A monitoring system for terrain deformation in the shallow subsurface natural gas hydrate area of the seabed

Abstract

The present invention discloses a system for monitoring topographic deformation in a shallow seabed natural gas hydrate region, relating to the technical field of seabed environmental monitoring. The system comprises a three-dimensional monitoring network, an underwater control center, a remote transmission unit, and an offshore intelligent prediction center. The three-dimensional monitoring network is used to collect monitoring array data; the underwater control center is used to obtain monitoring array data and output monitoring array operation instructions; the remote transmission unit is used to transmit the monitoring array data sent by the underwater control center to the offshore intelligent prediction center; and the offshore intelligent prediction center is used to reconstruct the topography of the shallow seabed natural gas hydrate region and predict the topographic trend of the shallow seabed natural gas hydrate region based on the monitoring array data. The present invention can monitor the shallow seabed natural gas hydrate mining environment.

Description

Technical Field

[0001] The present invention relates to the technical field of seabed environment monitoring, and in particular to a topographic deformation monitoring system for a shallow seabed natural gas hydrate area. Background Art

[0002] Natural gas hydrates, formed by methane and water molecules at low temperatures and high pressures, are a type of cage-like crystalline hydrate with broad application prospects. The presence of natural gas hydrates in shallow seabed layers is of great significance for the extraction of high-purity, bulk natural gas hydrates.

[0003] While natural gas hydrates offer the promise of new energy, they also pose a serious challenge to the human environment. The formation of shallow seafloor natural gas hydrates requires a high flux of methane gas. Furthermore, due to their close contact with seawater, these hydrates are susceptible to seawater temperature and pressure. When methane gas flux is high, natural gas hydrates form and aggregate near the seafloor, forming massive masses. This causes the volume of seafloor sediments to expand, forming shallow seafloor natural gas hydrate mounds. However, due to the influence of seawater, exposed seafloor natural gas hydrates are in a critical chemical stability state and are susceptible to decomposition. Decomposition alters the physical properties of sediments, significantly reducing their shear strength and softening the seafloor. This can lead to large-scale landslides, earthquakes, and collapses, and even threaten the safety of marine engineering facilities.

[0004] However, due to the complex physical and mechanical properties of sediments in the shallow seabed natural gas hydrate area, such as heterogeneity, anisotropy, structure, elastic-viscoplasticity, and other characteristics, there is a problem of complex and uneven deformation patterns. Therefore, it is difficult to capture and monitor the tiny deformation of sediments in the shallow seabed natural gas hydrate area. In addition, the massive data obtained by the in-situ monitoring system based on the sensor network has the characteristics of multi-source, multi-scale, dynamic, and spatiotemporal distribution. Moreover, the topography and formation deformation and soil mechanics parameters have complex nonlinear relationships. Therefore, the development of an in-situ monitoring system for topography and formation deformation in the shallow seabed natural gas hydrate area and an efficient prediction model is an inevitable requirement for realizing the monitoring and early warning of the shallow seabed natural gas hydrate mining environment. Summary of the Invention

[0005] The purpose of the present invention is to provide a system for monitoring the topographic deformation of a shallow seabed natural gas hydrate region, which can monitor the mining environment of shallow seabed natural gas hydrates.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A system for monitoring terrain deformation in a shallow seabed natural gas hydrate area, comprising: a three-dimensional monitoring network, an underwater control center, a remote transmission unit, and an offshore intelligent prediction center;

[0008] The three-dimensional monitoring network includes multiple monitoring arrays, namely horizontal arrays and vertical arrays; the horizontal arrays are laid parallel to the surface of the shallow seabed, and the vertical arrays are laid into the shallow seabed. The monitoring arrays are networked and connected by multiple monitoring nodes; the three-dimensional monitoring network is used to obtain monitoring array data; the monitoring array data is the coordinate information of the monitoring nodes and the environment in which the monitoring nodes are located;

[0009] The underwater control center is used to:

[0010] Acquiring the monitoring array data;

[0011] Output monitoring array working instructions; the monitoring array working instructions are used to achieve time synchronization of the monitoring array and control the working mode of the monitoring array;

[0012] The remote transmission unit is used to transmit the monitoring array data sent by the underwater control center to the offshore intelligent prediction center;

[0013] The offshore intelligent prediction center is used to reconstruct the topography of the shallow seabed natural gas hydrate area and predict the topography trend of the shallow seabed natural gas hydrate area based on the monitoring array data.

