalized rate of change of outdoor temperature × 0.6) + (normalized rate of change of solar radiation × 0.25) + (normalized rate of change of outdoor wind speed × 0.15). When the overall rate of change of outdoor driving parameters < 5% / 10min, and... When the denominator of the calculation formula is 0, the environment is considered to be in steady state, and the dynamic response index is determined. Marked as 1; when the comprehensive change rate of outdoor drive parameters is ≥5% / 10min, and When the denominator of the calculation formula is 0, it is determined that there is no response indoors, and the dynamic response index is... The value is marked as 0; when When the denominator of the calculation formula is negative, it is determined that the indoor and outdoor change trends are reversed, and the dynamic response index is affected. Marked as 3, it is included in the strong abnormal operating condition.

[0109] The chaotic correlation model is used to quantify the nonlinear coupling strength among multiple parameters, and calculates the maximum Lyapunov exponent through multivariable phase space reconstruction. ,in d is the time series step size, T is the time series length; d is the embedding dimension, which is determined using the pseudo nearest neighbor method; τ is the delay time, which is determined using the mutual information method; t represents time. This represents the distance between adjacent orbits of different parameter sequences in phase space;

[0110] when When the value is greater than 0 and the dynamic response index is within the normal range, it is determined that there is a nonlinear coupling relationship between the corresponding parameters. The larger the value, the stronger the coupling.

[0111] when When the dynamic response index is >0 and exceeds the normal range, it is determined to be a correlation driven by outdoor disturbances, not an indoor endogenous risk, and is only recorded.

[0112] when When ≤0, it is determined that there is no significant nonlinear coupling between the parameters.

[0113] Spatiotemporal Variation Verification and Optimization Module: Receives the analysis results and risk identification results output by the analysis model building module, performs multi-dimensional verification on the analysis results and risk identification results, dynamically adjusts the monitoring strategy based on the verification results, and sends the adjusted monitoring strategy back to the data acquisition module to complete the data iterative optimization process until the preset iteration termination conditions are met, and outputs the optimal monitoring result.

[0114] In this embodiment, it should be specifically noted that the multi-dimensional verification includes spatial heterogeneity verification, temporal periodicity verification, and comprehensive variation verification.

[0115] The spatial heterogeneity verification is based on the valid dataset and quantization results output by the constraint condition construction module. If... >0.6, marked as a region of high spatial variability;

[0116] The time-periodic verification: based on the valid dataset output by the constraint condition construction module and the quantization calculation results, if... >0.5 indicates a period of high temporal variability;

[0117] The comprehensive variation verification involves extracting the spatial and temporal variation coefficients from the data fusion module and calculating the spatiotemporal variation comprehensive index. ,in This indicates the weighting coefficient adjusted by the dynamic response index. When the dynamic response index is within the normal range, γ=1, the full weight of variation is included, and it is determined to be endogenous variation within the room; when the dynamic response index is within the normal range, γ=1, the variation weight is fully included, and it is determined to be endogenous variation within the room. When the value exceeds the normal range, γ=0.5, the variation weight is halved to reduce the impact of variations caused by outdoor disturbances and avoid false triggering; when the dynamic response index... When the environment is in steady state, γ = 0.2, and only basic monitoring is performed; α is the spatial variation weighting coefficient, and β is the temporal variation weighting coefficient; based on the natural variation characteristics of environmental parameters, they are divided into two categories: fast-changing parameters and slow-changing parameters. Fast-changing parameters include wind speed, wind direction, noise, illuminance, and solar radiation; slow-changing parameters include temperature and humidity, air pressure, CO2, PM2.5, PM10, and oxygen content. In steady state, α + β = 1, fast-changing parameters β = 0.7 and α = 0.3, slow-changing parameters α = 0.6 and β = 0.4. In densely populated areas, α is increased by 0.2, and β is decreased by 0.2 accordingly. The densely populated areas refer to indoor public areas with a designed personnel density ≥ 0.5 people / ㎡, including conference rooms, halls, atriums, restaurants, or lecture halls, etc.

[0118] The monitoring strategy:

[0119] When 0.4≤ When the value is ≤0.7, it is determined to be a medium variation area / time period. Fast-changing parameters and slow-changing parameters are collected at the conventional interval. The conventional collection interval is 10s for fast-changing parameters (wind speed, wind direction, noise, illuminance, solar radiation) and 60s for slow-changing parameters (temperature, humidity, air pressure, CO2, PM2.5, PM10, oxygen content).

