An intelligent indoor disinfection effect prediction method and system

Abstract

The application discloses an intelligent indoor disinfection effect prediction method and system, and the method comprises disinfection data collection, data preliminary optimization, disinfection data embedding, effect prediction model construction and disinfection effect prediction. The application relates to the technical field of disinfection effect data processing, and particularly discloses an intelligent indoor disinfection effect prediction method and system. The application obtains original disinfection effect data through disinfection data collection; adopts a data optimization method of missing value processing, space-time data alignment, data standardization and data set segmentation; adopts a material physical property characteristic embedding method, reduces the influence of local indoor material differences, and improves the prediction performance of the model through disinfection parameters and material physical property interaction; adopts a physical enhanced graph convolution network model as an effect prediction model, combines disinfectant forward diffusion process characteristics and disinfectant reverse residual process characteristics, and solves the deviation caused by disinfection effect lag through a time compensation mechanism.

Description

Technical Field

[0001] This invention relates to the field of disinfection effect data processing technology, specifically to an intelligent method and system for predicting indoor disinfection effects. Background Technology

[0002] Indoor disinfection effectiveness prediction refers to the prediction of the effect of disinfection treatment in an indoor environment through scientific methods and models. It aims to assess the effectiveness of disinfection methods in eliminating pathogenic microorganisms and the efficiency of disinfectant use. It can help to formulate scientific and reasonable disinfection strategies, improve the hygiene and safety of public places or home environments, and reduce the risk of infectious disease transmission. It has important application value, especially in disease prevention and control and hospital infection control.

[0003] However, traditional methods for predicting indoor disinfection effectiveness suffer from several drawbacks. They rely on a single temporal convolution kernel or a fixed period assumption, which fails to distinguish between the long-term trend and short-term fluctuations of disinfectant diffusion. Furthermore, spatial modeling depends on uniform grid partitioning, neglecting the impact of the non-uniform distribution of material properties within the region on the disinfection effect. Traditional methods also fail to model the reverse process of disinfectant degradation and cannot capture the nonlinear effects of environmental parameters such as UV intensity, temperature, and humidity on the lag in disinfection effectiveness, resulting in errors in time delay compensation. Summary of the Invention

[0004] To address the above issues and overcome the shortcomings of existing technologies, this invention provides an intelligent method and system for predicting indoor disinfection effects. Traditional methods for predicting indoor disinfection effects suffer from limitations such as using a single temporal convolution kernel or fixed-period assumptions, failing to distinguish between long-term trends and short-term fluctuations in disinfectant diffusion, and relying on uniform grid partitioning for spatial modeling, neglecting the impact of non-uniform distribution of material properties within the region on disinfection effectiveness. This solution creatively employs a method of embedding material property features. Through multi-dimensional temporal decomposition, multi-granular spatial representation, and embedded material properties, the influence of local indoor material differences is reduced, and the model's predictive performance is improved through the interaction between disinfection parameters and material properties. Furthermore, traditional methods for predicting indoor disinfection effects cannot model the reverse process of disinfectant degradation and cannot capture the nonlinear effects of environmental parameters such as UV intensity, temperature, and humidity on the lag in disinfection effectiveness, resulting in errors in time delay compensation. This solution creatively uses a physically enhanced graph convolutional network model as the effect prediction model. Under conditions conforming to physical laws, it fully combines the characteristics of the forward diffusion process and the reverse residue process of disinfectant, and adjusts for historical cumulative effects through a time compensation mechanism, thus solving the bias caused by the lag in disinfection effectiveness.

[0005] The technical solution adopted by this invention is as follows: This invention provides an intelligent method for predicting indoor disinfection effects, which includes the following steps:

[0006] Step S1: Disinfection data collection;

[0007] Step S2: Preliminary data optimization;

[0008] Step S3: Embedding disinfection data;

[0009] Step S4: Building the effect prediction model;

[0010] Step S5: Prediction of disinfection effect.

[0011] Further, in step S1, the disinfection data acquisition is used to collect the raw data required to predict the indoor disinfection effect. Specifically, it involves collecting disinfection data from the disinfection log system, building information system, and environmental sensor system to obtain the raw disinfection effect dataset. The raw disinfection effect dataset specifically includes a historical disinfection effect raw set and a real-time disinfection effect raw set. Both the historical disinfection effect raw set and the real-time disinfection effect raw set include spatiotemporal reference data, disinfection-related data, and environmental data. The historical disinfection effect raw set also includes historical disinfection effect label data.

