Method for determining a model of a physical property of a subsurface volume of interest

Abstract

A method for determining physical properties of a subsurface target volume is provided as well as a system and computer readable medium for performing the method. Unlocking insights from geologic data, the invention further relates to improvements in sustainability and environmental development: we co-create a safe and livable world. The method includes dividing a portion of the method into subtasks, wherein the subtasks are distributed among a network of computing devices of a network and executed in parallel. Unlocking insights from geologic data, the invention further relates to improvements in sustainability and environmental development: we co-create a safe and livable world.

Description

Technical Field

[0001] Unlocking insights from geological data, this invention relates to improvements in sustainability and environmental development: together we create a safe and livable world. More specifically, this disclosure relates to methods and systems for analyzing target volumes beneath the Earth's surface with improved efficiency. Background Technology

[0002] There is a pervasive and ongoing need for systems and methods to determine subsurface parameters. In particular, there is a need for systems and methods capable of modeling the properties of target volumes beneath the Earth's surface to provide useful information for infrastructure planning and foundation design. Determining subsurface properties early in the planning stages of a construction project reduces uncertainty in the project's location, design, and construction phases. This, in turn, reduces delays, cost overruns, and unnecessary use of material resources (such as concrete) throughout the construction and asset lifecycle.

[0003] A key parameter for determining the surface properties within a volume of interest is the shear modulus G and the shear velocity Vs. Shear velocity is the speed at which a shear wave travels through the material and is controlled by the material's shear modulus. The relationship between shear velocity and shear modulus is determined by… Limited, among which It refers to the density of the material. Therefore, measuring the shear rate provides valuable insights into the material properties of underground surface regions.

[0004] Surface wave spectral analysis (SASW) and surface wave multichannel analysis (MASW) are examples of techniques for acquiring surface wave information that can be used to determine material properties in subsurface volumes. In both techniques, surface-level vibrations, originating from passive sources (vibrations on the surface caused by environmental noise sources) or active sources (such as falling objects), are measured, and the dispersion of the recorded surface waves is recorded. ReMi (Refraction Micromotion) is another surface wave technique that performs surface measurements on waves from environmental seismic noise recorded at the surface to infer material properties in subsurface areas.

[0005] Invasive downhole and inter-well techniques can also be used to determine the material properties of target subsurface areas. In both methods, a receiver located in the borehole measures waves recorded from an active source located elsewhere. In downhole techniques, one source or receiver is located at a subsurface location within the borehole, while the other is located at the surface. In inter-well techniques, the source is located in a first borehole, and the receiver is located in a second borehole. In both techniques, the propagation of the recorded waves is studied to infer the properties of the material through which the waves originating from the source pass.

[0006] Invasive techniques for measuring material properties in subsurface areas often present logistical challenges, particularly in urban or hard-to-access environments, and are typically prohibitively expensive to obtain the required data volume relying solely on these techniques. These techniques require significant resources (equipment and personnel) and are associated with safety risks and negative environmental footprints. Conversely, current surface techniques may lack the accuracy and reliability of more invasive analytical techniques. Summary of the Invention

[0007] According to a first aspect of this disclosure, a method is provided for using a network of computing devices to determine a model of the physical properties of a subsurface target volume. The model includes a plurality of cells, each cell having a corresponding physical property value. The method includes receiving a plurality of signals detected by a plurality of receivers arranged on a surface above the subsurface target volume, wherein each corresponding signal of the plurality of signals is detected by a corresponding receiver of the plurality of receivers. The method includes cross-correlation of the signals between the plurality of receivers to obtain empirical time-of-flight data of surface waves between a plurality of receiver pairs. The method includes selecting a set of the plurality of receiver pairs, each corresponding receiver pair having a corresponding surface wave path between the receivers of the receiver pair. The method includes performing tomographic inversion on the empirical time-of-flight data between the receivers of each receiver pair in the set to obtain a corresponding physical property value for each cell. The method includes organizing the obtained physical property values ​​into the model. One or more of cross-correlation, selection, performing tomographic inversion, and organizing are divided into subtasks, wherein the subtasks are distributed among the computing devices of the network and executed in parallel.

[0008] A network can be configured to provide cloud-based computing, or in other words, the computing devices can be part of a cloud computing service. Multiple computing devices can be an entire collection of computing devices providing a cloud computing environment, or a subset of computing devices providing a cloud computing environment. Therefore, using a network of computing devices means that at least a portion of the method is performed by multiple computing devices connected via the network. For example, tomographic inversion can be performed by multiple computing devices in a network.

[0009] Receiving multiple signals detected by the plurality of receivers may include receiving the multiple signals from a user equipment communicating with a network (e.g., via the Internet) or by accessing a database storing the multiple signals thereon. In some examples, the method may further include detecting the multiple signals to be used in the method by the plurality of receivers, and may include preprocessing the detected signals for use in the method. The signals detected by the plurality of receivers may be environmental seismic signals.

[0010] The receiver used in this method can be a seismograph, accelerometer, seismograph, vibration sensor, and / or transducer. The receiver can acquire data over a considerably long time period. For example, ambient seismic noise can be continuously measured over a five-day period. This long recording time allows for the full extraction of surface wave information from the ambient seismic noise recorded at or near the surface of the target area. Furthermore, as described in more detail below, the (processed) surface wave information can be used for tomographic inversion to obtain a shear wave velocity model of the subsurface target volume. Therefore, the physical properties of the subsurface target volume can be represented by shear wave velocity.

[0011] Cross-correlation of signals between multiple receivers can include cross-correlation of a subset of the total number of receiver pairs (i.e., not every possible pair). As used herein, “empirical” data or information refers to data or information obtained empirically, i.e., real signals acquired from physical receivers located on a surface above the underground target volume.

[0012] The subtasks of cross-correlation, selection, execution tomography inversion, and finishing are independent, meaning each subtask can be completed without first completing other subtasks or requiring additional information from them, allowing for parallel execution. Therefore, parallel execution means that multiple (at least two, but possibly hundreds or more) subtasks are executed simultaneously. Subtasks can be executed anywhere on the network of the computing device (subject to network configuration and control). Each subtask can be associated with metadata to tagged it and / or provide information describing it. For example, metadata may include instructions on where to store the results. Each subtask may include writing the results to a specified storage location or sending the results to another function for storage or execution. The results of each subtask may be physical property values ​​or a collection of physical property values ​​to be finished into the model.

[0013] The efficiency of a method is improved by dividing one or more parts of it into subtasks that can be executed in parallel. For example, using environmental seismic signals typically requires longer measurement periods than signals from active testing. This results in the need to process large amounts of data to determine the desired model for the physical properties of subsurface target volumes (e.g., shear wave velocities). Using a network of computing devices as described above, dividing specific parts of the method into subtasks for parallel execution can significantly reduce the time required to determine the model (depending on available computing resources). Dividing into subtasks also improves efficiency by leveraging the independence of certain parts of the computation to schedule calculations.

[0014] Tomographic inversion is the computationally intensive part of this method, making it particularly suitable for improving computational speed and efficiency. Tomographic inversion can include a tomographic imaging component and an inversion component, either or both of which can be subdivided into subtasks. Additionally or alternatively, improved speed and efficiency can be achieved in any or any combination of cross-correlation, selection, performing tomographic inversion, and tidying. Receiving multiple signals can also be subdivided into subtasks in a similar manner. A given subtask can include a portion of cross-correlation, selection, performing tomographic inversion, and tidying; for example, performing all these operations for a frequency subrange. In some examples, each of one or more of cross-correlation, selection, performing tomographic inversion, and tidying can be subdivided into a separate subtask; for example, a subtask may only perform cross-correlation, and new subtasks may be created depending on whether the subtask is divided into selection or tomographic inversion.

[0015] In some examples, subtasks are frequency subtasks, where each frequency subtask corresponds to a corresponding frequency subrange of the total frequency range of multiple signals. The tomographic portion of a tomographic inversion can be divided into frequency subtasks. For surface waves traversing a subsurface target volume, the frequency carries depth information within that volume. This is because surface wave propagation is influenced by the physical properties of the subsurface volume up to approximately one wavelength deep; for example, low-frequency surface waves are more significantly affected by the physical properties at greater depths than high-frequency surface waves. Therefore, frequency subtasks correspond to calculations at different depths.

