Geographic Information Mapping Data Anomaly Detection Method and System Based on Data Analysis

Abstract

The present invention relates to the technical field of data processing. More specifically, the present invention relates to a method and system for detecting abnormal geographic information mapping data based on data analysis. The method includes: obtaining a deformation data sequence of a target monitoring point, calculating the edge degree of each data to screen edge points, selecting the point closest to the target edge point as the target point, screening edge points whose preference degree meets a preset condition from the remaining edge points and connecting them to the target point, and taking the connection point as a new target point, repeating the connection process based on the new target point until a termination condition is met to obtain an edge, taking the average value of the deformation data of all edge points in each edge as the clustering center corresponding to the edge, clustering all data according to the distance from each data to the clustering center, and monitoring the deformation anomaly of the target monitoring point based on the anomaly degree of the obtained clustering clusters. The present invention can achieve precise monitoring of the deformation anomaly of the monitoring point.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing. More specifically, the present invention relates to a method and system for detecting anomalies in geospatial mapping data based on data analysis. Background Art

[0002] In the field of geospatial mapping, deformation monitoring technology can, through high-precision dynamic monitoring means, obtain in real time the change information of potential geological risks such as ground settlement, building deformation, and ground fissures, greatly enhancing the state perception and safety assessment capabilities of infrastructure. Different from traditional geographic information systems (GIS) that rely on static data, deformation monitoring can provide accurate change process data through dynamic monitoring by combining sensor technology, remote sensing technology, and big data analysis, better reflecting the actual state of infrastructure. Therefore, deformation monitoring is an important part of geospatial mapping.

[0003] In the related art, for example, in the patent application document with the publication number CN113700053A, a method and system for monitoring and warning the deformation of a foundation pit during the construction process based on BIM are disclosed. The method includes: constructing a basic three-dimensional model of the target foundation pit based on BIM and the on-site measured data of the target foundation pit; arranging a plurality of measuring points on the surface of the target foundation pit in the basic three-dimensional model, establishing a spatial coordinate axis with any one of the measuring points as the origin, and obtaining the initial spatial coordinates of all the measuring points; screening out abnormal measuring points according to the initial spatial coordinates and the change of the spatial coordinates of the measuring points; predicting the movement change of the abnormal measuring points; and predicting the abnormal deformation state of the target foundation pit based on the movement change.

[0004] However, in the process of identifying abnormal measuring points in the above solution, it depends on the change amount of the spatial coordinates of the measuring points at different times, and does not fully consider the spatial distribution characteristics and similarities of the measuring points at different times. And only based on a single deformation amount for abnormal determination, it may lead to a low accuracy in identifying abnormal measuring points, thus affecting the screening effect of abnormal measuring points, and further unable to effectively reflect the true abnormal deformation state of the target monitoring object. Summary of the Invention

[0005] To solve the problem of being unable to accurately identify the abnormal deformation of the target monitoring object, the present invention provides a method and system for detecting anomalies in geospatial mapping data based on data analysis.

[0006] According to the first aspect of the present invention, there is provided a method for detecting anomalies in geospatial mapping data based on data analysis, including:

[0007] Obtaining a deformation data sequence when performing deformation monitoring on a target monitoring point;

[0008] For any data in the deformed data sequence, preset the first neighbor set of the data, calculate the edge degree of the data, where the edge degree represents the difference between the local density of the data and the local densities of the data in the first neighbor set, so as to screen edge points according to the edge degrees of the data;

[0009] Select a target edge point, take the edge point closest to the target edge point as the target point, use the direction from the target edge point to the target point as the polar axis, and use the preset direction as the positive direction of the angle to construct a polar coordinate system with the target point as the origin. Calculate the preference degree of the remaining edge points. The preference degree is negatively correlated with both the polar axis and the polar radius of the corresponding edge point in the polar coordinate system. Connect the connection point with the target point according to the preference degree and screen, and take the connection point as the new target point. Construct a polar coordinate system with the new target point as the coordinate origin and repeat the connection process until the preset termination condition is met to obtain an edge, so as to obtain all edges;

[0010] Take the mean value of the deformed data of all edge points in any edge as the clustering center of the edge, and cluster all data according to the distance from each data to the clustering center, so as to monitor the deformation anomaly of the target monitoring point based on the anomaly degree of the clustering cluster, where the anomaly degree represents the dispersion degree of the data inside the corresponding clustering cluster.

