An electromagnetic detection system based on seafloor crawler

Abstract

The present invention provides an electromagnetic detection system based on a seabed crawler, which belongs to the field of geophysical detection technology, and includes: a survey ship, a seabed crawler, a transmitting antenna, a receiving node array, a control chip and a power supply, wherein the survey ship is connected to the seabed crawler through a towing rope, a towing cable is provided at the tail of the seabed crawler, the transmitting antenna and the receiving node array are fixed on the towing cable, and the control chip and the power supply are both provided in the integrated electronic cabin of the seabed crawler; the power supply provides electric energy for the seabed crawler, the transmitting antenna, the receiving node array and the control chip, and the control chip is electrically connected to the transmitting antenna, the receiving node array and the power system of the seabed crawler; an electromagnetic detection interference compensation module is provided in the control chip, which is used to eliminate and compensate for the interference of the reflected waves received by the receiving node array. The technical problem of insufficient detection accuracy in the prior art is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of geophysical detection, and in particular, relates to an electromagnetic detection system based on a seafloor crawler. Background Art

[0002] Electromagnetic wave detection technology is widely used in the fields of marine geological exploration and seabed resource exploration. By transmitting electromagnetic waves and receiving reflected waves, seabed geological information can be obtained, providing an important basis for marine resource exploration and development. The existing seabed electromagnetic wave detection technologies mainly include the following four:

[0003] (1) Technical method 1: independent ocean electromagnetic data collection stations sink to the bottom, tow near the sea surface, or tow the excitation source deep near the seabed. The equipment cost of the floating and sinking ocean bottom electromagnetic instrument (OBEM) is high, the operation efficiency is low, and the risk of loss is high. Each OBEM costs hundreds of thousands of yuan, and each test requires dozens of devices. If you want to improve the data resolution and increase the equipment density, the number of devices required will be even larger, and the equipment cost will be higher.

[0004] (2) Technical method 2: The ocean electromagnetic data acquisition cable and the excitation cable are towed synchronously. The sea surface towing method is only suitable for shallow waters. In deep waters, the diffusion and attenuation of thick seawater will cause the electromagnetic response signal of the seabed sediments to have an extremely low signal-to-noise ratio, making it impossible to operate in deep waters. In addition, this technical method can only collect one electric field component along the cable direction, and does not collect any magnetic field component data. The data information is single, which is not conducive to data processing and interpretation.

[0005] (3) Technical method three: using both a deep-towed ocean electromagnetic excitation source near the seabed and a deep-towed ocean electromagnetic data acquisition node near the seabed. The technical solution of the Scripps Institution of Oceanography in the United States (the deep-towed excitation source and the deep-towed acquisition node are towed by one ship at the same time) is difficult to operate. It is difficult to tow multiple towed bodies with one cable and the failure rate is high.

[0006] (4) Technical method 4: seafloor towed electromagnetic detection technology, seafloor towed transmitting power supply + two pairs of electric field electrodes on the seafloor. The seafloor passive towing technology design is not suitable for work in areas with drastic terrain fluctuations.

[0007] That is to say, with the existing seabed electromagnetic detection technology, the propagation of electromagnetic waves in a complex seabed environment will be affected by many factors, such as water depth, water temperature, salinity, and the position changes of the seabed crawler itself. These interference factors will cause serious distortion of the reflected waves, thereby reducing the accuracy of electromagnetic detection. Summary of the invention

[0008] In view of this, the present invention provides an electromagnetic detection system based on a seabed crawler, which can solve the technical problem of insufficient detection accuracy caused by interference factors in existing seabed electromagnetic detection technology.

[0009] The present invention is achieved in that:

[0010] The present invention provides an electromagnetic detection system based on a seabed crawler, which includes: a survey ship, a seabed crawler, a transmitting antenna, a receiving node array, a control chip and a power supply. The survey ship is connected to the seabed crawler through a traction rope, and a towing cable is provided at the tail of the seabed crawler. The transmitting antenna and the receiving node array are fixed on the towing cable, and the control chip and the power supply are both arranged in an integrated electronic cabin of the seabed crawler; the power supply provides electrical energy to the seabed crawler, the transmitting antenna, the receiving node array and the control chip, and the control chip is electrically connected to the transmitting antenna, the receiving node array and the power system of the seabed crawler; an electromagnetic detection interference compensation module is provided in the control chip, which is used to eliminate and compensate for interference with reflected waves received by the receiving node array; the control chip is provided with a data transmission device, which is used to send the compensated electromagnetic detection results to the survey ship.

[0011] On the basis of the above technical solution, the electromagnetic detection system based on the seafloor crawler of the present invention can also be improved as follows:

[0012] Wherein, the seabed crawler is a crawler-type seabed crawler.

[0013] Furthermore, the data transmission device sends the compensated electromagnetic detection result to the survey ship via a data cable or a wireless channel.

[0014] The receiving node array is specifically a linear or two-dimensional array composed of multiple electromagnetic induction coils, and each receiving node includes a three-axis orthogonal induction coil for receiving electromagnetic field signals in different directions.

[0015] The spacing between the receiving nodes is adjustable to meet different detection depths and resolution requirements; the receiving node array is in contact with the seabed surface.

[0016] Furthermore, the transmitting antenna is a negative buoyancy antenna.

[0017] Furthermore, the module of the seabed crawler includes at least a depth meter and an altimeter.

[0018] Furthermore, it also includes an excitation source, which is fixedly arranged on the seabed crawler.

[0019] Furthermore, the excitation source, specifically a dipole transmitter composed of a pair of large-area electrode plates, can generate a low-frequency alternating current to excite an electromagnetic field in the seabed medium.

[0020] The electrode plate of the excitation source is in direct contact with the seabed surface. The electrode plate is made of a special alloy material and has good electrical conductivity and corrosion resistance.

[0021] Furthermore, the survey vessel is equipped with a satellite navigation system, an underwater positioning system, and a power supply unit.

[0022] Among them, the integrated electronic cabin is a cabin formed inside the body of the seabed crawler, the power supply supplies power to the transmitting antenna through an excitation source driving unit, and the power supply supplies power to the receiving node array through a receiving node driving unit. Specifically, an excitation source driving unit and a receiving node driving unit are provided in the electronic cabin, the excitation source driving unit is a high-power current transmitting circuit, which is used to provide energy for the transmitting antenna; the transmitting antenna is a negative buoyancy antenna; the receiving node driving unit is connected to the receiving node array.

[0023] Furthermore, a signal processing unit, an inertial measurement unit, a depth sensor, a temperature sensor, and a salinity sensor are installed in the integrated electronic cabin of the seafloor crawler.

[0024] Furthermore, the receiving node array and the seabed crawler transmit data in real time, the status information of the receiving node array and the received electromagnetic data are sent to the seabed crawler in real time, the seabed crawler records the electromagnetic data, and the commands of the seabed crawler are sent to the receiving node array in real time.

[0025] Optionally, the seabed crawler body is connected to the survey ship by a cable, the electric energy required by the seabed crawler is transmitted by the survey ship by cable, real-time data transmission is carried out between the seabed crawler and the survey ship, the status data of the seabed crawler and the received electromagnetic data are transmitted to the survey ship laboratory in real time, and the commands of the survey ship laboratory are sent to the seabed crawler in real time.

[0026] Specifically, the step S10 specifically includes: first, controlling the seafloor crawler to crawl in a predetermined detection area. During the crawling process, the seafloor crawler will collect and record seafloor environmental parameters including water depth, water temperature and salinity in real time, and obtain its own position, attitude, speed and acceleration and other posture information. At the same time, the electromagnetic wave transmitter on the seafloor crawler will transmit digitally modulated electromagnetic waves to the predetermined area, carrying digital information in the form of floating point number sequences. The electromagnetic wave receiver on the seafloor crawler will capture the reflected waves generated by these transmitted waves in the seafloor environment.

[0027] The step S20 specifically includes: demodulating and digitizing the captured reflected wave signal, extracting the digital information carried in the reflected wave, and comparing and analyzing the digital information with the digital information in the transmitted wave.

[0028] The step S30 specifically includes: firstly, aligning the transmission wave and the reflected wave in time, and then using the sliding time window method to calculate the time delay, amplitude attenuation and distinction between the reflected wave and the transmission wave in each time window. The distinction is defined as 1 minus the similarity between the reflected wave and the transmission wave in the sliding time window.

[0029] The step S40 specifically includes: inputting the information obtained in steps S10 to S30, including the transmitted wave, the reflected wave, and the time delay, amplitude attenuation and discrimination of each time window, into a pre-trained first reflected wave compensation model. The model adopts a neural network architecture with a multi-subnet structure, and after processing, a preliminarily compensated reflected wave can be obtained, which is recorded as the first compensated reflected wave.

