A data intelligent analysis system for ground fissure monitoring

Abstract

The present invention relates to the technical field of electronic digital data processing, and in particular to a data intelligent analysis system for ground fissure monitoring. The system obtains monitoring data and risk index values ​​of ground fissures; obtains change characteristic values ​​according to changes in monitoring data; obtains the overall impact value of monitoring data according to the difference between the change characteristic values ​​and the risk index values; screens out main monitoring data according to the fluctuations of the change characteristic values ​​and the risk index values, as well as the overall impact value; divides the main monitoring data into data categories according to the similarities between the main monitoring data, and screens out target data categories according to the fluctuation differences between the main monitoring data in each data category and each type of monitoring data between the last ground fissure end time and the current time, and then obtains the predicted risk index value to predict ground fissure disasters. The present invention accurately determines the target data category, and then accurately and efficiently predicts ground fissure disasters in a specified area.

Description

Technical Field

[0001] The invention relates to the technical field of electrical digital data processing, and in particular to a data intelligent analysis system for ground fissure monitoring. Background Art

[0002] Ground fissure disasters can cause great harm to humans and society. Ground fissures not only threaten the safety of buildings, but may also lead to secondary disasters such as ground subsidence and water resource pollution. Therefore, it is necessary to monitor various ground fissure data in real time, predict ground fissure disasters in a timely manner, make preparations in advance, and reduce the losses caused by ground fissure disasters.

[0003] In the existing methods, various monitoring data of ground fissures collected in real time in various regions are analyzed by intelligent analysis technology, and the ground fissure disaster situation in various regions is automatically predicted, which can timely discover ground fissure disasters and issue early warnings. However, in actual situations, in the process of monitoring various ground fissure data, due to the complexity of the geological environment, the diversity of monitoring data types, and the instability factors caused by the long-term operation of the monitoring system, it is easy to cause inaccurate monitoring data of various ground fissures, which in turn leads to the inability of intelligent analysis technology to timely and accurately predict ground fissure disasters, which may cause great harm to humans and society. Summary of the invention

[0004] In order to solve the technical problem that various monitoring data collected for ground fissures may be inaccurate, resulting in the inability of intelligent analysis technology to timely and accurately predict ground fissure disasters, the purpose of the present invention is to provide a data intelligent analysis system for ground fissure monitoring, and the technical solutions adopted are as follows:

[0005] The present invention proposes a data intelligent analysis system for ground fissure monitoring, which includes the following contents:

[0006] The data acquisition module is used to obtain each monitoring data at each time within the time period of each ground fissure occurrence in the specified area and each monitoring data at each time between the end time of the last ground fissure and the current time; and obtain the risk index value of each ground fissure in the specified area;

[0007] The overall impact value acquisition module is used to obtain the change characteristic value of each monitoring data under each ground fissure according to the change of each monitoring data in the time period of each ground fissure; according to the difference between the change characteristic value of each monitoring data under each ground fissure and the risk index value, and the distribution difference between the change characteristic value of each monitoring data and the risk index value, to obtain the overall impact value of the monitoring data;

[0008] The main monitoring data acquisition module is used to obtain the actual reference value of each monitoring data according to the fluctuation of the change characteristic value of each monitoring data and the fluctuation of the risk indicator value, as well as the overall impact value, and then screen out the main monitoring data;

[0009] The predicted risk index value acquisition module is used to obtain the similarity values ​​of different types of main monitoring data according to the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk index values, and then divide the main monitoring data into data categories; according to the fluctuation difference between each type of main monitoring data in each data category and each type of monitoring data between the end time of the last ground fissure and the current time, the target data category is screened out; based on the similarity values ​​corresponding to all main monitoring data in the target data category, the predicted risk index value is obtained;

[0010] The data processing module is used to predict ground fissure disasters in a designated area based on the predicted risk index value.

[0011] Furthermore, the method for obtaining the change characteristic value is:

[0012] For any kind of monitoring data within any time period of ground fissure occurrence, the change characteristic value of the monitoring data under the ground fissure is obtained based on the difference between the monitoring data at each moment within the time period of ground fissure occurrence and the mean value of the monitoring data within the time period of ground fissure occurrence, and the difference between the maximum preset critical value of the monitoring data and the maximum value of the monitoring data within the time period of ground fissure occurrence.

[0013] Furthermore, the calculation formula of the change characteristic value is:

[0014]

[0015] Where, T a,i is the characteristic value of the change of the a-th monitoring data under the i-th ground fissure; M i is the total number of times during the time period of the i-th ground fissure; U a,i,m is the ath type of monitoring data at the mth moment during the time period of the i-th ground fissure occurrence; is the mean value of the ath type of monitoring data during the time period of the i-th ground fissure occurrence; U′ a,max is the maximum preset critical value of the a-th type of monitoring data; U a,i,max is the largest a-th type of monitoring data during the time period of the i-th ground fissure; α is the first preset constant, which is greater than 0; || is the absolute value function.

[0016] Furthermore, the calculation formula of the overall impact value is:

[0017]

[0018]

[0019] Where, E is the overall impact value of the monitoring data; B is the total number of types of monitoring data; D is the total number of cracks occurring in the specified area; T b,d is the characteristic value of the change of the b-th monitoring data under the d-th ground fissure; γ d is the risk index value of the d-th ground fissure; is the relative entropy between the change characteristic value and the risk index value of the b-th monitoring data; is the distribution probability of the normalized change characteristic value of the b-th monitoring data under the d-th ground fissure among all the normalized change characteristic values ​​of the b-th monitoring data; is the distribution probability of the normalized risk index value of the d-th ground fissure among all normalized risk index values; ln is a logarithmic function with a natural constant as the base; β is a second preset constant, which is greater than 0; and norm is a normalization function.