[0011] The present invention classifies similar deformed data of the target monitoring point into one category by clustering, so that the anomaly in the deformed data sequence can be determined according to the distribution of the similar data, realizing accurate monitoring of the deformation anomaly of the target monitoring point; and when determining the clustering center, multiple aspects of data are combined, so that the accuracy of the determined clustering center can be guaranteed, and further the accuracy of the clustering result can be guaranteed.

[0012] Preferably, the difference between the local density of any data and the local densities of the data in the first neighbor set is a comprehensive difference. The method for obtaining the comprehensive difference includes:

[0013] Preset the second neighbor set of any data, and take the reciprocal of the average distance between the data and the data in the second neighbor set as the local density of the data; the comprehensive difference satisfies the following relational expression:

[0014] ;

[0015] In the formula, is the comprehensive difference between the local density of the th data and the local densities of the data in the first neighbor set; is the local density of the th data; is the local density of the th data in the first neighbor set of the th data; is the number of data in the first neighbor set; is a normalization function.

[0016] By calculating the local density of data, the present invention helps to capture the distribution characteristics of data points in the local area, so that the edge points can be screened out by using the feature that the distribution characteristics of the edge points and the remaining data are quite different.

[0017] Preferably, the method for obtaining the edge degree of any data includes:

[0018] Taking the normalized value of the range of the local densities of all data in the first nearest neighbor set of any data as the credibility of the comprehensive difference, and weighting the comprehensive difference with the credibility as the weight to obtain the edge degree of this data.

[0019] Preferably, each data in the deformed data sequence is a binary group data in a two-dimensional space coordinate system; wherein, the abscissa of the two-dimensional space coordinate system is the vibration frequency of the target monitoring point at the corresponding moment, and the ordinate is the distance between the spatial coordinates of the target monitoring point at the corresponding moment and the previous moment.

[0020] The vibration frequency collected by the present invention can reflect the change of the dynamic characteristics of the target monitoring point, and the change of the spatial coordinates can reflect the actual deformation condition of the target monitoring point, so as to provide a comprehensive, accurate and dynamic analysis perspective for the abnormal deformation monitoring of the target monitoring point.

[0021] Preferably, when screening the connection points of the target point or the new target point, the positive directions of the angles of all the constructed polar coordinate systems are the same, and the method for obtaining the positive direction of the angle includes:

[0022] Judging the magnitudes of the ordinates of the target edge point and the target point. If the ordinate of the target point is greater than or equal to the ordinate of the target edge point, then the clockwise direction is taken as the positive direction of the angle when constructing the polar coordinate system with the target point or the new target point as the coordinate origin;

[0023] If it is less than the ordinate of the target edge point, then the counterclockwise direction is taken as the positive direction of the angle when constructing the polar coordinate system with the target point or the new target point as the coordinate origin.

[0024] Preferably, when screening the connection points of the current target point, the preference degree of any remaining edge point satisfies the following relational expression:

[0025] ;

[0026] In the formula, is the preference degree of the th remaining edge point; , are respectively the polar angle and the polar radius of the th remaining edge point in the current polar coordinate system; is the natural exponential function.

[0027] Preferably, clustering all data according to the distances between each data and the clustering centers includes:

[0028] Calculating the similarity degree between any data and any clustering center according to the distance therebetween: ; where is the similarity degree between any data and any clustering center; is the absolute value of the difference in local density between the data and the clustering center, where the local density of the clustering center is the average of the local densities of all data within the minimum circumscribed rectangle corresponding to the edge; is the natural exponential function; is the distance between the data and the clustering center;

[0029] Dividing each data into the clustering cluster of the clustering center with the maximum similarity degree respectively to cluster all data.

[0030] Preferably, the method for obtaining the clustering range of any clustering center when clustering all data includes:

[0031] Obtaining the maximum value of the distances between any two clustering centers, constructing a square area with a side length equal to half of the maximum value centered on each clustering center, and taking the square area as the clustering range when clustering with the corresponding clustering center.

[0032] The way of determining the clustering range in the present invention can ensure clustering all data in the deformed data sequence.

