Multi-source environmental protection online monitoring data fusion analysis method and system

Abstract

The invention discloses a multi-source environmental protection online monitoring data fusion analysis method and system. The method comprises the following steps: S1, multi-source environmental protection online monitoring data acquisition and unified modeling; s2, performing adaptive quality evaluation on the multi-source monitoring data; and S3, carrying out data preprocessing and dynamic weighting based on quality scoring. According to the method, differential data preprocessing is realized to avoid low-credibility data interference by constructing a multi-dimensional data quality adaptive evaluation and dynamic weighting mechanism, a space-time correlation model is constructed by combining time sequence correlation and spatial neighborhood influence, a pollutant space-time evolution rule is accurately described, a model online updating strategy is designed, and a real-time evaluation result is obtained. According to the method, fusion analysis can dynamically adapt to environment change and equipment state fluctuation, the accuracy, reliability and real-time performance of a fusion result are remarkably improved, accurate data support is provided for environment quality evaluation, trend analysis and abnormal early warning, and the intelligent and refined level of intelligent environmental protection construction and environment supervision is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of environmental online monitoring data technology, and in particular to a method and system for fusion analysis of multi-source environmental online monitoring data. Background Technology

[0002] With the continuous strengthening of environmental protection supervision and the ongoing advancement of smart environmental protection construction, various types of environmental monitoring equipment, such as air quality, water environment, noise, and meteorology monitoring, have been widely deployed in cities, industrial parks, and key ecological areas, forming a multi-source, heterogeneous, and distributed online environmental monitoring network. By integrating and analyzing multi-source environmental monitoring data, a comprehensive perception and dynamic assessment of the environmental status can be achieved, which is an important technical means to improve the accuracy of environmental monitoring and the level of intelligent supervision. Existing multi-source environmental online monitoring data fusion and analysis systems typically collect data from different monitoring terminals centrally and use rule matching, fixed weights, or simple statistical analysis methods to fuse the multi-source data to obtain environmental status assessment results. However, in practical applications, existing technologies still have the following shortcomings: 1. Due to differences in sensor type, measurement accuracy, sampling frequency, and communication methods among different environmental monitoring terminals, there are significant differences in the time consistency, data integrity, and measurement stability of multi-source monitoring data. Existing technologies often use uniform or static data processing methods, making it difficult to effectively distinguish the quality differences between different data sources. When monitoring data is subject to noise interference, abnormal fluctuations, or missing data, low-reliability data is easily included in the fusion calculation, thereby reducing the fusion results. 1. The accuracy and reliability of the data are questionable. 2. Most existing multi-source environmental online monitoring data fusion and analysis methods focus on the statistical fusion of data at a single moment. The fusion process lacks comprehensive consideration of the temporal evolution characteristics and spatial distribution relationship of the monitoring data, making it difficult to depict the changes of pollutants over time and their spatial diffusion patterns. Especially in scenarios with densely distributed monitoring points or complex environmental conditions, ignoring the spatial correlation between monitoring points can easily lead to the fusion analysis results failing to truly reflect the dynamic changes in the environmental state. 3. Existing fusion strategies often rely on manually preset parameters or fixed model structures, lacking the ability to adaptively adjust according to environmental changes and monitoring data characteristics. When the monitoring environment changes or the status of monitoring equipment fluctuates, the fusion analysis effect is prone to significant fluctuations, making it difficult to meet the requirements of real-time performance, stability, and intelligence for environmental online monitoring. Based on the above, this application proposes a method and system for fusion analysis of multi-source environmental protection online monitoring data. Summary of the Invention

[0003] Based on the technical problems existing in the background technology, the present invention proposes a method and system for fusion analysis of multi-source environmental protection online monitoring data.