[0012] Further, in step S2, the data is initially optimized, which is used to preprocess the collected raw disinfection effect data to optimize the data, specifically including the following steps:

[0013] Step S21: Missing value processing, used to remove missing values, specifically to remove missing values ​​from the original set of historical disinfection effects and the original set of real-time disinfection effects, to obtain a coarsely processed historical dataset and a coarsely processed real-time dataset.

[0014] Step S22: Spatiotemporal data alignment, used to align spatiotemporal data, specifically converting the spatial data in the coarse-processed historical dataset and the coarse-processed real-time dataset into a unified coordinate system, and synchronizing the time data with a clock to obtain an aligned historical dataset and an aligned real-time dataset.

[0015] Step S23: Data standardization, used to standardize the aligned data, specifically by using the min-max standardization method to process the aligned historical dataset and the aligned real-time dataset to obtain the standardized historical dataset and the standardized real-time dataset;

[0016] Step S24: Dataset splitting, specifically splitting the standardized historical dataset into a prediction training set and a prediction test set.

[0017] Furthermore, in step S3, the disinfection data embedding is used to model the physical interaction between disinfection parameters and indoor building materials, specifically including the following steps;

[0018] Step S31: Time dimension decomposition, used to separate the multi-scale features of the time series, specifically by decomposing the time dimension features into trend, periodic and transient components;

[0019] Step S32: Spatial hierarchical coding, used to construct multi-granularity spatial representations, specifically spatial hierarchical coding at the grid level, region level, and global level;

[0020] Step S33: Parameter cross-embedding, used for interactive disinfection parameters and indoor building material physical parameters, the formula used is as follows:

[0021] ;

[0022] In the formula, Indicates the interactive characteristics of disinfection materials. This represents the softmax function. Indicates the characteristics of disinfection parameters. This indicates the property characteristics of indoor building materials, and their relationship with disinfection parameters. Same dimensions This indicates the feature dimensions of the disinfection parameters;

[0023] Step S34: Construct a feature set, specifically by processing the standardized real-time dataset, the prediction training set, and the prediction test set through the time dimension decomposition, the spatial hierarchical encoding, and the parameter cross-embedding to obtain a real-time feature set, a training feature set, and a test feature set.

[0024] Further, in step S4, the effect prediction model is constructed to build a model required to predict the indoor disinfection effect, specifically by constructing a physical augmentation graph convolutional network model as the effect prediction model;

[0025] The construction of the effect prediction model specifically includes the following steps:

[0026] Step S41: Design an activation function to introduce physical constraints, specifically by designing a physical enhancement activation function, the specific formula of which is as follows:

[0027] ;

[0028] In the formula, Let x represent the physics-enhanced activation function, and let x represent the independent variable of the physics-enhanced activation function. This represents the ventilation rate data for region j. This represents the absorption coefficient of the material in region j. This represents the sigmoid function. This represents the Euclidean distance from region m to region j. This represents the ambient humidity data for region m. This represents the dynamic sharpness adjustment coefficient for region j. Represents the ReLU function. Indicates the concentration of disinfectant in region j. Indicates the saturation concentration of the disinfectant;

[0029] Step S42: Design a forward diffusion process to simulate the active propagation of disinfectant under ventilation and material constraints. The steps include:

[0030] Step S421: Calculate the forward dynamic adjacency matrix using the following formula:

[0031] ;

[0032] In the formula, This represents a forward dynamic adjacency matrix. This indicates that the weight matrix is ​​calculated using the forward dynamic adjacency matrix. This represents the connectivity characteristics between region m and region j. This represents the airflow velocity data from region m to region j. Indicates the distance decay index. This represents the Euclidean distance from region m to region j. of Power of 1 This represents the connectivity characteristic from region m to region j. This represents an indicator function whose value is 1 when there is a connection between region m and region j, and whose value is 0 when there is no connection between region m and region j.

[0033] Step S422: Implement physical constraints, using the following formula:

[0034] ;

[0035] In the formula, This indicates the amount of disinfectant released in region m. This represents the volume of region m. Indicates the density of the disinfectant. This indicates a small value used to prevent division by zero. This represents the positive dynamic adjacency matrix under constraints, where j represents the index of the region. This represents the total number of regions that are connected to region m.

