Fast Marching Research Articles

Global path planning for inland vessels based on fast marching algorithm

Ship path planning is essential for enhancing navigation safety, shipping efficiency, cost savings and autonomy. This paper presents a path planning algorithm specifically designed for inland vessels, exemplified on the Western Scheldt waterway area from Hansweert to Bath in the Netherlands (known as the area of Schaar van Valkenisse). The algorithm utilizes the fast marching (FM) method to generate safe and cost-efficient paths for both autonomous and conventional navigation, considering factors such as water depth, distance to waterway borders, preferential navigation zones (e.g. starboard side and parallel routes), ship dimensions, and manoeuvring characteristics.The fast marching algorithm uses starting points and a velocity grid, generated based on the aforementioned factors, to produce an arrival time map. The velocity here represents preferences for deeper water, and wider and starboard navigation zones. Subsequently, a gradient descent algorithm is applied to determine the fastest route from the destination to the starting point on the arrival time map. To ensure the route is navigable, the optimal path is updated to a Dubins path that satisfies the ships’ dynamic constraints by using a so-called Dubins model filter.Case studies with various navigation scenarios demonstrate that the path planning tool can provide expected results with tunable parameters allowing for customized preferences. This work can further assist in predicting paths for a traffic study, designing fairways, aids to navigation and can possibly serve as an operational tool to assess the accessibility of a (alternative) fairway.

Impact of discontinuity data acquisition methods on rockfall susceptibility assessment using high-resolution 3D point cloud

This research evaluates the influence of various discontinuity data acquisition methods on three-dimensional rockfall susceptibility assessments, employing a high-resolution 3D point cloud of the Špičunak rock slope in Croatia. Discontinuity mapping was conducted utilising manual and semi-automated methods. Manual mapping was done in open-source software CloudCompare, while semi-automated mapping was done using Discontinuity Set Extractor and two CloudCompare plug-ins, qFacet Fast Marching and qFacet Kd-Tree algorithm. Significant differences were observed between the methods in terms of the number of identified discontinuity planes, sets, mean orientation values, and weighted densities associated with specific sets. Based on the mapping results, three-dimensional rockfall susceptibility maps were generated. The 3D rockfall susceptibility assessment employed a Kinematic Hazard Index (KHI) for planar, wedge, and flexural toppling failure mechanisms, which determines the percentage of discontinuity planes that meet the geometrical conditions specific to each type of failure. The findings underscore that the mapping results greatly affect the 3D rockfall susceptibility assessment, impacting the identification of potential rockfall source areas. The study emphasizes the importance of integrating manual and semi-automated mapping for reliable 3D rockfall susceptibility assessment. Additionally, this study evaluated the effect of varying the number of discontinuity orientation input data on rockfall susceptibility by comparing nine scenarios, each involving a reduction in the number of discontinuity orientation data. The results indicate that accurate discontinuity mapping is more critical than the sheer number of mapped discontinuities for reliable rockfall susceptibility assessment. Therefore, prioritizing the precise determination of discontinuity set numbers and their associated weighted densities is essential for reliable susceptibility evaluations.

Improved Transcranial Plane-Wave Imaging With Learned Speed-of-Sound Maps.

Although transcranial ultrasound plane-wave imaging (PWI) has promising clinical application prospects, studies have shown that variable speed-of-sound (SoS) would seriously damage the quality of ultrasound images. The mismatch between the conventional constant velocity assumption and the actual SoS distribution leads to the general blurring of ultrasound images. The optimization scheme for reconstructing transcranial ultrasound image is often solved using iterative methods like full-waveform inversion. These iterative methods are computationally expensive and based on prior magnetic resonance imaging (MRI) or computed tomography (CT) information. In contrast, the multi-stencils fast marching (MSFM) method can produce accurate time travel maps for the skull with heterogeneous acoustic speed. In this study, we first propose a convolutional neural network (CNN) to predict SoS maps of the skull from PWI channel data. Then, use these maps to correct the travel time to reduce transcranial aberration. To validate the performance of the proposed method, numerical, phantom and intact human skull studies were conducted using a linear array transducer (L11-5v, 128 elements, pitch = 0.3 mm). Numerical simulations demonstrate that for point targets, the lateral resolution of MSFM-restored images increased by 65%, and the center position shift decreased by 89%. For the cyst targets, the eccentricity of the fitting ellipse decreased by 75%, and the center position shift decreased by 58%. In the phantom study, the lateral resolution of MSFM-restored images was increased by 49%, and the position shift was reduced by 1.72 mm. This pipeline, termed AutoSoS, thus shows the potential to correct distortions in real-time transcranial ultrasound imaging, as demonstrated by experiments on the intact human skull.