[0004] The present invention proposes a method for fusion and analysis of multi-source environmental online monitoring data, comprising the following steps: S1: Multi-source environmental online monitoring data acquisition and unified modeling: Collect online monitoring data from different monitoring terminals such as air, water quality, noise, and meteorology, and unify the format, align the timestamps, and label the spatial coordinates of the collected data to form a multi-source raw monitoring dataset; S2: Adaptive quality assessment of multi-source monitoring data: For each monitoring data source, a data quality assessment model is constructed based on data integrity rate, stability, degree of abnormal fluctuation and historical consistency indicators, and the corresponding data quality score is output. S3: Data preprocessing and dynamic weighting based on quality scores: Based on the data quality scores, outlier correction and missing value compensation are performed on the monitoring data, and the data quality scores are mapped to fusion weight parameters to achieve dynamic weighting processing of different data sources; S4: Construction of spatiotemporal correlation features of multi-source monitoring data: Based on the spatial distribution relationship and time series characteristics of monitoring points, a spatiotemporal correlation model of multi-source monitoring data is constructed to extract the correlation features of pollutant changes over time and spatial diffusion; S5: Adaptive fusion analysis based on spatiotemporal correlation: Combining the dynamic weighting results with spatiotemporal correlation features, multi-source monitoring data are fused and analyzed to generate fused environmental state parameters; S6: Fusion Result Output and Online Update: The fusion analysis results are used for environmental status assessment, trend analysis, or early warning determination, and the quality assessment model and spatiotemporal correlation model are updated online to adapt to the dynamic changes in environmental and monitoring conditions.

[0005] Preferably, the specific logical steps of S1 are as follows: S101: Collect raw monitoring data from different environmental monitoring terminals. The raw monitoring data shall include at least the monitoring value, sampling time and spatial location information of the monitoring point. S102: Standardize the format and unify the units of monitoring data from different sources, and align the sampling time with timestamps so that data from different sampling frequencies are mapped to a unified time axis; S103: Label each monitoring data point with its corresponding spatial coordinate information, thereby unifying the multi-source monitoring data into a single model: ; in This represents the monitoring value of the i-th monitoring source at time t. This represents the spatial coordinates of the corresponding monitoring point, where i represents the monitoring data source or monitoring point number. This represents the monitoring value collected by the i-th monitoring data source at time t, where t represents the unified time identifier after timestamp alignment.

[0006] Preferably, the specific logical steps of S2 are as follows: S201: Within a preset time window, count the number of valid data points and the number of data points to be collected from each monitoring data source, and calculate the data integrity rate. ; in This represents the data integrity rate of the i-th monitoring data source. This indicates the number of valid monitoring data points within the preset time window. This indicates the total number of monitoring data to be collected within the preset time window; S202: Perform fluctuation analysis on the historical series of monitoring data, calculate its variance characteristics, and obtain stability indices accordingly. ; in This represents the data stability index of the i-th monitoring data source. This represents the standard deviation of the monitored values ​​of the i-th monitoring data source within the time window. This represents the preset maximum permissible standard deviation, used for normalization. S203: Compare the current monitored value with the historical average to calculate the degree of abnormal fluctuation. ; in This represents an index indicating the degree of abnormal fluctuation in the i-th monitored data source. This represents the monitoring value collected by the i-th monitoring data source at time t. This represents the historical average monitoring value within a preset time window. This indicates a small positive number that prevents the denominator from being zero; S204: Calculate the data quality score for the corresponding monitored data source based on comprehensive integrity rate, stability indicators, and degree of abnormal fluctuations. ; in , This represents the weighting coefficient of each quality indicator. This represents the data quality score of the i-th monitoring data source.

[0007] Preferably, the specific logical steps of S3 are as follows: S301: Scoring based on data quality It determines whether there are outliers or missing values ​​in the monitoring data, and performs smoothing correction on outliers and interpolation compensation on missing values. S302: Normalize the data quality scores to generate weight parameters for each monitoring data source in the fusion analysis: ; in This represents the weight of the i-th monitoring data source in the fusion analysis. The data source is denoted by n, where n represents the total number of monitoring data sources participating in the fusion, and j represents the index of the monitoring data source. S303: Based on the aforementioned weighting parameters, the monitoring data is dynamically weighted to obtain a weighted representation of the monitoring data: ; in This represents the weighted monitoring value of the i-th monitoring data source at time t. Indicates the fusion weight parameters. This represents the monitoring value collected by the i-th monitoring data source at time t.

