A method for generating a single channel three-dimensional river depositional facies model

Abstract

Embodiments of the present application disclose a method for generating a single-channel three-dimensional river sedimentary facies model. A specific embodiment of the method comprises: iteratively performing a correction step based on an initial channel center line, the correction step comprising: in each round of correction, performing a smoothing enhancement or migration enhancement adjustment operation according to a deviation direction; when detecting that the adjustment operation types of adjacent two rounds of correction steps are different, performing adjustment parameter attenuation; based on the initial channel center line after iterative correction, performing a preset number of migration steps; between adjacent time steps, judging whether to perform a channel straightening operation based on an adaptive threshold related to a local channel width; generating channel geometries of the number of time steps; and rendering the channel geometries of each time step to generate a single-channel three-dimensional river sedimentary facies model. Thus, the problem of insufficient bending control precision is solved, and real simulation of the channel migration process is achieved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method for generating a three-dimensional river sedimentary facies model of a single channel. Background Technology

[0002] Meandering rivers are among the most common sedimentary units in alluvial plain sedimentary systems, and their migration processes are of great significance for understanding river geomorphological evolution and predicting the distribution of underground oil and gas reservoirs. Related technologies generate river centerlines by simulating the physical evolution of river channels, but the inherent random disturbance characteristics limit the ability to accurately control the meandering of the river channel.

[0003] When the control precision of the river channel centerline curvature distribution cannot meet the preset requirements, it is difficult to obtain a training image dataset that conforms to the actual geological characteristics. Summary of the Invention

[0004] This disclosure is provided to briefly introduce the concepts, which will be described in detail in the subsequent Detailed Description section. This disclosure is not intended to identify key or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0005] In a first aspect, embodiments of the present invention provide a method for generating a three-dimensional river sedimentary facies model of a single channel. The method includes: using the channel initiation point as an initial node, evolving and generating nodes of an initial channel centerline; iteratively executing a correction step based on the initial channel centerline, the correction step including: in each correction round, calculating the deviation between the actual curvature of the current channel centerline and a preset target curvature, and performing a smoothing enhancement or migration enhancement adjustment operation according to the deviation direction; detecting the adjustment operation type of two adjacent correction rounds, and when the adjustment operation types of two adjacent correction rounds are detected to be different, performing adjustment parameter attenuation; based on the iteratively corrected initial channel centerline, performing a preset number of migration steps to obtain a channel centerline with a number of time steps, wherein the migration step includes: in each time step, calculating the normal displacement of each node of the centerline based on the weighted integral of the upstream curvature, and updating the centerline node coordinates according to the normal displacement; between adjacent time steps, determining whether to perform a channel straightening operation based on an adaptive threshold related to the local channel width. Based on the channel centerline of the specified number of time steps, channel geometry of the specified number of time steps is generated; the channel geometry of each time step is rendered to generate a single-channel three-dimensional fluvial sedimentary facies model.

[0006] In a second aspect, embodiments of the present invention provide a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for generating a single-channel three-dimensional fluvial sedimentary facies model as described in the first aspect. Attached Figure Description

[0007] The above and other features, advantages, and aspects of the various embodiments of the present invention will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0008] Figure 1 This is a flowchart of an embodiment of a method for generating a single-channel three-dimensional fluvial sedimentary facies model according to the present invention;

[0009] Figure 2 A schematic diagram showing the cross-sectional morphology and parameters of the sedimentary facies is presented;

[0010] Figure 3 A schematic diagram of the centerlines of multiple river channels is shown;

[0011] Figure 4 A schematic diagram of a scenario related to straightening a bend is shown;

[0012] Figure 5 This paper presents a comparative example of river migration under the influence of heterogeneous vegetation and migration without vegetation influence;

[0013] Figure 6 A schematic diagram illustrating the effect of rendering priority technology is shown;

[0014] Figure 7 The comparison between whether the voxel clearing mechanism is executed and not is shown;

[0015] Figure 8 This is a schematic diagram of the basic structure of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0016] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0017] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0018] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0019] It should be noted that the concepts of "first" and "second" mentioned in this invention are only used to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0020] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0021] In one or more embodiments of the present invention, a two-stage optimization method is proposed. This method combines initial physical evolution simulation with subsequent statistical correction, achieving effective decoupling between the evolution of the river centerline and the statistical control of meander. This framework eliminates random meander deviations through post-processing statistical correction techniques, achieving precise control of river meander while maintaining geological accuracy. This improves the control accuracy of the river centerline meander distribution, obtains a three-dimensional river sedimentary facies model with controllable distribution characteristics, and ultimately obtains an accurate training image dataset.

[0022] In one or more embodiments of the present invention, the river meandering phenomenon is considered as a deterministic oscillatory process. The recursive formula for the physical evolution process of this model is:

[0023]

[0024] θᵢ represents the river deflection angle at step i, b1 and b2 are preset coefficients, and εᵢ is the intensity of random disturbance, which is positively correlated with the set curvature value. This stage can generate an initial river centerline with physical evolution characteristics, but due to the randomness of εᵢ, the curvature of the generated river often deviates from the preset target value.

[0025] In one or more embodiments of the present invention, a bidirectional adjustment strategy is provided: when the actual curvature When Gaussian weighted neighborhood smoothing is used to reduce curvature; when At that time, coordinate migration based on local curvature enhances the curvature.

