Intelligent-based early warning system capable of automatically identifying abnormal carbon emission data

Abstract

The invention discloses an intelligent-based abnormal carbon emission data automatic identification early warning system, which belongs to the technical field of intellectualization and comprises a data acquisition preprocessing module, a high-precision space-time analysis module, an intelligent analysis module, an automatic abnormal identification module, a carbon footprint tracing module, a self-adaptive adjustment module and an intelligent emission prediction module. A carbon emission source and time-space distribution characteristics are accurately positioned through spatial positioning and time sequence analysis of the high-precision time-space analysis module, a spatial distribution diagram is more detailed and accurate through the optimized sensor position and an interpolation algorithm, and time sequence analysis is more accurate through timestamp correction and power spectrum density analysis. The method effectively evaluates the periodic intensity of the signal, analyzes a hidden mode and an association rule in the data through the intelligent analysis module integrating the spatial-temporal characteristics and the related information of the data acquisition and preprocessing module, analyzes the spatial-temporal association between variables through the calculation of a clustering center and the association intensity, and facilitates the discovery of a potential abnormal mode.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent technology, and specifically refers to an intelligent-based automatic recognition and early warning system for abnormal carbon emission data. Background Art

[0002] As global attention to climate change increases, accurately monitoring and managing carbon emissions has become crucial. However, traditional carbon emission monitoring methods have problems such as low data processing efficiency, poor accuracy, and difficulty in detecting anomalies in a timely manner.

[0003] However, the existing automatic identification of abnormal carbon emission data early warning system still has certain defects. The existing automatic identification of abnormal carbon emission data early warning system relies on the original sensor location information and lacks optimization algorithms to accurately locate the carbon emission source, resulting in the spatial distribution map not being detailed and accurate enough. It is relatively weak in timestamp correction and signal periodicity intensity assessment, and cannot effectively handle clock offsets or evaluate complex time series characteristics, which affects the accurate judgment of emission trends. It only relies on simple statistical analysis and fails to fully utilize the comprehensive spatiotemporal characteristics and multi-dimensional data for in-depth data mining, making it difficult to discover hidden patterns and association rules. It lacks effective distance calculation methods and cannot accurately quantify the similarity between different points, which affects the effect of cluster analysis. The existing anomaly scoring criteria are too simple and fail to cover multiple dimensions, resulting in incomplete and inaccurate anomaly identification. There is a lack of integrated data sets and advanced clustering results, and it cannot clearly present the spatiotemporal aggregation of abnormal patterns, which affects the precise positioning of carbon emission sources. For this reason, an intelligent automatic identification of abnormal carbon emission data early warning system is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent and automatic early warning system for identifying abnormal carbon emission data to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions: an intelligent and automatic recognition of abnormal carbon emission data early warning system, comprising a data acquisition and preprocessing module, a high-precision spatiotemporal analysis module, an intelligent analysis module, an automatic anomaly recognition module, a carbon footprint tracking and tracing module, an adaptive adjustment module, and an intelligent emission prediction module;

[0006] The data acquisition and preprocessing module is wirelessly connected to the high-precision spatiotemporal analysis module, the high-precision spatiotemporal analysis module is wirelessly connected to the intelligent analysis module, the intelligent analysis module is wirelessly connected to the automatic anomaly identification module, the automatic anomaly identification module is wirelessly connected to the carbon footprint tracking and tracing module, the automatic anomaly identification module is wirelessly connected to the adaptive adjustment module, the intelligent analysis module is wirelessly connected to the adaptive adjustment module, the intelligent analysis module is wirelessly connected to the intelligent emission prediction module, and the intelligent emission prediction module is wirelessly connected to the adaptive adjustment module;

[0007] The data acquisition and preprocessing module is used to obtain carbon emissions and environmental parameter information data in real time based on integrated sensors and Internet of Things devices, and preprocess the obtained data information;

[0008] The high-precision spatiotemporal analysis module is used to perform spatial positioning and time series analysis based on the processed data to accurately locate carbon emission sources and their spatiotemporal distribution characteristics;

[0009] The intelligent analysis module is used to analyze the hidden patterns and association rules in the data based on the data acquisition preprocessing data combined with the spatial and temporal dimension information of the high-precision spatiotemporal analysis module;

[0010] The automatic anomaly identification module is used to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module;

[0011] The carbon footprint tracking module is used to accurately locate the source of carbon emissions based on the identified abnormal patterns, combined with the distribution information of the high-precision spatiotemporal analysis module and the patterns analyzed by the intelligent analysis module, and formulate responsibility allocation and recommendations;

[0012] The intelligent emission prediction module is used to predict carbon emission trends based on the results of the intelligent analysis module and provide feedback to the adaptive adjustment module;

[0013] The adaptive adjustment module is used to dynamically adjust parameters according to the real-time analysis results, automatic anomaly identification results and prediction results of the intelligent analysis module, and feed back to the automatic anomaly identification module.

