Rock mass disaster early warning method and system based on cloud computing

By using a cloud computing-based rock mass disaster early warning method, the stress time series is segmented and the corrected local anomaly factor is calculated, which solves the problem of inaccurate rock mass disaster early warning in the existing technology and achieves higher early warning accuracy and reliability.

CN118692227BActive Publication Date: 2024-11-08SHANDONG GOLD MINING TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411170051.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-11-08
Estimated Expiration
2044-08-26

AI Technical Summary

Technical Problem

Existing technologies lack accuracy in rock mass disaster early warning. Local outlier detection methods (LOF algorithm) and methods based on impact energy indicators cannot accurately reflect the stress variation characteristics of anchoring components, leading to false warnings.

Method used

A cloud computing-based rock mass disaster early warning method is adopted. By acquiring the stress time series of anchoring components, dividing it into monotonically changing data segments, calculating the degree of fluctuation and the corrected local anomaly factor, and combining the outlier situation and fluctuation characteristics of stress data, the early warning degree is calculated and an early warning reminder is issued.

Benefits of technology

It improves the accuracy of rock mass disaster early warning, reduces false warnings caused by a single abnormal data point, and enhances the early warning capability for instability and deformation of anchoring components.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118692227B_ABST
    Figure CN118692227B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of data processing, and in particular to a rock disaster early warning method and system based on cloud computing. The method includes the steps of: obtaining a stress time series sequence carried by an anchoring component supporting a rock mass; dividing the stress time series sequence into a number of monotonically changing data segments as candidate analysis data segments, calculating the degree of fluctuation according to the candidate analysis data segments, and screening out candidate analysis data segments with a degree of fluctuation greater than a preset degree threshold as analysis data segments; calculating a corrected local anomaly factor according to the analysis data segments, taking stress data with a corrected local anomaly factor greater than a preset anomaly threshold as abnormal data, and calculating the early warning degree according to the abnormal data to assist in early warning reminders. In the abnormal analysis process, the abnormal detection index is corrected in combination with the characteristics of the rock disaster scene, so that the abnormal analysis is more accurate and the accuracy of disaster early warning is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a rock mass disaster early warning method and system based on cloud computing. Background Art

[0002] In the process of underground mining, with the excavation of the underground ore body, the original stress balance state of the surrounding rock is broken, resulting in stress redistribution. When the stress borne by the surrounding rock is greater than its tensile strength or compressive strength, disasters such as spalling, roof collapse and even rock burst will occur, seriously affecting the safe production of the mine. Therefore, in the mining process, with the excavation of the tunnel or the production of the mining site, it is necessary to use anchor components to support and reinforce the rock mass. At the same time, with the progress of production operations, the stress conditions of the anchor components used to support the rock mass will also change significantly during the stress changes of the surrounding rock and ore body. When drastic stress changes occur, it is easy for the anchor components to become unstable and deformed. Therefore, it is necessary to analyze the stress conditions of the anchor components during the mining process, so as to provide early warnings before rock disasters may occur. Before the anchor components become unstable and deformed, they are subjected to large or sudden changes in stress. This kind of stress data will show outlier characteristics, so outlier detection can be used to detect anomalies.

[0003] The Local Outlier Factor detection method (LOF algorithm for short) is often used for rock mass disaster detection. This method mainly performs outlier detection based on the density of the local area where each data point is located. This detection method cannot reflect the change characteristics of the data point. However, the change in the force of the anchor component is also the cause of the instability and deformation of the anchor component. Therefore, it is not accurate to directly use the LOF algorithm for rock mass disaster early warning analysis.

[0004] The patent application document with publication number CN112560135A discloses an early warning method for rock collapse based on impact energy index. This method uses rock vibration information to determine whether there is a collapse. Since vibration information is also generated during the excavation process, this vibration information will interfere with the collapse vibration judgment. Therefore, this method is not accurate enough for early warning analysis. Summary of the invention

[0005] In order to solve the problem of how to improve the accuracy of rock mass disaster early warning, the present invention provides a rock mass disaster early warning method and system based on cloud computing.