[0120] when When the value is greater than 0.7, it is identified as a high-variability area / time period, triggering encrypted monitoring: the acquisition interval for fast-changing parameters is shortened to 5 seconds, and the acquisition interval for slow-changing parameters is shortened to 30 seconds;

[0121] when When the value is less than 0.4, it is determined to be a stable region / time period, and frequency reduction monitoring is implemented: the acquisition interval of fast-changing parameters is extended to 20s, and the acquisition interval of slow-changing parameters is extended to 120s;

[0122] Furthermore, when the dynamic response index is <0.3 or >2.0, regardless of the spatiotemporal variation comprehensive index... To determine whether the standards are met, the driving-response parameters of the paired areas are triggered for encrypted monitoring, and the collection interval is shortened to half of the regular interval. The areas are then simultaneously marked as key traceability areas. The paired areas refer to a combination of indoor and outdoor monitoring areas that have a direct relationship of ventilation and heat exchange in space and whose parameter changes are causally related. The purpose is to bind "an indoor area directly affected by outdoor weather" and "an outdoor monitoring area that can represent the outdoor weather conditions affecting it" to form an independent analysis unit.

[0123] When the dynamic response index is 3, the highest level of encrypted monitoring is triggered, the interval between all parameter collections is shortened to 1 / 3 of the normal collection interval, and a yellow warning is triggered simultaneously.

[0124] The iteration termination condition is: within 30 consecutive iteration cycles. and When the overall rate of change is less than 0.5%, the iteration is terminated, and the result of the last iteration is output as the optimal monitoring result.

[0125] Specifically, the core action of this module is to conduct multi-dimensional compliance verification of the monitoring and risk identification results, dynamically adjust the monitoring strategy based on the verification results, complete the iterative optimization of data collection, until the preset iteration termination conditions are met, and output the optimal monitoring results. The specific implementation process is as follows:

[0126] First, after receiving the output of the analysis model building module, the multi-dimensional verification unit performs verification in three dimensions: First, spatial heterogeneity verification, based on the spatial distribution uniformity quantification results output by the constraint construction module, areas exceeding the anomaly threshold are marked as spatially high-variability areas; Second, temporal periodicity verification, based on the temporal fluctuation amplitude quantification results output by the constraint construction module, periods exceeding the anomaly threshold are marked as temporally high-variability periods; Third, comprehensive variation verification, combined with the calculation results of the dynamic response model, a correction coefficient is introduced to weight the variation results in the spatial and temporal dimensions to obtain a comprehensive spatiotemporal variation index. When the response sensitivity is within the normal range, the correction coefficient is taken at its full value, and the variation results are fully included in the comprehensive index, which is judged as indoor endogenous variation; when the response sensitivity exceeds the normal range, the correction coefficient is taken at half value to reduce the variation weight caused by outdoor disturbances and avoid false triggering; when it is judged as environmental steady state, the correction coefficient is taken at a low value, and only basic monitoring is performed without triggering any encryption strategies.

[0127] After completing multi-dimensional verification, the monitoring strategy dynamic adjustment unit adjusts the monitoring strategy differently for regions and time periods with different levels of variation based on the verification results: for regions and time periods with high variation, it triggers encrypted monitoring and shortens the collection cycle of the corresponding parameters; for regions and time periods with medium variation, it maintains the normal collection cycle; for regions and time periods with stable variation, it performs reduced-frequency monitoring and extends the collection cycle of the corresponding parameters; at the same time, for regions with abnormal response sensitivity, regardless of whether the comprehensive variation index meets the standard, it triggers encrypted monitoring of the corresponding indoor and outdoor paired parameters and marks them as key traceability areas; for reverse strong abnormal operating conditions, it triggers the highest level of encrypted monitoring of all parameters and pushes a yellow warning alert simultaneously.

[0128] Afterwards, the iterative optimization control unit sends the adjusted monitoring strategy back to the front-end IoT monitoring terminal. The IoT monitoring terminal performs data collection according to the new collection cycle, starting a new round of full-process data processing and verification. At the same time, iterative termination conditions are set. When the combined change of the spatial and temporal dimensions of the variation results is lower than the preset threshold in multiple consecutive iteration cycles, the iteration is terminated, the current monitoring strategy is locked as the optimal strategy, and the current monitoring dataset is taken as the optimal monitoring result and output to the subsequent environmental parameter evaluation module.

