Multi-source data collection method and system applied to urban and rural planning

Abstract

The application relates to the technical field of electric digital data processing, in particular to a multi-source data acquisition method and system applied to urban and rural planning. The application obtains multi-dimensional abnormal data of each urban and rural planning sub-region, analyzes abnormal probability and abnormal confidence of each analysis moment corresponding to each dimension abnormal data, analyzes similarity of different dimension abnormal data of the same sub-region, obtains abnormal consistency of multi-dimensional abnormal data of the same sub-region, obtains environmental risk indexes of each analysis moment of each dimension abnormal data, further divides data characteristic categories of abnormal data, analyzes frequency adjustment degrees under each data characteristic category, and finally obtains sampling frequencies of each sensor in the next monitoring period. The application solves the problems of large multi-source data volume, complex data and non-intercommunication between data of urban and rural planning, and adaptively adjusts the sampling frequencies of each sensor, avoids data redundancy, and improves transmission data quality and data effectiveness.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, specifically to a multi-source data acquisition method and system for urban and rural planning. Background Technology

[0002] Urban and rural planning is a comprehensive deployment of urban and rural economic structure, land use layout, environmental management, and integrated transportation based on the requirements of economic and social development in a certain period. It serves as the basis for urban and rural development, construction, and management, encompassing many aspects such as overall urban and rural layout, land use, infrastructure construction, allocation of public service facilities, and ecological environmental protection. Its aim is to achieve sustainable urban and rural development and create a livable, workable, and enjoyable environment. Environmental data affects people's normal lives, and analyzing environmental data can provide a rational basis for spatial layout. Therefore, the collection of multi-source environmental data is an indispensable part of urban and rural planning.

[0003] Data used for urban and rural planning includes multi-source and multi-type data. Generally, data from various monitoring points is progressively integrated and sent to higher levels to support urban and rural planning analysis and decision-making. However, environmental monitoring data is susceptible to noise interference, affecting the quality of multi-source data collection. Furthermore, current environmental monitoring primarily employs a fixed sampling frequency method. In situations involving hazardous environmental conditions or no significant environmental changes, a fixed sampling frequency cannot guarantee the timeliness of data collection, or results in significant temporal redundancy in the environmental data, impacting transmission efficiency and the accuracy of subsequent data analysis. Therefore, there is an urgent need to optimize the multi-source data collection process for urban and rural planning. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of this invention is to provide a multi-source data acquisition method and system for urban and rural planning.

[0005] According to a first aspect of the present invention, a multi-source data acquisition method for urban and rural planning is provided, the specific technical solution of which is as follows:

[0006] The urban and rural planning area is divided into several sub-regions, and multi-dimensional anomaly data of multi-dimensional environmental data are obtained for each sub-region;

[0007] Based on multidimensional anomaly data from the same sub-region, we analyze the anomaly probability and anomaly confidence at each analysis time corresponding to the anomaly data of each dimension.

[0008] Based on the anomaly probability and the anomaly credibility, the similarity of anomaly data in different dimensions of the same sub-region is analyzed to obtain the anomaly consistency of multi-dimensional anomaly data in the same sub-region.

[0009] According to the abnormal probability and the abnormal credibility, an environmental risk index corresponding to each analysis time of each dimension abnormal data is obtained;

[0010] According to the abnormal consistency and the environmental risk index, a data feature category of the abnormal data is divided, and a frequency adjustment degree under each data feature category is analyzed;

[0011] According to the frequency adjustment degree, a current sampling frequency of each dimension environmental data and a quantity of the abnormal data are combined, and a sampling frequency of each dimension environmental data in a next monitoring period is obtained.

[0012] In some embodiments of the present application, based on multi-dimensional abnormal data of the same sub-region, the abnormal probability and the abnormal credibility corresponding to each analysis time of each dimension abnormal data are analyzed, and the method comprises the following steps:

[0013] Based on multi-dimensional abnormal data of the same sub-region, a correlation feature between an analysis time corresponding to current analysis dimension abnormal data of the same sub-region and a correlation time corresponding to other dimension environmental data is analyzed, and the abnormal probability corresponding to each analysis time of each dimension abnormal data is obtained.

[0014] Based on multi-dimensional abnormal data of the same sub-region, an abnormal frequency of each dimension abnormal data in a monitoring period is analyzed, and the abnormal credibility corresponding to each analysis time of each dimension abnormal data is obtained in combination with the abnormal probability.

[0015] In some embodiments of the present application, according to the abnormal consistency and the environmental risk index, the data feature category of the abnormal data is divided, and the method comprises the following steps:

[0016] A consistency threshold and a risk index threshold are set.

[0017] According to the abnormal consistency and the environmental risk index, the data feature category of the abnormal data is divided in combination with the abnormal consistency threshold and the risk index threshold; wherein the data feature category comprises:

[0018] A consistency and risk index high category, that is, the abnormal consistency is higher than the consistency threshold, and the environmental risk index is higher than the risk index threshold;

[0019] A consistency high but risk index low category, that is, the abnormal consistency is higher than the consistency threshold, and the environmental risk index is not higher than the risk index threshold;

[0020] A consistency low category, that is, the abnormal consistency is not higher than the consistency threshold, and the environmental risk index is not higher than the risk index threshold.

[0021] In some embodiments of the present application, analyzing the frequency adjustment degree in the category of high consistency and high risk indicators comprises:

[0022] According to the abnormal consistency and the environmental risk indicators, the sampling frequency is adjusted in the increasing direction to obtain the frequency adjustment degree in the category of high consistency and high risk indicators.

[0023] In some embodiments of the present application, analyzing the frequency adjustment degree in the category of high consistency and high risk indicators comprises:

[0024] According to the abnormal consistency and the environmental risk indicators, the sampling frequency is adjusted in the decreasing direction to obtain the frequency adjustment degree in the category of high consistency and high risk indicators.

[0025] In some embodiments of the present application, analyzing the frequency adjustment degree in the category of low consistency comprises:

[0026] According to the abnormal consistency and the abnormal consistency threshold, the multi-dimensional abnormal data is divided into data capable of reflecting environmental changes and data failing to reflect environmental changes, and the multi-dimensional abnormal data is labeled;

[0027] Analyzing the consistency of the label marks of the abnormal data in the same sub-region to obtain the monitoring data stability of each sub-region;

[0028] Analyzing the consistency of the label marks of the abnormal data in the current sub-region and its adjacent sub-regions, combining the monitoring data stability, to obtain the authenticity of the environmental performance of each dimensional abnormal data in each sub-region;

[0029] According to the authenticity, the frequency adjustment degree in the category of low consistency is obtained.

[0030] In some embodiments of the present application, analyzing the consistency of the label marks of the abnormal data in the same sub-region to obtain the monitoring data stability of each sub-region comprises:

[0031] Analyzing the intersection and union of the label marks of the abnormal data in the same sub-region to obtain the monitoring data stability of each sub-region.