[0006] In a first aspect, the present invention provides a rock mass disaster early warning method based on cloud computing, which adopts the following technical solutions:

[0007] The rock mass disaster early warning method based on cloud computing includes the following steps:

[0008] Obtaining the stress time series borne by the anchoring components supporting the rock mass;

[0009] The stress time series is divided into several monotonically changing data segments as candidate analysis data segments, and the degree of fluctuation is calculated. The degree of fluctuation is positively correlated with the data variance and the number of data in the candidate analysis data segments; and the candidate analysis data segments with a degree of fluctuation greater than a preset degree threshold are selected as analysis data segments;

[0010] Calculate the corrected local anomaly factor: , Indicates the first A local anomaly factor of the stress data, wherein the local anomaly factor characterizes the outlier of the stress data, represents the risk level of the analysis data segment, wherein the risk level represents the degree of fluctuation of the analysis data segment, Indicates the maximum stress that the anchoring component can bear. Indicates the first Stress data, Indicates the first The corrected local anomaly factor of each stress data;

[0011] The stress data whose corrected local abnormal factor is greater than the preset abnormal threshold is taken as abnormal data, and the warning degree is calculated to assist in early warning reminder, wherein the warning degree is negatively correlated with the time interval between adjacent abnormal data.

[0012] In the process of acquiring and analyzing data segments, the present invention takes into account the monotony and fluctuation of the data sequence, thereby dividing the stress data in the stress change process of the anchoring component after each excavation of the ore body into one data segment, providing a basis for subsequent accurate analysis of abnormal conditions of the stress data. Furthermore, when calculating and correcting the local abnormality factor, not only the outliers of the stress data are considered, but also the differences between the stress fluctuations and the maximum stress that the stress data can bear are considered, thereby preventing the abnormal analysis from being inaccurate due to only considering the outliers of the stress data, and improving the accuracy of the rock disaster early warning analysis. Furthermore, when performing early warning reminders, by considering the interval of abnormal data, the phenomenon of false warnings caused by the erroneous judgment of a single abnormal data is eliminated, thereby improving the accuracy of the early warning.

[0013] Preferably, the method for obtaining the risk level includes:

[0014] ;

[0015] in, Indicates the first Stress data, Indicates the number of stress data in this analysis data segment, Represents the absolute value symbol.

[0016] The present invention introduces the variance of the fluctuations of all two adjacent stress data when calculating the risk level to reflect the sudden change of stress data, thereby comprehensively characterizing abnormal factors and improving the accuracy of abnormal detection; further, the mean growth rate of all stress data is also introduced when calculating the risk level to accurately reflect the comprehensive change of stress data, thereby comprehensively characterizing abnormal factors and improving the accuracy of abnormal detection.

[0017] Preferably, the calculation of the warning level includes:

[0018] ;

[0019] in, Indicates the stress time series The time of abnormal data, Indicates the stress time series The time of abnormal data, Indicates the number of abnormal data in the stress time series. represents an exponential function with a natural constant as base, Indicates the warning level.

[0020] When calculating the warning degree, the present invention takes into account the time interval of adjacent abnormal data, more accurately reflects the occurrence of abnormal data when rock disasters occur, prevents warning errors caused by misjudgment of single abnormal data, and improves the accuracy of warning.

[0021] Preferably, the method for obtaining the candidate analysis data segment includes:

[0022] The difference between each stress data and the previous stress data in the stress time series is calculated, and recorded as the first-order difference of each stress data; the first stress data is taken as the starting data, starting from the starting data, the first stress data with a first-order difference less than or equal to 0 is obtained and recorded as the cut-off data, and the data segment consisting of all stress data between the starting data and the cut-off data is recorded as the analysis data segment; in response to the cut-off data not being the last stress data, the next data of the cut-off data is taken as the new starting data, starting from the new starting data, the first stress data with a first-order difference less than or equal to 0 is obtained and recorded as the new cut-off data, and the data segment consisting of the stress data between the new starting data and the new cut-off data is recorded as the new analysis data segment; in response to the new cut-off data being the last stress data, all analysis data segments are obtained.