[0008] Preferably, the specific logical steps of S4 are as follows: S401: Perform time series analysis on historical monitoring data from a single monitoring point, and calculate the time correlation function under different time lag conditions. ; in This indicates that the i-th monitoring data source has a time lag. The correlation below, Indicates the time lag parameter. This represents the mathematical expectation operation. This represents the historical mean of the i-th monitoring data source. Indicates the variance of the monitoring data; S402: Construct a spatial neighborhood weighting function based on the spatial distance between monitoring points to describe the intensity of spatial influence. ; in This represents the spatial influence weight between monitoring point i and monitoring point j. This represents the spatial distance between monitoring point i and monitoring point j. This represents the spatial attenuation coefficient or distance scale parameter. S403: Combining temporal correlation and spatial neighborhood weights, the neighborhood monitoring data is weighted to generate spatiotemporal correlation features of the monitoring points: ; in This represents the spatiotemporal correlation characteristics of the i-th monitoring point at time t. Represents the set of spatial neighbors adjacent to monitoring point i. This represents the monitoring value of the neighboring monitoring points at the lag time.

[0009] Preferably, the specific logical steps of S5 are as follows: S501: Dynamically weighted monitoring data Corresponding spatiotemporal features Perform joint modeling; S502: Adaptively adjust the proportion of monitoring data and spatiotemporal features in the fusion analysis based on the strength of spatiotemporal correlation; S503: The fusion environment state parameters are calculated using the fusion model, and the formula used is as follows: ; in As a spatiotemporal balance factor, This represents the fused environment state parameters at time t. This represents the fusion weight of the i-th monitoring data source. This represents the monitoring value collected by the i-th monitoring data source at time t. This represents the spatiotemporal correlation characteristics, and n represents the number of monitoring data sources fused.

[0010] Preferably, the specific logical steps of S6 are as follows: S601: Output fusion environment status parameters Used for environmental quality assessment, trend analysis, or anomaly early warning; S602: Update the parameters of the data quality assessment model online based on the deviation between the fusion results and historical monitoring data. ; in Indicates the parameters of the data quality assessment model. Indicates the learning rate. The gradient of the data quality assessment error function is represented. S603: Based on fusion error feedback, the parameters of the spatiotemporal correlation feature model are updated, using the following formula: ; in Indicates the parameters of the spatiotemporal correlation model. Indicates the update coefficients. This represents the gradient of the error function of the spatiotemporal correlation model.

[0011] This invention also proposes a multi-source environmental protection online monitoring data fusion and analysis system, including a multi-source data acquisition module, a data quality assessment module, a dynamic weight processing module, a spatiotemporal correlation modeling module, a fusion analysis module, and a result output and model update module; The multi-source data acquisition module is used to collect environmental online monitoring data from different monitoring terminals and to perform unified processing of data format, time and spatial information. The data quality assessment module is used to perform real-time quality assessment of each monitored data source and generate corresponding data quality scores. The dynamic weighting processing module is used to dynamically weight the monitoring data according to the data quality score, and to complete anomaly correction and missing data compensation. The spatiotemporal correlation modeling module is used to construct a model of the temporal series correlation relationship and spatial neighborhood influence of multi-source monitoring data; The fusion analysis module is used to perform adaptive fusion analysis on multi-source environmental monitoring data based on dynamic weights and the spatiotemporal correlation model. The result output and model update module is used to output the fusion analysis results and update the data quality assessment model and the spatiotemporal correlation model online.