[0033] Preferably, monitoring the deformation anomaly of the target monitoring point based on the anomaly degree of the clustering cluster includes:

[0034] Calculating the anomaly degree of any clustering cluster in the clustering result, and the anomaly degree satisfies the following relational expression:

[0035] ;

[0036] where is the anomaly degree of the th clustering cluster; is the standard deviation of the distances between all data points and the clustering center in the th clustering cluster; is the standard deviation of the ordinal numbers of the sampling moments corresponding to all data points in the th clustering cluster; is the normalization function;

[0037] If the abnormality degrees of all clustering clusters are greater than a preset abnormality threshold, it is determined that the deformation amount of the target monitoring point is abnormal.

[0038] According to the second aspect of the present invention, there is provided a geographic information mapping data abnormality detection system based on data analysis. The system includes a memory and a processor. A computer program is stored on the memory, and the processor executes the computer program to implement the steps of the first aspect of the present invention.

[0039] The present invention has the following effects:

[0040] 1. When monitoring the deformation abnormality of the target monitoring point, the present invention can divide the deformation data similar to the monitoring point into one category, so that when calculating the abnormality degree of each clustering cluster by using the dispersion of the data inside each clustering cluster, the distribution characteristics of the similar deformation data can be combined, enabling accurate identification of the abnormal data in the deformation data sequence and realizing precise monitoring of the deformation abnormality of the target monitoring point.

[0041] 2. The present invention obtains the edge by using the feature that the edge of the clustering cluster is in a curve shape, thereby ensuring that the extracted edge corresponds to the true boundary of the clustering cluster, enabling accurate clustering centers to be obtained, guaranteeing the accuracy of the division of each data in the deformation data sequence, and thus accurately measuring the abnormality degree of each clustering cluster. Description of the Drawings

[0042] By referring to the following detailed description with reference to the drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understandable. In the drawings, several embodiments of the present invention are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, where:

[0043] Figure 1 is a schematic flowchart of the steps of the method for detecting the abnormality of geographic information mapping data based on data analysis according to an embodiment of the present invention;

[0044] Figure 2 is a schematic diagram of a polar coordinate system constructed with the target point as the coordinate origin in an embodiment of the present invention;

[0045] Figure 3 is another schematic diagram of a polar coordinate system constructed with the target point as the coordinate origin in an embodiment of the present invention. Detailed Embodiments

[0046] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0047] The following will describe in detail the specific implementation manners of the present invention with reference to the accompanying drawings.

[0048] Refer to Figure 1 , a method for detecting abnormal geospatial mapping data based on data analysis, includes steps S1 - S4, specifically as follows:

[0049] S1: Obtain a deformation data sequence for deformation monitoring of a target monitoring point.

[0050] It should be noted that through regular measurement of buildings, the ground, and engineering facilities, deformation monitoring can grasp deformation information in real time. During large-scale engineering construction, such as bridges and high-rise buildings, it is necessary to monitor the deformation of the structure to ensure the stability and safety of the project. Its core purpose is to detect and warn in advance of potential safety risks caused by geological or structural deformation. Deformation monitoring mainly includes two categories: horizontal monitoring and vertical monitoring. Specifically, means such as leveling measurement, satellite positioning technology, and total station surveying are used to conduct regular or continuous observations at the set monitoring points.

[0051] In an exemplary embodiment of the present invention, each data in the deformation data sequence is a binary group data in a two-dimensional space coordinate system; wherein, the abscissa of the two-dimensional space coordinate system is the vibration frequency of the target monitoring point at the corresponding moment, and the ordinate is the distance between the spatial coordinates of the target monitoring point at the corresponding moment and the previous moment.

[0052] Specifically, monitoring points (such as embedded RFID chips) can be arranged at key parts of the building to be monitored (such as load-bearing walls or columns), and then deformation monitoring of all monitoring points can be carried out using surveying techniques. The specific process is as follows: Select any monitoring point as the target monitoring point, use surveying equipment such as a total station to measure the horizontal and vertical directions of the target monitoring point, and collect the vibration frequency and spatial coordinate information of the target monitoring point at a predetermined frequency (for example, 0.2 Hz). Then, according to the data measured each time, construct the corresponding binary group data (vibration frequency, distance between the spatial coordinates at the current moment and the previous moment), thereby generating a deformation data sequence for deformation monitoring of the target monitoring point to effectively reflect the deformation trend and vibration characteristics of the target monitoring point.