[0014] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0015] The present invention provides a system for monitoring deformation of the terrain in the shallow seabed natural gas hydrate area, comprising: a three-dimensional monitoring network, an underwater control center, a remote transmission unit, and an offshore intelligent prediction center. By means of the mutual cooperation of the three-dimensional monitoring network, the underwater control center, the remote transmission unit, and the offshore intelligent prediction center, the terrain of the shallow seabed natural gas hydrate area is reconstructed and the terrain trend of the shallow seabed natural gas hydrate area is predicted, thereby realizing the monitoring of the shallow seabed natural gas hydrate mining environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in 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 paying any creative work.

[0017] Figure 1This is a structural block diagram of a system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to the present invention;

[0018] Figure 2 This is a flow chart of the control of the monitoring array by the underwater control center of the present invention;

[0019] Figure 3 This is a workflow diagram of the data preprocessing module of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Example 1

[0023] This embodiment provides a system for monitoring terrain deformation in a shallow seabed natural gas hydrate area. The system includes a three-dimensional monitoring network, an underwater control center, a remote transmission unit, and an offshore intelligent prediction center.

[0024] The three-dimensional monitoring network includes multiple monitoring arrays, namely horizontal arrays and vertical arrays; the horizontal arrays are laid parallel to the surface of the shallow seabed, and the vertical arrays are laid into the shallow seabed. The monitoring arrays are networked and connected by multiple monitoring nodes; the three-dimensional monitoring network is used to obtain monitoring array data; the monitoring array data is the coordinate information of the monitoring nodes and the environmental information of the monitoring nodes.

[0025] The underwater control center is used to:

[0026] Acquiring the monitoring array data;

[0027] Output monitoring array working instructions; the monitoring array working instructions are used to achieve time synchronization of the monitoring array and control the working mode of the monitoring array.

[0028] The remote transmission unit is used to transmit the monitoring array data sent by the underwater control center to the offshore intelligent prediction center.

[0029] The offshore intelligent prediction center is used to reconstruct the topography of the shallow seabed natural gas hydrate area and predict the topography trend of the shallow seabed natural gas hydrate area based on the monitoring array data.

[0030] In this embodiment, the monitoring node includes two types of data monitoring elements: a deformation monitoring unit and a non-deformation monitoring unit. The deformation monitoring unit includes a three-axis accelerometer, a three-axis gyroscope, and a three-axis electronic compass, and is used to obtain the monitoring node's coordinate information. The non-deformation monitoring unit includes a pressure sensor, a temperature sensor, and a methane concentration sensor, and is used to obtain environmental information about the monitoring node to predict changes in the monitoring node's coordinates.

[0031] The monitoring nodes are connected to the communication network via the RS485 bus, and the communication lines and power supply lines are wrapped with watertight materials to achieve interaction with the underwater control center.

[0032] The operating modes of the monitoring array are divided into automatic high-frequency acquisition mode, automatic low-frequency acquisition mode, manual acquisition mode and sleep mode. Among them, the automatic high-frequency acquisition mode and automatic low-frequency acquisition mode are suitable for long-term monitoring of the terrain in the shallow natural gas hydrate area on the seabed. The manual acquisition mode is suitable for targeted monitoring during scientific expeditions. The sleep mode is used for data transmission and monitoring array time synchronization.

[0033] In this embodiment, the underwater control center is used to provide power to the monitoring array, achieve time synchronization of the monitoring array, control the working mode of the monitoring array, and store real-time data. The underwater control center uses an SD card to store the data collected by the monitoring array and uses the FatFs file system to manage the data collected by the monitoring array. A separate file is created for the daily status monitoring data of each monitoring node to facilitate data search and management. The control process of the monitoring array by the underwater control center is as follows: Figure 2 shown.

[0034] In this embodiment, the monitoring array is powered by a soft-pack lithium-ion battery to withstand the high-pressure environment of shallow subsea natural gas hydrate areas. An RTC chip is used to synchronize the monitoring array's clock. Preferably, a high-performance, low-power real-time clock circuit with RAM, based on the DS1302, can be used.

[0035] In this embodiment, the remote transmission unit can transmit the data acquired by the monitoring array to the offshore intelligent prediction center through the underwater control center, and the offshore intelligent prediction center is used to run the terrain deformation prediction model of the shallow natural gas hydrate area on the seabed.

[0036] The remote transmission unit is connected to the control circuit of the underwater control center via a watertight cable (the first line), and exchanges instructions and data via the RS485 bus. The remote transmission unit exchanges information with the offshore intelligent prediction center via wireless communication (the third line), which can be satellite communication or shortwave communication.