[0129] Environmental parameter assessment module: Receives the optimal monitoring results output by the spatiotemporal variation verification and optimization module, calculates the deviation of environmental parameters and the comprehensive health index based on the optimal monitoring results, and outputs the corresponding building environment status assessment results and early warning signals.

[0130] In this embodiment, it is specifically necessary to explain the deviation of the environmental parameters. ,in, This represents the preprocessed real-time monitoring environmental parameter values. The values ​​represent the average parameters over a stable period of 7 consecutive days under normal operating conditions. These are the upper and lower limits of the corresponding national standard values, respectively. For parameters without a lower limit requirement according to national standards, these are the upper and lower limits. Set to 0;

[0131] The comprehensive health index: ,

[0132] in The data integration degree is represented by ω1, ω2, ω3, ω4, and ω5, which represent weights. The weights are allocated as follows: ω1 + ω2 + ω3 + ω4 + ω5 = 1. The weights for indoor and outdoor correlation are ω1 = 0.15, spatiotemporal variation are ω2 = 0.25, nonlinear coupling risk is ω3 = 0.25, dynamic response performance is ω4 = 0.2, and parameter compliance is ω5 = 0.15.

[0133] The warning signal:

[0134] Level I: EHI ≥ 0.85, and It is within the normal range, with no coupling risk, and the environmental condition is stable, outputting a green normal signal;

[0135] Level II: 0.7 ≤ EHI < 0.85, and The system is within the normal range, with no significant coupling risk, and environmental fluctuations are controllable, resulting in a blue warning signal being output.

[0136] Level III: 0.5 ≤ EHI < 0.7, or If the value is outside the normal range, there is a risk of weak nonlinear coupling, or the deviation of a single parameter exceeds the standard, a yellow warning signal will be output, triggering an environmental operation and maintenance prompt.

[0137] Level IV: EHI < 0.5, or Reverse anomalies indicate a risk of strong nonlinear coupling or simultaneous exceedance of multiple parameters. A red warning signal is output, and the parameters exceeding the standard, coupling pairs, and the location and time period of the exceedance are pushed simultaneously. Based on three-dimensional coordinates and spatial difference vectors, the spatial area where the risk occurs is located, and the core coupling parameter pairs are identified.

[0138] Specifically, the core function of this module is to calculate the degree of deviation of environmental parameters and the comprehensive health index based on the optimal monitoring results, complete the comprehensive assessment of the building environment status, and output the corresponding assessment results and graded early warning signals. The specific implementation process is as follows:

[0139] First, after receiving the optimal monitoring results, the environmental status assessment unit calculates the deviation of individual environmental parameters. Using the average value of parameters under stable operating conditions over several consecutive days as the benchmark value and the parameter limits stipulated by current national standards as the upper and lower boundaries, it calculates the relative deviation between the real-time monitoring value and the benchmark value. Then, it constructs a comprehensive building environment health index, integrating the results of five dimensions: correlation between indoor and outdoor parameters, spatiotemporal variation comprehensive index, multi-parameter coupling risk intensity, environmental response sensitivity, and parameter deviation. Fixed weights are assigned to each dimension to calculate a dimensionless comprehensive health index. The closer the comprehensive health index value is to 1, the better the health and comfort of the building environment.

[0140] Then, based on the numerical range of the comprehensive health index, the abnormality of response sensitivity, and the strength of coupling risk, a four-level environmental status classification and early warning judgment is executed: Level 1 is the excellent state, with the comprehensive health index in the highest range, normal response sensitivity, no coupling risk, and stable environmental status, outputting a green normal signal; Level 2 is the good state, with the comprehensive health index in the second highest range, normal response sensitivity, no significant coupling risk, and controllable environmental fluctuations, outputting a blue warning signal; Level 3 is the critical state, with the comprehensive health index in the middle range, or the response sensitivity exceeding the normal range, or the existence of weak coupling risk, or the deviation of a single parameter exceeding the standard, outputting a yellow warning signal and simultaneously triggering environmental operation and maintenance prompts; Level 4 is the out-of-standard state, with the comprehensive health index in the lowest range, or the response sensitivity showing a strong reverse anomaly, or the existence of strong coupling risk, or multiple parameters exceeding the standard simultaneously, outputting a red warning signal;

[0141] Finally, the early warning signal generation unit synchronously pushes the corresponding exceeding parameters, coupled correlation parameter pairs, exceeding location and time period for the early warning signal. Based on the three-dimensional coordinate system and spatial difference vector, it completes the accurate positioning of the risk area and identifies the core coupled risk parameter pairs. The result output unit synchronously pushes the complete building environment status assessment results, graded early warning signals, risk positioning and source tracing results to the building property operation and maintenance platform. At the same time, it can connect to the building automation system to trigger the automatic adjustment of the fresh air, air conditioning and lighting systems, and finally complete the entire process of building environment monitoring, analysis, optimization and evaluation.