[0036] Step S423: Forward graph convolution update;

[0037] Step S43: Design reverse residual transmission to simulate the passive diffusion of disinfectant under the influence of concentration gradient and personnel contact. The steps include:

[0038] Step S431: Calculate the half-life of the disinfectant component;

[0039] Step S432: Calculate the reverse dynamic adjacency matrix using the following formula:

[0040] ;

[0041] In the formula, Represents the reverse dynamic adjacency matrix. This indicates the concentration of disinfectant in area m. This indicates the half-life of the disinfectant component in region j. This indicates the half-life of the disinfectant component in region n. This represents the frequency of personnel contact between region m and region j. Let represent the surface area of ​​region j, and n represent the index of the region that is distinct from j.

[0042] Step S433: Inverse graph convolution update;

[0043] Step S44: Bidirectional gating fusion, used to balance forward propagation and backward residual features, specifically by designing a gating mechanism to fuse forward propagation features and backward residual features;

[0044] Step S45: Obtain the model output using the following formula:

[0045] ;

[0046] In the formula, This indicates the model's initial prediction output. Indicates the model output weights. Indicates fusion characteristics, express Time delay factor, This indicates that the delay coefficient is used to calculate the weight. Represents multi-scale temporal features. express UV intensity at any given time This represents the model's predicted output at time t, and Tm represents the total number of time points. express Preliminary prediction output from the time-matter model;

[0047] Step S46: Model construction and training, specifically, constructing a physical augmented graph convolutional network model through the design activation function, the design forward diffusion process, the design backward residual propagation, the bidirectional gating fusion, and the acquisition of model output; training the model based on the training feature set; verifying the model performance based on the test feature set; obtaining the physical augmented graph convolutional network model, which serves as the performance prediction model.

[0048] Further, in step S5, the disinfection effect prediction specifically involves using the effect prediction model to predict the indoor disinfection effect based on the real-time feature set, obtaining disinfection effect prediction reference data, and comprehensively evaluating the indoor disinfection effect based on the disinfection effect prediction reference data.

[0049] The present invention provides an intelligent indoor disinfection effect prediction system, including a disinfection data acquisition module, a preliminary data optimization module, a disinfection data embedding module, an effect prediction model construction module, and a disinfection effect prediction module;

[0050] The disinfection data acquisition module is used for disinfection data acquisition. Through disinfection data acquisition, the original disinfection effect dataset is obtained, and the original disinfection effect dataset is sent to the data preliminary optimization module.

[0051] The data preliminary optimization module is used for preliminary data optimization. Through preliminary data optimization, a standardized real-time dataset, a prediction training set, and a prediction test set are obtained, and the standardized real-time dataset, the prediction training set, and the prediction test set are sent to the disinfection data embedding module.

[0052] The disinfection data embedding module is used for disinfection data embedding. Through disinfection data embedding, the physical interaction between disinfection parameters and indoor building materials is modeled to obtain a real-time feature set, a training feature set, and a test feature set. The real-time feature set is sent to the disinfection effect prediction module, and the training feature set and the test feature set are sent to the effect prediction model construction module.

[0053] The effect prediction model construction module is used to construct the effect prediction model. It obtains the effect prediction model by constructing a physical augmentation graph convolutional network model and sends the effect prediction model to the disinfection effect prediction module.

[0054] The disinfection effect prediction module is used to predict the disinfection effect. By using the effect prediction model to predict the indoor disinfection effect, reference data for the disinfection effect prediction is obtained.

[0055] The beneficial effects achieved by adopting the above solution are as follows:

[0056] (1) Traditional indoor disinfection effect prediction methods have technical problems such as using a single temporal convolution kernel or fixed period assumption, which cannot distinguish the long-term trend and short-term fluctuation of disinfectant diffusion, and spatial modeling relies on uniform grid division, ignoring the impact of non-uniform distribution of material properties in the region on the disinfection effect. This solution creatively adopts the method of embedding material property features. Through multi-dimensional temporal decomposition, multi-granular spatial representation and embedding material properties, the influence of local indoor material differences is reduced, and the prediction performance of the model is improved through the interaction of disinfection parameters and material properties.