[0020] Furthermore, the method of obtaining the actual reference value of each monitoring data according to the fluctuation of the change characteristic value of each monitoring data and the fluctuation of the risk index value, as well as the overall impact value, and then screening out the main monitoring data is:

[0021] For any type of monitoring data, the ratio of the variance of the change characteristic value of the monitoring data to the variance of the risk index value is obtained as the impact reference value of the monitoring data;

[0022] The negative correlation and normalization result of the difference between the impact reference value of the monitoring data and the reference preset constant is used as the impact weight value of the monitoring data;

[0023] The product of the impact weight value of the monitoring data and the overall impact value will be used as the actual reference value of the monitoring data;

[0024] Obtain the actual reference value of each monitoring data and arrange it from small to large to obtain an actual reference value sequence;

[0025] The differences between two adjacent actual reference values ​​in the actual reference value sequence are both taken as the first differences;

[0026] The actual reference value sequence is divided into two parts by dividing between the two actual reference values ​​corresponding to the largest first difference, and the monitoring data corresponding to the actual reference values ​​in the latter part of the actual reference value sequence after the division are used as the main monitoring data.

[0027] Furthermore, the method of obtaining similarity values ​​of different types of main monitoring data according to the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk indicator values, and then dividing the main monitoring data into data categories is:

[0028] For any two main monitoring data, according to the difference in actual reference values ​​of the two main monitoring data and the difference in risk index values ​​between two different ground fissures, obtain similar reference values ​​of the two main monitoring data between the corresponding two ground fissures;

[0029] Obtain similarity reference values ​​of the two main monitoring data between any two ground fissures, and use the minimum similarity reference value as the similarity value of the two main monitoring data;

[0030] The similarity values ​​of any two main monitoring data are obtained, and the main monitoring data are clustered by the K-means clustering algorithm based on the similarity values ​​to obtain data categories.

[0031] Furthermore, the calculation formula of the similarity reference value is:

[0032]

[0033] In the formula, is the similarity reference value between the xth main monitoring data and the yth main monitoring data at the i-th ground fissure and the k-th ground fissure; F x is the actual reference value of the xth type of main monitoring data; F y is the actual reference value of the yth type of main monitoring data; i is the risk index value of the i-th ground fissure; k is the risk index value of the kth ground fissure; μ is the third preset constant, which is greater than 0.

[0034] Furthermore, the method for obtaining the target data category is:

[0035] The mean of the variance of all monitoring data between the last ground fissure ending time and the current time is taken as the first eigenvalue;

[0036] Obtain the variance of each main monitoring data in each data category under each ground fissure as the first value;

[0037] Obtaining the mean of the first values ​​of each type of main monitoring data in each data category as the first reference value of the corresponding type of main monitoring data in the corresponding data category;

[0038] Obtaining the mean of all first reference values ​​in each data category as the second characteristic value of the corresponding data category;

[0039] Obtaining a difference between the first eigenvalue and each of the second eigenvalues, each as a second difference;

[0040] The data category corresponding to the second eigenvalue corresponding to the smallest second difference is used as the target data category.

[0041] Furthermore, the method for obtaining the predicted risk index value is:

[0042] Obtain the similarity values ​​corresponding to the main monitoring data in the target data category as reference similarity values;

[0043] The risk index values ​​corresponding to all reference similarity values ​​are used as reference risk index values;

[0044] The duplicate reference risk index values ​​are removed and the mean of the retained reference risk index values ​​is obtained as the predicted risk index value for the specified area.

[0045] Furthermore, the method for obtaining the risk index value is:

[0046] Obtain the length, width, depth, number of buildings affected and number of casualties after each ground fissure occurs in the designated area, and manually score the length, width, depth, number of buildings affected and number of casualties after each ground fissure occurs;

[0047] Based on human scoring, a five-layer fully connected neural network model is used to obtain the risk index value of each ground fissure.

[0048] The present invention has the following beneficial effects:

[0049] According to the changes of each monitoring data in the time period of each ground fissure occurrence, the change characteristic value of each monitoring data under each ground fissure is obtained to accurately reflect the change of each monitoring data under each ground fissure, so as to accurately determine the main monitoring data in the subsequent period; then, according to the difference between the change characteristic value and the risk index value of each monitoring data under each ground fissure, as well as the distribution difference between the change characteristic value and the risk index value of each monitoring data, the overall impact value of the monitoring data is obtained to determine the impact of the monitoring data on the ground fissure disaster; further, according to the fluctuation of the change characteristic value of each monitoring data and the fluctuation of the risk index value, as well as the overall impact value, the actual reference value of each monitoring data is obtained to determine the importance of each monitoring data in representing the ground fissure disaster information, and then the main monitoring data is screened out to determine the data that can represent the ground fissure disaster. Monitoring data of ground fissure disaster information; then, according to the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk index values, obtain the similarity values ​​of different types of main monitoring data, determine the similarities between different types of main monitoring data, and then divide the main monitoring data into data categories according to the similarity values, so as to facilitate more accurate and efficient prediction of ground fissure disasters in the future; then, according to the fluctuation differences between each main monitoring data in each data category and each monitoring data between the end time of the last ground fissure and the current time, accurately screen out the target data category, and then based on the similarity values ​​corresponding to all the main monitoring data in the target data category, accurately obtain the predicted risk index value, accurately predict the ground fissure disaster situation in the specified area, and take preventive measures to reduce the losses caused by ground fissure disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A structural block diagram of a data intelligent analysis system for ground fissure monitoring provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a data intelligent analysis system for ground fissure monitoring proposed by the present invention in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0053] Unless defined otherwise, 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 invention belongs.

