Indoor environment monitoring system and method based on intelligent sensing technology

Abstract

The present application relates to the field of intelligent monitoring of indoor environments, and specifically discloses an indoor environment monitoring system and method based on intelligent sensing technology, which first collects the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored as input data, and then uses deep learning technology to extract and analyze these input data to obtain classification results indicating that the temperature in the room to be monitored should increase or decrease, and the humidity should increase or decrease. In this way, the indoor environment and indoor personnel can be monitored and comprehensively analyzed in real time, thereby realizing real-time monitoring and intelligent regulation of the indoor environment of the house. In this way, it is helpful to adjust the indoor comfort to the greatest extent.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring of indoor environments, and more specifically, to an indoor environment monitoring system and method based on intelligent sensing technology. Background Art

[0002] With the rapid development of Internet of Things technology and economy, people's requirements for the quality of living environment are increasing day by day. The quality of indoor environment directly affects the health of residents. Therefore, monitoring and adjusting the internal environment of the house has become an indispensable part of smart home. In order to improve the quality of indoor environment, it is necessary to effectively manage the monitoring and adjustment of the internal environment of the house.

[0003] At present, indoor environmental monitoring of houses mainly focuses on air quality, and indoor air is adjusted through fresh air systems to improve the indoor environment. However, there are some problems with the current indoor environmental monitoring and adjustment of houses: the adjustment is mainly based on the basic environmental values ​​monitored, and lacks consideration of the impact of factors such as the distribution, age, and behavior of indoor people on environmental adjustment. Therefore, the existing technology cannot achieve intelligent environmental adjustment, nor can it meet the needs of people of different ages and behavior types for environmental adjustment, and thus cannot achieve the most comfortable adjustment.

[0004] Therefore, we look forward to an indoor environment monitoring system and method based on intelligent sensing technology, which can conduct real-time monitoring and comprehensive analysis of indoor temperature, humidity and indoor personnel based on deep learning technology, so as to realize real-time monitoring and intelligent control of the indoor environment of the house. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an indoor environment monitoring system and method based on intelligent sensing technology, which first collects the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored as input data, and then uses deep learning technology to extract and analyze these input data to obtain classification results indicating that the temperature in the room to be monitored should increase or decrease, and the humidity should increase or decrease. In this way, it is possible to monitor and comprehensively analyze the indoor environment and indoor personnel in real time, thereby realizing real-time monitoring and intelligent control of the indoor environment of the house. In this way, it helps to adjust the indoor comfort to the greatest extent.

[0006] According to a first aspect of the present application, there is provided an indoor environment monitoring system based on intelligent sensing technology, which includes:

[0007] An indoor data acquisition module, used to collect the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored;

[0008] An indoor feature acquisition module, used for extracting and analyzing the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored, respectively, to obtain a temperature and humidity feature matrix and a personnel change feature vector;

[0009] An indoor feature fusion module, used for multiplying the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector;

[0010] The indoor feature classification result generating module is used to obtain a classification result based on the indoor environment feature vector, wherein the classification result is used to indicate that the temperature in the room to be monitored should be increased or decreased, and the humidity should be increased or decreased.

[0011] In combination with the first aspect of the present application, in an indoor environment monitoring system based on intelligent sensing technology in the first aspect of the present application, the indoor feature acquisition module includes: an indoor temperature feature extraction unit, which is used to arrange the temperatures of multiple predetermined time points in the room to be monitored according to the time dimension as a temperature input vector and then pass it through a temperature multi-scale feature extraction module to obtain a temperature multi-scale feature vector; an indoor humidity feature extraction unit, which is used to arrange the humidity of multiple predetermined time points in the room to be monitored according to the time dimension as a humidity input vector and then pass it through a humidity feature extraction module based on a time series encoder to obtain a humidity time series feature vector; a temperature and humidity correlation feature extraction unit, which is used to use a joint encoder to perform feature combination of the temperature multi-scale feature vector and the humidity time series feature vector to obtain the temperature and humidity feature matrix; an indoor personnel feature extraction unit, which is used to perform feature encoding on the personnel monitoring images of multiple predetermined time points in the room to be monitored to obtain the personnel change feature vector.

[0012] According to a second aspect of the present application, a method for indoor environment monitoring based on intelligent sensing technology is provided, which includes:

[0013] Collecting the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored;

[0014] Extracting and analyzing the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored to obtain a temperature and humidity feature matrix and a personnel change feature vector;

[0015] Multiplying the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector;

[0016] A classification result is obtained based on the indoor environment feature vector, and the classification result is used to indicate that the temperature in the room to be monitored should be increased or decreased, and the humidity should be increased or decreased.

