Environment smell analysis method based on intelligent terminal and related equipment

Abstract

The invention relates to an environment smell analysis method based on an intelligent terminal and related equipment, and the method comprises the following steps: carrying out the real-time collection of environment smell through a sensor network, obtaining an environment smell data flow, carrying out the preprocessing of the environment smell data flow, and obtaining the optimized environment smell data; performing feature extraction and decomposition on the optimized environmental odor data to obtain an environmental odor multi-dimensional vector; performing classification prediction on the environmental odor multi-dimensional vector to obtain an environmental odor prediction type; performing correlation analysis of environmental factors on the environmental odor prediction type to obtain an environmental odor-environmental factor thermodynamic distribution diagram; transmitting and analyzing the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram through an intelligent terminal to obtain an environmental odor diffusion path and an environmental odor diffusion range, and solving the problem that a traditional method is lack of dynamic prediction capability on the odor diffusion path and range; the actual condition of odor pollution cannot be comprehensively reflected.

Description

Technical Field

[0001] The present invention relates to the field of odor technology, and in particular to an environmental odor analysis method based on a smart terminal and related equipment. Background Art

[0002] In the process of modern urbanization, the monitoring and analysis of environmental odors are becoming increasingly important. Traditional odor monitoring methods mainly rely on artificial olfaction and laboratory analysis, which are not only time-consuming and costly, but also difficult to ensure the real-time and accuracy of data due to human factors and long sample collection cycles. In addition, with the advancement of industrialization and the increase in urban population, environmental pollution problems are becoming increasingly serious, and odor pollution has become an important factor affecting the quality of life of citizens. Therefore, it is particularly urgent to develop an efficient, accurate and real-time environmental odor analysis method.

[0003] Although existing odor analysis technologies can solve the problem of odor monitoring to a certain extent, they still have many limitations. For example, most methods rely on complex hardware equipment, which is cumbersome to operate and difficult to popularize. In addition, these methods usually lack the ability to dynamically predict the path and range of odor diffusion and cannot fully reflect the actual situation of odor pollution. The limitations of traditional methods make it difficult to achieve scientific and systematic management of odor pollution control and prevention.

[0004] The environmental odor analysis method based on smart terminals can effectively integrate sensor networks and data analysis technologies, and provide more accurate and timely odor monitoring information by collecting and processing odor data in real time. However, the current technology still has problems such as low data processing efficiency, insufficient classification prediction accuracy, and insufficient in-depth analysis of environmental factor associations. Therefore, it is necessary to further optimize the data collection and processing algorithms of sensor networks, improve the accuracy of feature extraction and classification prediction, and enhance the ability to analyze the association between environmental factors and odor data, so as to achieve a more comprehensive prediction of odor diffusion paths and ranges. Summary of the invention

[0005] The main purpose of the present invention is to provide an environmental odor analysis method and related equipment based on a smart terminal, which solves the technical problem that traditional methods lack the ability to dynamically predict the odor diffusion path and range, resulting in the inability to fully reflect the actual situation of odor pollution.

[0006] To achieve the above object, the present invention provides an environmental odor analysis method based on a smart terminal, comprising the following steps: Collecting environmental odors in real time through a sensor network to obtain an environmental odor data stream, and preprocessing the environmental odor data stream to obtain optimized environmental odor data; Performing feature extraction and decomposition on the optimized environmental odor data to obtain a multi-dimensional vector of the environmental odor; Classify and predict the multi-dimensional vector of environmental odor to obtain a predicted type of environmental odor; Performing correlation analysis of environmental factors on the predicted environmental odor type to obtain an environmental odor-environmental factor thermodynamic distribution diagram; Through the intelligent terminal, the environmental odor is transmitted and analyzed based on the environmental odor-environmental factor thermodynamic distribution diagram to obtain the environmental odor diffusion path and environmental odor diffusion range.

[0007] Furthermore, the environmental odor is collected in real time through a sensor network to obtain an environmental odor data stream, and the environmental odor data stream is preprocessed to obtain optimized environmental odor data, including: The odor in the environment is collected in real time through the sensor network to obtain the environmental odor data stream; Performing denoising processing on the environmental odor data stream to obtain preliminary purified data; Performing augmentation processing on the preliminary cleansed data to obtain enhanced data; Performing feature selection on the enhanced data to obtain selected feature data; Performing principal component analysis on the selected feature data to obtain key feature data; The key feature data is analyzed in the time domain to obtain time-frequency characteristic information, and the time-frequency characteristic information is used as the optimized environmental odor data.

[0008] Furthermore, feature extraction and decomposition are performed on the optimized environmental odor data to obtain a multi-dimensional vector of environmental odor, including: Normalizing the optimized environmental odor data to obtain normalized environmental odor data; Performing multi-dimensional feature extraction on the normalized environmental odor data to obtain multi-dimensional extracted features; wherein the multi-dimensional extracted features include spectrum energy features and spectrum entropy features; Decomposing the multi-dimensional extracted features in multiple scales and directions by using a preset curvelet transform technique to obtain a curve feature representation; Performing dimensionality reduction processing on the curve feature representation to obtain a compressed feature vector; The compressed feature vector is linearly transformed to obtain a multi-dimensional vector of environmental odor.

[0009] Furthermore, the multi-dimensional vector of the environmental odor is classified and predicted to obtain the environmental odor prediction type, including: Obtain a multi-dimensional vector of historical environmental odors, and use the multi-dimensional vector of historical environmental odors to embed in a high-dimensional space to obtain a feature space matrix; Perform hyperplane partitioning on the feature space matrix to obtain the environmental odor category boundary; Classifying the multi-dimensional vector of the environmental odor based on the environmental odor category boundary to obtain an environmental odor category label; The environmental odor category labels are fused by multimodal fusion technology to obtain environmental odor category confidence; Performing weighted scoring calculation on the environmental odor category labels based on the environmental odor category confidence to obtain a weighted ranking of environmental odor prediction types; The environmental odor category label with the highest weighted ranking among the environmental odor prediction types is used as the environmental odor prediction type.

[0010] Furthermore, the environmental odor prediction type is subjected to correlation analysis of environmental factors to obtain an environmental odor-environmental factor thermodynamic distribution diagram, including: Simulating the molecular structure corresponding to the environmental odor prediction type to obtain the molecular chemical characteristics of the environmental odor; wherein the molecular chemical characteristics of the environmental odor include the electron density distribution of the environmental odor molecules and the vibration frequency spectrum of the environmental odor molecules, and the molecular chemical characteristics of the environmental odor are used to characterize the interaction potential between the environmental odor molecules and environmental factors; Performing topological data analysis on the molecular chemical characteristics to obtain a molecular topological characteristic map; wherein the molecular topological characteristic map includes a manifold structure of a molecular configuration space; The molecular topological characteristic map and the environmental factors are subjected to dynamic evolution analysis to obtain a density matrix of the environmental odor transmission channel; wherein the density matrix of the environmental odor transmission channel is used to reflect the coherence and decoherence process of the molecules of the environmental odor under different environmental conditions; The diffusion behavior of the environmental odor under different environmental factors is calculated based on the density matrix of the environmental odor transmission channel through the vortex dynamics model; Performing Helmholtz-Hodge decomposition on the diffusion behavior to obtain the vortex field distribution of environmental odor diffusion; wherein the vortex field distribution of environmental odor diffusion includes vortex core position, vortex intensity and vortex interaction energy; Performing fractal dimension analysis on the vortex field distribution to obtain multi-scale indicators of environmental odor-environmental factors; Through the theory of non-extensive statistical thermodynamics, an environmental odor-environmental factor thermodynamic distribution map is constructed based on the multi-scale indicators; wherein the environmental odor-environmental factor thermodynamic distribution map includes a high concentration area, a transition area and a low concentration area.