[0030] The step S50 specifically includes: time-aligning the seabed environmental parameters and the position information of the seabed crawler obtained in step S10 with the transmission wave information in step S10 to ensure the time synchronization between various types of information.

[0031] The step S60 specifically includes: using the Bayesian network method to calculate the correlation between the seabed environmental parameters and the seabed crawler posture information and the time delay, amplitude attenuation and distinction of each time window, and obtain the influence of each interference factor on the reflected wave.

[0032] The step S70 specifically includes: according to the correlation calculated in step S60, using a preset relationship function, establishing a mathematical model covering multiple interference factors such as water depth, water temperature, salinity, crawler posture, speed and acceleration.

[0033] The step S80 specifically includes: collecting all the data obtained in steps S10-S70, constructing a fitting matrix, applying the least squares method and performing regularization processing, and finally obtaining the weight coefficient of each interference factor.

[0034] The step S90 specifically includes: inputting the first compensated reflected wave obtained in step S40, the weight coefficients of various interference factors obtained in step S80, and the seabed environmental parameters and the position information of the seabed crawler obtained in step S10 into a pre-trained second reflected wave compensation model, and after processing, the final compensated reflected wave can be obtained as the result of electromagnetic detection.

[0035] Among them, the digital information in the specific execution steps of the electromagnetic detection interference compensation module is a floating point number sequence.

[0036] Among them, the position information of the seabed crawler includes position (including x, y, z coordinates), attitude (including pitch angle, roll angle, heading angle), speed (including linear velocity and angular velocity), and acceleration (including linear acceleration and angular acceleration).

[0037] The first reflected wave compensation model adopts a neural network model with a multi-subnet structure, including a feature extraction module, a long-term feature subnet, a window feature subnet, an abnormal feature subnet and a fusion output subnet;

[0038] The feature extraction module is used to extract features from the input and classify them into long-term features, window features and abnormal features. The inputs are the time delay, amplitude attenuation and discrimination of the transmitted wave, the reflected wave and each sliding time window. The outputs are the long-term features of the reflected wave, the window features of the reflected wave and the abnormal features of the reflected wave.

[0039] The long-term feature sub-network is used to capture the long-term change trend of the reflected wave. The input is the long-term feature of the reflected wave, and the output is the error compensation vector corresponding to the long-term feature of the reflected wave, which is recorded as the first compensation vector. The structure is a multi-layer LSTM network.

[0040] The window feature subnetwork is used to process the reflection wave feature in the local time window, the input is the reflection wave window feature, the output is the error compensation vector corresponding to the reflection wave window feature, recorded as the second compensation vector, and the structure is a convolutional neural network;

[0041] The abnormal feature sub-network is used to detect and process abnormal patterns in the reflected wave, the input is the abnormal feature of the reflected wave, the output is the error compensation vector corresponding to the abnormal feature of the reflected wave, recorded as the third compensation vector, and the structure is an autoencoder network;

[0042] The fusion output sub-network is used to integrate the outputs of each sub-network and generate a final compensation result. The input is the first compensation vector, the second compensation vector, and the third compensation vector. It adopts a multi-head attention network structure and the output is the first compensation reflection wave.

[0043] The second reflected wave compensation model adopts a multi-branch neural network structure, including an environmental parameter branch network, a posture information branch network, a compensation reflected wave branch network, and a fusion branch network;

[0044] The environmental parameter branch network is used to process the influence of the seabed environmental parameters. The input is the seabed environmental parameters, and the output is the influence vector of the environmental parameters on the reflected wave. The structure is a multi-layer perceptron.

[0045] The posture information branch network is used to process the influence of the posture information of the seabed crawler. The input is the posture information of the seabed crawler, and the output is the influence vector of the posture information on the reflected wave. The structure is LSTM;

[0046] The compensation reflected wave branch network is used to further optimize the first compensation reflected wave, the input is the first compensation reflected wave, the output is the optimized compensation reflected wave, and the structure is a residual network;

[0047] The fusion branch network is used to integrate the outputs of each branch network and generate the final compensation result. The input is the output of each branch network, and the output is the final compensated reflected wave. The structure is a Transformer network combined with a fork attention mechanism.

[0048] Among them, the detailed steps of S80 are as follows:

[0049] a) Collect data: Use the data obtained in steps S10-S70, including:

[0050] Time delay, amplitude decay and discrimination of sliding time windows

[0051] Seabed environmental parameters (water depth, water temperature, salinity)

[0052] The position information of the seabed crawler (pitch angle, roll angle, heading angle, speed, acceleration)

[0053] b) Construct the fitting matrix: According to the interference equations established in S70, each interference factor is used as an independent variable, and time delay, amplitude attenuation and discrimination are used as dependent variables to construct the fitting matrix.

[0054] c) Apply the least squares method: Use the least squares method to solve the fitting matrix and preliminarily obtain the weight coefficients of each interference factor.

[0055] d) Regularization: To avoid overfitting, regularization methods (such as L1, L2 regularization or elastic network) are applied to the results of the least squares method.

[0056] e) Cross-validation: Use the k-fold cross-validation method to select the optimal regularization parameter.

[0057] f) Final fitting: Use the selected regularization parameters to perform a final fit on all data to obtain the weight coefficients of each interference factor.

[0058] g) Evaluate the quality of fit: Calculate indicators such as the coefficient of determination (R²) and the root mean square error (RMSE) to evaluate the quality of fit.

[0059] h) Output results: Output the final weight coefficients of various interference factors, including the weight coefficients of water depth, water temperature, salinity, crawler pitch angle, roll angle, heading angle, speed and acceleration.

[0060] The following is a description of the formulation of the interference equations:

[0061] Water depth interference equation:

[0062] ;

[0063] In the formula, is the degree of interference of water depth on electromagnetic waves; for water depth; For time; is the coefficient to be determined; is the first error term.

[0064] Parameter acquisition method:

[0065] It is directly measured by the depth sensor. Calculated through continuous depth measurement:

[0066] ;

[0067] In practical applications, a discrete approximation can be used:

[0068] ;

[0069] in is the sampling interval.

[0070] Water temperature interference equation:

[0071] ;

[0072] In the formula, is the degree of interference of water temperature on electromagnetic waves; for water temperature; for depth; is the angular frequency of temperature cycle change; is the coefficient to be determined; is the second error term.

[0073] Parameter acquisition method:

[0074] Directly measured by a temperature sensor. By measuring the temperature at different depths and calculating:

[0075] ;

[0076] in is the depth interval. It can be obtained by Fourier analysis of temperature time series data.

[0077] Salinity interference equation:

[0078] ;

[0079] In the formula, is the degree of interference of salinity on electromagnetic waves; is salinity; is the horizontal coordinate; is an imaginary unit; is the phase factor; is the coefficient to be determined; is the third error term.

[0080] Parameter acquisition method:

[0081] Directly measured by a salinity sensor. and By measuring the salinity at different horizontal positions and calculating:

[0082] ;

[0083] ;

[0084] in and is the horizontal interval. It can be determined by analyzing the periodic changes in salinity data.

[0085] Crawler pitch angle interference equation:

[0086] ;

[0087] In the formula, is the degree of interference of the pitch angle on electromagnetic waves; is the pitch angle; For time; is the coefficient to be determined; is the fourth error term.

[0088] Parameter acquisition method:

[0089] Obtained directly from the inertial measurement unit. Calculated through continuous pitch angle measurement:

[0090] ;

[0091] in is the sampling interval.

[0092] Crawler roll angle interference equation:

[0093] ;

[0094] In the formula, is the interference degree of the rolling angle on the electromagnetic wave; is the roll angle; For time; is the coefficient to be determined; is the fifth error term.

[0095] Parameter acquisition method:

[0096] Obtained directly from the inertial measurement unit. The roll angle is calculated by continuous measurement:

[0097] ;

[0098] in is the sampling interval.

[0099] Crawler heading angle interference equation:

[0100] ;

[0101] In the formula, is the interference degree of the heading angle on the electromagnetic wave; is the heading angle; For time; is an imaginary unit; is the coefficient to be determined; is the sixth error term.

[0102] Parameter acquisition method:

[0103] Obtained directly from the inertial measurement unit. Calculated through continuous heading angle measurement:

[0104] ;

[0105] in is the sampling interval.

[0106] Crawler velocity disturbance equation:

[0107] ;

[0108] In the formula, is the degree of interference of speed on electromagnetic waves; is the velocity vector; For time; is the coefficient to be determined; is the seventh error term.