[0142] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A building environment monitoring system based on the Internet of Things, characterized in that, include: Data acquisition module: A three-dimensional coordinate system is established based on the building structure. Monitoring points are selected within the coordinate system and IoT monitoring terminals are installed. The IoT monitoring terminals collect multidimensional indoor and outdoor geographic data and environmental parameters of the building to form a raw dataset of monitoring points with spatial location and time stamp. Constraint Condition Construction Module: Receives the original dataset of monitoring points output by the data acquisition module, performs data preprocessing to obtain a valid dataset, completes quantitative calculation based on the valid dataset, sets constraint thresholds, and forms quantitative monitoring constraints containing the valid dataset, quantitative calculation results, and constraint thresholds, and outputs them. Data fusion module: Receives the quantitative monitoring constraints output by the constraint construction module, and performs spatial feature extraction, temporal feature extraction and multi-source data spatiotemporal fusion processing on the effective dataset in sequence, using the constraint threshold in the quantitative monitoring constraints as the boundary, and outputs a standardized fusion dataset; Analysis Model Building Module: Receives the standardized fusion dataset output by the data fusion module, performs multi-dimensional analysis on the standardized fusion dataset, and outputs the corresponding analysis results and risk identification results; Spatiotemporal variation verification and optimization module: Receives the analysis results and risk identification results output by the analysis model building module, retrieves the effective dataset from the constraint construction module, performs multi-dimensional verification on the analysis results and risk identification results, dynamically adjusts the monitoring strategy based on the verification results, and sends the adjusted monitoring strategy back to the data acquisition module to complete the data iterative optimization process until the preset iteration termination condition is met, and outputs the optimal monitoring result; Environmental parameter assessment module: Receives the optimal monitoring results output by the spatiotemporal variation verification and optimization module, calculates the deviation of environmental parameters and the comprehensive health index based on the optimal monitoring results, and outputs the building environment status assessment results and early warning signals.

2. The building environment monitoring system based on the Internet of Things according to claim 1, characterized in that, The three-dimensional coordinate system has its origin O (0, 0, 0) at the main entrance of the building, with the x-axis pointing due east, the y-axis pointing due north, and the z-axis pointing vertically upwards. The geographic data specifically includes: the three-dimensional coordinates (x, y, z) of each monitoring point inside and outside the building obtained through positioning devices; the environmental parameters include outdoor and indoor parameters. The outdoor parameters include 10 items: outdoor temperature, humidity, wind speed, wind direction, solar radiation, noise, PM2.5, PM10, air pressure, and illuminance; the indoor parameters include 9 items: indoor wind speed, wind direction, illuminance, air temperature, air humidity, air pressure, PM2.5, CO2 concentration, and oxygen content; and several IoT monitoring terminals are deployed indoors and outdoors of the building structure according to a tiered deployment rule.

3. The building environment monitoring system based on the Internet of Things according to claim 2, characterized in that, The tiered deployment rules are as follows: For enclosed indoor spaces: 1 IoT monitoring terminal is deployed for every single space with a building area ≤ 50㎡ and a floor height ≤ 3m; for every additional 30㎡ of building area or every additional 2m of floor height, 1 additional IoT monitoring terminal is deployed; for densely populated indoor areas, [the number of terminals is not specified]. One IoT monitoring terminal is installed in the space, with a horizontal spacing of ≤8m between IoT monitoring terminals; Outdoor monitoring areas are divided into background monitoring points, building influence areas, and pedestrian influence areas. Background monitoring points are installed in unobstructed areas ≥30m away from the building body, with 2 units installed. Building influence areas are installed in areas 10m-20m away from the building exterior wall, with 2-3 units evenly distributed, in the downwind direction of the prevailing wind. Pedestrian influence areas are installed within 10m of the main entrance of the building and the entrance / exit of the underground parking garage, with 1-2 units installed. The spacing between outdoor IoT monitoring terminals is ≤15m. All IoT monitoring terminals are installed at a uniform height of 1.2m-1.5m, avoiding ventilation openings, heat sources, pollution sources, and obstructions.

