Ground and in-well time-frequency electromagnetic measurement system and method based on AI big data model

Abstract

The invention discloses a ground and in-well time-frequency electromagnetic measurement system and method based on an AI big data model, and the method comprises the steps: processing an artificial big data model through AI-trained time-frequency electromagnetic data, and carrying out the AI processing of the time-frequency electromagnetic data evaporated from the artificial big data model; in combination with a three-dimensional resistivity model and a polarizability model of a time-frequency electromagnetic data acquisition area established through underground three-dimensional resistivity and polarizability modeling software, AI processing is performed on time-frequency electromagnetic data acquired on the ground or / and underground, and an underground real three-dimensional resistivity and polarizability model is obtained through an AI-based three-dimensional inversion technology. Based on the distribution and change rules of the high-resistivity anomalous body and the high-polarizability anomalous body in the underground reservoir three-dimensional space, the oil-gas saturation occurrence state and the oil-water or gas-water interface migration change state in the reservoir space are judged respectively; comprehensive evaluation on oil and gas resources of an underground reservoir is realized or residual oil and gas and residual oil and gas resources are searched in an oil and gas resource development area.

Description

Technical Field

[0001] The present invention relates to the application of AI artificial intelligence large model technology in the field of time-frequency electromagnetic exploration technology. More specifically, it relates to a ground and borehole time-frequency electromagnetic measurement system and method based on an AI big data model. Background Art

[0002] Artificial Intelligence (AI) is an important driving force for the new round of scientific and technological revolution and industrial transformation. It is a new technical science that studies, develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence. Artificial Intelligence (AI) is an interdisciplinary and emerging discipline that is based on computer science and is cross-integrated by multiple disciplines such as computer science, psychology, and philosophy. It is a new technical science that studies, develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence, attempting to understand the essence of intelligence and produce a new intelligent machine that can respond in a way similar to human intelligence. The research in this field includes robots, speech recognition, image recognition, natural language processing, and expert systems, etc. Artificial Intelligence is an important part of the intelligent discipline, attempting to understand the essence of intelligence and produce a new intelligent machine that can respond in a way similar to human intelligence. Artificial Intelligence is a very broad science, including robots, speech recognition, image recognition, natural language processing, expert systems, machine learning, computer vision, etc.

[0003] A large model refers to a machine learning model with large-scale parameters and a complex computing structure. These models are usually constructed by deep neural networks and have billions or even hundreds of billions of parameters. The design purpose of large models is to improve the model's expressive ability and prediction performance, and they can handle more complex tasks and data. Large models have a wide range of applications in various fields, including natural language processing, computer vision, speech recognition, and recommendation systems, etc. Large models learn complex patterns and features by training on massive amounts of data and have a stronger generalization ability, enabling them to make accurate predictions on unseen data.

[0004] Big data analysis refers to the process of processing, analyzing, and mining large, high-speed, multi-source, and multi-type data to discover valuable information and knowledge. The core technologies of big data analysis include data storage, data processing, data mining, data analysis, and data visualization, etc. With the development of technologies such as the Internet, artificial intelligence, and the Internet of Things, the scale and complexity of data have been continuously increasing, and big data analysis technology has become an indispensable part of enterprises and organizations.

[0005] In the field of deep learning, model compression and deployment is an important research topic, and model distillation is one of the effective methods. Model Distillation was initially proposed by Hinton et al. in 2015. Its core idea is to transfer the knowledge of a complex large model (teacher model) to a relatively simple small model (student model) through knowledge transfer. Briefly, it uses the predicted probability distribution of the teacher model as soft labels to train the student model, thus greatly reducing the complexity and computational resource requirements of the model while maintaining high prediction performance, achieving the lightweight and high efficiency of the model.

[0006] The time-frequency electromagnetic method is a new method emerging in the field of petroleum exploration. It adopts a working mode similar to that of large-offset seismic exploration, supplies a strong current to the earth to stimulate oil and gas exploration targets, and measures the secondary electromagnetic field and electromagnetic field spectrum formed by the discharge of pore media in the oil and gas reservoir; this technology simultaneously obtains time-domain and frequency-domain signals, and through the joint processing of time-domain and frequency-domain signals, accurately reconstructs the underground physical property model and obtains the resistivity and polarization rate anomalies of oil and gas exploration targets.

[0007] The time-frequency electromagnetic technology combines frequency-domain sounding and time-domain sounding in one system. It can select different frequencies and different types of excitation waveforms according to the depth of the exploration target, and can not only provide resistivity information but also provide induced polarization information. Therefore, it can detect its oil and gas content while studying the electrical structure. The time-domain sounding processing uses quasi-two-dimensional resistivity inversion to obtain resistivity information, and the frequency-domain sounding processing introduces the Cole-Cole model to extract induced polarization information. The time-frequency electromagnetic method uses a large device to conduct sounding of targets at different depths by changing the waveform length and frequency.

[0008] For the field construction of the time-frequency electromagnetic method, the equatorial dipole device is adopted, which is divided into two parts: transmitting and receiving. The transmitting field source consists of multiple parallel copper wires to form a finite-length wire source grounded at both ends. A high-power transmitter is used to send a series of square-wave currents with different periods to the ground at different frequencies. At the receiving end, the electric component E is measured through the grounded dipole MN X and the vertical magnetic induction component (dBz / dt) is measured by a highly sensitive magnetic rod. Summary of the Invention

[0009] The purpose of the present invention is to provide a surface and borehole time-frequency electromagnetic measurement system and data processing method based on an AI big data model, an artificial big data model for time-frequency electromagnetic data processing trained by AI, and a time-frequency electromagnetic data AI processing model extracted from the artificial big data model for time-frequency electromagnetic data processing. By combining a three-dimensional resistivity model and a three-dimensional polarization rate model of the underground in the time-frequency electromagnetic data acquisition work area established by an AI-based underground three-dimensional resistivity and polarization rate modeling software, AI processing is performed on the time-frequency electromagnetic data collected on the surface or / and in the borehole. Through an AI-based three-dimensional inversion technology, the true underground three-dimensional resistivity model and the true underground three-dimensional polarization rate model corresponding to the collected time-frequency electromagnetic data are obtained, and a comprehensive evaluation of the oil and gas resources in the underground reservoir is carried out, or residual oil and gas and remaining oil and gas resources are searched for in the oil and gas resource development area.

[0010] The above technical object of the present invention is achieved through the following technical solutions:

[0011] A surface and borehole time-frequency electromagnetic measurement system based on an AI big data model includes surface time-frequency electromagnetic data acquisition stations arranged in a grid pattern on the ground of the work area, borehole array-type time-frequency electromagnetic data acquisition short joints arranged in the borehole, and a surface current source emission device; the surface current source emission device is two current source emission antennas arranged in parallel on both sides of the periphery of the work area or four current source emission antennas arranged around the periphery of the work area, or a circular or rectangular emission coil surrounding the periphery of the work area, or a dipole current source emission antenna arranged at the wellhead and far from the wellhead. It also includes a high-power time-frequency electromagnetic emission source connected to the surface current source emission device.