[0109] Parameter acquisition method:

[0110] Calculated by a combination of the inertial measurement unit and GPS (if near the water surface). Calculated from continuous speed measurements:

[0111] ;

[0112] in is the sampling interval.

[0113] Crawler acceleration disturbance equation:

[0114] ;

[0115] In the formula, is the degree of interference of acceleration on electromagnetic waves; is the acceleration vector; For time; is the coefficient to be determined; is the eighth error term.

[0116] Parameter acquisition method:

[0117] Obtained directly from the inertial measurement unit. Calculated through continuous acceleration measurement:

[0118] ;

[0119] in is the sampling interval.

[0120] The step of constructing the training data set of the first reflected wave compensation model specifically includes:

[0121] A large number of electromagnetic wave transmission and reception experiments were carried out under different seabed environmental conditions to collect transmission wave and reflection wave data.

[0122] The collected data are preprocessed, including denoising, normalization, and time alignment.

[0123] Using a sliding time window technique, the time delay, amplitude decay, and discrimination of each window are calculated.

[0124] The processed data is divided into training set, validation set and test set.

[0125] The training steps of the first reflected wave compensation model specifically include:

[0126] Step 101: Initialize the parameters of each sub-network.

[0127] Step 102: Input the training data into a feature extraction module to obtain long-term features, window features, and abnormal features.

[0128] Step 103: Input these features into the corresponding sub-networks for forward propagation.

[0129] Step 104: Use the fusion output sub-network to synthesize the outputs of each sub-network.

[0130] Step 105: Calculate the loss function between the output and the actual compensation reflected wave.

[0131] Step 106: Use the back propagation algorithm to calculate the gradient and update the network parameters.

[0132] Step 107: Repeat steps 102-106 until a predetermined number of iterations is reached or a convergence condition is met.

[0133] Step 108: Use the validation set to evaluate the model performance and adjust hyperparameters if necessary.

[0134] Step 109: Finally, use the test set to evaluate the generalization ability of the model.

[0135] The step of constructing the training data set of the second reflected wave compensation model specifically includes:

[0136] Collect seabed environmental parameters, seabed crawler position information, and corresponding electromagnetic wave emission and reception data.

[0137] The collected data are preliminarily processed using the first reflection wave compensation model.

[0138] Calculate the weight coefficient of each interference factor.

[0139] The processed data, weight coefficients, and original environment parameters and pose information are combined into training samples.

[0140] Divide the dataset into training set, validation set, and test set.

[0141] The training steps of the second reflected wave compensation model specifically include:

[0142] Step 201: Initialize the parameters of each branch network.

[0143] Step 202: input the training data into each branch network for forward propagation.

[0144] Step 203: Use the fusion branch network to integrate the outputs of each branch network.

[0145] Step 204: Calculate the loss function between the output and the actual final compensated reflected wave.

[0146] Step 205: Use the back propagation algorithm to calculate the gradient and update the network parameters.

[0147] Step 206: Repeat steps 202-205 until a predetermined number of iterations is reached or a convergence condition is met.

[0148] Step 207: Use the validation set to evaluate the model performance and adjust the hyperparameters if necessary.

[0149] Step 208: Finally, use the test set to evaluate the generalization ability of the model.

[0150] Compared with the prior art, the electromagnetic detection system based on the seafloor crawler provided by the present invention has the following beneficial effects:

[0151] 1. Comprehensive consideration of various interference factors: The method of the present invention not only considers seabed environmental parameters such as water depth, water temperature and salinity, but also includes dynamic factors such as the position, attitude, speed and acceleration of the seabed crawler. By establishing a mathematical model covering multiple interference sources, the impact of various interferences on the reflected waves can be more comprehensively analyzed and compensated.

[0152] 2. Adopting adaptive compensation mechanism: The present invention adopts a multi-level neural network model to adaptively compensate for reflected waves. The model can automatically learn the influence of various interference factors on reflected waves based on the real-time measured environmental parameters and posture information, and generate targeted compensation results. Compared with empirical formulas or static models, this method has stronger adaptive ability and versatility.

[0153] 3. Improve detection accuracy: By performing multi-level compensation on reflected waves, the method of the present invention can effectively suppress the influence of various interference factors on electromagnetic detection results.

[0154] In summary, the present invention solves the technical problem that the existing seabed electromagnetic detection technology may have insufficient detection accuracy due to interference factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0155] Figure 1 A schematic diagram of the composition of an electromagnetic detection system based on a seafloor crawler provided by the present invention;

[0156] Figure 2 A flow chart of steps executed by an electromagnetic detection interference compensation module;

[0157] Figure 3 This is a schematic diagram of the system composition in Example 1. DETAILED DESCRIPTION

[0158] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0159] like Figure 1 The figure shows a schematic diagram of the composition of an electromagnetic detection system based on a seabed crawler provided by the present invention. The system includes: a survey ship, a seabed crawler, a transmitting antenna, a receiving node array, a control chip and a power supply. The survey ship is connected to the seabed crawler through a towing rope. A towing cable is provided at the tail of the seabed crawler. The transmitting antenna and the receiving node array are fixed on the towing cable. The control chip and the power supply are both arranged in the integrated electronic cabin of the seabed crawler. The power supply provides electrical energy to the seabed crawler, the transmitting antenna, the receiving node array and the control chip. The control chip is electrically connected to the transmitting antenna, the receiving node array and the power system of the seabed crawler. An electromagnetic detection interference compensation module is provided in the control chip for eliminating interference and compensating for reflected waves received by the receiving node array. The control chip is provided with a data transmission device for sending the compensated electromagnetic detection results to the survey ship. The specific composition of the system is described in detail as follows:

[0160] 1. Overall system composition

[0161] 1.1 Seabed crawler: The seabed crawler is the core carrier of this system. It adopts a sealed metal shell structure and integrates electronic equipment, power system and various sensors. The bottom of the crawler is equipped with a crawler or wheel structure for moving on the seabed. The top of the crawler is equipped with a connection interface for connecting to the composite cable.

[0162] 1.2 Electromagnetic transmitting antenna: The electromagnetic transmitting antenna adopts a flexible design and can be folded and stored in a special storage device inside or outside the crawler. The antenna is made of conductive material and covered with insulating material on the outside. It can conduct large currents in seawater and transmit electromagnetic signals.

[0163] 1.3 Receiving node array: The receiving node array consists of multiple independent receiving nodes, each of which contains an electric field sensor and a magnetic field sensor. The nodes are connected by flexible connecting wires and can be stored in a special storage device inside or outside the crawler.

[0164] 1.4 Composite cable: Composite cable is used to connect the seabed crawler and the survey ship, and contains power transmission lines and optical fiber communication lines. The composite cable has sufficient strength to withstand the pressure and tension of the underwater environment.

[0165] 1.5 Survey vessel: The survey vessel is equipped with a dynamic positioning system, winch system, lifting equipment, monitoring system and data processing system, etc.

[0166] 2. Detailed structure of the submarine crawler

[0167] 2.1 Shell structure: The crawler shell is made of pressure-resistant material and can withstand the high pressure of the deep sea environment. There are sealed hatches on the shell to facilitate maintenance and replacement of internal equipment.

[0168] 2.2 Power system: The power system includes multiple motors and thrusters, which can realize the crawler's forward, backward, turning and other actions. The thrusters adopt a sealed design to prevent seawater from infiltrating.

[0169] 2.3 Electronics cabin: The electronics cabin is located inside the crawler and contains core electronic equipment such as the main controller, power management system, and communication system. The electronics cabin adopts a special heat dissipation design to ensure that the equipment works normally under high-voltage environment.

[0170] 2.4 Excitation source drive unit: The excitation source drive unit is used to control the electromagnetic transmitting antenna and can generate electromagnetic signals of different waveforms, frequencies and intensities. The unit adopts a high-efficiency power module and can output large current.

[0171] 2.5 Receiving Node Driver Unit: The receiving node driver unit is responsible for controlling the receiving node array, including data acquisition, storage and transmission functions. This unit uses a high-precision analog-to-digital converter to ensure the accuracy of data acquisition.

[0172] 2.6 Sensor system: The sensor system includes depth meter, altimeter, attitude sensor, acoustic positioning system, etc. These sensors are used to monitor the position, attitude and surrounding environment parameters of the crawler.

[0173] 2.7 Lighting and camera system: The crawler is equipped with high-brightness LED lighting and high-definition cameras for seabed environment observation and operation monitoring.

[0174] 3. Electromagnetic transmitting antenna

[0175] 3.1 Structural design: The electromagnetic transmitting antenna adopts a long strip design and can be folded for storage. The antenna is composed of multiple sections of flexible conductors, each section is connected by a flexible connector, which is easy to unfold and store.

