Geological disaster landslide monitoring and anti-theft monitoring method and system

Abstract

The invention discloses a geological disaster landslide monitoring and anti-theft monitoring method and system, and relates to the technical field of geological disaster monitoring, and the method comprises the steps: obtaining the inclination angle information, acceleration information, attitude information and position information of the system; constructing a first feature vector based on the inclination angle information, the acceleration information, the attitude information and the position information; inputting the first feature vector into a trained first logistic regression model to obtain a state monitoring result of the system; the state monitoring result comprises any one of the following items: a normal state, a state in a landslide process and a stolen state. According to the invention, the technical problem that the existing automatic detection early warning equipment has the risk of being stolen is relieved.

Description

Technical Field

[0001] The present invention relates to the technical field of geological disaster monitoring, and in particular to a method and system for geological disaster landslide monitoring and anti-theft monitoring. Background Art

[0002] Since the Ministry of Natural Resources launched the "Geological Disaster Monitoring and Early Warning Experiment" in 2020, more than 56,000 potential risk points across the country have been installed with 290,000 sets of automated monitoring and early warning equipment. Automated monitoring and early warning equipment has played an important role in disaster prevention and mitigation. It is understood that in the multiple heavy rainfall processes since 2024, automated monitoring has successfully warned and forecasted 68 disaster risks, effectively avoiding 216 possible casualties. According to statistics, nearly half of these dangerous situations occurred at night, and landslides induced by short-term heavy rainfall are often sudden, leaving a very limited time window for early warning and evacuation. In this context, automated monitoring equipment has demonstrated its unique advantages of "all-weather, quasi-real-time", and effectively coordinated with group monitoring and group prevention.

[0003] Automated monitoring and early warning equipment is often installed on natural slopes around human activities or artificial slopes caused by self-built houses, that is, "in front of and behind houses", which are within the reach of ordinary people. Although the equipment has played a great role in geological disaster early warning, it is easy to be stolen by criminals due to the problem of installation location. The loss of each automated monitoring equipment means that the monitoring point is at risk of being out of control, and it also means that the lives and property of residents under the landslide are unprotected. Slope safety monitoring equipment can monitor the safety status data of the slope in real time. After the equipment is lost, it is difficult to obtain these data and accurately assess the stability of the slope. Due to the lack of equipment monitoring data, the dangerous condition of the slope cannot be discovered in advance. Once a disaster such as landslide or collapse occurs on the slope, it may cause greater losses to surrounding buildings, roads, personnel, etc. The purchase, installation, and commissioning of these monitoring equipment all require certain expenses. Repurchasing and installing equipment will increase costs, and additional safety measures may be required due to the monitoring gaps during the period when the equipment is lost, such as increasing manual patrols, which will also bring additional expenses.

[0004] The existing automatic monitoring and early warning equipment has the following problems in terms of anti-theft:

[0005] Some equipment uses an external power supply system. The thieves disconnected the external power supply system, causing the equipment to lose power and shut down, and the data to become invalid.

[0006] Some devices use a switch design, which is easy to find and turn off, shutting down the device and invalidating the data;

[0007] Since disaster monitoring equipment requires a long service life, data collection and uploading frequency is 1-2 hours / time. After theft and damage, it is impossible to send information in time;

[0008] Some equipment has built-in positioning function, but the moving distance range is large when landslide occurs, especially for large landslides, the sliding distance exceeds more than ten meters or even hundreds of meters, and setting up electronic fences cannot effectively determine;

[0009] Landslide monitoring is basically deployed on slopes near residential areas. Animal and human activities may falsely trigger data changes. At the same time, natural deformation of landslides will also cause data changes. It is challenging to identify data information and ensure the accuracy of monitoring data and anti-theft alarm data. Summary of the invention

[0010] In order to solve the above technical problems existing in the prior art, the embodiment of the present invention provides a method and system for monitoring landslides and preventing theft caused by geological disasters. The technical solution is as follows:

[0011] On the one hand, a geological disaster landslide monitoring and anti-theft monitoring method is provided, which is applied to a geological disaster landslide monitoring and anti-theft monitoring system; the method includes: obtaining the inclination angle information, acceleration information, posture information and position information of the system; constructing a first feature vector based on the inclination angle information, the acceleration information, the posture information and the position information; inputting the first feature vector into a trained first logistic regression model to obtain a status monitoring result of the system; the status monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being in a stolen state.