[0011] Furthermore, the intelligent terminal performs transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram to obtain the environmental odor diffusion path and environmental odor diffusion range, including: By using the odor analysis algorithm preset in the smart terminal, manifold learning is performed on the environmental odor-environmental factor thermodynamic distribution map to obtain an environmental odor transmission topology structure, wherein the environmental odor transmission topology structure includes an odor source point, a transmission node, and a convergence point; Based on the environmental odor transmission topological structure, a non-equilibrium thermodynamic model is performed on the transmission process of the environmental odor to obtain a transmission kinetic equation group, wherein the transmission kinetic equation group includes a mass conservation equation, a momentum conservation equation, and an energy conservation equation; Numerically solving the transmission dynamics equations to obtain the spatiotemporal evolution data of the ambient odor concentration field, wherein the spatiotemporal evolution data of the ambient odor concentration field includes a concentration gradient, a flow velocity field, and a temperature field; Based on the spatiotemporal evolution data of the ambient odor concentration field, a particle Brownian motion simulation of the ambient odor is performed to obtain a set of ambient odor particle trajectories, wherein the set of ambient odor particle trajectories includes a main path, a secondary path, and an abnormal path; Performing path integral analysis and optimal transmission theory calculation on the set of environmental odor particle trajectories to obtain an optimal diffusion network path for the environmental odor, wherein the optimal diffusion network path for the environmental odor includes a trunk path, a branch path, and a critical path; Based on the optimal environmental odor diffusion network path, the environmental odor diffusion range boundary is identified and predicted to obtain the environmental odor diffusion path and the environmental odor diffusion range.

[0012] Furthermore, the transmission dynamics equations are numerically solved to obtain the spatiotemporal evolution data of the ambient odor concentration field, including: Discretizing the transmission dynamics equations to obtain a discretized equation group; Iteratively solving the discretized equation group to obtain an initial concentration field; Advancing the time step of the initial concentration field to obtain a time-evolved concentration field; Performing spatial grid refinement on the time-evolved concentration field to obtain a grid-refined concentration field; Performing stability check on the refined concentration field to obtain a stable concentration field; The stable concentration field is subjected to a spatiotemporal evolution analysis to obtain the spatiotemporal evolution data of the ambient odor concentration field.

[0013] The present invention also provides an environment odor analysis device based on a smart terminal, comprising: A collection module, used to collect environmental odors in real time through a sensor network to obtain an environmental odor data stream, and pre-process the environmental odor data stream to obtain optimized environmental odor data; An extraction module, used for performing feature extraction and decomposition on the optimized environmental odor data to obtain a multi-dimensional vector of the environmental odor; A prediction module, used to classify and predict the multi-dimensional vector of environmental odor to obtain a predicted type of environmental odor; A first analysis module is used to perform correlation analysis of environmental factors on the environmental odor prediction type to obtain an environmental odor-environmental factor thermodynamic distribution diagram; The second analysis module is used to perform transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram through the smart terminal to obtain the environmental odor diffusion path and environmental odor diffusion range.

[0014] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0015] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0016] The environmental odor analysis method based on the intelligent terminal provided by the present invention comprises the following steps: real-time collection of environmental odors through a sensor network to obtain an environmental odor data stream, and preprocessing the environmental odor data stream to obtain optimized environmental odor data; feature extraction and decomposition of the optimized environmental odor data to obtain an environmental odor multi-dimensional vector; classification prediction of the environmental odor multi-dimensional vector to obtain an environmental odor prediction type; correlation analysis of environmental factors on the environmental odor prediction type to obtain an environmental odor-environmental factor thermodynamic distribution map; through the intelligent terminal, transmission analysis of the environmental odor based on the environmental odor-environmental factor thermodynamic distribution map to obtain an environmental odor diffusion path and an environmental odor diffusion range; through the above technical solution, the technical problem that the traditional method lacks the ability to dynamically predict the odor diffusion path and range, resulting in the inability to fully reflect the actual situation of odor pollution is solved, and the beneficial effect of timely discovering and responding to environmental odor pollution problems and taking effective measures to improve environmental quality is achieved through accurate odor monitoring and analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of the steps of an environmental odor analysis method based on a smart terminal in one embodiment of the present invention; Figure 2 is a structural block diagram of an environmental odor analysis device based on a smart terminal in one embodiment of the present invention; Figure 3 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0018] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] like Figure 1 As shown, Figure 1 This is a schematic diagram of the steps of an environmental odor analysis method based on a smart terminal in one embodiment of the present invention; In one embodiment of the present invention, a method for analyzing environmental odor based on a smart terminal is provided, comprising the following steps: Step S1, collecting environmental odor in real time through a sensor network to obtain an environmental odor data stream, and preprocessing the environmental odor data stream to obtain optimized environmental odor data.

[0021] Specifically, to collect environmental odors in real time through sensor networks, it is first necessary to deploy multiple highly sensitive sensors in the target environment. These sensors can capture different types of odor molecules and convert them into analyzable data signals. These sensors may include gas sensors, chemical sensors, etc., which can detect the concentration and composition of specific gases. For example, in an industrial area, sensors can be installed at different locations to capture various gases emitted from factories, including volatile organic compounds (VOCs), hydrogen sulfide, ammonia, etc. When the sensors capture the odor information in the environment, this information will be transmitted through the sensor network to form an environmental odor data stream. This data stream is a continuous and dynamically updated data set that reflects the real-time odor situation. Then, preprocessing the environmental odor data stream is a very critical step. Preprocessing includes steps such as data cleaning, denoising, and normalization to ensure the quality and consistency of the data. Data cleaning can remove outliers and missing values, denoising can reduce the impact of sensor errors and environmental interference, and normalization can convert data of different dimensions to the same scale for subsequent analysis. For example, in the application scenario of the above-mentioned industrial area, the odor data captured by the sensor may contain noise, such as electromagnetic interference generated by the operation of the machine or fluctuations caused by weather changes. Through the preprocessing step, these noises can be filtered out to make the data purer and more reliable. Optimizing environmental odor data is the ultimate goal of preprocessing, which means that the processed data is more suitable for feature extraction and analysis. Optimizing data not only improves the accuracy of the data, but also improves the efficiency and effectiveness of subsequent analysis. Therefore, the environmental odor is collected in real time through the sensor network, and the environmental odor data stream is preprocessed to finally obtain optimized environmental odor data. This step plays a fundamental and key role in the entire environmental odor analysis method. The optimized data will provide a solid data foundation for subsequent feature extraction, classification prediction and environmental factor association analysis, ensuring the reliability and accuracy of the entire system.

[0022] Step S2, extracting and decomposing features of the optimized environmental odor data to obtain a multi-dimensional vector of the environmental odor.