[0054] The specific scheme of the data intelligent analysis system for ground fissure monitoring provided by the present invention is described in detail below with reference to the accompanying drawings.

[0055] See also Figure 1 , which shows a structural block diagram of a data intelligent analysis system for ground fissure monitoring provided by an embodiment of the present invention, the system includes: a data acquisition module 10, an overall impact value acquisition module 20, a main monitoring data acquisition module 30, a predicted risk index value acquisition module 40, and a data processing module 50.

[0056] The data acquisition module 10 is used to obtain each monitoring data at each time within the time period of each ground fissure occurrence in the specified area and each monitoring data at each time between the end time of the last ground fissure and the current time; and obtain the risk index value of each ground fissure in the specified area.

[0057] Specifically, the embodiment of the present invention takes any region as an example for explanation, and any region selected is used as the designated region. Other types of sensors such as surface deformation sensors, groundwater level monitoring sensors, and earthquake detection sensors are deployed in the designated region to obtain each type of monitoring data of ground fissures at each moment in the designated region. The embodiment of the present invention sets the time interval between two adjacent moments to 0.1 seconds. The implementer can set the size of the time interval between two adjacent moments according to actual conditions, which is not limited here. Each type of monitoring data at each moment in the time period of each ground fissure disaster in the designated region is obtained, stored and backed up, a database of historical ground fissure data is established in the data intelligent analysis system, and a query interface is provided. Each type of monitoring data at each time when a ground fissure disaster occurred in the designated region in history is retrieved through the query interface. Each type of monitoring data at each moment between the end moment of the last ground fissure and the current moment is obtained, so as to facilitate the subsequent prediction of the ground fissure disaster situation in the designated region.

[0058] In order to more accurately predict the ground fissure disaster situation in a specified area, the embodiment of the present invention analyzes the consequences of each ground fissure disaster in history, so as to facilitate the subsequent accurate screening of the main monitoring data that can characterize the ground fissure disaster information, so as to accurately and efficiently predict the ground fissure disaster situation in the specified area. The embodiment of the present invention obtains the crack length, crack width, crack depth, number of buildings affected and number of casualties after each ground fissure occurs in the specified area, and manually scores the crack length, crack width, crack depth, number of buildings affected and number of casualties after each ground fissure occurs; wherein, the higher the score of the manual score, the greater the degree of loss caused by the ground fissure disaster, and then obtains the risk index value of each ground fissure based on the manual score through a five-layer fully connected neural network model. wherein, the loss function of the five-layer fully connected neural network model is a cross entropy function, and the five-layer fully connected neural network model is trained by the gradient descent method. the manual scores of the crack length, crack width, crack depth, number of buildings affected and number of casualties after each ground fissure occurs are input into the trained five-layer fully connected neural network model, and the output result is the risk index value of the corresponding ground fissure. The larger the risk index value, the more serious the corresponding ground fissure disaster. So far, the risk index value of each ground fissure occurring in the specified area is obtained.

[0059] It should be noted that after each ground fissure disaster occurs, staff will repair the ground fissures in the designated area.

[0060] The overall impact value acquisition module 20 is used to obtain the change characteristic value of each monitoring data under each ground fissure according to the change of each monitoring data during the time period of each ground fissure occurrence; and to obtain the overall impact value of the monitoring data according to the difference between the change characteristic value of each monitoring data under each ground fissure and the risk index value, and the distribution difference between the change characteristic value of each monitoring data and the risk index value.

[0061] Specifically, in actual situations, the severity of each ground fissure disaster is different. In order to accurately determine the degree of influence of each monitoring data on the ground fissure disaster, the change of each monitoring data in each ground fissure occurrence time period is analyzed to obtain the change characteristic value of each monitoring data under each ground fissure, wherein the larger the change characteristic value, the greater the degree of change of the corresponding monitoring data in the corresponding ground fissure occurrence time period. In order to accurately obtain the degree of influence of each monitoring data on the ground fissure disaster, the embodiment of the present invention analyzes the similarity between the change of the change characteristic value of each monitoring data and the change of the risk index value. When the change of the change characteristic value of a certain monitoring data and the change of the risk index value are more similar, it means that the influence of this monitoring data on the ground fissure disaster is greater. At the same time, when the change characteristic value of a certain monitoring data under each ground fissure is more similar to the risk index value of the same ground fissure, it means that the change of this monitoring data is more consistent with the occurrence of each ground fissure, indicating that this monitoring data can better represent the information of the ground fissure disaster, and indirectly indicates that the influence of this monitoring data on the ground fissure disaster is greater. Therefore, according to the difference between the change characteristic value and the risk index value of each type of monitoring data under each ground fissure, as well as the distribution difference between the change characteristic value and the risk index value of each type of monitoring data, the overall impact value of all types of monitoring data on ground fissures is obtained. The larger the overall impact value, the greater the impact of the monitoring data on ground fissure disasters, and the more accurate the prediction result of ground fissure disasters in a specified area based on the monitoring data. Among them, the specific method for obtaining the overall impact value of monitoring data is as follows:

[0062] (1) Obtain the change characteristic value.