4. The building environment monitoring system based on the Internet of Things according to claim 1, characterized in that, The data preprocessing includes denoising and normalization. The denoising method uses a 5-point window and a second-order polynomial Savitzky-Golay filter to remove random noise from the data. Normalization uses the min-max normalization method to uniformly map all types of environmental parameters to the [0,1] interval, eliminating the dimensional differences between different parameters; The quantitative calculations are as follows: based on the effective dataset, spatial coefficient of variation, temporal coefficient of variation, and data fusion degree are calculated respectively. The spatial coefficient of variation is used to quantify the uniformity of the distribution of environmental parameters in the building space at the same time, identify local parameter abrupt change areas, and obtain the quantitative result of spatial distribution uniformity, which is used to reflect the degree of distribution difference of environmental parameters in the building space. The temporal coefficient of variation is used to quantify the fluctuation amplitude of environmental parameters at a single monitoring point in the time dimension, identify the period of abnormal parameter fluctuation, and obtain the quantitative result of temporal fluctuation amplitude, which is used to reflect the degree of drastic change of environmental parameters in the time dimension. Data fusion degree calculation is a quantification of the spatiotemporal correlation of indoor and outdoor environmental parameters of the same type, used to reflect the synchronicity of changes in indoor and outdoor parameters of the same type; After completing all quantitative calculations, quantitative constraint thresholds are set based on current national building environment standards and the building's intended use.

5. The building environment monitoring system based on the Internet of Things according to claim 4, characterized in that, The spatial variability coefficient ,in, This represents the standard deviation of the same parameter value at all monitoring points at the same time. This represents the average value of the same parameter at all monitoring points at the same time. When the value is 1.5 times lower than the corresponding sensor's detection limit, Set to 0; The coefficient of time variation ,in, The standard deviation of a single point parameter time series. This represents the mean of a single point parameter time series, when When the value is 1.5 times lower than the corresponding sensor's detection limit, Set to 0; The degree of data fusion ,in, B represents the natural constant, and B represents the total number of pairs of parameters for the same physical quantity in indoor and outdoor environments. This represents the min-max normalized difference of the k-th pair of parameters at time t. This represents the global maximum value of the normalized difference of the paired parameters in the k-th group within the monitoring period. , These represent the timestamps for the collection of the k-th parameter group indoors and outdoors, respectively. Represents the time decay constant; The constraint thresholds include a data fusion degree threshold, a spatial coefficient of variation threshold, and a temporal coefficient of variation threshold. The data fusion degree threshold is: >0.7, the threshold for spatial variation coefficient is >0.6, the threshold for the coefficient of variation over time is >0.

5.

6. The building environment monitoring system based on the Internet of Things according to claim 5, characterized in that, The process of extracting a standardized fused dataset from a valid dataset: The first step, based on building indoor and outdoor geographic data, is to define the criteria for determining adjacent monitoring points: for indoor monitoring points on the same floor and of the same type, those with a three-dimensional spatial distance within a preset range are considered adjacent monitoring points; for outdoor monitoring points in the same area, those with a three-dimensional spatial distance within a preset range are also considered adjacent monitoring points. Then, for each monitoring point, the normalized difference value of its parameters of the same type with all adjacent points is calculated to form a spatial difference vector, based on a spatial variation coefficient threshold. >0.6, filter out high-variability regions from the effective dataset where spatial distribution differences exceed the threshold, and remove low-value data with uniform spatial distribution and no significant changes; The second step involves using a 30-minute fixed-duration sliding window to calculate the mean, standard deviation, and coefficient of variation of the single-parameter time series, based on the time coefficient of variation threshold. >0.5, in the high-variability area data selected in the first step, further filter out the high-variability period data with time fluctuation exceeding the threshold, and remove low-value data with stable time dimension; The third step is to obtain the data integration degree. In the high spatiotemporal variability data selected in the second step, identify Strongly correlated parameter pairs with a value greater than 0.7 are used to form a standardized fusion dataset.