[0026] In one or more embodiments of the present invention, an oscillation detection mechanism and a parameter decay mechanism are provided to prevent continuous oscillations between smoothing and transition operations. When two consecutive operations of different types occur, it is recorded as a jump. The system maintains a smoothing intensity parameter and a transition distance parameter respectively, and implements different decays based on the jump direction. When smoothing jumps to transition, the smoothing intensity parameter is decayed to prevent over-smoothing; when transitioning jumps to smoothing, the transition distance parameter is decayed to prevent over-transition. The decay parameter can be set to a fixed decay factor; for example, the parameter decreases by 15% for each jump.

[0027] In one or more embodiments of the present invention, the dynamic threshold judgment based on "local channel width" rather than fixed distance truncation makes the formation of oxbow lakes conform to the laws of fluid dynamics.

[0028] In one or more embodiments of the present invention, such as Figure 2 As shown, Figure 2 A schematic diagram of the cross-sectional morphology and parameters of the sedimentary facies is shown. An alternative geometry method is used for the cross-section of the natural levee. The lower part of the natural levee exhibits a linear depth decrease from the centerline to the boundary, with the depth parameter extending downwards from the centerline and transitioning linearly to the outer boundary. This eliminates unreasonable gaps between the bottom space and the channel when rendering the natural levee voxels.

[0029] In one or more embodiments of the present invention, a vegetation heterogeneity distribution mechanism can be incorporated to generate training images of meandering rivers. The spatial heterogeneity of vegetation distribution further increases the complexity of the river migration process, with different vegetation cover areas exhibiting varying erosion resistance, resulting in a non-uniform migration rate distribution. Before performing the migration, the vegetation influence field is pre-calculated, and then the vegetation influence is used as a constraint on the migration distance during the migration update phase.

[0030] Please refer to Figure 1 This illustrates a flowchart of an embodiment of a method for generating a single-channel three-dimensional fluvial sedimentary facies model according to the present invention. Figure 1 The method for generating a single-channel three-dimensional fluvial sedimentary facies model, as shown, includes the following steps:

[0031] Step 101: Using the starting point of the river channel as the initial node, generate each node of the initial river channel centerline.

[0032] In this embodiment, the execution entity (e.g., a server and / or terminal device) of the method for generating a single-channel three-dimensional river sedimentary facies model can use the starting point of the river channel as the initial node, and based on a deterministic oscillatory recursive relationship and a random perturbation term, evolve and generate various nodes of the initial channel centerline to generate an initial channel centerline with physical evolution characteristics by simulating the physical evolution process of the river channel. Optionally, the random perturbation term is positively correlated with the target curvature value.

[0033] Step 102: Iteratively perform correction steps based on the initial channel centerline. The correction steps include: in each round of correction, calculating the deviation between the actual curvature of the current channel centerline and the preset target curvature, and performing smoothing enhancement or migration enhancement adjustment operations according to the direction of the deviation.

[0034] Step 103: Detect the adjustment operation type of two adjacent calibration steps. When the adjustment operation types of two adjacent calibration steps are different, perform adjustment parameter attenuation.

[0035] Step 104: Based on the initial channel centerline that has been iteratively corrected, perform a preset number of migration steps to obtain the channel centerline with the number of time steps.

[0036] Step 105: Between adjacent time steps, determine whether to perform the straightening operation of the river channel based on an adaptive threshold related to the local river width.

[0037] Step 106: Generate the river geometry for the specified number of time steps based on the river centerline of the specified number of time steps.

[0038] Step 107: Render the channel geometry at each time step to generate a single-channel three-dimensional river sedimentary facies model.

[0039] In this embodiment, a complete initial channel centerline can include the coordinate sequence of all nodes. After this complete initial centerline is generated, migration calculations are performed to generate a series of historical centerlines; at this time, only coordinate data is generated, and voxel rendering is not involved.

[0040] The river channel migration is based on the position of the entire centerline from the previous time step. Specifically, the centerline coordinates at step T+1 are obtained by lateral movement based on the coordinates at step T, according to the calculated normal migration rate (e.g., influenced by curvature, flow velocity, and vegetation). Migration stopping conditions can include a stopping state determined by a preset number of migration steps. The simulation stops after a set number of iterations (e.g., 60). For an example, please refer to [reference needed]. Figure 3 , Figure 3 A schematic diagram of the centerlines of multiple river channels is shown.

[0041] Optionally, the generation of the river geometry can be performed after each migration step.

[0042] Between adjacent migration steps, it can be determined whether to perform a bend straightening operation. If the determination result is to perform bend straightening, then bend straightening is performed. For an example, please refer to [link to example]. Figure 4 , Figure 4 A schematic diagram of a scenario related to straightening a bend is shown.

[0043] Once all geometric simulations (including migrations) are complete, the system obtains a set of centerline data. Only then does it invoke the rendering module to convert these lines into voxels in a 3D mesh.