[0037] The remote transmission center includes a first relay and a second relay; the remote transmission unit includes three lines, wherein the first line is a watertight cable directly connecting the first relay and the underwater control center, the second line is an umbilical cable connecting the first relay and the second relay, and the third line is wireless communication for information exchange between the second relay and the offshore intelligent prediction center.

[0038] The first relay is connected to the control circuit of the underwater control center via a first line, and exchanges commands and data via the RS485 bus. The first relay is connected to the second relay via a second line, and the second relay can send commands to the first relay and receive data back from the first relay via the Modbus protocol. The second relay is connected to the offshore intelligent prediction center via a third line, which can be satellite communication or shortwave communication, to exchange information between the two modules.

[0039] The first relay is an underwater acoustic communication device or an underwater optical communication device; the function of the first relay is to make no physical connection between the first line and the second line, so as to ensure that the umbilical cable does not drag the monitoring array.

[0040] Furthermore, the first relay is a pair of underwater acoustic communicators that can realize full-duplex communication. The underwater acoustic communicator is connected to the underwater control center through a first line, and the underwater acoustic communicator is connected to the second relay through a second line.

[0041] The second relay is a scientific research vessel or a floating base station on the sea surface, which can send data to the offshore intelligent prediction center through a third line.

[0042] In this embodiment, a target prediction model is provided in the offshore intelligent prediction center; the target prediction model is a terrain deformation prediction model for the shallow seabed natural gas hydrate area.

[0043] The target prediction model is used to:

[0044] The pre-processed monitoring array data is processed to obtain trend component matrix, spatial component matrix and environmental component matrix.

[0045] The topography of the shallow seabed natural gas hydrate area is reconstructed according to the spatial component matrix.

[0046] The topographic trend of the shallow seabed natural gas hydrate area is predicted based on the trend component matrix and the environment component matrix.

[0047] In terms of the monitoring array data preprocessing, the target prediction model is used to:

[0048] performing missing value processing on the monitoring array data;

[0049] Performing out-of-range elimination processing on the monitoring array data after missing value processing; the out-of-range processing is beyond the range collected by the sensors in the monitoring array;

[0050] Perform outlier elimination processing on the monitoring array data after out-of-range elimination processing;

[0051] The monitoring array data after outlier elimination is subjected to denoising to obtain preprocessed monitoring array data.

[0052] In the aspect of processing the pre-processed monitoring array data to obtain the trend component matrix, the spatial component matrix and the environmental component matrix, the target prediction model is used to:

[0053] Determining a trend component matrix based on first data; the first data being monitoring node coordinate information in the monitoring array data preprocessed at a first moment and monitoring node coordinate information in the monitoring array data preprocessed at a second moment; the first moment being a moment before the second moment;

[0054] determining a spatial component matrix based on the first data;

[0055] An environmental component matrix is determined based on second data; the second data is environmental information of the monitoring node in the preprocessed monitoring array data.

[0056] The trend component matrix is composed of multiple elements; the elements represent the terrain change trend at the location of any monitoring node in any monitoring array; wherein the elements in the trend component matrix at the next time t+1 are Equal to the element corresponding to the current time t Slope at current moment t and the random component at the next moment t+1 The sum of the slope at the next moment t+1 Equal to the slope at the current moment t and the random component at the next moment t+1 sum.

[0057] In the aspect of predicting the topographic trend of the shallow seabed natural gas hydrate area based on the trend component matrix and the environment component matrix, the target prediction model is used to:

[0058] According to the trend component matrix, the topographic trend of the shallow seabed natural gas hydrate area is preliminarily determined; according to the environmental component matrix, the preliminarily determined topographic trend of the shallow seabed natural gas hydrate area is corrected, and finally the predicted topographic trend of the shallow seabed natural gas hydrate area is obtained.

[0059] An example is:

[0060] The target prediction model includes a data preprocessing module, a component decomposition prediction module and a parameter correction module

[0061] Due to the complex seabed environment, sensor data is prone to measurement deviation and data loss, so it is necessary to preprocess the data directly obtained by the sensor. Data preprocessing includes the following four steps. The workflow diagram of the data preprocessing module is as follows: Figure 3 shown.

[0062] Step 1: Missing Value Handling: Sensor data may be missing due to environmental disturbances and short periods of system maintenance. Simply deleting the rows containing missing data may compromise the integrity of the dataset. Therefore, the following method is used to address missing values.