7. The building environment monitoring system based on the Internet of Things according to claim 6, characterized in that, The multi-dimensional analysis includes a spatiotemporal feature analysis model, a dynamic response model, and a chaotic correlation model. The spatiotemporal feature analysis model is used to calculate spatial difference vectors and time series statistical features, quantify and standardize the spatial heterogeneity and temporal fluctuation patterns of all parameters in the fusion dataset, and output a spatial heterogeneity distribution map and a temporal fluctuation trend map of the building's overall environmental parameters, as well as a preliminary list of highly variable areas / time periods. The spatial difference vector ,in, Indicates monitoring point The m-th parameter, The parameters represent adjacent points j; the statistical characteristics of the time series are calculated using the time variation coefficient. The dynamic response model is used to calculate the dynamic response index. ,in, This represents the number of time series data points, which is determined by the monitoring duration. To fix the calculation step size; This represents the normalized rate of change of the outdoor drive parameters; Indicates the number of monitoring cycles. Outdoor wind speed after normalization at each sampling time; Indicates the number of monitoring cycles. Solar radiation after normalization at each sampling time; This is expressed as the normalized rate of change of the indoor response parameters; The table represents the number of monitoring periods. Indoor wind speed after normalization at each sampling time; Indicates the number of monitoring cycles. Indoor illuminance after normalization at each sampling time; Sensitivity of indoor thermal environment to outdoor weather disturbances used to quantify the response of indoor thermal environment: Normal range: 0.3≤ ≤2.0 indicates that the indoor thermal environment exhibits reasonable response attenuation to outdoor meteorological disturbances, which is within the normal operating condition; low response is abnormal. <0.3 indicates that the indoor thermal environment does not change with outdoor weather conditions, which may be due to excessive airtightness of the building envelope, a malfunction in the fresh air system, or over-adjustment of the air conditioning system; Highly sensitive anomaly: >2.0, Indoor thermal environment is excessively sensitive to changes in outdoor weather, corresponding to insufficient thermal insulation performance of the building envelope, air leakage through doors and windows, or malfunction of the air conditioning system; Reverse anomaly: =3, the indoor and outdoor environmental trends are completely opposite, which corresponds to abnormal heating of indoor equipment or failure of cold / heat source; The weighted average of the normalized rates of change of the three core driving parameters—outdoor temperature, solar radiation, and wind speed—is used to obtain the comprehensive rate of change of the outdoor driving parameters. When the comprehensive rate of change of the outdoor driving parameters is <5% / 10min, and... When the denominator of the calculation formula is 0, the environment is considered to be in steady state, and the dynamic response index is determined. Marked as 1; when the comprehensive change rate of outdoor drive parameters is ≥5% / 10min, and When the denominator of the calculation formula is 0, it is determined that there is no response indoors, and the dynamic response index is... The value is marked as 0; when When the denominator of the calculation formula is negative, it is determined that the indoor and outdoor change trends are reversed, and the dynamic response index is affected. Marked as 3, included in the strongly abnormal operating condition; The chaotic correlation model is used to quantify the nonlinear coupling strength among multiple parameters. It calculates the maximum cross-Lyapunov exponent through multivariable phase space reconstruction. ,in The time series step size is T, and the time series length is T. To determine the embedding dimension, a pseudo nearest neighbor method is used. ; The mutual information method is used to determine the delay time. t represents time; The distance between adjacent orbits of different parameter sequences in phase space; when When the value is greater than 0 and the dynamic response index is within the normal range, it is determined that there is a nonlinear coupling relationship between the corresponding parameters; the larger the value, the stronger the coupling. When the dynamic response index is >0 and exceeds the normal range, it is determined to be a correlation driven by outdoor disturbances, not an inherent indoor risk, and is only recorded; when When ≤0, it is determined that there is no significant nonlinear coupling between the parameters.