[0023] The present invention divides the stress time series into a plurality of monotonically changing data segments, thereby ensuring that the stress data of the anchoring component in the stress change process after each ore body excavation is divided into one data segment.

[0024] Preferably, the calculation of the degree of fluctuation includes:

[0025] The product of the number of data in the candidate analysis data segment and the variance of all data in the candidate analysis data segment is taken as the fluctuation degree of the candidate analysis data segment.

[0026] When calculating the degree of fluctuation, the present invention not only considers the change of data, but also considers the amount of data in the data segment, thereby eliminating the interference of data in the stress change balance process of the anchoring component after each excavation of the ore body.

[0027] Preferably, the method for obtaining the local abnormal factor includes:

[0028] The LOF algorithm is used to perform anomaly detection on all stress data in the stress time series to obtain the local anomaly factor of each stress data.

[0029] Preferably, the auxiliary early warning reminder includes:

[0030] If the warning level is greater than the preset warning level threshold, a warning reminder will be issued; if the warning level is less than or equal to the preset warning level threshold, no warning reminder will be issued.

[0031] In a second aspect, the present invention provides a rock mass disaster early warning system based on cloud computing, which adopts the following technical solutions:

[0032] The rock mass disaster early warning system based on cloud computing includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the rock mass disaster early warning method based on cloud computing is implemented.

[0033] By adopting the above technical solution, the above cloud computing-based rock disaster warning method is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is made according to the memory and the processor for easy use.

[0034] The present invention has the following technical effects:

[0035] In the process of acquiring and analyzing data segments, the present invention takes into account the monotony and fluctuation of the data sequence, thereby dividing the stress data in the stress change process of the anchoring component after each excavation of the ore body into one data segment, providing a basis for subsequent accurate analysis of abnormal conditions of the stress data. Furthermore, when calculating and correcting the local abnormality factor, not only the outliers of the stress data are considered, but also the differences between the stress fluctuations and the maximum stress that the stress data can bear are considered, thereby preventing the abnormal analysis from being inaccurate due to only considering the outliers of the stress data, and improving the accuracy of the rock disaster early warning analysis. Furthermore, when performing early warning reminders, by considering the interval of abnormal data, the phenomenon of false warnings caused by the erroneous judgment of a single abnormal data is eliminated, thereby improving the accuracy of the early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0037] Figure 1 It is a flow chart of a method in a rock mass disaster early warning method based on cloud computing in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0039] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0040] The embodiment of the present invention discloses a rock mass disaster early warning method based on cloud computing, referring to Figure 1 , comprising steps S1 to S4:

[0041] S1: Obtain the stress time series borne by the anchor components supporting the rock mass.

[0042] It should be noted that the anchoring components are the main supporting components of the surrounding rock mass. The stress of the surrounding rock mass and the supporting force of the anchoring components offset each other, thereby preventing the rock mass from sliding. Therefore, the stress of the anchoring components can reflect the stress changes of the rock mass, and thus the possible rock mass disasters can be judged by the stress of the anchoring components supporting the rock mass.

[0043] Specifically, a stress sensor is installed on the anchoring component of the supporting rock mass, and the stress sensor is used to collect stress data every n seconds, and the data is collected Q times, where n represents a preset collection interval, and Q represents a preset number of times. The stress data collected at all times are arranged in time series to obtain a stress time series sequence. In this embodiment, n is 1 and Q is 100 as an example for description, and other embodiments may take other values, which are not specifically limited in this embodiment.

[0044] S2: Divide the stress time series into several monotonically changing data segments as candidate analysis data segments, calculate the degree of fluctuation, and the degree of fluctuation is positively correlated with the data variance and the number of data in the candidate analysis data segments; select the candidate analysis data segments whose degree of fluctuation is greater than a preset degree threshold as the analysis data segments.