[0012] Compared with existing technologies, the beneficial effects of this invention are: 1. By constructing a multi-dimensional data quality assessment model that includes data integrity, stability, abnormal fluctuation degree, and historical consistency, the model performs real-time quality scoring on each monitoring data source and maps the scores to dynamic fusion weights. At the same time, it completes outlier correction and missing value compensation based on the quality scores. Compared with the uniform or static data processing methods in the existing technology, it can effectively distinguish the differences in data source quality caused by differences in sensor accuracy, sampling frequency, and communication methods of different monitoring terminals, avoid the interference of low-reliability data on the fusion results, improve the accuracy of fusion calculation from the data source, and ensure that the fusion results can reflect the true reliability of the monitoring data. 2. By calculating the temporal correlation function of monitoring data and constructing a spatial neighborhood weighting function, the temporal series correlation features and spatial diffusion correlation features of multi-source monitoring data are comprehensively extracted to form a spatiotemporal correlation model. In the fusion analysis, a spatiotemporal balance factor is introduced to adaptively adjust the proportion of monitoring data and spatiotemporal correlation features. This model can accurately depict the changing trend of pollutants over time and their spatial neighborhood influence. Especially in scenarios with densely distributed monitoring points and complex environmental conditions, the fusion analysis results can truly reflect the spatiotemporal dynamic changes of the environmental state, achieving a comprehensive perception of the environmental state. 3. By designing an online model update mechanism, after fusion analysis, based on the deviation between the fusion results and historical monitoring data, the parameters of the data quality assessment model and the spatiotemporal correlation feature model are updated in real time using the gradient descent method. Compared with the existing fusion strategies that rely on manually preset parameters and fixed model structures, this approach can autonomously adjust model parameters according to the dynamic changes in the monitoring environment and the fluctuations in the status of monitoring equipment. This ensures that the fusion analysis strategy always adapts to the actual monitoring conditions, avoids fluctuations in the fusion effect caused by changes in the environment or equipment, meets the technical requirements of real-time performance and stability for environmental online monitoring, and improves the intelligence level of the monitoring system. This invention constructs a multi-dimensional adaptive data quality assessment and dynamic weighting mechanism to achieve differentiated data preprocessing to avoid interference from low-reliability data. It combines time series correlation and spatial neighborhood influence to construct a spatiotemporal correlation model, accurately depicting the spatiotemporal evolution of pollutants. At the same time, it designs an online model update strategy to enable the fusion analysis to dynamically adapt to environmental changes and equipment status fluctuations, significantly improving the accuracy, reliability, and real-time performance of the fusion results. This provides precise data support for environmental quality assessment, trend analysis, and anomaly early warning, effectively enhancing the intelligence and refinement of smart environmental protection construction and environmental supervision. Attached Figure Description

[0013] Figure 1 This is a flowchart of a multi-source environmental protection online monitoring data fusion and analysis method proposed in this invention; Figure 2 This is a block diagram of a multi-source environmental protection online monitoring data fusion and analysis system proposed in this invention. Detailed Implementation

[0014] The present invention will be further explained below with reference to specific embodiments.