Effect of Using Fractals as Grain Configuration on the Performance of Solid Rocket Motors

Grain design is a very crucial consideration for determining the performance of solid propellant rocket motors. A rocket’s performance and combustion characteristics (thrust profile, chamber pressure variation, propellant mass flow rate, and sliver among other important performance parameters) depend very strongly on the grain configuration. We propose fractal shapes (Koch star, Koch snowflake, Gosper, and Cesaro) as new port geometries and theoretically predict their performance using simulations based on the fast-marching method implemented in the open-source software Open-Motor. The initial thrust for fractal geometry grains was found to be significantly higher than the commonly used industry standard BATES and Finosyl ports. The fractal grains also showed a significant decrease in sliver as compared to the Finocyl port that is generally used to increase performance as compared to the circular BATES ports. Especially, The Koch snowflake port had a drastically lower sliver. These new grain geometries proposed in this work can be used for applications in which low initiation and ignition times along with high initial thrust are crucial. Moreover, these geometries can improve the performance of the solid rocket motor (SRM) owing to their better regression characteristics which promise complete utilization of the entire grain web.

Improved FMM for well locations optimization in in-situ leaching areas of sandstone uranium mines

Rapidly obtaining the coverage characteristics of leaching solution in In-situ Leaching Area of Sandstone Uranium Mines is a necessary condition for optimizing well locations reasonably. In the presented study, the improved algorithm of the Fast Marching Method (FMM) was studied for rapidly solving coverage characteristics to replace the groundwater numerical simulator. First, the effectiveness of the FMM was verified by simulating diffusion characteristics of the leaching solution in In-situ Leaching Area. Second, based on the radial flow pressure equation and the interaction mechanism of the front diffusion of production and injection well flow field, an improved FMM which is suitable for In-situ Leaching Mining, was developed to achieve the co-simulation of production and injection well. Finally, the improved algorithm was applied to engineering practice to guide the design and production. The results show that the improved algorithm can efficiently solve the coverage characteristics of leaching solution, which is consistent with those obtained from traditional numerical simulators. In engineering practice, the improved FMM can be used to rapidly analyze the leaching process, delineate Leaching Blind Spots, and evaluate the rationality of well pattern layout. Furthermore, it can help to achieve iterative optimization and rapid decision-making of production and injection well locations under large-scale mining area models.

Trajectory Planning and Singularity Avoidance Algorithm for Robotic Arm Obstacle Avoidance Based on an Improved Fast Marching Tree

The quest for efficient and safe trajectory planning in robotic manipulation poses significant challenges, particularly in complex obstacle environments where the risk of encountering singularities and obstacles is high. Addressing this critical issue, our study presents a novel enhancement of the Fast Marching Tree (FMT) algorithm, ingeniously designed to navigate the complex terrain of Cartesian space with an unprecedented level of finesse. At the heart of our approach lies a sophisticated two-stage path point sampling strategy, ingeniously coupled with a singularity avoidance mechanism that leverages geometric perception to assess and mitigate the risk of encountering problematic configurations. This innovative method not only facilitates seamless obstacle navigation but also adeptly circumvents the perilous zones of singularity, ensuring a smooth and uninterrupted path for the robotic arm. To further refine the trajectory, we incorporate a quasi-uniform cubic B-spline curve, optimizing the path for both efficiency and smoothness. Our comprehensive simulation experiments underscore the superiority of our algorithm, showcasing its ability to consistently achieve shorter, more efficient paths while steadfastly avoiding obstacles and singularities. The practical applicability of our method is further corroborated through successful implementation in real-world robotic arm trajectory planning scenarios, highlighting its potential to revolutionize the field with its robustness and adaptability.