[0045] It should be noted that as the mining of the ore body proceeds, each process of excavating the ore body will cause the stress of the surrounding rock body to change and quickly restore balance. This stress change of the rock body will be transmitted to the anchoring component and presented as the stress of the anchoring component. The instability and deformation of the anchoring component mainly occur in the process of rock stress change in each process of excavating the ore body, so it is necessary to analyze the stress change of the anchoring component after each process of excavating the ore body. Therefore, it is necessary to segment the data and divide the data of the stress change process of the anchoring component after each action of excavating the ore body in the stress time series into one data segment.

[0046] It should be further explained that as the mining of the ore body proceeds, more and more surrounding rock bodies without support points are generated, so the stress of this surrounding rock body will gradually increase, and this stress will be transmitted to the anchoring components, causing the stress of the anchoring components to gradually increase. At the same time, the excavation operation is a discrete process. After each ore body excavation operation, there will be a change process in which the anchoring components gradually increase and reach equilibrium. Based on this feature, the data of the stress change process of the anchoring components after each excavation action can be divided into one data segment.

[0047] S20: Divide the stress time series into a number of monotonically changing data segments and record them as candidate analysis data segments.

[0048] Preferably, as an example, the stress time series is divided into a number of monotonically changing data segments recorded as candidate analysis data segments, including:

[0049] The difference between each stress data and the previous stress data in the stress time series is calculated, and recorded as the first-order difference of each stress data; the first stress data is taken as the starting data, starting from the starting data, the first stress data with a first-order difference less than or equal to 0 is obtained and recorded as the cut-off data, and the data segment consisting of all stress data between the starting data and the cut-off data is recorded as the analysis data segment; in response to the cut-off data not being the last stress data, the next data of the cut-off data is taken as the new starting data, starting from the new starting data, the first stress data with a first-order difference less than or equal to 0 is obtained and recorded as the new cut-off data, and the data segment consisting of the stress data between the new starting data and the new cut-off data is recorded as the new analysis data segment; in response to the new cut-off data being the last stress data, all analysis data segments are obtained.

[0050] It is understandable that this segmentation process will segment the monotonically changing data segments together. Since the instability and deformation of the anchoring component mainly occur during the stress increase process, the stress data during the equilibrium process is not the research object, so the data segments during the equilibrium process need to be removed from all monotonically changing data segments.

[0051] S21: Calculate the degree of fluctuation, which is positively correlated with the data variance and the amount of data in the candidate analysis data segment.

[0052] It should be noted that the variance of the stress data in the balancing process is small, and the amount of data with monotonous changes in the balancing process is small, so the data segments in the balancing process can be removed based on this.

[0053] Preferably, as an example, calculating the degree of fluctuation includes:

[0054] The product of the number of data in the candidate analysis data segment and the variance of all data in the candidate analysis data segment is taken as the fluctuation degree of the candidate analysis data segment.

[0055] S22: Filter out candidate analysis data segments whose fluctuation degree is greater than a preset degree threshold as analysis data segments.

[0056] It should be noted that the stress data of the data segment during the balancing process changes slightly, and the amount of monotonically changing stress data in the data segment during the balancing process is small, so the fluctuation degree of the data segment during the balancing process is small.

[0057] Preferably, as an example, screening out candidate analysis data segments whose fluctuation degree is greater than a preset degree threshold as analysis data segments includes:

[0058] The candidate analysis data segments with fluctuation degree greater than the preset degree threshold are taken as analysis data segments. This embodiment is described by taking the preset degree threshold of 0.75 as an example. Other embodiments may take other values, and this embodiment does not make specific restrictions.

[0059] S3: Calculate the corrected local anomaly factor.

[0060] It should be noted that before the anchor component loses stability and deforms, the stress data it bears mainly shows that the stress data is greatly different from other stress data, the stress data fluctuates greatly, and the stress borne by the anchor component reaches its stress bearing limit. Therefore, the abnormal situation of the stress data can be analyzed based on this. This embodiment reflects the abnormal situation of the stress data through the corrected local abnormal factor.