[0023] Specifically, feature extraction and decomposition of the optimized environmental odor data is to further analyze the preprocessed optimized environmental odor data in order to extract a multi-dimensional vector that can reflect the odor characteristics. In this process, it is first necessary to select and design appropriate feature extraction methods, which may include principal component analysis (PCA), independent component analysis (ICA), linear discriminant analysis (LDA), etc., which can extract the most representative feature information from the optimized environmental odor data. In practical applications, such as in the odor monitoring scene of industrial areas, the optimized environmental odor data may contain concentration values ​​and time series of multiple gases. Through feature extraction, these complex raw data can be converted into a few key parameters that can describe the odor characteristics. For example, through principal component analysis (PCA), the high-dimensional data of multiple gas concentration values ​​can be reduced to a lower-dimensional feature space to obtain the main components of the environmental odor. These main components constitute the multi-dimensional vector of the environmental odor. The data after feature extraction is often a high-dimensional vector, and these vectors can describe the characteristics of the environmental odor from multiple angles. In order to better understand and utilize these high-dimensional vectors, we can further decompose them. For example, a high-dimensional vector can be decomposed into several sub-vectors using methods such as singular value decomposition (SVD), each of which represents a different dimension of odor characteristics. These sub-vectors can be used for subsequent classification prediction and correlation analysis of environmental factors. In the example of an industrial area, through feature extraction and decomposition, we can obtain a multi-dimensional vector of environmental odor, in which each dimension may correspond to a specific gas component, such as VOC concentration, hydrogen sulfide concentration, etc. In addition, some time characteristics, such as the rate and trend of change of odor, can also be included. Through these multi-dimensional vectors, the characteristics of environmental odor can be described more comprehensively and accurately, thereby providing a solid data foundation for subsequent odor classification and prediction. In short, feature extraction and decomposition of the optimized environmental odor data is a process of converting complex raw data into multi-dimensional vectors that can describe odor characteristics. This process not only improves the interpretability of the data, but also provides high-quality data input for subsequent analysis. Through reasonable feature extraction and decomposition methods, important information in environmental odor data can be better captured and utilized to ensure the accuracy and effectiveness of the entire environmental odor analysis system.

[0024] Step S3, classify and predict the multi-dimensional vector of environmental odor to obtain a predicted type of environmental odor.

[0025] Specifically, the classification prediction of the multi-dimensional vector of environmental odor is to further determine the specific type of environmental odor by analyzing the features in these multi-dimensional vectors. This step can be achieved through a variety of machine learning algorithms, such as support vector machine (SVM), random forest, neural network, etc. The selection of a suitable classification algorithm can be determined according to the characteristics of the data and the specific application scenario. In the odor monitoring scenario of an industrial area, the multi-dimensional vector of environmental odor may include the concentration values ​​and time characteristics of multiple gas components. In order to perform classification prediction, these multi-dimensional vectors need to be trained first, that is, a classification model is established through a machine learning algorithm using a training data set with known labels. Each multi-dimensional vector of environmental odor in the training data set will have a corresponding odor type label, such as "industrial waste gas", "agricultural odor", "natural odor", etc. When a new multi-dimensional vector of environmental odor is predicted by a trained classification model, the model will output an environmental odor prediction type based on the learned features and rules. For example, in the scenario of an industrial area, if a new odor data is input into the model, the model may judge that the odor belongs to the "industrial waste gas" type based on its characteristics. This process can not only quickly identify the type of odor in the environment, but also assist managers in environmental monitoring and decision-making. In order to ensure the accuracy of classification predictions, the model training process requires a large amount of data and repeated verification. Common methods include cross-validation and grid search to adjust model parameters and evaluate their performance. In practical applications, new odor data can be continuously collected and the model can be updated to improve the accuracy and robustness of classification predictions. For example, in the odor monitoring project in an industrial area, it is assumed that a large amount of environmental odor data has been collected and the corresponding multi-dimensional vectors have been extracted. These vectors are used to train a support vector machine (SVM) model that learns the characteristics of different types of odors. When new odor data is collected by the sensor and feature extraction and decomposition are performed, the resulting multi-dimensional vector of environmental odor will be input into the trained SVM model. The model will predict the type of odor, such as "industrial waste gas", based on the features learned during the training process. In this way, abnormal odors in the environment can be monitored and identified in real time, which helps to take timely countermeasures to ensure environmental quality and safety. In short, the classification and prediction of the multi-dimensional vector of environmental odor is a process of analyzing and judging the features in the multi-dimensional vector using a machine learning algorithm to obtain the predicted type of environmental odor. Through reasonable model training and continuous data updating, accurate classification and prediction of environmental odor can be achieved, providing important technical support for environmental monitoring and management.

[0026] Step S4, performing correlation analysis of environmental factors on the predicted environmental odor type to obtain an environmental odor-environmental factor thermodynamic distribution diagram.

[0027] Specifically, the association analysis of environmental factors for the environmental odor prediction type is to further explore the relationship between environmental odor and various environmental factors on the basis of the environmental odor prediction type, so as to obtain an environmental odor-environmental factor thermodynamic distribution map. In this process, it is first necessary to clarify the environmental factors that need to be analyzed, including temperature, humidity, wind speed, wind direction, atmospheric pressure, etc. In practical applications, for example, in the odor monitoring scene of an industrial area, the environmental odor prediction type includes different odor categories, such as rancid odor, aromatic odor, chemical odor, etc. In order to perform the association analysis of environmental factors, it is first necessary to collect and organize long-term monitoring data, which include environmental odor prediction types and environmental factor data at the corresponding time. Next, statistical analysis methods, such as correlation analysis, regression analysis, etc., can be used to calculate the association between environmental odor prediction types and various environmental factors. For example, the linear relationship between a certain odor type and temperature and humidity can be quantified by the Pearson correlation coefficient. Once the association between each environmental factor and the environmental odor prediction type is obtained, the environmental odor-environmental factor thermodynamic distribution map can be further constructed. This distribution map can be presented in the form of a heat map, where the horizontal and vertical axes represent different environmental factors and odor types, respectively, and the depth of the color indicates the strength of the association. For example, in the environmental odor-environmental factor thermodynamic distribution map, if a certain odor type has a strong positive correlation with the environmental factors of high temperature and high humidity, the corresponding heat map area will be displayed in dark color. Such visualization can help quickly identify which environmental factors have the greatest impact on a specific odor type. In the example of an industrial area, through this thermodynamic distribution map, the distribution of certain odor types under specific environmental conditions can be intuitively seen. For example, a rancid odor may occur more frequently under conditions of high humidity and low wind speed, while a chemical odor may be more significant under conditions of high temperature and high wind speed. Through such association analysis, it can not only help to better understand the source and diffusion law of environmental odor, but also provide a scientific basis for environmental management and pollution control. In short, the association analysis of environmental factors for the environmental odor prediction type is to reveal the relationship between environmental odor and environmental factors through statistical and visualization methods. The environmental odor-environmental factor thermodynamic distribution map finally obtained in this process not only improves the interpretability and visualization of the data, but also provides important decision support for environmental monitoring and management. Through reasonable correlation analysis methods, environmental odor data can be better understood and utilized to ensure the scientificity and effectiveness of the entire environmental odor analysis system.

[0028] Step S5, using the smart terminal, performing transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram to obtain an environmental odor diffusion path and an environmental odor diffusion range.

[0029] Specifically, the intelligent terminal is used to perform transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution map. This process aims to use the intelligent terminal and the environmental odor-environmental factor thermodynamic distribution map to effectively analyze and predict the diffusion path and diffusion range of the odor. First, the intelligent terminal needs to receive and process the data collected by the environmental odor sensor. These data include the concentration, composition and related environmental factors of the odor, such as temperature, humidity, wind speed and wind direction. Then, based on these data, the intelligent terminal can generate an environmental odor-environmental factor thermodynamic distribution map, which shows the distribution and changes of the odor under different environmental conditions.

[0030] In practical applications, such as in the odor monitoring scenario in industrial areas, the environmental odor-environmental factor thermodynamic distribution map can show the concentration changes of certain specific gases in different time periods and different locations. By analyzing the thermodynamic distribution map, the intelligent terminal can use a series of gas diffusion models and thermodynamic equations, combined with current environmental factors, to predict the diffusion path and range of the odor. Commonly used diffusion models include Gaussian diffusion model, Lagrangian particle trajectory model, etc. These models use the data in the thermodynamic distribution map to simulate the diffusion behavior of odor in the air. For example, in the actual scenario of the industrial area, if a factory emits a certain amount of volatile organic compounds (VOCs), the intelligent terminal can predict in which direction these VOCs will diffuse and how large the diffusion range is by analyzing the environmental odor-environmental factor thermodynamic distribution map, combined with the wind speed and wind direction information at the time. This prediction can help relevant departments take corresponding measures in a timely manner, such as adjusting emission strategies or setting warning areas to reduce the impact on the surrounding environment and residents. Odor transmission analysis through the intelligent terminal can not only monitor and predict the diffusion path and range of odors in real time, but also provide a scientific basis for environmental management and decision-making. The core of this process is to transform complex environmental odor and environmental factor data into a visual thermodynamic distribution map, and use the map to perform precise transmission analysis. Ultimately, this method can effectively improve the responsiveness and accuracy of the environmental odor monitoring system, ensuring that potential environmental problems are discovered and dealt with in a timely manner.