[0014] Among them, the data acquisition and preprocessing module is used to obtain carbon emission and environmental parameter information data in real time based on integrated sensors and Internet of Things devices, and preprocess the obtained data information; the Internet of Things devices obtain carbon emission parameter data and environmental parameter data from integrated sensors in real time through a set high-frequency acquisition frequency, and perform data cleaning on the obtained parameter data, including format verification, outlier extraction and processing of missing values. The cleaned parameter data is format converted and normalized, and the integrated data is stored in the database.

[0015] The high-precision spatiotemporal analysis module is used to perform spatial positioning and time series analysis based on the processed data, accurately locate the carbon emission source and the spatiotemporal distribution characteristics; optimize according to the position information obtained by the sensor, and assume that the position information obtained by the sensor is P i , the total number of sensors is N, and the weight factor is defined as w i (t) = exp(-λ||P i -P ref (t)|| 2 ), λ represents the adjustment parameter, P ref (t) represents the reference position, and the objective function J(P) is the weighted sum of square errors between all sensor positions and the optimized position:

[0016]

[0017] Assume that the optimized sensor position P opt (t), minimize the objective function, and the implementation formula is:

[0018]

[0019] Interpolate according to the optimized sensor position, and define the kernel function as K h (d,t)=exp(-γd 2 / h(t) 2 ), d represents the distance, h(t) represents the bandwidth that changes with time, and γ represents the adjustment parameter. The interpolation implementation formula is:

[0020]

[0021] In the formula, v(x,t) represents the interpolation value, z j (t) represents the measurement value of the jth sensor at time t, M represents the number of neighboring points involved in interpolation, P opt,j (t) represents the precise position coordinate of the jth sensor at time t after optimization, and t represents the current time point. Interpolation is performed based on the optimized sensor position to generate a more detailed spatial distribution map.

[0022] In the time series analysis, let t i is the timestamp of the i-th sensor. The timestamp is corrected and the implementation formula is:

[0023]

[0024] In the formula, t i ′ represents the corrected timestamp, r i (t) represents the relative clock offset rate, dt represents the time differential;

[0025] Let the original measurement value be z i (t i ) is mapped to the corrected timestamp t′ i , construct a new time series z i (t i ′), perform Fourier transform on the new time series to get the frequency domain representation Z(f), and let the power spectrum density p(f) reflect the energy distribution of the signal at different frequencies, which is defined as:

[0026] P(f)=|Z(f)| 2 ;

[0027] The spectral entropy is calculated based on the power spectral density to evaluate the periodicity strength of the signal. The implementation formula is:

[0028] S(t)=-∑ f p(f)logp(f);

[0029] Assume that the mass of the emission source is m i , the feature extraction implementation formula is:

[0030]

[0031] In the formula, G(s,t) represents the extracted spatiotemporal features, s represents the spatial position, t represents the time point, α represents the spatial attenuation coefficient, β represents the temporal attenuation coefficient, and s i represents the spatial location of the i-th emission source, t i represents the time point of the i-th emission source.

[0032] The intelligent analysis module is used to analyze the hidden patterns and association rules in the data based on the data collection preprocessing data combined with the spatial and temporal dimension information of the high-precision spatiotemporal analysis module; the comprehensive spatiotemporal features G(s, t), the optimized geographical location P opt (t) and the relevant data in the data acquisition preprocessing module are the comprehensive feature vector F i (t), according to the comprehensive feature vector analysis, let the comprehensive feature distance D ij (t), measures the similarity between different points, and the implementation formula is:

[0033]

[0034] In the formula, D ij (t) represents the comprehensive feature distance between the i-th and j-th sensors at time t, m represents the dimension of the comprehensive feature vector, F ik (t) and F jk (t) represents the kth characteristic component of the i-th and j-th sensor at time t, respectively.