[0061] Preferably, as an example, the modified local anomaly factor satisfies the relationship:

[0062]

[0063] in, Indicates the first The local anomaly factor of the stress data characterizes the outlier of the stress data. The larger the value, the greater the difference between the stress data and other stress data, and the greater the possibility of instability and deformation of the anchor component. Indicates the risk level of the analysis data segment. The risk level characterizes the fluctuation level of the analysis data segment. The larger the value, the greater the fluctuation level of the stress data. The greater the stress fluctuation, the greater the risk of instability and deformation of the anchor component. Indicates the maximum stress that the anchoring component can bear. Indicates the first Stress data, This value reflects whether the stress on the anchor component has reached the stress limit. The larger the value, the farther the stress on the anchor component is from its stress limit, and the risk of instability and deformation of the anchor component is smaller. Therefore, the smaller the abnormality of the stress data, Indicates the first The corrected local anomaly factor of the stress data. Represents an exponential function with a natural constant as base, which is used to normalize the data.

[0064] The above embodiments involve risk levels and local abnormal factors. The following describes the method for determining the risk levels and local abnormal factors.

[0065] S30: Calculate the risk level.

[0066] Preferably, as an example, calculating the risk level includes:

[0067] ;

[0068] in, Indicates the first Stress data, Indicates the first Stress data, It reflects the variance of the changes of all two adjacent stress data in the analysis data segment. The larger the value, the greater the difference in the changes of all two adjacent stress data. Therefore, the greater the possibility of stress mutation, the greater the possibility that stress mutation will cause instability and deformation of the anchor component, and thus the greater the risk. Indicates the number of stress data in this analysis data segment, It reflects the mean change rate of all stress data segments in the analysis data segment. The larger the value, the greater the change of all stress data in the analysis data segment. The greater the stress change, the greater the possibility of instability and deformation of the anchor component, and thus the greater the risk. Indicates the degree of risk, Represents the absolute value symbol.

[0069] S31: Obtain a local abnormality factor.

[0070] Preferably, as an example, obtaining the local abnormal factor of the stress data includes:

[0071] The LOF algorithm is used to perform anomaly detection on all stress data in the stress time series to obtain the local anomaly factor of each stress data.

[0072] It should be noted that the use of the LOF algorithm to calculate the local anomaly factor of stress data is an existing technology and will not be described in detail here.

[0073] S4: The stress data whose corrected local abnormal factor is greater than the preset abnormal threshold is regarded as abnormal data, and the warning degree is calculated to assist the early warning reminder, wherein the warning degree is negatively correlated with the time interval between adjacent abnormal data.

[0074] S40: The stress data whose corrected local abnormal factor is greater than the preset abnormal threshold is taken as abnormal data, and the warning degree is calculated.

[0075] It should be noted that, since abnormal data frequently appear before the anchoring component becomes unstable, in order to prevent inaccurate abnormality determination caused by relying on only one abnormal data, the frequency of occurrence of abnormal data needs to be analyzed.

[0076] Preferably, as an example, the stress data whose corrected local abnormal factor is greater than a preset abnormal threshold is taken as abnormal data, and the warning degree is calculated, including:

[0077] The corrected local anomaly factor is normalized using the maximum and minimum method to obtain the normalized corrected local anomaly factor. For ease of description, the normalized corrected local anomaly factor will still be referred to as the corrected local anomaly factor in the following.

[0078] The stress data whose corrected local abnormal factor is greater than the preset abnormal threshold is regarded as abnormal data.

[0079] In this embodiment, the preset abnormal threshold value is 0.75 for description. Other embodiments may take other values, and this embodiment does not make any specific limitation.

[0080] Calculate the warning level threshold:

[0081]

[0082] in, Indicates the stress time series The time of abnormal data, Indicates the stress time series The time of abnormal data, It reflects the time interval between two adjacent abnormal data. The larger the value, the smaller the frequency of abnormal data, and the smaller the risk of instability and deformation of the anchor component. Indicates the number of abnormal data in the stress time series. represents an exponential function with a natural constant as the base, which is used for normalization. Indicates the warning level.

[0083] S41: To assist in early warning reminders.

[0084] Preferably, as an example, to assist in early warning reminders, including:

[0085] If the warning level is greater than the preset warning level threshold, a warning reminder is issued; if the warning level is less than or equal to the preset warning level threshold, no warning reminder is issued. This embodiment takes the preset warning level of 0.8 as an example for description, and other embodiments may take other values, which are not specifically limited in this embodiment.