[0057] (2) In view of the technical problems that traditional indoor disinfection effect prediction methods cannot model the reverse process of disinfectant degradation and cannot capture the nonlinear effects of environmental parameters such as ultraviolet intensity, temperature and humidity on the lag of disinfection effect, and have errors in time delay compensation, this solution creatively adopts the physical enhancement graph convolutional network model as the effect prediction model. Under the condition of conforming to physical laws, it can fully combine the characteristics of the forward diffusion process of disinfectant and the characteristics of the reverse residue process of disinfectant, and adjust the historical cumulative effect through the time compensation mechanism, thus solving the deviation caused by the lag of disinfection effect. Attached Figure Description

[0058] Figure 1 A flowchart illustrating an intelligent method for predicting indoor disinfection effects provided by this invention;

[0059] Figure 2 A schematic diagram of a module for predicting the effect of indoor disinfection provided by the present invention;

[0060] Figure 3 A flowchart illustrating the initial data optimization process in step S2;

[0061] Figure 4 A flowchart illustrating the process of embedding disinfection data in step S3;

[0062] Figure 5 A flowchart illustrating the process of building the effect prediction model for step S4.

[0063] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0064] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0065] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0066] Example 1, see Figure 1 The technical solution adopted by this invention is as follows: This invention provides an intelligent method for predicting indoor disinfection effects, which includes the following steps:

[0067] Step S1: Disinfection data collection;

[0068] Step S2: Preliminary data optimization;

[0069] Step S3: Embedding disinfection data;

[0070] Step S4: Building the effect prediction model;

[0071] Step S5: Prediction of disinfection effect.

[0072] Example 2, see Figure 1 and Figure 2 In step S1, the disinfection data acquisition is used to collect the raw data required to predict the indoor disinfection effect. Specifically, it involves collecting disinfection data from the disinfection log system, building information system, and environmental sensor system to obtain the raw disinfection effect dataset. The raw disinfection effect dataset specifically includes a historical disinfection effect raw set and a real-time disinfection effect raw set. Both the historical disinfection effect raw set and the real-time disinfection effect raw set include spatiotemporal reference data, disinfection-related data, and environmental data. The historical disinfection effect raw set also includes historical disinfection effect label data. The spatiotemporal reference data specifically includes disinfection operation timestamp data, indoor area spatial division data, door and window connection data, and ventilation duct layout data. The disinfection-related data specifically includes disinfectant type data and disinfectant spraying data. The environmental data specifically includes indoor building material property data, ambient temperature data, ambient humidity data, ambient airflow velocity data, and indoor ventilation rate data.

[0073] Example 3, see Figure 1 , Figure 2 and Figure 3 This embodiment is based on the above embodiment. In step S2, the data is initially optimized to preprocess the collected raw disinfection effect data to optimize the data. Specifically, it includes the following steps:

[0074] Step S21: Missing value processing, used to remove missing values, specifically to remove missing values ​​from the original set of historical disinfection effects and the original set of real-time disinfection effects, to obtain a coarsely processed historical dataset and a coarsely processed real-time dataset.

[0075] Step S22: Spatiotemporal data alignment, used to align spatiotemporal data, specifically converting the spatial data in the coarse-processed historical dataset and the coarse-processed real-time dataset into a unified coordinate system, and synchronizing the time data with a clock to obtain an aligned historical dataset and an aligned real-time dataset.

[0076] Step S23: Data standardization, used to standardize the aligned data, specifically by using the min-max standardization method to process the aligned historical dataset and the aligned real-time dataset to obtain the standardized historical dataset and the standardized real-time dataset;

[0077] Step S24: Dataset splitting, specifically splitting the standardized historical dataset into a prediction training set and a prediction test set.

[0078] Example 4, see Figure 1 , Figure 2 and Figure 4 This embodiment is based on the above embodiment. In step S3, the disinfection data embedding is used to model the physical interaction between disinfection parameters and indoor building materials, specifically including the following steps;

[0079] Step S31: Time dimension decomposition, used to separate the multi-scale features of the time series, specifically decomposes the time dimension features into trend, periodic, and transient components, using the following formula:

[0080] ;

[0081] In the formula, Representing time-dimensional features, Indicates the trend term. Represents a periodic term. Represents the transient term. This represents the average pooling function. This indicates that the first k dominant frequency components are extracted using Fast Fourier Transform. Represents multi-scale temporal features;

[0082] Step S32: Spatial hierarchical encoding, used to construct a multi-granularity spatial representation, specifically, spatial hierarchical encoding is performed at the grid level, region level, and global level, using the following formula:

[0083] ;

[0084] In the formula, Represents the spatial characteristics of position i. Represents grid-level spatial features. Indicates regional spatial characteristics. Represents global spatial features. This represents the grid-level linear transformation weight matrix. Global represents the set of locations in region j, and Global represents the global set of locations.

