Energy-saving building indoor air quality monitoring system and method

Abstract

The invention discloses an energy-saving building indoor air quality monitoring system and method. The system comprises a sensor module, a data processing module, a quality evaluation module and an alarm module. The sensor module is used for monitoring initial air data in a building; the data processing module is used for preprocessing the initial air data and performing spatial-temporal feature fusion on the preprocessed data to obtain fused features; the quality evaluation module evaluates the air quality based on the fused features to obtain an evaluation result; and the alarm module gives an alarm based on the evaluation result. The key air quality indexes are monitored in real time through the integrated distributed sensor network, data preprocessing and feature fusion are carried out through the advanced data processing module, intelligent evaluation is carried out through the LSTM-AE deep learning model, the monitoring precision and efficiency are effectively improved, and remarkable technical advancement and economic benefits are achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent buildings, and in particular relates to an energy-saving building indoor air quality monitoring system and method. Background Art

[0002] In the indoor air quality monitoring system of energy-saving buildings, a large number of sensor nodes need to be deployed because multiple areas inside the building need to be monitored in real time. However, with the expansion of the monitoring range and the increase in the number of sensors, the complexity of data collection and processing also increases. The traditional centralized data processing method is difficult to meet the requirements of real-time and scalability, and is prone to problems such as data processing delays and single point failures.

[0003] In order to solve this problem, it is necessary to use parallel monitoring technology to distribute data collection and processing tasks to multiple nodes for simultaneous execution. However, when implementing parallel monitoring, there is the challenge of how to coordinate data interaction and task synchronization between multiple nodes. If data transmission and task scheduling between different nodes cannot be effectively coordinated, it will lead to system performance degradation and resource waste.

[0004] In addition, in the process of parallel monitoring, data fusion and consistency between different sensors need to be considered. Since the sampling frequency, accuracy and stability of different sensors may vary, how to ensure data consistency and reliability in the process of parallel processing is also a technical problem that needs to be solved urgently. At the same time, parallel monitoring must also take into account the fault tolerance and robustness of the system to ensure that when some nodes fail, the system can still maintain normal operation and repair or replace the failed nodes in a timely manner. Summary of the invention

[0005] The present invention aims to solve the deficiencies of the prior art and provides the following solutions:

[0006] An energy-saving building indoor air quality monitoring system comprises: a sensor module, a data processing module, a quality assessment module and an alarm module;

[0007] The sensor module is used to monitor initial air data in the building;

[0008] The data processing module is used to preprocess the initial air data and perform spatiotemporal feature fusion on the preprocessed data to obtain fused features;

[0009] The quality assessment module assesses the air quality based on the fused features to obtain an assessment result;

[0010] The alarm module issues an alarm based on the evaluation result.

[0011] Preferably, the sensor module is a distributed sensor network;

[0012] The distributed sensor network includes: a temperature sensor, a humidity sensor, a carbon dioxide sensor, a volatile organic compound sensor, a particulate matter sensor, and a data collector;

[0013] The temperature sensor is used to measure indoor temperature data, the humidity sensor is used to measure indoor humidity data, the carbon dioxide sensor is used to measure indoor carbon dioxide concentration data, the volatile organic compound sensor is used to measure volatile organic compound concentration data, the particulate matter sensor is used to measure PM2.5 concentration data and PM10 concentration data, and the data collector is used to collect the data collected by each sensor and package it into the initial air data.

[0014] Preferably, the data processing module includes: a data preprocessing unit, a feature extraction unit, and a feature fusion unit;

[0015] The data preprocessing unit is used to perform denoising, outlier processing, and normalization processing on the initial air data to obtain preprocessed data;

[0016] The feature extraction unit is used to perform spatio-temporal alignment and feature extraction on the preprocessed data to obtain multi-source data features of different source data;

[0017] The feature fusion unit is used to perform spatio-temporal feature fusion on the multi-source data features to obtain the fused features.

[0018] Preferably, the quality assessment module includes: a model construction unit, a model training unit, and an evaluation unit;

[0019] The model construction unit is used to construct an LSTM-AE model;

[0020] The model training unit trains the LSTM-AE model based on historical air quality data and historical evaluation results to obtain a quality assessment model;

[0021] The evaluation unit inputs the fused features into the quality assessment model for control quality assessment to obtain the evaluation results.