[0017] In combination with the second aspect of the present application, in an indoor environment monitoring method based on intelligent sensing technology in the second aspect of the present application, a classification result is obtained based on the indoor environment feature vector, and the classification result is used to indicate that the temperature in the monitored room should increase or decrease, and the humidity should increase or decrease, including: an indoor environment feature extraction unit, used to pass the indoor environment feature vector through an indoor environment feature extraction module based on a convolutional neural network model to obtain an indoor environment classification feature vector; an indoor environment feature optimization unit, used to perform expected activation of the indoor environment classification feature vector in the parameter space to obtain an optimized indoor environment classification feature vector; an indoor environment feature classification unit, used to pass the optimized indoor environment classification feature vector through a classifier to obtain the classification result.

[0018] Compared with the prior art, the present application provides an indoor environment monitoring system and method based on intelligent sensing technology, which first collects the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored as input data, and then uses deep learning technology to extract and analyze these input data to obtain classification results indicating that the temperature in the room to be monitored should increase or decrease, and the humidity should increase or decrease. In this way, the indoor environment and indoor personnel can be monitored and comprehensively analyzed in real time, thereby realizing real-time monitoring and intelligent control of the indoor environment of the house. In this way, it helps to adjust the indoor comfort to the greatest extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 The figure illustrates a schematic block diagram of an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application.

[0021] Figure 2The figure illustrates a schematic block diagram of an indoor feature acquisition module in an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application.

[0022] Figure 3 The figure illustrates a schematic block diagram of an indoor personnel feature extraction unit in an indoor feature acquisition module in an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application.

[0023] Figure 4 The figure illustrates a schematic block diagram of an indoor feature classification result generation module in an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application.

[0024] Figure 5 The figure illustrates a flow chart of an indoor environment monitoring method based on intelligent sensing technology according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0026] Exemplary Systems

[0027] Figure 1 FIG. 1 is a schematic block diagram of an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application. Figure 1 As shown, according to the indoor environment monitoring system 100 based on intelligent sensing technology according to the embodiment of the present application, it includes: an indoor data acquisition module 110, which is used to collect the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored; an indoor feature acquisition module 120, which is used to extract and analyze the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored, respectively, to obtain a temperature and humidity feature matrix and a personnel change feature vector; an indoor feature fusion module 130, which is used to multiply the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector; an indoor feature classification result generation module 140, which is used to obtain a classification result based on the indoor environment feature vector, and the classification result is used to indicate that the temperature in the room to be monitored should increase or decrease, and the humidity should increase or decrease.

[0028] With the rapid development of IoT technology and economy, people's expectations for the quality of living environment are constantly increasing. The quality of indoor environment directly affects the health of residents. Therefore, in smart home systems, monitoring and regulating indoor environment has become a vital part. In order to improve the quality of indoor environment, it is necessary to effectively manage the monitoring and regulation of the internal environment of the house.

[0029] At present, the internal environmental monitoring of houses mainly focuses on air quality, and the indoor air is regulated through the fresh air system to improve the overall indoor environmental quality. However, there are some problems with the current monitoring and regulation of indoor house environments: the regulation is mainly based on the monitored basic environmental values, and lacks consideration of the impact of factors such as the distribution, age, and behavior of indoor personnel on environmental regulation. Therefore, the existing technology cannot achieve intelligent environmental regulation, nor can it meet the needs of people of different ages and behavior types for environmental regulation, and thus cannot achieve the greatest degree of comfort regulation. Therefore, an indoor environment monitoring system and method based on intelligent sensing technology is expected to monitor and comprehensively analyze indoor temperature, humidity, and indoor personnel in real time based on deep learning technology to achieve real-time monitoring and intelligent regulation of the indoor environment of the house.

[0030] In particular, deep learning is a branch of artificial intelligence, a machine learning technology that mimics the neural network structure of the human brain. It learns the abstract features of data through a multi-level neural network structure to achieve complex pattern recognition and decision-making tasks. Deep learning models can automatically learn features and patterns in data and adapt to various complex tasks and data types. Deep learning technology has a wide range of applications in various fields, can handle complex data and tasks, and provides strong support for the development of artificial intelligence.

[0031] In the indoor environment monitoring system based on intelligent sensing technology in the embodiment of the present application, deep learning technology plays an important role. For example, the indoor temperature and humidity data are monitored and analyzed in real time using a deep learning algorithm, abnormal temperature and humidity conditions are identified, and control measures are taken in a timely manner. Therefore, in the specific embodiment of the present application, the input data is feature extracted and analyzed using deep learning technology.