[0016] In some examples, the subtasks are model element subtasks, where each model element subtask corresponds to a specific element of the model. For example, each subtask is associated with an element of the model to be determined, and cross-correlation and / or selection can be partitioned such that each model element subtask selects the receiver pair based on which receiver pairs' surface wave paths pass through the relevant element of the model. The inversion portion of the tomographic inversion can then be performed individually for each element of the model to obtain the corresponding physical property values.

[0017] In some examples, the subtasks are path subtasks, where each path subtask corresponds to a specific surface wave path. For instance, tomographic inversion can be performed over the entire surface wave path between receiver pairs, and the results can then be used to determine or modify the physical property values ​​of the cells the wave path traverses in the model. By using this approach, fewer tomographic inversion calculations are required, thus reducing computation time.

[0018] In some examples, the subtasks are travel-time subtasks, where each travel-time subtask corresponds to a corresponding acquired empirical travel-time data. For example, selecting and / or performing tomographic inversion can be divided into subtasks based on the travel time obtained during cross-correlation.

[0019] In some examples, subtasks are receiver subtasks, where each receiver subtask corresponds to one receiver among the plurality of receivers. For example, cross-correlation and / or selection can be divided into subtasks based on the receivers.

[0020] In some examples, a subtask can be a combination of frequency subtasks, model unit subtasks, path subtasks, and travel time subtasks. The overall approach can use several different types of subtasks; for example, different types of subtasks can be used for different parts of the approach, or different types of subtasks can be used within the same part. A subtask can be categorized into more than one of the following: frequency subtask, model unit subtask, path subtask, and travel time subtask, i.e., for a specific frequency subrange and / or a wave path, etc.

[0021] In some examples, subtasks are recorded in a work queue, where each subtask has a position in the work queue, and the result of each completed subtask is recorded at a result position in a result queue, corresponding to the position of the completed subtask in the work queue. For example, the positions in the work queue can be sequentially numbered, and the corresponding result position is the position number in the result queue that is equal to the position number in the work queue. In another example, the result position can be determined by an algorithm based on the position in the work queue. By preserving the structure that reflects the results of the input work queue as described above, values ​​can be organized into the model particularly efficiently because it reduces or avoids the time spent searching for relevant values ​​to be recorded in the model.

[0022] In some examples, performing tomographic inversion includes: obtaining an initial model with initial physical property values ​​for multiple elements; determining a simulated travel time for each receiver pair in the set using the initial physical property values, which are associated with the elements of the model traversed by the surface wave path between the receivers of the receiver pair; and determining new physical property values ​​for the multiple elements based on the simulated travel time and the empirical travel time. In some examples, the initial model may be selected based on a coarser model (i.e., with larger elements) previously determined using less precise methods, or based on approximate expected physical property values ​​(e.g., based on typical geological properties of the location), or default values. Determining new physical property values ​​may include comparing the simulated travel time and the empirical travel time and refining the model to reduce the discrepancy between the simulated travel time and the empirical travel time, for example, by using the least squares method or the Markov chain Monte Carlo method.

[0023] According to a second aspect of this disclosure, a method is provided for modeling the physical properties of an underground target volume. This method can be performed by a user equipment. The method includes: transmitting to a network of computing devices a plurality of signals detected by a plurality of receivers arranged on a surface above the underground target volume, wherein each corresponding signal of the plurality of signals is detected by a corresponding receiver of the plurality of receivers. The method includes: transmitting an instruction to the network that causes the network to perform the method described above with respect to a first aspect of this disclosure, and optionally any examples as described above. The method includes: receiving from the network a model of the physical properties of the underground target volume. The transmission may be from the user equipment to the network, and / or the reception may be received at the user equipment from the network.

[0024] In some examples, the instructions include computational parameters that include one or more of the following: the dimensions of the subsurface target volume; the dimensions of multiple elements of the model; the target precision of the physical property values; the target computation time; the amount of computational resources to be used; the number of subtasks to be created; and the type of subtasks to be created. The dimensions of the subsurface target volume can be the full height, width, and depth of the volume, and may include the shape of the volume if the volume is not a cuboid. The dimensions of the multiple elements of the model may include the desired level of detail of the model, which controls the precision of the resulting model as well as the computational cost and the required time (where a more detailed model with smaller elements will require more computation). The dimensions of the multiple elements may also be indirectly defined by the total number of elements, or the total number of columns and rows (which defines a two-dimensional surface mesh above the target volume, where the value of each element(s) in a given row and column includes the value variation at the depth below that element). The target computation time will optionally be used in combination with the model dimensions and element dimensions to determine how much computational resources (e.g., the amount of processing power) are required to determine the method within the target computation time. The amount of computational resources to be used can directly define the number and power of the processing elements. The number and type of subtasks set parallelization parameters to improve the speed and efficiency of the method. These can be specified directly in the instructions, or the instructions can provide a set of conditions to be applied.

[0025] In some examples, the method includes receiving from the network a request for further instructions due to the network reaching a programming checkpoint. Accordingly, the method described above with respect to the first aspect may include sending to the user equipment a request for further instructions due to the network reaching a programming checkpoint. The further instructions may be provided by the user equipment based on information stored locally on the user equipment, or may be entered by the user through a user interface on the user equipment. By using programming checkpoints, this method can reduce the wasted time on error-based assumptions, providing only error results, which then require the user to restart the entire method.

[0026] According to a third aspect of this disclosure, a user equipment is provided. The user equipment includes one or more processors and one or more memories, the one or more memories storing computer-readable instructions configured to cause the one or more processors to perform operations including those described above according to the second aspect and optionally any of the examples described above.

[0027] According to a fourth aspect of this disclosure, a computer-readable medium is provided. The computer-readable medium includes instructions that, when executed by a user equipment as described above according to a third aspect, cause the one or more processors to perform operations including those described according to a second aspect (and optionally any examples as described above). The computer-readable medium may be transient or non-transient.

[0028] According to a fifth aspect of this disclosure, a system is provided that includes a network of computing devices configured to perform the methods described above according to the first aspect, and optionally any of the examples described above. In some examples, the system includes user equipment as described above according to the third aspect.

[0029] According to a sixth aspect of this disclosure, a computer-readable medium is provided. The computer-readable medium includes instructions that, when executed by a system as described above according to the fifth aspect (and optionally any examples as described above for doing so), cause the network to perform operations including those described according to the first aspect (and optionally any examples as described above for doing so). The computer-readable medium may be transient or non-transient. Attached Figure Description

[0030] The disclosed embodiments will now be described by way of example to illustrate various aspects of this disclosure, with reference to the accompanying drawings, in which: Figure 1 A cross-sectional view of the underground target volume is shown; Figure 2 It shows multiple receivers arranged on the surface above the underground target volume; Figure 3 This is a schematic diagram of the shear wave velocity model; Figure 4A It is a top view of the straight wave path on the surface above the underground target volume, with a shear wave velocity model superimposed on it. Figure 4B It is a top view of the curved wave path on the surface above the underground target volume, with a shear wave velocity model superimposed on it. Figure 5 This is a schematic diagram of a cloud computing environment; Figure 6This is a schematic diagram of a method for determining the physical properties of the volume of an underground target; Figure 7 yes Figure 6 A schematic diagram of the process used in the method; Figure 8 This is a schematic diagram of a method for determining the physical properties of the volume of an underground target; Figure 9 This is an example of a shear wave velocity model generated by the method described in this paper; Figure 10 A schematic diagram of a computing device suitable for performing the methods described herein; and Figure 11 This is a schematic diagram of a system suitable for performing the methods described herein. Detailed Implementation

[0031] This detailed description is for reference. Figure 1 Figure 4 illustrates a method for measuring the structural properties of ground volume using a receiver. Next, refer to... Figures 5 to 9 This describes a method for using a network of computing devices to determine the physical properties of underground target volumes. Finally, refer to... Figure 10 and Figure 11 User devices and systems that can be used to perform the disclosed methods are described.

[0032] The following examples are described in the context of the receiver array on a surface above a subsurface volume to aid understanding. However, it should be understood that the disclosed systems and methods are applicable to a wide variety of receiver types, including but not limited to receivers, accelerometers, velocity meters, seismometers, vibration sensors, and / or transducers. The disclosed methods can be applied to any suitable set of signals.