[0012] Furthermore, the method further comprises: if the state monitoring result is the state of being in a landslide process, issuing a landslide alarm message; if the state monitoring result is the state of being stolen, issuing a theft alarm message.

[0013] Furthermore, the first feature vector includes x1=[θ, a, p, δ]; wherein θ is the change in tilt angle, a is the change in acceleration, p is the change in posture, and δ is the change in position.

[0014] Furthermore, the status monitoring result also includes: equipment damage status, and the status of being disturbed by animals or pedestrians.

[0015] Furthermore, the method further includes: training the first logistic regression model based on a first preset training set to obtain the trained first logistic regression model.

[0016] Furthermore, the first logistic regression model includes:

[0017]

[0018] In the formula, The input feature vector x belongs to category C k The probability of w k is corresponding to category C k The weight vector, b k is corresponding to category C k The bias term of , k is the category index;

[0019] The loss function for training the first logistic regression model includes:

[0020]

[0021] Where W is the set of all weight vectors, b is the set of all bias terms, N is the number of training samples, and y ik is an indicator variable.

[0022] On the other hand, a geological disaster landslide monitoring and anti-theft monitoring method is also provided, which is applied to a geological disaster landslide monitoring and anti-theft monitoring system; the method includes: obtaining the system's inclination angle information, acceleration information, posture information, position information and rainfall information of the environment where the system is located; constructing a second feature vector based on the inclination angle information, the acceleration information, the posture information, the position information and the rainfall information; inputting the second feature vector into a trained second logistic regression model to obtain a status monitoring result of the system; the status monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being in a stolen state.

[0023] Furthermore, the second characteristic vector includes x2=[θ, a, p, δ, R]; wherein θ is the change in tilt angle, a is the change in acceleration, p is the change in posture, δ is the change in position, and R is the rainfall level.

[0024] On the other hand, a geological disaster landslide monitoring and anti-theft monitoring system is also provided, including: a sensor module, a building module and a monitoring module; wherein the sensor module is used to obtain the inclination angle information, acceleration information, posture information and position information of the system; the building module is used to construct a first feature vector based on the inclination angle information, the acceleration information, the posture information and the position information; the monitoring module is used to input the first feature vector into a trained first logistic regression model to obtain a status monitoring result of the system; the status monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being in a stolen state.

[0025] Furthermore, the sensor module is also used to obtain rainfall information of the environment where the system is located; the construction module is also used to construct a second feature vector based on the tilt angle information, the acceleration information, the posture information, the position information and the rainfall information; the monitoring module is also used to input the second feature vector into a trained second logistic regression model to obtain a status monitoring result of the system.

[0026] The embodiment of the present invention provides a geological disaster landslide monitoring and anti-theft monitoring method and system, which adopts integrated monitoring of accelerometers, attitude sensors and positioning sensors, combines rainfall sensors to monitor rainfall levels, and uses logistic regression to simultaneously perform landslide disaster prediction and anti-theft monitoring, thereby alleviating the technical problem that existing automated detection and early warning equipment has the risk of theft. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0028] Figure 1 It is a flow chart of a geological disaster landslide monitoring and anti-theft monitoring method provided by an embodiment of the present invention;

[0029] Figure 2 It is a flow chart of another geological disaster landslide monitoring and anti-theft monitoring method provided by an embodiment of the present invention;

[0030] Figure 3 It is a schematic diagram of a geological disaster landslide monitoring and anti-theft monitoring system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0032] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0033] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0034] Embodiment 1

[0035] Figure 1 1 is a flow chart of a geological disaster landslide monitoring and anti-theft monitoring method provided according to an embodiment of the present invention, and the method is applied to a geological disaster landslide monitoring and anti-theft monitoring system. Figure 1 As shown, the method specifically comprises the following steps:

[0036] Step S102, obtaining the system's tilt angle information, acceleration information, attitude information and position information.