[0032] In the embodiment of the present application, the indoor data acquisition module 110 is used to collect the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored. It should be understood that temperature and humidity are two important factors affecting human comfort. Appropriate temperature and humidity levels can improve human comfort and make people feel more comfortable. That is, the monitoring and regulation of indoor temperature and humidity are crucial. Therefore, in order to better monitor and analyze indoor temperature and humidity, first collect the temperature at multiple predetermined time points in the room to be monitored and the humidity at multiple predetermined time points. Further, it should be understood that different indoor population densities may lead to uneven indoor temperature distribution; people of different age groups have different sensitivities to temperature and humidity; and people with high activity intensity will generate more heat, thereby affecting the indoor temperature. That is, the distribution, age, behavior and other factors of indoor personnel have a certain influence on the regulation of the indoor environment. Therefore, in order to more accurately adjust the indoor temperature and humidity and improve the comfort of the human body, it is also necessary to collect personnel monitoring images at multiple predetermined time points in the room to be monitored.

[0033] Specifically, when collecting the temperature at multiple predetermined time points in the room to be monitored and the humidity at multiple predetermined time points in the room to be monitored, temperature and humidity sensors can be installed at key locations in the room to be monitored. When collecting personnel monitoring images at multiple predetermined time points in the room to be monitored, cameras can be installed at appropriate locations in the room to be monitored. Personnel monitoring images at multiple predetermined time points can be collected by sensors or timed triggering to ensure that the system obtains comprehensive personnel activity information.

[0034] In the embodiment of the present application, the indoor feature acquisition module 120 is used to extract and analyze the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored to obtain a temperature and humidity feature matrix and a personnel change feature vector. It should be understood that after collecting these input data, the input data is further subjected to feature extraction and analysis tasks.

[0035] Specifically, Figure 2 FIG. 1 is a schematic block diagram of an indoor feature acquisition module in an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application. Figure 2As shown, the indoor feature acquisition module 120 includes: an indoor temperature feature extraction unit 121, which is used to arrange the temperatures of multiple predetermined time points in the room to be monitored according to the time dimension as a temperature input vector and then pass it through a temperature multi-scale feature extraction module to obtain a temperature multi-scale feature vector; an indoor humidity feature extraction unit 122, which is used to arrange the humidity of multiple predetermined time points in the room to be monitored according to the time dimension as a humidity input vector and then pass it through a humidity feature extraction module based on a time series encoder to obtain a humidity time series feature vector; a temperature and humidity correlation feature extraction unit 123, which is used to use a joint encoder to perform feature combination of the temperature multi-scale feature vector and the humidity time series feature vector to obtain the temperature and humidity feature matrix; an indoor personnel feature extraction unit 124, which is used to perform feature encoding on the personnel monitoring images of multiple predetermined time points in the room to be monitored to obtain the personnel change feature vector.

[0036] First, the temperature at multiple predetermined time points in the room to be monitored is feature extracted and analyzed. It should be understood that temperature data usually has obvious time correlation, and the temperature values ​​at previous and subsequent time points may affect each other. Through time series arrangement, the time relationship between the data can be retained, which is conducive to the deep learning model to learn the dynamic change law of time series data. Therefore, the temperatures at multiple predetermined time points in the room to be monitored are first arranged as temperature input vectors according to the time dimension. Furthermore, considering that temperature data may have different time scale characteristics, such as short-term fluctuations and long-term trend changes. Therefore, in order to capture feature information at different time scales, the temperature multi-scale feature extraction module is used to perform multi-scale feature extraction operations on the temperature input vector to obtain a temperature multi-scale feature vector. In this way, it helps the model to understand the temperature data more comprehensively.

[0037] In a specific embodiment of the present application, the indoor temperature feature extraction unit 121 is used to: use the first convolution layer of the temperature multi-scale feature extraction module to perform one-dimensional convolution encoding on the temperature input vector to obtain the temperature first-scale feature vector; use the second convolution layer of the temperature multi-scale feature extraction module to perform one-dimensional convolution encoding on the temperature input vector to obtain the temperature second-scale feature vector; and cascade the temperature first-scale feature vector and the temperature second-scale feature vector to obtain the temperature multi-scale feature vector.

[0038] Then, the humidity at multiple predetermined time points in the room to be monitored is feature extracted and analyzed. It should be understood that the indoor humidity changes over time, that is, the humidity in the room to be monitored has time series variation characteristics. Therefore, in order to better retain these time series characteristics, the humidity at multiple predetermined time points in the room to be monitored is first arranged as a humidity input vector according to the time dimension. Further, considering that the time series encoder can help the system better capture the time series information in the time series data, including the time dependency between data points and the patterns in the sequence. Therefore, in order to more effectively analyze and understand the time series characteristics of indoor humidity data and provide more accurate and reliable support for real-time monitoring and intelligent control, a time series encoder is used to extract and analyze the time series characteristics of the humidity input vector to obtain a humidity time series feature vector.