[0053] In another embodiment, a deformation data sequence for deformation monitoring of the target monitoring point can also be constructed according to other deformation data of the target monitoring point, such as displacement data, tilt angle data, and temperature data, etc. The type of deformation data is not particularly limited in this embodiment.

[0054] S2: For any data in the deformation data sequence, preset the first nearest neighbor set of this data, and calculate the edge degree of this data. The edge degree represents the difference between the local density of this data and the local densities of the data in the first nearest neighbor set, so as to screen edge points according to the edge degrees of each data.

[0055] It should be noted that when performing deformation monitoring on the target monitoring point, by comparing the spatial coordinate differences at different times, the deformation degree of the target monitoring point can be reflected. When the spatial coordinate differences of the target monitoring point at different time points are small, it indicates that its deformation is small, and vice versa, it indicates a large deformation degree. Therefore, by monitoring the anomalies in the deformation data sequence composed of several binary data (vibration frequency, distance between the spatial coordinates at the current moment and the previous moment), the abnormal changes in the deformation data sequence can be effectively identified, thereby helping to judge the deformation degree of the target monitoring point. And the clustering algorithm can cluster the data into multiple clusters according to the similarity of the data, so that the anomalies in the deformation data sequence can be effectively identified based on the anomalies of the clusters, realizing the precise monitoring of the deformation anomalies of the target monitoring point.

[0056] It should be further noted that when the deformation degree of the target monitoring point is small, when performing deformation monitoring on the target monitoring point, the spatial positions of the target monitoring point obtained are relatively close. Therefore, the present invention utilizes this feature to calculate the local density of each data in the deformation data sequence and uses the difference between the local density of each data and the local density of the neighboring data to screen the data at the edge when clustering the deformation data sequence, that is, the edge points.

[0057] Optionally, the first data selected in ascending order of the distance from any data in the deformation data sequence can be obtained to get the first nearest neighbor set of this data. The size of in this embodiment is not particularly limited.

[0058] In an exemplary embodiment of the present invention, the difference between the local density of any data and the local densities of the data in the first nearest neighbor set is a comprehensive difference, and the determination of the comprehensive difference between the local density of any data and the local densities of the data in the first nearest neighbor set can be realized through the following steps:

[0059] Step 1: Preset the second nearest neighbor set of any data, and take the reciprocal of the average distance between this data and the data in the second nearest neighbor set as the local density of this data.

[0060] Optionally, the first data selected in ascending order of the distance from any data in the deformed data sequence is obtained to form the second nearest neighbor set of the data. In this embodiment, there is no special limitation on the size of .

[0061] Specifically, the local density of any data in the deformed data sequence satisfies the following relational expression:

[0062] ;

[0063] In the formula, is the local density of the th data; is the distance between the th data and the th data in the second nearest neighbor set; is the number of data in the second nearest neighbor set. In this embodiment, .

[0064] Among them, the smaller the average value of the distances between any data in the deformed data sequence and the data in the second nearest neighbor set, the greater the local density of the data.

[0065] Optionally, the Euclidean distance can be used to measure the distance between any two data, or the Manhattan distance can be used to measure the distance between any two data. In this embodiment, there is no special limitation on the selected distance type.

[0066] It should be noted that in each step of the present invention, when measuring the distance between data, the Euclidean distance is used. Therefore, when the distance between data is involved in the subsequent steps, the type of the selected distance will not be described in detail.

[0067] Step 2: Calculate the comprehensive difference of the data.

[0068] Specifically, the comprehensive difference between the local density of any data and the local densities of the data in the first nearest neighbor set satisfies the following relational expression:

[0069] ;

[0070] In the formula, is the comprehensive difference between the local density of the th data and the local densities of the data in the first nearest neighbor set; is the local density of the th data; is the local density of the th data in the first nearest neighbor set of the th data; is the number of data in the first nearest neighbor set. In this embodiment, ; is a normalization function.

[0071] Among them, reflects the difference between the local density of the -th data and the local densities of the data in the first nearest neighbor set. The larger this value is, it indicates that the difference between the local density of the -th data and the local densities of its neighboring data is relatively large, which further indicates that this data is more likely to be an edge point, and the corresponding edge degree of this data is relatively large.