[0063] Preferably, the method for obtaining the change characteristic value is as follows: for any type of monitoring data within any ground fissure occurrence time period, the greater the difference between the type of monitoring data at each moment within the ground fissure occurrence time period and the mean value of the type of monitoring data within the ground fissure occurrence time period, the greater the fluctuation degree of the type of monitoring data within the ground fissure occurrence time period; in order to limit the degree of change of each type of monitoring data, a preset critical value will be set for each type of monitoring data in actual situations, and the implementer can set the preset critical value of each type of monitoring data according to actual situations, which is not limited here. In the embodiment of the present invention, the maximum preset critical value of each type of monitoring data is taken as an example. When the maximum type a monitoring data within the i-th ground fissure occurrence time period is equal to the maximum preset critical value of the type a monitoring data, it means that the fluctuation degree of the type a monitoring data within the i-th ground fissure occurrence time period is greater. Therefore, the embodiment of the present invention obtains the change characteristic value of the monitoring data under the ground fissure based on the absolute value of the difference between the monitoring data at each moment within the time period when the ground fissure occurs and the mean value of the monitoring data within the time period when the ground fissure occurs, and the difference between the maximum preset critical value of the monitoring data and the maximum monitoring data within the time period when the ground fissure occurs.

[0064] As an example, taking the ath type of monitoring data within the time period of the i-th ground fissure as an example, the calculation formula for obtaining the change characteristic value of the ath type of monitoring data under the i-th ground fissure is:

[0065]

[0066] Where, T a,i is the characteristic value of the change of the a-th monitoring data under the i-th ground fissure; M i is the total number of times during the time period of the i-th ground fissure; U a,i,m is the ath type of monitoring data at the mth moment during the time period of the i-th ground fissure occurrence; is the mean value of the ath type of monitoring data during the time period of the i-th ground fissure occurrence; U′ a,max is the maximum preset critical value of the a-th type of monitoring data; U a,i,max is the largest a-th type of monitoring data during the time period of the i-th ground fissure; α is the first preset constant, which is greater than 0; || is the absolute value function.

[0067] In the embodiment of the present invention, α is set to 1 to avoid the denominator being 0. The implementer may set the value of α according to actual conditions, which is not limited here.

[0068] It should be noted that The larger the U a,i,m and The greater the difference, the greater the fluctuation of the a-th monitoring data during the i-th ground fissure occurrence period. The larger the T a,i The larger the a,max ―U a,i,max ) is greater than or equal to 0, when (U′ a,max ―U a,i,max ) is smaller, U a,i,max with U′ a,max The more equal they are, the greater the fluctuation of the a-th monitoring data during the i-th ground fissure occurrence period, U′ a,max ―Ua ,i,max Must be greater than or equal to 0, The larger the T a,i The larger the a,i The larger it is, the greater the fluctuation of the a-th monitoring data during the time period of the i-th ground fissure.

[0069] According to the method for obtaining the change characteristic value of the a-th type of monitoring data under the i-th ground fissure, the change characteristic value of each type of monitoring data under each ground fissure is obtained.

[0070] (2) Obtain the overall impact value.

[0071] In order to accurately obtain the degree of influence of each monitoring data on the ground fissure disaster, the embodiment of the present invention arranges the change characteristic value and the risk index value of each monitoring data in chronological order and performs curve fitting, obtains the fitting curve of the change characteristic value of each monitoring data and the fitting curve of the risk index value, when the change of the fitting curve of the change characteristic value of a certain monitoring data and the fitting curve of the risk index value are closer, it means that the degree of influence of this monitoring data on the ground fissure disaster is greater. In actual situations, the change characteristic value and the risk index value of each monitoring data may not be within a unified dimension, resulting in the inability to accurately analyze the similarity between the fitting curve of the change characteristic value of each monitoring data and the fitting curve of the risk index value, making it impossible to accurately analyze the influence of each monitoring data on the ground fissure disaster, therefore, the embodiment of the present invention normalizes the change characteristic value and the risk index value of each monitoring data, so that the change characteristic value and the risk index value of each monitoring data are within a unified dimension, and at the same time, according to the normalized change characteristic value and the risk index value, curve fitting is performed, that is, the fitting curve of the normalized change characteristic value of each monitoring data and the fitting curve of the normalized risk index value are obtained. Wherein, the method for obtaining the fitting curve is a known technology and will not be repeated. The embodiment of the present invention determines the similarity between the change characteristic value of each monitoring data and the change of the risk index value by obtaining the relative entropy between the fitting curve of the normalized change characteristic value of each monitoring data and the fitting curve of the normalized risk index value. When the relative entropy is smaller, it means that the change trend of the change characteristic value of the corresponding type of monitoring data is more consistent with the change trend of the risk index value, which indirectly indicates that the corresponding type of monitoring data has a greater impact on the ground fissure disaster. Among them, the method for obtaining the relative entropy is a well-known technology and will not be repeated here. Therefore, according to the difference between the normalized change characteristic value of each monitoring data under each ground fissure and the normalized risk index value of the same ground fissure, and the relative entropy between the fitting curve of the normalized change characteristic value of each monitoring data and the fitting curve of the normalized risk index value, the overall impact value of all types of monitoring data on the ground fissure disaster is obtained. Among them, the calculation formula for obtaining the overall impact value is:

[0072]

[0073]

[0074] Where, E is the overall impact value of the monitoring data; B is the total number of types of monitoring data; D is the total number of cracks occurring in the specified area; T b,d is the characteristic value of the change of the b-th monitoring data under the d-th ground fissure; γ d is the risk index value of the d-th ground fissure; is the relative entropy between the change characteristic value and the risk index value of the b-th monitoring data; is the distribution probability of the normalized change characteristic value of the b-th monitoring data under the d-th ground fissure among all the normalized change characteristic values ​​of the b-th monitoring data; is the distribution probability of the normalized risk index value of the d-th ground fissure among all normalized risk index values; ln is a logarithmic function with a natural constant as the base; β is a second preset constant, which is greater than 0; and norm is a normalization function.