[0015] Example Reference Figure 1 This embodiment proposes a method for fusion and analysis of multi-source environmental online monitoring data, including the following steps: S1: Multi-source environmental online monitoring data acquisition and unified modeling: Collect online monitoring data from different monitoring terminals such as air, water quality, noise, and meteorology, and unify the format, align the timestamps, and label the spatial coordinates of the collected data to form a multi-source raw monitoring dataset; The specific logical steps are as follows: S101: Collect raw monitoring data from different environmental monitoring terminals. The raw monitoring data shall include at least the monitoring value, sampling time and spatial location information of the monitoring point. S102: Standardize the format and unify the units of monitoring data from different sources, and align the sampling time with timestamps so that data from different sampling frequencies are mapped to a unified time axis; S103: Label each monitoring data point with its corresponding spatial coordinate information, thereby unifying the multi-source monitoring data into a single model: ; in This represents the monitoring value of the i-th monitoring source at time t. This represents the spatial coordinates of the corresponding monitoring point, where i represents the monitoring data source or monitoring point number. This represents the monitoring value collected by the i-th monitoring data source at time t, where t represents the unified time identifier after timestamp alignment. S2: Adaptive quality assessment of multi-source monitoring data: For each monitoring data source, a data quality assessment model is constructed based on data integrity rate, stability, degree of abnormal fluctuation and historical consistency indicators, and the corresponding data quality score is output. The specific logical steps are as follows: S201: Within a preset time window, count the number of valid data points and the number of data points to be collected from each monitoring data source, and calculate the data integrity rate. ; in This represents the data integrity rate of the i-th monitoring data source. This indicates the number of valid monitoring data points within the preset time window. This indicates the total number of monitoring data to be collected within the preset time window; S202: Perform fluctuation analysis on the historical series of monitoring data, calculate its variance characteristics, and obtain stability indices accordingly. ; in This represents the data stability index of the i-th monitoring data source. This represents the standard deviation of the monitored values ​​of the i-th monitoring data source within the time window. This represents the preset maximum permissible standard deviation, used for normalization. S203: Compare the current monitored value with the historical average to calculate the degree of abnormal fluctuation. ; in This represents an index indicating the degree of abnormal fluctuation in the i-th monitored data source. This represents the monitoring value collected by the i-th monitoring data source at time t. This represents the historical average monitoring value within a preset time window. This indicates a small positive number that prevents the denominator from being zero; S204: Calculate the data quality score for the corresponding monitored data source based on comprehensive integrity rate, stability indicators, and degree of abnormal fluctuations. ; in , This represents the weighting coefficient of each quality indicator. This represents the data quality score of the i-th monitored data source; S3: Data preprocessing and dynamic weighting based on quality scores: Based on the data quality scores, outlier correction and missing value compensation are performed on the monitoring data, and the data quality scores are mapped into fusion weight parameters to achieve dynamic weighting processing of different data sources; The specific logical steps are as follows: S301: Scoring based on data quality It determines whether there are outliers or missing values ​​in the monitoring data, and performs smoothing correction on outliers and interpolation compensation on missing values. S302: Normalize the data quality scores to generate weight parameters for each monitoring data source in the fusion analysis: ; in This represents the weight of the i-th monitoring data source in the fusion analysis. The data source is denoted by n, where n represents the total number of monitoring data sources participating in the fusion, and j represents the index of the monitoring data source. S303: Based on the aforementioned weighting parameters, the monitoring data is dynamically weighted to obtain a weighted representation of the monitoring data: ; in This represents the weighted monitoring value of the i-th monitoring data source at time t. Indicates the fusion weight parameters. This represents the monitoring value collected by the i-th monitoring data source at time t; S4: Construction of spatiotemporal correlation features of multi-source monitoring data: Based on the spatial distribution relationship and time series characteristics of monitoring points, a spatiotemporal correlation model of multi-source monitoring data is constructed to extract the correlation features of pollutant changes over time and spatial diffusion; The specific logical steps are as follows: S401: Perform time series analysis on historical monitoring data from a single monitoring point, and calculate the time correlation function under different time lag conditions. ; in This indicates that the i-th monitoring data source has a time lag. The correlation below, Indicates the time lag parameter. This represents the mathematical expectation operation. This represents the historical mean of the i-th monitoring data source. Indicates the variance of the monitoring data; S402: Construct a spatial neighborhood weighting function based on the spatial distance between monitoring points to describe the intensity of spatial influence. ; in This represents the spatial influence weight between monitoring point i and monitoring point j. This represents the spatial distance between monitoring point i and monitoring point j. This represents the spatial attenuation coefficient or distance scale parameter. S403: Combining temporal correlation and spatial neighborhood weights, the neighborhood monitoring data is weighted to generate spatiotemporal correlation features of the monitoring points: ; in This represents the spatiotemporal correlation characteristics of the i-th monitoring point at time t. Represents the set of spatial neighbors adjacent to monitoring point i. This indicates the monitoring value of neighboring monitoring points at the lag time. S5: Adaptive fusion analysis based on spatiotemporal correlation: Combining dynamic weighting results with spatiotemporal correlation characteristics, multi-source monitoring data are fused and analyzed to generate fused environmental state parameters; The specific logical steps are as follows: S501: Dynamically weighted monitoring data Corresponding spatiotemporal features Perform joint modeling; S502: Adaptively adjust the proportion of monitoring data and spatiotemporal features in the fusion analysis based on the strength of spatiotemporal correlation; S503: The fusion environment state parameters are calculated using the fusion model, and the formula used is as follows: ; in As a spatiotemporal balance factor, This represents the fused environment state parameters at time t. This represents the fusion weight of the i-th monitoring data source. This represents the monitoring value collected by the i-th monitoring data source at time t. This represents the spatiotemporal correlation characteristics, where n represents the number of monitoring data sources fused. S6: Fusion Result Output and Online Update: The fusion analysis results are used for environmental status assessment, trend analysis or early warning determination, and the quality assessment model and spatiotemporal correlation model are updated online to adapt to the dynamic changes in the environment and monitoring conditions. The specific logical steps are as follows: S601: Output fusion environment status parameters Used for environmental quality assessment, trend analysis, or anomaly early warning; S602: Update the parameters of the data quality assessment model online based on the deviation between the fusion results and historical monitoring data. ; in Indicates the parameters of the data quality assessment model. Indicates the learning rate. The gradient of the data quality assessment error function is represented. S603: Based on fusion error feedback, the parameters of the spatiotemporal correlation feature model are updated, using the following formula: ; in Indicates the parameters of the spatiotemporal correlation model. Indicates the update coefficients. This represents the gradient of the error function of the spatiotemporal correlation model.