[0072] Optionally, when calculating the local density of data, selecting more neighboring data can accurately measure the density of the data distribution around the corresponding data; while when calculating the edge degree of each data, selecting neighboring data that are closer as a reference can ensure the accuracy of the calculation result while reducing the calculation amount.

[0073] Furthermore, after determining the difference between the local density of any data and the local densities of the data in the first nearest neighbor set, the edge degree of this data can be calculated.

[0074] In an exemplary embodiment of the present invention, the determination of the edge degree of any data can be achieved through the following steps:

[0075] Take the normalized value of the range of the local densities of all data in the first nearest neighbor set of any data as the credibility of the comprehensive difference, and use the credibility as a weight to weight the comprehensive difference to obtain the edge degree of this data.

[0076] Specifically, the edge degree of any data satisfies the following relational expression:

[0077] ;

[0078] In the formula, is the edge degree of the -th data; is the comprehensive difference between the local density of the -th data and the local densities of the data in the first nearest neighbor set; is the maximum value of the local densities of all data in the first nearest neighbor set of the -th data; is the minimum value of the local densities of all data in the first nearest neighbor set of the -th data; is a normalization function.

[0079] Among them, reflects the range of the local densities of all data in the first nearest neighbor set of the -th data. The larger this value is, it indicates that the value range of the local densities of the data in the first nearest neighbor set of this data is relatively large. If When it is larger, it indicates that the larger one has a higher credibility, so that the difference between the data and the local density of the data in the first nearest neighbor set can be accurately measured, ensuring the accuracy of the determined edge degree.

[0080] Furthermore, after determining the edge degrees of the data in the deformed data sequence, the edge degrees of the data can be compared with the size of a preset threshold (such as 0.8), so as to determine all the edge points. The size of the preset threshold in this embodiment is not particularly limited.

[0081] S3: Select a target edge point, take the edge point closest to the target edge point as the target point, use the direction from the target edge point to the target point as the polar axis, and use the preset direction as the positive direction of the angle to construct a polar coordinate system with the target point as the origin, calculate the preference degree of the remaining edge points. The preference degree is negatively correlated with both the polar axis and the polar radius of the corresponding edge point in the polar coordinate system. Connect the connection points to the target point according to the preference degree for screening, and take the connection points as new target points. Construct a polar coordinate system with the new target point as the coordinate origin and repeat the connection process until the preset termination condition is met to obtain an edge, so as to obtain all the edges.

[0082] It should be noted that before clustering the deformed data sequence, the clustering center needs to be determined first, and the selection of the clustering center directly affects the quality and accuracy of the clustering result. Therefore, in the present invention, by first determining the edge points and then obtaining all the edges based on the positional relationship between the edge points, the obtained edges can be used as the basis for further selecting the clustering center to ensure the accuracy of the clustering center.

[0083] Among them, the target edge point refers to a randomly selected edge point; the preference degree refers to the possibility of taking any edge point as the next connection point of the current target point. For example, when the preference degree of any remaining edge point is relatively high, the possibility of taking this edge point as the next connection point of the current target point is relatively large.

[0084] In an exemplary embodiment of the present invention, when screening the connection points of the target point or the new target point, the positive directions of the angles of all the constructed polar coordinate systems are the same. The determination of the positive direction of the angle when constructing the polar coordinate system with the target point or the new target point as the coordinate origin can be achieved through the following steps:

[0085] Judge the size of the ordinate of the target edge point and the target point. If the ordinate of the target point is greater than or equal to the ordinate of the target edge point, then use the clockwise direction as the positive direction of the angle when constructing the polar coordinate system with the target point or the new target point as the coordinate origin; if it is less than the ordinate of the target edge point, then use the counterclockwise direction as the positive direction of the angle when constructing the polar coordinate system with the target point or the new target point as the coordinate origin.

[0086] Exemplarily, if the ordinate of the target point is greater than or equal to the ordinate of the target edge point, the polar coordinate system constructed with the target point as the coordinate origin is as shown in Figure 2 ; if the ordinate of the target point is less than the ordinate of the target edge point, the polar coordinate system constructed with the target point as the coordinate origin is as shown in Figure 3 .