[0037] Step S104: construct a first feature vector based on the tilt angle information, acceleration information, posture information and position information.

[0038] Step S106, inputting the first feature vector into the trained first logistic regression model to obtain a system status monitoring result; the status monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being stolen.

[0039] Specifically, if the status monitoring result is a state of landslide, a landslide alarm message is issued; if the status monitoring result is a stolen state, a theft alarm message is issued.

[0040] In an optional implementation provided by an embodiment of the present invention, the status monitoring result further includes: a device damage status, and a device being disturbed by animals or pedestrians.

[0041] In an optional implementation provided by an embodiment of the present invention, the first eigenvector includes x1=[θ, a, p, δ]; wherein θ is the change in tilt angle, a is the change in acceleration, p is the change in posture, and δ is the change in position.

[0042] Specifically, the method provided in the embodiment of the present invention further includes a training step, including:

[0043] The first logistic regression model is trained based on the first preset training set to obtain a trained first logistic regression model.

[0044] The following is an analysis and summary of the data characteristics of the four indicators of the system: tilt angle information, acceleration information, posture information, and position information, which have formed sample characteristics:

[0045] (1) The state of landslide:

[0046] (a) Tilt angle information:

[0047] Characteristics: Continuous and gradually increasing angle changes, which may be followed by brief periods of stability.

[0048] Data sample characteristics: The angle gradually deviates from the initial set value, and the rate of change may be slow at the beginning and accelerate later.

[0049] (b) Acceleration information:

[0050] Characteristics: Small acceleration changes may be detected in the early stage, followed by significant acceleration changes when the landslide occurs.

[0051] Data sample characteristics: The acceleration data increases first and then decreases or fluctuates periodically, and the direction may be consistent with the direction of gravity.

[0052] (c) Posture information:

[0053] Characteristics: The device's posture changes significantly, possibly accompanied by rotation or tumbling.

[0054] Data sample characteristics: The change trend of posture angle is consistent with that of tilt angle, but it provides a specific direction in three-dimensional space.

[0055] (d) Location information:

[0056] Characteristics: A relatively large displacement of the device position occurs.

[0057] Data sample characteristics: The displacement distance and direction are consistent with the direction of landslide movement, and the displacement speed increases rapidly when the landslide occurs.

[0058] (2) In stolen state:

[0059] (a) Tilt angle information:

[0060] Characteristics: Rapid changes in angle, possibly accompanied by sudden tilt or inversion.

[0061] Data sample characteristics: The angle changes greatly and quickly, which is significantly different from the change pattern during normal use.

[0062] (b) Acceleration information:

[0063] Features: During the theft process, dramatic acceleration changes may be detected, such as acceleration during shaking or transportation.

[0064] Data sample characteristics: The acceleration data shows irregular and violent fluctuations and may contain acceleration components in multiple directions.

[0065] (c) Posture information:

[0066] Characteristics: The device posture may change rapidly and irregularly.

[0067] Data sample characteristics: Posture data shows frequent and discontinuous changes, which contrasts with the stable posture during normal use.

[0068] (d) Location information:

[0069] Characteristics: Rapid and discontinuous changes in equipment position that may extend beyond the landslide.

[0070] Data sample characteristics: The position data jumps, and the displacement speed and direction are significantly different from those in normal use.

[0071] (3) Equipment damage status:

[0072] (a) Tilt angle information:

[0073] Characteristics: The angle may change suddenly and then remain constant, or may fluctuate in small, irregular amplitudes.

[0074] Data sample characteristics: After the angle changes, it may remain at a fixed value for a long time, or there may be small random fluctuations.