Learning and generalization of task-parameterized skills through few human demonstrations

In the field of robotics, the demand for adaptable skills capable of effectively handling diverse situations has surpassed the reliance on repetitive tasks. To enhance the generalization of motion policies, task-parameterized learning has emerged as a valuable approach, as it encodes pertinent contextual information in task parameters, facilitating the flexible execution of tasks. This process requires the collection of multiple demonstrations in various situations. Generate a set of diverse situations covering all possible cases is a complex task. Due to this, the search for training with fewer demonstrations becomes highly desirable. In this article, we present a novel algorithm focused on generating synthetic information to facilitate the generalization of parameterized tasks. This algorithm makes use of Kinesthetic Fast Marching Learning, a Learning from Demonstration (LfD) algorithm that allows obtaining optimal movement paths based on velocity fields. The algorithm enables autonomous data generation, producing demonstrations that are equal to or more optimal than those generated by users themselves. Evaluation is done through a metric based on Wasserstein distance that takes into account the probabilistic data of the generated paths. The algorithm has been evaluated through tests in simulated environments, comparing its efficiency against two widely used algorithms (where it has shown greater efficiency in generalization and the generation of more optimal paths) and with real-world tests in a task-oriented environment (sweeping) carried out by the ADAM robot.

Study of 3-D velocity structure characteristics in Dadu river Houziyan reservoir area at different impoundment stages

High-resolution three-dimensional VP, VS and VP/VS images in the Houziyan Reservoir Area were obtained by using Fast Marching Tomography Package (FMTOMO) with the travel time data from 6330 seismic events monitored by the Houziyan Reservoir Seismic Network. This analysis yielded the 3-D velocity structure, including longitudinal wave velocity (VP), shear wave velocity (VS), and the ratio of longitudinal and shear wave velocity (VP/VS) at different impoundment stages. The data changes at various impoundment times, depths of sections, and directions of profiles were analyzed to obtain these results. The final findings demonstrate the following results: 1) Through tomographic analysis, it was determined that the underground velocity structure in the Houziyan reservoir area was anisotropic before impoundment. 2) The area of high wave velocity increases in stage 1, stage 3, and stage 4. The area of low wave velocity increases in stage 2, especially in depth, indicating significant changes in the underground velocity structure at different impoundment stages. 3) Compared to the changes in underground velocity structures in other reservoirs after impoundment, the Houziyan reservoir exhibited a unique pattern. 4) In general, the underground velocity structure displayed an overall increasing trend after impoundment. However, it also exhibited instances of decreasing velocity, reflecting continuous dynamic adjustments to the underground velocity structure after impoundment. These conclusions highlight the impact of impoundment in the reservoir area on the underground velocity structure and provide scientific theoretical support for seismic risk assessment following impoundment in the reservoir area.

Topology optimization for additive manufacturing of CFRP structures

Combining design optimization and additive manufacturing (AM) enables the full exploitation of the potential for heightening the performance of carbon fiber reinforced plastic (CFRP) structures. This study simultaneously conducts topology optimization and fiber path design by employing the radial basis function (RBF) based level set function (LSF). Fiber paths are determined instinctively for the inherent advantages of the LSF, and fiber orientations are parameterized accordingly. Manufacturing drawbacks such as gaps and overlaps can be avoided by introducing a fast-marching method. To verify the effectiveness of the optimization method, three groups of optimized and empirical designs are fabricated by the AM technique, respectively, and the experimental tests are further carried out. Finite element (FE) models are also reconstructed based on the printing schemes, and then the FE simulation is validated by the experimental tests. With the proposed optimization method, stiffnesses for all three groups of the optimal samples are significantly improved compared with the empirical counterparts. The FE modeling technique is capable of reproducing the experimental results. This study paves a new way to develop an integrated framework of optimization, additive manufacturing, experimentation, and validation to deliver high-performance CFRP structures.