[0176] 3.2 Material selection: The antenna conductor is made of highly conductive materials, such as copper or aluminum alloy. The outer layer is covered with insulating materials resistant to seawater corrosion, such as special polymers.

[0177] 3.3 Connection method: One end of the antenna is connected to the excitation source drive unit of the crawler, and the other end is equipped with a buoyancy device or a weight to adjust the posture of the antenna in the water.

[0178] 4. Receiving node array

[0179] 4.1 Node structure: Each receiving node is a sealed spherical or cylindrical structure, which integrates electric field sensor, magnetic field sensor, data acquisition circuit and memory.

[0180] 4.2 Sensor configuration: Each node contains a three-component electric field sensor and a three-component magnetic field sensor, which can measure the electromagnetic field in all directions.

[0181] 4.3 Node connection: Nodes are connected by flexible cables, which contain power lines and data transmission lines. The entire array can be folded for storage, making it easy to deploy and recycle.

[0182] 4.4 Power supply mode: The receiving node can be powered by a built-in battery or it can obtain power from the crawler through a connecting cable.

[0183] 5. Composite cable

[0184] 5.1 Structural design: The composite cable consists of a multi-layer structure, including a central reinforcement core, a power transmission layer, an optical fiber communication layer and an outer protective cover.

[0185] 5.2 Power transmission: The power transmission layer uses multi-strand copper conductors, which can carry high-power power transmission and provide continuous and stable power supply for the crawler.

[0186] 5.3 Data communication: The fiber optic communication layer contains multiple optical fibers for high-speed data transmission, ensuring real-time communication between the crawler and the survey vessel.

[0187] 5.4 Strength design: The composite cable has sufficient tensile strength to withstand the pressure and tension of the underwater environment while maintaining good flexibility.

[0188] 6. Survey vessel equipment

[0189] 6.1 Dynamic Positioning System: The survey vessel is equipped with a dynamic positioning system, which can maintain a stable position at sea and facilitate the deployment and recovery of equipment.

[0190] 6.2 Winch system: The winch system is used for laying and recovering composite cables and has tension monitoring and automatic adjustment functions.

[0191] 6.3 Lifting equipment: The survey vessel is equipped with a crane for the deployment and recovery of the seafloor crawler.

[0192] 6.4 Monitoring system: The monitoring system includes a multi-screen display, an operating console and a computer system for real-time monitoring of the status of subsea operations.

[0193] 6.5 Data processing system: The data processing system consists of high-performance computers and large-capacity storage devices, and is used to receive, store and preliminarily process seabed electromagnetic detection data.

[0194] like Figure 2 As shown, it is a flowchart of the specific execution steps of the electromagnetic detection interference compensation module. The specific implementation methods of the above steps are described in detail below:

[0195] The specific implementation method of step S10 is: first, control the seabed crawler to crawl in the predetermined detection area. During the crawling process, the seabed crawler will collect and record seabed environmental parameters in real time, including water depth, water temperature, salinity and other information. At the same time, the seabed crawler will also obtain its own posture information, including position (x, y, z coordinates), attitude (pitch angle, roll angle, heading angle), speed (linear velocity and angular velocity) and acceleration (linear acceleration and angular acceleration). At the same time, the electromagnetic wave transmitter installed on the seabed crawler will emit electromagnetic waves to the predetermined area. These emission waves are digitally modulated and carry digital information in the form of floating point number sequences. The electromagnetic waves emitted by the electromagnetic wave transmitter will be reflected in the seabed environment, and the electromagnetic wave receiver on the seabed crawler will capture these reflected waves.

[0196] The specific implementation of step S20 is: demodulating and digitizing the captured reflected wave signal to extract the digital information carried in the reflected wave. This digital information will be compared and analyzed with the digital information in the transmitted wave.

[0197] The specific implementation method of step S30 is: first, the transmission wave and the reflected wave are time-aligned, and then the time delay, amplitude attenuation and distinction between the reflected wave and the transmission wave in each time window are calculated by using the sliding time window method. The distinction is defined as 1 minus the similarity between the reflected wave and the transmission wave in the sliding time window. These characteristic parameters will play an important role in the subsequent reflected wave compensation.

[0198] The specific implementation method of step S40 is: input the information obtained in steps S10 to S30, including the transmission wave, the reflection wave, and the time delay, amplitude attenuation and discrimination of each time window, into the pre-trained first reflection wave compensation model. The model adopts a neural network architecture with a multi-subnet structure, including a feature extraction module, a long-term feature sub-network, a window feature sub-network, an abnormal feature sub-network and a fusion output sub-network. After processing by this model, a preliminarily compensated reflection wave can be obtained, which is recorded as the first compensated reflection wave.

[0199] The specific implementation method of step S50 is: time-aligning the seabed environmental parameters (water depth, water temperature, salinity) obtained in step S10 and the position information (position, attitude, speed, acceleration) of the seabed crawler with the transmission wave information in step S10 to ensure the time synchronization between various types of information.

[0200] The specific implementation method of step S60 is: using the Bayesian network method, calculating the correlation between the seabed environmental parameters (water depth, water temperature, salinity) and the position information (pitch angle, roll angle, heading angle, speed, acceleration) of the seabed crawler and the time delay, amplitude attenuation and discrimination of each time window. In this way, the influence of each interference factor on the reflected wave can be obtained.

[0201] The specific implementation method of step S70 is: according to the correlation calculated in step S60, a series of pre-set relationship functions are used to establish interference equations such as water depth interference equation, water temperature interference equation, salinity interference equation, crawler attitude interference equation (pitch angle, roll angle, heading angle), crawler speed interference equation and crawler acceleration interference equation. These relationship functions include linear relationship, inverse relationship, logarithmic relationship, exponential relationship and power function relationship. The specific expression of the interference equation group is as follows:

[0202] Water depth interference equation:

[0203] ;

[0204] in, is the interference degree of water depth on electromagnetic waves, For water depth, For time, is the coefficient to be determined, is the first error term. Water depth It is directly measured by the depth sensor, It is calculated through continuous depth measurement.

[0205] Water temperature interference equation:

[0206] ;

[0207] in, is the degree of interference of water temperature on electromagnetic waves, For water temperature, For depth, is the angular frequency of the temperature cycle, is the coefficient to be determined, is the second error term. Water temperature It is directly measured by the temperature sensor, It is calculated by measuring the temperature at different depths.

[0208] Salinity interference equation:

[0209] ;

[0210] in, is the degree of interference of salinity on electromagnetic waves, is the salinity, is the horizontal coordinate, is an imaginary unit, is the phase factor, is the coefficient to be determined, is the third error term. Salinity It is directly measured by the salinity sensor. and It is calculated by measuring the salinity at different horizontal positions.

[0211] Crawler pitch angle interference equation:

[0212] ;

[0213] in, is the interference degree of the pitch angle on the electromagnetic wave, is the pitch angle, For time, is the coefficient to be determined, is the fourth error term. Pitch angle It is directly measured by the inertial measurement unit, and It is calculated through continuous pitch angle measurements.

[0214] Crawler roll angle interference equation:

[0215] ;

[0216] in, is the interference degree of the rolling angle on the electromagnetic wave, is the roll angle, For time, is the coefficient to be determined, is the fifth error term. Roll angle It is directly measured by the inertial measurement unit, and It is calculated through continuous roll angle measurement.

[0217] Crawler heading angle interference equation:

[0218] ;

[0219] in, is the interference degree of heading angle on electromagnetic waves, is the heading angle, For time, is an imaginary unit, is the coefficient to be determined, is the sixth error term. Heading angle It is directly measured by the inertial measurement unit, and It is calculated through continuous heading angle measurement.

[0220] Crawler velocity disturbance equation:

[0221] ;

[0222] in, is the degree of interference of speed on electromagnetic waves, is the velocity vector, For time, is the coefficient to be determined, is the seventh error term. Speed It is calculated by the inertial measurement unit and GPS (if it is near the water surface), and It is calculated through continuous speed measurement.

[0223] Crawler acceleration disturbance equation:

[0224] ;

[0225] in, is the degree of interference of acceleration on electromagnetic waves, is the acceleration vector, For time, is the coefficient to be determined, is the eighth error term. Acceleration It is directly measured by the inertial measurement unit, and It is calculated through continuous acceleration measurement.