[0075] (b) Acceleration information:

[0076] Characteristics: The acceleration data may suddenly fluctuate violently and then disappear, or continue to fluctuate slightly abnormally.

[0077] Data sample characteristics: Acceleration data may present a sudden large value, then return to zero or show irregular small fluctuations.

[0078] (c) Posture information:

[0079] Characteristics: The device posture may change dramatically and then remain unchanged, or become completely unresponsive.

[0080] Data sample characteristics: Posture data may suddenly change to a fixed value, or remain unchanged.

[0081] (d) Location information:

[0082] Characteristics: Location information may remain constant, but if the device is moved, sudden changes in location data may occur.

[0083] Data sample characteristics: Position data may remain unchanged for a long time, or may have displacements that are inconsistent with those before the destruction.

[0084] (4) Being disturbed by animals or pedestrians:

[0085] (a) Tilt angle information:

[0086] Characteristics: Small, brief changes in angle may occur.

[0087] Data sample characteristics: The angle changes in small amplitude and short duration, and quickly returns to the initial state.

[0088] (b) Acceleration information:

[0089] Characteristics: Acceleration data may fluctuate slightly, but usually not too drastic.

[0090] Data sample characteristics: Acceleration data shows small, short-term changes, usually within the normal range.

[0091] (c) Posture information:

[0092] Features: The device posture may change slightly, but remains stable overall.

[0093] Data sample characteristics: Posture data show small fluctuations, and the change pattern is related to animal or pedestrian activities.

[0094] (d) Location information:

[0095] Characteristics: The location of the device does not usually change significantly.

[0096] Data sample characteristics: The location data remains basically unchanged or changes very slightly.

[0097] In order to distinguish the four situations, a logistic regression model is used to predict the category probability by combining features with weights and applying the sigmoid function. The following is the discrimination formula combining tilt angle information, acceleration information, posture information, and position information:

[0098] First, define the first eigenvector x1=[θ,a,p,δ], where each feature represents:

[0099] θ: change in tilt angle;

[0100] a: the change in acceleration;

[0101] p: the amount of change in posture (may be a direction or angle indicating the change in posture);

[0102] δ: the change in position;

[0103] Then, define the weight vector w=[w θ ,w a ,w p ,w δ ] and bias term b. The logistic regression model predicts the probability P(C k |x) as follows:

[0104]

[0105] in: The input feature vector x belongs to category C k The probability of w k is corresponding to category Ck The weight vector, b k is corresponding to category C k The bias term, T represents the transposed vector, k is the category index, and the value range is 1 to 4, corresponding to the normal state, the state in the process of landslide, the stolen state, the equipment damaged state, and the state disturbed by animals or pedestrians.

[0106] In order to obtain the discriminant formula for each category, the present invention performs a k There is a weight vector w k and a bias term b k During training, these parameters are learned by optimizing the loss function:

[0107]

[0108] Where: W is the set of all weight vectors, b is the set of all bias terms, N is the number of training samples, y ik is an indicator variable. If sample i belongs to category C k , then y ik =1, otherwise y ik =0.

[0109] After training is completed, use the following formula to predict the category of the new sample x, and select the category with the highest probability as the prediction result:

[0110]

[0111] Embodiment 2

[0112] Figure 2 1 is a flow chart of another geological disaster landslide monitoring and anti-theft monitoring method provided according to an embodiment of the present invention, the method is applied to a geological disaster landslide monitoring and anti-theft monitoring system. Figure 2 As shown, the method specifically comprises the following steps:

[0113] Step S202, obtaining the system's tilt angle information, acceleration information, posture information, position information, and rainfall information of the environment where the system is located.

[0114] Step S204: construct a second feature vector based on the tilt angle information, acceleration information, posture information, position information and rainfall information.

[0115] Step S206, input the second feature vector into the trained second logistic regression model to obtain the system status monitoring result; the status monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being stolen.

[0116] In an optional implementation provided by an embodiment of the present invention, the status monitoring result further includes: a device damage status, and a device being disturbed by animals or pedestrians.