[0016] Reference Figure 2 This embodiment also proposes a multi-source environmental protection online monitoring data fusion and analysis system, including a multi-source data acquisition module, a data quality assessment module, a dynamic weight processing module, a spatiotemporal correlation modeling module, a fusion analysis module, and a result output and model update module; The multi-source data acquisition module is used to collect environmental online monitoring data from different monitoring terminals and to perform unified processing of data format, time and spatial information; The data quality assessment module is used to perform real-time quality assessments on each monitored data source and generate corresponding data quality scores. The dynamic weighting module is used to dynamically weight the monitoring data according to the data quality score and to complete anomaly correction and missing data compensation. The spatiotemporal correlation modeling module is used to construct a model of the time-series correlation and spatial neighborhood influence of multi-source monitoring data; The fusion analysis module is used to perform adaptive fusion analysis on multi-source environmental monitoring data based on dynamic weights and spatiotemporal correlation models. The results output and model update module is used to output the fusion analysis results and update the data quality assessment model and spatiotemporal correlation model online. This embodiment constructs a multi-dimensional data quality adaptive assessment and dynamic weighting mechanism to achieve differentiated data preprocessing to avoid interference from low-reliability data. It combines time series correlation and spatial neighborhood influence to construct a spatiotemporal correlation model, accurately depicting the spatiotemporal evolution of pollutants. At the same time, it designs an online model update strategy to enable the fusion analysis to dynamically adapt to environmental changes and equipment status fluctuations, significantly improving the accuracy, reliability, and real-time performance of the fusion results. This provides precise data support for environmental quality assessment, trend analysis, and anomaly early warning, effectively enhancing the intelligence and refinement of smart environmental protection construction and environmental supervision.