[0032] In some embodiments of the present application, the urban and rural planning area is divided into a plurality of sub-regions, and the multi-dimensional abnormal data of the multi-dimensional environmental data of each sub-region is obtained, comprising:

[0033] The urban and rural planning area is divided into a plurality of sub-regions;

[0034] The multi-dimensional environmental data of each sub-region is collected;

[0035] The abnormal data in the multi-dimensional environment data is monitored through the abnormality monitoring model, and multi-dimensional abnormal data is obtained.

[0036] According to a second aspect of the embodiments of the present application, a multi-source data acquisition system applied to urban and rural planning is provided, comprising a memory and a processor, wherein:

[0037] The memory is configured to store program codes.

[0038] The processor is configured to read the program codes stored in the memory and execute the method according to the first aspect of the embodiments of the present application.

[0039] In some embodiments of the present application, the processor comprises:

[0040] The abnormal data acquisition module is configured to divide the urban and rural planning area into a plurality of sub-regions, and acquire multi-dimensional abnormal data of multi-dimensional environment data of each sub-region.

[0041] The frequency adjustment degree analysis module is configured to analyze the abnormal probability and abnormal confidence of each analysis time corresponding to each dimension abnormal data based on the multi-dimensional abnormal data of the same sub-region, analyze the similarity of different dimension abnormal data of the same sub-region based on the abnormal probability and the abnormal confidence, obtain the abnormal consistency of the multi-dimensional abnormal data of the same sub-region, and obtain the environmental risk index of each analysis time corresponding to each dimension abnormal data according to the abnormal probability and the abnormal confidence, and divide the data feature category of the abnormal data according to the abnormal consistency and the environmental risk index, and analyze the frequency adjustment degree under each data feature category.

[0042] The sampling frequency determination module is configured to obtain the sampling frequency of each dimension environment data in the next monitoring period according to the frequency adjustment degree, in combination with the current sampling frequency of each dimension environment data and the number of abnormal data.

[0043] Compared with the prior art, the multi-source data acquisition method and system applied to urban and rural planning provided by the present application has the following beneficial effects:

[0044] The present application solves the problems of large amount of multi-source data, complex data, and non-intercommunication between data in urban and rural planning, and adaptively adjusts the sampling frequency of each sensor to avoid data redundancy and improve the quality and effectiveness of the transmission data.

[0045] The present application divides the urban and rural planning area into a plurality of sub-regions, and acquires multi-dimensional abnormal data of multi-dimensional environment data of each sub-region, each sub-region corresponding to an edge server, so that each edge node uploads a small amount of data, reduces the load and calculation amount of the edge server, and realizes efficient data processing.

[0046] The application analyzes the abnormal probability and abnormal reliability of each dimension abnormal data corresponding to each analysis time, and further analyzes the similarity of different dimension abnormal data of the same sub-region, obtains the abnormal consistency of multi-dimensional abnormal data of the same sub-region, and obtains the environmental risk index of each dimension abnormal data corresponding to each analysis time, verifies the collected data through the intercommunication of the collected data of each sensor, and can effectively judge the accuracy of the sensor collection frequency.

[0047] The application divides the data characteristic categories of abnormal data, analyzes the frequency adjustment degree under each data characteristic category, determines the sampling frequency of each dimension environmental data in the next monitoring period, and adaptively adjusts the sampling frequency of each sensor, so that data redundancy is avoided, and the data transmission quality and data effectiveness are improved.

[0048] The application provides a basis for adjusting the sampling frequency of the sensor by combining the characteristics that the abnormal environment data will affect multiple sub-regions, and the monitoring data between different sub-regions is also associated, so as to accurately divide the data characteristic categories and improve the accuracy of the frequency adjustment degree. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0050] Figure 1 A basic flowchart of a multi-source data collection method applied to urban and rural planning provided by an embodiment of the present application;

[0051] Figure 2 A basic composition diagram of a multi-source data collection system applied to urban and rural planning provided by an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the multi-source data collection method and system applied to urban and rural planning according to the present application, its specific implementation, structure, features and effects are described in detail as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The use of the terms "at least" and "one or more of should be taken as open-ended, for example, in a disclosure of "comprising at least one of A, B, and C" one or more of A, B, and C alone are also included. The use of terminology "comprising" includes the instances of "consisting essentially of and "consisting of.

[0054] The application provides a multi-source data acquisition method applied to urban and rural planning.

[0055] Please refer to Figure 1 , which shows the basic flow of a multi-source data acquisition method applied to urban and rural planning provided by an embodiment of the application.

[0056] As Figure 1 shown, the multi-source data acquisition method applied to urban and rural planning provided by an embodiment of the application specifically includes:

[0057] S100: Divide the urban and rural planning area into a plurality of sub-regions, and obtain multi-dimensional abnormal data of multi-dimensional environmental data of each sub-region.

[0058] The land of the urban and rural planning area needs to be divided into a plurality of sub-regions, a plurality of monitoring sensors are arranged in each sub-region, and multi-dimensional environmental data of each sub-region is collected by different monitoring sensors at different sampling frequencies. The multi-dimensional environmental data includes air quality related data and hydrology related data, and the air quality related data includes the concentration of SO2, PM10, PM 2.5 and other air pollutants, and the hydrology related data includes heavy metal content, turbidity, etc. In the initial state, each sensor collects environmental data at a fixed frequency; a collection period is set, such as 2 hours, and the multi-dimensional environmental data obtained in the 2-hour period is taken as an analysis object, which is used to determine the sampling frequency of the multi-dimensional environmental data in the future collection period. It should be noted that because the sampling frequencies of different sensors are different, the environmental data of different dimensions obtained are not aligned in time, that is, some environmental data is not collected at the same time.

[0059] Each single sensor is regarded as an edge node, and the environmental data collected by the sensors is uploaded to the upper level and transmitted to the edge server. The environmental data of each dimension (edge node) is analyzed on the edge server, such as the status of each pollutant of the air quality data and the status of the water pollutant of the hydrology data. There is an edge server in each sub-region.

[0060] An abnormality monitoring model, such as an xgboost model, is arranged on the edge node. Through the abnormality monitoring model, abnormal data in the multi-dimensional environment data is monitored to obtain multi-dimensional abnormal data. Specifically, in the abnormality monitoring model, the real-time collected environment data of each sensor is input, and a data label is output to divide the environment data into normal data and abnormal data, to obtain multi-dimensional abnormal data corresponding to the multi-dimensional environment data, and to record the timestamp of the abnormal data. The abnormal data obtained after processing is uploaded to the edge server through the edge node, and the edge server can obtain multi-dimensional abnormal data corresponding to different sensors in each sub-region, thereby completing the acquisition of the current analysis data. It should be noted that if the environment data is normal data, it is not uploaded, and only abnormal data is uploaded to reduce the load and processing amount of the server. Each sensor corresponds to one dimension of environment data and one dimension of abnormal data. Subsequent analysis of the sensor is the analysis of the corresponding dimension abnormal data or corresponding dimension environment data unless otherwise specified.