[0035] Among them, according to the comprehensive feature distance D ij (t) Calculate each cluster center and the implementation formula is:

[0036]

[0037] In the formula, C k represents the kth cluster center feature vector, c k represents all points in the kth cluster, D i,F (t) represents the comprehensive feature distance between the i-th sensor and the candidate center F at time t;

[0038] Let the correlation strength be I XY (s, t), the formula for measuring the correlation between variables X and Y in time and space is:

[0039]

[0040] In the formula, K represents the number of samples, x k and y k are the k-th observation values ​​of variables X and Y, respectively. and represents the mean of variables X and Y, γ and δ represent the spatial and temporal attenuation coefficients, s and t represent the current spatial position and time, s k and t k represents the spatial position and time of the k-th observation.

[0041] Among them, the automatic anomaly identification module is used to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module; define multi-dimensional anomaly scoring standards based on the comprehensive feature vectors, clustering results and association rules of the intelligent analysis module, and dynamically adjust the weights of each scoring dimension according to the characteristics of different time periods and regions. In the time dimension, it identifies short-term and long-term abnormal trends; in the spatial dimension, it identifies abnormal clustering phenomena in specific areas; in the association dimension, it identifies abnormal relationships with other variables; through pattern matching with the abnormal pattern library, newly discovered anomalies are matched, and thresholds are adaptively set according to the abnormal score and classification results. When the abnormal score exceeds the set threshold, an early warning response is issued and feedback is given to management personnel.

[0042] Among them, the carbon footprint tracking and tracing module is used to accurately locate the source of carbon emissions based on the identified abnormal patterns and combine the distribution information of the high-precision spatiotemporal analysis module with the patterns analyzed by the intelligent analysis module, and formulate responsibility allocation and suggestions; obtain the abnormal pattern data of the automatic abnormality identification module, combine the spatiotemporal characteristics provided by the high-precision spatiotemporal analysis module and the integrated optimized geographic coordinates into a unified data set, analyze the clustering of abnormal patterns in time and space based on the data set, identify abnormal clustering phenomena within a certain time period and within a geographical area, find the possible emission source location and its impact range through clustering results and association rules, track the propagation path of abnormal emissions from the source to the detection point, combine meteorological conditions to reversely deduce the emission source, and display the spatial distribution and temporal evolution of abnormal emissions through spatiotemporal heat maps to intuitively reflect the emission path, integrate multiple data sources, cross-validate the tracing results, and analyze the historical emission records, operating conditions and management measures of the responsible party based on the results of the tracing analysis, evaluate the degree of its impact on abnormal emissions and make responsibilities, and put forward improvement suggestions based on the causes and characteristics of abnormal emissions.

[0043] Among them, the intelligent emission prediction module is used to predict carbon emission trends based on the results of the intelligent analysis module and feed back to the adaptive adjustment module; obtain the comprehensive feature vector, clustering results and association rule information of the intelligent analysis module, make predictions based on the results of the analysis of the intelligent analysis module, and make long-term predictions based on historical trends, generate a detailed prediction report based on the prediction results and feed back to the adaptive adjustment module.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. This invention precisely locates the source of carbon emissions and their spatiotemporal distribution characteristics through spatial positioning and time series analysis using a high-precision spatiotemporal analysis module. Its optimized sensor positioning and interpolation algorithm make the spatial distribution map more detailed and accurate, helping to accurately track the source of carbon emissions. Time series analysis effectively assesses the periodicity strength of the signal through corrected timestamps and power spectral density analysis.

[0046] 2. The present invention uses an intelligent analysis module to integrate spatiotemporal features and relevant information from the data acquisition and preprocessing module to analyze the hidden patterns and association rules in the data. The integrated feature vector and distance calculations quantify the similarities between different points. Furthermore, by calculating cluster centers and association strengths, the spatiotemporal correlations between variables are analyzed, which helps to discover potential abnormal patterns.

[0047] 3. The present invention uses an automatic anomaly recognition module to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module. The multi-dimensional anomaly scoring standard makes anomaly identification more accurate and comprehensive, and the dynamic adjustment of weights adapts to the characteristics of different time periods and regions. By matching with the anomaly pattern library and adaptively setting thresholds, it can provide timely warnings and feedback to management personnel, helping to take timely measures to reduce carbon emissions.