[0022] Preferably, the LSTM-AE model includes: an input layer, an LSTM layer, an encoder layer, a decoder layer, and an output layer;

[0023] The input layer is used to receive the fused features;

[0024] The LSTM layer is used to capture the change features and change patterns of air quality in the fused features;

[0025] The encoder layer is composed of multiple fully-connected layers, and the encoder layer is used to compress the output of the LSTM layer into a low-dimensional feature representation;

[0026] The decoder layer is used to reconstruct the low-dimensional feature representation to obtain high-dimensional output data;

[0027] The output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain the evaluation result based on the reconstruction error.

[0028] The present invention also provides an energy-saving building indoor air quality monitoring method, which is applied to the monitoring system described in any one of the above, and includes the following steps:

[0029] Monitor the initial air data in the building;

[0030] Preprocess the initial air data, and perform spatio-temporal feature fusion on the preprocessed data to obtain fused features;

[0031] Evaluate the air quality based on the fused features to obtain an evaluation result;

[0032] Send an alarm based on the evaluation result.

[0033] Preferably, a distributed sensor network is constructed to monitor the initial air data;

[0034] The distributed sensor network includes: a temperature sensor, a humidity sensor, a carbon dioxide sensor, a volatile organic compound sensor, a particulate matter sensor, and a data collector;

[0035] The temperature sensor is used to measure indoor temperature data, the humidity sensor is used to measure indoor humidity data, the carbon dioxide sensor is used to measure indoor carbon dioxide concentration data, the volatile organic compound sensor is used to measure volatile organic compound concentration data, the particulate matter sensor is used to measure PM2.5 concentration data and PM10 concentration data, and the data collector is used to collect the data collected by each sensor and package it into the initial air data.

[0036] Preferably, the method for obtaining the fused features includes:

[0037] Denoise, process outliers, and standardize the initial air data to obtain preprocessed data;

[0038] Perform spatio-temporal alignment and feature extraction on the preprocessed data to obtain multi-source data features of different source data;

[0039] Perform spatio-temporal feature fusion on the multi-source data features to obtain the fused features.

[0040] Preferably, the method for obtaining the evaluation result includes:

[0041] Construct an LSTM-AE model;

[0042] Train the LSTM-AE model based on historical air quality data and historical evaluation results to obtain a quality evaluation model;

[0043] Input the fused features into the quality evaluation model for control quality evaluation to obtain the evaluation result.

[0044] Preferably, the LSTM-AE model includes: an input layer, an LSTM layer, an encoder layer, a decoder layer, and an output layer;

[0045] The input layer is used to receive the fused features;

[0046] The LSTM layer is used to capture the change features and change patterns of air quality in the fused features;

[0047] The encoder layer is composed of multiple fully connected layers, and the encoder layer is used to compress the output of the LSTM layer into a low-dimensional feature representation;

[0048] The decoder layer is used to reconstruct the low-dimensional feature representation to obtain high-dimensional output data;

[0049] The output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain the evaluation result based on the reconstruction error.

[0050] Compared with the prior art, the beneficial effects of the present invention are:

[0051] The present invention effectively improves the monitoring accuracy and efficiency by integrating a distributed sensor network to real-time monitor key air quality indicators, using an advanced data processing module for data preprocessing and feature fusion, and adopting an LSTM-AE deep learning model for intelligent evaluation, comprehensively covering multiple key parameters of indoor air quality, realizing a real-time alarm mechanism, thus ensuring the physical health of indoor personnel, improving the living comfort, and at the same time reducing energy consumption, with remarkable technological advancement and economic benefits. Description of the Drawings

[0052] To more clearly illustrate the technical solutions of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0053] Figure 1 It is a schematic diagram of the system structure of an embodiment of the present invention. Specific embodiments

[0054] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0055] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] Embodiment 1

[0057] In this embodiment, as Figure 1 shown, an indoor air quality monitoring system for energy-saving buildings includes: a sensor module, a data processing module, a quality assessment module, and an alarm module.

[0058] The sensor module is used to monitor the initial air data in the building.

[0059] The sensor module is a distributed sensor network. The distributed sensor network includes: a temperature sensor, a humidity sensor, a carbon dioxide sensor, a volatile organic compound sensor, a particulate matter sensor, and a data collector. The temperature sensor is used to measure the indoor temperature data, the humidity sensor is used to measure the indoor humidity data, the carbon dioxide sensor is used to measure the indoor carbon dioxide concentration data, the volatile organic compound sensor is used to measure the volatile organic compound concentration data, the particulate matter sensor is used to measure the PM2.5 concentration data and the PM10 concentration data, and the data collector is used to collect the data collected by each sensor and package it into the initial air data.