[0039] In a specific embodiment of the present application, the indoor humidity feature extraction unit 122 is used to: use the fully connected layer of the time series encoder to fully connect the humidity input vector to extract the high-dimensional implicit features of the eigenvalues ​​at each position in the humidity input vector; and use the one-dimensional convolution layer of the time series encoder to perform one-dimensional convolution encoding on the humidity input vector to extract the high-dimensional implicit correlation features of the correlation between the eigenvalues ​​at each position in the humidity input vector.

[0040] Then, the correlation features between the temperature and humidity in the monitored room are extracted and analyzed. It should be understood that temperature and humidity are two important parameters in the indoor environment, and there is usually a certain correlation between the two. Fusion of temperature and humidity features can provide more comprehensive and integrated indoor environment features, which helps to more accurately describe and understand the state of the indoor environment. Therefore, in order to better monitor the indoor environment, the temperature multi-scale feature vector and the humidity time series feature vector are feature fused. Further, considering that the joint encoder can process temperature and humidity features at the same time, extract the correlation features between the two, so as to obtain a feature representation with more information richness and characterization capability. Therefore, the joint encoder is used to perform feature fusion on the temperature multi-scale feature vector and the humidity time series feature vector. Specifically, in an embodiment of the present application, a joint encoder is used to perform feature fusion on the temperature multi-scale feature vector and the humidity time series feature vector to obtain the temperature and humidity feature matrix.

[0041] In a specific embodiment of the present application, the temperature and humidity associated feature extraction unit 123 is used to: use the joint encoder to jointly encode the temperature multi-scale feature vector and the humidity time series feature vector using the following formula to generate the temperature and humidity feature matrix; wherein the formula is: in Represents vector multiplication, M a Represents the temperature and humidity feature matrix, Va1 represents the temperature multi-scale eigenvector, V a2 represents the humidity time series feature vector, represents the transpose of the humidity time series feature vector.

[0042] Finally, feature extraction and analysis are performed on the personnel monitoring images at multiple predetermined time points in the room to be monitored. Specifically, Figure 3 The figure shows a schematic block diagram of an indoor personnel feature extraction unit in an indoor feature acquisition module in an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application. Figure 3 As shown, the indoor personnel feature extraction unit 124 includes: an image denoising subunit 124-1, which is used to pass the personnel monitoring images of multiple predetermined time points in the monitored room through an image denoising module based on an autoencoder to obtain multiple denoised personnel monitoring images; a personnel feature extraction subunit 124-2, which is used to pass the multiple denoised personnel monitoring images through a personnel feature extractor based on a deep and shallow feature fusion module to obtain multiple personnel feature matrices; a personnel change feature extraction subunit 124-3, which is used to arrange the multiple personnel feature matrices into personnel change input tensors according to the time dimension and then pass them through a personnel change feature extraction module based on a spatial attention mechanism to obtain the personnel change feature vector.

[0043] It should be understood that monitoring the indoor environment requires accurate and clear personnel monitoring images for analysis and identification. Noise reduction processing can remove noise and interference in the image and improve the clarity and quality of the image. Therefore, firstly, an image denoising module based on an autoencoder is used to perform noise reduction processing on the personnel monitoring images at multiple predetermined time points in the room to be monitored. Among them, the autoencoder has a certain generalization ability when learning data representation, and can adapt to different types and degrees of noise, so that noise reduction processing can be effectively performed in different environments.

[0044] In a specific embodiment of the present application, the image denoising subunit 124-1 is used to: use the feature encoder of the image denoising module to extract multiple image features from the personnel monitoring images of multiple predetermined time points in the monitored room; and use the feature decoder of the image denoising module to decode the multiple image features to obtain the multiple denoised personnel monitoring images.

[0045] Next, it should be understood that the denoised personnel monitoring image contains not only shallow features such as image edges, textures, and colors, but also more advanced and abstract deep features such as body posture, facial expressions, and walking movements of the personnel. Therefore, in order to more comprehensively capture the feature information of indoor personnel, a personnel feature extractor based on a deep and shallow feature fusion module is used to extract and analyze the deep and shallow fusion features of the multiple denoised personnel monitoring images to obtain the multiple personnel feature matrices. In this way, the features can be made more complementary through fusion, and the representation ability and robustness of the features can be improved.