[0048] 4. The present invention accurately locates the source of carbon emissions based on the identified abnormal patterns and information from the high-precision spatiotemporal analysis module through the carbon footprint tracking and tracing module, and formulates responsibility allocation and suggestions. The integrated data set and clustering results enable the spatiotemporal aggregation of abnormal patterns to be clearly presented, which helps to track the propagation path of abnormal emissions. The spatiotemporal heat map and cross-validation tracing results can intuitively reflect the emission path, and based on the results of the tracing analysis, improvement suggestions are put forward, which helps to improve the level of carbon emission management. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a schematic diagram of the structure of the intelligent automatic recognition of abnormal carbon emission data early warning system of the present invention;

[0050] Figure 2 This is a flowchart of the operation of the high-precision spatiotemporal analysis module of the intelligent automatic identification of abnormal carbon emission data early warning system of the present invention;

[0051] Figure 3 This is a flowchart of the operation of the intelligent analysis module of the intelligent automatic identification of abnormal carbon emission data early warning system of the present invention;

[0052] Figure 4 This is a flowchart of the operation of the automatic anomaly identification module of the intelligent automatic identification of abnormal carbon emission data early warning system of the present invention. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0054] Example

[0055] See also Figure 1-Figure 4 As shown, the present invention provides a technical solution: including a data acquisition and preprocessing module, a high-precision spatiotemporal analysis module, an intelligent analysis module, an automatic anomaly recognition module, a carbon footprint tracking and tracing module, an adaptive adjustment module and an intelligent emission prediction module;

[0056] The data acquisition and preprocessing module is wirelessly connected to the high-precision spatiotemporal analysis module, the high-precision spatiotemporal analysis module is wirelessly connected to the intelligent analysis module, the intelligent analysis module is wirelessly connected to the automatic anomaly identification module, the automatic anomaly identification module is wirelessly connected to the carbon footprint tracking and tracing module, the automatic anomaly identification module is wirelessly connected to the adaptive adjustment module, the intelligent analysis module is wirelessly connected to the adaptive adjustment module, the intelligent analysis module is wirelessly connected to the intelligent emission prediction module, and the intelligent emission prediction module is wirelessly connected to the adaptive adjustment module;

[0057] The data acquisition and preprocessing module is used to obtain carbon emissions and environmental parameter information data in real time based on integrated sensors and Internet of Things devices, and preprocess the obtained data information;

[0058] The high-precision spatiotemporal analysis module is used to perform spatial positioning and time series analysis based on the processed data to accurately locate carbon emission sources and their spatiotemporal distribution characteristics;

[0059] The intelligent analysis module is used to analyze the hidden patterns and association rules in the data based on the data acquisition preprocessing data combined with the spatial and temporal dimension information of the high-precision spatiotemporal analysis module;

[0060] The automatic anomaly identification module is used to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module;

[0061] The carbon footprint tracking module is used to accurately locate the source of carbon emissions based on the identified abnormal patterns, combined with the distribution information of the high-precision spatiotemporal analysis module and the patterns analyzed by the intelligent analysis module, and formulate responsibility allocation and recommendations;

[0062] The intelligent emission prediction module is used to predict carbon emission trends based on the results of the intelligent analysis module and provide feedback to the adaptive adjustment module;

[0063] The adaptive adjustment module is used to dynamically adjust parameters according to the real-time analysis results, automatic anomaly identification results and prediction results of the intelligent analysis module, and feed back to the automatic anomaly identification module.

[0064] Among them, the data acquisition and preprocessing module is used to obtain carbon emission and environmental parameter information data in real time based on integrated sensors and Internet of Things devices, and preprocess the obtained data information; the Internet of Things devices obtain carbon emission parameter data and environmental parameter data from integrated sensors in real time through a set high-frequency acquisition frequency, and perform data cleaning on the obtained parameter data, including format verification, outlier extraction and processing of missing values. The cleaned parameter data is format converted and normalized, and the integrated data is stored in the database.