[0033] The methods and systems disclosed herein generally relate to processing signals detected by receivers on a surface. In one example, the receivers are arranged on the surface, and the detected signal is ambient seismic noise. Processing these signals provides useful insights into the structure of the surface on which the receivers are placed and the subsurface target volume, as described in more detail below. By dividing the method into subtasks, which are distributed among computing devices in a network and executed in parallel, computational speed and efficiency are improved, making it feasible to process very large quantities and durations of signals even over practical time periods.

[0034] Before discussing the details of the disclosed method, some background is provided on using receivers to determine surface and subsurface properties. Shear modulus is a measure of a material's elastic shear stiffness, representing the deformation of a solid when subjected to a force parallel to one of its surfaces while its opposite surface experiences an opposing force. Obtaining these forces and their effects within a target subsurface volume is an important parameter studied before and during the design of building and infrastructure projects. To determine the shear modulus of a volume, the shear rate Vs is determined. This, in turn, indicates the stiffness of the subsurface material and its ability to support structures located above and / or through the volume.

[0035] In the context of terrestrial research, two types of waves are typically distinguished: P-waves, in which particles in a volume oscillate in the direction of wave propagation, causing compression and decompression of the ground as the wave propagates. S-waves are shear waves, in which particles oscillate in a direction perpendicular to the direction of wave propagation.

[0036] P-waves and S-waves are volume waves that propagate through the bulk of a volume in all directions. The interaction of P-waves and S-waves with the Earth's surface produces surface waves that propagate along that surface. Several types of surface waves can be distinguished. In the system and method described herein, Rayleigh waves are measured and studied because it is convenient to measure the vertical component of surface vibrations. However, it should be understood that other surface waves, such as Love waves and Skolter waves, can also be measured and utilized in the system and method described herein.

[0037] Because surface waves propagate in two dimensions (on a surface), they decay more slowly than volume waves (which propagate in three dimensions). Surface waves typically exist at a depth of one wavelength above the surface, generally travel more slowly, and have significantly lower frequencies than volume waves.

[0038] The lower attenuation, slower travel time, and lower frequency of surface waves make them particularly attractive for determining shear velocity Vs. Because of their lower attenuation, surface waves maintain signal strength better over longer travel distances. Therefore, the resulting measurements typically have higher signal quality (signal-to-noise ratio) than those obtained from volume wave studies.

[0039] Now for reference Figure 1 The diagram shows a cross-sectional view of the subsurface volume 100. P-waves and S-waves travel through volume 100 as body waves. Surface 102 extends above the subsurface volume and is defined by the x and y planes. Surface waves propagate along surface 102.

[0040] At one example point, a schematic representation of particle oscillations (due to Rayleigh wave propagation) at a surface above an underground target volume is shown. As shown, the oscillation of particle P is partially perpendicular and partially parallel to the propagation direction. Therefore, the resulting particle motion is essentially elliptical.

[0041] A plurality of receivers 104 are arranged on surface 102 above volume 100. The receivers 104 located on surface 102 can be configured to measure the vertical component of the oscillation schematically shown at the example point.

[0042] Receivers 104 are arranged in a grid array on the surface, extending in two directions. It should be noted that the surface above the target region may not be planar in many cases. Therefore, the array of receivers 104 may not be truly “two-dimensional”, as each receiver may be offset in the z-dimensional from its adjacent receivers in the grid. However, for simplicity, this grid arrangement of receivers will be referred to herein as a two-dimensional array.

[0043] The surface wave traveling on surface 102 will cause vertical motion at multiple receivers 104 as it travels across the surface.

[0044] To determine the shear velocity Vs from observations of surface waves (especially Rayleigh waves), the dispersion behavior of the surface waves can be measured. Surface waves are dispersed, meaning their velocity depends on the frequency. Typically, earthquake velocities increase with depth. Therefore, normal surface wave dispersion manifests as a decrease in surface wave velocity with increasing frequency. By studying the behavior of surface waves at the surface above a subsurface target volume, the material properties of the volume can be determined.

[0045] There are two methods to measure the velocity of dispersive surface waves and to distinguish between the determination of group velocity or phase velocity.

[0046] The group velocity of a wave is the speed at which the overall envelope shape of the wave's amplitude (called the wave's modulation or envelope) propagates through space. Group velocity is equivalent to the speed at which the wave's energy propagates through a volume. It is measured by determining the wave propagation between (synthesized) transducer pairs and is a frequency-dependent point property within the volume, depending on depth. Group velocity is a measure of the travel time between a (virtual) source and receiver as a function of frequency (i.e., the dispersion function).

[0047] Phase velocity is the speed at which a specific frequency component of a wave travels. Therefore, phase velocity is expressed as a function of frequency. To measure phase velocity, at least two measurement nodes are selected to measure the wave propagating through a volume to determine the relative travel time between receivers at different frequencies. The result is the phase velocity as a function of frequency, averaged over the volume between the two measurement nodes. Phase velocity is obtained as a point in a two-dimensional phase space (dispersion spectrum), which is obtained by performing a two-dimensional transformation (e.g., tilt superposition, Radon transform, FK, etc.) on an array of recorded waveforms (time-distance space).

[0048] Refer again Figure 1Each receiver 104 provides a measurement node for measuring the vertical component of the surface wave. Receiver 104 can be configured to measure vibrations due to environmental seismic noise, i.e., the background wave field due to natural or anthropogenic noise (rather than pulse points such as explosions or hammering used in active methods).

[0049] By cross-correlating the ambient noise signals measured at a pair of receivers, the Green's function of the pair of receivers can be obtained. This Green's function represents the wave field as if one receiver in the pair were a virtual source and the other receiver in the pair were a real receiver.

[0050] Figure 2 Multiple pairs of virtual receivers are shown on a surface above the underground target volume. Wave paths 202 between receiver pairs are indicated, specifically the wave paths from a single central receiver near the center of the receiver array to every other receiver in the array. Background shading and contour loops indicate the travel time field from the central receiver to the other receivers. Corresponding wave paths also exist between each receiver and all other receivers, i.e., between each pair of receivers; for simplicity, Figure 2 These wave paths are not depicted in the text.

[0051] Each receiver pair can provide cross-correlation between the signal at the first location and the signal at the second location to reconstruct the virtual receiver pair using interferometry principles. Specifically, in Figure 1 The cross-correlation of the corresponding receiver at the surface with the measured ambient noise can be used to reproduce the response from the underground target volume as if it were caused by a pulse point source, and the response is equal to the Green's function.

[0052] In some examples, Figure 1 Receivers 104a and 104b form the first receiver pair in the array at surface 102. By cross-correlating the signals received at receivers 104a and 104b, receivers 104a and 104b can act as (virtual) source-receiver pairs, where each receiver in the pair records a signal as if the signal originated from the other receiver in the pair.

[0053] In other words, the response received by cross-correlating the records of the two receivers can be interpreted as a response measured at one of the receiver locations, as if a source existed at the other receiver location. Various methods for determining the Green's function of a virtual receiver pair are known, and an overview of these methods is described in "Tutorial on Seismic Interferometry: Part 1 – Basic Principles and Applications", GEOPHYSICS. Vol. 75, No 5 (Sept-Oct 2010; P.75A195075A209, Waphanaar et al.).

[0054] exist Figure 1 In the diagram, only one pair of receivers (104a, 104b) is labeled. However, it should be understood that for each receiver 104 in the array, each other receiver in the array can act as the other half of the source-receiver pair. In this way, the Green's function for each source-receiver pair can be obtained. The Green's function over multiple virtual source-receiver pairs is studied to determine the dispersion behavior of surface waves.

[0055] refer to Figure 3 The first shear wave velocity model 300 has a cell grid m spanning the surface above the underground target volume. xy , including column m extending in the x direction x1 m x2 m x3 Equal to the line m extending in the y direction 1y m 2y m 3y Each element defines the area of ​​the surface. For example, each element can define a 5m × 5m square; other example options include a 1m × 1m square or a 10m × 10m square. In other words, the shear wave velocity model comprises multiple elements arranged in a two-dimensional grid 300. The two-dimensional grid spans at least the area of ​​the surface above the subsurface target volume. Choosing smaller element areas increases the model's resolution. Each element also includes a volume extending vertically below the surface area. The model 300 can extend infinitely below the surface area or to a predetermined depth for which the shear wave velocity value has a significant impact on the propagation of surface waves on the surface. For example, the model 300 can define depths up to 50m, 100m, 200m, or 300m.