[0061] S200: Based on the multi-dimensional abnormal data of the same sub-region, the abnormal probability and abnormal confidence of each analysis time corresponding to each dimension abnormal data are analyzed.

[0062] Based on the multi-dimensional abnormal data of the same sub-region, the abnormal probability and abnormal confidence of each analysis time corresponding to each dimension abnormal data are analyzed. Further, it includes:

[0063] First, based on the multi-dimensional abnormal data of the same sub-region, the correlation feature between the analysis time corresponding to the current analysis dimension abnormal data of the same sub-region and the correlation time corresponding to the abnormal data of other dimensions is analyzed to obtain the abnormal probability of each analysis time corresponding to each dimension abnormal data. The specific implementation is:

[0064] Based on the multi-dimensional abnormal data of the same sub-region, for the abnormal data of the current analysis dimension (i.e. the abnormal data uploaded by the corresponding sensor of the current analysis edge node), the time corresponding to the dimension abnormal data is obtained as the analysis time, denoted as ; the most recent sampling time is selected from the other dimension environment data of the same sub-region as the analysis time; it is judged whether the dimension environment data also behaves as abnormal data at the most recent sampling time; if yes, the sampling time is taken as the correlation time in the other dimension environment data as the analysis time, denoted as ; if not, the analysis time does not exist in the correlation time in the other dimension environment data, and is filled with NAN or null value processing, and the dimension is marked. The role of the abnormal data value or attribute mark is to quickly judge whether the dimension environment data is abnormal without traversing all environment data and time of the dimension. The above analysis is performed on each other dimension.

[0065] In the same sub-region, if the abnormal data time of other dimensions is close to the abnormal data time of the current analysis dimension, and the sampling frequency of the other dimension environmental data (other types of sensors) is high, it means that the abnormality of the environmental data is monitored by other sensors near the analysis time. In the same sub-region, if each analysis dimension and all other dimensions have environmental data abnormalities, it is considered that there is a real abnormal change in the environment at the analysis time, such as air pollutant emissions, sewage emissions, etc.

[0066] Therefore, first, based on the multi-dimensional abnormal data of the same sub-region, the time difference between the analysis time corresponding to the abnormal data of the current analysis dimension of the same sub-region and the associated time corresponding to the environmental data of other dimensions is calculated, and the sampling frequency of the environmental data of other dimensions is combined to obtain the abnormal performance ability of the associated time of the environmental data of other dimensions to the analysis time of the abnormal data of the analysis dimension:

[0067]

[0068] In the formula, represents the abnormal performance ability of the associated time of the i-th other dimension environmental data to the analysis time of the abnormal data of the analysis dimension; represents the timestamp of the analysis time; represents the timestamp of the associated time of the i-th other dimension environmental data; represents the sampling frequency of the sensor corresponding to the i-th other dimension environmental data; represents the rounding up. represents the number of missing sample points (the number of normal sample points) between the analysis time and its associated time, then The smaller the value is, the greater the value is, the fewer the number of missing sample points is, that is, the smaller the number of normal sample points is, and it means that there is a real environmental abnormality at the analysis time of the abnormal data of the analysis dimension.

[0069] It should be noted that if the i-th other dimension environmental data does not have an associated time, that is, the dimension exists mark, then . Calculate the mean value of all

[0070] , denoted as , that is, if , it does not participate in the calculation.

[0071]

[0072] ​​​​​Then, count the number of times all other dimensions of environmental data show anomalies at the corresponding time points (the number of times the environmental data at the corresponding time point is anomalous). The number of dimensions), denoted as And calculate the number of abnormalities. The number of dimensions of multidimensional environmental data within the same sub-region The ratio of (the number of edge nodes connected to the same edge server) to the number of edge nodes connected to the same edge server. Number of dimensions With all dimensions ratio .

[0073] Finally, according to all mean , combined Number of dimensions With all dimensions ratio The probability of environmental anomalies at the analysis time corresponding to the abnormal data in the current analysis dimension is obtained as follows:

[0074]

[0075] In the formula, This indicates the probability of environmental anomalies at the time of analysis corresponding to the abnormal data in the current analysis dimension; This indicates the number of anomalous data points at the corresponding time points for all other dimensions of environmental data (the number of anomalous environmental data points at the corresponding time point). (number of dimensions); This indicates the number of dimensions of multidimensional environmental data within the same sub-region (the number of edge nodes connected to the same edge server). Indicates all The mean.

[0076] The larger the value, the greater the real environmental anomaly in other dimensions of environmental data, indicating that there is a real environmental anomaly in the environment at the current analysis time.

[0077] The above processing is performed on the abnormal data uploaded by each edge node sensor on the edge server to obtain the environmental anomaly probability represented by the abnormal data of each sensor, that is, to obtain the anomaly probability of each analysis time corresponding to the abnormal data of each dimension.

[0078] Then, based on multidimensional anomaly data from the same sub-region, the frequency of anomalies in each dimension during the monitoring period is analyzed. Combined with the anomaly probability, the anomaly reliability at each analysis time corresponding to each dimension is obtained. The specific implementation method is as follows:

[0079] Based on the multi-dimensional abnormal data of the same sub-region, if the abnormal data of the same sensor appears more frequently in the monitoring period, and has a high environmental abnormality probability in the monitoring period, the current abnormal data belongs to the true data reflecting environmental abnormality, rather than inaccurate monitoring data caused by noise interference. Therefore, the abnormal frequency of the analysis dimension abnormal data in the monitoring period is calculated and analyzed, combined with the sampling frequency and abnormal probability corresponding to the analysis dimension abnormal data, to obtain the abnormal confidence of the analysis time corresponding to the analysis dimension abnormal data:

[0080]

[0081] In the formula, represents the abnormal confidence of the analysis time corresponding to the analysis dimension abnormal data; represents the sampling frequency of the sensor corresponding to the analysis dimension abnormal data; represents the number of neighborhoods, represents the abnormal data near the analysis time of the analysis dimension abnormal data (because normal data is not uploaded), The value of can be 5; represents the time span of the abnormal data in the neighborhood; represents the average environmental abnormality probability of the abnormal data near the analysis time of the analysis dimension.

[0082] represents the abnormal frequency of the analysis dimension abnormal data (analysis sensor) in the time neighborhood of the monitoring period. The larger the value, the higher the frequency of the abnormal data of the sensor in the monitoring period, indicating that the environmental abnormality is monitored in the monitoring period where the analysis time is located; The larger the value, the higher the environmental abnormality probability of the analysis dimension abnormal data in the neighborhood time of the analysis time, indicating that the environmental abnormality is monitored near the analysis time; The value is the confidence of the environmental abnormality. The larger the value, the more the environmental abnormality is monitored near the analysis time, and the higher the environmental abnormality confidence of the analysis time.