Rayleigh wave group velocities in North-West Iran: SOLA Backus-Gilbert vs. Fast Marching tomographic methods

In this study, we focus on Northwest Iran and exploit a dataset of Rayleigh-wave group-velocity measurements obtained from ambient noise cross-correlations and earthquakes.We build group-velocity maps using the recently developed SOLA Backus-Gilbert linear tomographic scheme as well as the more traditional Fast-marching Surface-wave Tomography method.The SOLA approach produces robust, unbiased local averages of group velocities with detailed information on their local resolution and uncertainty; however, it does not as yet allow ray-path updates in the inversion process. The Fast-marching method, on the other hand, does allow ray-path updates, although it does not provide information on the resolution and uncertainties of the resulting models (at least not without great computational cost) and may suffer from bias due to model regularisation.The core of this work consists in comparing these two tomographic methods, in particular how they perform in the case of strong vs. weak seismic-velocity contrasts and good vs. poor data coverage. We demonstrate that the only case in which the Fast-marching inversion outperforms the SOLA inversion is for strong anomaly contrasts in regions with good path coverage; in all other configurations, the SOLA inversion produces more coherent anomalies with fewer artefacts.

Continuous fiber reinforced meta-composites with tailorable Poisson’s ratio and effective elastic modulus: Design and experiment

Auxetic metamaterials, i.e., with negative Poisson’s ratio are attracting increasing attention. This study develops a novel method to design and investigate continuous fiber-reinforced meta-composites with tailorable negative Poisson’s ratio and effective elastic modulus. First, the design scheme based on parametric level set method, is proposed by combining fast marching method, filtering technology and matrix rotation for attaining the two-dimensional isotropic composites. A novel objective function considering both negative Poisson’s ratio and effective elastic modulus is introduced and the result demonstrate that it is effective to not only accelerate the numerical convergence but also achieve a larger negative Poisson’s ratio when compared with that using the currently-used objective function. By tailoring a weight ratio for balancing the negative Poisson’s ratio and effective elastic modulus, a general contradictive trend was captured. Experimentally, the specimens are accordingly fabricated via 3D printing. Then, the axial compression tests are performed with the aid of digital image correlation technology to characterize the deformation behaviors and evaluate the mechanical properties of the designed meta-composites. Experimental results show that, as the weight ratio increases from 0.02 to 1.57, the effective elastic modulus increases from 180.5 MPa to 310.3 MPa (enhanced by 71.9 %) while the negative Poisson’s ratio decreases from − 0.41 to − 0.036 (reduced by 91.4 %), proving the effectiveness of the design method. The design methods and experimental results can provide effective approaches and important information for designing novel continuous fiber reinforced meta-composites with tailorable mechanical performance.

Improving understanding of groundwater flow in an alpine karst system by reconstructing its geologic history using conduit network model ensembles

Abstract. Reconstructing the geologic history of a karst area can advance understanding of the system's present-day hydrogeologic functioning and help predict the location of unexplored conduits. This study tests competing hypotheses describing past conditions controlling cave formation in an alpine karst catchment, by comparing an ensemble of modeled networks to the observed network map. The catchment, the Gottesacker karst system (Germany and Austria), is drained by three major springs and a paleo-spring and includes the partially explored Hölloch cave, which consists of an active section whose formation is well-understood and an inactive section whose formation is the subject of debate. Two hypotheses for the formation of the inactive section are the following: (1) glaciation obscured the three present-day springs, leaving only the paleo-spring, or (2) the lowest of the three major springs (Sägebach) is comparatively young, so its subcatchment previously drained to the paleo-spring. These hypotheses were tested using the pyKasso Python library (built on anisotropic fast-marching methods) to generate two ensembles of networks, one representing each scenario. Each ensemble was then compared to the known cave map. The simulated networks generated under hypothesis 2 match the observed cave map more closely than those generated under hypothesis 1. This supports the conclusion that the Sägebach spring is young, and it suggests that the cave likely continues southwards. Finally, this study extends the applicability of model ensemble methods from situations where the geologic setting is known but the network is unknown to situations where the network is known but the geologic evolution is not.