[0226] The specific implementation method of step S80 is: first collect all the data obtained in steps S10-S70, including the time delay, amplitude attenuation and discrimination of the sliding time window, as well as the seabed environmental parameters (water depth, water temperature, salinity) and the posture information of the seabed crawler (pitch angle, roll angle, heading angle, speed, acceleration). Then, according to the interference equations established in step S70, each interference factor is used as an independent variable, and the time delay, amplitude attenuation and discrimination are used as dependent variables to construct a fitting matrix. Next, the least squares method is applied to solve this fitting matrix to preliminarily obtain the weight coefficients of each interference factor. In order to avoid overfitting, the results of the least squares method need to be regularized, such as L1 regularization or L2 regularization. Then the k-fold cross-validation method is used to select the optimal regularization parameter. Finally, the selected regularization parameter is used to perform a final fitting on all the data to obtain the final weight coefficients of each interference factor. In order to evaluate the quality of the fitting, it is also necessary to calculate indicators such as the coefficient of determination (R²) and the root mean square error (RMSE). Through the above steps, the weight coefficients of various interference factors such as water depth, water temperature, salinity, crawler pitch angle, roll angle, heading angle, speed and acceleration can be obtained.

[0227] The specific implementation of step S90 is: the first compensation reflection wave obtained in step S40, the weight coefficients of each interference factor obtained in step S80, and the seabed environmental parameters and the position and posture information of the seabed crawler obtained in step S10 are input into the pre-trained second reflection wave compensation model. The model adopts a multi-branch neural network structure, including an environmental parameter branch network, a position and posture information branch network, a compensation reflection wave branch network, and a fusion branch network. After processing by this model, the final compensated reflection wave can be obtained as the final result of electromagnetic detection.

[0228] In general, the invention proposes an electromagnetic detection method based on a seafloor crawler. Through real-time measurement of seafloor environmental parameters and crawler posture information, as well as multi-level compensation of reflected waves, the influence of various interference factors on electromagnetic detection results can be effectively suppressed, thereby improving detection accuracy. Among them, the key technologies include: 1) using digital modulation to transmit electromagnetic waves, and demodulating and extracting features from reflected waves; 2) establishing a mathematical model covering multiple interference factors, and solving the weight coefficients of each interference factor through the least squares method; 3) using a neural network model with a multi-subnet structure to perform multi-level compensation for reflected waves.

[0229] Specifically, the principle of the present invention is:

[0230] First, the seafloor crawler crawls in the predetermined detection area, and collects environmental parameters such as water depth, water temperature, salinity, and posture information such as position, attitude, speed, and acceleration in real time through various sensors installed. At the same time, the electromagnetic wave transmitter on the crawler will transmit digitally modulated electromagnetic waves to the predetermined area, carrying digital information in the form of floating-point number sequences. These transmitted waves will be reflected in the seafloor environment, and the electromagnetic wave receiver on the crawler will capture these reflected wave signals. Since the excitation source is in direct contact with the seafloor surface, the electromagnetic waves can be effectively excited, and more electromagnetic waves directly enter the sedimentary strata, reducing the diffusion and attenuation of the excited electromagnetic waves by the highly conductive seawater. Since the receiving array is in direct contact with the seafloor surface, the electromagnetic waves can be effectively received.

[0231] Next, the captured reflected wave signal is demodulated and digitized to extract the digital information carried in it. Then, the sliding time window method is used to calculate the characteristic parameters such as time delay, amplitude attenuation and discrimination between the reflected wave and the transmitted wave in each time window. These parameters can reflect the impact of various interference factors on the reflected wave.

[0232] Based on the above characteristic parameters, the present invention adopts a neural network model with a multi-subnet structure to perform preliminary reflected wave compensation. The model includes a feature extraction module, a long-term feature subnetwork, a window feature subnetwork, an abnormal feature subnetwork, and a fusion output subnetwork. Through the collaborative work of these subnetworks, the long-term change trend of the reflected wave, the characteristics within the local time window, and the abnormal pattern can be captured, and the preliminary compensation results can be comprehensively generated.

[0233] In order to further improve the compensation effect, the present invention also establishes a multi-branch neural network model. The model includes an environmental parameter branch network, a posture information branch network, a compensation reflection wave branch network, and a fusion branch network. The environmental parameter branch network and the posture information branch network are used to process the influence of environmental parameters and posture information on the reflection wave respectively, the compensation reflection wave branch network further optimizes the preliminary compensation result, and the fusion branch network integrates the output of each branch to generate the final compensation result.

[0234] It is worth noting that before establishing the above compensation model, the method of the present invention also calculates the correlation of various interference factors on the reflected wave through the Bayesian network, and establishes a mathematical model covering various factors such as water depth, water temperature, salinity, crawler posture, speed and acceleration, that is, the interference equation group. This provides important prior knowledge and parameter basis for the subsequent neural network model.

[0235] In general, the method of the present invention makes full use of the dynamic measurement capability of the seabed crawler, establishes a mathematical model covering a variety of interference factors, and adopts an adaptive neural network compensation mechanism, which can effectively suppress the influence of various interferences on the electromagnetic detection results and improve the detection accuracy.

[0236] In order to better understand and implement the present invention, a specific embodiment 1 of the hardware composition and operation of the present invention is provided below: Figure 3 As shown, the electromagnetic detection system based on the seafloor crawler of Example 1 is mainly composed of the following parts: a seafloor crawler, an electromagnetic transmitting antenna, a receiving node array, a composite cable and a survey ship, which is divided into a surface survey ship support platform and a seafloor crawler platform. The specific implementation methods and working processes of each component will be described in detail below.

[0237] 1. Seafloor crawler

[0238] The seabed crawler is the core carrier of this system. The deep-sea crawler model is ROV3000, and its main parameters are as follows:

[0239] Maximum operating depth: 3000 meters;

[0240] Size: 2.5 meters long, 1.8 meters wide, 1.2 meters high;

[0241] Weight: about 1500 kg in air, about 50 kg underwater (slightly positive buoyancy);

[0242] Maximum crawling speed: 2 knots (about 1 m / s);

[0243] Power system: 6 400W thrusters for all-round movement;

[0244] Power supply: DC380V, maximum power 15kW;

[0245] The crawler body integrates the following main components:

[0246] a) Electronics cabin: Made of titanium alloy, it contains the main controller, power management system, communication system and other core electronic equipment. The main controller uses Intel Core i7 10700 processor with a main frequency of 3.8 GHz, 16 GB RAM, and 512 GB SSD storage.

[0247] b) Excitation source drive unit: It adopts high-efficiency IGBT (insulated gate bipolar transistor) power module with maximum output current of 1000A and voltage range of 01000V, which can generate electromagnetic signals of various waveforms.

[0248] c) Receiving node driver unit: uses 24-bit high-precision ADC (analog-to-digital converter), with a sampling rate of up to 192kHz and a dynamic range of 144dB.

[0249] d) Sensor module: including depth meter (accuracy ±0.01%FS), altimeter (accuracy ±1cm), attitude sensor (accuracy ±0.1°), and acoustic positioning system (accuracy ±0.1m).

[0250] e) Lighting system: 4 500W LED lamps, total luminous flux 80,000 lumens.

[0251] f) Camera system: 2 4K ultra-high-definition cameras, 360° rotatable.

[0252] 2. Electromagnetic transmitting antenna

[0253] The electromagnetic transmitting antenna adopts negative buoyancy design to ensure that it can maintain good contact with the ground on the seabed. The specific parameters are as follows:

[0254] Length: 100 meters; Diameter: 50 mm; Material: Polyurethane sheath on the outside, copper conductor inside; Resistance: 0.01Ω / m; Maximum operating current: 1000A; Frequency range: 0.1Hz10kHz;

[0255] 3. Receiving node array

[0256] The receiving node array consists of 10 receiving nodes, with a spacing of 10 meters between each node and a total length of 100 meters. Each receiving node contains the following sensors:

[0257] Electric field sensor: 3-component orthogonal electric field sensor; Magnetic field sensor: 3-component magnetometer; Inclinometer: accuracy 0.1°;

[0258] The main parameters of the receiving node array are:

[0259] Sampling rate: up to 192kHz, adjustable; Resolution: 24 bits; Storage capacity: 256GB per node; Battery: Li-ion battery, 100Wh capacity, can work continuously for 72 hours;

[0260] 4. Composite cable

[0261] The composite cable is used to connect the seafloor crawler and the survey ship, providing power transmission and data communication functions. The main parameters are as follows:

[0262] Length: 5000 meters; Diameter: 32 mm; Weight: 1.2 kg / m (in air); Tensile strength: 150 kN; Power transmission: 3-phase 380 VAC, maximum power 20 kW; Data communication: optical fiber communication, bandwidth 10 Gbps; Material: The outer layer is a polyurethane sheath, and the interior contains copper conductors and optical fibers;

[0263] 5. Survey Ship

[0264] The survey vessel is a 3,000-ton oceanographic survey vessel equipped with the following main equipment:

[0265] DP2 level dynamic positioning system; composite cable winch: capacity 5000 meters, maximum tension 200kN; deck crane: maximum lifting capacity 10 tons; integrated monitoring host: adopts high-performance workstation, equipped with multi-screen display system; data storage system: adopts RAID5 array, total capacity 100TB; uninterruptible power supply system: capacity 100kVA, sustainable power supply for 2 hours.