[0056] In some examples, the first shear wave velocity model 300 is not a grid of square cells, but includes cells with rectangular, rhomboid, or other tessellated shapes that include non-uniform shapes or combinations of different shapes.

[0057] Each element is associated with a shear wave velocity value (an example of a physical property value), which represents the expected shear wave velocity within the actual subsurface target volume of interest. The shear wave velocity value carries depth information; either the shear wave value is constant throughout the entire volume below the element area, or the shear wave value varies with depth (…). Figure 3 The shear wave velocity varies (in the z-direction). For example, the defined shear wave value for each cell can be explicitly a function of depth, and can be a continuous function or a series of values ​​with an associated depth range for each value. In another example, the shear wave velocity value can be a function of frequency, which corresponds to depth information, because surface wave propagation is influenced by the physical properties of the subsurface volume up to approximately one wavelength deep. In other words, surface waves at lower frequencies are more influenced by the physical properties at deeper depths than surface waves at higher frequencies.

[0058] In other examples, the model can be a physical property different from the shear wave velocity, such as compressive wave velocity, density, elastic modulus, shear modulus, or, if a viscoelastic model is used, optionally include the viscosity quality factor Q. s and Q p Typically, a model can specify multiple physical property values.

[0059] refer to Figures 4A to 4B This depicts a representation of the wave path between a (virtual) source and a receiver. The wave path is the path taken by a surface wave as it travels from the source to the receiver. See below for reference. Figure 4A and Figure 4B The wave path can be represented by a straight or curved wave path connecting the source and receiver. Alternatively, the wave path can be represented by an elliptical Fresnel zone (also known as the Fresnel nucleus) between the source and receiver pair. As another alternative, the wave path can be represented by a curved Fresnel zone.

[0060] refer to Figure 4A The wave path is defined as a straight line (sometimes called a "ray" path according to ray theory) between two receivers at surface locations A and B on model 300. The straight-line wave path model assumes a constant transverse velocity (in other words, the velocity varies only with depth). In many cases, the straight-line wave path model provides a sufficiently accurate approximation of the path a wave travels between the source and receiver. The straight-line wave path approximation is also mathematically and computationally efficient compared to other more complex techniques. The wave path represents the motion of a surface wave according to wave theory as it travels from A to B (or from B to A). Specifically, the wave path is defined as the direction of propagation of the surface wave, either perpendicular to the wavefront in wave theory or perpendicular to the contour lines of travel time. Figure 4AThe wave path shown passes through seven elements of model 300, numbered 1 to 7. This wave path comprises seven bands, each with a length L depending on the length of the wave path travels through each element. i ,Right now Figure 4A L1 to L7 in the middle.

[0061] refer to Figure 4B The wave path is defined as the curve between the two receivers at surface locations C and D. Figure 4B The wavepath shown passes through six elements of model 300, numbered 1 to 6. This wavepath comprises six bands, each with a length L based on the length of the wavepath passing through each element. i ,Right now Figure 4B L1 to L6 in the model. The trajectory of the wave path can be calculated using the minimum time principle (Fermat's principle) based on the shear wave velocity values ​​of the elements in model 300 between positions C and D.

[0062] refer to Figure 5 The methods described herein can be executed in a cloud computing environment 500. In some examples, the cloud computing environment 500 has a virtual state machine 510 for coordinating computational processes within the user-defined cloud computing environment 500; a virtual machine 520; virtual memory 530; and accessible memory 540. The virtual machine 520 emulates a processor and communicates with the virtual state machine 510 to receive instructions to be executed and to send updates on progress and results, such as status reports and / or the location of result data. The virtual memory 530 emulates memory with which the virtual machine 520 interacts to perform its computations. The virtual memory 530 may include one or more virtual SSDs (Solid State Drives) and / or one or more virtual HDDs (Hard Disk Drives) depending on the requirements of the processes running by the virtual machine 520. The accessible memory 540 is object storage, which can be accessed by user devices, for example, via the Internet through an API (Application Programming Interface) gateway. Users of the cloud computing environment 500 can utilize the network of the computing devices on which the cloud computing environment 500 is based (e.g., referenced below). Figure 11 The method described herein is performed by defining a virtual state machine 510, selecting the number and type of virtual machine 520 and virtual memory 530, inputting the input data (i.e., multiple signals detected by multiple receivers above the underground target volume), and extracting the result data (i.e., a model of the physical properties of the underground target volume).

[0063] refer to Figure 6 A method 600 for determining one or more physical properties of an underground target volume includes receiving 602 multiple signals detected by receivers arranged on a surface, such as reference signals. Figure 1As described above. Typically, one signal is received from each receiver, although in some cases it may be possible to receive signals from only a subset of the total number of receivers. The signals represent the vertical oscillations of the ground measured at the respective receivers, caused by surface waves from environmental seismic noise or active sources, as referenced above. Figure 1 The signal may include metadata about the location and time of the recording.

[0064] The signal can be received directly from the receiver or through one or more intermediate devices. In some examples, additional processing steps may be performed on the signal before or after receiving the 602 signal to optimize it for further processing.

[0065] Example receivers include speedometers or accelerometers. Examples of specific mechanisms for such receivers include a ferromagnetic mass on a spring that moves within an electromagnetic coil in response to surface movement, thereby inducing a current proportional to the ground velocity, which can then be measured.

[0066] Method 600 further includes cross-correlating (604) the signals between a plurality of receivers arranged on a surface to obtain empirical travel time data of surface waves between a plurality of virtual receiver pairs. As described above, this step involves cross-correlating ambient noise signals measured at a pair of receivers to obtain an associated Green's function, which represents the wave field as if one receiver in the pair were a virtual source and the other receiver in the pair were a receiver. As will be well understood by those skilled in the art, information indicating phase velocity as a function of frequency and information indicating group velocity as a function of frequency can be derived from the cross-correlation process. Cross-correlating (604) the signals between the plurality of receivers is to obtain group and / or phase travel time data at multiple frequencies. The frequencies can be continuous, thus obtaining group velocity and / or phase velocity dispersion information, or travel time data can be obtained for a finite number of frequencies. Group travel time refers to the time it takes for a group wave to travel from a virtual source to a receiver. Phase travel time refers to the time it takes for a particular phase component of a wave to travel from a virtual source to a receiver. Since the distance between the virtual source and receiver is known, the group and / or phase velocities or slowness (the slowness being the reciprocal of the velocity) at multiple frequencies can be derived from the correlated Green's function of the receiver pair. Slowness and travel time are practically equivalent, with their scaling factor provided by the distance between the source and receiver.

[0067] A model for the physical properties of an underground target volume can be determined based on at least multiple frequencies of group and / or phase travel time data. (See reference above.) Figure 3The model comprises multiple cells arranged in a two-dimensional grid. Each cell of the model has a corresponding physical property value (e.g., shear wave velocity), which may include a depth distribution. Therefore, the cells of the model collectively provide a three-dimensional model of the physical properties, as each cell provides a one-dimensional model as a function of depth.

[0068] The model for determining the physical properties of the underground target volume includes: selecting a set of multiple receiver pairs (606) for which empirical travel time information has been obtained. (See above reference.) Figure 4A and Figure 4B As described, the wave path between each receiver pair typically traverses two or more cells of the model. In principle, any and all receiver pairs may be suitable for determining the model. However, increasing the number of wave paths used will increase the required computational power or result in longer computation time. Therefore, in practice, it is often beneficial to select a subset of receiver pairs, where the number to be selected depends on the computational power intended to be used or other practical considerations. Several principles can be used to limit the number of receiver pairs selected. As a preliminary problem, due to the reciprocity of cross-correlation, the wave path from position A to position B is the same as the wave path from position B to position A, so only paired receivers need to be selected, regardless of which is considered a virtual source. One method of excluding receiver pairs is to omit those receiver pairs that do not have suitable travel time data, i.e., the cross-correlation of the signals from the pair of receivers does not show an identifiable wave traversing both receivers that can be used to find the travel time between them. Another method of selecting wave paths is to omit receiver pairs with low-quality pickups, such as those with low signal-to-noise ratios, large uncertainties, or those that are outliers compared to adjacent pickups. Any picks that would lead to non-physical or non-geological results are excluded, as are any picks that are not credible for the specific subsurface target volume being investigated. Finally, if the remaining number of receiver pairs still exceeds the desired number, a subsample of receiver pairs can be selected, such as every nth pair or randomly. In this case, other suitable subsamples of receiver pairs can be used consecutively to increase the total number of pairs used. Regardless of the method used to limit the number of receiver pairs (if necessary), the end result is the selection of receiver pairs and the corresponding wave paths, for which empirical group and / or phase travel time information has been obtained, such as multiple velocity or dispersion functions at different frequencies (e.g., group velocity dispersion function and / or phase velocity dispersion function).