[0046] In a specific embodiment of the present application, the personnel feature extraction subunit 124-2 is used to: extract a shallow feature matrix from the i-th layer of the personnel feature extractor, the i-th layer being the first to sixth layers of the personnel feature extractor; extract a deep feature matrix from the j-th layer of the personnel feature extractor, the ratio between the j-th layer and the i-th layer being greater than or equal to 5; and use the shallow and deep feature fusion module to fuse the shallow feature matrix and the deep feature matrix to obtain each of the multiple personnel feature matrices.

[0047] Next, considering the evolution of personnel behavior over time, based on this, multiple personnel feature matrices are first arranged into personnel change input tensors according to the time dimension. In this way, the characteristic change information of personnel at different time points can be retained, which helps to capture the evolution and change trend of personnel behavior over time. Then, a personnel change feature extraction module based on the spatial attention mechanism is used to extract features from the personnel change input tensor to obtain the personnel change feature vector. Among them, the spatial attention mechanism can dynamically adjust the attention weights of different positions according to different contexts and needs, so that the model can better adapt to the personnel change feature extraction needs in different situations.

[0048] In a specific embodiment of the present application, the personnel change feature extraction subunit 124-3 is used to: use the convolution encoding part of the personnel change feature extraction module to perform deep convolution encoding on the personnel change input tensor to obtain an initial convolution feature map; input the initial convolution feature map into the spatial attention part of the personnel change feature extraction module to obtain a spatial attention map; pass the spatial attention map through a Softmax activation function to obtain a spatial attention feature map; calculate the position point multiplication of the spatial attention feature map and the initial convolution feature map to obtain a personnel change feature map; and perform a global pooling operation based on a feature matrix on the personnel change feature map to obtain the personnel change feature vector.

[0049] In an embodiment of the present application, the indoor feature fusion module 130 is used to multiply the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector. It should be understood that factors such as the distribution, age, and behavior of indoor personnel have a certain influence on environmental regulation. By fusing the temperature and humidity feature matrix and the personnel change feature vector, the correlation between temperature and humidity and personnel behavior can be better analyzed, and the potential laws and influences between the two can be discovered. Therefore, in order to obtain more comprehensive and accurate indoor environmental information and provide more powerful support for real-time monitoring and intelligent control, the temperature and humidity feature matrix and the personnel change feature vector are multiplied to obtain an indoor environment feature vector.

[0050] In the embodiment of the present application, the indoor feature classification result generating module 140 is used to obtain a classification result based on the indoor environment feature vector, and the classification result is used to indicate whether the temperature in the room to be monitored should be increased or decreased, and whether the humidity should be increased or decreased. Specifically, Figure 4 FIG. 1 is a schematic block diagram of an indoor feature classification result generation module in an indoor environment monitoring system based on intelligent sensing technology according to an embodiment of the present application. Figure 4 As shown, the indoor feature classification result generation module 140 includes: an indoor environment feature extraction unit 141, used to pass the indoor environment feature vector through an indoor environment feature extraction module based on a convolutional neural network model to obtain an indoor environment classification feature vector; an indoor environment feature optimization unit 142, used to perform expected activation activation in parameter space on the indoor environment classification feature vector to obtain an optimized indoor environment classification feature vector; an indoor environment feature classification unit 143, used to pass the optimized indoor environment classification feature vector through a classifier to obtain the classification result.

[0051] It should be understood that only feature fusion may make the feature expression not rich enough, affecting the system's accurate understanding and classification of the indoor environment. Therefore, in order to capture more advanced abstract features, a convolutional neural network model is used to extract features from the indoor environment feature vector to obtain the indoor environment classification feature vector. Among them, the convolutional neural network model gradually extracts abstract features of the data through multi-layer convolution and pooling operations, which helps to capture complex patterns in the data. In this way, by using the convolutional neural network model for feature extraction, the system can more effectively extract meaningful features from the indoor environment feature vector, providing more powerful support for subsequent indoor environment classification and intelligent control.

[0052] In particular, in the technical solution of the present application, it is considered that if there is noise, missing or inconsistency in the data of temperature, humidity or personnel monitoring images during the acquisition process, these quality problems will directly affect the accuracy of feature extraction, thereby causing local abnormal distribution in the indoor environment classification feature vector. The temperature multi-scale feature extraction module, the humidity feature extraction module based on the temporal encoder, and the personnel feature extraction module based on the deep and shallow feature fusion module may not be able to fully capture all relevant environmental changes and personnel behavior patterns. This limitation may lead to the loss of some important information in the feature vector, resulting in local anomalies. The local abnormal distribution may mislead the classifier, causing it to be unable to accurately judge the state of the indoor environment. For example, abnormal features may be mistakenly interpreted as a signal of some environmental change, resulting in incorrect classification results. The indoor environment classification feature vector is key information for guiding indoor temperature and humidity adjustment decisions. If there is an abnormality in the indoor environment classification feature vector, it may lead to wrong adjustment decisions, such as adjusting the temperature or humidity when it is not necessary, or failing to respond in time when adjustment is required. Based on this, in order to solve the problem of local abnormal distribution in the indoor environment classification feature vector, in the technical solution of the present application, the indoor environment classification feature vector is activated in the parameter space.