[0075] In the embodiment of the present invention, β is set to 1 to avoid the denominator being 0. The implementer may set the value of β according to actual conditions, which is not limited here.

[0076] It should be noted that (norm(T b,d )―norm(γ d )) 2 The smaller the norm(T b,d ) and norm(γ d ), the closer they are, the more synchronized the changing trend of the characteristic value of the b-type monitoring data is with the changing trend of the risk index value, which indirectly indicates that the monitoring data has a greater impact on ground fissure disasters. The smaller it is, the larger E is; The smaller it is, the more similar the fitting curve of the normalized change characteristic value of the b-th monitoring data is to the fitting curve of the normalized risk index value, which indirectly indicates that the monitoring data has a greater impact on ground fissure disasters, and the larger E is. The value range of is greater than or equal to 0; the larger the E is, the greater the impact of the monitoring data on ground fissure disasters is, and the more accurate the prediction of ground fissure disasters in the specified area based on the monitoring data is.

[0077] The main monitoring data acquisition module 30 is used to obtain the actual reference value of each monitoring data according to the fluctuation of the change characteristic value of each monitoring data and the fluctuation of the risk index value, as well as the overall impact value, and then screen out the main monitoring data.

[0078] Specifically, when the changing trend of the characteristic value of a certain monitoring data is more similar to that of the risk index value, it means that the monitoring data can better reflect the information of ground fissure disasters. Therefore, the embodiment of the present invention obtains the ratio of the variance of the characteristic value of the change of each monitoring data to the variance of the risk index value as the influence reference value of the corresponding monitoring data. When the influence reference value is closer to 1, it means that the corresponding monitoring data can better represent the information of ground fissure disasters. Therefore, the embodiment of the present invention obtains the actual reference value of each monitoring data according to the fluctuation of the characteristic value of the change of each monitoring data and the fluctuation of the risk index value, as well as the overall influence value. When the actual reference value is larger, it means that the corresponding monitoring data is more likely to represent the information of ground fissure disasters, and the corresponding monitoring data is more important; therefore, according to the actual reference value of each monitoring data, the main monitoring data that can accurately represent the ground fissure disasters are screened out, and then the ground fissure disasters in the specified area are accurately and efficiently predicted based on the main monitoring data. Among them, the specific method for obtaining the main monitoring data is as follows:

[0079] (1) Obtain the actual reference value.

[0080] Preferably, the method for obtaining the actual reference value is: for any kind of monitoring data, the ratio of the variance of the change characteristic value of the monitoring data to the variance of the risk index value is obtained as the influence reference value of the monitoring data; when the influence reference value is closer to 1, it means that the monitoring data can better represent the information of ground fissure disaster. In the embodiment of the present invention, the reference preset constant is set to 1, and the implementer can set the size of the reference preset constant according to the actual situation, which is not limited here. The result of negative correlation and normalization of the difference between the influence reference value of the monitoring data and the reference preset constant is used as the influence weight value of the monitoring data; the larger the influence weight value, the more likely it is that the monitoring data is important monitoring data that represents the information of ground fissure disaster, and then the product of the influence weight value of the monitoring data and the overall influence value is obtained as the actual reference value of the monitoring data. Among them, the larger the actual reference value, the more likely it is that the monitoring data can represent the information of ground fissure disaster, and the more likely it is to be the main monitoring data.

[0081] As an example, taking the ath type of monitoring data in step S2 as an example, the calculation formula for obtaining the actual reference value of the ath type of monitoring data is:

[0082]

[0083] In the formula, F a is the actual reference value of the monitoring data of type a; E is the overall impact value of the monitoring data; is the variance of the characteristic value of the a-th monitoring data; γ 2 is the variance of the risk indicator value; is a reference preset constant, which is set to 1 in the embodiment of the present invention; || is an absolute value function; exp is an exponential function with a natural constant as the base; is the impact reference value of the a-th type of monitoring data; is the impact weight value of the ath type of monitoring data.

[0084] It should be noted that The smaller, The closer it is to 1, the more consistent the changing trend of the characteristic value of the a-type monitoring data is with the changing trend of the risk index value, which indirectly indicates that the a-type monitoring data can better represent the information of ground fissure disasters. The larger the F a The larger the E, the greater the impact of the monitoring data on ground fissure disasters, which indirectly means that the monitoring data can more accurately reflect the ground fissure disaster information. a The larger the value, the larger the F a The larger it is, the more the a-th type of monitoring data can represent the information of ground fissure disasters, and the more important the a-th type of monitoring data is in the process of analyzing ground fissure disasters.

[0085] According to the method for obtaining the actual reference value of the ath type of monitoring data, the actual reference value of each type of monitoring data is obtained.

[0086] (2) Obtain key monitoring data.

[0087] It is known that the larger the actual reference value is, the more the corresponding monitoring data can represent the information of ground fissure disasters. Because there are many types of monitoring data for ground fissure disasters, if the ground fissure disasters in a specified area are predicted directly based on all types of monitoring data, it will generate too much calculation and easily lead to inaccurate prediction of ground fissure disasters in the specified area. Therefore, the embodiment of the present invention screens out the main monitoring data that can accurately represent the information of ground fissure disasters based on the actual reference value of the monitoring data.