[0044] It should be noted that the method provided in this embodiment ensures the geological rationality of the initial river channel morphology through physical evolution simulation, providing an initial geometric shape that conforms to the laws of mechanics for subsequent migration simulation; iterative correction accurately converges the actual curvature to the preset target value; and the weighted integral of the upstream curvature history characterizes the memory effect of the river system, realizing the nonlocal simulation of river channel migration. This ensures that the migration at the current location is not only controlled by the local curvature but also influenced by the upstream historical curvature, accurately reflecting the inertial characteristics and hysteresis response mechanism of natural rivers. The adaptive threshold is more in line with the laws of fluid mechanics and can dynamically adjust the cutoff standard according to the width of the river channel, effectively preventing unreasonable self-intersections or narrow necks during the migration process.

[0045] The bidirectional adjustment strategy (smoothing enhancement and migration enhancement) in the calibration step is a potential source of oscillations. An oscillation detection mechanism monitors the continuity of operation types in real time. Once a jump between smoothing and migration is detected, parameter decay is immediately triggered to reduce the aggressiveness of subsequent adjustments. This constraint ensures the stability of the calibration step and prevents algorithm divergence due to over-adjustment. In other words, through oscillation detection and parameter decay mechanisms, the system guarantees algorithm convergence in large-scale iterations, completely avoiding the divergence inherent in traditional physical simulation models, and ensuring the robustness and repeatability of the generation process.

[0046] The adaptive necking mechanism and the migration mechanism complement each other. Migration increases curvature, leading to excessive curvature that triggers adaptive truncation, thereby resetting the river channel morphology. This ensures that the river channel does not become infinitely entangled during long-term simulations, maintaining geological consistency.

[0047] By generating and rendering geometric bodies in multiple time-series phases, the system fully reproduces the geological history of river channel migration and evolution, forming a three-dimensional sedimentary facies model that conforms to the actual geological conditions. Through the synergistic effect of the above-mentioned technical features, the system achieves a realistic simulation of river channel sedimentary facies, improves the consistency between the single-channel three-dimensional river sedimentary facies model and the actual geological conditions, and provides high-quality training images with both geological authenticity and statistical representativeness for deep learning applications.

[0048] In step 101, the initial river centerline can be generated by taking the starting point of the river channel as the initial node and superimposing random disturbance terms on a deterministic oscillation recursive relationship, so as to generate an initial river centerline with physical evolution characteristics by simulating the physical evolution process of the river channel.

[0049] The deterministic oscillation recursion determines the deflection angle of the current node based on the deflection angles of the preceding nodes and the model coefficients. An example of the recursion expression is shown below:

[0050]

[0051] in, Let represent the river channel deflection angle at step i, and b1 and b2 are preset coefficients. This recursive relationship treats the river meandering phenomenon as a deterministic oscillation process, which can generate a river centerline shape with physical evolution characteristics.

[0052] The intensity of the random disturbance is determined by superimposing a deterministic oscillation recursive relationship. This is used as a random disturbance term to simulate the uncertainties in the river channel evolution process. It is positively correlated with the target curvature value, that is, the greater the target curvature, the greater the intensity of the random disturbance term.

[0053] It should be noted that the initial channel centerline generated by superimposing a random perturbation term onto a deterministic oscillatory recursive relationship possesses both physical evolution characteristics and a certain degree of randomness. However, due to the presence of the random perturbation term, the curvature of the generated channel often deviates from the preset target value. Experiments show that when the target curvature is set between 1.3 and 1.7, the actual generated curvature differs from the target by an average of 14.77%.

[0054] Step 102 provides a specific implementation method for iterative correction of curvature. This method uses statistical correction to iteratively correct the curvature of the current river channel centerline, achieving precise target curvature control.

[0055] In some embodiments, step 102 may include: calculating the numerical deviation between the actual curvature of the current river channel centerline and the preset target curvature; when the actual curvature is greater than the target curvature, performing Gaussian weighted neighborhood smoothing enhancement; when the actual curvature is less than the target curvature, performing coordinate migration enhancement based on local curvature.

[0056] First, calculate the numerical deviation between the actual curvature of the current river channel centerline and the preset target curvature.

[0057] Here, meander is defined as the ratio of the length of the river channel centerline to the straight-line distance of the valley, and is a quantitative indicator of the degree of river meandering.

[0058] Then, corresponding adjustment operations are performed based on the direction of the deviation. These adjustment operations may include smoothing enhancement operations and migration enhancement operations.

[0059] When the actual curvature exceeds the target curvature, a Gaussian weighted neighborhood smoothing operation (i.e., smoothing enhancement) is performed to reduce the curvature. Gaussian weighted neighborhood smoothing reduces the local fluctuation amplitude of the river channel by using a Gaussian function to weight the node coordinates within the neighborhood.

[0060] When the actual curvature is less than the target curvature, a coordinate migration based on local curvature (i.e., migration enhancement) operation is performed to increase the curvature. Based on the magnitude and direction of the local curvature, coordinate migration is performed on the river centerline nodes to enhance the local curvature of the river.

[0061] It should be noted that the problem of insufficient curvature control accuracy is solved by introducing a statistical correction method. Experiments show that after adopting the two-stage optimization method of this invention, the curvature deviation can be reduced to below 3%. The parameter decay mechanism effectively prevents continuous oscillations during the iteration process, ensuring the convergence stability of the algorithm.

[0062] In some embodiments, step 103 may include: detecting the type of two consecutive adjustment operations; when a jump from smooth enhancement to migration enhancement is detected, attenuating the smoothness intensity parameter by a preset attenuation coefficient; when a jump from migration enhancement to smooth enhancement is detected, attenuating the migration distance parameter by a preset attenuation coefficient.