[0063] Select missing data points The first three data (τ1, S1), (τ2, S2), (τ3, S3) and the last three data (τ4, S4), (τ5, S5), (τ6, S6) are solved to get the interpolation polynomial: and will Substitute the above polynomial to find the missing value

[0064] Where τ represents time and S represents the data collected by the sensor. For example, (τ1, S1) means that the data collected by the sensor at time τ1 is S1.

[0065] The data points without wavy lines represent the normally collected data, and the data points with wavy lines ~ represent missing data, which are null values but can be solved by interpolation.

[0066] Step 2: Beyond the reasonable range; the data obtained by each type of sensor has a measurement range. When data beyond the sensor's measurement range is obtained, the data is discarded and returned to step 1 and processed as missing values.

[0067] For example: the monitoring range of the accelerometer is ±2g, the monitoring range of the gyroscope is ±250° / s, and the monitoring range of the magnetometer is ±4800U.

[0068] Step 3: Anomaly detection based on multivariate Gaussian distribution. Since each monitoring node is composed of different sensors, when a node fails, such as insufficient power supply voltage or reference voltage drift, the measurement data of the entire node will become abnormal. Follow the steps below to detect and process abnormal data. Each detection node contains 6 different sensors, and the data vector collected by the sensors is: V = [V1, V2, V3, V4, V5, V6] T The data follows the following multivariate Gaussian distribution:

[0069]

[0070] Where μ is the mean vector and Σ is the covariance matrix.

[0071] The above two parameter matrices μ and Σ are trained based on m data samples processed by steps 1 and 2 within a certain monitoring period:

[0072]

[0073]

[0074] Substituting the above two parameters into formula (1), we can get the distribution of the data set.

[0075] Substituting each data value into the distribution of the data set can obtain the probability that the data is an outlier, and eliminate data with a probability of more than 90% being an outlier.

[0076] Step 4: Sensor Noise Removal

[0077] Underwater terrain generally deforms slowly, so the data collected by sensors is low-frequency. Sensors are susceptible to ambient noise and high-frequency interference. Furthermore, random errors are generated during signal triggering, sensing, and transmission, and these noise frequencies are generally high. Using digital filters to eliminate noise requires fewer operations and is simple to design, making them suitable for real-time data processing and transmission systems.

[0078] An IIR digital low-pass filter based on BP neural network is used to reduce the noise of sensor data. The frequency response of the filter is:

[0079]

[0080] Among them, a i and b i is the transfer parameter in the digital filter, P and Q are the orders of the digital filter, and ω is the frequency.

[0081] After the abnormal data is extracted from the data vector collected by the sensor in step (3), the obtained data vector can be expressed as:

[0082] V'=[V1',V'2,V'3,V'4,V'5,V'6] T ;

[0083] After processing by the IIR digital filter, the frequency response of the sensor data can be expressed as:

[0084] Y i i (jω)=V i '(jω)·H(jω),i=1,2,3,4,5,6;

[0085] Where i is the number of sensor vectors, H(jw) is the digital filter frequency response function, V' i (jw) is the data vector of the sensor before filtering.

[0086] Therefore, the filtered sensor data vector can be expressed as:

[0087] Y=[Y1,Y2,Y3,Y4,Y5,Y6] T .

[0088] The preprocessed monitoring array data is used as the input of the component decomposition prediction module. According to the sensor type and the factors affecting terrain changes, the terrain changes are divided into trend component matrix, spatial component matrix, environmental component matrix and random component matrix. The expression is as follows:

[0089] y t =α t +β t +δ t +ε t ;

[0090] where y t is the predicted value at the current time t, α t is the trend component matrix at the current time t, β t is the spatial component matrix at the current time t, δ t is the environmental component matrix at the current time t, ε t is the random component matrix at the current time t, and the elements of the random component matrix follow the normal distribution.

[0091] The prediction data source of the trend component matrix comes from the deformation monitoring unit of the monitoring array. Among them, the single element of the trend component matrix at the current time t is It can be seen as an extension of a linear regression model, α [i,j] It can be seen as an intercept that changes with time, v [i,j]It can be seen as a slope that changes over time. The relationship between the trend component matrix at the next moment t+1 and the trend component matrix at the current moment t is:

[0092]

[0093] Among them, each element represents the changing trend of the terrain of different monitoring nodes at the current time t.