[0088] Preferably, the method for obtaining the main monitoring data is: obtain the actual reference value of each monitoring data and arrange them from small to large to obtain an actual reference value sequence; the absolute value of the difference between two adjacent actual reference values ​​in the actual reference value sequence is used as the first difference; the larger the first difference, the greater the difference in ground fissure disaster information represented by the two monitoring data corresponding to the corresponding two adjacent actual reference values, and the monitoring data corresponding to the larger actual reference value is more likely to be the main monitoring data. Therefore, the largest first difference is obtained, and the actual reference value sequence is divided between the two actual reference values ​​corresponding to the largest first difference, and the actual reference value sequence is divided into two parts, and the monitoring data corresponding to the actual reference value in the latter part of the actual reference value sequence after the division is used as the main monitoring data, that is, the monitoring data corresponding to the larger actual reference value in the actual reference value sequence is used as the main monitoring data.

[0089] The predicted risk index value acquisition module 40 is used to obtain the similarity values ​​of different types of main monitoring data based on the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk index values, and then divide the main monitoring data into data categories; filter out the target data category based on the fluctuation differences between each type of main monitoring data in each data category and each type of monitoring data between the end time of the last ground fissure and the current time; and obtain the predicted risk index value based on the similarity values ​​corresponding to all main monitoring data in the target data category.

[0090] Specifically, in order to accurately and efficiently predict the ground fissure conditions in a specified area based on the main monitoring data, an embodiment of the present invention obtains similarity values ​​of different types of main monitoring data based on the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk indicator values, clusters the main monitoring data, and divides the main monitoring data with similar information on ground fissure disasters into the same data category. The fluctuation of each type of main monitoring data in each data category under each ground fissure disaster is analyzed to accurately obtain the changes in the main monitoring data in each data category. Obtain the fluctuation of all types of monitoring data between the last time the ground fissure ended and the current time, and compare the fluctuation of all types of monitoring data between the last time the ground fissure ended and the current time with the changes in the main monitoring data in each data category. When the fluctuation of all types of monitoring data between the last time the ground fissure ended and the current time is more similar to the changes in the main monitoring data in a certain data category, it means that the changes in the main monitoring data in the data category are more similar to the change trend of the monitoring data between the last time the ground fissure ended and the current time, and the data category is more likely to be the target data category. Then, based on the similarity values ​​corresponding to all the main monitoring data in the target data category, obtain the predicted risk index value of the specified area, and accurately predict the ground fissure disaster situation in the specified area. Among them, the specific method for obtaining the predicted risk index value is as follows:

[0091] (1) Obtain similarity value.

[0092] Preferably, the method for obtaining the similarity value is: for any two main monitoring data, according to the difference in actual reference values ​​of the two main monitoring data and the difference in risk index values ​​between two different ground fissures, obtain the similarity reference value of the two main monitoring data between the corresponding two ground fissures; the smaller the similarity reference value, the more similar the two main monitoring data are between the corresponding two ground fissures. Therefore, the embodiment of the present invention obtains the similarity reference value of the two main monitoring data between any two ground fissures, and takes the smallest similarity reference value as the similarity value of the two main monitoring data.

[0093] As an example, taking the xth type of main monitoring data and the yth type of main monitoring data as examples, the calculation formula for obtaining the similarity reference value between the xth type of main monitoring data and the yth type of main monitoring data at the i-th ground fissure and the k-th ground fissure is:

[0094]

[0095] In the formula, is the similarity reference value between the xth main monitoring data and the yth main monitoring data at the i-th ground fissure and the k-th ground fissure; F x is the actual reference value of the xth type of main monitoring data; F y is the actual reference value of the yth type of main monitoring data; i is the risk index value of the i-th ground fissure; k is the risk index value of the kth ground fissure; μ is the third preset constant, which is greater than 0.

[0096] In the embodiment of the present invention, μ is set to 0.1 to avoid ((F x ―F y ) 2 +μ) and ((γ i ―γ k ) 2 +μ) is 0, and the implementer can set the size of μ according to the actual situation, which is not limited here.

[0097] It should be noted that ((F x ―F y ) 2 +μ) is smaller, F x With F y The closer they are, the more consistent the impact of the xth and yth main monitoring data on ground fissure disasters is; the more similar the xth and yth main monitoring data are, The smaller; ((γ i ―γ k ) 2 +μ) is smaller, γ i With γ k The more equal they are, the more similar the changes of the main monitoring data between the i-th ground fissure and the k-th ground fissure are. The smaller; therefore, The smaller it is, the more similar the changes of the x-th main monitoring data and the y-th main monitoring data between the ith ground fissure and the k-th ground fissure are; the more similar the x-th main monitoring data and the y-th main monitoring data are, the more likely they are to be classified into the same data category.

[0098] According to the method of obtaining the similarity reference value between the xth type of main monitoring data and the yth type of main monitoring data at the ith ground fissure and the kth ground fissure, the similarity reference value between the xth type of main monitoring data and the yth type of main monitoring data between any two ground fissures is obtained, and the minimum similarity reference value is used as the similarity value between the xth type of main monitoring data and the yth type of main monitoring data, so as to accurately reflect the similarity between the xth type of main monitoring data and the yth type of main monitoring data.

[0099] According to the method for obtaining the similarity value of the x-th type of main monitoring data and the y-th type of main monitoring data, the similarity value of any two main monitoring data is obtained.

[0100] (2) Obtain data category.

[0101] According to the similarity values ​​of any two main monitoring data, the main monitoring data are clustered by the K-means clustering algorithm to obtain data categories. Among them, the k value in the K-means clustering algorithm is obtained by the elbow method. Among them, the K-means clustering algorithm and the elbow method are both well-known technologies and will not be described in detail.