STExplorer: A Hierarchical Autonomous Exploration Strategy with Spatio-temporal Awareness for Aerial Robots

The autonomous exploration task we consider requires Unmanned Aerial Vehicles (UAVs) to actively navigate through unknown environments with the goal of fully perceiving and mapping the environments. Some existing exploration strategies suffer from rough cost budgets, ambiguous Information Gain (IG), and unnecessary backtracking exploration caused by Fragmented Regions (FRs). In our work, a hierarchical spatio-temporal-aware exploration framework is proposed to alleviate these problems. At the local exploration level, the Asymmetrical Traveling Salesman Problem (ATSP) is solved by comprehensively considering exploration time, IG, and heading consistency to avoid blindly exploring. Specifically, the exploration time is reasonably budgeted by fast marching in an artificial potential field. Meanwhile, a transformer-based map occupancy predictor is designed to assist in IG calculation by imagining spatial clues out of the Field of View (FoV), facilitating the prescient exploration. We verify that our local exploration is effective in alleviating the unnecessary back-and-forth movements caused by FRs and the interference of potential obstacle occlusion on the IG calculation. At the global exploration level, the classical Next Best View Points (NBVP) are generalized to Next Best Sub-Regions (NBSR) to choose informative sub-regions for further forward-looking exploration based on a well-designed utility function. Safe flight paths and dynamically feasible trajectories are reasonably generated throughout the exploration process by fast marching and B-spline curve optimization. Comparative simulations and benchmark tests demonstrate that our proposed exploration strategy is quite competitive in terms of exploration path length, total exploration time, and exploration ratio.

Ambient noise tomography of the Aegean region of Türkiye from Rayleigh wave group velocity

We have studied Rayleigh wave group velocities beneath the Aegean region of Türkiye using ambient noise tomography. Noise data were gathered from 43 broadband seismic stations belonging to three permanent broad-band arrays. The cross-correlation method was used to estimate empirical Green’s functions. Group velocities of the fundamental mode Rayleigh waves were determined using multiple filter technique. We measured dispersion for each station pair in a period range of 2–12 s and computed maps of group velocity distribution using the fast marching surface tomography method. The group velocity maps correlate well with the geological and tectonic features of the region, displaying low velocities for the Quaternary alluvial basins, moderate velocities for Cretaceous rocks, and high velocities for the regions where gneiss and granitoid rocks outcrop. Low velocity anomalies may be associated with the grabens and horsts formed by faulting and to deep mantle flow. We detect potential offshore geothermal zones in Izmir and Candarli Bay. The information from broad-band networks improves the resolution of crustal surface wave velocity structures, particularly for shallow depths. This improvement will support the assessment of seismic hazard, as the shallow layers of the crust contribute significantly to seismic activity.

Automatic pulmonary artery-vein separation in CT images using a twin-pipe network and topology reconstruction.

With the wide application of CT scanning, the separation of pulmonary arteries and veins (A/V) based on CT images plays an important role for assisting surgeons in preoperative planning of lung cancer surgery. However, distinguishing between arteries and veins in chest CT images remains challenging due to the complex structure and the presence of their similarities. We proposed a novel method for automatically separating pulmonary arteries and veins based on vessel topology information and a twin-pipe deep learning network. First, vessel tree topology is constructed by combining scale-space particles and multi-stencils fast marching (MSFM) methods to ensure the continuity and authenticity of the topology. Second, a twin-pipe network is designed to learn the multiscale differences between arteries and veins and the characteristics of the small arteries that closely accompany bronchi. Finally, we designed a topology optimizer that considers interbranch and intrabranch topological relationships to optimize the results of arteries and veins classification. The proposed approach is validated on the public dataset CARVE14 and our private dataset. Compared with ground truth, the proposed method achieves an average accuracy of 90.1% on the CARVE14 dataset, and 96.2% on our local dataset. The method can effectively separate pulmonary arteries and veins and has good generalization for chest CT images from different devices, as well as enhanced and noncontrast CT image sequences from the same device.