[0094] Specifically, = is the topographic change trend of the monitoring node j in the monitoring array i at the current time t. If it is a positive value, it means that the topographic change trend of the monitoring node j in the monitoring array i at the current time t is an upward trend. If it is a negative value, it means that the topographic change trend of the monitoring node j in the monitoring array i at the current time t is a subsidence trend. The update method of each element in the trend component matrix at the current time t is shown in formula (2). The elements in the trend component matrix at the next time t+1 are Equal to the element corresponding to the current time t Slope at current moment t and the random component at the next moment t+1 The sum of the slope at the next moment t+1 Equal to the slope at the current moment t and the random component at the next moment t+1 sum.

[0095] Among them, the random component Represents the update trend component The random fluctuations added when Indicates the update slope The random fluctuations added when , the mean and variance of the two are different.

[0096] The predicted data of the spatial component comes from the deformation monitoring unit of the monitoring array and is used to characterize the spatial distribution of the terrain in the monitoring area. First, the IMU converts the data output by the three-axis accelerometer, three-axis gyroscope, and three-axis electronic compass into quaternary data. The quaternary data is then used to calculate the attitude and reconstruct the spatial component. The specific steps are as follows:

[0097] (1) Determine the initial quaternion based on the initial attitude angles ψ0, θ0, and γ0:

[0098]

[0099] (2) Posture matrix update:

[0100]

[0101] (3) Attitude angle calculation:

[0102]

[0103] T 13 (t) is the first row and third column element of the attitude matrix at time t, and the rest are similar.

[0104] (4) Substitute the calculated attitude angle into the arc model or straight line model to reconstruct the topography of the shallow natural gas hydrate area on the seabed.

[0105] The prediction data of the environmental component matrix comes from the non-deformation monitoring units of the monitoring array, and the prediction is mainly based on the impact of pressure, temperature and methane concentration on terrain changes.

[0106] Assume that the normalized environmental vector of pressure, temperature and methane concentration at a certain moment is: x = [x1, x2, x3] T , let the terrain difference between this moment and the next moment be y, and the hypothetical function of the terrain difference and the environment vector be: h θ (x; θ)=θ0+θ1x1+θ2x2+θ3x3.

[0107] Based on m training samples, the gradient descent method is used to optimize the following cost function to obtain the impact function of environmental factors on terrain changes. The cost function is:

[0108]

[0109] Where m is the number of training samples, i is the serial number of the sample, x is the input feature of the sample, and y is the output value of the sample. θ is the hypothesis function of the terrain difference and the environment vector, and θ0, θ1, θ2, and θ3 are the parameters of the hypothesis function. J is the cost function, where θ0, θ1, θ2, and θ3 are used as decision variables to minimize the value of J(θ).

[0110] The elements of the random component matrix are white noise.

[0111] The parameter correction module inputs the preprocessed monitoring array data of the previous time period into the component decomposition prediction module, obtains the prediction result, compares the prediction result with the subsequent corresponding actual monitoring situation, feeds back the error to the various parameter matrices in the data preprocessing module and corrects the parameter matrix.

[0112] Assuming that the error between the prediction result and the actual monitoring situation is δ, the correction of the mean vector μ and the covariance matrix Σ is: where κ i is the error conversion coefficient.

[0113] Compared with the prior art, the present invention has the following technical effects:

[0114] 1. It can realize three-dimensional monitoring of terrain and strata with multiple parameters.

[0115] 2. The impact of environmental factors of shallow seabed natural gas hydrates on topography and stratum changes is combined to make the prediction more accurate.

[0116] 3. A series of preprocessing is performed on the sensor data to avoid prediction errors caused by errors or faults in the sensor measurement data.

[0117] 4. The underwater control center has a clock synchronization function to ensure that the sampling time of the sensor network is accurate and unified.

[0118] 5. Data can be transmitted wirelessly over long distances, avoiding the umbilical cable dragging the monitoring network during wired transmission.