[0012] The surface time-frequency electromagnetic data acquisition stations and the borehole array-type time-frequency electromagnetic data acquisition short joints are connected to a surface time-frequency electromagnetic data acquisition control and processing computer workstation through cables or optoelectronic composite cables; an artificial big data model for time-frequency electromagnetic data processing trained by AI and a time-frequency electromagnetic data AI processing model extracted from the artificial big data model for time-frequency electromagnetic data processing are installed in the surface time-frequency electromagnetic data acquisition control and processing computer workstation; by using an AI-based underground three-dimensional resistivity and polarization rate modeling software, inputting the three-dimensional geological and structural model, three-dimensional resistivity model, three-dimensional polarization rate model, all resistivity logging data, other surface or / and borehole time-frequency electromagnetic data, and resistivity and polarization rate measurement data of core specimens in the time-frequency electromagnetic data acquisition work area, a three-dimensional resistivity model and a three-dimensional polarization rate model of the underground in the time-frequency electromagnetic data acquisition work area are established;

[0013] The surface time-frequency electromagnetic data acquisition station may be a surface time-frequency electromagnetic data acquisition station with a wired or wireless conventional two-component electric field sensor plus a three-component magnetic field sensor, or an optical fiber surface time-frequency electromagnetic data acquisition station with an optical fiber three-component electric field sensor plus an optical fiber three-component magnetic field sensor. The in-well array-type time-frequency electromagnetic data acquisition short section arranged in the well may be an array-type conventional time-frequency electromagnetic data acquisition short section composed of a pair of conventional vertical electric field sensors plus a three-component magnetic field sensor, where the pair of conventional vertical electric field sensors are arranged at the upper and lower ends of the three-component magnetic field sensor, or an array-type optical fiber time-frequency electromagnetic data acquisition short section composed of an optical fiber three-component electric field sensor plus an optical fiber three-component magnetic field sensor.

[0014] A high-power time-frequency electromagnetic transmitter is connected to the middle of each of the current source transmitting antennas, and both ends of the current source transmitting antennas are grounded through copper rod electrodes. A high-power time-frequency electromagnetic transmitter is connected to the tail end of each of the transmitting coils.

[0015] One end of the dipole current source transmitting antenna is connected to the metal casing at the wellhead, and the other end is located several kilometers away from the wellhead and is grounded through a copper rod electrode. A high-power time-frequency electromagnetic transmitter is connected between the end connected to the metal casing at the wellhead and the end far from the wellhead of the dipole current source transmitting antenna.

[0016] The surface time-frequency electromagnetic data acquisition station and the in-well array-type time-frequency electromagnetic data acquisition short section are both used to separately collect the time-frequency electromagnetic signals generated by the current source transmitting antenna, the transmitting coil on the surface, or the dipole current source transmitting antenna arranged on the wellhead casing and at the far end.

[0017] The current source transmitting antenna is a large-current power supply wire with a length of 5 km to 10 km, and copper rod electrodes for grounding are connected to both ends of the current source transmitting antenna.

[0018] The conventional surface time-frequency electromagnetic data acquisition station on the surface includes two pairs of mutually orthogonal and vertical conventional non-polarizing electric field sensor pairs, and also includes a three-component magnetic field sensor that is mutually orthogonal and vertical. The three-component magnetic field sensor is located at the center position of the two groups of non-polarizing electric field sensor pairs. For each of the conventional non-polarizing electric field sensor pairs, it can be one of copper sulfate, silver chloride, nanomaterials, tantalum capacitor non-polarizing electrodes. The distance between the two electric field sensors in the conventional non-polarizing electric field sensor pair is 10 m to 100 m.

[0019] The three-component magnetic field sensors of multiple said ground conventional time-frequency electromagnetic data acquisition stations and downhole arrayed conventional time-frequency electromagnetic data acquisition short joints are one of an induction magnetic field sensor, a fluxgate magnetic field sensor, a MEMS magnetic field sensor, a cold atom magnetic field sensor, and a superconducting magnetic field sensor; the three-component electric field sensors of the ground fiber optic ground time-frequency electromagnetic data acquisition stations and downhole arrayed fiber optic time-frequency electromagnetic data acquisition short joints are three-component fiber optic electric field sensors, and the three-component magnetic field sensors are three-component fiber optic magnetic field sensors.

[0020] When a wired conventional ground time-frequency electromagnetic data acquisition station is deployed on the ground or / and an arrayed conventional time-frequency electromagnetic data acquisition short joint is deployed downhole, it further includes a connection cable, and a ground time-frequency electromagnetic data acquisition control and processing computer workstation for controlling and receiving the data acquired by the wired conventional ground time-frequency electromagnetic data acquisition station on the ground or / and the arrayed conventional time-frequency electromagnetic data acquisition short joint downhole. When a fiber optic ground time-frequency electromagnetic data acquisition station is deployed on the ground or / and an arrayed fiber optic time-frequency electromagnetic data acquisition short joint is deployed downhole, it further includes a connection optical cable, and a fiber optic time-frequency electromagnetic data acquisition control, modulation, and demodulation instrument for controlling and receiving the data acquired by the fiber optic ground time-frequency electromagnetic data acquisition station on the ground or / and the arrayed fiber optic time-frequency electromagnetic data acquisition short joint downhole.

[0021] A three-component electronic attitude sensor or a three-component fiber optic attitude sensor is placed inside each of the ground conventional ground time-frequency electromagnetic data acquisition station or the ground fiber optic ground time-frequency electromagnetic data acquisition station, and the downhole arrayed conventional time-frequency electromagnetic data acquisition short joint or the downhole arrayed fiber optic time-frequency electromagnetic data acquisition short joint.

[0022] The AI-trained time-frequency electromagnetic data processing artificial big data model uses various underground three-dimensional geological and structural models, three-dimensional resistivity models, three-dimensional polarization models, and according to the deployment methods of the ground time-frequency electromagnetic data acquisition stations, the downhole arrayed time-frequency electromagnetic data acquisition short joints in the well, two current source transmitting antennas arranged in parallel on both sides of the periphery of the work area or four current source transmitting antennas arranged around the periphery of the work area, or a circular or rectangular transmitting coil surrounding the periphery of the work area, and dipole current source transmitting antennas arranged at the wellhead and away from the wellhead, to forward simulate the time-frequency electromagnetic data recorded on the ground or / and in the well for any underground geological structure and geoelectric model, as well as other existing measured ground or / and well time-frequency electromagnetic data, and then conducts AI training on the time-frequency electromagnetic data processing artificial big data model.

[0023] The time-frequency electromagnetic data AI processing model is obtained by distillation on the basis of the time-frequency electromagnetic data processing artificial big data model, and is used to perform AI processing on the acquired ground or / and well time-frequency electromagnetic data in the ground time-frequency electromagnetic data acquisition control and processing computer workstation.