[0069] As referenced above Figure 4A and Figure 4B The wave path between the (virtual) source and receiver pair can be a straight line extending between the source and receiver. In some examples, at least some wave paths extend along curves between the respective source and receiver, as referenced above. Figure 4BThe selected wave paths will pass through at least two cells of the first model, having a path segment length in each cell they pass through (including the cells where the wave path begins and ends). The curved wave paths can be determined based on the initial physical property values ​​of the model's cells. By using curved wave paths determined according to the physical property values ​​of the model's cells, the wave paths more closely follow the actual wave path that the wave will travel between the source and receiver. Therefore, this approach improves the accuracy of the results of subsequent methodological processes because it more closely corresponds to the physical reality of surface waves passing through the subsurface target volume.

[0070] As referenced above Figure 4A and Figure 4B In other examples, wave paths are represented by Fresnel zones, which provide a region of many possible paths a wave takes as it propagates from the source to the receiver. A Fresnel zone provides a region covering two or more cells through which a wave might pass, where each cell has a corresponding region of the Fresnel zone with an associated sensitivity value defined by the Fresnel zone. The Fresnel zone can be determined based on initial physical property values ​​of the cells in the model. Because the Fresnel zone defines the region of possible paths a surface wave takes as it travels from the source to the receiver, this approach more closely corresponds to the physical reality of surface waves traversing and scattering within a subsurface target volume.

[0071] The model for determining the physical properties of the subsurface target volume also includes performing a 608-stage tomographic inversion on the time-of-travel data for each of the selected receiver pairs. The tomographic inversion is used to derive values ​​for the physical properties of each element of the model from the group and / or phase time-of-travel data obtained for each of the multiple receiver pairs. As mentioned above, the physical property values ​​for each element can be single values, distributions, or physical properties provided as a function of depth. It should be understood that any appropriate procedure for performing the tomographic inversion can be followed. The tomographic inversion can be performed based on empirical group time-of-travel information or empirical phase dispersion information. In the context of phase dispersion, phase velocity or slowness information can be converted into equivalent time-of-travel information, allowing tomography to be performed on such phase time-of-travel information.

[0072] In some examples, performing tomographic inversion involves two phases: a tomographic phase followed by an inversion phase. In some examples, the tomographic phase is divided into frequency subtasks. Alternatively or additionally, the inversion phase is divided into unit subtasks.

[0073] The tomographic imaging stage tomographically maps the travel time information from each receiver pair into cells of a model of physical properties. Therefore, the result of this process is an empirical model of group and / or phase velocities (depending on the travel time data used), where each cell of the model has a corresponding group or phase velocity value. This process is performed for each of several frequencies to obtain the group or phase velocity value for each cell at each frequency. This is the first stage of the two-stage tomographic inversion (i.e., the tomographic imaging stage).

[0074] The tomographic process involves obtaining an initial velocity model. This initial velocity model includes initial values ​​for the group velocity and / or phase velocity for each cell across multiple frequencies. The initial velocity model sets initial (group or phase) velocity values ​​for each cell (for each given frequency), and the tomographic process refines these initial values ​​iteratively using empirical time-of-travel data. Therefore, the initial velocity model and the corresponding initial velocity values ​​do not need to be highly accurate or high-resolution, although a more accurate initial model may allow the tomographic process to map time-of-travel to each cell more quickly or accurately. A more accurate initial model also reduces the risk of finding local minima rather than global minima in iterative gradient descent, although this problem can be addressed using Monte Carlo methods. In some examples, the initial velocity model is determined based on user input, such as historical data or map information indicating possible physical property values ​​across the subsurface target volume. Alternatively, arbitrarily chosen typical group / phase velocity values ​​can be used as a starting point for each cell. In these examples, any model can be chosen based on estimates of the physical properties of the subsurface target volume.

[0075] The tomographic imaging process also includes determining the simulated group or phase travel time for each of the selected source-receiver pairs using an initial group or phase velocity model. This is performed by identifying the wave path from the source to the receiver (e.g., as a straight ray, a curved ray, or an elliptical / curved Fresnel band) and identifying the cells through which the wave path passes. The simulated travel time based on the initial velocity model is then determined based on the known distance between the source and receiver and the velocity values ​​of each cell through which the wave path passes.

[0076] The tomographic imaging process also includes determining an error value that indicates the difference between the simulated travel time and the empirical travel time for each of the selected source-receiver pairs. For example, for each source-receiver pair, at each frequency, the error value can be the simple difference between the simulated travel time and the empirical travel time, also known as the residual. In some examples, the error value can be a combination of all the differences between the simulated travel time and the empirical travel time for each source-receiver pair. Typically, determining the error value is part of an iterative process; for example, in least-squares inversion methods, the square of the residual is calculated to be minimized through iteration. Other forms of inversion processes use different error values ​​to provide feedback to the initial group velocity model.

[0077] The tomographic process also includes determining an updated velocity model using the error values. The updated model is typically identical to the initial velocity model in all respects, except for the new velocity values ​​associated with at least some cells. In other words, the updated model is an updated version of the initial velocity model that takes into account the error values ​​determined between the empirical or phase travel times and the simulated or phase travel times for each source-receiver pair. This feedback process can involve least squares methods, Markov chain Monte Carlo methods, or other inversion techniques to iteratively update the initial model based on the updated model. The tomographic process can be repeated for each of a finite number of frequencies.

[0078] Determining the simulated travel time, the error value, and the updated velocity model are typically all parts of the subroutine of the first-stage tomography process, which is then iteratively performed using an inversion method (e.g., least squares inversion). The iterative nature of the tomography process means that the determined updated model is used to determine a new simulated travel time for each source-receiver pair. In other words, each time an updated model is determined using the error value obtained from the initial model and the simulated travel time for each source-receiver pair, the resulting updated model is then used as the initial model for the next iteration. Iteration continues until the error value reaches a termination condition, for example, the error value is below a threshold absolute value of the difference between the simulated travel time and the empirical travel time, or below a threshold of the proportional difference between the simulated travel time and the empirical travel time. Another termination condition, which can be used alone or in combination with the threshold error value termination condition, is that the change in the initial model to produce the updated model for the iteration is below a threshold amount or proportion. Thus, if the iteration reaches a stable minimum error value, the iteration can end because further iterations will not bring significant accuracy improvements.

[0079] Once the first-stage tomography process has been performed to obtain a velocity model containing group or phase velocity values ​​for each element (for each frequency), the tomography inversion can proceed to the second stage, a two-stage tomography inversion process, to obtain a resulting model of the physical properties of the subsurface target volume. Similar to the first-stage tomography process, the second-stage inversion process is performed for each of the multiple frequencies to obtain group or phase velocity values ​​for each element at each frequency.

[0080] The inversion process includes obtaining an initial physical property model. This initial physical property model is a preliminary model of the physical properties of the subsurface volume, where the initial model includes the initial physical property values ​​for each of the first plurality of elements. The initial physical model can be as shown in the reference above. Figure 3 The shear wave velocity model 300, and / or the model with reference above. Figure 3 The initial physical model describes any features or variations thereof. The initial model sets initial physical property values ​​for each cell of the model. The inversion process uses the group / phase velocity model obtained from the iterative tomography process described above and further iterative processes to refine these initial physical property values. Therefore, the initial physical property model and initial physical property values ​​do not necessarily need to be a highly accurate or high-resolution model of the subsurface target volume, although a more accurate initial model can increase the expected accuracy of the final result or reduce the computation time required to achieve the final result.