[0266] Workflow:

[0267] 1. Preparation before operation

[0268] 1. Equipment inspection: On the deck of the survey ship, the technicians first conducted a comprehensive inspection of the seafloor crawler: check the appearance of the crawler to ensure that there is no obvious damage or corrosion; turn on the power of the crawler to verify that all systems start up normally; test the working status of the six thrusters to ensure that the speed and thrust are normal; check the sealing of the electronic compartment to ensure that there is no risk of water seepage; calibrate the depth gauge, altimeter and attitude sensor; test the camera system and lighting system; then, the technicians inspected the electromagnetic transmitting equipment: check the appearance of the transmitting antenna to ensure that there is no damage; measure the resistance of the transmitting antenna to ensure that it is around 0.01Ω / m; test the excitation source drive unit to verify the different energy levels (10 Check the output of the receiving array and the crawler (0A, 500A, 1000A); check the connection interface between the transmitting antenna and the crawler; then, inspect the seabed electromagnetic receiving array: check the appearance of 10 receiving nodes; test the battery power of each node to ensure sufficient; calibrate the electric field sensor, magnetic field sensor and inclinometer of each node; test the data acquisition and storage functions; check the connection interface between the receiving array and the crawler; finally, check the composite cable: check whether the outer sheath of the composite cable is intact; test the power transmission function to ensure that the 380VAC power supply is normal; test the optical fiber communication to verify that the 10Gbps bandwidth is available; check the connection interface between the composite cable and the crawler and the survey ship;

[0269] 2. System connection

[0270] After confirming that all equipment has been inspected correctly, the technicians began to connect the system: connect the electromagnetic transmitting antenna to the excitation source drive unit of the crawler; connect the receiving node array to the receiving node drive unit of the crawler; connect one end of the composite cable to the crawler and the other end to the equipment on the survey ship;

[0271] 3. System testing

[0272] After the connection is completed, the overall system test is carried out: the communication between the crawler and the survey ship is tested to ensure stable data transmission; the power supply is tested to ensure that the power supply from the survey ship to the crawler is normal; and the simulation test of electromagnetic transmission and reception is carried out to verify that the entire system is working properly;

[0273] 2. Equipment deployment

[0274] 1. Survey vessel positioning: After the survey vessel arrives at the operating sea area, it uses the DP2 dynamic positioning system to maintain the ship's position and reduces the ship's speed to about 2 knots for steady sailing.

[0275] 2. Receiving array deployment: The technicians first put the far end of the receiving node array (the 10th node) into the sea. As the survey ship slowly moves forward, the remaining nodes are gradually deployed. During the deployment process, the status information of each node is monitored to ensure normal operation.

[0276] 3. Transmitting antenna deployment: When the receiving array is nearly completed, start to deploy the electromagnetic transmitting antenna; start entering the water from the far end of the antenna and gradually deploy the entire antenna;

[0277] 4. Crawler deployment: When the deployment of the transmitting antenna is nearly completed, use the deck crane to lift the crawler; slowly lower the crawler into the water and lay the composite cable at the same time; monitor the crawler's depth gauge and altimeter data and adjust the composite cable release speed; when the crawler is about 10 meters away from the seabed, slow down the cable release speed; after the crawler touches the bottom, keep a little excess composite cable to facilitate the movement of the crawler.

[0278] 3. Seabed electromagnetic detection operations

[0279] 1. Parameter settings

[0280] The laboratory technicians of the survey vessel set the following parameters through the integrated monitoring host:

[0281] Crawler moving path: A grid path is designed according to the exploration area, with each survey line length of 1,000 meters and a line spacing of 100 meters;

[0282] Crawling speed: set to 0.5 m / s;

[0283] Excitation source parameters: excitation current: 500A; excitation waveform: square wave; fundamental frequency: 0.25Hz; excitation duration: 4 seconds; interval time: 16 seconds; collection parameters: sampling rate: 4096Hz; sampling length: continuous sampling.

[0284] 2. Operation starts: Start the crawler and start moving along the set path; activate the electromagnetic launch system and start emitting electromagnetic signals according to the set parameters; the receiving node array starts to continuously collect data;

[0285] 3. Real-time monitoring

[0286] The laboratory technicians of the survey ship monitor the following information in real time through the integrated monitoring host: the position, depth and attitude of the crawler; the working status of the transmitting system, including the output current and voltage waveform; the working status of the receiving node, including the signal strength and data quality; the status of the composite cable, including the tension and length;

[0287] 4. Data transmission and storage: The received electromagnetic data is transmitted to the survey ship in real time via a composite cable; the data storage system on the survey ship saves the data in a RAID5 array; at the same time, the receiving node also stores a copy of the data locally as a backup;

[0288] 5. Emergency treatment

[0289] If the following situations occur, the technicians will take corresponding measures: The crawler deviates from the planned path: adjust the direction of the crawler through the integrated monitoring host; The launch system is abnormal: try to adjust the launch parameters, such as reducing the launch power; If the problem persists, suspend the launch; The receiving node is abnormal: If a single node is abnormal, continue the operation but mark the node data; If multiple nodes are abnormal, consider repairing the recovery equipment; The composite cable tension is abnormal: immediately stop the crawler movement and adjust the ship position to reduce the tension.

[0290] (IV) Completion of operation and equipment recovery

[0291] 1. Operation end procedure: After completing the scheduled survey line, stop the crawler movement; turn off the electromagnetic transmission system; stop receiving node data collection; and perform a quick quality check on the collected data;

[0292] 2. Equipment recovery: Start the composite cable winch and begin to recover the composite cable; monitor the ascent process of the crawler and maintain a slow and stable speed; when the crawler approaches the sea surface, use the deck crane to assist in lifting it onto the deck; continue to recover the transmitting antenna, paying attention to avoid entanglement; finally, recover the receiving node array and check the status of each node one by one.

[0293] In order to further better understand and implement the present invention, a specific embodiment 2 of the electromagnetic detection interference compensation module of the present invention is provided below. The specific steps of this embodiment 2 are described in detail as follows:

[0294] The specific implementation of step S10 is as follows:

[0295] First, the seafloor crawler is controlled to crawl in the predetermined detection area. During the crawling process, the seafloor crawler will collect and record seafloor environmental parameters in real time. Specifically, the seafloor crawler is equipped with depth sensors, temperature sensors and salinity sensors, through which the water depth can be obtained. , water temperature and salinity At the same time, the seabed crawler is also equipped with an inertial measurement unit and GPS (if it is near the water surface), which can obtain the location of the crawler. ,attitude ,speed and acceleration Etc. posture information.

[0296] At the same time, the seafloor crawler is equipped with an electromagnetic wave transmitter and receiver. The electromagnetic wave transmitter will transmit digitally modulated electromagnetic waves to the predetermined area. These transmitted waves can be expressed as:

[0297] ;

[0298] in, is the amplitude modulated signal, is the phase modulated signal, is the carrier frequency. The transmitted wave adopts digital modulation and carries digital information in the form of floating point number sequence.

[0299] These transmitted waves will be reflected in the seabed environment, and the electromagnetic wave receivers on the seabed crawler will capture these reflected waves, which can be expressed as:

[0300] ;

[0301] in, is the number of reflected wave paths, For the The complex amplitude attenuation coefficient of the path, For the The time delay of each path, For noise.

[0302] The specific implementation of step S20 is as follows:

[0303] The captured reflected wave signal Demodulation and digital processing are performed to extract the digital information carried in the reflected wave. Specifically, the reflected wave signal can be demodulated by synchronous detection method to obtain the demodulated digital information.

[0304] The specific implementation of step S30 is as follows:

[0305] First, the emission wave and reflected waves To perform time alignment, for example, the time delay of the reflected wave can be estimated by the least squares method ,Then align the transmitted and reflected waves.

[0306] Next, the sliding time window method is used to calculate the characteristic parameters between the reflected wave and the transmitted wave in each time window. Specifically, let the time window length be , the sliding step length is , then The characteristic parameters within a time window include:

[0307] Time Delay :

[0308] ;

[0309] Amplitude attenuation :

[0310] ;

[0311] Distinction :

[0312] ;

[0313] in, Distinction It is defined as 1 minus the correlation coefficient between the reflected wave and the transmitted wave in the sliding time window, reflecting the similarity between the two.