[0024] Data acquisition and processing method for surface and borehole time-frequency electromagnetic measurement system based on AI big data model, including the following specific steps:

[0025] (1) Arrange the surface current source emission device, and use two current source emission antennas arranged in parallel on both sides of the periphery of the work area or four current source emission antennas arranged around the periphery of the work area to form a surface or borehole time-frequency electromagnetic data acquisition area around the work area. The middle of each current source emission antenna is connected to a high-power time-frequency electromagnetic emission source, and both ends of each current source emission antenna are grounded through copper rod electrodes; or use an emission coil;

[0026] (2) For the acquisition of borehole-to-surface time-frequency electromagnetic data, connect one end of the dipole current source emission antenna to the metal casing at the wellhead, and the other end of the dipole current source emission antenna is arranged at a position several kilometers away from the wellhead and grounded through a copper rod electrode. The middle of the dipole current source emission antenna is connected to a high-power time-frequency electromagnetic emission source;

[0027] (3) Arrange multiple surface time-frequency electromagnetic data acquisition stations in the time-frequency electromagnetic data acquisition area in a designed grid pattern or centered on the well. For each surface time-frequency electromagnetic data acquisition station, arrange two sets of non-polarized electric field sensor pairs in the north-south direction and the east-west direction or in mutually perpendicular orthogonal directions, and arrange mutually orthogonal three-component magnetic field sensors at the center positions of the two sets of non-polarized electric field sensor pairs;

[0028] (4) For the acquisition of borehole time-frequency electromagnetic data, arrange the borehole array-type conventional time-frequency electromagnetic data acquisition short section or the borehole array-type fiber optic time-frequency electromagnetic data acquisition short section to the depth position near the reservoir at the bottom of the well through armored cable or armored optical cable;

[0029] (5) Connect the wired conventional surface time-frequency electromagnetic data acquisition stations arranged on the surface or / and the array-type conventional time-frequency electromagnetic data acquisition short sections arranged underground to the time-frequency electromagnetic data acquisition control system through connecting cables; connect the fiber optic surface time-frequency electromagnetic data acquisition stations arranged on the surface or / and the array-type fiber optic time-frequency electromagnetic data acquisition short sections arranged underground to the fiber optic time-frequency electromagnetic data acquisition control modulation and demodulation instrument through connecting optical cables;

[0030] (6) Use the underground three-dimensional resistivity and polarizability modeling software based on AI, input the three-dimensional geological and structural model, three-dimensional resistivity model, three-dimensional polarizability model, all resistivity logging data, other surface or / and underground time-frequency electromagnetic data, and resistivity and polarizability measurement data of core specimens in the time-frequency electromagnetic data acquisition work area, and establish a three-dimensional resistivity model and a three-dimensional polarizability model underground in the time-frequency electromagnetic data acquisition work area;

[0031] (7) Smooth all logging data in the time-frequency electromagnetic data acquisition work area to obtain the porosity, resistivity, polarizability, and hydrocarbon saturation of the fluid in the reservoir of the underground rock formation along the well trajectory. Establish the relationship between the hydrocarbon saturation, resistivity, and polarizability in the reservoir through AI processing.

[0032] (8) Start the high-power time-frequency electromagnetic emission source and the ground time-frequency electromagnetic data acquisition control and processing computer workstation. Use two non-polarizing electric field sensors on the ground to obtain the two horizontal electric field component data of the time-frequency electromagnetic field excited by the high-power time-frequency electromagnetic emission source, and use a ground three-component magnetic field sensor to obtain the three-component magnetic field data of the underground time-frequency electromagnetic field. The ground time-frequency electromagnetic signal includes two mutually orthogonal horizontal electric field component data and three mutually orthogonal three-component magnetic field data.

[0033] (9) Start the high-power time-frequency electromagnetic emission source and the ground time-frequency electromagnetic data acquisition control and processing computer workstation. Use two vertical non-polarizing electric field sensors in the well to obtain the vertical electric field component data of the time-frequency electromagnetic field excited by the high-power time-frequency electromagnetic emission source, and use a downhole three-component magnetic field sensor to obtain the three-component magnetic field data of the underground time-frequency electromagnetic field. The downhole time-frequency electromagnetic signal includes vertical electric field component data and three mutually orthogonal three-component magnetic field data.

[0034] (10) When a fiber optic ground time-frequency electromagnetic data acquisition station is deployed on the ground or / and an arrayed fiber optic time-frequency electromagnetic data acquisition short section is deployed in the well, the fiber optic time-frequency electromagnetic data acquisition control, modulation, and demodulation instrument is connected to the fiber optic ground time-frequency electromagnetic data acquisition station or / and the arrayed fiber optic time-frequency electromagnetic data acquisition short section through a connecting optical cable.

[0035] (11) Start the fiber optic time-frequency electromagnetic data acquisition control, modulation, and demodulation instrument, and start collecting the three mutually orthogonal three-component electric field component data of the time-frequency electromagnetic field on the ground or / and in the well and the mutually orthogonal three-component magnetic field data of the time-frequency electromagnetic field.

[0036] (12) Synchronously start the ground time-frequency electromagnetic data acquisition control and processing computer workstation installed with the time-frequency electromagnetic data processing artificial big data model and the time-frequency electromagnetic data AI processing model trained with AI.

[0037] (13) Perform on-site processing of the ground or / and downhole time-frequency electromagnetic data collected in the time-frequency data acquisition work area in the ground time-frequency electromagnetic data acquisition control and processing computer workstation.

[0038] (14) Use the AI - based time - frequency electromagnetic data pre - processing software installed in the ground time - frequency electromagnetic data acquisition control and processing computer workstation to perform denoising processing on the collected ground or / and downhole time - frequency electromagnetic data, consistency processing of surface or / and downhole excitation sources, and consistency processing of surface or / and downhole time - frequency electromagnetic data;

[0039] (15) Input the time - frequency electromagnetic data processed in step (14) and the three - dimensional resistivity model and three - dimensional polarizability model established in step (6) into the distilled time - frequency electromagnetic data processing artificial big data model;

[0040] (16) The time - frequency electromagnetic data processing artificial big data model in the ground time - frequency electromagnetic data acquisition control and processing computer workstation, based on the time - frequency electromagnetic data processed in step (14), combined with the three - dimensional resistivity model and three - dimensional polarizability model established in step (6), obtains the true underground three - dimensional resistivity model and true underground three - dimensional polarizability model corresponding to the collected time - frequency electromagnetic data through AI - based three - dimensional inversion technology;

[0041] (17) When performing the AI - based three - dimensional inversion processing in step (16), strongly constrain and calibrate the inversion results in step (16) at the well trajectory position using the burial depth and thickness parameters of the reservoir geoelectric model established in step (6) and the relationship between the hydrocarbon - bearing saturation in the reservoir obtained in step (7) and resistivity and polarizability;