[0086] The embodiment of the present invention also discloses a rock disaster warning system based on cloud computing, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a rock disaster warning method based on cloud computing according to the present invention is implemented.

[0087] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0088] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, an enhanced dynamic random access memory, a high bandwidth memory, a hybrid storage cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device.

[0089] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0090] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A rock mass disaster early warning method based on cloud computing, characterized in that: Includes steps: Obtaining the time series of stresses borne by the anchoring components supporting the rock mass; The stress time series is divided into several monotonically changing data segments as candidate analysis data segments, and the degree of fluctuation is calculated. The degree of fluctuation is positively correlated with the data variance and the number of data in the candidate analysis data segments; and the candidate analysis data segments with a degree of fluctuation greater than a preset degree threshold are selected as analysis data segments; Calculate the corrected local anomaly factor: L0 i represents the local abnormal factor of the i-th stress data in the analysis data segment, the local abnormal factor characterizes the outlier of the stress data, F represents the risk level of the analysis data segment, the risk level characterizes the fluctuation level of the analysis data segment, T represents the maximum stress that the anchor component can bear, Y i Indicates the i-th stress data in the analysis data segment, L1 i represents the corrected local anomaly factor of the i-th stress data in the analysis data segment; The stress data whose corrected local abnormal factor is greater than the preset abnormal threshold is regarded as abnormal data, and the warning degree is calculated to assist in early warning reminder, wherein the warning degree is negatively correlated with the time interval between adjacent abnormal data; The calculation of the warning level includes: Among them, t j+1 Indicates the time of the j+1th abnormal data in the stress time series, t j It represents the time of the jth abnormal data in the stress time series, M represents the number of abnormal data in the stress time series, exp() represents the exponential function with natural constant as the base, and S represents the warning degree.

2. The rock mass disaster early warning method based on cloud computing according to claim 1 is characterized in that: The method for obtaining the risk level includes: Among them, Y i+1 It represents the i+1th stress data in an analysis data segment, N represents the number of stress data in the analysis data segment, and || represents the absolute value sign.

3. The rock mass disaster early warning method based on cloud computing according to claim 1 is characterized in that: The method for obtaining the candidate analysis data segment comprises: The difference between each stress data and the previous stress data in the stress time series is calculated, and recorded as the first-order difference of each stress data; the first stress data is taken as the starting data, starting from the starting data, the first stress data with a first-order difference less than or equal to 0 is obtained and recorded as the cut-off data, and the data segment consisting of all stress data between the starting data and the cut-off data is recorded as the analysis data segment; in response to the cut-off data not being the last stress data, the next data of the cut-off data is taken as the new starting data, starting from the new starting data, the first stress data with a first-order difference less than or equal to 0 is obtained and recorded as the new cut-off data, and the data segment consisting of the stress data between the new starting data and the new cut-off data is recorded as the new analysis data segment; in response to the new cut-off data being the last stress data, all analysis data segments are obtained.

4. The rock mass disaster early warning method based on cloud computing according to claim 1 is characterized in that: The calculation of the degree of fluctuation includes: The product of the number of data in the candidate analysis data segment and the variance of all data in the candidate analysis data segment is taken as the fluctuation degree of the candidate analysis data segment.

5. The rock mass disaster early warning method based on cloud computing according to claim 1, characterized in that: The method for obtaining the local abnormal factor includes: The LOF algorithm is used to perform anomaly detection on all stress data in the stress time series to obtain the local anomaly factor of each stress data.

6. The rock mass disaster early warning method based on cloud computing according to claim 1, characterized in that: The auxiliary early warning reminder includes: If the warning level is greater than the preset warning level threshold, a warning reminder will be issued; if the warning level is less than or equal to the preset warning level threshold, no warning reminder will be issued.

7. The rock mass disaster early warning system based on cloud computing is characterized by: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a cloud computing-based rock disaster early warning method according to any one of claims 1 to 6 is implemented.