[0060] In this embodiment, the distributed sensor network includes: (1) Temperature sensors: Temperature sensors of model DHT22 are used. To obtain accurate data on indoor temperature, the temperature sensors are evenly installed in various areas of the building, including the center, corners of rooms, and positions near windows and doors, to ensure the comprehensiveness and representativeness of temperature data; (2) Humidity sensors: Humidity sensors of model HIH-4000 are used. The humidity sensors are evenly installed in various areas of the building, especially in crowded areas and near ventilation outlets, to monitor the impact of humidity changes on indoor air quality; (3) Carbon dioxide sensors: Carbon dioxide sensors of model MH-Z19B are used. The carbon dioxide sensors are installed in areas with frequent human activities, such as meeting rooms, offices, and lounges, as well as near ventilation systems, to monitor changes in carbon dioxide concentration; (4) Volatile organic compound (VOC) sensors: Volatile organic compound (VOC) sensors of model MP502 are used. The VOC sensors are arranged in areas where chemical substances may be released, such as printing rooms, laboratories, and storage rooms, as well as near building materials and furniture, to monitor the concentration of VOCs; (5) Particle sensors: Particle sensors of model GP2Y1010AU0F are used. They are installed near air conditioner outlets and return air vents, as well as in areas with frequent human activities, to monitor the concentration of airborne particles; (6) Data collectors: Data collectors of model Arduino Mega 2560 are used. The data collectors, as the central hub of the sensor module, are placed in the central control room or computer room of the building to collect data from all sensors, and perform preliminary processing and packaging to obtain initial air data.

[0061] The data processing module is used to preprocess the initial air data and perform spatio-temporal feature fusion on the preprocessed data to obtain the fused features.

[0062] The data processing module includes: a data preprocessing unit, a feature extraction unit, and a feature fusion unit. The data preprocessing unit is used to denoise, handle outliers, and perform normalization on the initial air data to obtain the preprocessed data; the feature extraction unit is used to perform spatio-temporal alignment and feature extraction on the preprocessed data to obtain multi-source data features from different source data; the feature fusion unit is used to perform spatio-temporal feature fusion on the multi-source data features to obtain the fused features.

[0063] In this embodiment, the working process of the data preprocessing unit includes: using moving average filtering to denoise the initial air data:

[0064]

[0065] Among them, x’ represents the denoised data at time point t, x represents the initial air data, N represents the size of the smoothing window, and i represents a natural number; then the Z-score method is used to identify outliers:

[0066]

[0067] Among them, μ represents the mean value, σ represents the standard deviation. In this embodiment, data points with a Z value greater than 3 or less than -3 are regarded as outliers. After identifying the outliers, the outliers are removed; the data after outlier removal is standardized to obtain the preprocessed data. The workflow of the feature extraction unit includes: the preprocessed data contains data at different time points, and interpolation and upsampling are required to align them to a unified time series:

[0068] F i (t) = interpolate(F(t1), F(t2),..., F(t n ),t)

[0069] Among them, F i (t) represents the aligned feature vector at time point t, F(t) represents the preprocessed data at different time points, and interpolate represents the interpolation and upsampling operation; then statistics that help describe the data characteristics are extracted from the time-aligned data, including the feature mean and the feature standard deviation:

[0070]

[0071] Among them, μ i represents the feature mean, σ i represents the feature standard deviation, and n represents a natural number. The workflow of the feature fusion unit includes: performing spatio-temporal feature fusion on the multi-source data features to obtain the fused features, and this fused feature is represented by a feature matrix:

[0072]

[0073] Among them, k represents the number of sensors.

[0074] The quality assessment module evaluates the air quality based on the fused features to obtain an evaluation result.

[0075] The quality assessment module includes: a model construction unit, a model training unit, and an evaluation unit. The model construction unit is used to construct an LSTM-AE model; the model training unit trains the LSTM-AE model based on historical air quality data and historical evaluation results to obtain a quality assessment model; the evaluation unit inputs the fused features into the quality assessment model to perform air quality assessment and obtains an evaluation result.

[0076] The LSTM-AE model includes: an input layer, an LSTM layer, an encoder layer, a decoder layer, and an output layer. The input layer is used to receive the fused features; the LSTM layer is used to capture the changing features and patterns of air quality in the fused features; the encoder layer consists of multiple fully connected layers, and the encoder layer is used to compress the output of the LSTM layer into a low-dimensional feature representation; the decoder layer is used to reconstruct the low-dimensional feature representation to obtain high-dimensional output data; the output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain an evaluation result based on the reconstruction error.