[0042] (18) According to the accurate resistivity distribution and accurate polarizability distribution of the underground three - dimensional reservoir medium, combined with the true underground three - dimensional resistivity model and true underground three - dimensional polarizability model corresponding to the time - frequency electromagnetic data obtained in step (16), and based on the distribution and variation laws of high - resistivity anomalies and high - polarizability anomalies in the underground three - dimensional reservoir space, respectively judge the occurrence state of hydrocarbon - bearing saturation and the migration and change state of the oil - water interface or gas - water interface in the underground three - dimensional reservoir space in the time - frequency electromagnetic data acquisition work area. Use the relationship between the hydrocarbon - bearing saturation in the reservoir and resistivity and polarizability in step (7), combined with the high - resistivity anomaly and high - polarizability anomaly data and reservoir porosity parameters inverted in step (17) to calculate the total content of oil and gas in the reservoir, fully evaluate the underground oil and gas resource prospects in the oil and gas exploration area, comprehensively investigate the residual or remaining oil and gas resources in the underground reservoir in the oil and gas resource production area, and accordingly optimize the layout of infill wells or adjustment wells. Description of the Drawings

[0043] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, form a part of this application, and do not limit the embodiments of the present invention. In the drawings:

[0044] Figure 1Schematic layout diagram of the four-direction excitation surface and downhole wired time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0045] Figure 2 Schematic layout diagram of the four-direction excitation surface wireless node type time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0046] Figure 3 Schematic layout diagram of the two-direction excitation surface and downhole wired time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0047] Figure 4 Schematic layout diagram of the two-direction excitation surface wireless node type time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0048] Figure 5 Schematic layout diagram of the loop coil excitation surface and downhole wired time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0049] Figure 6 Schematic layout diagram of the loop coil excitation surface wireless node type time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0050] Figure 7 Schematic layout diagram of the four-direction and downhole casing excitation surface wireless node type time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0051] Figure 8 Schematic layout diagram of the four-direction excitation surface and downhole optical fiber time-frequency electromagnetic data measurement system in the embodiment of the present invention;

[0052] Figure 9 Schematic layout diagram of the loop coil excitation surface and downhole optical fiber time-frequency electromagnetic data measurement system in the embodiment of the present invention.

[0053] Marks in the drawings and corresponding component names:

[0054] 1 - Surface time-frequency electromagnetic data acquisition station 1, 2 - Downhole array type time-frequency electromagnetic data acquisition short joint, 3 - Current source transmitting antenna, 4 - Transmitting coil, 5 - Dipole current source transmitting antenna, 6 - High-power time-frequency electromagnetic transmitting source, 7 - Surface time-frequency electromagnetic data acquisition control and processing computer workstation, 8 - Time-frequency electromagnetic data processing artificial big data model, 9 - Time-frequency electromagnetic data AI processing model, 10 - Three-dimensional resistivity model and three-dimensional polarization rate model, 11 - Copper rod electrode, 12 - Optical fiber time-frequency electromagnetic data acquisition control modulation and demodulation instrument. Detailed implementation manners

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. Components of the embodiments of the present invention usually described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations.

[0056] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0057] It should be noted that like reference numerals and letters denote like items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0058] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed during use, it is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0059] In addition, terms such as "horizontal", "vertical", "hanging" do not mean that the component is required to be absolutely horizontal or hanging, but can be slightly inclined. For example, "horizontal" only means that its direction is more horizontal relative to "vertical", and does not mean that the structure must be completely horizontal, but can be slightly inclined.

[0060] In the description of the embodiments of the present invention, "a plurality of" represents at least two.

[0061] In the description of the embodiments of the present invention, it should also be noted that unless otherwise clearly specified and limited, if terms such as "set", "installed", "connected", "connected" are understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0062] Example 1: Schematic layout diagram of a four-direction excitation surface and downhole wired time-frequency electromagnetic data measurement system, as shown in Figure 1 .

[0063] Example 2: Schematic layout diagram of a four-direction excitation surface wireless node type time-frequency electromagnetic data measurement system, as shown in Figure 2 .

[0064] Example 3: Schematic layout diagram of a two-direction excitation surface and downhole wired time-frequency electromagnetic data measurement system, as shown in Figure 3 .

[0065] Example 4: Schematic layout diagram of a two-direction excitation surface wireless node type time-frequency electromagnetic data measurement system, as shown in Figure 4 .

[0066] Example 5: Figure 5 This is the schematic layout diagram of the loop coil excitation surface and downhole wired time-frequency electromagnetic data measurement system in the embodiment of the present invention, as shown in Figure 5 .

[0067] Example 6: Schematic layout diagram of the loop coil excitation surface wireless node type time-frequency electromagnetic data measurement system, as shown in Figure 6 .

[0068] Example 7: Schematic layout diagram of the four-direction and downhole casing excitation surface wireless node type time-frequency electromagnetic data measurement system, as shown in Figure 7 .

[0069] Example 8: Schematic layout diagram of the four-direction excitation surface and downhole fiber optic time-frequency electromagnetic data measurement system, as shown in Figure 8 .

[0070] Example 9: Schematic layout diagram of the loop coil excitation surface and downhole fiber optic time-frequency electromagnetic data measurement system, as shown in Figure 9 .

[0071] This embodiment provides a surface and downhole time-frequency electromagnetic measurement system based on an AI big data model, including a surface time-frequency electromagnetic data acquisition station 1 arranged in a grid pattern on the surface of the work area, a downhole array type time-frequency electromagnetic data acquisition short joint 2 arranged in the well, and a surface current source emission device;

[0072] The surface current source emission device is two current source emission antennas 3 arranged in parallel on both sides of the periphery of the work area or four current source emission antennas 3 arranged around the periphery of the work area, or a circular or rectangular emission coil 4 around the periphery of the work area, or a dipole current source emission antenna 5 arranged at the wellhead and away from the wellhead;

[0073] It also includes a high-power time-frequency electromagnetic emission source 6 connected to the ground current source emission device. The ground time-frequency electromagnetic data acquisition station 1 and the downhole array time-frequency electromagnetic data acquisition short section 2 are connected to the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7 through cables or fiber-optic composite cables. An artificial big data model 8 for processing time-frequency electromagnetic data trained by AI is installed in the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7, and there is also a time-frequency electromagnetic data AI processing model 9 extracted from the artificial big data model 8 for processing time-frequency electromagnetic data. Using AI-based underground three-dimensional resistivity and polarizability modeling software, inputting the three-dimensional geological and structural model, three-dimensional resistivity model, three-dimensional polarizability model, all resistivity logging data, other ground or / and downhole time-frequency electromagnetic data, and resistivity and polarizability measurement data of core specimens in the time-frequency electromagnetic data acquisition work area, a three-dimensional resistivity model and a three-dimensional polarizability model 10 underground in the time-frequency electromagnetic data acquisition work area are established.

[0074] The ground time-frequency electromagnetic data acquisition station 1 can be a ground time-frequency electromagnetic data acquisition station with a wired or wireless conventional two-component electric field sensor plus a three-component magnetic field sensor, or a fiber-optic ground time-frequency electromagnetic data acquisition station with a fiber-optic three-component electric field sensor plus a fiber-optic three-component magnetic field sensor.