[0083] S300: Based on the abnormal probability and the abnormal confidence, the similarity of the different dimension abnormal data of the same sub-region is analyzed, and the abnormal consistency of the multi-dimensional abnormal data of the same sub-region is obtained.

[0084] Based on the abnormal probability and the abnormal confidence, the similarity of the different dimension abnormal data of the same sub-region is analyzed, and the abnormal consistency of the multi-dimensional abnormal data of the same sub-region is obtained. The specific implementation is:

[0085] According to the abnormal probability of the analysis time corresponding to the current analysis dimension abnormal data and abnormal credibility This yields the environmental characteristics at the analysis time corresponding to the abnormal data in the current analysis dimension. Simultaneously, it obtains the environmental characteristics of other dimensions of environmental data in the same sub-region at the corresponding time points. If there is no abnormal data at the corresponding time point for a certain dimension of environmental data, then ,thereby This forms a set of environmental features corresponding to the associated time points in different dimensions of environmental data. Therefore, by calculating the similarity between the environmental features at the associated time points corresponding to any two other-dimensional anomaly data points within the same sub-region, the anomaly consistency of multi-dimensional anomaly data within the same sub-region is obtained as follows:

[0086]

[0087] In the formula, This indicates abnormal consistency in multidimensional outlier data within the same sub-region. This indicates the number of dimensions of multidimensional environmental data within the same sub-region (the number of edge nodes connected to the same edge server). Indicates the first Environmental characteristics at the corresponding time points corresponding to each dimension of environmental data; No. Environmental characteristics at the corresponding time points corresponding to each dimension of environmental data; This represents the cosine similarity function.

[0088] The greater the similarity between the environmental features of any two other-dimensional anomalous data corresponding to the associated time in the same sub-region, the greater the similarity of the anomalous probability and anomalous credibility of the multi-dimensional anomalous data in the same sub-region. In other words, the multi-dimensional anomalous data in the same sub-region all exhibit the same characteristics.

[0089] S400: Based on the anomaly probability and anomaly confidence level, obtain the environmental risk indicators for each analysis time corresponding to the abnormal data in each dimension.

[0090] Based on the anomaly probability and anomaly confidence level, the environmental risk indicators for each analysis time point corresponding to the anomaly data in each dimension are as follows:

[0091]

[0092] In the formula, This represents the environmental risk indicator at the time of analysis corresponding to abnormal data in the analysis dimension. This indicates the probability of environmental anomalies at the time of analysis corresponding to the abnormal data in the current analysis dimension; This indicates the credibility of the anomaly at the analysis time corresponding to the abnormal data in the analysis dimension; represents a linear normalization function.

[0093] S500: According to the abnormal consistency and environmental risk indicators, the data feature categories of the abnormal data are divided, and the frequency adjustment degree under each data feature category is analyzed.

[0094] According to the abnormal consistency and environmental risk indicators, the data feature categories of the abnormal data are divided, and the frequency adjustment degree under each data feature category is analyzed. Further, including:

[0095] Firstly, set the consistency threshold and risk indicator threshold, wherein the consistency threshold can be 0.9, and the risk indicator threshold can be 0.9.

[0096] Then, according to the abnormal consistency and environmental risk indicators, the data feature categories of the abnormal data are divided in combination with the abnormal consistency threshold and the risk indicator threshold; wherein the data feature categories include:

[0097] The consistency and risk indicator high category, that is, the abnormal consistency is higher than the consistency threshold, and the environmental risk indicator is higher than the risk indicator threshold. It shows that the current analysis sensor (dimensional abnormal data) monitors the environment abnormality at the analysis moment, and the abnormality can be verified by the environmental abnormality on the monitoring data of other sensors, which shows that the sampling frequency of the sensor at this time can be used to reflect the abnormal change of the environment, so that the sampling frequency of the sensor is appropriate, and it is not necessary to make a large sampling frequency adjustment. Therefore, according to the abnormal consistency and environmental risk indicators, the sampling frequency is adjusted in the increasing direction, that is, the adjustment direction is positive (the sampling frequency needs to be appropriately increased, greater than 1), and the frequency adjustment degree of the current analysis sensor (dimensional abnormal data) corresponding to the analysis moment is obtained as:

[0098]

[0099] In the formula, represents the frequency adjustment degree of the dimensional abnormal data (sensor) corresponding to the consistency and risk indicator high category; represents the abnormal consistency of the multi-dimensional abnormal data in the same sub-region; represents the environmental risk indicator of the analysis moment corresponding to the analysis dimensional abnormal data; represents an exponential function with a natural constant as the base.

[0100] The higher the abnormal consistency, and the higher the environmental risk indicator, the smaller the corresponding frequency adjustment degree.

[0101] The consistency is high but the risk index is low category, that is, the abnormal consistency is higher than the consistency threshold, and the environmental risk index is not higher than the risk index threshold. It shows that the correlation between the abnormal data monitored by different sensors in the same sub-region is weak, and the time span between the abnormal data of the same sensor is large, or there is no uploading of abnormal data, and all sensors in the same sub-region show such characteristics, which indicates that there is a large amount of normal data monitored at the sampling moment in the sub-region at this time. At this time, the sampling frequency can be appropriately reduced, and the higher the consistency and the lower the risk index, the lower the frequency of sampling can be used. Therefore, according to the abnormal consistency and the environmental risk index, the sampling frequency is adjusted in the direction of reduction, that is, the adjustment direction is negative, and the frequency adjustment degree of the consistency and risk index high category is:

[0102]

[0103] In the formula, represents the frequency adjustment degree of the consistency high but risk index low category of the dimensional abnormal data (sensor); represents the abnormal consistency of the multi-dimensional abnormal data in the same sub-region; represents the environmental risk index of the analysis moment corresponding to the analysis dimensional abnormal data; represents an exponential function with a natural constant as the base.

[0104] The higher the abnormal consistency and the lower the risk index, the greater the corresponding frequency adjustment degree, and since the sampling frequency is adjusted in the direction of reduction, a lower frequency of sampling is used.

[0105] The consistency is low category, that is, the abnormal consistency is not higher than the consistency threshold, and the environmental risk index is not higher than the risk index threshold. It shows that there is a large difference between the abnormal data of the current analysis sensor and the abnormal data of the remaining sensors at the correlation moment, which indicates that the environmental characteristics shown by the sensor are different from the environmental characteristics shown by the remaining sensors. At this time, it can be further divided into two cases:

[0106] The first case: the monitoring data of the current analysis sensor is insufficient due to insufficient monitoring frequency, resulting in weak performance of the sensor to the environmental abnormal change, for example, the remaining sensors all monitor obvious environmental abnormalities near the time of the abnormal data, but the sensor is insufficient due to insufficient monitoring frequency, resulting in weak performance of the abnormal data. At this time, the sampling frequency of the sensor to the environment should be increased.