[0065] The high-precision spatiotemporal analysis module is used to perform spatial positioning and time series analysis based on the processed data, accurately locate the carbon emission source and the spatiotemporal distribution characteristics; optimize according to the position information obtained by the sensor, and assume that the position information obtained by the sensor is P i , the total number of sensors is N, and the weight factor is defined as w i (t) = exp(-λ||P i -P ref (t)|| 2 ), λ represents the adjustment parameter, P ref (t) represents the reference position, and the objective function J(P) is the weighted sum of square errors between all sensor positions and the optimized position:

[0066]

[0067] Assume that the optimized sensor position P opt (t), minimize the objective function, and the implementation formula is:

[0068]

[0069] Interpolate according to the optimized sensor position, and define the kernel function as K h (d,t)=exp(-γd 2 / h(t) 2 ), d represents the distance, h(t) represents the bandwidth that changes with time, and γ represents the adjustment parameter. The interpolation implementation formula is:

[0070]

[0071] In the formula, v(x,t) represents the interpolation value, z j (t) represents the measurement value of the jth sensor at time t, M represents the number of neighboring points involved in interpolation, P opt,j (t) represents the precise position coordinate of the jth sensor at time t after optimization, and t represents the current time point. Interpolation is performed based on the optimized sensor position to generate a more detailed spatial distribution map.

[0072] In the time series analysis, let t i is the timestamp of the i-th sensor. The timestamp is corrected and the implementation formula is:

[0073]

[0074] In the formula, t i ′ represents the corrected timestamp, r i (t) represents the relative clock offset rate, dt represents the time differential;

[0075] Let the original measurement value be z i (t i ) is mapped to the corrected timestamp t i ′, construct a new time series z i (t i ′), perform Fourier transform on the new time series to get the frequency domain representation Z(f), and let the power spectrum density p(f) reflect the energy distribution of the signal at different frequencies, which is defined as:

[0076] P(f)=|Z(f)| 2 ;

[0077] The spectral entropy is calculated based on the power spectral density to evaluate the periodicity strength of the signal. The implementation formula is:

[0078] S(t)=-∑ f p(f)logp(f);

[0079] Assume that the mass of the emission source is m i , the feature extraction implementation formula is:

[0080]

[0081] In the formula, G(s,t) represents the extracted spatiotemporal features, s represents the spatial position, t represents the time point, α represents the spatial attenuation coefficient, β represents the temporal attenuation coefficient, and s i represents the spatial location of the i-th emission source, t i represents the time point of the i-th emission source.

[0082] The intelligent analysis module is used to analyze the hidden patterns and association rules in the data based on the data collection preprocessing data combined with the spatial and temporal dimension information of the high-precision spatiotemporal analysis module; the comprehensive spatiotemporal features G(s, t), the optimized geographical location P opt (t) and the relevant data in the data acquisition preprocessing module are the comprehensive feature vector F i (t), according to the comprehensive feature vector analysis, let the comprehensive feature distance D ij (t), measures the similarity between different points, and the implementation formula is:

[0083]

[0084] In the formula, D ij (t) represents the comprehensive feature distance between the i-th and j-th sensors at time t, m represents the dimension of the comprehensive feature vector, F ik (t) and F jk (t) represents the kth characteristic component of the i-th and j-th sensor at time t, respectively.

[0085] Among them, according to the comprehensive feature distance D ij (t) Calculate each cluster center and the implementation formula is:

[0086]

[0087] In the formula, C k represents the kth cluster center feature vector, c k represents all points in the kth cluster, D i,F (t) represents the comprehensive feature distance between the i-th sensor and the candidate center F at time t;

[0088] Let the correlation strength be I XY (s, t), the formula for measuring the correlation between variables X and Y in time and space is:

[0089]

[0090] In the formula, K represents the number of samples, x k and y k are the k-th observation values ​​of variables X and Y, respectively. and represents the mean of variables X and Y, γ and δ represent the spatial and temporal attenuation coefficients, s and t represent the current spatial position and time, s k and t k represents the spatial position and time of the k-th observation.

[0091] Among them, the automatic anomaly identification module is used to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module; define multi-dimensional anomaly scoring standards based on the comprehensive feature vectors, clustering results and association rules of the intelligent analysis module, and dynamically adjust the weights of each scoring dimension according to the characteristics of different time periods and regions. In the time dimension, it identifies short-term and long-term abnormal trends; in the spatial dimension, it identifies abnormal clustering phenomena in specific areas; in the association dimension, it identifies abnormal relationships with other variables; through pattern matching with the abnormal pattern library, newly discovered anomalies are matched, and thresholds are adaptively set according to the abnormal score and classification results. When the abnormal score exceeds the set threshold, an early warning response is issued and feedback is given to management personnel.