[0075] The downhole array time-frequency electromagnetic data acquisition short section 2 can be an array-type conventional time-frequency electromagnetic data acquisition short section composed of a pair of conventional vertical electric field sensors plus a three-component magnetic field sensor, with the pair of conventional vertical electric field sensors arranged at the upper and lower ends of the three-component magnetic field sensor, or an array-type fiber-optic time-frequency electromagnetic data acquisition short section composed of a fiber-optic three-component electric field sensor plus a fiber-optic three-component magnetic field sensor.

[0076] For the ground current source emission device, a high-power time-frequency electromagnetic emission source 6 is connected to the middle of each current source emission antenna 3, and both ends of the current source emission antenna 3 are grounded through copper rod electrodes 11.

[0077] The tail ends of the transmitting coils 4 are all connected to a high-power time-frequency electromagnetic emission source 6.

[0078] One end of the dipole current source emission antenna 5 is connected to the metal casing of the wellhead, and the other end is located several kilometers away from the wellhead and grounded through a copper rod electrode 11. A high-power time-frequency electromagnetic emission source 6 is connected between the end of the dipole current source emission antenna 5 connected to the metal casing of the wellhead and the end far from the wellhead.

[0079] Both the ground time-frequency electromagnetic data acquisition station 1 and the downhole array time-frequency electromagnetic data acquisition short section 2 are used to collect the time-frequency electromagnetic signals generated underground by the ground current source emission device respectively.

[0080] The current source transmitting antenna is a large current supply wire with a length ranging from 5 km to 10 km, and copper rod electrodes 11 for grounding are connected to both ends of the current source transmitting antenna 5.

[0081] When the ground time-frequency electromagnetic data acquisition station 1 is a conventional ground time-frequency electromagnetic data acquisition station on the ground, it includes two pairs of mutually orthogonal and perpendicular conventional non-polarized electric field sensor pairs, and also includes a three-component magnetic field sensor that is mutually orthogonal and perpendicular. The three-component magnetic field sensor is located at the center position of the two non-polarized electric field sensor pairs. For each of the conventional non-polarized electric field sensor pairs, it can be one of copper sulfate, silver chloride, nanomaterials, tantalum capacitor non-polarized electrodes, and the distance between the two electric field sensors in the conventional non-polarized electric field sensor pair is from 10 m to 100 m.

[0082] When the ground time-frequency electromagnetic data acquisition station 1 is a conventional ground time-frequency electromagnetic data acquisition station on the ground and the downhole arrayed time-frequency electromagnetic data acquisition short section 2 is a downhole arrayed conventional time-frequency electromagnetic data acquisition short section, the three-component magnetic field sensor is one of an induction magnetic field sensor, a fluxgate magnetic field sensor, a MEMS magnetic field sensor, a cold atom magnetic field sensor, a superconducting magnetic field sensor; when the ground time-frequency electromagnetic data acquisition station 1 is a ground fiber optic ground time-frequency electromagnetic data acquisition station and the downhole arrayed time-frequency electromagnetic data acquisition short section 2 is a downhole arrayed fiber optic time-frequency electromagnetic data acquisition short section, the three-component electric field sensor is a three-component fiber optic electric field sensor, and the three-component magnetic field sensor is a three-component fiber optic magnetic field sensor.

[0083] When the ground time-frequency electromagnetic data acquisition station 1 is a ground-deployed conventional ground time-frequency electromagnetic data acquisition station and the downhole arrayed time-frequency electromagnetic data acquisition short section 2 is a downhole arrayed conventional time-frequency electromagnetic data acquisition short section, it is directly connected to the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7; when the ground time-frequency electromagnetic data acquisition station 1 is a ground fiber optic ground time-frequency electromagnetic data acquisition station and the downhole arrayed time-frequency electromagnetic data acquisition short section 2 is a downhole arrayed fiber optic time-frequency electromagnetic data acquisition short section, it further includes a fiber optic time-frequency electromagnetic data acquisition control modulation and demodulation instrument 12.

[0084] Attitude sensors 10 are installed inside both the ground time-frequency electromagnetic data acquisition station 1 and the downhole arrayed time-frequency electromagnetic data acquisition short section 2, and the attitude sensors 10 are three-component electronic attitude sensors or three-component fiber optic attitude sensors.

[0085] The artificial big data model 8 for time-frequency electromagnetic data processing uses various underground three-dimensional geological and structural models, three-dimensional resistivity models, three-dimensional polarizability models, and based on the layout of the ground time-frequency electromagnetic data acquisition stations 1, the downhole array-type time-frequency electromagnetic data acquisition short joints 2, and the ground current source emission devices, forward simulates the time-frequency electromagnetic data recorded on the ground or / and in the well for any underground geological structure and geoelectric model, as well as other existing measured ground or / and downhole time-frequency electromagnetic data, and then conducts AI training on the artificial big data model 8 for time-frequency electromagnetic data processing.

[0086] The AI processing model 9 for time-frequency electromagnetic data is obtained by distillation on the basis of the artificial big data model 8 for time-frequency electromagnetic data processing, and is used to perform AI processing on the collected ground or / and downhole time-frequency electromagnetic data in the ground time-frequency electromagnetic data acquisition control processing computer workstation 7.

[0087] The data acquisition and processing method of the ground and downhole time-frequency electromagnetic measurement system based on the AI big data model includes the following specific steps:

[0088] (a) Layout the ground current source emission devices, and use two current source emission antennas 3 arranged in parallel on both sides of the periphery of the work area or four current source emission antennas 3 arranged around the periphery of the work area to form a ground or downhole time-frequency electromagnetic data acquisition area around the work area. The middle of each current source emission antenna 3 is connected to a high-power time-frequency electromagnetic emission source 6, and both ends of each current source emission antenna 3 are grounded through copper rod electrodes 11; or use a transmitting coil (4);

[0089] (b) For the acquisition of well-to-ground time-frequency electromagnetic data, connect one end of the dipole current source emission antenna 5 to the metal casing at the wellhead, and arrange the other end of the dipole current source emission antenna 5 at a position several kilometers away from the wellhead and ground it through a copper rod electrode 11. The middle of the dipole current source emission antenna 5 is connected to a high-power time-frequency electromagnetic emission source 6;

[0090] (c) Arrange multiple ground time-frequency electromagnetic data acquisition stations 1 in the time-frequency electromagnetic data acquisition area in a designed grid manner or centered on the well. For each ground time-frequency electromagnetic data acquisition station 1, arrange two groups of non-polarized electric field sensor pairs in the north-south direction and the east-west direction or in mutually perpendicular orthogonal directions, and arrange mutually orthogonal and perpendicular three-component magnetic field sensors at the center positions of the two groups of non-polarized electric field sensor pairs;

[0091] (d) For the acquisition of downhole time-frequency electromagnetic data, arrange the downhole array-type time-frequency electromagnetic data acquisition short joint 2 to a depth position near the reservoir at the bottom of the well through armored cables or armored optical cables;