[0117] In an optional implementation provided by an embodiment of the present invention, the second eigenvector includes x2=[θ,a,p,δ,R]; wherein θ is the change in tilt angle, a is the change in acceleration, p is the change in posture, δ is the change in position, and R is the rainfall level.

[0118] Specifically, the method provided in the embodiment of the present invention further includes a training step, including:

[0119] The second logistic regression model is trained based on the second preset training set to obtain a trained second logistic regression model.

[0120] In an optional implementation provided by an embodiment of the present invention, rainfall is the main cause of landslide disasters. Rainfall is divided into light rain, moderate rain, heavy rain, and rainstorm. When heavy to rainstorm occurs, thieves will not go to the scene to steal equipment for their own safety. However, the possibility of landslide data may increase, and the possibility of animal interference or gravel rolling is also increased. In order to further improve the judgment result, the rainfall data (obtained through the network or collected by the rainfall sensor data) is used as a parameter and input into the device by the platform for calculation. According to actual application analysis, rainfall has a greater impact on the landslide process and animal or pedestrian interference, but has a smaller impact on equipment theft and equipment damage.

[0121] The rainfall level is represented by a discrete variable R, which is divided into one of the following four levels, namely light rain R1, moderate rain R2, heavy rain R3, and torrential rain R4. The rainfall level is added as an additional feature to the feature vector x to form a new second feature vector x2=[θ,a,p,δ,R].

[0122] Then, modify the discriminant formula to take into account the rainfall level R:

[0123]

[0124] in:

[0125] w k and b k is corresponding to category C k The weight vector and bias term of .

[0126] x is the second eigenvector containing the tilt angle, acceleration, attitude, position change, and rain level.

[0127] During training, the loss function needs to be updated to take into account the new features R:

[0128]

[0129] In this way, the model will learn the impact of rainfall level R on different categories and will be able to make more accurate predictions based on this information.

[0130] Embodiment 3

[0131] Figure 3 Schematic diagram of a geological disaster landslide monitoring and anti-theft monitoring system provided according to an embodiment of the present invention. Figure 3 As shown, the system includes: a sensor module 10 , a construction module 20 and a monitoring module 30 .

[0132] Specifically, the sensor module 10 is used to obtain the tilt angle information, acceleration information, attitude information and position information of the system;

[0133] A construction module 20, configured to construct a first feature vector based on the tilt angle information, the acceleration information, the posture information and the position information;

[0134] The monitoring module 30 is used to input the first feature vector into the trained first logistic regression model to obtain the system status monitoring result; the status monitoring result includes any of the following: being in a normal state, being in a landslide process, being stolen, being damaged, or being disturbed by animals or pedestrians.

[0135] In an optional implementation provided by the embodiment of the present invention, the sensor module 10 is also used to obtain rainfall information of the environment where the system is located;

[0136] The construction module 20 is further used to construct a second feature vector based on the tilt angle information, the acceleration information, the posture information, the position information and the rainfall information;

[0137] The monitoring module 30 is further used to input the second feature vector into the trained second logistic regression model to obtain a system status monitoring result.

[0138] Optionally, the sensor module 10 includes: an accelerometer 11 , a posture sensor 12 , a positioning sensor 13 and a rain sensor 14 .

[0139] Specifically, the accelerometer 11 is used to obtain acceleration information of the system;

[0140] The attitude sensor 12 is used to obtain the tilt angle information and attitude information of the system;

[0141] The positioning sensor 13 is used to obtain the position information of the system.

[0142] The rainfall sensor 14 is used to obtain rainfall information of the environment where the system is located.

[0143] In an optional implementation provided by an embodiment of the present invention, the overall structure of the system provided by the present invention adopts an integrated and miniaturized design, a built-in lithium battery power supply system, supports external charging, and wireless data transmission functions. The device is integrated with high-sensitivity, low-power sensors such as MEMS accelerometers, gyroscopes (i.e., attitude sensors), and Beidou positioning (i.e., positioning sensors). The system can be used independently as a monitoring device for indicators such as tilt angle, tilt direction, and acceleration during landslide monitoring, and can also be used as a monitoring and anti-theft module, integrated into other sensor devices. The system can work for more than 5 years without external charging conditions. A hidden magnetic switch (i.e., software switch) is designed to avoid shutting down after cutting off the external power supply and shutting down by manual switch. The external data interaction interface of the device is used to output anti-theft data and monitoring data.