[0053] In a specific embodiment of the present application, the indoor environment feature optimization unit 142 includes: calculating the natural exponential function value of the eigenvalue at the predetermined position of the indoor environment classification feature vector raised to the power and then subtracting one to obtain an adjustment index eigenvalue; calculating the adjustment index eigenvalue divided by the eigenvalue at the predetermined position of the indoor environment classification feature vector, and taking the absolute value of the calculation result of the adjustment index eigenvalue to obtain an adjustment index eigenvalue ratio; activating the adjustment index eigenvalue ratio through a sigmoid function to obtain an activated adjustment index eigenvalue ratio; calculating the natural exponential function value of the opposite number of the eigenvalue at the predetermined position of the indoor environment classification feature vector raised to the power. The method comprises the following steps: calculating the eigenvalue of a predetermined position of the indoor environment classification feature vector, dividing the eigenvalue by the exponential attenuation coefficient of the eigenvalue and then subtracting one, and taking the absolute value of the calculated result of the exponential attenuation coefficient to obtain the adjusted attenuation difference of the eigenvalue; activating the adjusted attenuation difference of the eigenvalue by a sigmoid function to obtain the adjusted attenuation difference of the activated eigenvalue; calculating the weighted sum of the activated adjusted exponential eigenvalue ratio and the activated eigenvalue adjusted attenuation difference, and multiplying the weighted calculated result by the eigenvalue of the predetermined position of the indoor environment classification feature vector to obtain the eigenvalue of the predetermined position of the optimized indoor environment classification feature vector.

[0054] Here, in view of the local abnormal distribution in the indoor environment classification feature vector, in the technical solution of the present application, the indoor environment classification feature vector is activated in the parameter space of the expected activation, which activates the feature vector elements that have an important impact on the classification decision in an expected manner by dynamically adjusting the parameter space. When implementing the expected activation, not only the importance of a single feature vector element is considered, but also the interaction and synergy between the feature vector elements are considered. Specifically, by performing the expected activation of the parameter space of the indoor environment classification feature vector, it is possible to automatically identify and enhance those feature vector elements that contribute more to the model performance, and dynamically adjust the parameter space according to the current model's learning state and the activation of the feature vector elements to maintain the activity and diversity of the feature representation. In this way, the problem of reduced classification performance due to local abnormal distribution can be effectively alleviated, and the sensitivity and adaptability of the model to key features in the data can also be improved, so that a high classification accuracy and robustness can be maintained in various complex data environments.

[0055] Next, considering that the classifier can classify according to the input feature vector, it helps the system make decisions, such as judging whether the state of the indoor environment needs to be adjusted, and the adjustment method. Therefore, the classifier is used to perform feature classification on the optimized indoor environment classification feature vector to obtain a classification result, and the classification result is used to indicate that the temperature in the room to be monitored should be increased or decreased, and the humidity should be increased or decreased. In this way, according to the classification result, intelligent regulation of the indoor environment of the house can be achieved.

[0056] In a specific embodiment of the present application, the indoor environment feature classification unit 143 is used to: use the classifier to process the optimized indoor environment classification feature vector using the following formula to obtain the classification result;

[0057] Wherein, the formula is: O = softmax{(W b ,B n ):…:(W 1 ,B 1 )|V}, where W 1 To W n is the weight matrix, B 1 To B n is a bias vector, V is an optimized indoor environment classification feature vector, softmax represents a normalized exponential function, and O represents the classification result.

[0058] It is worth mentioning that, in addition to using a classifier to perform feature classification on the optimized indoor environment classification feature vector, a regression model can also be used. The regression model can be used to predict continuous numerical outputs, and is suitable for situations where numerical results need to be predicted.

[0059] The specific implementation process is as follows: 1. Data preparation: Collect and prepare the monitored optimized indoor environment feature vector data and the corresponding temperature and humidity adjustment values. 2. Model training: Use regression models (such as linear regression, support vector regression, etc.) to train the feature vectors and the corresponding temperature and humidity adjustment values, and establish the relationship between the feature vectors and the temperature and humidity adjustment values. 3. Model prediction: When the system monitors a new optimized indoor environment feature vector, use the trained regression model to predict and obtain the corresponding temperature and humidity adjustment values. 4. Control implementation: According to the temperature and humidity adjustment values ​​predicted by the regression model, the system can automatically or manually adjust the indoor temperature and humidity to achieve intelligent control of the indoor environment.