[0085] Step S33: Parameter cross-embedding, used for interactive disinfection parameters and indoor building material physical parameters, the formula used is as follows:

[0086] ;

[0087] In the formula, Indicates the interactive characteristics of disinfection materials. This represents the softmax function. Indicates the characteristics of disinfection parameters. This indicates the property characteristics of indoor building materials, and their relationship with disinfection parameters. Same dimensions This indicates the feature dimensions of the disinfection parameters;

[0088] Step S34: Construct a feature set, specifically by processing the standardized real-time dataset, the prediction training set, and the prediction test set through the time dimension decomposition, the spatial hierarchical encoding, and the parameter cross-embedding to obtain a real-time feature set, a training feature set, and a test feature set.

[0089] By performing the above operations, this solution addresses the technical problems of traditional indoor disinfection effect prediction methods, which rely on a single temporal convolution kernel or fixed period assumptions, making it impossible to distinguish between the long-term trend and short-term fluctuations of disinfectant diffusion, and whose spatial modeling depends on uniform grid partitioning, ignoring the impact of the non-uniform distribution of material properties within the region on the disinfection effect. This solution creatively adopts a method of embedding material property features. Through multi-dimensional temporal decomposition, multi-granular spatial representation, and embedding material properties, the influence of local indoor material differences is reduced, and the predictive performance of the model is improved through the interaction between disinfection parameters and material properties.

[0090] Example 5, see Figure 1 , Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S4, the effect prediction model is constructed to build a model required to predict the indoor disinfection effect. Specifically, a physical augmentation graph convolutional network model is constructed as the effect prediction model.

[0091] The construction of the effect prediction model specifically includes the following steps:

[0092] Step S41: Design an activation function to introduce physical constraints, specifically by designing a physical enhancement activation function, the specific formula of which is as follows:

[0093] ;

[0094] In the formula, Let x represent the physics-enhanced activation function, and let x represent the independent variable of the physics-enhanced activation function. This represents the ventilation rate data for region j. This represents the absorption coefficient of the material in region j. This represents the sigmoid function. This represents the Euclidean distance from region m to region j. This represents the ambient humidity data for region m. This represents the dynamic sharpness adjustment coefficient for region j. Represents the ReLU function. Indicates the concentration of disinfectant in region j. Indicates the saturation concentration of the disinfectant;

[0095] Step S42: Design a forward diffusion process to simulate the active propagation of disinfectant under ventilation and material constraints. The steps include:

[0096] Step S421: Calculate the forward dynamic adjacency matrix using the following formula:

[0097] ;

[0098] In the formula, This represents a forward dynamic adjacency matrix. This indicates that the weight matrix is ​​calculated using the forward dynamic adjacency matrix. This represents the connectivity characteristics between region m and region j. This represents the airflow velocity data from region m to region j. Indicates the distance decay index. This represents the Euclidean distance from region m to region j. of Power of 1 This represents the connectivity characteristic from region m to region j. This represents an indicator function whose value is 1 when there is a connection between region m and region j, and whose value is 0 when there is no connection between region m and region j.

[0099] Step S422: Implement physical constraints, using the following formula:

[0100] ;

[0101] In the formula, This indicates the amount of disinfectant released in region m. This represents the volume of region m. Indicates the density of the disinfectant. This indicates a small value used to prevent division by zero. This represents the positive dynamic adjacency matrix under constraints, where j represents the index of the region. This represents the total number of regions that are connected to region m.