[0031] In a specific embodiment, the environmental odor is collected in real time through a sensor network to obtain an environmental odor data stream, and the environmental odor data stream is preprocessed to obtain optimized environmental odor data, including: The odor in the environment is collected in real time through the sensor network to obtain the environmental odor data stream; Performing denoising processing on the environmental odor data stream to obtain preliminary purified data; Performing augmentation processing on the preliminary cleansed data to obtain enhanced data; Performing feature selection on the enhanced data to obtain selected feature data; Performing principal component analysis on the selected feature data to obtain key feature data; The key feature data is analyzed in the time domain to obtain time-frequency characteristic information, and the time-frequency characteristic information is used as the optimized environmental odor data.

[0032] Specifically, the environmental odor is collected in real time through the sensor network to obtain an environmental odor data stream, and the environmental odor data stream is preprocessed to obtain optimized environmental odor data. This process involves multiple steps, including data collection, denoising, data augmentation, feature selection, principal component analysis and time domain analysis. First, the odor in the environment is collected in real time through the sensor network to obtain an environmental odor data stream. This step is the starting point of the entire process. The sensor network can collect various odor components in the air in real time at different locations and times. For example, in the monitoring scene of the industrial area, the sensor can detect the concentration of volatile organic compounds (VOC), hydrogen sulfide, ammonia and other gases. Next, the environmental odor data stream is denoised to obtain preliminary purified data. Since the collected environmental odor data stream may contain various noises and interferences, such as sensor errors, changes in ambient temperature and humidity, etc., it is necessary to filter out these interferences through denoising to make the data purer and more reliable. Common denoising methods include using filters, statistical methods or machine learning algorithms, such as Kalman filtering, wavelet transform, etc. After obtaining the preliminary purified data, the preliminary purified data is augmented to obtain enhanced data. Data augmentation is to increase the diversity and representativeness of data, especially when the amount of data is insufficient or the data distribution is unbalanced. Data augmentation can generate more diverse samples, thereby improving the generalization ability and robustness of the model. Augmentation methods can include data rotation, translation, scaling, adding noise and other operations to make the enhanced data more representative. Subsequently, feature selection is performed on the enhanced data to obtain selected feature data. Feature selection is to select the most representative features from a large amount of enhanced data, remove redundant and irrelevant information, thereby reducing the data dimension and improving computational efficiency and model performance. Common feature selection methods include filtering, wrapping, embedding, etc., such as using information gain, chi-square test, L1 regularization and other methods for feature selection. Then, principal component analysis is performed on the selected feature data to obtain key feature data. Principal component analysis (PCA) is a dimensionality reduction technique that can retain the most important information in the data and remove noise and redundancy by mapping high-dimensional data to low-dimensional space. Principal component analysis is very effective in processing high-dimensional data, which can significantly reduce the dimension of the data, improve computational efficiency, and retain the main features of the data. Through PCA, the selected feature data can be converted into several principal components, which constitute the key feature data. Finally, the key feature data is analyzed in the time domain to obtain time-frequency characteristic information, and the time-frequency characteristic information is used as the optimized environmental odor data. Time domain analysis is to extract useful information from time series data. Common methods include Fourier transform, wavelet transform, etc. Time domain analysis can reveal the frequency components and time characteristics of the data. The time-frequency characteristic information includes not only the concentration value of the odor component, but also its changing trend and periodicity in different time periods.This information can help better understand and predict the changing patterns of environmental odor. In the example of an industrial zone, through the above steps, optimized environmental odor data is finally obtained, which can be used for applications such as environmental monitoring, pollution source tracing, and odor classification. For example, by real-time monitoring and analysis of odor data, abnormal odor events can be discovered and warned in a timely manner, helping management departments to take corresponding countermeasures, thereby effectively controlling and improving environmental quality. The entire process not only improves the accuracy and reliability of the data, but also provides a solid data foundation for subsequent analysis and decision-making.

[0033] In a specific embodiment, feature extraction and decomposition are performed on the optimized environmental odor data to obtain a multi-dimensional vector of environmental odor, including: Normalizing the optimized environmental odor data to obtain normalized environmental odor data; Performing multi-dimensional feature extraction on the normalized environmental odor data to obtain multi-dimensional extracted features; wherein the multi-dimensional extracted features include spectrum energy features and spectrum entropy features; Decomposing the multi-dimensional extracted features in multiple scales and directions by using a preset curvelet transform technique to obtain a curve feature representation; Performing dimensionality reduction processing on the curve feature representation to obtain a compressed feature vector; The compressed feature vector is linearly transformed to obtain a multi-dimensional vector of environmental odor.

[0034] Specifically, in a specific embodiment, by extracting and decomposing features of the environmental odor data, a multi-dimensional vector of the environmental odor can be obtained. First, the optimized environmental odor data is normalized. Normalization is to convert the data into a unified scale range so that different features are comparable and the influence of the dimension is eliminated. For example, the maximum-minimum normalization can be used to map the data value to the [0,1] interval, or the Z-score standardization can be used to convert the data into a distribution form with a mean of 0 and a standard deviation of 1. The normalized data is normalized environmental odor data. Next, multi-dimensional feature extraction is performed on the normalized environmental odor data to obtain multi-dimensional extraction features. In this process, the spectral energy features and spectral entropy features of the odor data can be extracted. The spectral energy feature is obtained by Fourier transforming the odor data to obtain a frequency domain representation, from which the energy distribution of each frequency band is calculated, which can reflect the frequency characteristics of the odor components and their changes. The spectral entropy feature uses the concept of entropy to measure the complexity and uncertainty of the odor data spectrum and better capture the structural information of the odor signal. Then, the multi-dimensional extracted features are decomposed in multiple scales and multiple directions by the preset curvelet transform technology to obtain the curve feature representation. Curvelet transform is an advanced signal processing technology that can decompose signals in multiple scales and multiple directions to capture the local characteristics and geometric structure of the signal. Compared with traditional Fourier transform and wavelet transform, curvelet transform has better performance when processing signals with complex geometric structures. In the specific implementation, a suitable curvelet basis function can be selected to perform multi-scale decomposition to obtain feature representations at different scales, and decompose in different directions at the same time to extract directional features. The curve feature representation obtained in this way can better describe the details and global characteristics of the odor data. Next, the curve feature representation is subjected to dimensionality reduction processing to obtain a compressed feature vector. Dimensionality reduction processing is to reduce the dimension of the feature, reduce the computational complexity, and retain the main information of the data. In the specific implementation, dimensionality reduction techniques such as principal component analysis (PCA) can be used to project the high-dimensional curve feature representation into a low-dimensional space to obtain a compressed feature vector. PCA achieves dimensionality reduction by finding the main change direction of the data and taking the projection of the data in these directions as a new feature. Finally, the compressed feature vector is linearly transformed to obtain a multi-dimensional vector of environmental odor. Linear transformation is to transform the compressed feature vector into a new feature space through matrix operation, so that different features are more independent and discriminative. For example, techniques such as linear discriminant analysis (LDA) can be used to optimize the inter-class distance and intra-class distance to obtain a feature vector with better classification performance. Through this series of steps, the final multi-dimensional vector of environmental odor can fully and accurately describe the characteristics of environmental odor.For example, in the application scenario of industrial areas, the odor data collected in real time through the sensor network can be processed through the above processing steps to obtain normalized environmental odor data. Through multi-dimensional feature extraction, the spectrum energy characteristics and spectrum entropy characteristics representing the odor components are extracted, and then multi-scale and multi-directional decomposition is performed through the curvelet transform technology to obtain curve feature representation. These feature representations are processed by dimensionality reduction to obtain more concise compressed feature vectors, and finally through linear transformation, the multi-dimensional vectors of environmental odor are obtained. These multi-dimensional vectors can be used for applications such as environmental monitoring, pollution source tracing and odor classification. For environmental monitoring, abnormal odor events can be discovered in time by analyzing the changing trend of multi-dimensional vectors, and corresponding countermeasures can be taken; for pollution source tracing, the location and nature of the pollution source can be determined by comparing the multi-dimensional vectors of different locations; for odor classification, the multi-dimensional vectors can be classified by machine learning algorithms to identify different types of odor components. For example, at a monitoring point in an industrial area, by analyzing the multi-dimensional vectors of odor collected in real time, abnormal harmful gas emissions can be found and quickly located to nearby factories, so as to take measures to reduce the impact of pollution on the environment.