[0092] (e) Connect the wired conventional surface time-frequency electromagnetic data acquisition station 1 deployed on the ground and / or the conventional downhole array time-frequency electromagnetic data acquisition short section 2 to the surface time-frequency electromagnetic data acquisition control and processing computer workstation 7 through a connecting cable; connect the fiber optic surface time-frequency electromagnetic data acquisition station 1 deployed on the ground and / or the fiber optic downhole array time-frequency electromagnetic data acquisition short section 2 to the fiber optic time-frequency electromagnetic data acquisition control, modulation and demodulation instrument 12 through a connecting optical cable;

[0093] (f) Use the AI-based underground three-dimensional resistivity and polarizability modeling software, input the three-dimensional geological and structural model, three-dimensional resistivity model, three-dimensional polarizability model, all resistivity logging data, other surface and / or downhole time-frequency electromagnetic data, and resistivity and polarizability measurement data of core specimens in the time-frequency electromagnetic data acquisition work area, and establish the three-dimensional resistivity model and three-dimensional polarizability model 10 underground in the time-frequency electromagnetic data acquisition work area;

[0094] (g) Smooth all the logging data in the time-frequency electromagnetic data acquisition work area to obtain the porosity, resistivity, polarizability, and hydrocarbon saturation of the underground rock formations and reservoirs along the well trajectory. Through AI processing, establish the relationship between the hydrocarbon saturation and resistivity and polarizability in the reservoir;

[0095] (h) Start the high-power time-frequency electromagnetic emission source 6 and the surface time-frequency electromagnetic data acquisition control and processing computer workstation 7, and use two non-polarizing electric field sensors on the ground to obtain the two horizontal electric field component data of the time-frequency electromagnetic field excited by the high-power time-frequency electromagnetic emission source 6, and use the ground three-component magnetic field sensor to obtain the three-component magnetic field data of the underground time-frequency electromagnetic field. The surface time-frequency electromagnetic signal includes two mutually orthogonal horizontal electric field component data and three mutually orthogonal three-component magnetic field data;

[0096] (i) Start the high-power time-frequency electromagnetic emission source 6 and the surface time-frequency electromagnetic data acquisition control and processing computer workstation 7, and use two vertical non-polarizing electric field sensors downhole to obtain the vertical electric field component data of the time-frequency electromagnetic field excited by the high-power time-frequency electromagnetic emission source 6, and use the downhole three-component magnetic field sensor to obtain the three-component magnetic field data of the underground time-frequency electromagnetic field. The downhole time-frequency electromagnetic signal includes the vertical electric field component data and three mutually orthogonal three-component magnetic field data;

[0097] (j) When the fiber optic surface time-frequency electromagnetic data acquisition station 1 is deployed on the ground and / or the fiber optic downhole array time-frequency electromagnetic data acquisition short section 2 is deployed downhole, the fiber optic time-frequency electromagnetic data acquisition control, modulation and demodulation instrument 12 is connected to the fiber optic surface time-frequency electromagnetic data acquisition station 1 and / or the fiber optic downhole array time-frequency electromagnetic data acquisition short section 2 through a connecting optical cable.

[0098] (k) Start the fiber optic time-frequency electromagnetic data acquisition control modulation and demodulation instrument 12 to begin collecting data on the three mutually orthogonal three-component electric field components of the time-frequency electromagnetic field on the ground and / or in the well, and data on the mutually orthogonal three-component magnetic field of the time-frequency electromagnetic field.

[0099] (l) Synchronously start the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7 installed with the time-frequency electromagnetic data processing artificial big data model 8 with AI training and the time-frequency electromagnetic data AI processing model 9.

[0100] (m) Perform on-site processing of the time-frequency electromagnetic data collected in the time-frequency data acquisition work area on the ground or / and in the well in the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7.

[0101] (n) Use the AI-based time-frequency electromagnetic data preprocessing software installed in the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7 to perform denoising processing, consistency processing of surface and / or downhole excitation sources, and consistency processing of surface and / or downhole time-frequency electromagnetic data on the collected time-frequency electromagnetic data on the ground or / and in the well.

[0102] (o) Input the time-frequency electromagnetic data processed in step (n) and the three-dimensional resistivity model and three-dimensional polarizability model 10 established in step (f) into the distilled time-frequency electromagnetic data processing artificial big data model 8.

[0103] (p) The time-frequency electromagnetic data processing artificial big data model (8) in the ground time-frequency electromagnetic data acquisition control and processing computer workstation 7, based on the time-frequency electromagnetic data processed in step (n) and combined with the three-dimensional resistivity model and three-dimensional polarizability model 10 established in step (f), obtains the true underground three-dimensional resistivity model and true underground three-dimensional polarizability model corresponding to the collected time-frequency electromagnetic data through AI-based three-dimensional inversion technology.

[0104] (q) When performing the AI-based three-dimensional inversion processing in step (p), strongly constrain and calibrate the inversion results in step (p) at the well trajectory position using the burial depth and thickness parameters of the reservoir geoelectric model established in step (f) and the relationship between the hydrocarbon saturation in the reservoir and the resistivity and polarizability obtained in step (g).

[0105] (r) Based on the accurate resistivity distribution and the accurate polarizability distribution of the underground three-dimensional reservoir medium, combining the true three-dimensional resistivity model and the true three-dimensional polarizability model of the underground space corresponding to the time-frequency electromagnetic data obtained in step (p), and based on the distribution and variation laws of high-resistivity anomalies and high-polarizability anomalies in the three-dimensional space of the underground reservoir, respectively judge the occurrence state of the oil and gas saturation and the migration and change state of the oil-water interface or gas-water interface in the three-dimensional reservoir space underground in the time-frequency electromagnetic data acquisition work area. Use the relational formula between the oil and gas saturation, resistivity, and polarizability in the reservoir in step (g), combine the high-resistivity anomaly and high-polarizability anomaly data and reservoir porosity parameters inverted in step (q) to calculate the total content of oil and gas in the reservoir, fully evaluate the underground oil and gas resource prospects in the oil and gas exploration area, conduct a comprehensive investigation of the residual oil and gas resources or remaining oil and gas resources in the underground reservoir in the oil and gas resource production area, and accordingly optimize the layout of infill wells or adjustment wells.

[0106] The above specific implementation manners have further elaborated on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific implementation manners of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A surface and borehole time-frequency electromagnetic measurement system based on an AI big data model, characterized in that It includes a surface time-frequency electromagnetic data acquisition station (1) arranged in a grid pattern on the surface of the work area, an in-well arrayed time-frequency electromagnetic data acquisition short section (2) arranged in the well, and a surface current source transmitting device; The surface current source transmitting device is connected to a high-power time-frequency electromagnetic transmitter (6); Both the surface time-frequency electromagnetic data acquisition station (1) and the in-well arrayed time-frequency electromagnetic data acquisition short section (2) are used to acquire the time-frequency electromagnetic signals generated underground by the surface current source transmitting device; The surface time-frequency electromagnetic data acquisition station (1) and the in-well arrayed time-frequency electromagnetic data acquisition short section (2) are connected to a surface time-frequency electromagnetic data acquisition control and processing computer workstation (7) through cables or fiber-optic composite cables; an artificial big data model (8) for time-frequency electromagnetic data processing trained by AI is installed in the surface time-frequency electromagnetic data acquisition control and processing computer workstation (7), and there is also a time-frequency electromagnetic data AI processing model (9) derived from the artificial big data model (8) for time-frequency electromagnetic data processing; using AI-based underground three-dimensional resistivity and polarizability modeling software, inputting the three-dimensional geological and structural model, three-dimensional resistivity model, three-dimensional polarizability model, all resistivity logging data, other surface or / and downhole time-frequency electromagnetic data, and resistivity and polarizability measurement data of core specimens in the time-frequency electromagnetic data acquisition work area, a three-dimensional resistivity model and a three-dimensional polarizability model (10) of the underground in the time-frequency electromagnetic data acquisition work area are established.