[0063] Here, the type of two consecutive adjustment operations is detected during the iteration process. When a jump from a smoothing operation to a transition operation, or from a transition operation to a smoothing operation, is detected, it is recorded as an oscillating jump.

[0064] The intensity parameters include a smoothing intensity parameter and a migration distance parameter. When a jump from a smoothing operation to a migration operation is detected, the smoothing intensity parameter is reduced by a preset attenuation factor; when a jump from a migration operation to a smoothing operation is detected, the migration distance parameter is reduced by a preset attenuation factor. As an example, the preset attenuation factor can be set to 15%.

[0065] It should be noted that when the operation direction is reversed in two consecutive iterations (i.e., an oscillation jump occurs), the system will automatically trigger bidirectional decay, reducing the corresponding intensity parameters (such as mobility or smoothness) by 15%. Through the dynamic damping mechanism, the algorithm can be guaranteed to converge in large-scale iterations without falling into an infinite loop, which has extremely high practical value.

[0066] Optionally, step 104 provides an implementation method for simulating river channel migration using a weighted integral of the upstream curvature history. This allows for control of the river centerline migration, reflecting the nonlocal properties and memory effect of the river system.

[0067] As an example, the formula for calculating the normal displacement of a river channel is as follows:

[0068]

[0069] Where n is the normal displacement of the river channel, E0 is the erosion coefficient, and G is the weighting function. 's' represents the upstream curvature history, and 's' represents the arc length coordinates along the river channel. This formula reflects that river channel migration is not only controlled by local curvature, but more importantly by the comprehensive historical control of upstream curvature.

[0070] In other words, for each node on the centerline, local geometric quantities are first extracted, including local curvature, unit normal vector, and channel width. Then, a weighted integral is performed on the curvature within the upstream window, with the integral weight decreasing exponentially with distance. The integral length is adaptively determined based on a multiple of the average channel width and can be calculated by multiplying the average channel width by the integral length factor.

[0071] In some embodiments, the migration direction and distance of each centerline node are calculated in the following ways.

[0072] For the i-th node, local geometric quantities (curvature, unit normal vector, channel width W, etc.) are first extracted. A weighted integral of the curvature is then performed within the upstream window to characterize the memory effect. The integral decay and integral direction can be expressed as:

[0073]

[0074]

[0075] in, It is the composite curvature at point i. It is the local curvature at point j. It is an exponentially decaying weight. It refers to the number of integration points. It is the composite direction vector (unit vector) of the i-th point. It is the local normal vector (unit vector) of point j. It is the absolute value of the curvature at point j.

[0076] The integral length is determined through adaptive calculation based on the average width of the river channel (as shown in the following formula):

[0077]

[0078] in, It is the average width of the river channel. It is the integral length factor.

[0079] During the migration process, we employ GPU parallel processing.

[0080] Optionally, when vegetation impact is enabled, migration distance limits can be implemented based on the vegetation impact coefficient and the original migration distance.

[0081] Therefore, it can accurately reflect the nonlocal properties and memory effect of the river system, making the simulated river migration process more consistent with actual geological laws. The weighted integral mechanism of upstream curvature takes into account the historical dependence of river migration, and the adaptive calculation of the integral length ensures the consistency of migration simulation for rivers of different widths.

[0082] In one or more embodiments of the present invention, when generating the river channel geometry in step 106, a trend factor can be introduced. Through mathematical modeling, the geometric pattern is embedded into a stochastic simulation process to achieve spatial trend control of the river channel geometric parameters. This allows for the realistic simulation of the significant downstream trend changes exhibited by natural river systems during certain geological evolution processes.

[0083] The downstream variations in river geometry parameters follow a comprehensive approach combining modified power-law and linear incremental models. Relative position parameterization is used to normalize the river channel coordinates to [0, 1], where 0 represents the upstream starting point and 1 represents the downstream ending point, providing a unified basis for trend evolution control.

[0084] Expression for downstream trend of river width:

[0085]

[0086] Expression for downstream trend of channel thickness:

[0087]

[0088] in, and It refers to the width and thickness of the river channel at location 's' along the river. and It refers to the initial width and thickness of the river channel at its starting point. and It is the strength coefficient for the increase in width and thickness. and It is the growth index, where s is the arc length coordinate along the centerline of the river channel. It is the total length of the river channel.

[0089] Expression for downstream trend of riverbed elevation:

[0090]

[0091] in, It is the elevation of the top of the riverbed at location 's' along the river channel. It is the initial elevation at the starting point of the river channel. It is the gradient of the riverbed (unit: m / km).

[0092] In some embodiments of the present invention, a method for calculating the impact of vegetation on river channel migration can be provided. The impact of vegetation on river channel migration has become an important research topic in river geomorphology. Vegetation significantly affects riverbank stability and evolutionary trajectory through mechanisms such as root reinforcement, hydrodynamic resistance, and sediment capture.

[0093] In some embodiments, please refer to Figure 5 , Figure 5 A comparative example of river migration under the influence of heterogeneous vegetation and migration without vegetation influence is shown. The method further includes: pre-calculating the vegetation influence field based on the spatial heterogeneity of vegetation, and determining the vegetation influence coefficient corresponding to different vegetation patches in the vegetation influence field; correspondingly, step 104 may include: determining the vegetation-constrained normal displacement based on the normal displacement obtained by weighted integration based on upstream curvature and the vegetation influence coefficient.