[0035] In a specific embodiment, the multi-dimensional vector of the environmental odor is classified and predicted to obtain the predicted type of the environmental odor, including: Obtain a multi-dimensional vector of historical environmental odors, and use the multi-dimensional vector of historical environmental odors to embed in a high-dimensional space to obtain a feature space matrix; Perform hyperplane partitioning on the feature space matrix to obtain the environmental odor category boundary; Classifying the multi-dimensional vector of the environmental odor based on the environmental odor category boundary to obtain an environmental odor category label; The environmental odor category labels are fused by multimodal fusion technology to obtain environmental odor category confidence; Performing weighted scoring calculation on the environmental odor category labels based on the environmental odor category confidence to obtain a weighted ranking of environmental odor prediction types; The environmental odor category label with the highest weighted ranking among the environmental odor prediction types is used as the environmental odor prediction type.

[0036] Specifically, in the process of classifying and predicting the multi-dimensional vector of environmental odor to obtain the predicted type of environmental odor, it is first necessary to obtain the multi-dimensional vector of historical environmental odor. These multi-dimensional vectors of historical environmental odor are collected from past odor monitoring data. By analyzing and processing these historical data, reference and basis can be provided for current and future odor prediction. By using these multi-dimensional vectors of historical environmental odor to embed in high-dimensional space, these multi-dimensional vectors can be mapped into a feature space matrix, which contains the key feature information of environmental odor. In the odor monitoring scenario of industrial areas, the multi-dimensional vector of historical environmental odor may include the concentration values ​​of various odor components collected in different time periods and the corresponding time characteristics. Through high-dimensional space embedding, these multi-dimensional vectors can be converted into a feature space matrix, in which each row in the matrix represents a multi-dimensional vector of historical environmental odor, and each column represents a feature dimension. Next, the feature space matrix is ​​hyperplane partitioned to obtain the boundary of environmental odor categories. Hyperplane partitioning is a machine learning algorithm that separates different environmental odor categories by finding the best hyperplane in the feature space. This process can be implemented using classification algorithms such as support vector machines (SVM). By dividing the feature space matrix by hyperplane, the boundaries of each environmental odor category can be determined, and these boundaries can help us classify the new environmental odor multidimensional vector in the feature space. In the classification process, based on the environmental odor category boundary, we can classify the new environmental odor multidimensional vector to obtain the environmental odor category label. This process is to map the new environmental odor multidimensional vector to the feature space matrix and determine which odor category the vector belongs to based on the predetermined environmental odor category boundary. For example, in the odor monitoring scenario of an industrial area, if a new environmental odor multidimensional vector is mapped to the feature space matrix and falls in an area indicating a high VOC concentration, then the vector may be classified as a "VOC odor" category. In order to improve the accuracy and robustness of the classification, we can fuse the environmental odor category labels through multimodal fusion technology to obtain the environmental odor category confidence. Multimodal fusion technology is a method that comprehensively utilizes multiple data sources and classification models. By fusing different data and model results, the reliability and accuracy of the classification results can be improved. For example, in odor monitoring in industrial areas, multiple sensor data and different classification models can be combined to fuse odor category labels to obtain more accurate environmental odor category confidence. Based on the environmental odor category confidence, we can perform weighted scoring calculations on the environmental odor category labels to obtain weighted rankings of environmental odor prediction types. Weighted scoring calculation is a method of weighted summation of the confidences of different category labels. Through this method, the comprehensive score of each environmental odor category label can be determined and sorted according to the score.In the odor monitoring scenario of industrial areas, through weighted scoring calculation, those odor category labels with higher confidence can be given priority, thereby improving the accuracy of the prediction results. Finally, the environmental odor category label with the highest weighted ranking of the environmental odor prediction type is used as the environmental odor prediction type. This process is to select the odor category label with the highest weighted score as the final environmental odor prediction result. For example, in industrial area odor monitoring, if the VOC odor category label has the highest weighted score, then the final environmental odor prediction type will be "VOC odor". Through the above process, we can realize the classification prediction of multi-dimensional vectors of environmental odors and obtain accurate and reliable environmental odor prediction types. This process not only improves the accuracy of odor monitoring, but also provides important data support for environmental management and decision-making.

[0037] In a specific embodiment, the environmental odor prediction type is subjected to correlation analysis of environmental factors to obtain an environmental odor-environmental factor thermodynamic distribution diagram, including: Simulating the molecular structure corresponding to the environmental odor prediction type to obtain the molecular chemical characteristics of the environmental odor; wherein the molecular chemical characteristics of the environmental odor include the electron density distribution of the environmental odor molecules and the vibration frequency spectrum of the environmental odor molecules, and the molecular chemical characteristics of the environmental odor are used to characterize the interaction potential between the environmental odor molecules and environmental factors; Performing topological data analysis on the molecular chemical characteristics to obtain a molecular topological characteristic map; wherein the molecular topological characteristic map includes a manifold structure of a molecular configuration space; The molecular topological characteristic map and the environmental factors are subjected to dynamic evolution analysis to obtain a density matrix of the environmental odor transmission channel; wherein the density matrix of the environmental odor transmission channel is used to reflect the coherence and decoherence process of the molecules of the environmental odor under different environmental conditions; The diffusion behavior of the environmental odor under different environmental factors is calculated based on the density matrix of the environmental odor transmission channel through the vortex dynamics model; Performing Helmholtz-Hodge decomposition on the diffusion behavior to obtain the vortex field distribution of environmental odor diffusion; wherein the vortex field distribution of environmental odor diffusion includes vortex core position, vortex intensity and vortex interaction energy; Performing fractal dimension analysis on the vortex field distribution to obtain multi-scale indicators of environmental odor-environmental factors; Through the theory of non-extensive statistical thermodynamics, an environmental odor-environmental factor thermodynamic distribution map is constructed based on the multi-scale indicators; wherein the environmental odor-environmental factor thermodynamic distribution map includes a high concentration area, a transition area and a low concentration area.