[0314] The specific implementation of step S40 is as follows:

[0315] The information obtained in steps S10 to S30, including the transmission wave , reflected wave and the time delay for each time window , Amplitude attenuation and Distinction , and input it into the pre-trained first reflection wave compensation model. After being processed by this model, a preliminary compensated reflection wave can be obtained. , recorded as the first compensation reflection wave.

[0316] The specific implementation of step S50 is as follows:

[0317] The seabed environmental parameters obtained in step S10 and the position information of the seafloor crawler , and the transmission wave in step S10 Perform time alignment to ensure time synchronization between various types of information.

[0318] The specific implementation of step S60 is as follows:

[0319] Using Bayesian network method to calculate seabed environmental parameters and the seafloor crawler position information The time delay with each time window , Amplitude attenuation and Distinction The correlation between them.

[0320] The specific implementation of step S70 is as follows:

[0321] According to the correlation calculated in step S60, a series of pre-set relationship functions are used to establish a mathematical model covering various interference factors such as water depth, water temperature, salinity, crawler posture, speed and acceleration, that is, an interference equation group.

[0322] Water depth interference equation:

[0323] ;

[0324] in, is the interference degree of water depth on electromagnetic waves, For water depth, For time, is the coefficient to be determined, is the first error term. Water depth It is directly measured by the depth sensor, Then it is calculated through continuous depth measurement:

[0325] ;

[0326] Water temperature interference equation:

[0327] ;

[0328] in, is the degree of interference of water temperature on electromagnetic waves, For water temperature, For depth, is the angular frequency of the temperature cycle, is the coefficient to be determined, is the second error term. Water temperature It is directly measured by the temperature sensor, By measuring the temperature at different depths and calculating:

[0329] ;

[0330] Salinity interference equation:

[0331] ;

[0332] in, is the degree of interference of salinity on electromagnetic waves, is the salinity, is the horizontal coordinate, is an imaginary unit, is the phase factor, is the coefficient to be determined, is the third error term. Salinity It is directly measured by the salinity sensor. and Then the salinity is measured at different horizontal positions and calculated:

[0333] ;

[0334] ;

[0335] Crawler pitch angle interference equation:

[0336] ;

[0337] in, is the interference degree of the pitch angle on the electromagnetic wave, is the pitch angle, For time, is the coefficient to be determined, is the fourth error term. Pitch angle It is directly measured by the inertial measurement unit, and Then it is calculated by continuous pitch angle measurement:

[0338] ;

[0339] Crawler roll angle interference equation:

[0340] ;

[0341] Crawler heading angle interference equation:

[0342] ;

[0343] in, is the interference degree of heading angle on electromagnetic waves, is the heading angle, For time, is an imaginary unit, is the coefficient to be determined, is the sixth error term. Heading angle It is directly measured by the inertial measurement unit, and Then it is calculated through continuous heading angle measurement:

[0344] ;

[0345] Crawler velocity disturbance equation:

[0346] ;

[0347] in, is the degree of interference of speed on electromagnetic waves, is the velocity vector, For time, is the coefficient to be determined, is the seventh error term. Speed It is calculated by the inertial measurement unit and GPS (if it is near the water surface), and Then the following is calculated by continuous speed measurement:

[0348] ;

[0349] Crawler acceleration disturbance equation:

[0350] ;

[0351] in, is the degree of interference of acceleration on electromagnetic waves, is the acceleration vector, For time, is the coefficient to be determined, is the eighth error term. Acceleration It is directly measured by the inertial measurement unit, and Then it is calculated through continuous acceleration measurement:

[0352] ;

[0353] The above interference equations cover various interference factors such as water depth, water temperature, salinity, crawler attitude (pitch angle, roll angle, heading angle), speed and acceleration. By establishing these mathematical models, we can better analyze and quantify the impact of various interference factors on electromagnetic wave propagation.

[0354] The specific implementation of step S80 is as follows:

[0355] Collect all data obtained in steps S10-S70, including the time delay of the sliding time window , Amplitude attenuation and Distinction , and seabed environmental parameters and the position information of the seafloor crawler .

[0356] According to the interference equations established in step S70, each interference factor is used as an independent variable, and the time delay , Amplitude attenuation and Distinction As the dependent variable, a fitting matrix is ​​constructed.

[0357] The least squares method is used to solve the fitting matrix and the weight coefficients of each interference factor are preliminarily obtained. Specifically, the loss function can be defined as:

[0358] ;

[0359] in, is the value predicted by the interference equations. By minimizing the loss function , the preliminary weight coefficients of each interference factor can be obtained.

[0360] To avoid overfitting, the results of the least squares method need to be regularized. L1 regularization or L2 regularization can be used to add a regularization term to the loss function:

[0361] ;

[0362] or

[0363] ;

[0364] in, is the weight coefficient of each interference factor, is the regularization parameter.

[0365] Use the k-fold cross-validation method to select the optimal regularization parameter , making the model perform best on the validation set.

[0366] Finally, all the data are fitted using the selected regularization parameters to obtain the final weight coefficients of each interference factor.

[0367] Finally, the final weight coefficients of each interference factor are output, including water depth , water temperature ,salinity , Crawler pitch angle , Roll Angle , heading angle ,speed and acceleration The weight coefficient of .

[0368] In general, the purpose of step S80 is to establish and fit the interference equation group to obtain the influence of various interference factors on electromagnetic wave propagation, which provides an important basis for the subsequent reflection wave compensation. This involves multiple steps such as data collection, fitting matrix construction, least squares solution, regularization processing, cross-validation and model evaluation.

[0369] The specific implementation of step S90 is as follows:

[0370] The first compensation reflected wave obtained in step S40 , the weight coefficients of the various interference factors obtained in step S80, and the seabed environmental parameters obtained in step S10 and the position information of the seafloor crawler , and are input together into the pre-trained second reflection wave compensation model.

[0371] Through the processing of this multi-branch neural network model, the final compensated reflected wave can be obtained. , as the result of electromagnetic detection. The model comprehensively considers the influence of seabed environmental parameters and crawler posture information on reflected waves, and further optimizes the preliminary compensation results, which can better suppress various interference factors and improve the accuracy of electromagnetic detection.

[0372] In order to further better understand and implement the present invention, the following provides Example 3 of a specific application scenario of the present invention:

[0373] A marine geological exploration organization decided to use the electromagnetic detection method based on the seafloor crawler proposed in this invention to conduct seafloor resource exploration in a sea area in the South China Sea. The water depth in this sea area is about 3,000 meters, the water temperature ranges from 4 to 8°C, the salinity is between 34 and 36‰, and the geological structure is relatively complex. In order to obtain detailed geological information of this sea area, the organization invested in a seafloor crawler and carried out relevant modifications and equipment integration on the equipment.

[0374] 1. Equipment integration of seafloor crawler

[0375] The main parameters of the seafloor crawler and the integrated sensor equipment are shown in the following table:

[0376] Table 1 Main parameters and integrated sensors of the seafloor crawler

[0377]

[0378] The seafloor crawler is wheel-driven and equipped with 6 independent drive wheels, which can maneuver flexibly in complex seafloor terrain. The crawler is integrated with a self-developed electromagnetic wave transmitter and receiver, which operates at a frequency of 500kHz and can transmit and receive digitally modulated electromagnetic wave signals. At the same time, the crawler is equipped with depth sensors, temperature sensors, salinity sensors, inertial measurement units and GPS receivers, which can collect seafloor environmental parameters and its own position information in real time. These sensor data will provide a basis for subsequent interference compensation.

[0379] 2. Seabed Exploration Implementation Process

[0380] The agency set up a 100km×100km detection area in a certain area of ​​the South China Sea and carried out detailed seabed topography mapping in the area to provide a basis for the crawler's path planning. Before officially starting the detection, the staff first conducted exploratory measurements at several representative points in the area and collected preliminary environmental parameters and electromagnetic wave reflection data to lay the foundation for training the compensation model.

[0381] After entering the formal detection phase, the seafloor crawler first moves slowly along the predetermined path, while recording its own dynamic information such as position, posture, speed and acceleration in real time. During the crawling process, the crawler will periodically emit electromagnetic waves and receive reflected wave signals. At the same time, the supporting environmental sensors will continue to collect parameters such as water depth, water temperature and salinity. These real-time measurement data will serve as input for subsequent interference analysis and compensation.