[0081] In some examples, the initial physical property model is determined based on user input, such as historical data or map information indicating possible physical property values ​​across the subsurface target volume. Alternatively, the initial physical model can be determined using received signals, for example, by inverting the group velocity or phase velocity dispersion curves between receivers, which can be calculated from the cross-correlation of signals as described above, to find out using a coarser grid or a faster method. As a last resort, a typical value of any chosen physical property can be used as a starting point for each cell, which can be based on the estimated physical properties of the subsurface target volume. In some examples, the initial physical model can be determined based on empirical phase dispersion data between source-receiver pairs. Phase velocity information is obtained as a point in a two-dimensional phase space (dispersion spectrum), which can be obtained through a two-dimensional transformation of a waveform array (time-distance space), such transformation as Radon transform, tilt stacking, FK, or any other suitable method obvious to those skilled in the art. The generation of dispersion curves showing phase velocity as a function of frequency can provide a preliminary three-dimensional model of shear velocity in a three-dimensional model. However, since the acquisition of phase velocity is essentially an average frequency-dependent velocity between measurement locations, the resolution of this model is relatively low. Therefore, a physical model derived solely from empirical phase dispersion data can be used as the initial model for inversion, which also takes into account empirical group dispersion data to obtain a more accurate final three-dimensional model of the subsurface target volume.

[0082] The inversion process also includes determining the simulated surface wave velocity for each element based on the initial model. More specifically, using forward modeling methods, the phase and / or group velocities of each element in the initial physical model are derived using the corresponding physical property values ​​of the given elements. For example, the physical property value for each element could be the shear wave velocity. The shear wave velocity value for each element can be used to calculate the corresponding phase velocity dispersion function (and thus the phase velocity at a given frequency) using the propagation matrix method introduced by Thomson (1950, “Transmission of elastic waves through a stratified solid medium”, Journal of Applied Physics, 21(2), 89-93) and Haskell (1953, “The dispersion of surface waves on multilayered media”, Bulletin of the Seismological Society of America, 43(1), 17-34). Those skilled in the art will be well aware of this and other methods that can be used to forward model groups or phase velocities (or dispersion functions) from shear wave velocity values ​​(or depth functions).

[0083] The inversion process also includes determining an error value indicating the difference between the simulated velocity and the empirical velocity for each element of the model. This error value can be determined for each element in a manner similar to that described above regarding the tomographic process, where the error between the simulated travel time and the empirical travel time for each source-receiver pair is determined. As with the tomographic process, determining the error value is typically part of an iterative process; for example, in the least squares inversion method, the square of the residual is calculated to minimize it through iteration. Other forms of inversion processes use different error values ​​to provide feedback to the initial model of the physical properties of the subsurface target volume.

[0084] The inversion process also includes determining an updated physical model based on the determined error values. The updated physical property model is typically identical to the initial physical model in all respects, except for new physical property values ​​associated with at least some elements. In other words, the updated model is an updated version of the initial physical model that takes into account the error values ​​determined between the empirical surface wave velocity and the simulated surface wave velocity for each element of the model. This feedback process may involve least squares methods, Markov chain Monte Carlo methods, or other inversion techniques to iteratively update the initial physical model based on the updated physical model.

[0085] Determining the simulated surface wave velocity, determining the error value, and determining the updated physical property model are typically all parts of the subroutine of the second-stage inversion process, which is then iterated according to an inversion method (e.g., least squares inversion). The iterative nature of the inversion process means that the determined updated model is used to determine the new surface wave velocity for each element. In other words, each time an updated model is determined using the error value obtained from the initial model and the simulated velocity of each element, the resulting updated model is then used as the initial model for the next iteration. Iteration continues until the error value reaches a termination condition, such as the termination criterion described above regarding the tomographic imaging process.

[0086] In an alternative to two-stage tomographic inversion, performing 608 tomographic inversion can be a single-stage process, where tomography and inversion are combined. This means that the iterative inversion stage, which finds physical property values ​​for the model, is performed for each wavepath between the corresponding source-receiver pairs, rather than for each cell of the model. This means that in a single-stage process, there is no need for a first-stage tomographic process that maps surface wave velocities between source-receiver pairs to cells, because the physical property model inversion process performs inversion for each wavepath, not for each cell. In other words, two-stage tomographic inversion first uses time-of-travel tomography to map the time-of-travel information along the wavepath to surface wave velocity information in the cell grid. Then, two-stage tomographic inversion performs cell-by-cell inversion of the surface wave velocity information (surface wave velocity as a function of frequency) to obtain the corresponding shear wave velocity function (as a function of depth) for each cell. In contrast, the single-stage process performs a direct inversion of surface wave velocity information for each wave path (for each source-receiver pair) to obtain the shear wave velocity function for that wave path, and then tomographically maps it to each cell of the model.

[0087] As referenced above Figure 6 The aforementioned cross-correlation, selection, execution of tomographic inversion, and sorting involve one or more parallel processing routines in a cloud computing environment 500, such as those mentioned above. Figure 5 As described above. In some examples, parallel processing routines divide a portion of the method into subtasks based on frequency, as referenced below. Figure 7 As stated above.

[0088] refer to Figure 7 The frequency-based parallel processing routine 700 includes multiple received signals ( Figure 6 Within 602 (as described above), multiple frequency sub-ranges f1, f2, f3, ..., f are determined. NSubranges can be provided by the user equipment while receiving multiple signals, for example, in the metadata accompanying the multiple signals and selected by the user sending the multiple signals to the cloud computing environment 500. In some examples, the virtual state machine 502 has been programmed with a protocol to determine frequency subranges based on the multiple signals, such as dividing the total frequency range of the multiple signals into N (e.g., 100 or some other predetermined number) subranges of equal bandwidth, or subranges on a logarithmic scale. In some examples, the number of subranges and their bandwidth are determined based on the expected computation time and / or the expected amount of computational resources to be used.

[0089] Frequency parallel processing routine 700 divides a portion of the method into subtasks T1, T2, T3, ..., T N Each subtask is associated with multiple frequency subranges f1, f2, f3, ..., f N One of the related elements is mentioned. Subtask T is recorded in work queue 704, and the position of each subtask in work queue 704 is based on its subtask number. Work queue 704 can be a one-dimensional queue or an array with two or more dimensions.

[0090] Each subtask includes any information required to independently execute a portion of the method divided into subtasks for its specified frequency subrange. In some examples, a portion of the method divided into subtasks includes all of cross-correlation, selection, execution of tomography inversion, and finishing. In this case, these portions of the method can be referenced above. Figure 6 The same approach is performed for a specific frequency subrange. In some examples, the subtask includes information about where the results of the subtask should be recorded in the resulting physical property model 300 during the finishing process 610. Since surface waves are affected by the physical properties of the medium within a depth range of approximately one wavelength from the surface, as mentioned above, the results of each frequency subtask provide physical property values ​​for different depths in the model (e.g., comparing the results of two frequency subranges indicates the influence of layers in the subsurface target volume that only significantly affect the lower frequency subrange).

[0091] The virtual machine 520 in the cloud computing environment 500 can execute multiple processes simultaneously (using virtual memory 530 for any short-term storage during process execution) because it is created by a distributed network of computing devices. This can be understood as having multiple workers W1, W2, W3, ..., W pA worker pool 706 is used, where P represents the number of independent processes that virtual machine 520 can execute concurrently. The number of workers P can be determined by virtual state machine 510 at the start of the method, or it can be determined by user input based on how much computing resources the user wishes to use. Each worker W in worker pool 706 picks up a subtask T, performs the operation associated with that subtask, and records the result R in result queue 708. Therefore, compared to conventional techniques, computation time is reduced by utilizing the large amount of processing resources available across the network of computing devices in a cloud computing environment. Once all subtasks T are completed, result queue 708 will be filled with results R1, R2, ..., R N The worker W selects a subtask T to execute by choosing the next available subtask (the one with the lowest subtask number) from the work queue 704 for each idle worker, or by some other protocol, or even by random selection. As a worker completes its subtask, it selects a new subtask from the work queue until all subtasks are completed. The result R is recorded in the result queue 708 at the position corresponding to the subtask number, thus preserving the order and structure of the subtasks in the work queue 704. This means that retrieving results from the result queue 708 is faster and simpler. In some examples, each result R may include multiple values ​​from a model that will be used to determine physical properties.