[0107] The second case: since not every different sensor can be reflected in the monitoring data when the environment appears abnormal, because the influence of the environment appearing abnormal on different sensors is different, for example, when the water source is polluted, the influence on the hydrological monitoring data is greater, at this time the data of air pollutants may still belong to normal monitoring data, at this time the air pollutants sensor edge node exists less abnormal data upload than the rest of the sensors, at this time the sampling frequency of the air pollutants sensor is in the normal state, even if the sampling frequency of the sensor is increased, the environment appearing abnormal cannot be monitored.

[0108] Based on the above analysis, in some embodiments of the present application, the degree of frequency adjustment in the low consistency category is analyzed. Specifically, it includes:

[0109] First, according to the abnormal consistency and the abnormal consistency threshold, the multi-dimensional abnormal data with abnormal consistency greater than the abnormal consistency threshold is divided into data that can reflect environmental changes, denoted as set , and the multi-dimensional abnormal data with abnormal consistency not greater than the abnormal consistency threshold is divided into data that cannot reflect environmental changes, denoted as set , and the multi-dimensional abnormal data is labeled, wherein the data that can reflect environmental changes is labeled as , denoted as label , and the data that cannot reflect environmental changes is labeled as , denoted as label . Obviously, the current analysis dimension abnormal data belongs to the data that cannot reflect environmental changes.

[0110] Then, the current analysis time of the current analysis dimension abnormal data is selected. Obviously, since the environmental abnormal state will have an impact on multiple urban planning sub-regions, each sub-region is not isolated, when the monitoring data of the sensor in a certain sub-region shows environmental abnormality, the sensors corresponding to the adjacent sub-regions will also reflect the abnormality of the monitoring data. Therefore, the abnormal data of the sensor is obtained, the label classification of the current analysis dimension abnormal data (data that can reflect environmental changes or data that cannot reflect environmental changes) is stored, and the current analysis time (time stamp) corresponding to the current analysis dimension abnormal data is stored as data to be uploaded (transmission label classification and time stamp).

[0111] The edge server of the sub-region obtains the data to be uploaded (the label classification and timestamp corresponding to the current analysis dimension abnormal data), and then continues to upload the processing data of the edge server to the upper-level server as a secondary edge server. After the secondary edge server obtains the timestamp, it sends a data upload command to multiple adjacent sub-regions (for example, sub-regions within 1 km) of the sub-region to require the remaining sub-regions to upload all abnormal data (limited to the same type of sensor, that is, monitoring the same type of data as the current analysis dimension abnormal data, such as monitoring PM2.5 of air) in the timestamp neighborhood time range (the neighborhood time range can be selected within 2 minutes, which can be adjusted according to the sampling frequency of the current analysis dimension abnormal data and the distance between two sub-regions). The upload form is the same as that of the edge node, and the label classification and timestamp of each current analysis dimension abnormal data (because the environmental data obtained by the sensors of the adjacent sub-region will also be processed, and these data will also exist, except that the number, size and corresponding timestamp of the abnormal data in the neighborhood time range are different) are represented by an empty set.

[0112] On the secondary edge server, the label of the current analysis dimension abnormal data of the current analysis sub-region is represented by two sets (the above 、 ). And the label of the abnormal data of the same dimension of the adjacent sub-region and the current analysis dimension abnormal data (there may be multiple groups of labels of the same type of other sensors because the abnormal data in the neighborhood time range is uploaded, and the number of abnormal data may have multiple) is recorded as , which represents the label classification of the th sensor of the th adjacent sub-region.

[0113] If the label division of the monitoring data (all dimension abnormal data) of the same sub-region remains consistent within adjacent time periods, it indicates that the current analysis dimension abnormal data of the sub-region has the stability of the monitoring result, and the authenticity of the current analysis dimension abnormal data is higher (it is not necessary to reflect whether the current analysis dimension abnormal data is monitored accurately, only the stability of the monitoring result is required).

[0114] Therefore, in some embodiments of the present application, the stability of the monitoring data of each sub-region is obtained by analyzing the consistency of the label marks of the abnormal data in the same sub-region. The specific implementation is to analyze the intersection and union of the label marks of the abnormal data in the same sub-region, and more specifically, for each adjacent sub-region of the current analysis sub-region, the intersection and union of the label marks of the monitoring data (all dimension abnormal data) in each adjacent sub-region are calculated to obtain the stability of the monitoring data of each sub-region. Wherein:

[0115] the All data tags (labels) that reflect environmental changes in the monitoring data of each neighborhood sub-region. ) for parameters The intersection can be represented as:

[0116]

[0117] In the formula, Indicates the first All data tags (labels) that reflect environmental changes in the monitoring data of each neighborhood sub-region. ) Regarding parameters The intersection of the two results only corresponds to the first one. It relates to a neighborhood sub-region and is used to represent a set of sub-regions whose labels are consistent in time series. Indicates the first The first neighborhood sub-region The label classification of each sensor.

[0118] No. All data tags (labels) that reflect environmental changes in the monitoring data of each neighborhood sub-region. ) for parameters The union of sets can be represented as:

[0119]

[0120] In the formula, Indicates the first All data tags (labels) that reflect environmental changes in the monitoring data of each neighborhood sub-region. ) Regarding parameters The union of; Indicates the first The first neighborhood sub-region The label classification of each sensor.

[0121] Data labels (tags) that can reflect environmental changes The stability of ) is:

[0122]

[0123] In the formula, Indicates the first Data labels (tags) that reflect environmental changes in monitoring data from each neighborhood sub-region The stability of ); Indicates the first All data tags (labels) that reflect environmental changes in the monitoring data of each neighborhood sub-region. ) Regarding parameters The intersection; Indicates the first All data tags (labels) that reflect environmental changes in the monitoring data of each neighborhood sub-region. ) Regarding parameters The union of; This indicates the number of elements to be retrieved from the set.

[0124] Similarly, we obtain , .

[0125] Then the first The stability of the labeling of monitoring data in each neighborhood sub-region is:

[0126]

[0127] In the formula, Indicates the first The stability of labeling of monitoring data in each neighborhood sub-region; Indicates the first Data labels (tags) that reflect environmental changes in monitoring data from each neighborhood sub-region The stability of ); Indicates the first Data labels (tags) that fail to reflect environmental changes in monitoring data from individual neighborhood sub-regions. The stability of ).

[0128] It can reflect the first All dimensions of outlier data in each neighborhood sub-region are considered authentic. The larger the value, the more significant the effect. The greater the likelihood of anomalies across all dimensions within a neighborhood sub-region, the more likely they are to be genuine. It should be noted that... The authenticity of the representation depends only on the sub-region.