[0017] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for fusion and analysis of multi-source environmental online monitoring data, characterized in that, Includes the following steps: S1: Multi-source environmental online monitoring data acquisition and unified modeling: Collect online monitoring data from different monitoring terminals such as air, water quality, noise, and meteorology, and unify the format, align the timestamps, and label the spatial coordinates of the collected data to form a multi-source raw monitoring dataset; S2: Adaptive quality assessment of multi-source monitoring data: For each monitoring data source, a data quality assessment model is constructed based on data integrity rate, stability, degree of abnormal fluctuation and historical consistency indicators, and the corresponding data quality score is output. S3: Data preprocessing and dynamic weighting based on quality scores: Based on the data quality scores, outlier correction and missing value compensation are performed on the monitoring data, and the data quality scores are mapped to fusion weight parameters to achieve dynamic weighting processing of different data sources; S4: Construction of spatiotemporal correlation features of multi-source monitoring data: Based on the spatial distribution relationship and time series characteristics of monitoring points, a spatiotemporal correlation model of multi-source monitoring data is constructed to extract the correlation features of pollutant changes over time and spatial diffusion; S5: Adaptive fusion analysis based on spatiotemporal correlation: Combining the dynamic weighting results with spatiotemporal correlation features, multi-source monitoring data are fused and analyzed to generate fused environmental state parameters; S6: Fusion Result Output and Online Update: The fusion analysis results are used for environmental status assessment, trend analysis, or early warning determination, and the quality assessment model and spatiotemporal correlation model are updated online to adapt to the dynamic changes in environmental and monitoring conditions.

2. The method for fusion and analysis of multi-source environmental online monitoring data according to claim 1, characterized in that, The specific logical steps of S1 are as follows: S101: Collect raw monitoring data from different environmental monitoring terminals. The raw monitoring data shall include at least the monitoring value, sampling time and spatial location information of the monitoring point. S102: Standardize the format and unify the units of monitoring data from different sources, and align the sampling time with timestamps so that data from different sampling frequencies are mapped to a unified time axis; S103: Label each monitoring data point with its corresponding spatial coordinate information, thereby unifying the multi-source monitoring data into a single model: ; in This represents the monitoring value of the i-th monitoring source at time t. This represents the spatial coordinates of the corresponding monitoring point, where i represents the monitoring data source or monitoring point number. This represents the monitoring value collected by the i-th monitoring data source at time t, where t represents the unified time identifier after timestamp alignment.

3. The method for fusion and analysis of multi-source environmental online monitoring data according to claim 1, characterized in that, The specific logical steps of S2 are as follows: S201: Within a preset time window, count the number of valid data points and the number of data points to be collected from each monitoring data source, and calculate the data integrity rate. ; in This represents the data integrity rate of the i-th monitoring data source. This indicates the number of valid monitoring data points within the preset time window. This indicates the total number of monitoring data to be collected within the preset time window; S202: Perform fluctuation analysis on the historical series of monitoring data, calculate its variance characteristics, and obtain stability indices accordingly. ; in This represents the data stability index of the i-th monitoring data source. This represents the standard deviation of the monitored values ​​of the i-th monitoring data source within the time window. This represents the preset maximum permissible standard deviation, used for normalization. S203: Compare the current monitored value with the historical average to calculate the degree of abnormal fluctuation. ; in This represents an index indicating the degree of abnormal fluctuation in the i-th monitored data source. This represents the monitoring value collected by the i-th monitoring data source at time t. This represents the historical average monitoring value within a preset time window. This indicates a small positive number that prevents the denominator from being zero; S204: Calculate the data quality score for the corresponding monitored data source based on comprehensive integrity rate, stability indicators, and degree of abnormal fluctuations. ; in , This represents the weighting coefficient of each quality indicator. This represents the data quality score of the i-th monitoring data source.

4. The method for fusion and analysis of multi-source environmental online monitoring data according to claim 1, characterized in that, The specific logical steps of S3 are as follows: S301: Scoring based on data quality It determines whether there are outliers or missing values ​​in the monitoring data, and performs smoothing correction on outliers and interpolation compensation on missing values. S302: Normalize the data quality scores to generate weight parameters for each monitoring data source in the fusion analysis: ; in This represents the weight of the i-th monitoring data source in the fusion analysis. The data source is denoted by n, where n represents the total number of monitoring data sources participating in the fusion, and j represents the index of the monitoring data source. S303: Based on the aforementioned weighting parameters, the monitoring data is dynamically weighted to obtain a weighted representation of the monitoring data: ; in This represents the weighted monitoring value of the i-th monitoring data source at time t. Indicates the fusion weight parameters. This represents the monitoring value collected by the i-th monitoring data source at time t.