[0092] In some examples, the worker W can further divide its subtasks into smaller parts; for example, if each worker corresponds to a virtual CPU with four cores, then four processes can be executed simultaneously.

[0093] In some examples, the frequency-parallel processing routine 700 includes processing the results R1, R2, ..., R from the work queue. N Combine 710 into a single result. This can be part of the general method 600, specifically the refinement of 610.

[0094] Although the above example of a parallel processing routine is frequency parallel processing routine 700, the same principle applies to creating subtasks based on model cells, wave paths, travel times, or other parameters of the method 600 used to determine the physical model.

[0095] For example, when creating subtasks based on model elements, the number of subtasks will correspond to the number of elements in the model to be determined. This could be in the thousands, such as 5000. By dividing the process into subtasks executed in parallel by workers W, the time required to determine the physical property values ​​of the model is significantly reduced. For example, in an example with approximately 500 workers, the computation time would be approximately 500 times faster. In some examples, each model element subtask includes selecting receiver pairs and corresponding cross-correlation signals whose surface wave paths traverse the designated model element for that subtask. The travel time information can then be combined to generate group and / or phase velocity information for that element, as referenced above. Figure 6 The inversion calculation is performed to obtain the corresponding physical property values ​​of the unit. After all model unit subtasks have been completed, the results queue 708 will contain the final information of the model, including the physical values ​​of the underground target values.

[0096] In some examples, when the surface wave path is divided into subtasks, each path subtask can be assigned to a pair of receivers. This can be done before cross-correlation 604, so that each cross-correlation is performed by a separate path subtask, or after selection 606, only for the selected receiver pair (and its corresponding surface wave path). Path subtasks are well-suited for the single-stage tomographic inversion process described above because the inversion calculation is performed over the entire surface wave path, and the calculated values ​​are then applied to each model cell through which the surface wave path passes (optionally based on the proportion of the surface wave path in each model cell, such as...). Figure 4A and Figure 4B The L shown i (Weighted).

[0097] Typically, subtasks, which divide one or more parts of a method into, can be any set of instructions that, when executed, facilitate the execution of a method based on a model used to determine the volume of a subsurface target. Subtasks may include relevant data (e.g., a signal, a portion of a signal, or an intermediate result of the method) or a location from which data to be processed in the subtask is retrieved. Subtasks may also include labels used to identify the subtask and how it should be processed.

[0098] refer to Figure 8 A method 800 for determining the physical properties of an underground target volume can be executed via a user device interacting with a cloud computing environment 500 created by the network of computing devices. This method includes sending multiple signals to the network of computing devices. In some examples, this is achieved via an API gateway or by storing the multiple signals in accessible storage 540 within the cloud computing environment 500.

[0099] This method includes sending instructions to network commands to perform actions as described in the reference above. Figure 6The method described herein. This can be implemented via an orchestrator that manages HTTP requests to and from the cloud computing environment 500. Instructions may include file addresses where multiple signals for performing the method are located, if stored on accessible memory 540. Instructions may also include details of the model to be determined (size and cell size) and / or an initial model for initiating tomographic inversion. Instructions may also include user input that defines the desired level of accuracy (e.g., how small the difference between simulated travel time and empirical travel time needs to be to end the inversion process), the amount and type of computing resources to be used (e.g., the number of virtual processors for running on virtual machine 520, the number of subtasks T to be used, and / or the number of workers W to be used), or the desired computation time, based on which virtual state machine 510 calculates the amount of computing resources required for deployment to achieve that computation time.

[0100] The method also includes receiving the determined model from the network in response to the network's computing device performing the above-mentioned reference in a cloud computing environment 500. Figure 6 The method described herein. Therefore, the method on the user device controls the parallelization parameters in the cloud computing environment 500 to achieve fast and efficient model determination.

[0101] In some examples, user devices operate through a virtual private cloud, which provides secure network traffic to and from computing devices.

[0102] In some examples, the virtual state machine 510 in the cloud computing environment 500 has pre-programmed checkpoints for user input. This can be useful for testing the parallelization methods described herein, as individual parts of the overall process can be executed and reviewed without the rest. In some examples, checkpoints are triggered by unexpected events identified in the process, at which point the virtual state machine 510 can pause the process and request further instructions from the user device on how to continue.

[0103] In some examples, the method includes preprocessing multiple signals into a format for use by the network, or the instructions include how to preprocess multiple signals.

[0104] refer to Figure 9An example output of the method described in this paper is a final shear wave velocity model of the subsurface target volume. The subsurface target volume extends to a depth of 100 m in the x and y directions (z direction). The shear wave velocity values ​​are represented by shading in the figure, and the transitions between regions of different shear wave velocities are visible, indicating the different components or structures of parts of the subsurface target region. Using the method described in this paper, the final shear wave velocity model can be determined with high resolution and accuracy, and with shorter computation time. Such a subsurface model can be used to better understand the suitability of the subsurface target volume for man-made structures supporting the top or interior of the subsurface target volume.

[0105] Alternatively, additional physical properties of the subsurface target volume can be determined from the final shear wave velocity model, for example, using calculations from one or more of Equations 1 and 2. These further physical properties can be recorded in the shear wave velocity model itself or output separately to an output device.

[0106] In addition to or instead of shear wave velocity, other physical properties associated with shear wave velocity can be used. For example, the longitudinal wave (P-wave) velocity Vp and the shear wave (or transverse wave / S-wave) velocity Vs are expressed by the following equations from linear elasticity theory with the elastic modulus. shear modulus and density Related to other physical properties of the ground: Equation 1 (P-wave velocity): Equation 2 (Shear wave velocity): Figure 10 A block diagram of one embodiment of a computing device 1000 is shown, within which a set of instructions can be executed to cause the computing device to perform, as described above, as a user device. Figure 8The method described herein. In alternative embodiments, the computing device may be connected (e.g., networked) to other machines in a local area network (LAN), intranet, extranet, or the Internet. The computing device may operate at the capacity of a server or client machine in a client-server network environment, or at the capacity of a peer-to-peer (or distributed) network environment. The computing device may be a personal computer (PC), tablet computer, set-top box (STB), personal digital assistant (PDA), cellular phone, network device, server, network router, switch, or bridge, or any machine capable of executing a set of instructions (sequential or non-sequential) specifying the operations to be performed by the machine. Furthermore, although only a single computing device is shown, the term "computing device" should also be understood to include any collection of machines (e.g., computers) that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more methods discussed herein.

[0107] The example computing device 1000 includes a processor 1002, a main memory 1004 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM), such as synchronous DRAM (SDRAM) or Rambus DRAM (RDRAM), etc.), a static memory 1006 (e.g., flash memory, static random access memory (SRAM), etc.), and auxiliary memory (e.g., data storage device 1018), which communicate with each other via a bus 1030.

[0108] Processor 1002 represents one or more general-purpose processors, such as microprocessors, central processing units, etc. More specifically, processor 1002 may be a Complex Instruction Set Computing (CISC) microprocessor, a Reduced Instruction Set Computing (RISC) microprocessor, a Very Long Instruction Word (VLIW) microprocessor, a processor implementing other instruction sets, or a processor implementing a combination of instruction sets. Processor 1002 may also be one or more special-purpose processors, such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), network processors, etc. Processor 1002 is configured to execute processing logic (instructions 1022) for performing the operations and steps discussed herein.

[0109] The computing device 1000 may further include a network interface device 1008. The computing device 1000 may also include a video display unit 1010 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), an alphanumeric input device 1012 (e.g., a keyboard or a touch screen), a cursor control device 1014 (e.g., a mouse or a touch screen), and an audio device 1016 (e.g., a speaker).

[0110] Obviously, Figure 10Some features of the computer device 1000 shown may be absent. For example, one or more computing devices 1000 may not require a display device 1010 (or any associated adapter). This may be the case, for example, for a specific server-side computer device 1000 that is used only for its processing power and does not need to display information to a user. Similarly, a user input device 1012 may not be necessary. In its simplest form, the computer device 1000 includes a processor 1002 and a memory 1004.