[0102] (3) Obtain the target data category.

[0103] The mean of the variance of all monitoring data between the last ground fissure ending time and the current time is obtained as the first eigenvalue. When the first eigenvalue is larger, it means that the degree of fluctuation of the monitoring data between the last ground fissure ending time and the current time is larger. In order to accurately predict the ground fissure disaster in the specified area according to the first eigenvalue, the embodiment of the present invention obtains the variance of each main monitoring data in each data category under each ground fissure as the first value; obtains the mean of the first value of each main monitoring data in each data category as the first reference value of the corresponding main monitoring data in the corresponding data category, and accurately reflects the fluctuation of the corresponding main monitoring data under the ground fissure disaster through the first reference value; obtains the mean of all first reference values ​​in each data category as the second eigenvalue of the corresponding data category, reflecting the change of the main monitoring data in the corresponding data category. The difference between the first eigenvalue and each second eigenvalue is obtained as the second difference; when the second difference is smaller, it means that the main monitoring data in the data category corresponding to the second eigenvalue is more similar to the change trend of all monitoring data between the last ground fissure ending time and the current time. Therefore, the data category corresponding to the second eigenvalue corresponding to the smallest second difference is used as the target data category. If the smallest second difference corresponds to two data categories, one of the data categories is selected as the target data category.

[0104] (4) Obtain the predicted risk index value.

[0105] Preferably, the method for obtaining the predicted risk index value is: obtaining the similarity value corresponding to the main monitoring data in the target data category as the reference similarity value; taking the risk index values ​​corresponding to all the reference similarity values ​​as the reference risk index value; wherein, the reference risk index values ​​are repeated, that is, the same ground fissure corresponds to multiple reference risk index values, therefore, the repeated reference risk index values ​​are removed to ensure that only one reference risk index value corresponds to one ground fissure. Obtain the mean of the retained reference risk index values ​​as the predicted risk index value for the specified area.

[0106] The data processing module 50 is used to predict the ground fissure disasters in the designated area according to the predicted risk index value.

[0107] Specifically, the predicted risk index value is normalized. When the normalized predicted risk index value is greater than or equal to the preset risk index value threshold, a serious ground fissure disaster may occur in the designated area, and the staff needs to take measures to prevent ground fissure disaster losses in the designated area to ensure people's safety and reduce the losses caused by ground fissure disasters. In the embodiment of the present invention, the preset risk index value threshold is set to 0.3. The implementer can set the size of the preset risk index value threshold according to the actual situation, which is not limited here.

[0108] Providing early warning of ground fissure disasters in designated areas based on predicted risk index values ​​can help reduce the losses caused by ground fissure disasters and improve the accuracy and credibility of the intelligent analysis system for ground fissure monitoring data.

[0109] The embodiment of the present invention takes a designated region as an example for analysis, and the present invention can be used to analyze other regions.

[0110] In summary, the embodiment of the present invention obtains the monitoring data and risk index value of ground fissures; according to the change of the monitoring data, obtains the change characteristic value; according to the difference between the change characteristic value and the risk index value, obtains the overall impact value of the monitoring data; according to the fluctuation of the change characteristic value and the risk index value, and the overall impact value, the main monitoring data is screened out; according to the similarity between the main monitoring data, the main monitoring data is divided into data categories, and according to the fluctuation difference between the main monitoring data in each data category and each monitoring data between the end time of the last ground fissure to the current time, the target data category is screened out, and then the predicted risk index value is obtained to predict the ground fissure disaster. The present invention accurately determines the target data category, and then accurately and efficiently predicts the ground fissure disaster in the specified area.

[0111] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A data intelligent analysis system for ground fissure monitoring, characterized in that: The system includes the following steps: The data acquisition module is used to obtain each monitoring data at each time within the time period of each ground fissure occurrence in the specified area and each monitoring data at each time between the end time of the last ground fissure and the current time; and obtain the risk index value of each ground fissure in the specified area; The overall impact value acquisition module is used to obtain the change characteristic value of each monitoring data under each ground fissure according to the change of each monitoring data in the time period of each ground fissure; according to the difference between the change characteristic value of each monitoring data under each ground fissure and the risk index value, and the distribution difference between the change characteristic value of each monitoring data and the risk index value, to obtain the overall impact value of the monitoring data; The main monitoring data acquisition module is used to obtain the actual reference value of each monitoring data according to the fluctuation of the change characteristic value of each monitoring data and the fluctuation of the risk indicator value, as well as the overall impact value, and then screen out the main monitoring data; The predicted risk index value acquisition module is used to obtain the similarity values ​​of different types of main monitoring data according to the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk index values, and then divide the main monitoring data into data categories; according to the fluctuation difference between each type of main monitoring data in each data category and each type of monitoring data between the end time of the last ground fissure and the current time, the target data category is screened out; based on the similarity values ​​corresponding to all main monitoring data in the target data category, the predicted risk index value is obtained; A data processing module is used to predict ground fissure disasters in a designated area based on the predicted risk index value; The calculation formula of the overall impact value is: Where, E is the overall impact value of the monitoring data; B is the total number of types of monitoring data; D is the total number of cracks occurring in the specified area; T b，d is the characteristic value of the change of the b-th monitoring data under the d-th ground fissure; γ d is the risk index value of the d-th ground fissure; is the relative entropy between the change characteristic value and the risk index value of the b-th monitoring data; is the distribution probability of the normalized change characteristic value of the b-th monitoring data under the d-th ground fissure among all the normalized change characteristic values ​​of the b-th monitoring data; is the distribution probability of the normalized risk index value of the d-th ground fissure among all normalized risk index values; ln is a logarithmic function with a natural constant as the base; β is a second preset constant, which is greater than 0; and norm is a normalization function.