[0129] Obtain the stability of all neighboring sub-regions of the current analysis sub-region, normalize them, and make the sum of the stability of these neighboring sub-regions equal to 1. Each normalized stability is denoted as . .

[0130] Then, by analyzing the consistency of the labels of outlier data in the current sub-region and its neighboring sub-regions, and combining this with the stability of the labels in the neighboring sub-regions, the authenticity of the environmental representation of outlier data in each dimension within each sub-region is obtained. The formula for calculating the authenticity of the environmental representation of outlier data in the current analysis dimension within the current analysis sub-region is as follows:

[0131]

[0132] In the formula, the realness of the current analysis dimension abnormal data in the current analysis sub-region to the environment performance; the normalized stability of the label marking of the current analysis dimension abnormal data of the first neighborhood sub-region; the intersection of all data label markings (label ) in the monitoring data of the first neighborhood sub-region that can reflect the environmental change for the parameter ; the number of elements in the set composed of data that can reflect the environmental change; the intersection of all data label markings (label ) in the monitoring data of the first neighborhood sub-region that cannot reflect the environmental change for the parameter ; the number of elements in the set composed of data that cannot reflect the environmental change; the number of elements in the set ;

[0133] the consistency of the sensors that show environmental abnormalities in the first neighborhood sub-region and the sensors that show environmental abnormalities in the current analysis sub-region; the consistency of the label category division of the current analysis sub-region and its neighborhood sub-regions, reflecting whether the label classification of each sensor in the current analysis sub-region is consistent, the greater the value, the higher the consistency of the environmental monitoring abnormalities on multiple urban planning sub-regions, indicating that the monitoring abnormalities represent real environmental abnormalities, i.e., the realness of the current analysis dimension abnormal data in the current analysis sub-region to the environment performance; the stability of the label marking of the current analysis dimension abnormal data of the first neighborhood sub-region as a weight, the greater the value, the greater the stability of the label marking of the current analysis dimension abnormal data of the first neighborhood sub-region, indicating that the realness of the current analysis dimension abnormal data of the first

[0134] neighborhood sub-region to the environment performance is greater, and therefore a greater weight is given to it. the realness of the current analysis dimension abnormal data in the current analysis sub-region to the environment performance the greater the value, the more normal it is for the current analysis dimension abnormal data of the current analysis sub-region belonging to the set The smaller the value, the more the current analysis dimension abnormal data of the current analysis sub-region reflects the environmental abnormal change, which is due to the insufficient sampling frequency of the sensor, belonging to the "first case", therefore, The smaller the value, the greater the amplitude of the sampling frequency adjustment should be. Therefore, according to the authenticity, the frequency adjustment of the dimension abnormal data (sensor) belonging to the low consistency category is obtained, and the frequency adjustment degree of the low consistency category is:

[0135]

[0136] In the formula, The frequency adjustment degree of the dimension abnormal data (sensor) corresponding to the low consistency category is represented; The authenticity of the current analysis dimension abnormal data in the current analysis sub-region to the environment is represented; The natural constant is taken as the base of the exponential function.

[0137] S600: According to the frequency adjustment degree, the current sampling frequency of each dimension environmental data and the number of abnormal data are combined to obtain the sampling frequency of each dimension environmental data in the next monitoring period.

[0138] The frequency adjustment degree in the three data feature categories is determined through step S500, that is, the frequency adjustment degree of each abnormal data of each sensor at the analysis time point is determined. Then, according to the frequency adjustment degree, the current sampling frequency of each dimension environmental data and the number of abnormal data are combined to obtain the sampling frequency of each dimension environmental data in the next monitoring period. The specific implementation is: according to the frequency adjustment degree of each abnormal data of the sensor at the analysis time point, a distribution curve of the frequency adjustment degree (the horizontal axis is the frequency adjustment degree, and the vertical axis is the ratio of the number of analysis time points corresponding to the same frequency adjustment degree to the total number of analysis time points) is obtained; at the same time, the number of normal data (theoretical sampling number in the sampling time-the number of uploaded abnormal data) is obtained; the sampling frequency of the current analysis dimension abnormal data (sensor) of the current analysis sub-region in the next monitoring period is obtained as:

[0139]

[0140] In the formula, The sampling frequency of the current analysis dimension abnormal data (sensor) of the current analysis sub-region in the next monitoring period is represented; The number of normal data of the current analysis dimension environmental data of the current analysis sub-region is represented; The number of abnormal data of the current analysis dimension environmental data of the current analysis sub-region is represented; A current sampling frequency of the current analysis dimension abnormal data (sensor) of the current analysis sub-region; A distribution representing the frequency adjustment degree, and a proportion of the number of abnormal data under the same frequency adjustment degree; A frequency adjustment degree of the dimension abnormal data (sensor), comprising 、 and .

[0141] A sensor sampling frequency determined by normal data quantity weighting, A sensor sampling frequency determined by abnormal data quantity weighting, and a sampling frequency of the current analysis dimension abnormal data (sensor) of the current analysis sub-region in the next monitoring period is obtained by comprehensively considering the two.

[0142] The edge server obtains the sampling frequency of each sensor corresponding to the next monitoring period, and transmits the sampling frequency to each corresponding edge node (sensor), and the edge node samples the environment according to the sampling frequency transmitted by the edge server in the next monitoring period.

[0143] Based on the same inventive concept as the above method, the embodiment also provides a multi-source data acquisition system applied to urban and rural planning.

[0144] Please refer to Figure 2 , which shows a basic composition of a multi-source data acquisition system applied to urban and rural planning provided by an embodiment of the application.

[0145] As shown in Figure 2 , a multi-source data acquisition system applied to urban and rural planning comprises a memory 10 and a processor 20, wherein:

[0146] The memory 10 is used for storing program codes;

[0147] The processor 20 is configured to read the program code stored in the memory 10, and perform the following steps: dividing a rural and urban planning area into a plurality of sub-regions, obtaining multi-dimensional abnormal data of multi-dimensional environmental data of each sub-region; based on the multi-dimensional abnormal data of the same sub-region, analyzing abnormal probability and abnormal confidence of each analysis time corresponding to each dimensional abnormal data; based on the abnormal probability and the abnormal confidence, analyzing similarity of different dimensional abnormal data of the same sub-region, obtaining abnormal consistency of the multi-dimensional abnormal data of the same sub-region; obtaining environmental risk indexes of each analysis time corresponding to each dimensional abnormal data according to the abnormal probability and the abnormal confidence; dividing data feature categories of the abnormal data according to the abnormal consistency and the environmental risk indexes, and analyzing frequency adjustment degrees under each data feature category; and obtaining sampling frequencies of each dimensional environmental data in a next monitoring period according to the frequency adjustment degrees, and combining a current sampling frequency of each dimensional environmental data and a number of abnormal data.