[0144] When used independently, the device is an integrated monitoring sensor for inclination, acceleration, and attitude, and uses wireless transmission to send data directly to the cloud platform. When used as a module, it supports data transmission to other sensors through an external data interface. After being integrated, other sensors can receive and configure the device data through an external data interaction interface. After detecting a landslide disaster or theft, emergency data is sent independently to the cloud platform.

[0145] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0146] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0147] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0148] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0149] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0150] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A geological disaster landslide monitoring and anti-theft monitoring method, characterized in that: Applied to geological disaster landslide monitoring and anti-theft monitoring system; the method includes: Obtaining tilt angle information, acceleration information, attitude information and position information of the system; Constructing a first feature vector based on the tilt angle information, the acceleration information, the posture information and the position information; The first feature vector is input into a trained first logistic regression model to obtain a state monitoring result of the system; the state monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being stolen.

2. The method according to claim 1, characterized in that The method further comprises: if the state monitoring result is the state of being in the landslide process, issuing a landslide warning message; If the status monitoring result is that the device is in the stolen state, a theft alarm message is issued.

3. The method according to claim 1, characterized in that The first feature vector includes x1=[θ, a, p, δ], wherein θ is the change in tilt angle, a is the change in acceleration, p is the change in posture, and δ is the change in position.

4. The method according to claim 1, characterized in that: The state monitoring results also include: equipment damage status, and the state of being disturbed by animals or pedestrians.

5. The method according to claim 1, characterized in that The method further includes: training the first logistic regression model based on a first preset training set to obtain the trained first logistic regression model.

6. The method according to claim 5, characterized in that The first logistic regression model includes: ； In the formula, The input feature vector x belongs to category C k The probability of w k is corresponding to category C k The weight vector, b k is corresponding to category C k The bias term of , k is the category index; The loss function for training the first logistic regression model includes: ； Where W is the set of all weight vectors, b is the set of all bias terms, N is the number of training samples, and y ik is an indicator variable.

7. A geological disaster landslide monitoring and anti-theft monitoring method, characterized in that: Applied to geological disaster landslide monitoring and anti-theft monitoring system; the method includes: Obtaining tilt angle information, acceleration information, attitude information, position information of the system and rainfall information of the environment where the system is located; constructing a second feature vector based on the tilt angle information, the acceleration information, the posture information, the position information and the rainfall information; The second feature vector is input into the trained second logistic regression model to obtain the state monitoring result of the system; the state monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being stolen.

8. The method according to claim 7, characterized in that The second characteristic vector includes x2=[θ, a, p, δ, R]; wherein θ is the change in tilt angle, a is the change in acceleration, p is the change in posture, δ is the change in position, and R is the rainfall level.

9. A geological disaster landslide monitoring and anti-theft monitoring system, characterized in that: include: Sensor module, building module and monitoring module; among them, The sensor module is used to obtain the tilt angle information, acceleration information, attitude information and position information of the system; The construction module is used to construct a first feature vector based on the tilt angle information, the acceleration information, the posture information and the position information; The monitoring module is used to input the first feature vector into a trained first logistic regression model to obtain a status monitoring result of the system; the status monitoring result includes any one of the following: being in a normal state, being in a landslide process, and being stolen.

10. The system according to claim 9, characterized in that The sensor module is also used to obtain rainfall information of the environment where the system is located; The construction module is further used to construct a second feature vector based on the tilt angle information, the acceleration information, the posture information, the position information and the rainfall information; The monitoring module is further used to input the second feature vector into a trained second logistic regression model to obtain a status monitoring result of the system.

Images