[0102] Step S423: Forward graph convolution update, using the following formula:

[0103] ;

[0104] In the formula, This represents the forward graph convolutional features of the (l+1)th layer. This represents the graph convolution operation function. This represents the forward convolutional features of the l-th layer;

[0105] Step S43: Design reverse residual transmission to simulate the passive diffusion of disinfectant under the influence of concentration gradient and personnel contact. The steps include:

[0106] Step S431: Calculate the half-life of the disinfectant component using the following formula:

[0107] ;

[0108] In the formula, This indicates the half-life of the disinfectant component in region j. Indicates the standard half-life of disinfectant components. Indicates the influence coefficient of ultraviolet radiation intensity. This represents the ultraviolet intensity data for region j. Indicates the temperature of region j;

[0109] Step S432: Calculate the reverse dynamic adjacency matrix using the following formula:

[0110] ;

[0111] In the formula, Represents the reverse dynamic adjacency matrix. This indicates the concentration of disinfectant in area m. This indicates the half-life of the disinfectant component in region n. This represents the frequency of personnel contact between region m and region j. Let represent the surface area of ​​region j, and n represent the index of the region that is distinct from j.

[0112] Step S433: Inverse graph convolution update, using the following formula:

[0113] ;

[0114] In the formula, This represents the inverse graph convolutional features of the (l+1)th layer. This represents the inverse graph convolutional features of the l-th layer;

[0115] Step S44: Bidirectional gating fusion, used to balance forward propagation and backward residual features. Specifically, a gating mechanism is designed to fuse forward propagation features and backward residual features. The formula used is as follows:

[0116] ;

[0117] In the formula, Gw represents the gating weight. This indicates the weights used in calculating the gating weights. Indicates positive propagation characteristics. Indicates reverse residual characteristics, Indicates fusion characteristics;

[0118] Step S45: Obtain the model output using the following formula:

[0119] ;

[0120] In the formula, This indicates the model's initial prediction output. Indicates the model output weights. express Time delay factor, This indicates that the delay coefficient is used to calculate the weight. express UV intensity at any given time This represents the model's predicted output at time t, and Tm represents the total number of time points. express Preliminary prediction output from the time-matter model;

[0121] Step S46: Model construction and training, specifically, constructing a physical augmented graph convolutional network model through the design activation function, the design forward diffusion process, the design backward residual propagation, the bidirectional gating fusion, and the acquisition of model output; training the model based on the training feature set; verifying the model performance based on the test feature set; obtaining the physical augmented graph convolutional network model, which serves as the performance prediction model.

[0122] By performing the above operations, this solution addresses the technical problems of traditional indoor disinfection effect prediction methods, which cannot model the reverse process of disinfectant degradation, cannot capture the nonlinear effects of environmental parameters such as ultraviolet intensity, temperature, and humidity on the lag in disinfection effect, and have errors in time delay compensation. This solution creatively adopts a physically enhanced graph convolutional network model as the effect prediction model. Under the condition of conforming to physical laws, it can fully combine the characteristics of the forward diffusion process of disinfectant and the characteristics of the reverse residue process of disinfectant, and adjust the historical cumulative effect through a time compensation mechanism, thus solving the deviation caused by the lag in disinfection effect.

[0123] Example 6, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S5, the disinfection effect prediction specifically involves using the effect prediction model to predict the indoor disinfection effect based on the real-time feature set, obtaining disinfection effect prediction reference data, and comprehensively evaluating the indoor disinfection effect based on the disinfection effect prediction reference data.

[0124] Example 7, see Figure 1 and Figure 2 Based on the above embodiments, this embodiment provides an intelligent indoor disinfection effect prediction system, including a disinfection data acquisition module, a preliminary data optimization module, a disinfection data embedding module, an effect prediction model construction module, and a disinfection effect prediction module.

[0125] The disinfection data acquisition module is used for disinfection data acquisition. Through disinfection data acquisition, the original disinfection effect dataset is obtained, and the original disinfection effect dataset is sent to the data preliminary optimization module.

[0126] The data preliminary optimization module is used for preliminary data optimization. Through preliminary data optimization, a standardized real-time dataset, a prediction training set, and a prediction test set are obtained, and the standardized real-time dataset, the prediction training set, and the prediction test set are sent to the disinfection data embedding module.

[0127] The disinfection data embedding module is used for disinfection data embedding. Through disinfection data embedding, the physical interaction between disinfection parameters and indoor building materials is modeled to obtain a real-time feature set, a training feature set, and a test feature set. The real-time feature set is sent to the disinfection effect prediction module, and the training feature set and the test feature set are sent to the effect prediction model construction module.

[0128] The effect prediction model construction module is used to construct the effect prediction model. It obtains the effect prediction model by constructing a physical augmentation graph convolutional network model and sends the effect prediction model to the disinfection effect prediction module.