[0092] Among them, the carbon footprint tracking and tracing module is used to accurately locate the source of carbon emissions based on the identified abnormal patterns and combine the distribution information of the high-precision spatiotemporal analysis module with the patterns analyzed by the intelligent analysis module, and formulate responsibility allocation and suggestions; obtain the abnormal pattern data of the automatic abnormality identification module, combine the spatiotemporal characteristics provided by the high-precision spatiotemporal analysis module and the integrated optimized geographic coordinates into a unified data set, analyze the clustering of abnormal patterns in time and space based on the data set, identify abnormal clustering phenomena within a certain time period and within a geographical area, find the possible emission source location and its impact range through clustering results and association rules, track the propagation path of abnormal emissions from the source to the detection point, combine meteorological conditions to reversely deduce the emission source, and display the spatial distribution and temporal evolution of abnormal emissions through spatiotemporal heat maps to intuitively reflect the emission path, integrate multiple data sources, cross-validate the tracing results, and analyze the historical emission records, operating conditions and management measures of the responsible party based on the results of the tracing analysis, evaluate the degree of its impact on abnormal emissions and make responsibilities, and put forward improvement suggestions based on the causes and characteristics of abnormal emissions.

[0093] Among them, the intelligent emission prediction module is used to predict carbon emission trends based on the results of the intelligent analysis module and feed back to the adaptive adjustment module; obtain the comprehensive feature vector, clustering results and association rule information of the intelligent analysis module, make predictions based on the results of the analysis of the intelligent analysis module, and make long-term predictions based on historical trends, generate a detailed prediction report based on the prediction results and feed back to the adaptive adjustment module.

[0094] Working Principle: The data collection and preprocessing module integrates sensors and IoT devices to collect carbon emissions and environmental parameter information in real time. The collected data is cleaned, including format verification, outlier removal, missing value processing, format conversion and normalization. The cleaned data is then integrated and stored in the database.

[0095] According to the location information obtained by the sensor, the sensor position is accurately determined through the optimization algorithm, the timestamp is corrected, and a new time series is constructed. The periodic intensity of the signal is evaluated through Fourier transform and power spectral density analysis, and the spatiotemporal characteristics are extracted. According to the optimized sensor position and interpolation algorithm, a more detailed spatial distribution map is generated to show the spatiotemporal distribution characteristics of the carbon emission source. The spatiotemporal characteristics, the optimized geographical location and the data in the data acquisition preprocessing module are combined to construct a comprehensive feature vector. By calculating the comprehensive feature distance, the similarity between different points is measured. Cluster analysis is performed based on the comprehensive feature vector, and the correlation strength between variables is calculated. The hidden patterns and association rules in the data are analyzed. According to the comprehensive feature vector, clustering results and association rules of the intelligent analysis module, a multi-dimensional anomaly scoring standard is defined. According to the characteristics of different time periods and regions, the weight of each scoring dimension is dynamically adjusted. Abnormal patterns are identified in time, space and association dimensions. Pattern matching is performed with the abnormal pattern library. When an abnormal When the score exceeds the set threshold, an early warning response is issued and feedback is given to management personnel. The abnormal pattern data of the automatic anomaly identification module is obtained, and the spatiotemporal features provided by the high-precision spatiotemporal analysis module and the integrated and optimized geographic coordinates are combined into a unified data set to analyze the aggregation of abnormal patterns in time and space, identify the possible emission source locations and their impact range, track the propagation path of abnormal emissions from the source to the detection point, and display the spatial distribution and temporal evolution of abnormal emissions through spatiotemporal heat maps. According to the results of the traceability analysis, the historical emission records, operating conditions and management measures of the responsible party are analyzed, the degree of their impact on abnormal emissions is evaluated and responsibilities are allocated, and improvement suggestions are put forward. Predictions are made based on the results of the intelligent analysis module, and long-term predictions are made based on historical trends. A detailed forecast report is generated and fed back to the adaptive adjustment module. According to the real-time analysis results, automatic anomaly identification results and forecast results of the intelligent analysis module, parameters are dynamically adjusted to optimize carbon emission monitoring and management effects.

[0096] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

[0097] The present invention and its embodiments are described above. Such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by the above, and does not deviate from the purpose of the present invention, without inventive design, a structure and embodiment similar to the technical solution should fall within the scope of protection of the present invention.