[0077] In this embodiment, the input layer receives the fused features from the sensor module, and the dimension of the input features depends on the number and type of sensors; the LSTM layer consists of multiple LSTM units, and each LSTM unit contains an input gate, a forget gate, and an output gate, which control the flow of information, enabling the LSTM to learn long-term dependencies. The purpose of the LSTM layer is to capture the features and patterns in time series data, such as the changes in air quality at different times of the day; the encoder layer consists of multiple fully connected layers (Dense Layers), which further compress the output of the LSTM layer into a low-dimensional feature representation. Each fully connected layer may be followed by an activation function, such as ReLU, to increase non-linearity and help the network learn complex feature representations; the decoder layer structure is symmetric to the encoder layer but has the opposite function. It reconstructs the low-dimensional feature representation back into high-dimensional data, and each fully connected layer in the decoder layer may also be followed by an activation function to help reconstruct the original data. The output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain an evaluation result based on the reconstruction error. Usually, the mean squared error (MSE) is used as the loss function to measure the difference between the reconstructed data and the original data.

[0078] The alarm module issues an alarm based on the evaluation result.

[0079] When the evaluation result finds that the data of one of the sensors is abnormal, an alarm and notification are issued accordingly.

[0080] Embodiment 2

[0081] In this embodiment, an energy-saving building indoor air quality monitoring method includes the following steps:

[0082] S1. Monitor the initial air data in the building.

[0083] Build a distributed sensor network to monitor initial air data. The distributed sensor network includes: a temperature sensor, a humidity sensor, a carbon dioxide sensor, a volatile organic compound sensor, a particulate matter sensor, and a data collector. The temperature sensor is used to measure indoor temperature data, the humidity sensor is used to measure indoor humidity data, the carbon dioxide sensor is used to measure indoor carbon dioxide concentration data, the volatile organic compound sensor is used to measure volatile organic compound concentration data, the particulate matter sensor is used to measure PM2.5 concentration data and PM10 concentration data, and the data collector is used to collect the data collected by each sensor and package it into initial air data.

[0084] S2. Preprocess the initial air data, and perform spatio-temporal feature fusion on the preprocessed data to obtain the fused features.

[0085] The method for obtaining the fused features includes: denoising, outlier processing, and normalization processing on the initial air data to obtain the preprocessed data; performing spatio-temporal alignment and feature extraction on the preprocessed data to obtain multi-source data features of different source data; and performing spatio-temporal feature fusion on the multi-source data features to obtain the fused features.

[0086] S3. Evaluate the air quality based on the fused features to obtain the evaluation result.

[0087] The method for obtaining the evaluation result includes: constructing an LSTM-AE model; training the LSTM-AE model based on historical air quality data and historical evaluation results to obtain a quality evaluation model; and inputting the fused features into the quality evaluation model for control quality evaluation to obtain the evaluation result.

[0088] The LSTM-AE model includes: an input layer, an LSTM layer, an encoder layer, a decoder layer, and an output layer. The input layer is used to receive the fused features; the LSTM layer is used to capture the change features and change patterns of air quality in the fused features; the encoder layer is composed of multiple fully connected layers, and the encoder layer is used to compress the output of the LSTM layer into a low-dimensional feature representation; the decoder layer is used to reconstruct the low-dimensional feature representation to obtain high-dimensional output data; and the output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain the evaluation result based on the reconstruction error.

[0089] S4. Issue an alarm based on the evaluation result.

[0090] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. An indoor air quality monitoring system for energy-saving buildings, characterized in that, Comprising: A sensor module, a data processing module, a quality assessment module, and an alarm module; The sensor module is used to monitor the initial air data in the building; The data processing module is used to preprocess the initial air data and perform spatio-temporal feature fusion on the preprocessed data to obtain the fused features; The quality assessment module assesses the air quality based on the fused features to obtain an assessment result; The alarm module issues an alarm based on the assessment result.

2. The indoor air quality monitoring system for an energy-saving building according to claim 1, wherein The sensor module is a distributed sensor network; The distributed sensor network includes: a temperature sensor, a humidity sensor, a carbon dioxide sensor, a volatile organic compound sensor, a particulate matter sensor, and a data collector; The temperature sensor is used to measure the indoor temperature data, the humidity sensor is used to measure the indoor humidity data, the carbon dioxide sensor is used to measure the indoor carbon dioxide concentration data, the volatile organic compound sensor is used to measure the volatile organic compound concentration data, the particulate matter sensor is used to measure the PM2.5 concentration data and the PM10 concentration data, and the data collector is used to collect the data collected by each sensor and package it into the initial air data.