[0129] The disinfection effect prediction module is used to predict the disinfection effect. By using the effect prediction model to predict the indoor disinfection effect, reference data for the disinfection effect prediction is obtained.

[0130] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0131] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

[0132] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An intelligent indoor disinfection effect prediction method, characterized in that: The method comprises the following steps: S1: disinfection data collection, through disinfection data collection, an original disinfection effect data set is obtained, the original disinfection effect data set specifically comprises a historical disinfection effect original set and a real-time disinfection effect original set; S2: data preliminary optimization, the collected original data is preprocessed to optimize the data, to obtain a standardized real-time data set, a prediction training set and a prediction test set; S3: disinfection data embedding, used for modeling the physical interaction of disinfection parameters and indoor building materials, through time dimension decomposition, space level coding and parameter cross embedding, a real-time feature set, a training feature set and a test feature set are obtained; S4: effect prediction model construction, used for constructing a model required for predicting indoor disinfection effect, specifically, a physical enhanced graph convolutional network model is constructed as the effect prediction model; In step S4, step S41: design an activation function, used for introducing physical constraints, specifically, a physical enhanced activation function is designed, and the specific formula is as follows: ； wherein, represents a physical enhancement activation function, x represents a physical enhancement activation function argument, represents ventilation rate data for zone j, represents an absorption coefficient for the material of zone j, represents a sigmoid function, represents the Euclidean distance from zone m to zone j, represents ambient humidity data for zone m, represents a dynamic sharpness adjustment coefficient for zone j, represents a ReLU function, represents the concentration of disinfectant in zone j, represents the saturation concentration of disinfectant; S5: disinfection effect prediction, specifically, indoor disinfection effect prediction is performed through the effect prediction model, to obtain disinfection effect prediction reference data.

2. The intelligent indoor disinfection effect prediction method according to claim 1, characterized in that: The effect prediction model construction specifically comprises the following steps: Step S41: design an activation function; Step S42: design a forward diffusion process, used for simulating the active propagation of disinfectants under the constraints of ventilation and material, and the steps comprise: Step S421: calculate a forward dynamic adjacency matrix, and the formula used is as follows: ； wherein, represents a forward dynamic adjacency matrix, represents a forward dynamic adjacency matrix to compute a weight matrix, represents a connectedness feature between region m to region j, represents a flow velocity data from region m to region j, represents a distance decay exponent, represents a Euclidean distance from region m to region j to the power of n, represents a connectedness existence feature from region m to region j, represents an indicator function that takes a value of 1 when there is a connectedness between region m to region j and a value of 0 when there is no connectedness between region m to region j.​ Step S422: implement physical constraints, and the formula used is as follows: ； wherein represents the amount of disinfectant released by region m, represents the volume of region m, represents the density of disinfectant, represents a small value used to prevent division by zero, represents the forward dynamic adjacency matrix under constraints, j represents the index of the region, represents the total number of regions that have a communicative relationship with region m; Step S423: forward graph convolution update; Step S43: design a reverse residual propagation, used for simulating the passive diffusion of disinfectants under the influence of concentration gradient and personnel contact, and the steps comprise: Step S431: calculate the half-life of disinfectant components; Step S432: calculate a reverse dynamic adjacency matrix, and the formula used is as follows: ； wherein represents the reverse dynamic adjacency matrix, represents the zone m disinfectant concentration, represents the zone j disinfectant component half-life, represents the zone n disinfectant component half-life, represents the zone m to zone j personnel contact frequency, represents the zone j surface area, n represents an index of a zone different from j; Step S433: reverse graph convolution update; Step S44: bidirectional gate fusion, used for balancing forward propagation and reverse residual, specifically, a gate mechanism is designed to fuse forward propagation features and reverse residual features; Step S45: obtain model output, and the formula used is as follows: ； In the formula, denotes the model preliminary prediction output, denotes the model output weight, denotes the fusion feature, denotes the time delay coefficient, denotes the delay coefficient calculation weight, denotes the multi-scale time feature, denotes the time ultraviolet intensity, denotes the model prediction output at t, and Tm denotes the total number of times, denotes the model preliminary prediction output at t. Step S46: model construction and training, specifically, through the design activation function, the design forward diffusion process, the design reverse residual propagation, the bidirectional gate fusion and the model output, the physical enhanced graph convolutional network model is constructed, the model is trained based on the training feature set, the model performance is verified based on the test feature set, to obtain the physical enhanced graph convolutional network model as the effect prediction model. 3.The intelligent indoor disinfection effect prediction method according to claim 1, characterized in that: The disinfection data embedding specifically comprises the following steps: Step S31: time dimension decomposition, used for separating the multi-scale features of time series, specifically, the time dimension features are decomposed into trend items, periodic items and transient items; Step S32: space level coding, used for constructing multi-granularity space representation, specifically, space level coding is performed according to grid level, region level and global level; Step S33: parameter cross embedding, used for interacting disinfection parameters and indoor building material physical parameters, and the formula used is as follows: ； wherein, represents disinfection material interaction features, represents a softmax function, represents disinfection parameter features, represents indoor building material attribute features, which are combined with disinfection parameter features are of the same dimension, represents disinfection parameter feature dimension; Step S34: constructing a feature set, specifically, processing the standardized real-time data set, the prediction training set and the prediction test set through the time dimension decomposition, the spatial hierarchy coding and the parameter cross-embedding to obtain a real-time feature set, a training feature set and a test feature set.