[0094] In some embodiments, the step of pre-calculating the vegetation influence field based on the spatial heterogeneity of vegetation and determining the vegetation influence coefficients corresponding to different vegetation patches in the vegetation influence field includes: generating natural fractal noise at multiple scales, wherein the noise amplitude and frequency at each scale are determined according to a preset amplitude coefficient and frequency multiplication factor; superimposing the fractal noise at each scale to obtain a basic vegetation distribution field; constructing a multidimensional feature vector containing spatial coordinate features, distance from the center features, and angle features; performing K-means clustering on the basic vegetation distribution field based on the multidimensional feature vector to divide the space into multiple vegetation patches; generating a spatial vegetation coefficient field based on the clustering results, wherein different vegetation patches correspond to different vegetation influence intensity coefficients; and mapping the spatial vegetation coefficient field to each centerline node, wherein the vegetation influence coefficient is used to constrain the normal displacement of the mapped node.

[0095] Here, the vegetation influence field is pre-calculated. Natural fractal noise is generated at multiple scales, with the noise amplitude and frequency at each scale determined according to preset amplitude coefficients and frequency multiplication factors. The fractal noise at each scale is superimposed to obtain the basic vegetation distribution field.

[0096] A multidimensional feature vector containing spatial coordinate features, distance from the center features, and angular features is constructed. Based on this multidimensional feature vector, K-means clustering is performed on the basic vegetation distribution field, dividing the space into multiple vegetation patches. A spatial vegetation coefficient field is generated based on the clustering results, with different vegetation patches corresponding to different vegetation influence intensity coefficients.

[0097] As an example, as shown in the following formula, the vegetation patch generation algorithm employs multi-scale superposition of natural fractal noise.

[0098]

[0099] Among them, Ak It is the amplitude coefficient at the k-th scale. λ is the frequency factor, and λ is the frequency multiplication factor.

[0100] As an example, K-means clustering is used for patch generation, which realizes the realistic distribution of vegetation patches by combining spatial coordinates and feature dimensions, as shown in the following equation:

[0101]

[0102] in, It is a spatial coordinate feature. It is a feature of distance from the center. It is an angular feature.

[0103] After completing the vegetation pre-calculation, the migration direction and distance of each centerline node can be calculated.

[0104] The spatial vegetation coefficient field is mapped to each centerline node. In the normal displacement calculation, the original migration distance is divided by the vegetation influence coefficient to obtain the migration distance constrained by vegetation. In areas with higher vegetation influence intensity, the river migration distance is more restricted, thus simulating the constraining effect of vegetation on river migration.

[0105] It should be noted that the heterogeneous vegetation impact mechanism enables the simulation results to reflect the real influence of vegetation on river channel migration. By multi-scale superposition of fractal noise and K-means clustering, vegetation patches with realistic spatial distribution characteristics can be generated. Different vegetation cover areas exhibit differentiated erosion resistance, thus producing a non-uniform migration rate distribution, enhancing the geological realism of the simulation.

[0106] Optionally, an initial heterogeneous vegetation distribution can be pre-generated before migration. As an example, the migration steps are set to 60, and the vegetation distribution is updated every 15 steps. The impact of vegetation constraints on channel migration is quantified using a migration damping coefficient.

[0107] In some implementations, optional implementations of step 105 are provided. Please refer to the examples provided. Figure 4 This prevents unreasonable self-intersections or narrow necks caused by migration by performing adaptive topology truncation detection between adjacent time steps.

[0108] In some embodiments, step 105 may include: employing a multi-scale progressive search strategy to gradually expand the search radius at scales from small to large, calculating the minimum inter-segment distance between different segments of the river centerline at each scale; when the minimum inter-segment distance is less than an adaptive threshold related to the local river width, performing straightening and topology reconstruction on the river centerline; wherein the adaptive threshold is the sum of the river widths at two comparison locations multiplied by a cutoff factor, and the cutoff factor has a value greater than or equal to 0.4 and less than or equal to 0.6.

[0109] Therefore, a multi-scale progressive search strategy is adopted, gradually expanding the search radius from small to large scales, and calculating the minimum inter-segment distance between different segments of the river channel centerline at each scale. This progressive search strategy can efficiently detect potential neck regions while avoiding missing small necks at large scales.

[0110] As an example, to prevent unreasonable self-intersections or narrow necks caused by migration, adaptive topology reconstruction and cutoff detection are performed after each update. Neck detection employs a multi-scale progressive search strategy, gradually expanding the search radius from small to large scales. Calculate the minimum inter-segment distance d at each scale. min The geometric cutoff criterion is based on an adaptive threshold for the river channel width:

[0111]

[0112] in, It is the cutoff factor, W i and W j These are the widths of the river channel at point i and point j, respectively. It is the minimum segment threshold.

[0113] When the minimum inter-segment distance is less than an adaptive threshold based on the local channel width, a bend truncation and straightening procedure is determined, and topology reconstruction is performed. The adaptive threshold is the sum of the channel widths at the two comparison points multiplied by a truncation factor, with the truncation factor ranging from 0.4 to 0.6. The adaptive threshold based on channel width can adapt to the neck-cutting requirements of channels with different widths.