[0111] Data storage device 1018 may include one or more computer-readable storage media (or more particularly, one or more non-transient computer-readable storage media) 1028 on which one or more sets of instructions 1022 are stored, the instructions embodying any one or more methods or functions described herein. Instructions 1022 may also reside wholly or at least partially within main memory 1004 and / or processor 1002 during execution by computer system 1000, which also constitute computer-readable storage media.

[0112] The above reference Figure 8 The described methods can be implemented by a computer program. The computer program may include computer code arranged to instruct a computer to perform the functions of these methods. The computer program and / or code for performing these methods may be provided to a device such as a computer on one or more computer-readable media or more generally, a computer program product. The computer-readable media may be transient or non-transient. The one or more computer-readable media may be, for example, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, or propagation media for data transmission, such as for downloading code via the Internet. Alternatively, the one or more computer-readable media may take the form of one or more physical computer-readable media, such as semiconductor or solid-state memory, magnetic tape, removable computer disk, random access memory (RAM), read-only memory (ROM), rigid disk, and optical disk, such as CD-ROM, CD-R / W, or DVD.

[0113] In one implementation, the modules, components and other features described herein may be implemented as discrete components or integrated into the functionality of hardware components such as ASICs, FPGAs, DSPs or similar devices.

[0114] A "hardware component" is a tangible (e.g., non-transient) physical component (e.g., a collection of one or more processors) capable of performing certain operations and which can be configured or arranged in a physical manner. A hardware component may include dedicated circuitry or logic permanently configured to perform certain operations. A hardware component may be or include dedicated processors, such as field-programmable gate arrays (FPGAs) or ASICs. A hardware component may also include programmable logic or circuitry temporarily configured by software to perform certain operations.

[0115] Therefore, the phrase “hardware component” should be understood to include tangible entities that can be physically constructed, permanently configured (e.g., hardwired) or temporarily configured (e.g., programmed) to operate in a certain way or perform certain operations described herein.

[0116] Furthermore, modules and components can be implemented as firmware or functional circuitry within a hardware device. Additionally, modules and components can be implemented as any combination of hardware devices and software components, or solely as software (e.g., code stored in or otherwise embodied on a computer-readable medium or transmission medium).

[0117] Figure 11 A block diagram of one embodiment of system 1100 is shown, within which a set of instructions can be executed to cause a network of computing devices to perform the methods described above as user equipment. For example, the network includes multiple computing devices 1110 to provide, according to any standard cloud computing approach, as referenced above. Figure 5 The cloud computing environment 500 is described above. The computing device 1110 may have a specific computer-readable storage medium 1128, which has the features described in the above reference. Figure 10 Any features of the computer-readable storage medium 1028 described.

[0118] The computing devices 1110 forming the network are connected via a communication connection 1130, which, depending on any typical cloud computing setup, can be a local connection (e.g., via a local area network (LAN)), an intranet, an extranet, a virtual private cloud, or the Internet.

[0119] The system may also include a physical storage device 1120 for storing data and accessible to a network of computing devices via a communication connection 1130.

[0120] System 1100 may also include the above references. Figure 10 The computing device 1000 and as referenced above Figure 8 The computer-readable storage medium 1028 and computing device 1000 perform operations according to the above reference. Figure 8 The user equipment described in the method.

[0121] In some examples, the network connectivity of computing device 1110 is separate from the connectivity of user devices (computing device 1000) and / or physical storage devices 1120.

[0122] Unless otherwise expressly stated, it will be apparent from the following discussion that, throughout the description, discussions using terms such as “receive,” “determine,” “compare,” “calculate,” “average,” “identify,” “update,” “solve,” and “output” refer to the actions and processes of a computer system or similar electronic computing device that manipulates and converts data represented as physical (electronic) quantities in the registers and memories of the computer system into other data similarly represented as physical quantities in the computer system’s memory or registers or other such information storage, transmission, or display devices.

[0123] It should be understood that the above description is intended to be illustrative and not restrictive. Many other embodiments will be apparent to those skilled in the art upon reading and understanding the above description. Although this disclosure has been described with reference to specific exemplary embodiments, it should be recognized that this disclosure is not limited to the described embodiments but can be implemented with modifications and changes within the spirit and scope of the appended claims. Therefore, the specification and drawings should be considered illustrative and not restrictive. Consequently, the scope of this disclosure should be determined by reference to the appended claims and the full scope of their equivalents.

Claims

1. A method for determining the physical properties of an underground target volume using a network of computing devices, wherein, The model comprises multiple units, each with corresponding physical property values, and the method includes: Receive multiple signals detected by multiple receivers arranged on a surface above the underground target volume, wherein each of the multiple signals is detected by a corresponding receiver among the multiple receivers; Cross-correlation is performed on the signals between the multiple receivers to obtain empirical travel time data of surface waves between multiple receiver pairs; Select a set of the plurality of receiver pairs, each corresponding receiver pair having a corresponding surface wave path between the receivers of the receiver pair; Tomographic inversion is performed on the empirical travel time data between receivers for each receiver pair in the set to obtain the corresponding physical property values ​​for each cell; and The obtained physical property values ​​are then incorporated into the model. The process involves dividing one or more of the following tasks into subtasks: cross-correlation, selection, execution of tomographic inversion, and sorting. These subtasks are distributed among the computing devices of the network and executed in parallel.

2. The method according to claim 1, wherein, The subtasks are frequency subtasks, wherein each frequency subtask corresponds to a corresponding frequency subrange of the total frequency range of the plurality of signals.

3. The method according to any of the preceding claims, wherein, The subtask is a model unit subtask, wherein each model unit subtask corresponds to a corresponding unit of the model.

4. The method according to any of the preceding claims, wherein, The subtasks are path subtasks, where each path subtask corresponds to a specific surface wave path.

5. The method according to any of the preceding claims, wherein, The subtasks are travel time subtasks, where each travel time subtask corresponds to the corresponding obtained experience travel time data.

6. The method according to any of the preceding claims, wherein, The subtasks are recorded in a work queue, wherein each subtask has a position in the work queue, and the result of each completed subtask is recorded at a result position in a result queue, which corresponds to the position of the completed subtask in the work queue.

7. The method according to any of the preceding claims further includes storing the status of executed subtasks and work records in a central database via an application programming interface.

8. The method according to any of the preceding claims further includes using a central service that runs a daemon process that observes subtasks and their working status to take appropriate action to execute new computational tasks when conditions are met.

9. The method according to any of the preceding claims, wherein, Performing tomographic inversion includes: An initial model with the initial physical property values ​​of the plurality of units is obtained; For each receiver pair in the set, the simulated travel time is determined using the initial physical property values, which are associated with the elements of the model traversed by the surface wave path between the receivers of the receiver pair; and New physical property values ​​are determined for the multiple units based on simulated travel time and empirical travel time.

10. A method for determining the physical properties of the volume of an underground target using a model, the method comprising: The network of computing devices transmits multiple signals detected by multiple receivers arranged on a surface above the underground target volume, wherein each of the multiple signals is detected by a corresponding receiver of the multiple receivers; Sending instructions to the network, the instructions causing the network to perform the method according to any of the preceding claims; and A model of the physical properties of the underground target volume received from the network.

11. The method according to claim 10, wherein, The instruction includes calculation parameters, which include one or more of the following: The dimensions of the underground target volume; The dimensions of the plurality of units of the model; The target precision of the physical property values; Target computation time; The amount of computing resources required; The number of subtasks to be created; The type of subtask to be created.

12. The method according to claim 10 or 11, wherein, The method includes: Receive a request from the network for further instructions because the network has reached a programming checkpoint.

13. A user equipment, comprising: One or more processors; as well as One or more memories thereon store computer-readable instructions configured to cause the one or more processors to perform operations including the method according to any one of claims 10 to 12.

14. A computer-readable medium comprising instructions that, when executed by a user equipment according to claim 14, cause the one or more processors to perform operations including the method according to any one of claims 10 to 12.

15. A system comprising: A network of computing devices configured to perform the method according to any one of claims 1 to 9.

16. The system of claim 15, comprising: The user equipment according to claim 13.

17. A computer-readable medium comprising instructions that, when executed by a system according to claim 15 or 16, cause the network to perform operations including the method according to any one of claims 1 to 9.

Images