4. The intelligent indoor disinfection effect prediction method according to claim 1, characterized in that: The disinfection data collection is used to collect original data required for predicting indoor disinfection effect, specifically, by collecting disinfection data from a disinfection log system, a building information system and an environmental sensor system to obtain an original disinfection effect data set, the original disinfection effect data set specifically includes a historical disinfection effect original set and a real-time disinfection effect original set, both the historical disinfection effect original set and the real-time disinfection effect original set include space-time reference data, disinfection related data and environmental data, and the historical disinfection effect original set further includes historical disinfection effect label data.

5. The intelligent indoor disinfection effect prediction method according to claim 1, characterized in that: The data preliminary optimization specifically includes the following steps: Step S21: missing value processing, used to remove missing values, specifically, removing missing values in the historical disinfection effect original set and the real-time disinfection effect original set to obtain a rough processed historical data set and a rough processed real-time data set; Step S22: space-time data alignment, used to align space-time data, specifically, converting spatial data in the rough processed historical data set and the rough processed real-time data set into a unified coordinate system, and clock synchronizing time data to obtain an aligned historical data set and an aligned real-time data set; Step S23: data standardization, used to standardize the aligned data, specifically, using the minimum-maximum standardization method to process the aligned historical data set and the aligned real-time data set to obtain a standardized historical data set and a standardized real-time data set; Step S24: data set segmentation, used to segment data sets, specifically, segmenting the standardized historical data set into a prediction training set and a prediction test set.

6. The intelligent indoor disinfection effect prediction method according to claim 1, characterized in that: The disinfection effect prediction specifically includes predicting indoor disinfection effect based on the real-time feature set using the effect prediction model to obtain disinfection effect prediction reference data, and comprehensively evaluating indoor disinfection effect based on the disinfection effect prediction reference data.

7. An intelligent indoor disinfection effect prediction system for implementing the intelligent indoor disinfection effect prediction method according to any one of claims 1-6, characterized in that: The disinfection data collection module, the data preliminary optimization module, the disinfection data embedding module, the effect prediction model construction module and the disinfection effect prediction module. 8.The intelligent indoor disinfection effect prediction system according to claim 7, characterized in that: The disinfection data collection module is used for disinfection data collection, and the original disinfection effect data set is obtained through disinfection data collection, and the original disinfection effect data set is sent to the data preliminary optimization module; The data preliminary optimization module is used for data preliminary optimization, and the standardized real-time data set, the prediction training set and the prediction test set are obtained through data preliminary optimization, and the standardized real-time data set, the prediction training set and the prediction test set are sent to the disinfection data embedding module; The disinfection data embedding module is configured to perform disinfection data embedding, model physical interaction between disinfection parameters and indoor building materials, obtain a real-time feature set, a training feature set, and a test feature set, and send the real-time feature set to the disinfection effect prediction module and send the training feature set and the test feature set to the effect prediction model construction module. The effect prediction model construction module is configured to perform effect prediction model construction, construct a physical enhanced graph convolutional network model, obtain an effect prediction model, and send the effect prediction model to the disinfection effect prediction module. The disinfection effect prediction module is configured to perform disinfection effect prediction, perform indoor disinfection effect prediction by using the effect prediction model, and obtain disinfection effect prediction reference data.

Images