[0114] It should be noted that the adaptive topology truncation detection mechanism effectively prevents unreasonable self-intersections or narrow neck problems during migration. The multi-scale progressive search strategy ensures the accuracy and efficiency of neck detection; the adaptive threshold based on channel width can adapt to the neck detection requirements of channels with different widths, ensuring the geological rationality of topology reconstruction.

[0115] It should be noted that the adaptive necking mechanism and the migration mechanism complement and reinforce each other. Migration increases curvature, leading to excessive curvature that triggers adaptive truncation, thereby resetting the river channel morphology. This ensures that the river channel does not become infinitely entangled during long-term simulations, maintaining geological consistency.

[0116] During GPU parallel rendering, concurrent write conflicts can arise when multiple rivers simultaneously affect the same voxel. Even with preset priority rules, time window conflicts can still lead to rendering chaos. Some embodiments of this invention propose a two-stage GPU rendering architecture of Dynamic Candidate Collection and Delayed Decision (DCCD). This architecture constructs a dynamic candidate collection mechanism to gather river attribute information from all rivers attempting to write to the target voxel, awaiting subsequent conflict resolution. This resolves the rendering priority chaos caused by time window conflicts in GPU parallel rendering.

[0117] In one or more embodiments of this application, under the information collection and delayed decision-making mechanism, candidate points of river channels and natural dikes are calculated (collected) in the GPU, and then in the delayed decision-making stage, the final color of the voxel is determined according to the preset priority rules.

[0118] In some embodiments, step 108 may include: rendering the channel geometry at each time step to generate a single-channel three-dimensional river sedimentary facies model, including: pre-calculating the distance field from voxels in three-dimensional space to the channel centerline at each time step, wherein the distance field records the shortest distance from each voxel to the nearest channel centerline position; counting the number of candidate sedimentary objects for each voxel based on the distance field, and dynamically allocating a candidate buffer according to the number of candidate sedimentary objects; collecting attribute information of candidate sedimentary objects attempting to be written to the target voxel, and storing it in the candidate buffer, wherein the attribute information includes time step identifier, channel identifier, top elevation value, and sedimentary facies type, wherein the sedimentary facies type includes channel type, natural levee type, and breach fan type; performing conflict resolution based on the time step identifier of the candidate sedimentary objects, determining the target sedimentary object of the voxel in the three-dimensional space, and rendering the target sedimentary object to the target voxel.

[0119] During the candidate collection phase, the distance field from all voxels to the river centerline at each time step is pre-calculated. The distance field records the shortest distance from each voxel to the nearest river centerline position. This matrix is ​​pre-calculated using a distance field algorithm, decoupling distance calculation from voxel rendering, eliminating redundant calculations, and significantly improving computational efficiency.

[0120] Then, based on the distance field, the number of candidate sedimentary objects for each voxel is counted. The candidate sedimentary objects include the channel location and its associated sedimentary bodies (natural dikes, breach fans) at different time steps. The candidate buffer is dynamically allocated according to the number of candidate sedimentary objects, and the memory space is precisely allocated according to the actual number of candidates using the Compacted Sparse Row (CSR) format.

[0121] Collect all sedimentary object attribute information that attempts to write to the target voxel and store it in the candidate buffer. The attribute information may include time step identifier, channel identifier, top elevation value, and sedimentary facies type. The time step identifier is used to identify the channel location at different time steps, and the channel identifier is used to establish a two-way mapping relationship between voxels and channels.

[0122] In the conflict resolution phase, in some embodiments, the conflict resolution based on the time step identifier of the candidate depositional objects, determining the target depositional object of the voxels in the three-dimensional space, and rendering the target depositional object to the target voxels, includes: when multiple channel voxels overlap spatially, comparing their channel identifier sizes and retaining the voxel with the larger channel identifier, wherein the channel centerline of each time step is assigned an increasing channel identifier as the channel migrates; for candidate depositional objects with the same channel identifier, determining the coverage relationship according to the hierarchical order of breach fan priority over channel body priority over natural levee.

[0123] A voxel-level parallelization strategy can be used, leveraging the GPU to process all candidates for all voxel locations in parallel, and selecting the optimal candidate through a global priority rule. For voxel V(x, y, z), the candidate with the largest channel identifier is selected, reflecting the coverage advantage of the subsequently formed channel.

[0124] In other words, in river migration simulations, the river centerline at each time step is assigned an increasing number of river markers, with smaller markers representing earlier channels and larger markers representing later channels. Delayed conflict resolution based on the size of the time step markers can include: when multiple channel voxels overlap spatially, comparing their channel marker sizes and retaining the voxel with the larger marker. This priority rule reflects the coverage advantage of later channels over earlier channels and aligns with geological evolution patterns.

[0125] As an example, the rendering priority algorithm ensures that the spatial stacking of sedimentary bodies conforms to geological evolution, which is crucial for accurately characterizing the distribution of porosity and other parameters. For further examples, please refer to... Figure 6 , Figure 6 A schematic diagram illustrating the effect of rendering priority technology is shown, demonstrating the effectiveness of priority rules under different modes through cross-sections.