[0060] In summary, the indoor environment monitoring system based on intelligent sensing technology according to the embodiment of the present application is explained, which first collects the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored as input data, and then uses deep learning technology to extract and analyze these input data to obtain classification results indicating that the temperature in the room to be monitored should increase or decrease, and the humidity should increase or decrease. In this way, it is possible to monitor and comprehensively analyze the indoor environment and indoor personnel in real time, thereby realizing real-time monitoring and intelligent control of the indoor environment of the house. In this way, it is helpful to adjust the indoor comfort to the greatest extent.

[0061] As described above, the indoor environment monitoring system 100 based on intelligent sensing technology according to the embodiment of the present application can be implemented in various wireless terminals, such as a server for indoor environment monitoring based on intelligent sensing technology. In one example, the indoor environment monitoring system 100 based on intelligent sensing technology according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the indoor environment monitoring system 100 based on intelligent sensing technology can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the indoor environment monitoring system 100 based on intelligent sensing technology can also be one of the many hardware modules of the wireless terminal.

[0062] Alternatively, in another example, the indoor environment monitoring system 100 based on intelligent sensing technology and the wireless terminal may also be separate devices, and the indoor environment monitoring system 100 based on intelligent sensing technology may be connected to the wireless terminal via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0063] Exemplary Methods

[0064] Figure 5The figure shows a flow chart of an indoor environment monitoring method based on intelligent sensing technology according to an embodiment of the present application. Figure 5 As shown, according to the indoor environment monitoring method based on intelligent sensing technology according to the embodiment of the present application, the method includes: S1, collecting the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored; S2, respectively extracting and analyzing the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored to obtain a temperature and humidity feature matrix and a personnel change feature vector; S3, multiplying the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector; S4, obtaining a classification result based on the indoor environment feature vector, and the classification result is used to indicate that the temperature in the room to be monitored should increase or decrease, and the humidity should increase or decrease.

[0065] Here, those skilled in the art can understand that the specific functions and operations of each step in the indoor environment monitoring method based on intelligent sensing technology have been referred to above. Figure 1 The indoor environment monitoring system based on intelligent sensing technology has been introduced in detail, and therefore, its repeated description will be omitted.

[0066] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, apparatuses and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0067] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0068] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0069] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0070] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or systems stated in a system claim can also be implemented by one unit or system through software or hardware. Secondary words are used to indicate names, but not to indicate any particular order.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit of the technical solution of the present invention.

Claims

1. An indoor environment monitoring system based on intelligent sensing technology, characterized in that: include: An indoor data acquisition module, used to collect the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored; An indoor feature acquisition module, used for extracting and analyzing the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored, respectively, to obtain a temperature and humidity feature matrix and a personnel change feature vector; An indoor feature fusion module, used for multiplying the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector; An indoor feature classification result generating module, used to obtain a classification result based on the indoor environment feature vector, wherein the classification result is used to indicate whether the temperature in the room to be monitored should be increased or decreased, and whether the humidity should be increased or decreased; Wherein, the indoor feature acquisition module includes: An indoor temperature feature extraction unit, used for arranging the temperatures of a plurality of predetermined time points in the room to be monitored according to the time dimension as a temperature input vector and then passing it through a temperature multi-scale feature extraction module to obtain a temperature multi-scale feature vector; An indoor humidity feature extraction unit, used for arranging the humidity of a plurality of predetermined time points in the room to be monitored according to the time dimension as a humidity input vector and then passing it through a humidity feature extraction module based on a time series encoder to obtain a humidity time series feature vector; A temperature and humidity associated feature extraction unit, configured to use a joint encoder to perform feature combination on the temperature multi-scale feature vector and the humidity time series feature vector to obtain the temperature and humidity feature matrix; An indoor personnel feature extraction unit, used for performing feature encoding on the personnel monitoring images at multiple predetermined time points in the room to be monitored to obtain the personnel change feature vector; Wherein, the indoor feature classification result generation module includes: An indoor environment feature extraction unit, used for passing the indoor environment feature vector through an indoor environment feature extraction module based on a convolutional neural network model to obtain an indoor environment classification feature vector; An indoor environment feature optimization unit, used for performing expected activation of the parameter space of the indoor environment classification feature vector to obtain an optimized indoor environment classification feature vector; An indoor environment feature classification unit, used for passing the optimized indoor environment classification feature vector through a classifier to obtain the classification result; Wherein, the indoor environment characteristic optimization unit comprises: Calculate a natural exponential function value raised to a power of the eigenvalue at a predetermined position of the indoor environment classification eigenvector and then subtract one to obtain an adjusted exponential eigenvalue; Calculating the adjustment index eigenvalue divided by the eigenvalue of a predetermined position of the indoor environment classification eigenvector, and taking an absolute value of the calculation result of the adjustment index eigenvalue to obtain an adjustment index eigenvalue ratio; Activating the adjustment index eigenvalue ratio through a sigmoid function to obtain an activated adjustment index eigenvalue ratio; Calculate the natural exponential function value of the inverse number of the eigenvalue at the predetermined position of the indoor environment classification eigenvector to obtain the exponential decay coefficient of the eigenvalue; Calculate the characteristic value of the predetermined position of the indoor environment classification characteristic vector, divide it by the exponential decay coefficient of the characteristic value and then subtract one, and take the absolute value of the calculation result of the exponential decay coefficient to obtain the adjusted attenuation difference of the characteristic value; Activating the adjusted attenuation difference of the eigenvalues ​​through a sigmoid function to obtain an activated eigenvalue adjusted attenuation difference; A weighted sum of the activation adjustment index eigenvalue ratio and the activation eigenvalue adjustment attenuation difference is calculated, and the weighted calculation result is multiplied by the eigenvalue of the predetermined position of the indoor environment classification feature vector to obtain the eigenvalue of the predetermined position of the optimized indoor environment classification feature vector.