[0038] Specifically, the environmental odor prediction type is subjected to correlation analysis of environmental factors to obtain an environmental odor-environmental factor thermodynamic distribution map. This process involves multiple complex steps and methods to fully reveal the behavioral characteristics of environmental odor under different environmental factors. First, the molecular structure corresponding to the environmental odor prediction type is simulated to obtain the molecular chemical characteristics of the environmental odor. This step is achieved through molecular dynamics simulation or quantum chemical calculation, and the electron density distribution and vibration frequency spectrum of the environmental odor molecules can be obtained. These molecular chemical characteristics are used to characterize the interaction potential between environmental odor molecules and environmental factors. For example, in the monitoring scenario of industrial areas, the molecular structure of volatile organic compounds (VOCs) can be simulated to obtain their electron density distribution and vibration frequency spectrum, so as to understand the interaction between VOCs and other components in the environment. Next, the molecular chemical characteristics are subjected to topological data analysis to obtain a molecular topological feature map. This step uses topological data analysis methods, such as persistent homology, to reveal the manifold structure of the molecular configuration space. The molecular topological feature map can intuitively display the topological relationship of molecules in different configuration spaces, so that we can better understand the structural characteristics of molecules and their association with environmental factors. Subsequently, the molecular topological characteristic map and the environmental factors are subjected to dynamic evolution analysis to obtain the density matrix of the environmental odor transmission channel. Through dynamic evolution analysis, the motion trajectory and behavior characteristics of environmental odor molecules under different environmental conditions can be simulated, and the density matrix is ​​used to reflect the coherence and decoherence process of molecules under different environmental conditions. For example, in the application of industrial areas, the transmission path of VOC molecules under different temperature, humidity and wind speed conditions can be simulated to obtain its density matrix, so as to understand the diffusion and sedimentation behavior of VOC under these conditions. Through the vortex dynamics model, the diffusion behavior of environmental odor under different environmental factors is calculated based on the density matrix of the environmental odor transmission channel. This step uses the vortex dynamics model to obtain the diffusion behavior of environmental odor under different environmental factors by calculating the rotation and diffusion characteristics of molecular motion. For example, the diffusion path of VOC molecules in the wind field can be simulated to analyze its diffusion speed and diffusion range. The diffusion behavior is subjected to Helmholtz-Hodge decomposition to obtain the vortex field distribution of environmental odor diffusion. Helmholtz-Hodge decomposition is a mathematical method that can decompose a vector field into an irrotational field and a vortex field to obtain a vortex field distribution. The vortex field distribution includes the vortex core position, vortex strength and vortex interaction energy. This information can help us understand the diffusion and aggregation characteristics of environmental odor molecules. For example, in industrial areas, the vortex field distribution of VOC molecules under different wind field conditions can be obtained by Helmholtz-Hodge decomposition, so as to understand its diffusion and aggregation areas. Next, the vortex field distribution is analyzed by fractal dimension to obtain multi-scale indicators of environmental odor-environmental factors.Fractal dimension analysis is a mathematical method for describing scale invariance in complex systems. Through fractal dimension analysis, the complexity and multi-scale characteristics of vortex field distribution can be quantified. For example, the vortex field distribution of VOC molecules under different wind speed and temperature conditions can be analyzed to obtain its fractal dimension, so as to understand the complexity of its diffusion behavior. Finally, through the non-extensive statistical thermodynamics theory, the environmental odor-environmental factor thermodynamic distribution diagram is constructed based on the multi-scale index. Non-extensive statistical thermodynamics theory is a statistical method for describing non-equilibrium systems, through which multi-scale indicators can be converted into a thermodynamic distribution diagram of environmental odor-environmental factors. The environmental odor-environmental factor thermodynamic distribution diagram includes a high concentration area, a transition area, and a low concentration area, which can intuitively show the distribution characteristics of environmental odor under different environmental factors. For example, in an industrial area, a thermodynamic distribution diagram of VOC under different temperature, humidity and wind speed conditions can be constructed to understand its concentration distribution and diffusion law under different conditions. In short, through the comprehensive analysis of the above steps, the correlation analysis of environmental factors for the environmental odor prediction type can be realized, and the environmental odor-environmental factor thermodynamic distribution diagram can be obtained. This process not only reveals the molecular chemical characteristics and topological characteristics of environmental odor molecules, but also obtains the density matrix of its transmission channel through dynamic evolution analysis, analyzes its diffusion behavior and vortex field distribution using vortex dynamics model and Helmholtz-Hodge decomposition, and finally constructs the environmental odor-environmental factor thermodynamic distribution map through non-extensive statistical thermodynamics theory. These analysis results can provide scientific basis and technical support for applications such as environmental monitoring, pollution source tracing and odor classification, and help to better understand and control the diffusion and impact of environmental odors.

[0039] In a specific embodiment, the intelligent terminal performs transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram to obtain the environmental odor diffusion path and environmental odor diffusion range, including: By using the odor analysis algorithm preset in the smart terminal, manifold learning is performed on the environmental odor-environmental factor thermodynamic distribution map to obtain an environmental odor transmission topology structure, wherein the environmental odor transmission topology structure includes an odor source point, a transmission node, and a convergence point; Based on the environmental odor transmission topological structure, a non-equilibrium thermodynamic model is performed on the transmission process of the environmental odor to obtain a transmission kinetic equation group, wherein the transmission kinetic equation group includes a mass conservation equation, a momentum conservation equation, and an energy conservation equation; Numerically solving the transmission dynamics equations to obtain the spatiotemporal evolution data of the ambient odor concentration field, wherein the spatiotemporal evolution data of the ambient odor concentration field includes a concentration gradient, a flow velocity field, and a temperature field; Based on the spatiotemporal evolution data of the ambient odor concentration field, a particle Brownian motion simulation of the ambient odor is performed to obtain a set of ambient odor particle trajectories, wherein the set of ambient odor particle trajectories includes a main path, a secondary path, and an abnormal path; Performing path integral analysis and optimal transmission theory calculation on the set of environmental odor particle trajectories to obtain an optimal diffusion network path for the environmental odor, wherein the optimal diffusion network path for the environmental odor includes a trunk path, a branch path, and a critical path; Based on the optimal environmental odor diffusion network path, the environmental odor diffusion range boundary is identified and predicted to obtain the environmental odor diffusion path and the environmental odor diffusion range.