Rapid Inference of Reservoir Permeability from Inversion of Traveltime Data Under a Fast Marching Method-Based Deep Learning Framework

Summary The fast marching method (FMM) is a highly efficient numerical algorithm used to solve the Eikonal equation. It calculates traveltime from the source point to different spatial locations and provides a geometric description of the advancing front in anisotropic and heterogeneous media. As the Eikonal solution, the diffusive time of flight (DTOF) can be used to formulate an asymptotic approximation to the pressure diffusivity equation to describe transient flow behavior in subsurface porous media. For the infinite-acting flow that occurs in porous media with smoothly varying heterogeneity, traveltime of the pressure front from the active production or injection well to the observation well can be directly estimated from the DTOF using the concept of radius (or depth) of investigation (ROI or DOI), which is defined as the moment when a maximum magnitude of the partial derivative of pressure to time occurs. Based on the ROI or DOI definition, we propose a deep neural network called the inversion neural network (INN) to inversely estimate heterogeneous reservoir permeability by inverting the traveltime data. The INN is trained by traveltime data created for a large data set of distinct permeability fields from FMM simulations, which can be two orders of magnitude faster than conventional reservoir simulators. A convolutional neural network (CNN), the U-Net architecture, is incorporated into the INN, which establishes a nonlinear mapping between the heterogeneous permeability fields and the traveltime data collected at sparse observation wells. The loss function used for the INN is defined as the root mean square error (RMSE) between the logarithm of the predicted permeability and the logarithm of the true permeability. The performance of the INN is tested on reservoir models with both smoothly varying heterogeneity and high-contrast media properties. For the 2D smoothly varying heterogeneous models with a grid size of 49×49, the permeability predicted by the INN has an average estimation error of 8.73% when a set of 7×7 uniformly distributed observation wells is used to collect “observational” traveltime data from the FMM simulation. For models with the same grid size and observation well density but with high-contrast media properties, the INN can still capture the general heterogeneity distribution, although with reduced prediction accuracy. Using a graphics processing unit (GPU) for training and prediction allows the entire inverse modeling process for a 2D 49×49 reservoir model to be completed within 7 minutes.

Application of fast marching method and quality map to well trajectory optimization with a novel well parametrization

This paper proposes a novel parametrization technique for deviated and horizontal wells and investigates the potential of combining the Fast Marching Method (FMM) with Particle Swarm Optimization (PSO) to facilitate optimization of the trajectory of vertical, deviated, and horizontal wells. The parametrization technique is based on practical drilling parameters, which can be controlled easily. Results showed that the PSO-FMM could find similar solutions to the direct optimization (PSO-DIRECT) on simple, synthetic cases with few number wells, but not on complex cases with a large number of wells. The FMM was hybridized with a reservoir simulator (PSO-HYBRID) that resolved the shortcomings of the PSO-FMM by filtering the solutions of the PSO-FMM based on the net present value (NPV). Also, initializing the optimizer by a quality map (QM) and the PSO-FMM method significantly improved the solutions. Initializing the PSO-DIRECT method by the PSO-FMM increased the NPV by 12% compared to the PSO-FMM method, and an average growth of 8.4% compared to the PSO-HYBRID. Also, the PSO-DIRECT + QM could find better results than the direct optimization, with an average increase of 1.4% in the NPV; however, its computational cost was two times the computational cost of the PSO-DIRECT + FMM and even higher than the computational cost of the direct optimization. The novelty of this paper is proposing a new well trajectory parametrization technique and comprehensively investigating the application of the FMM in well location and trajectory optimization on synthetic and real-field cases. It was shown that it is better to use the FMM as a helper function for the reservoir simulator rather than using it to completely substitute the reservoir simulator.