[0126] like Figure 6Part (e) shown is the sediment body obtained by the multiphase model after 60 migrations. According to the multiphase sediment priority rendering algorithm under the aforementioned migration mode, the later generated sediment body will cover the earlier generated sediment body. After the voxel cleaning mechanism, the simulated channel sediment body model is obtained. Figure 6 Part (f) shown is a single-phase channel migration model, which illustrates the channel facies sediments and rendering priorities.

[0127] It should be noted that the dynamic candidate collection and delayed decision architecture can solve the problem of concurrent write conflicts in GPU parallel rendering. By decoupling distance calculation from voxel rendering, redundant calculations are eliminated; and by using time step-based delayed conflict resolution, the rendering results are ensured to conform to the geological patterns of river migration.

[0128] In some embodiments of the present invention, this application also provides a voxel cleaning mechanism. The voxel cleaning mechanism effectively ensures the geological rationality of the spatial stacking relationship of sedimentary bodies. For example, please refer to... Figure 7 , Figure 7 The comparison shows whether the voxel cleaning mechanism is executed or not.

[0129] Optionally, the method further includes: after voxel rendering is completed, based on the bidirectional mapping relationship between voxels and channel markers, identifying natural levee voxels and / or breach fan voxels with smaller channel markers above late-stage channel voxels. Deleting the identified natural levee voxels and / or breach fan voxels; identifying and deleting natural levee voxels with the same channel marker above breach fan voxels.

[0130] After voxel rendering is completed, the target deposition object corresponding to the voxel has a channel identifier, thus forming a two-way mapping relationship between the channel identifier and the voxel.

[0131] Based on the two-way mapping relationship between voxels and channel markers, early natural dikes and breach fan voxels with smaller channel markers within the spatial range above late-stage channel voxels were identified and removed. This cleanup operation accurately reflects the erosive and modifying effect of late-stage channels on early associated sedimentary bodies.

[0132] Based on the bidirectional mapping relationship between voxels and channel markers, all natural levee voxels with the same ancestral channel markers above all breach fan voxels are identified and deleted, thus eliminating the penetration phenomenon of homologous sedimentary bodies.

[0133] This eliminates two types of geological inconsistencies that existed during the rendering of sedimentary bodies.

[0134] The following is for reference. Figure 8This document illustrates a structural diagram of an electronic device (e.g., a terminal device or a server) suitable for implementing embodiments of the present invention. The terminal device in the embodiments of the present invention may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 8 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0135] like Figure 8 As shown, the electronic device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage device 808 into a random access memory (RAM) 803. The RAM 803 also stores various programs and data required for the operation of the electronic device 800. The processing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0136] Typically, the following devices can be connected to I / O interface 805: input devices 806 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 807 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 808 including, for example, magnetic tapes, hard disks, etc.; and communication devices 809. Communication device 809 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have instead.

[0137] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 809, or installed from a storage device 808, or installed from a ROM 802. When the computer program is executed by a processing device 801, it performs the functions defined in the methods of the embodiments of the present invention.

[0138] It should be noted that the computer-readable medium described above in this invention can be a computer-readable signal medium, a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0139] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0140] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0141] The aforementioned computer-readable medium carries one or more programs. When the aforementioned one or more programs are executed by the electronic device, the electronic device causes the following: iteratively generates nodes of an initial river centerline, starting from the river's starting point; iteratively executes correction steps based on the initial river centerline, the correction steps including: in each correction round, calculating the deviation between the actual curvature of the current river centerline and a preset target curvature, and performing smoothing enhancement or migration enhancement adjustments according to the direction of the deviation; detecting the adjustment operation types of two adjacent correction rounds, and when different adjustment operation types are detected between two adjacent correction rounds, performing adjustment parameter attenuation; based on the iteratively corrected initial river centerline, performing a preset number of migration steps to obtain a river centerline with a number of time steps, wherein the migration steps include: in each time step, calculating the normal displacement of each node of the centerline based on the weighted integral of the upstream curvature, and updating the centerline node coordinates based on the normal displacement; between adjacent time steps, determining whether to perform a river straightening operation based on an adaptive threshold related to the local river width. Based on the channel centerline of the specified number of time steps, channel geometry of the specified number of time steps is generated; the channel geometry of each time step is rendered to generate a single-channel three-dimensional fluvial sedimentary facies model.

[0142] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0143] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0144] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in sequential order. In certain circumstances, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the invention. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0145] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for generating a three-dimensional fluvial sedimentary facies model of a single channel, characterized in that, include: Starting from the river's starting point, the nodes of the initial river centerline are generated. The correction steps are performed iteratively based on the initial channel centerline. The correction steps include: in each round of correction, calculating the deviation between the actual curvature of the current channel centerline and the preset target curvature, and performing smoothing enhancement or migration enhancement adjustment operations according to the direction of the deviation. The adjustment operation type of two adjacent calibration steps is detected. When the adjustment operation types of two adjacent calibration steps are different, the adjustment parameter is decayed. Based on the initial channel centerline after iterative correction, a preset number of migration steps are performed to obtain a channel centerline with a number of time steps. The migration steps include: in each time step, calculating the normal displacement of each node of the centerline based on the weighted integral of the upstream curvature, and updating the coordinates of the centerline nodes according to the normal displacement. Between adjacent time steps, based on an adaptive threshold related to the local river width, it is determined whether to perform the river straightening operation. Based on the channel centerline of the specified number of time steps, channel geometry of the specified number of time steps is generated; the channel geometry of each time step is rendered to generate a single-channel three-dimensional fluvial sedimentary facies model.