2. The surface and borehole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, wherein, The described surface current source transmitting device is two current source transmitting antennas (3) arranged in parallel on both sides of the periphery of the work area or four current source transmitting antennas (3) arranged around the periphery of the work area, or a circular or rectangular transmitting coil (4) surrounding the periphery of the work area, or a dipole current source transmitting antenna (5) arranged at the wellhead and far from the wellhead; The middle of each of the current source transmitting antennas (3) is connected to a high-power time-frequency electromagnetic transmitter (6), and both ends of the current source transmitting antenna (3) are grounded through copper rod electrodes (11); the current source transmitting antenna (3) is a large-current power supply wire with a length of 5 km to 10 km; The tail end of the transmitting coil (4) is connected to a high-power time-frequency electromagnetic transmitter (6); One end of the dipole current source transmitting antenna (5) is connected to the metal casing of the wellhead, and the other end is located several kilometers away from the wellhead and is grounded through a copper rod electrode (11); a high-power time-frequency electromagnetic transmitter (6) is connected between the end connected to the metal casing of the wellhead and the end far from the wellhead of the dipole current source transmitting antenna (5).

3. The ground and borehole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, wherein The surface time-frequency electromagnetic data acquisition station (1) is a surface time-frequency electromagnetic data acquisition station with wired or wireless conventional two-component electric field sensors and three-component magnetic field sensors. The two-component electric field sensors are two pairs of conventional non-polarizing electric field sensors that are orthogonal to each other. The three-component magnetic field sensors are orthogonal to each other, and the three-component magnetic field sensors are located at the center of the two pairs of conventional non-polarizing electric field sensors; the conventional non-polarizing electric field sensor pairs are one of copper sulfate, silver chloride, nanomaterials, and tantalum capacitor non-polarizing electrodes. The distance between the two electric field sensors in the conventional non-polarizing electric field sensor pairs is 10 m to 100 m; Alternatively, the surface time-frequency electromagnetic data acquisition station (1) is a fiber-optic surface time-frequency electromagnetic data acquisition station composed of a fiber-optic three-component electric field sensor and a fiber-optic three-component magnetic field sensor.

4. The surface and downhole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, characterized in that, The downhole arrayed time-frequency electromagnetic data acquisition sub-section (2) is an arrayed conventional time-frequency electromagnetic data acquisition sub-section composed of a pair of conventional vertical electric field sensors and a three-component magnetic field sensor, with the pair of conventional vertical electric field sensors arranged at the upper and lower ends of the three-component magnetic field sensor; or an arrayed fiber-optic time-frequency electromagnetic data acquisition sub-section composed of a fiber-optic three-component electric field sensor and a fiber-optic three-component magnetic field sensor.

5. The ground and borehole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, wherein, The three-component magnetic field sensors of the surface time-frequency electromagnetic data acquisition station (1) and the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2) are one of an induction magnetic field sensor, a fluxgate magnetic field sensor, a MEMS magnetic field sensor, a cold-atom magnetic field sensor, and a superconducting magnetic field sensor; the three-component electric field sensors of the surface fiber-optic surface time-frequency electromagnetic data acquisition station (1) and the downhole arrayed fiber-optic time-frequency electromagnetic data acquisition sub-section (2) are three-component fiber-optic electric field sensors, and the three-component magnetic field sensors are three-component fiber-optic magnetic field sensors.

6. The ground and downhole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, characterized in that, When the surface time-frequency electromagnetic data acquisition station (1) is a wired conventional surface time-frequency electromagnetic data acquisition station or / and the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2) is a downhole-arranged arrayed conventional time-frequency electromagnetic data acquisition sub-section, the surface time-frequency electromagnetic data acquisition station (1) or / and the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2) are directly connected to the surface time-frequency electromagnetic data acquisition control and processing computer workstation (7), which is used to control and receive the data collected by the surface time-frequency electromagnetic data acquisition station (1) and the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2); when the surface time-frequency electromagnetic data acquisition station (1) is a fiber-optic surface time-frequency electromagnetic data acquisition station or / and the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2) is an arrayed fiber-optic time-frequency electromagnetic data acquisition sub-section, a fiber-optic time-frequency electromagnetic data acquisition control, modulation, and demodulation instrument (12) for collecting data is also included.

7. The ground and downhole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, characterized in that Attitude sensors (10) are installed in both the surface time-frequency electromagnetic data acquisition station (1) and the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2), and the attitude sensors (10) are three-component electronic attitude sensors or three-component fiber-optic attitude sensors.

8. The ground and downhole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, wherein, The time-frequency electromagnetic data processing artificial big data model (8) uses various underground three-dimensional geological and structural models, three-dimensional resistivity models, three-dimensional polarization models, and based on the layout methods of the surface time-frequency electromagnetic data acquisition station (1), the downhole arrayed time-frequency electromagnetic data acquisition sub-section (2), and the surface current source emission device, forward simulates the time-frequency electromagnetic data recorded on the surface or / and in the well for any underground geological structure and geoelectric model, as well as other existing measured surface or / and downhole time-frequency electromagnetic data, and then conducts AI training on the time-frequency electromagnetic data processing artificial big data model (8).

9. The ground and downhole time-frequency electromagnetic measurement system based on the AI big data model according to claim 1, wherein, The time-frequency electromagnetic data AI processing model (9) is obtained by distillation based on the artificial big data model (8) for time-frequency electromagnetic data processing, and is used to perform AI processing on the collected surface and / or downhole time-frequency electromagnetic data in the surface time-frequency electromagnetic data acquisition control processing computer workstation (7).