8. The building environment monitoring system based on the Internet of Things according to claim 7, characterized in that, The multi-dimensional verification includes spatial heterogeneity verification, temporal periodicity verification, and comprehensive variation verification; The spatial heterogeneity verification is based on the valid dataset and quantization results output by the constraint condition construction module. If... >0.6, marked as a region of high spatial variability; The time-periodic verification: based on the valid dataset output by the constraint condition construction module and the quantization calculation results, if... >0.5 indicates a period of high temporal variability; The comprehensive variation verification involves extracting the spatial and temporal variation coefficients from the data fusion module and calculating the spatiotemporal variation comprehensive index. ,in This indicates the weighting coefficient adjusted by the dynamic response index. When the dynamic response index is within the normal range, γ=1, the full weight of variation is included, and it is determined to be endogenous variation within the room; when the dynamic response index is within the normal range, γ=1, the variation weight is fully included, and it is determined to be endogenous variation within the room. When the value exceeds the normal range, γ=0.5, the variation weight is halved to reduce the impact of variations caused by outdoor disturbances and avoid false triggering; when the dynamic response index... When the environment is in steady state, γ = 0.2; α is the spatial variation weighting coefficient, and β is the time variation weighting coefficient. Based on the natural variation characteristics of environmental parameters, they are divided into two categories: fast-changing parameters and slow-changing parameters. When the environment is in steady state, α + β = 1. For fast-changing parameters, β = 0.7 and α = 0.

3. For slow-changing parameters, α = 0.6 and β = 0.

4. In densely populated areas, α is increased by 0.2, and β is decreased by 0.

2. The densely populated areas refer to indoor public areas with a design personnel density ≥ 0.5 people / ㎡.

9. The building environment monitoring system based on the Internet of Things according to claim 8, characterized in that, The monitoring strategy: When 0.4≤ When the value is ≤0.7, it is determined to be a medium variation region / time period. The fast variation parameter and the slow variation parameter adopt the conventional acquisition interval. The conventional acquisition interval is 10s for fast variation parameter and 60s for slow variation parameter. when When the value is greater than 0.7, it is identified as a high-variability area / time period, triggering encrypted monitoring: the acquisition interval for fast-changing parameters is shortened to 5 seconds, and the acquisition interval for slow-changing parameters is shortened to 30 seconds; when When the value is less than 0.4, it is determined to be a stable region / time period, and frequency reduction monitoring is implemented: the acquisition interval for fast-changing parameters is extended to 20s, and the acquisition interval for slow-changing parameters is extended to 120s. Furthermore, when the dynamic response index is <0.3 or >2.0, regardless of the spatiotemporal variation comprehensive index... To determine whether the criteria are met, encrypted monitoring is triggered on the drive-response parameters of the paired area, the collection interval is shortened to half of the normal interval, and the area is simultaneously marked as a key traceability area; When the dynamic response index is 3, the highest level of encrypted monitoring is triggered, and the collection interval of all parameters is shortened to 1 / 3 of the normal collection interval; The iteration termination condition is: within 30 consecutive iteration cycles. and When the overall rate of change is less than 0.5%, the iteration is terminated and the optimal monitoring result is output.

10. The building environment monitoring system based on the Internet of Things according to claim 9, characterized in that, The deviation of the environmental parameters ,in, This represents the preprocessed real-time monitoring environmental parameter values. The values ​​represent the average parameters over a stable period of 7 consecutive days under normal operating conditions. These are the upper and lower limits of the national standard for the corresponding parameters. For parameters without a lower limit requirement in the national standard, the value is 0. The comprehensive health index ,in The data integration degree is represented by ω1, ω2, ω3, ω4, and ω5, which represent weights. The weights are allocated as follows: ω1 + ω2 + ω3 + ω4 + ω5 = 1. The weights for indoor and outdoor correlation are ω1 = 0.15, spatiotemporal variation are ω2 = 0.25, nonlinear coupling risk is ω3 = 0.25, dynamic response performance is ω4 = 0.2, and parameter compliance is ω5 = 0.

15. The warning signal: Level I: EHI ≥ 0.85, and It is within the normal range, with no coupling risk, and the environmental condition is stable, outputting a green normal signal; Level II: 0.7 ≤ EHI < 0.85, and The system is within the normal range, with no significant coupling risk, and environmental fluctuations are controllable, resulting in a blue warning signal being output. Level III: 0.5 ≤ EHI < 0.7, or If the value is outside the normal range, there is a risk of weak nonlinear coupling, or the deviation of a single parameter exceeds the standard, a yellow warning signal will be output, triggering an environmental operation and maintenance prompt. Level IV: EHI < 0.5, or If the system is in a reverse anomaly, indicating a strong risk of nonlinear coupling or multiple parameters exceeding the standard simultaneously, a red warning signal will be output. The system will simultaneously push out the parameters exceeding the standard, the coupling pairs, the location and time period of the exceeding the standard, and locate the spatial area where the risk occurs based on three-dimensional coordinates and spatial difference vectors, and identify the core coupling parameter pairs.

Images