[0040] Specifically, through the intelligent terminal, the environmental odor is transmitted and analyzed based on the environmental odor-environmental factor thermodynamic distribution map to obtain the environmental odor diffusion path and environmental odor diffusion range, including the odor analysis algorithm preset in the intelligent terminal, the environmental odor-environmental factor thermodynamic distribution map is subjected to manifold learning to obtain the environmental odor transmission topology structure, wherein the environmental odor transmission topology structure includes odor source points, transmission nodes and convergence points. In this process, manifold learning is a nonlinear dimensionality reduction technology, which can extract the key features of odor transmission from the high-dimensional environmental odor-environmental factor thermodynamic distribution map to form a transmission topology structure, which is used to represent the transmission path and convergence mode of odor in a complex environment. Based on the environmental odor transmission topology structure, the transmission process of the environmental odor is subjected to non-equilibrium thermodynamic modeling to obtain a transmission kinetic equation group, wherein the transmission kinetic equation group includes mass conservation equations, momentum conservation equations and energy conservation equations. This step aims to simulate the mass, momentum and energy changes of odor molecules during the transmission process by establishing a non-equilibrium thermodynamic model. The mass conservation equation is used to describe the concentration change of odor molecules, the momentum conservation equation is used to describe the motion state of odor molecules, and the energy conservation equation is used to describe the energy change of odor molecules during the transmission process. The transmission dynamics equation group is numerically solved to obtain the spatiotemporal evolution data of the environmental odor concentration field, wherein the spatiotemporal evolution data of the environmental odor concentration field includes concentration gradient, flow velocity field and temperature field. Through numerical solution, the concentration distribution, flow velocity and temperature change of odor molecules at different time and space positions can be obtained, and these data can intuitively reflect the propagation characteristics of odor in the environment. Based on the spatiotemporal evolution data of the environmental odor concentration field, the particle Brownian motion simulation of the environmental odor is performed to obtain the environmental odor particle trajectory set, wherein the environmental odor particle trajectory set includes the main path, the secondary path and the abnormal path. At this stage, through the particle Brownian motion simulation, the motion trajectories of a large number of odor molecules can be tracked, thereby obtaining the actual path of odor diffusion. These trajectory data include the main transmission path, the secondary diffusion path and some abnormal transmission paths, which can help to better understand the complex movement pattern of odor in the environment. The path integral analysis and optimal transmission theory calculation are performed on the set of environmental odor particle trajectories to obtain the optimal diffusion network path of the environmental odor, wherein the optimal diffusion network path of the environmental odor includes a trunk path, a branch path, and a critical path. Through path integral analysis, the optimal path of the odor molecules during the transmission process can be calculated. These optimal paths can most effectively connect the odor source point and the convergence point, thereby forming an optimal diffusion network for the environmental odor. These paths include trunk paths (main transmission paths), branch paths (auxiliary transmission paths), and critical paths (key transmission paths), which are the backbone of the odor transmission network.Based on the optimal diffusion network path of the environmental odor, the boundary of the environmental odor diffusion range is identified and predicted to obtain the environmental odor diffusion path and the environmental odor diffusion range. By analyzing and predicting these optimal diffusion paths, the boundary range of the odor diffusion in the environment can be identified, thereby obtaining the actual diffusion path and coverage of the environmental odor. This information can be used for environmental monitoring and pollution control. For example, in the odor monitoring scene of an industrial area, by real-time analysis of the odor diffusion path and range, abnormal odor events can be discovered and warned in time, helping the management department to take corresponding countermeasures to effectively control and improve the environmental quality. In summary, through the above steps, based on the intelligent terminal and the preset odor analysis algorithm, the environmental odor can be transmitted in detail and analyzed to obtain the environmental odor diffusion path and range. This process not only improves the analysis accuracy of the odor data, but also provides a scientific basis for environmental monitoring and decision-making, thereby realizing effective management and control of environmental quality. For example, in a monitoring scene in an industrial area, through these steps, the propagation path and impact range of harmful gases can be monitored and predicted in real time, helping relevant departments to take countermeasures in time to protect the health and safety of the public.

[0041] In a specific embodiment, the transmission dynamics equations are numerically solved to obtain the spatiotemporal evolution data of the ambient odor concentration field, including: Discretizing the transmission dynamics equations to obtain a discretized equation group; Iteratively solving the discretized equation group to obtain an initial concentration field; Advancing the time step of the initial concentration field to obtain a time-evolved concentration field; Performing spatial grid refinement on the time-evolved concentration field to obtain a grid-refined concentration field; Performing stability check on the refined concentration field to obtain a stable concentration field; The stable concentration field is subjected to a spatiotemporal evolution analysis to obtain the spatiotemporal evolution data of the ambient odor concentration field.

[0042] Specifically, numerically solving the transmission dynamics equations to obtain the spatiotemporal evolution data of the ambient odor concentration field is a complex and delicate process, which includes multiple key steps. First, the transmission dynamics equations are discretized to obtain a discretized equation set. The purpose of this step is to convert the continuous partial differential equation into a discrete form that can be processed on a computer. Common discretization methods include finite difference method, finite element method and finite volume method, which can divide the continuous space and time domain into finite discrete points so that the equation set can be calculated at these discrete points. After the discretization is completed, the discretized equation set is iteratively solved to obtain the initial concentration field. Iterative solution methods generally include explicit iteration methods and implicit iteration methods. Selecting a suitable iteration method can ensure the stability and convergence of the solution. For example, by an explicit time stepping method, the odor concentration distribution at the initial moment can be gradually calculated as the basis for subsequent calculations. Next, the initial concentration field is time-stepped to obtain a time-evolved concentration field. Time step advancement refers to simulating the change of odor concentration over time by gradually advancing the time step under known initial conditions and boundary conditions. In this process, it is necessary to select an appropriate time step to ensure the stability and accuracy of the numerical solution. Too large a time step may cause the numerical solution to diverge, while too small a time step will increase the amount of calculation and calculation time. The time evolution concentration field is spatially meshed to obtain a mesh-refined concentration field. Spatial mesh refinement is to improve spatial resolution and make the simulation results more refined and accurate. By refining the spatial grid, more subtle changes and local features in the odor concentration field can be captured. For example, in the odor monitoring scenario of an industrial area, the refined grid can more accurately describe the odor diffusion near the pollution source. In order to ensure the reliability of the calculation results, the refined concentration field is stability checked to obtain a stable concentration field. Stability checks include convergence and consistency checks of the numerical solution, which can ensure that the numerical solution remains stable and reliable throughout the calculation process. For example, residual analysis and error estimation can be used to determine whether the numerical solution meets the expected accuracy requirements. Finally, the spatiotemporal evolution analysis of the stable concentration field is performed to obtain the spatiotemporal evolution data of the environmental odor concentration field. The purpose of spatiotemporal evolution analysis is to extract useful information from the numerical solution, such as the diffusion speed, diffusion direction and concentration change trend of the odor. These spatiotemporal evolution data include not only the spatial distribution of the odor concentration, but also its dynamic change characteristics over time. In the example of the industrial zone, through the above steps, the spatiotemporal evolution data of the environmental odor concentration field are finally obtained. These data can be used for applications such as environmental monitoring, pollution source tracking and odor diffusion prediction. For example, through the spatiotemporal evolution data, the dynamic changes of odor pollution can be monitored in real time, abnormal odor events can be discovered and warned in time, and management departments can be helped to take corresponding countermeasures, thereby effectively controlling and improving environmental quality.The whole process not only improves the accuracy and reliability of the data, but also provides a solid data foundation for subsequent analysis and decision-making. Through these detailed steps and methods, the accuracy and scientificity of environmental odor transmission analysis can be ensured, and finally comprehensive and reliable spatiotemporal evolution data of environmental odor concentration field can be obtained.

[0043] The above describes the environment odor analysis method based on the smart terminal in the embodiment of the present invention. The following describes the environment odor analysis device based on the smart terminal in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, an environment odor analysis device based on a smart terminal includes: The collection module 21 is used to collect environmental odors in real time through a sensor network to obtain an environmental odor data stream, and pre-process the environmental odor data stream to obtain optimized environmental odor data; An extraction module 22, used for performing feature extraction and decomposition on the optimized environmental odor data to obtain a multi-dimensional vector of the environmental odor; A prediction module 23, used for classifying and predicting the multi-dimensional vector of the environmental odor to obtain a predicted type of the environmental odor; A first analysis module 24 is used to perform correlation analysis of environmental factors on the environmental odor prediction type to obtain an environmental odor-environmental factor thermodynamic distribution diagram; The second analysis module 25 is used to perform transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram through the smart terminal to obtain the environmental odor diffusion path and environmental odor diffusion range.

[0044] In this embodiment, for the specific implementation of each unit in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0045] Reference Figure 3 The present invention also provides a computer device in an embodiment, wherein the internal structure of the computer device can be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0046] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0047] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0048] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0049] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0050] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for analyzing environmental odor based on a smart terminal, characterized in that: The following steps are involved: Collecting environmental odors in real time through a sensor network to obtain an environmental odor data stream, and preprocessing the environmental odor data stream to obtain optimized environmental odor data; Performing feature extraction and decomposition on the optimized environmental odor data to obtain a multi-dimensional vector of the environmental odor; Classify and predict the multi-dimensional vector of environmental odor to obtain a predicted type of environmental odor; Performing correlation analysis of environmental factors on the predicted environmental odor type to obtain an environmental odor-environmental factor thermodynamic distribution diagram; Through the intelligent terminal, the environmental odor is transmitted and analyzed based on the environmental odor-environmental factor thermodynamic distribution diagram to obtain the environmental odor diffusion path and environmental odor diffusion range.