[0148] Further, the processor 20 includes an abnormal data acquisition module 21, a frequency adjustment degree analysis module 22, and a sampling frequency determination module 23. Wherein:

[0149] The abnormal data acquisition module 21 is configured to divide the rural and urban planning area into a plurality of sub-regions, and obtain multi-dimensional abnormal data of multi-dimensional environmental data of each sub-region.

[0150] The frequency adjustment degree analysis module 22 is configured to analyze abnormal probability and abnormal confidence of each analysis time corresponding to each dimensional abnormal data based on the multi-dimensional abnormal data of the same sub-region; and based on the abnormal probability and the abnormal confidence, analyze similarity of different dimensional abnormal data of the same sub-region, obtain abnormal consistency of the multi-dimensional abnormal data of the same sub-region, and obtain environmental risk indexes of each analysis time corresponding to each dimensional abnormal data; and divide data feature categories of the abnormal data according to the abnormal consistency and the environmental risk indexes, and analyze frequency adjustment degrees under each data feature category.

[0151] The sampling frequency determination module 23 is configured to obtain sampling frequencies of each dimensional environmental data in a next monitoring period according to the frequency adjustment degrees, and combine a current sampling frequency of each dimensional environmental data and a number of abnormal data.

[0152] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0153] Various embodiments are described herein with reference to the drawings. The various embodiments are described in the following progressive manner, wherein the same or similar components are referenced the same or similarly, and each embodiment highlights the differences from other embodiments.

Claims

1. A multi-source data acquisition method for urban and rural planning, characterized in that, The method includes: The urban and rural planning area is divided into several sub-regions, and multi-dimensional anomaly data of multi-dimensional environmental data are obtained for each sub-region; Based on multidimensional anomaly data from the same sub-region, we analyze the anomaly probability and anomaly confidence at each analysis time corresponding to the anomaly data of each dimension. Based on the anomaly probability and the anomaly credibility, the similarity of anomaly data in different dimensions of the same sub-region is analyzed to obtain the anomaly consistency of multi-dimensional anomaly data in the same sub-region. Based on the anomaly probability and the anomaly confidence level, the environmental risk indicators for each analysis time corresponding to the anomaly data of each dimension are obtained; Based on the aforementioned anomaly consistency and the aforementioned environmental risk indicators, the data feature categories of the abnormal data are divided, and the degree of frequency adjustment under each data feature category is analyzed. Based on the frequency adjustment level, combined with the current sampling frequency of environmental data in each dimension and the number of abnormal data, the sampling frequency of environmental data in each dimension in the next monitoring period is obtained; Based on multidimensional anomaly data from the same sub-region, the probability and reliability of anomalies at each analysis time corresponding to the anomaly data in each dimension are analyzed, including: Based on multidimensional anomaly data from the same sub-region, the correlation characteristics between the analysis time corresponding to the current analysis dimension anomaly data and the correlation time corresponding to other dimension environmental data in the same sub-region are analyzed to obtain the anomaly probability of each analysis time corresponding to the anomaly data of each dimension, specifically including: Based on multidimensional anomaly data from the same sub-region, for the anomaly data in the current analysis dimension, the time corresponding to the anomaly data in that dimension is obtained as the analysis time, denoted as t; from the environmental data of other dimensions in the same sub-region, the sampling time closest to the analysis time is selected; it is determined whether the environmental data of that dimension also exhibits anomaly data at that closest sampling time; if so, this sampling time is taken as the associated time of the analysis time in the environmental data of those other dimensions, denoted as t. ; Based on multidimensional anomaly data from the same sub-region, the time difference between the analysis time corresponding to the current analysis dimension anomaly data and the correlation time corresponding to other dimension environmental data in the same sub-region is calculated. Combined with the sampling frequency of other dimension environmental data, the anomaly performance of the correlation time of other dimension environmental data on the analysis time of the analysis dimension anomaly data is obtained as follows: In the formula, Indicates the first The ability of the correlation moments of other dimensions of environmental data to reflect the abnormal performance of the analysis moments of abnormal data in the analysis dimension; The timestamp indicates the moment of analysis; Indicates the first The timestamps of other dimensions of environmental data that are associated with the time of analysis; Indicates the first The sampling frequency of the sensor corresponds to other dimensions of environmental data; Indicates rounding up; If the first If there is no related time for other dimensions of environmental data, then Calculate all The mean, denoted as ; The number of times that the correlation between all other dimensions of environmental data shows anomalies is denoted as . And calculate the number of abnormalities. The number of dimensions of multidimensional environmental data within the same sub-region ratio Therefore, the probability of environmental anomalies at the analysis time corresponding to the abnormal data in the current analysis dimension is: In the formula, This indicates the probability of environmental anomalies at the time of analysis corresponding to the abnormal data in the current analysis dimension; Based on multidimensional anomaly data from the same sub-region, the frequency of anomalies in each dimension during the monitoring period is analyzed. Combined with the anomaly probability, the anomaly reliability at each analysis time corresponding to each dimension is obtained, specifically including: The frequency of anomalies in the analytical dimension's abnormal data within the monitoring period is calculated. Combined with the sampling frequency and anomaly probability corresponding to the abnormal data in the analytical dimension, the anomaly confidence level at the analysis time corresponding to the abnormal data in the analytical dimension is obtained as follows: In the formula, This indicates the credibility of the anomaly at the analysis time corresponding to the abnormal data in the analysis dimension; This indicates the sampling frequency of the sensor corresponding to the abnormal data in the analysis dimension; This represents the number of neighborhoods, indicating the analysis time near the point of analysis for the abnormal data in the analysis dimension. One abnormal data point; Indicates the neighborhood The time span of the anomalous data; This indicates the time frame near the analysis dimension. The average probability of environmental anomalies at the time of analysis corresponding to each abnormal data point.

2. The multi-source data acquisition method for urban and rural planning according to claim 1, characterized in that, Based on the aforementioned anomaly consistency and the aforementioned environmental risk indicators, the data characteristic categories of the abnormal data are divided, including: Set consistency thresholds and risk indicator thresholds; Based on the aforementioned anomaly consistency and the aforementioned environmental risk indicators, and combining the anomaly consistency threshold and the risk indicator threshold, the data feature categories of the abnormal data are classified; wherein, the data feature categories include: Both consistency and risk indicators are in the high category, meaning that the abnormal consistency is higher than the consistency threshold, and the environmental risk indicator is higher than the risk indicator threshold. The category with high consistency but low risk index is defined as follows: the abnormal consistency is higher than the consistency threshold, while the environmental risk index is not higher than the risk index threshold. The low consistency category means that the abnormal consistency is not higher than the consistency threshold, and the environmental risk indicator is not higher than the risk indicator threshold.