Claims

1. An intelligent and automatic early warning system for identifying abnormal carbon emission data, characterized by: It includes data acquisition and preprocessing module, high-precision spatiotemporal analysis module, intelligent analysis module, automatic anomaly identification module, carbon footprint tracking and tracing module, adaptive adjustment module and intelligent emission prediction module; The data acquisition and preprocessing module is used to obtain carbon emissions and environmental parameter information data in real time based on integrated sensors and Internet of Things devices, and preprocess the obtained data information; The high-precision spatiotemporal analysis module is used to perform spatial positioning and time series analysis based on the processed data to accurately locate carbon emission sources and their spatiotemporal distribution characteristics; The intelligent analysis module is used to analyze the hidden patterns and association rules in the data based on the data acquisition preprocessing data combined with the spatial and temporal dimension information of the high-precision spatiotemporal analysis module; The automatic anomaly identification module is used to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module; The carbon footprint tracking module is used to accurately locate the source of carbon emissions based on the identified abnormal patterns, combined with the distribution information of the high-precision spatiotemporal analysis module and the patterns analyzed by the intelligent analysis module, and formulate responsibility allocation and recommendations; The intelligent emission prediction module is used to predict carbon emission trends based on the results of the intelligent analysis module and provide feedback to the adaptive adjustment module; The adaptive adjustment module is used to dynamically adjust parameters according to the real-time analysis results, automatic anomaly identification results and prediction results of the intelligent analysis module, and feed back to the automatic anomaly identification module.

2. The intelligent automatic identification of abnormal carbon emission data early warning system according to claim 1 is characterized by: The data acquisition and preprocessing module is used to obtain carbon emission and environmental parameter information data in real time based on integrated sensors and Internet of Things devices, and preprocess the obtained data information; the Internet of Things devices obtain carbon emission parameter data and environmental parameter data from the integrated sensors in real time through a set high-frequency acquisition frequency, and perform data cleaning on the obtained parameter data, including format verification, outlier extraction and missing value processing. The cleaned parameter data is format converted and normalized, and the integrated data is stored in the database.

3. The intelligent automatic identification of abnormal carbon emission data early warning system according to claim 1 is characterized by: The high-precision spatiotemporal analysis module is used to perform spatial positioning and time series analysis based on the processed data, accurately locate the carbon emission source and the spatiotemporal distribution characteristics; optimize according to the position information obtained by the sensor, assuming that the position information obtained by the sensor is P i , the total number of sensors is N, and the weight factor is defined as w i (t) = exp(-λ||P i -P ref (t)|| 2 ), λ represents the adjustment parameter, P ref (t) represents the reference position, and the objective function J(P) is the weighted sum of square errors between all sensor positions and the optimized position: Assume that the optimized sensor position P opt (t), minimize the objective function, and the implementation formula is: Interpolate according to the optimized sensor position, and define the kernel function as K h (d,t)=exp(-γd 2 / h(t) 2 ), d represents the distance, h(t) represents the bandwidth that changes with time, and γ represents the adjustment parameter. The interpolation implementation formula is: In the formula, v(x,t) represents the interpolation value, z j (t) represents the measurement value of the jth sensor at time t, M represents the number of neighboring points involved in interpolation, and P opt,j (t) represents the precise position coordinate of the jth sensor at time t after optimization, and t represents the current time point. Interpolation is performed based on the optimized sensor position to generate a more detailed spatial distribution map.

4. The intelligent automatic identification of abnormal carbon emission data early warning system according to claim 3 is characterized by: In the time series analysis, let t i is the timestamp of the i-th sensor. The timestamp is corrected and the implementation formula is: In the formula, t i ′ represents the corrected timestamp, r i (t) represents the relative clock offset rate, dt represents the time differential; Let the original measurement value be z i (t i ) is mapped to the corrected timestamp t i ′, construct a new time series z i (t i ′), perform Fourier transform on the new time series to get the frequency domain representation Z(f), and let the power spectrum density p(f) reflect the energy distribution of the signal at different frequencies, which is defined as: P(f)=|Z(f)| 2 ; The spectral entropy is calculated based on the power spectral density to evaluate the periodicity strength of the signal. The implementation formula is: S(t)=-∑ f p(f)logp(f); Assume that the mass of the emission source is m i , the feature extraction implementation formula is: In the formula, G(s,t) represents the extracted spatiotemporal features, s represents the spatial position, t represents the time point, α represents the spatial attenuation coefficient, β represents the temporal attenuation coefficient, and s i represents the spatial location of the i-th emission source, t i represents the time point of the i-th emission source.