[0119] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0120] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A system for monitoring terrain deformation in a shallow seabed natural gas hydrate area, characterized in that: include: Three-dimensional monitoring network, underwater control center, remote transmission unit and offshore intelligent prediction center; The three-dimensional monitoring network includes multiple monitoring arrays, namely horizontal arrays and vertical arrays; the horizontal arrays are laid parallel to the surface of the shallow seabed, and the vertical arrays are laid into the shallow seabed. The monitoring arrays are networked and connected by multiple monitoring nodes; the three-dimensional monitoring network is used to obtain monitoring array data; the monitoring array data is the coordinate information of the monitoring nodes and the environment in which the monitoring nodes are located; The underwater control center is used to: Acquiring the monitoring array data; Output monitoring array working instructions; the monitoring array working instructions are used to achieve time synchronization of the monitoring array and control the working mode of the monitoring array; The underwater control center provides power to the monitoring array; The power supply adopts soft-pack lithium-ion battery; The remote transmission unit is used to transmit the monitoring array data sent by the underwater control center to the offshore intelligent prediction center; the remote transmission unit includes a first relay and a second relay; the remote transmission unit includes three lines, wherein the first line is a watertight cable directly connecting the first relay and the underwater control center, the second line is an umbilical cable connecting the first relay and the second relay, and the third line is a wireless communication line for information exchange between the second relay and the offshore intelligent prediction center; the first relay is an underwater acoustic communication device or an underwater optical communication device; The offshore intelligent prediction center is used to reconstruct the topography of the shallow seabed natural gas hydrate area and predict the topography trend of the shallow seabed natural gas hydrate area based on the monitoring array data; the offshore intelligent prediction center is provided with a target prediction model; the target prediction model is a topography deformation prediction model of the shallow seabed natural gas hydrate area; the target prediction model is used to: process the preprocessed monitoring array data to obtain a trend component matrix, a spatial component matrix, and an environmental component matrix; and reconstruct the topography of the shallow seabed natural gas hydrate area based on the spatial component matrix; The topographic trend of the shallow seabed natural gas hydrate area is predicted based on the trend component matrix and the environment component matrix.

2. A system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: The monitoring node includes a deformation variable monitoring unit and a non-deformation variable monitoring unit; the deformation variable monitoring unit includes a three-axis accelerometer, a three-axis gyroscope and a three-axis electronic compass, and the deformation variable monitoring unit is used to obtain the coordinate information of the monitoring node; the non-deformation variable monitoring unit includes a pressure sensor, a temperature sensor and a methane concentration sensor, and the non-deformation variable monitoring unit is used to obtain the environmental information of the monitoring node.

3. The system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: The working modes of the monitoring array are divided into automatic high-frequency acquisition mode, automatic low-frequency acquisition mode, manual acquisition mode and sleep mode; among them, the automatic high-frequency acquisition mode and the automatic low-frequency acquisition mode are used for long-term monitoring of the terrain in the shallow natural gas hydrate area of the seabed; the manual acquisition mode is used for targeted monitoring of the terrain in the shallow natural gas hydrate area of the seabed during scientific expeditions; the sleep mode is used for data transmission and monitoring array time synchronization.

4. The system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: The remote transmission unit is connected to the control circuit of the underwater control center through a watertight cable; the remote transmission unit exchanges information with the offshore intelligent prediction center through wireless communication.

5. The system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: In terms of the monitoring array data preprocessing, the target prediction model is used to: perform missing value processing on the monitoring array data; perform out-of-range elimination processing on the monitoring array data after the missing value processing; the out-of-range is beyond the range collected by the sensors in the monitoring array; and perform outlier elimination processing on the monitoring array data after the out-of-range elimination processing; The monitoring array data after outlier elimination is subjected to denoising to obtain preprocessed monitoring array data.

6. The system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: In the aspect of processing the preprocessed monitoring array data to obtain a trend component matrix, a spatial component matrix and an environmental component matrix, the target prediction model is used to: determine the trend component matrix based on first data; the first data is the monitoring node coordinate information in the monitoring array data preprocessed at a first moment and the monitoring node coordinate information in the monitoring array data preprocessed at a second moment; the first moment is the previous moment of the second moment; determine the spatial component matrix based on the first data; determine the environmental component matrix based on the second data; the second data is the environmental information of the monitoring node in the preprocessed monitoring array data.

7. The system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: The trend component matrix is composed of multiple elements; the elements represent the terrain change trend at the location of any monitoring node in any monitoring array; among them, the elements in the trend component matrix at the next time t+1 are equal to the sum of the slope of the corresponding element at the current time t and the random component at the next time t+1; the slope at the next time t+1 is equal to the sum of the slope at the current time t and the random component at the next time t+1.

8. The system for monitoring terrain deformation in a shallow seabed natural gas hydrate area according to claim 1, characterized in that: In the aspect of predicting the topographic trend of the shallow seabed natural gas hydrate region based on the trend component matrix and the environment component matrix, the target prediction model is used to: preliminarily determine the topographic trend of the shallow seabed natural gas hydrate region based on the trend component matrix; The initially determined topographic trend of the shallow seabed natural gas hydrate area is corrected according to the environmental component matrix, and finally a predicted topographic trend of the shallow seabed natural gas hydrate area is obtained.

Images