3. The multi-source data acquisition method for urban and rural planning according to claim 2, characterized in that, The analysis includes the degree of frequency adjustment under categories with both high consistency and risk indicators, including: Based on the abnormal consistency and the environmental risk index, the sampling frequency is adjusted in the direction of increase to obtain the frequency adjustment degree under the category where both the consistency and risk index are high.

4. The multi-source data acquisition method for urban and rural planning according to claim 2, characterized in that, The analysis includes the degree of frequency adjustment under categories with both high consistency and risk indicators, including: Based on the anomaly consistency and the environmental risk index, the sampling frequency is adjusted in the direction of reduction to obtain the frequency adjustment degree under the category where both the consistency and risk index are high.

5. The multi-source data acquisition method for urban and rural planning according to claim 2, characterized in that, The analysis of frequency modulation under the low consistency category includes: Based on the anomaly consistency and the anomaly consistency threshold, the multidimensional anomaly data is divided into data that can reflect environmental changes and data that cannot reflect environmental changes, and the multidimensional anomaly data is labeled. By analyzing the consistency of the labeling of the abnormal data within the same sub-region, the stability of the monitoring data in each sub-region can be obtained. By analyzing the consistency of the labels of the abnormal data in the current sub-region and its neighboring sub-regions, and combining this with the stability of the monitoring data, the authenticity of the abnormal data in each dimension of each sub-region in terms of environmental performance is obtained. Based on the stated authenticity, the degree of frequency adjustment under the low consistency category is obtained.

6. The multi-source data acquisition method for urban and rural planning according to claim 5, characterized in that, Analyzing the consistency of labeling of the abnormal data within the same sub-region yields the stability of the monitoring data for each sub-region, including: The stability of the monitoring data for each sub-region is obtained by analyzing the intersection and union of the labels of the abnormal data within the same sub-region.

7. The multi-source data acquisition method for urban and rural planning according to claim 1, characterized in that, The urban and rural planning area is divided into several sub-regions, and multidimensional anomaly data of the multidimensional environmental data of each sub-region is obtained, including: The urban and rural planning area is divided into several sub-areas; Collect multidimensional environmental data for each sub-region; Anomaly monitoring model is used to monitor anomalous data in the multidimensional environmental data to obtain multidimensional anomalous data.

8. A multi-source data acquisition system for urban and rural planning, characterized in that, The system includes: a memory and a processor, wherein: The memory is used to store program code; The processor is configured to read program code stored in the memory and execute the method as described in any one of claims 1 to 7.

9. The multi-source data acquisition system for urban and rural planning according to claim 8, characterized in that, The processor includes: The abnormal data acquisition module is used to divide the urban and rural planning area into several sub-regions and obtain multi-dimensional abnormal data of the multi-dimensional environmental data of each sub-region; The frequency regulation degree analysis module is used to analyze the probability and reliability of anomalies in each dimension of multidimensional anomaly data at each analysis time based on multidimensional anomaly data in the same sub-region; and based on the probability and reliability of anomalies, to analyze the similarity of anomaly data in different dimensions in the same sub-region, to obtain the anomaly consistency of multidimensional anomaly data in the same sub-region; and based on the probability and reliability of anomalies, to obtain the environmental risk index for each dimension of anomaly data at each analysis time; and based on the anomaly consistency and the environmental risk index, to classify the data feature categories of the anomaly data, and to analyze the frequency regulation degree under each data feature category. The sampling frequency determination module is used to determine the sampling frequency of each dimension of environmental data in the next monitoring period based on the frequency adjustment degree, combined with the current sampling frequency of each dimension of environmental data and the number of abnormal data. Based on multidimensional anomaly data from the same sub-region, the probability and reliability of anomalies at each analysis time corresponding to the anomaly data in each dimension are analyzed, including: Based on multidimensional anomaly data from the same sub-region, the correlation characteristics between the analysis time corresponding to the current analysis dimension anomaly data and the correlation time corresponding to other dimension environmental data in the same sub-region are analyzed to obtain the anomaly probability of each analysis time corresponding to the anomaly data of each dimension, specifically including: Based on multidimensional anomaly data from the same sub-region, for the anomaly data in the current analysis dimension, the time corresponding to the anomaly data in that dimension is obtained as the analysis time, denoted as t; from the environmental data of other dimensions in the same sub-region, the sampling time closest to the analysis time is selected; it is determined whether the environmental data of that dimension also exhibits anomaly data at that closest sampling time; if so, this sampling time is taken as the associated time of the analysis time in the environmental data of those other dimensions, denoted as t. ; Based on multidimensional anomaly data from the same sub-region, the time difference between the analysis time corresponding to the current analysis dimension anomaly data and the correlation time corresponding to other dimension environmental data in the same sub-region is calculated. Combined with the sampling frequency of other dimension environmental data, the anomaly performance of the correlation time of other dimension environmental data on the analysis time of the analysis dimension anomaly data is obtained as follows: In the formula, Indicates the first The ability of the correlation moments of other dimensions of environmental data to reflect the abnormal performance of the analysis moments of abnormal data in the analysis dimension; The timestamp indicates the moment of analysis; Indicates the first The timestamps of other dimensions of environmental data that are associated with the time of analysis; Indicates the first The sampling frequency of the sensor corresponds to other dimensions of environmental data; Indicates rounding up; If the first If there is no related time for other dimensions of environmental data, then Calculate all The mean, denoted as ; The number of times that the correlation between all other dimensions of environmental data shows anomalies is denoted as . And calculate the number of abnormalities. The number of dimensions of multidimensional environmental data within the same sub-region ratio Therefore, the probability of environmental anomalies at the analysis time corresponding to the abnormal data in the current analysis dimension is: In the formula, This indicates the probability of environmental anomalies at the time of analysis corresponding to the abnormal data in the current analysis dimension; Based on multidimensional anomaly data from the same sub-region, the frequency of anomalies in each dimension during the monitoring period is analyzed. Combined with the anomaly probability, the anomaly reliability at each analysis time corresponding to each dimension is obtained, specifically including: The frequency of anomalies in the analytical dimension's abnormal data within the monitoring period is calculated. Combined with the sampling frequency and anomaly probability corresponding to the abnormal data in the analytical dimension, the anomaly confidence level at the analysis time corresponding to the abnormal data in the analytical dimension is obtained as follows: In the formula, This indicates the credibility of the anomaly at the analysis time corresponding to the abnormal data in the analysis dimension; This indicates the sampling frequency of the sensor corresponding to the abnormal data in the analysis dimension; This represents the number of neighborhoods, indicating the analysis time near the point of analysis for the abnormal data in the analysis dimension. One abnormal data point; Indicates the neighborhood The time span of the anomalous data; This indicates the time frame near the analysis dimension. The average probability of environmental anomalies at the time of analysis corresponding to each abnormal data point.

Images