2. The method according to claim 1, characterized in that, The detection of adjustment operation types in two adjacent calibration steps, when different adjustment operation types are detected in two adjacent calibration steps, involves performing adjustment parameter attenuation, including: Detect the type of two consecutive adjustment operations; When a jump from smooth enhancement to migration enhancement is detected, the smoothing intensity parameter is attenuated by a preset attenuation factor; When a jump from migration enhancement to smooth enhancement is detected, the migration distance parameter is attenuated by a preset attenuation factor.

3. The method according to claim 2, characterized in that, The detection of adjustment operation types in two adjacent calibration steps, when different adjustment operation types are detected in two adjacent calibration steps, involves performing adjustment parameter attenuation, including: Calculate the numerical deviation between the actual curvature of the current river channel centerline and the preset target curvature; When the actual curvature is greater than the target curvature, Gaussian weighted neighborhood smoothing enhancement is performed; When the actual curvature is less than the target curvature, coordinate migration enhancement based on local curvature is performed.

4. The method according to claim 1, characterized in that, The method further includes: Based on the spatial heterogeneity of vegetation, the vegetation influence field is pre-calculated, and the vegetation influence coefficient corresponding to different vegetation patches in the vegetation influence field is determined; and The calculation of the normal displacement of each node on the centerline based on the weighted integral of the upstream curvature at each time step, and the updating of the centerline node coordinates based on the normal displacement, includes: The vegetation-constrained normal displacement is determined based on the normal displacement obtained by weighted integral based on upstream curvature and the vegetation influence coefficient.

5. The method according to claim 4, characterized in that, The process of pre-calculating the vegetation influence field based on the spatial heterogeneity of vegetation, and determining the vegetation influence coefficient corresponding to different vegetation patches within the vegetation influence field, includes: Natural fractal noise is generated for multiple scales, wherein the noise amplitude and frequency at each scale are determined according to a preset amplitude coefficient and frequency multiplication factor. The basic vegetation distribution field is obtained by superimposing the fractal noise at various scales. Construct a multidimensional feature vector that includes spatial coordinate features, distance from the center features, and angular features; Based on the multidimensional feature vectors, K-means clustering is performed on the basic vegetation distribution field to divide the space into multiple vegetation patches. A spatial vegetation coefficient field is generated based on the clustering results, where different vegetation patches correspond to different vegetation influence intensity coefficients; The spatial vegetation coefficient field is mapped to each centerline node, wherein the vegetation influence coefficient is used to constrain the normal displacement of the mapped node.

6. The method according to claim 1, characterized in that, The step of determining whether to perform a river straightening operation based on an adaptive threshold related to the local river width between adjacent time steps includes: A multi-scale progressive search strategy is adopted to gradually expand the search radius from small to large scales, and the minimum inter-segment distance between different segments of the river centerline is calculated at each scale. When the minimum inter-segment distance is less than an adaptive threshold related to the local channel width, straightening of bends is performed, and topology reconstruction is performed on the channel centerline. The adaptive threshold is the sum of the river widths at the two comparison positions multiplied by a cutoff factor, wherein the cutoff factor has a value range of greater than or equal to 0.4 and less than or equal to 0.

6.

7. The method according to any one of claims 1 to 6, characterized in that, The process of rendering the channel geometry at each time step to generate a single-channel three-dimensional fluvial sedimentary facies model includes: The distance field from voxels in three-dimensional space to the river centerline at each time step is pre-calculated, wherein the distance field records the shortest distance from each voxel to the nearest river centerline position; The number of candidate deposition objects for each voxel is counted based on the distance field, and a candidate buffer is dynamically allocated according to the number of candidate deposition objects. Collect attribute information of candidate sedimentary objects that attempt to be written to the target voxel and store it in the candidate buffer. The attribute information includes time step identifier, channel identifier, top elevation value and sedimentary facies type, wherein the sedimentary facies type includes channel type, natural levee type and breach fan type. Conflict resolution is performed based on the time step identifier of the candidate deposition object, the target deposition object of the voxel in the three-dimensional space is determined, and the target deposition object is rendered to the target voxel.

8. The method according to claim 7, characterized in that, The process of performing conflict resolution based on the time step identifier of candidate deposition objects, determining the target deposition object of the voxel in the three-dimensional space, and rendering the target deposition object to the target voxel includes: When multiple river voxels overlap in space, their river label sizes are compared, and the voxel with the larger river label is retained. In this case, the river centerline is assigned an increasing river label at each time step as the river migrates. For candidate sedimentary objects with the same channel identifier, the cover relationship is determined according to the hierarchical order of breach fans taking precedence over the channel body, and the channel body taking precedence over natural dikes.

9. The method according to claim 8, characterized in that, The method further includes: After voxel rendering is completed, based on the two-way mapping relationship between voxels and channel markers, identify natural levee voxels and / or breach fan voxels with smaller channel markers above late channel voxels; delete the identified natural levee voxels and / or breach fan voxels; identify and delete natural levee voxels with the same channel markers above breach fan voxels.

10. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-9.

Images