[0382] To ensure the time synchronization of the data, all sensor devices on the crawler are connected to a central control unit and time-calibrated via GPS. Every 5 seconds, the control unit records the data of all sensors to form a time series. The following is some of the data collected by the crawler during an actual detection:

[0383] Table 2 Part of the data collected in real time by the seabed crawler

[0384]

[0385] Through high-frequency sampling every 5 seconds, the crawler recorded about 72,000 time series data in the entire detection process. These data provide sufficient samples for subsequent interference analysis and reflection wave compensation.

[0386] 3. Interference analysis and compensation

[0387] With the above real-time collected environmental parameters and crawler posture data, we can then analyze and model various interference factors in the propagation of electromagnetic waves. The specific steps are as follows:

[0388] (1) Sliding window feature extraction

[0389] First, the transmitted wave and reflected waves Perform time alignment, and then use the sliding time window method to calculate the time delay between the reflected wave and the transmitted wave in each time window , Amplitude attenuation and Distinction These parameters can reflect the impact of various interference factors on the reflected wave. The following are the characteristic parameters of some time windows:

[0390] Table 3 Reflection wave characteristic parameters of some time windows

[0391]

[0392] (2) Interference factor modeling

[0393] Based on the characteristic parameters extracted in step (1) and the environmental parameters and posture information collected in real time in step 2, a mathematical model covering various interference factors, namely, a set of interference equations, can be established.

[0394] Taking the water depth interference equation as an example, according to the above formula:

[0395]

[0396] Substituting the measured water depth data into the least squares fitting method, we can get:

[0397]

[0398] in, is a random error term.

[0399] Similarly, the interference equations of water temperature, salinity, crawler attitude, velocity and acceleration are also established. These model parameters will provide a basis for the subsequent reflection wave compensation.

[0400] (3) Multi-level compensation

[0401] Based on the above interference analysis, a two-level neural network model is used to compensate for the reflected wave.

[0402] The first layer is a reflection wave compensation model with a multi-subnet structure. The model includes a feature extraction module, a long-term feature sub-network, a window feature sub-network, an abnormal feature sub-network, and a fusion output sub-network. Through the collaborative work of these sub-networks, the long-term change trend of the reflection wave, the characteristics within the local time window, and the abnormal pattern can be captured, and preliminary compensation results can be generated. The second layer is a multi-branch reflection wave compensation model. The model includes an environmental parameter branch network, a posture information branch network, a compensation reflection wave branch network, and a fusion branch network. The environmental parameter branch network and the posture information branch network are used to process the influence of environmental parameters and posture information on the reflection wave respectively. The compensation reflection wave branch network further optimizes the preliminary compensation results, and the fusion branch network integrates the outputs of each branch to generate the final compensation result.

[0403] Through this multi-level compensation mechanism, we can make full use of the real-time measured environmental parameters and posture information, and effectively suppress the influence of various interference factors on the electromagnetic detection results.

[0404] 4. Experimental results and analysis

[0405] Based on the above implementation steps, a one-week electromagnetic detection test was conducted in a certain area of ​​the South China Sea. During the detection process, the seabed crawler traveled a total of 20 km and collected a large amount of environmental parameters, position information, and electromagnetic wave emission and reflection data.

[0406] In order to verify the performance of the method of the present invention, several typical geological structure points in the sea area were selected, and the detection results obtained by using the method of the present invention and the traditional method (without considering interference factors) were compared and analyzed. The following is the detection result of one of the representative locations:

[0407] Table 4 Comparison of detection results using different methods

[0408]

[0409] As can be seen from Table 4, the thickness of the sediment layer obtained by the method of the present invention is 30 meters higher than that of the traditional method, and signs of oil and gas are found, which is consistent with the geological structure characteristics of the sea area. Further analysis shows that the method of the present invention can effectively suppress the influence of interference factors such as water depth, water temperature, salinity and crawler posture changes, and the detection results are more accurate and reliable.

[0410] In order to comprehensively evaluate the performance of the method of the present invention, 20 representative locations were selected in the entire detection area, and the traditional method and the method of the present invention were used for comparative analysis. The results showed that the detection accuracy of the method of the present invention was improved by an average of 23.7% compared with the traditional method. At the same time, the method of the present invention can better identify the lithology, structural characteristics and resource signs of the seabed, providing a more reliable basis for subsequent marine geological exploration.

[0411] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An electromagnetic detection system based on a seafloor crawler, characterized in that: include: A survey ship, a seabed crawler, a transmitting antenna, a receiving node array, a control chip and a power supply, wherein the survey ship is connected to the seabed crawler through a towing rope, a towing cable is provided at the tail of the seabed crawler, the transmitting antenna and the receiving node array are fixed on the towing cable, and the control chip and the power supply are both provided in an integrated electronic cabin of the seabed crawler; the power supply provides electrical energy for the seabed crawler, the transmitting antenna, the receiving node array and the control chip, and the control chip is electrically connected to the transmitting antenna, the receiving node array and the power system of the seabed crawler; an electromagnetic detection interference compensation module is provided in the control chip, which is used to eliminate and compensate for interference with reflected waves received by the receiving node array; the control chip is provided with a data transmission device, which is used to send the compensated electromagnetic detection results to the survey ship; the electromagnetic detection interference compensation module specifically performs the following steps: S10, controlling the seafloor crawler to crawl in a predetermined detection area, obtaining seafloor environmental parameters, position information of the seafloor crawler, and simultaneously transmitting electromagnetic waves and receiving reflected waves, wherein the seafloor environmental parameters include water depth, water temperature, and salinity; wherein the transmitted electromagnetic waves are loaded with digital information in a modulation manner, which is recorded as transmitted waves; S20, demodulating and digitizing the reflected wave to obtain demodulated digital information; S30, aligning the reflected wave with the transmitted wave according to the demodulated digital information, and calculating the time delay, amplitude attenuation and distinction of each sliding time window of the reflected wave by using a sliding time window, wherein the distinction is defined as: 1-the similarity between the reflected wave and the transmitted wave within the sliding time window; S40, inputting the transmission wave, the reflected wave and the time delay, amplitude attenuation and discrimination of each sliding time window into a pre-trained first reflected wave compensation model to obtain a first compensated reflected wave; S50, aligning the seabed environmental parameters and the position information of the seabed crawler with the transmission wave according to time; S60, using a Bayesian network to calculate the correlation between the seabed environmental parameters and the position information of the seabed crawler and the time delay, amplitude attenuation and discrimination of each sliding time window; S70, according to the calculated correlation, using a preset relationship function combination, to establish an interference equation group, including a water depth interference equation, a water temperature interference equation, a salinity interference equation, a crawler pitch angle interference equation, a crawler roll angle interference equation, a crawler heading angle interference equation, a crawler speed interference equation, and a crawler acceleration interference equation; S80, fitting the interference equation group and calculating the weight coefficient of each interference factor; the interference factors include seabed environmental parameters and position information of the seabed crawler; S90, input the first compensated reflected wave, the weight coefficients of various interference factors, the seabed environmental parameters and the position information of the seabed crawler into a pre-trained second reflected wave compensation model to obtain the final compensated reflected wave as the electromagnetic detection result.

2. The electromagnetic detection system based on the seafloor crawler according to claim 1 is characterized in that: The seabed crawler is a crawler-type seabed crawler.

3. The electromagnetic detection system based on the seafloor crawler according to claim 2 is characterized in that: The data transmission device sends the compensated electromagnetic detection result to the survey ship via a data cable or a wireless channel.

4. The electromagnetic detection system based on the seafloor crawler according to claim 1 is characterized in that: The receiving node array is specifically a linear or two-dimensional array composed of multiple electromagnetic induction coils, and each receiving node includes a three-axis orthogonal induction coil for receiving electromagnetic field signals in different directions.

5. The electromagnetic detection system based on the seafloor crawler according to claim 4 is characterized in that: The transmitting antenna is a negative buoyancy antenna.

6. The electromagnetic detection system based on the seafloor crawler according to claim 5 is characterized in that: The module of the seafloor crawler at least includes a depth meter and an altimeter.

7. The electromagnetic detection system based on the seafloor crawler according to claim 6 is characterized in that: It also includes an excitation source, which is fixedly arranged on the seabed crawler.

8. The electromagnetic detection system based on the seafloor crawler according to claim 7, characterized in that: The excitation source is specifically a dipole transmitter composed of a pair of large-area electrode plates, which can generate a low-frequency alternating current to excite an electromagnetic field in the seabed medium.

9. The electromagnetic detection system based on the seafloor crawler according to claim 8, characterized in that: The survey vessel is equipped with a satellite navigation system, an underwater positioning system, and a power supply unit.

Images