2. A data intelligent analysis system for ground fissure monitoring as claimed in claim 1, characterized in that: The method for obtaining the change characteristic value is: For any kind of monitoring data within any time period of ground fissure occurrence, the change characteristic value of the monitoring data under the ground fissure is obtained based on the difference between the monitoring data at each moment within the time period of ground fissure occurrence and the mean value of the monitoring data within the time period of ground fissure occurrence, and the difference between the maximum preset critical value of the monitoring data and the maximum value of the monitoring data within the time period of ground fissure occurrence.

3. A data intelligent analysis system for ground fissure monitoring as claimed in claim 1, characterized in that: The calculation formula of the change characteristic value is: Where, T a，i is the characteristic value of the change of the a-th monitoring data under the i-th ground fissure; M i is the total number of moments in the time period during which the i-th ground fissure occurs; U a，i，m is the ath type of monitoring data at the mth moment during the time period of the i-th ground fissure occurrence; is the mean value of the ath type of monitoring data during the time period of the i-th ground fissure occurrence; U′ a，max is the maximum preset critical value of the a-th type of monitoring data; U a，i，max is the largest a-th type of monitoring data during the time period of the i-th ground fissure; α is the first preset constant, which is greater than 0; || is the absolute value function.

4. A data intelligent analysis system for ground fissure monitoring as claimed in claim 1, characterized in that: The method for obtaining the actual reference value of each monitoring data according to the fluctuation of the change characteristic value of each monitoring data, the fluctuation of the risk index value, and the overall impact value, and then screening out the main monitoring data is: For any type of monitoring data, the ratio of the variance of the change characteristic value of the monitoring data to the variance of the risk index value is obtained as the impact reference value of the monitoring data; The negative correlation and normalization result of the difference between the impact reference value of the monitoring data and the reference preset constant is used as the impact weight value of the monitoring data; The product of the impact weight value of the monitoring data and the overall impact value will be used as the actual reference value of the monitoring data; Obtain the actual reference value of each monitoring data and arrange it from small to large to obtain an actual reference value sequence; The differences between two adjacent actual reference values ​​in the actual reference value sequence are both taken as the first differences; The actual reference value sequence is divided into two parts by dividing between the two actual reference values ​​corresponding to the largest first difference, and the monitoring data corresponding to the actual reference values ​​in the latter part of the actual reference value sequence after the division are used as the main monitoring data.

5. The data intelligent analysis system for ground fissure monitoring according to claim 1, characterized in that: The method of obtaining the similarity values ​​of different types of main monitoring data according to the differences in actual reference values ​​of different types of main monitoring data and the differences in different risk indicator values, and then dividing the main monitoring data into data categories is: For any two main monitoring data, according to the difference in actual reference values ​​of the two main monitoring data and the difference in risk index values ​​between two different ground fissures, obtain similar reference values ​​of the two main monitoring data between the corresponding two ground fissures; Obtain similarity reference values ​​of the two main monitoring data between any two ground fissures, and use the minimum similarity reference value as the similarity value of the two main monitoring data; The similarity values ​​of any two main monitoring data are obtained, and the main monitoring data are clustered by the K-means clustering algorithm based on the similarity values ​​to obtain data categories.

6. A data intelligent analysis system for ground fissure monitoring as claimed in claim 5, characterized in that: The calculation formula of the similarity reference value is: In the formula, is the similarity reference value between the xth main monitoring data and the yth main monitoring data at the i-th ground fissure and the k-th ground fissure; F x is the actual reference value of the xth type of main monitoring data; F y is the actual reference value of the yth type of main monitoring data; i is the risk index value of the i-th ground fissure; k is the risk index value of the kth ground fissure; μ is the third preset constant, which is greater than 0.

7. A data intelligent analysis system for ground fissure monitoring as claimed in claim 1, characterized in that: The method for obtaining the target data category is: The mean of the variance of all monitoring data between the last ground fissure ending time and the current time is taken as the first eigenvalue; Obtain the variance of each main monitoring data in each data category under each ground fissure as the first value; Obtaining the mean of the first values ​​of each type of main monitoring data in each data category as the first reference value of the corresponding type of main monitoring data in the corresponding data category; Obtaining the mean of all first reference values ​​in each data category as the second characteristic value of the corresponding data category; Obtaining a difference between the first eigenvalue and each of the second eigenvalues, each as a second difference; The data category corresponding to the second eigenvalue corresponding to the smallest second difference is used as the target data category.

8. The data intelligent analysis system for ground fissure monitoring according to claim 1, characterized in that: The method for obtaining the predicted risk index value is: Obtain the similarity values ​​corresponding to the main monitoring data in the target data category as reference similarity values; The risk index values ​​corresponding to all reference similarity values ​​are used as reference risk index values; The duplicate reference risk index values ​​are removed and the mean of the retained reference risk index values ​​is obtained as the predicted risk index value for the specified area.

9. The data intelligent analysis system for ground fissure monitoring according to claim 1, characterized in that: The method for obtaining the risk index value is: Obtain the length, width, depth, number of buildings affected and number of casualties after each ground fissure occurs in the designated area, and manually score the length, width, depth, number of buildings affected and number of casualties after each ground fissure occurs; Based on human scoring, a five-layer fully connected neural network model is used to obtain the risk index value of each ground fissure.

Images