2. The method for analyzing environmental odor based on a smart terminal according to claim 1 is characterized in that: The environmental odor is collected in real time through a sensor network to obtain an environmental odor data stream, and the environmental odor data stream is preprocessed to obtain optimized environmental odor data, including: The odor in the environment is collected in real time through the sensor network to obtain the environmental odor data stream; Performing denoising processing on the environmental odor data stream to obtain preliminary purified data; Performing augmentation processing on the preliminary cleansed data to obtain enhanced data; Performing feature selection on the enhanced data to obtain selected feature data; Performing principal component analysis on the selected feature data to obtain key feature data; The key feature data is analyzed in the time domain to obtain time-frequency characteristic information, and the time-frequency characteristic information is used as the optimized environmental odor data.

3. The method for analyzing environmental odor based on a smart terminal according to claim 1 is characterized in that: Feature extraction and decomposition are performed on the optimized environmental odor data to obtain a multi-dimensional vector of environmental odor, including: Normalizing the optimized environmental odor data to obtain normalized environmental odor data; Performing multi-dimensional feature extraction on the normalized environmental odor data to obtain multi-dimensional extracted features; wherein the multi-dimensional extracted features include spectrum energy features and spectrum entropy features; Decomposing the multi-dimensional extracted features in multiple scales and directions by using a preset curvelet transform technique to obtain a curve feature representation; Performing dimensionality reduction processing on the curve feature representation to obtain a compressed feature vector; The compressed feature vector is linearly transformed to obtain a multi-dimensional vector of environmental odor.

4. The method for analyzing environmental odor based on a smart terminal according to claim 1 is characterized in that: Classify and predict the multi-dimensional vector of the environmental odor to obtain the predicted type of the environmental odor, including: Obtain a multi-dimensional vector of historical environmental odors, and use the multi-dimensional vector of historical environmental odors to embed in a high-dimensional space to obtain a feature space matrix; Perform hyperplane partitioning on the feature space matrix to obtain the environmental odor category boundary; Classifying the multi-dimensional vector of the environmental odor based on the environmental odor category boundary to obtain an environmental odor category label; The environmental odor category labels are fused by multimodal fusion technology to obtain environmental odor category confidence; Performing weighted scoring calculation on the environmental odor category labels based on the environmental odor category confidence to obtain a weighted ranking of environmental odor prediction types; The environmental odor category label with the highest weighted ranking among the environmental odor prediction types is used as the environmental odor prediction type.

5. The method for analyzing environmental odor based on a smart terminal according to claim 1 is characterized in that: The environmental odor prediction type is subjected to correlation analysis of environmental factors to obtain an environmental odor-environmental factor thermodynamic distribution diagram, including: Simulating the molecular structure corresponding to the environmental odor prediction type to obtain the molecular chemical characteristics of the environmental odor; wherein the molecular chemical characteristics of the environmental odor include the electron density distribution of the environmental odor molecules and the vibration frequency spectrum of the environmental odor molecules, and the molecular chemical characteristics of the environmental odor are used to characterize the interaction potential between the environmental odor molecules and environmental factors; Performing topological data analysis on the molecular chemical characteristics to obtain a molecular topological characteristic map; wherein the molecular topological characteristic map includes a manifold structure of a molecular configuration space; The molecular topological characteristic map and the environmental factors are subjected to dynamic evolution analysis to obtain a density matrix of the environmental odor transmission channel; wherein the density matrix of the environmental odor transmission channel is used to reflect the coherence and decoherence process of the molecules of the environmental odor under different environmental conditions; The diffusion behavior of the environmental odor under different environmental factors is calculated based on the density matrix of the environmental odor transmission channel through the vortex dynamics model; Performing Helmholtz-Hodge decomposition on the diffusion behavior to obtain the vortex field distribution of environmental odor diffusion; wherein the vortex field distribution of environmental odor diffusion includes vortex core position, vortex intensity and vortex interaction energy; Performing fractal dimension analysis on the vortex field distribution to obtain multi-scale indicators of environmental odor-environmental factors; Through the theory of non-extensive statistical thermodynamics, an environmental odor-environmental factor thermodynamic distribution map is constructed based on the multi-scale indicators; wherein the environmental odor-environmental factor thermodynamic distribution map includes a high concentration area, a transition area and a low concentration area.

6. The method for analyzing environmental odor based on a smart terminal according to claim 1, characterized in that: Through the intelligent terminal, the environmental odor is transmitted and analyzed based on the environmental odor-environmental factor thermodynamic distribution diagram to obtain the environmental odor diffusion path and environmental odor diffusion range, including: By using the odor analysis algorithm preset in the smart terminal, manifold learning is performed on the environmental odor-environmental factor thermodynamic distribution map to obtain an environmental odor transmission topology structure, wherein the environmental odor transmission topology structure includes an odor source point, a transmission node, and a convergence point; Based on the environmental odor transmission topological structure, a non-equilibrium thermodynamic model is performed on the transmission process of the environmental odor to obtain a transmission kinetic equation group, wherein the transmission kinetic equation group includes a mass conservation equation, a momentum conservation equation, and an energy conservation equation; Numerically solving the transmission dynamics equations to obtain the spatiotemporal evolution data of the ambient odor concentration field, wherein the spatiotemporal evolution data of the ambient odor concentration field includes a concentration gradient, a flow velocity field, and a temperature field; Based on the spatiotemporal evolution data of the ambient odor concentration field, a particle Brownian motion simulation of the ambient odor is performed to obtain a set of ambient odor particle trajectories, wherein the set of ambient odor particle trajectories includes a main path, a secondary path, and an abnormal path; Performing path integral analysis and optimal transmission theory calculation on the set of environmental odor particle trajectories to obtain an optimal diffusion network path for the environmental odor, wherein the optimal diffusion network path for the environmental odor includes a trunk path, a branch path, and a critical path; Based on the optimal environmental odor diffusion network path, the environmental odor diffusion range boundary is identified and predicted to obtain the environmental odor diffusion path and the environmental odor diffusion range.

7. The method for analyzing environmental odor based on a smart terminal according to claim 6 is characterized in that: The transmission dynamics equations are numerically solved to obtain the spatiotemporal evolution data of the ambient odor concentration field, including: Discretizing the transmission dynamics equations to obtain a discretized equation group; Iteratively solving the discretized equation group to obtain an initial concentration field; Advancing the time step of the initial concentration field to obtain a time-evolved concentration field; Performing spatial grid refinement on the time-evolved concentration field to obtain a grid-refined concentration field; Performing stability check on the refined concentration field to obtain a stable concentration field; The stable concentration field is subjected to a spatiotemporal evolution analysis to obtain the spatiotemporal evolution data of the ambient odor concentration field.

8. An environmental odor analysis device based on a smart terminal, characterized in that: include: A collection module, used to collect environmental odors in real time through a sensor network to obtain an environmental odor data stream, and pre-process the environmental odor data stream to obtain optimized environmental odor data; An extraction module, used for performing feature extraction and decomposition on the optimized environmental odor data to obtain a multi-dimensional vector of the environmental odor; A prediction module, used to classify and predict the multi-dimensional vector of environmental odor to obtain a predicted type of environmental odor; A first analysis module is used to perform correlation analysis of environmental factors on the environmental odor prediction type to obtain an environmental odor-environmental factor thermodynamic distribution diagram; The second analysis module is used to perform transmission analysis on the environmental odor based on the environmental odor-environmental factor thermodynamic distribution diagram through the smart terminal to obtain the environmental odor diffusion path and environmental odor diffusion range.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images