5. The intelligent automatic identification and early warning system for abnormal carbon emission data according to claim 1 is characterized by: The intelligent analysis module is used to analyze the hidden patterns and association rules in the data based on the data collection preprocessing data combined with the spatial and temporal dimension information of the high-precision spatiotemporal analysis module; the comprehensive spatiotemporal features G(s, t) and the optimized geographical location P opt (t) and the relevant data in the data acquisition preprocessing module are the comprehensive feature vector F i (t), according to the comprehensive feature vector analysis, let the comprehensive feature distance D ij (t), measures the similarity between different points, and the implementation formula is: In the formula, D ij (t) represents the comprehensive feature distance between the i-th and j-th sensors at time t, m represents the dimension of the comprehensive feature vector, F ik (t) and F jk (t) represents the kth characteristic component of the i-th and j-th sensor at time t, respectively.

6. The intelligent automatic identification of abnormal carbon emission data early warning system according to claim 5 is characterized by: According to the comprehensive feature distance D ij (t) Calculate each cluster center and the implementation formula is: In the formula, C k represents the kth cluster center feature vector, c k represents all points in the kth cluster, D i,F (t) represents the comprehensive feature distance between the i-th sensor and the candidate center F at time t; Let the correlation strength be I XY (s, t), the formula for measuring the correlation between variables X and Y in time and space is: In the formula, K represents the number of samples, x k and y k are the k-th observation values ​​of variables X and Y, respectively. and represents the mean of variables X and Y, γ and δ represent the spatial and temporal attenuation coefficients, s and t represent the current spatial position and time, s k and t k represents the spatial position and time of the k-th observation.

7. The intelligent automatic identification of abnormal carbon emission data early warning system according to claim 1 is characterized by: The automatic anomaly identification module is used to automatically identify abnormal patterns in carbon emission data based on the analysis results of the intelligent analysis module; define multi-dimensional anomaly scoring standards based on the comprehensive feature vectors, clustering results and association rules of the intelligent analysis module, dynamically adjust the weights of each scoring dimension according to the characteristics of different time periods and regions, identify short-term and long-term abnormal trends in the time dimension, identify abnormal clustering phenomena in specific areas in the spatial dimension, and identify abnormal relationships with other variables in the association dimension. By matching newly discovered anomalies with the abnormal pattern library, adaptively set thresholds based on the anomaly score and classification results, and issue an early warning response when the anomaly score exceeds the set threshold, and provide feedback to management personnel.

8. The intelligent automatic identification and early warning system for abnormal carbon emission data according to claim 1 is characterized by: The carbon footprint tracking module is used to accurately locate the source of carbon emissions based on the identified abnormal patterns, combined with the distribution information of the high-precision spatiotemporal analysis module and the patterns analyzed by the intelligent analysis module, and formulate responsibility allocation and recommendations; Acquire abnormal pattern data from the automatic anomaly identification module, combine the spatiotemporal features provided by the high-precision spatiotemporal analysis module and the integrated optimized geographic coordinates to form a unified data set, analyze the temporal and spatial aggregation of abnormal patterns based on the data set, identify abnormal aggregation phenomena within a certain time period and geographical area, find possible emission source locations and their impact ranges through clustering results and association rules, trace the propagation path of abnormal emissions from the source to the detection point, reversely deduce the emission source based on meteorological conditions, and display the spatial distribution and temporal evolution of abnormal emissions through spatiotemporal heat maps to intuitively reflect the emission path, integrate multiple data sources, cross-validate the traceability results, analyze the historical emission records, operating conditions and management measures of the responsible party based on the results of the traceability analysis, evaluate their impact on abnormal emissions and assign responsibilities, and propose improvement suggestions based on the causes and characteristics of abnormal emissions.

9. The intelligent automatic identification and early warning system for abnormal carbon emission data according to claim 1 is characterized by: The intelligent emission prediction module is used to predict carbon emission trends based on the results of the intelligent analysis module and feed back to the adaptive adjustment module; obtain the comprehensive feature vector, clustering results and association rule information of the intelligent analysis module, make predictions based on the results of the analysis by the intelligent analysis module, and make long-term predictions based on historical trends; generate a detailed prediction report based on the prediction results and feed back to the adaptive adjustment module.

Images