10. A ground and borehole time-frequency electromagnetic measurement method based on an AI big data model, characterized in that, Applied to the surface and downhole time-frequency electromagnetic measurement system based on the AI big data model according to any one of claims 1-9, it includes the following specific steps: (a) Arrange the surface current source transmitting device; use the current source transmitting antenna (3) to form a surface or downhole time-frequency electromagnetic data acquisition area around the work area. The middle of each current source transmitting antenna (3) is connected to a high-power time-frequency electromagnetic transmitter (6), and both ends of each current source transmitting antenna (3) are grounded through copper rod electrodes (11); or use a transmitting coil (4); (b) For the acquisition of well-to-surface time-frequency electromagnetic data, connect one end of the dipole current source transmitting antenna (5) to the metal casing at the wellhead, and arrange the other end of the dipole current source transmitting antenna (5) at a position several kilometers away from the wellhead and ground it through a copper rod electrode (11); connect a high-power time-frequency electromagnetic transmitter (6) in the middle of the dipole current source transmitting antenna (5); (c) Arrange multiple surface time-frequency electromagnetic data acquisition stations (1) in the time-frequency electromagnetic data acquisition area in a designed grid pattern or centered on the well. For each surface time-frequency electromagnetic data acquisition station (1), arrange two sets of non-polarized electric field sensor pairs in the north-south direction and the east-west direction or in mutually perpendicular directions, and arrange the mutually perpendicular three-component magnetic sensors at the center positions of the two sets of non-polarized electric field sensor pairs; (d) For the acquisition of downhole time-frequency electromagnetic data, arrange the downhole array time-frequency electromagnetic data acquisition sub-section (2) to a depth position near the reservoir at the bottom of the well through armored cables or armored optical cables; (e) Connect the surface time-frequency electromagnetic data acquisition station (1) and / or the downhole array time-frequency electromagnetic data acquisition sub-section (2) to the surface time-frequency electromagnetic data acquisition control processing computer workstation (7) through connecting cables; (f) Use the AI-based underground three-dimensional resistivity and polarizability modeling software, input the three-dimensional geological and structural model, three-dimensional resistivity model, three-dimensional polarizability model, all resistivity logging data, other surface and / or downhole time-frequency electromagnetic data, and resistivity and polarizability measurement data of core specimens in the time-frequency electromagnetic data acquisition work area, and establish a three-dimensional resistivity model and a three-dimensional polarizability model (10) underground in the time-frequency electromagnetic data acquisition work area; (g) Smooth all the logging data in the time-frequency electromagnetic data acquisition work area to obtain the porosity, resistivity, polarizability, and hydrocarbon saturation of the underground rock formations and reservoirs along the well trajectory, and establish a relationship between the hydrocarbon saturation, resistivity, and polarizability in the reservoir through AI processing; (h) Start the high-power time-frequency electromagnetic emission source (6) and the ground time-frequency electromagnetic data acquisition, control, and processing computer workstation (7), and use two non-polarizing electric field sensors on the ground to obtain data of two horizontal electric field components of the time-frequency electromagnetic field excited by the high-power time-frequency electromagnetic emission source (6), and use a three-component magnetic field sensor on the ground to obtain data of three components of the underground time-frequency electromagnetic field. The ground time-frequency electromagnetic signal includes data of two mutually orthogonal horizontal electric field components and data of three mutually orthogonal three-component magnetic fields. (i) Start the high-power time-frequency electromagnetic emission source (6) and the ground time-frequency electromagnetic data acquisition, control, and processing computer workstation (7), and use two vertical non-polarizing electric field sensors in the well to obtain data of the vertical electric field component of the time-frequency electromagnetic field excited by the high-power time-frequency electromagnetic emission source (6), and use a three-component magnetic field sensor in the well to obtain data of three components of the underground time-frequency electromagnetic field. The time-frequency electromagnetic signal in the well includes data of the vertical electric field component and data of three mutually orthogonal three-component magnetic fields. (j) When the ground time-frequency electromagnetic data acquisition station (1) is an optical fiber ground time-frequency electromagnetic data acquisition station, and the in-well array time-frequency electromagnetic data acquisition short section (2) is an in-well deployed array optical fiber time-frequency electromagnetic data acquisition short section, the optical fiber time-frequency electromagnetic data acquisition, control, modulation, and demodulation instrument (12) is connected to the optical fiber ground time-frequency electromagnetic data acquisition station (1) and / or the in-well array time-frequency electromagnetic data acquisition short section (2) through a connecting optical cable. (k) Start the optical fiber time-frequency electromagnetic data acquisition, control, modulation, and demodulation instrument (12) to start collecting data of three mutually orthogonal three-component electric field components of the time-frequency electromagnetic field on the ground and / or in the well and data of mutually orthogonal three-component magnetic fields of the time-frequency electromagnetic field. (l) Synchronously start the ground time-frequency electromagnetic data acquisition, control, and processing computer workstation (7) installed with the time-frequency electromagnetic data processing artificial big data model (8) trained with AI and the time-frequency electromagnetic data AI processing model (9). (m) Perform on-site processing of the ground and / or in-well time-frequency electromagnetic data collected in the time-frequency data acquisition work area in the ground time-frequency electromagnetic data acquisition, control, and processing computer workstation (7). (n) Use the AI-based time-frequency electromagnetic data preprocessing software installed in the ground time-frequency electromagnetic data acquisition, control, and processing computer workstation (7) to perform denoising processing, consistency processing of surface and / or downhole excitation sources, and consistency processing of surface and / or downhole time-frequency electromagnetic data on the collected ground and / or in-well time-frequency electromagnetic data. (o) Input the time-frequency electromagnetic data processed in step (n) and the three-dimensional resistivity model and three-dimensional polarizability model (10) established in step (f) into the distilled time-frequency electromagnetic data processing artificial big data model (8). (p) The time-frequency electromagnetic data processing artificial big data model (8) in the ground time-frequency electromagnetic data acquisition control and processing computer workstation (7) obtains the true underground three-dimensional resistivity model and the true underground three-dimensional polarization rate model corresponding to the collected time-frequency electromagnetic data through the AI-based three-dimensional inversion technology, based on the time-frequency electromagnetic data processed in step (n) and combined with the three-dimensional resistivity model and the three-dimensional polarization rate model (10) established in step (f); (q) When performing the AI-based three-dimensional inversion processing in step (p), the burial depth and thickness parameters of the reservoir geoelectric model established in step (f) and the relationship between the oil-gas saturation in the reservoir obtained in step (g) and the resistivity and polarization rate are used to strongly constrain and calibrate the inversion results in step (p) along the well trajectory position; (r) According to the accurate resistivity distribution and accurate polarization rate distribution of the underground three-dimensional reservoir medium, combined with the true underground three-dimensional resistivity model and the true underground three-dimensional polarization rate model corresponding to the time-frequency electromagnetic data obtained in step (p), and based on the distribution and variation laws of high-resistivity anomalies and high-polarization rate anomalies in the underground three-dimensional reservoir space, the occurrence state of the oil-gas saturation and the migration and variation state of the oil-water interface or gas-water interface in the underground three-dimensional reservoir space in the time-frequency electromagnetic data acquisition work area are respectively judged. Using the relationship between the oil-gas saturation in the reservoir and the resistivity and polarization rate in step (g), combined with the high-resistivity anomaly and high-polarization rate anomaly data and reservoir porosity parameters inverted in step (q), the total content of oil and gas in the reservoir is calculated, the underground oil and gas resource prospects in the oil and gas exploration area are fully evaluated, the residual oil and gas resources or remaining oil and gas resources in the underground reservoir in the oil and gas resource production area are comprehensively investigated, and the optimal layout of infill wells or adjustment wells is carried out accordingly.

Images