[0087] Optionally, when constructing the polar coordinate system with the target point or the new target point as the coordinate origin, the determination methods of the positive direction of the angle and the polar axis can ensure the connected domain of the edge points, that is, the edge is a curve, so as to ensure that the obtained edge is the edge of the clustering cluster.

[0088] Furthermore, after determining the polar coordinate system constructed with the current target point as the coordinate origin, the connection points of the current target point can be screened according to the preference degree of the remaining edge points, so as to obtain a new target point and continue the connection process until the termination condition is met, and an edge is obtained.

[0089] In an exemplary embodiment of the present invention, when screening the connection points of the current target point, the preference degree of any remaining edge point satisfies the following relational expression:

[0090] ;

[0091] In the formula, is the preference degree of the th remaining edge point; , are respectively the polar angle and the polar radius of the th remaining edge point in the current polar coordinate system; is the natural exponential function, where the natural exponential function refers to the exponential function with the natural constant as the base.

[0092] Among them, is the distance from the th remaining edge point to the coordinate origin in the current polar coordinate system; when is smaller, and is relatively small, then the preference degree of taking the th remaining edge point as the next connection point of the current target point is relatively large.

[0093] In another embodiment, the preference degree of any remaining edge point can also be calculated through other calculation formulas. For example, the normalized value of the reciprocal of the product of the polar angle and the polar radius of any remaining edge point can be used as the preference degree of the remaining edge point.

[0094] Further, after determining the preference degrees of the remaining edge points, an edge point with the maximum preference degree can be selected from all the remaining edge points as the connection point of the target point, and then the connection process is repeated until the maximum value of the preference degrees of all the edge points among the current remaining edge points is less than a preference degree threshold, such as 0.8, to obtain an edge, and then all the edges can be obtained. The size of the preference degree threshold in this embodiment is not particularly limited.

[0095] Next, the process of obtaining the edges will be described in detail:

[0096] Step 1: Randomly select a target edge point and determine the target point of the target edge point;

[0097] Step 2: Use the ray starting from the target point in the direction from the target edge point to the target point as the polar axis, obtain the positive direction of the angle determined according to the ordinate sizes of the target edge point and the target point, construct a polar coordinate system with the target point as the coordinate origin, and then calculate the preference degrees of the corresponding remaining edge points according to the polar angles and polar radii of the remaining edge points in this polar coordinate system, so as to determine the connection point of the target point and make a connection;

[0098] Step 3: Take the connection point as the new target point, use the ray starting from the new target point in the direction from the target point to the new target point as the polar axis, use the positive direction of the angle obtained in Step 2, construct a polar coordinate system with the new target point as the coordinate origin, and obtain a new connection point according to the preference degrees of the remaining edge points for connection; if the maximum value of the preference degrees of all the calculated remaining edge points is less than 0.8, the connection process is terminated, and the current connected domain is taken as an edge; if it is greater than 0.8, then continue to repeat Step 3 until the maximum value of the preference degrees of the remaining edge points is less than 0.8, so as to obtain an edge, and then all the edges can be obtained.

[0099] S4: Take the average value of the deformation data of all the edge points in any edge as the clustering center of the edge, and cluster all the data according to the distances from the data to the clustering center, so as to monitor the deformation anomaly of the target monitoring point based on the anomaly degree of the clustering cluster, and the anomaly degree represents the dispersion degree of the data inside the corresponding clustering cluster.

[0100] It should be noted that since the deformation data sequence obtained in the present invention is a sequence of several binary groups (vibration frequency, distance between the spatial coordinates at the current moment and the previous moment) data, the clustering center corresponding to each edge is a binary group data composed of the average values of the values of all the edge points in the corresponding edge in each dimension of data.

[0101] In an exemplary embodiment of the present invention, the clustering of all the data can be achieved through the following steps:

[0102] Calculate the similarity degree between any data and any cluster center according to the distance between them, and divide each data into the cluster of the corresponding cluster center with the greatest similarity degree, so as to cluster all data.

[0103] Specifically, the similarity degree between any data and any cluster center satisfies the following relational expression:

[0104] ;

[0105] In the formula, is the similarity degree between any data and any cluster center; is the absolute value of the difference in local density between the data and the cluster center. Among them, the local density of the cluster center is the average value of the local densities of all data within the minimum circumscribed rectangle of the corresponding edge; is the distance between the data and the cluster center; is the natural exponential function.