3. The indoor air quality monitoring system for an energy-saving building according to claim 1, wherein The data processing module includes: a data preprocessing unit, a feature extraction unit, and a feature fusion unit; The data preprocessing unit is used to perform denoising, outlier processing, and normalization processing on the initial air data to obtain the preprocessed data; The feature extraction unit is used to perform spatio-temporal alignment and feature extraction on the preprocessed data to obtain multi-source data features of different source data; The feature fusion unit is used to perform spatio-temporal feature fusion on the multi-source data features to obtain the fused features.

4. The indoor air quality monitoring system for an energy-saving building according to claim 1, characterized in that, The quality assessment module includes: a model construction unit, a model training unit, and an evaluation unit; The model construction unit is used to construct an LSTM-AE model; The model training unit trains the LSTM-AE model based on historical air quality data and historical assessment results to obtain a quality assessment model; The evaluation unit inputs the fused features into the quality assessment model for control quality assessment to obtain the assessment result.

5. The indoor air quality monitoring system for an energy-saving building according to claim 4, wherein The LSTM-AE model includes: an input layer, an LSTM layer, an encoder layer, a decoder layer, and an output layer; The input layer is used to receive the fused features; The LSTM layer is used to capture the change features and change patterns of the air quality in the fused features; The encoder layer is composed of multiple fully connected layers, and the encoder layer is used to compress the output of the LSTM layer into a low-dimensional feature representation; The decoder layer is used to reconstruct the low-dimensional feature representation to obtain high-dimensional output data; The output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain the assessment result based on the reconstruction error.

6. A method for monitoring indoor air quality in an energy-saving building, the method being applied to the monitoring system according to any one of claims 1-5, characterized in that, Including the following steps: Monitoring the initial air data in the building; Preprocessing the initial air data and performing spatio-temporal feature fusion on the preprocessed data to obtain the fused features; Assessing the air quality based on the fused features to obtain the assessment result; Issue an alarm based on the evaluation result.

7. The method for monitoring the indoor air quality of an energy-saving building according to claim 6, characterized in that, Construct a distributed sensor network to monitor the initial air data; The distributed sensor network includes: a temperature sensor, a humidity sensor, a carbon dioxide sensor, a volatile organic compound sensor, a particulate matter sensor, and a data collector; The temperature sensor is used to measure indoor temperature data, the humidity sensor is used to measure indoor humidity data, the carbon dioxide sensor is used to measure indoor carbon dioxide concentration data, the volatile organic compound sensor is used to measure volatile organic compound concentration data, the particulate matter sensor is used to measure PM2.5 concentration data and PM10 concentration data, and the data collector is used to collect the data collected by each sensor and package it into the initial air data.

8. The indoor air quality monitoring method for an energy-saving building according to claim 6, characterized in that, The method for obtaining the fused features includes: Perform denoising, outlier processing, and normalization on the initial air data to obtain preprocessed data; Perform spatio-temporal alignment and feature extraction on the preprocessed data to obtain multi-source data features of different source data; Perform spatio-temporal feature fusion on the multi-source data features to obtain the fused features.

9. The method for monitoring indoor air quality of an energy-saving building according to claim 6, wherein, The method for obtaining the evaluation result includes: Construct an LSTM-AE model; Train the LSTM-AE model based on historical air quality data and historical evaluation results to obtain a quality evaluation model; Input the fused features into the quality evaluation model for control quality evaluation to obtain the evaluation result.

10. The method for monitoring indoor air quality of an energy-saving building according to claim 9, characterized in that, The LSTM-AE model includes: an input layer, an LSTM layer, an encoder layer, a decoder layer, and an output layer; The input layer is used to receive the fused features; The LSTM layer is used to capture the change features and change patterns of air quality in the fused features; The encoder layer is composed of multiple fully connected layers, and the encoder layer is used to compress the output of the LSTM layer into a low-dimensional feature representation; The decoder layer is used to reconstruct the low-dimensional feature representation to obtain high-dimensional output data; The output layer is used to compare the high-dimensional output data with the air quality threshold, calculate the reconstruction error, and obtain the evaluation result based on the reconstruction error.

Images