2. The indoor environment monitoring system based on intelligent sensing technology according to claim 1 is characterized in that: The temperature and humidity associated feature extraction unit is used to: use the joint encoder to jointly encode the temperature multi-scale feature vector and the humidity time series feature vector according to the following formula to generate the temperature and humidity feature matrix; Wherein, the formula is: in Represents vector multiplication, M a Represents the temperature and humidity feature matrix, V a1 represents the temperature multi-scale eigenvector, V a2 represents the humidity time series feature vector, represents the transpose of the humidity time series feature vector.

3. The indoor environment monitoring system based on intelligent sensing technology according to claim 2 is characterized in that: The indoor personnel feature extraction unit comprises: An image denoising subunit, configured to pass the personnel monitoring images at a plurality of predetermined time points in the room to be monitored through an image denoising module based on an autoencoder to obtain a plurality of denoised personnel monitoring images; A personnel feature extraction subunit, configured to respectively pass the plurality of denoised personnel monitoring images through a personnel feature extractor based on a deep and shallow feature fusion module to obtain a plurality of personnel feature matrices; The personnel change feature extraction subunit is used to arrange the multiple personnel feature matrices into a personnel change input tensor according to the time dimension and then obtain the personnel change feature vector through a personnel change feature extraction module based on the spatial attention mechanism.

4. The indoor environment monitoring system based on intelligent sensing technology according to claim 3 is characterized in that: The image denoising subunit is used for: Extracting multiple image features from the personnel monitoring images at multiple predetermined time points in the room to be monitored using the feature encoder of the image noise reduction module; as well as The plurality of image features are decoded using a feature decoder of the image denoising module to obtain the plurality of denoised personnel monitoring images.

5. The indoor environment monitoring system based on intelligent sensing technology according to claim 4 is characterized in that: The personnel feature extraction subunit is used to: Extracting a shallow feature matrix from the i-th layer of the personnel feature extractor, wherein the i-th layer is the first to sixth layers of the personnel feature extractor; Extracting a deep feature matrix from the jth layer of the personnel feature extractor, wherein a ratio between the jth layer and the ith layer is greater than or equal to 5; as well as The shallow feature fusion module is used to fuse the shallow feature matrix and the deep feature matrix to obtain each of the multiple personnel feature matrices.

6. A method for indoor environment monitoring based on intelligent sensing technology, using the indoor environment monitoring system based on intelligent sensing technology according to claim 1, characterized in that ,include: Collecting the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored; Extracting and analyzing the temperature at multiple predetermined time points in the room to be monitored, the humidity at multiple predetermined time points in the room to be monitored, and the personnel monitoring images at multiple predetermined time points in the room to be monitored to obtain a temperature and humidity feature matrix and a personnel change feature vector; Multiplying the temperature and humidity feature matrix and the personnel change feature vector to obtain an indoor environment feature vector; A classification result is obtained based on the indoor environment feature vector, and the classification result is used to indicate that the temperature in the room to be monitored should be increased or decreased, and the humidity should be increased or decreased.

Images