5. The method for fusion and analysis of multi-source environmental protection online monitoring data according to claim 1, characterized in that, The specific logical steps of S4 are as follows: S401: Perform time series analysis on historical monitoring data from a single monitoring point, and calculate the time correlation function under different time lag conditions. ; in This indicates that the i-th monitoring data source has a time lag. The correlation below, Indicates the time lag parameter. This represents the mathematical expectation operation. This represents the historical mean of the i-th monitoring data source. Indicates the variance of the monitoring data; S402: Construct a spatial neighborhood weighting function based on the spatial distance between monitoring points to describe the intensity of spatial influence. ; in This represents the spatial influence weight between monitoring point i and monitoring point j. This represents the spatial distance between monitoring point i and monitoring point j. This represents the spatial attenuation coefficient or distance scale parameter. S403: Combining temporal correlation and spatial neighborhood weights, the neighborhood monitoring data is weighted to generate spatiotemporal correlation features of the monitoring points: ; in This represents the spatiotemporal correlation characteristics of the i-th monitoring point at time t. Represents the set of spatial neighbors adjacent to monitoring point i. This represents the monitoring value of the neighboring monitoring points at the lag time.

6. The method for fusion and analysis of multi-source environmental online monitoring data according to claim 1, characterized in that, The specific logical steps of S5 are as follows: S501: The dynamically weighted monitoring data... Corresponding spatiotemporal features Perform joint modeling; S502: Adaptively adjust the proportion of monitoring data and spatiotemporal features in the fusion analysis based on the strength of spatiotemporal correlation; S503: The fusion environment state parameters are calculated using the fusion model, and the formula used is as follows: ; in As a spatiotemporal balance factor, This represents the fused environment state parameters at time t. This represents the fusion weight of the i-th monitoring data source. This represents the monitoring value collected by the i-th monitoring data source at time t. This represents the spatiotemporal correlation characteristics, and n represents the number of monitoring data sources fused.

7. The method for fusion and analysis of multi-source environmental online monitoring data according to claim 1, characterized in that, The specific logical steps of S6 are as follows: S601: Output fusion environment status parameters Used for environmental quality assessment, trend analysis, or anomaly early warning; S602: Update the parameters of the data quality assessment model online based on the deviation between the fusion results and historical monitoring data. ; in Indicates the parameters of the data quality assessment model. Indicates the learning rate. The gradient of the data quality assessment error function is represented. S603: Based on fusion error feedback, the parameters of the spatiotemporal correlation feature model are updated, using the following formula: ; in Indicates the parameters of the spatiotemporal correlation model. Indicates the update coefficients. This represents the gradient of the error function of the spatiotemporal correlation model.

8. A multi-source environmental protection online monitoring data fusion and analysis system, used to implement the method described in any one of claims 1-7, characterized in that, It includes a multi-source data acquisition module, a data quality assessment module, a dynamic weight processing module, a spatiotemporal correlation modeling module, a fusion analysis module, and a result output and model update module; The multi-source data acquisition module is used to collect environmental online monitoring data from different monitoring terminals and to perform unified processing of data format, time and spatial information. The data quality assessment module is used to perform real-time quality assessment of each monitored data source and generate corresponding data quality scores. The dynamic weighting processing module is used to dynamically weight the monitoring data according to the data quality score, and to complete anomaly correction and missing data compensation. The spatiotemporal correlation modeling module is used to construct a model of the temporal series correlation relationship and spatial neighborhood influence of multi-source monitoring data; The fusion analysis module is used to perform adaptive fusion analysis on multi-source environmental monitoring data based on dynamic weights and the spatiotemporal correlation model. The result output and model update module is used to output the fusion analysis results and update the data quality assessment model and the spatiotemporal correlation model online.

Images