Computing geodesic paths encoding a curvature prior for curvilinear structure tracking

In this paper, we introduce an efficient method for computing curves minimizing a variant of the Euler-Mumford elastica energy, with fixed endpoints and tangents at these endpoints, where the bending energy is enhanced with a user-defined and data-driven scalar-valued term referred to as the curvature prior. In order to guarantee that the globally optimal curve is extracted, the proposed method involves the numerical computation of the viscosity solution to a specific static Hamilton-Jacobi-Bellman (HJB) partial differential equation (PDE). For that purpose, we derive the explicit Hamiltonian associated with this variant model equipped with a curvature prior, discretize the resulting HJB PDE using an adaptive finite difference scheme, and solve it in a single pass using a generalized fast-marching method. In addition, we also present a practical method for estimating the curvature prior values from image data, designed for the task of accurately tracking curvilinear structure centerlines. Numerical experiments on synthetic and real-image data illustrate the advantages of the considered variant of the elastica model with a prior curvature enhancement in complex scenarios where challenging geometric structures appear.

Comparative study of advanced image reconstruction algorithms for complex arbitrary components

Accurate ultrasonic imaging and inspection of specimens with non-planar complex surfaces are often challenging from both transducer coupling and imaging perspectives. Several inspection strategies have been developed to inspect such complex interfaced components such as flat arrays with flexible wedges or water-filled chambers with a flexible membrane. In these methods, firstly surface profiles are estimated and then Fermat’s principle or other algorithms are applied to calculate delay laws. The main drawback of these methods is the longer computation time and therefore real time inspection is limited, also a coupling problem due to the complex geometries. Alternative approaches have been developed to compute the ultrasonic beam paths such as the Fast-Marching Method (FMM), Dijkstra’s algorithms, numerical analysis, bisection and root-finding algorithms for the inspection of complex surface components. These algorithms can be applied in two and three-dimensional domains resulting in high computation time. This work presents an experimental comparison of an iterative beamforming method for arbitrary shaped surface profile imaging against both Full Matrix Capture-Total Focusing Method (FMC-TFM) and CIVA simulation on an aluminium calibration block with Side drilled holes (SDHs) at various depths in the range of 15-80 mm. The performance of the iterative beamforming method is evaluated in terms of sensitivity, Signal to Noise Ratio (SNR) and Array Performance Indicators (API). The method provides a 2 dB improved SNR when compared to FMC-TFM.

Fast marching method assisted permeability upscaling using a hybrid deep learning method coupled with particle swarm optimization

Geo-cellar models contain millions of gridblocks and are very time-consuming to simulate. This challenge necessitates upscaling geo-cellular models to obtain fit-for-purpose models for simulation. Simulation-based upscaling methods are computationally demanding, especially if there is a large number of geological realizations. On the other hand, static methods do not take into account the dynamics of the flow. In this paper, a novel hybrid deep learning-based upscaling method is proposed that can handle highly channelized and layered reservoir models. The proposed method combines the ConvLSTM network with the fast marching method (FMM), which is a computationally efficient method to get the dynamic response of the reservoir without the need for full-physics flow simulation. Since reservoir models have deposited during different geological periods, they have a temporal distribution along the third dimension. ConvLSTM layers can simultaneously extract spatiotemporal dependencies that other networks in the literature were not able to do so. The proposed method follows a two-step training process. In the first step, a variational autoencoder with ConvLSTM blocks (ConvLSTM-VAE) is trained and validated unsupervised using 10,000 permeability realizations. This network learns the data-generating distribution of the realizations and improves the overall upscaling process. Then, the acquired weights are used to initialize the ConvLSTM-FMM network, which is fine-tuned by the particle swarm optimization to accomplish the upscaling task by minimizing the difference between the pressure drop of the fine-scale and upscaled realizations calculated by the FMM. Validity of the FMM to estimate the BHP drop of the wells was verified on the utilized reservoir model. The advantage of the proposed data-driven method is that it can be trained once and used to upscale new realizations, contrary to conventional simulation-based methods that require running the entire process whenever the fine realizations change. The proposed network was applied to upscaling permeability realizations since absolute permeability is one of the most challenging properties to upscale because of its directional nature. Results showed the outstanding performance of the proposed method with a mean absolute percentage error, mean squared error, and coefficient of determination of 15.38, 0.18, and 0.81 on the test realizations, and outperformed the geometric mean and the simulation-based upscaling methods.