[0106] Furthermore, after determining the similarity degrees between each data in the deformed data sequence and each cluster center, each data can be divided into the cluster of the corresponding cluster center with the greatest similarity degree, so as to realize the clustering of the deformed data sequence.

[0107] In another embodiment, the K-means algorithm can also be directly used to cluster all data based on the distances between each data and each cluster center, so as to realize the clustering of the deformed data sequence.

[0108] In an exemplary embodiment of the present invention, when clustering all data, the determination of the clustering range of any cluster center can be realized through the following steps:

[0109] Obtain the maximum value of the distances between any two cluster centers, construct a square area with a side length equal to half of the maximum value centered on each cluster center, and use the square area as the clustering range when clustering with the corresponding cluster center.

[0110] It should be noted that the clustering range determined by the present invention can ensure the clustering of all data, thereby further ensuring the accuracy of the clustering result.

[0111] In an exemplary embodiment of the present invention, the monitoring of the deformation anomaly of the target monitoring point can be realized through the following steps:

[0112] Calculate the anomaly degree of any cluster in the clustering result. If the anomaly degrees of all clusters are greater than the preset anomaly threshold, it is determined that the deformation amount of the target monitoring point is abnormal.

[0113] Specifically, the anomaly degree of any cluster satisfies the following relational expression:

[0114] ;

[0115] In the formula, is the degree of abnormality of the th clustering cluster; is the standard deviation of the distances from all data points in the th clustering cluster to the cluster center; is the standard deviation of the ordinal numbers of the sampling times corresponding to all data points in the th clustering cluster; is a normalization function.

[0116] Among them, when is relatively large and is relatively large, it indicates that the distribution of the data in the th clustering cluster is relatively dispersed and the sampling time interval is relatively large, then the degree of abnormality of this clustering cluster is relatively large.

[0117] It should be noted that in order to avoid the influence of the collected noise data, the preset abnormality threshold is 0.8. When the degrees of abnormality of all clustering clusters are greater than 0.8, it is determined that the deformation amount of the target monitoring point is abnormal and an alarm is issued to timely remind relevant personnel that the deformation amount at the target monitoring point exceeds the preset safety threshold, so that corresponding measures can be taken in time for processing.

[0118] The present invention also provides a geographic information mapping data anomaly detection system based on data analysis. The system includes a memory and a processor, and a computer program is stored on the memory. The computer program integrates the functions of the geographic information mapping data anomaly detection method based on data analysis. When the computer program is executed, the present invention can achieve precise monitoring of the deformation anomaly of the monitoring point through the geographic information mapping data anomaly detection method based on data analysis.

[0119] In the description of this specification, the meanings of "a plurality of" and "several" are at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0120] Although this specification has shown and described multiple embodiments of the present invention, it is obvious to those skilled in the art that such embodiments are provided only by way of example. Those skilled in the art will think of many changes, alterations and alternative ways without departing from the spirit and idea of the present invention. It should be understood that various alternative solutions to the embodiments of the present invention described herein can be adopted in the practice of the present invention.

Claims

1. A method for detecting anomalies in geographic information surveying and mapping data based on data analysis, characterized in that: include: Acquire a deformation data sequence when deformation monitoring is performed on a target monitoring point; For any data in the deformed data sequence, a first neighbor set of the data is preset, and the edge degree of the data is calculated, where the edge degree represents the difference between the local density of the data and the local density of each data in the first neighbor set, so as to filter edge points according to the edge degree of each data; Select the target edge point, take the edge point closest to the target edge point as the target point, take the direction from the target edge point to the target point as the polar axis, take the preset direction as the positive direction of the angle, construct a polar coordinate system with the target point as the coordinate origin, calculate the preference degree of the remaining edge points, the preference degree is negatively correlated with the polar axis and polar diameter of the corresponding edge point in the polar coordinate system, select the connection point and connect it with the target point according to the preference degree, take the connection point as the new target point, construct a polar coordinate system with the new target point as the coordinate origin and repeat the connection process until the preset termination condition is met to obtain an edge, so as to obtain all edges; The mean value of the deformation data of all edge points in any edge is taken as the cluster center of the edge, and all data are clustered according to the distance from each data to the cluster center, so as to monitor the deformation anomaly of the target monitoring point based on the abnormal degree of the cluster cluster. The abnormal degree represents the discrete degree of the data inside the corresponding cluster cluster. The method for obtaining the edge degree of any data includes: The difference between the local density of any data and the local density of each data in the first nearest neighbor set is taken as the comprehensive difference; The normalized value of the extreme difference of the local density of all data in the first nearest neighbor set of any data is used as the credibility of the comprehensive difference, and the credibility is used as a weight to weight the comprehensive difference to obtain the marginal degree of the data.

2. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 1, characterized in that: The method for obtaining the comprehensive difference includes: A second nearest neighbor set of any data is preset, and the reciprocal of the average distance between the data and each data in the second nearest neighbor set is taken as the local density of the data; the comprehensive difference satisfies the following relationship: ； In the formula, For the The comprehensive difference between the local density of each data and the local density of each data in the first nearest neighbor set; For the The local density of the data; For the The first neighbor set of the data The local density of the data; is the number of data in the first nearest neighbor set; is the normalization function.

3. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 1, characterized in that: Each data in the deformation data sequence is a binary data in a two-dimensional space coordinate system; wherein the horizontal coordinate of the two-dimensional space coordinate system is the vibration frequency of the target monitoring point at the corresponding moment, and the vertical coordinate is the distance between the spatial coordinates of the target monitoring point at the corresponding moment and the previous moment.

4. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 3, characterized in that: When selecting the connection points of the target point or the new target point, the positive directions of the angles of all the polar coordinate systems constructed are the same. The method for obtaining the positive directions of the angles includes: Determine the size of the ordinates of the target edge point and the target point. If the ordinate of the target point is greater than or equal to the ordinate of the target edge point, the clockwise direction is used as the positive direction of the angle when constructing a polar coordinate system with the target point or the new target point as the coordinate origin. If it is less than the ordinate of the target edge point, the counterclockwise direction is used as the positive direction of the angle when constructing a polar coordinate system with the target point or the new target point as the coordinate origin.

5. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 4, characterized in that: When screening the connection points of the current target point, the preference degree of any remaining edge point satisfies the following relationship: ； In the formula, For the The degree of preference of the remaining edge points; , Respectively The polar angle and polar diameter of the remaining edge points in the current polar coordinate system; is a natural exponential function.

6. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 1, characterized in that: The clustering of all data according to the distance between each data and the cluster center includes: According to the distance between any data and any cluster center, calculate the similarity between the data and the cluster center: ; In the formula, is the similarity between any data and any cluster center; is the absolute value of the difference between the local density of the data and the cluster center, where the local density of the cluster center is the average value of the local density of all data within the minimum circumscribed rectangle of the corresponding edge; is the natural exponential function; is the distance between the data and the cluster center; Each data is divided into the corresponding clusters with the largest similarity of the cluster center to cluster all the data.

7. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 1, characterized in that: When clustering all data, the method for obtaining the clustering range of any cluster center includes: The maximum value of the distance between any two cluster centers is obtained, a square area with a side length of half the maximum value is constructed with each cluster center as the center, and the square area is used as the clustering range when clustering is performed with the corresponding cluster center.

8. The method for detecting anomalies in geographic information surveying and mapping data based on data analysis according to claim 1, characterized in that: The abnormal degree monitoring of the deformation abnormality of the target monitoring point based on the clustering cluster includes: Calculate the abnormality degree of any cluster in the clustering result, and the abnormality degree satisfies the following relationship: ； In the formula, For the The abnormality of each cluster; For the The standard deviation of the distance between all data points in a cluster and the cluster center; For the The standard deviation of the sampling time sequence numbers corresponding to all data points in the cluster; is the normalization function; If the abnormal degree of all clusters is greater than the preset abnormal threshold, it is determined that the deformation of the target monitoring point is abnormal.

9. The geographic information surveying and mapping data anomaly detection system based on data analysis is characterized by: The geographic information surveying and mapping data anomaly detection system based on data analysis includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the geographic information surveying and mapping data anomaly detection method based on data analysis as described in any one of claims 1-8.

Images