Velocity Model Parameters Research Articles

A Novel Asymmetric Car Following Model for Driver-Assist Enabled Vehicle Dynamics

Adaptive cruise control (ACC) vehicles are proving to be the first generation of driver-assist enabled vehicles. In order to study the impacts of ACC vehicles on string stability and traffic flow characteristics, accurately calibrating microscopic car following models is crucial to simulate inter-vehicle dynamics. While many car following models have been used to simulate car following behavior, a single, continuous function may not describe both acceleration and braking realistically. We propose an asymmetric model which is based on the symmetric optimal velocity relative velocity (OVRV) model and switch parameters under different conditions to realize and reproduce car following dynamics of ACC vehicles. We conduct an analytical string stability analysis and the string stability criterion is derived. The calibration and simulation results show that the proposed asymmetric ACC model reduces model spacing error by up to 38% compared with the symmetric OVRV model. Compared with other commonly used asymmetric car following models in the transportation community, the proposed asymmetric ACC model can reduce spacing error by 44.8%. Furthermore, we validate the derived string stability criterion with a numerical test simulating with a string of vehicles. We conclude that an asymmetric car following model shows more accurate performance in the capture of ACC car following behavior.

Physics-guided deep autoencoder to overcome the need for a starting model in full-waveform inversion

Full-waveform inversion (FWI) is a popular technique to obtain high-resolution estimates of earth model parameters using all information present in seismic data. Thus, it can provide important information about the subsurface. The FWI algorithm is formulated as a data-fitting minimization problem that iteratively updates an initial velocity model using the gradient of the misfit until an acceptable match is obtained between the real and synthetic data under a tolerance level based on noise in the data. The inversion is computationally expensive and can converge to a local minimum if the starting model used is not close enough to an optimal model. Here, we propose an alternative approach using a combination of machine learning and the physics of the forward model. Unlike conventional supervised machine learning, known answers are not required to train our network. The shot gathers are input to a convolutional neural network-based autoencoder, the output of which is used as the velocity model that is used to compute synthetic seismograms. The synthetic data are compared against observed input data, and the misfit is estimated. The gradient of the misfit with respect to the velocity model parameters is calculated using the adjoint state method. The adjoint state gradient is then used to update the network weights using the automatic differentiation technique. Once the misfit term converges, the neural network can generate velocity models consistent with the observed data. We observe that the neural network can capture spatial correlations at different scales and thus can introduce regularization in our inverse problem. Experiments with the Marmousi model and SEG Advanced Modeling Corporation Phase 1 salt model suggest that the proposed method can overcome local minima, requires no starting model, and produces robust results in the presence of noise and complex salt body structures.

Uncertainty quantification of phase velocity surface waves multy-modal inversion using machine learning

The paper is devoted to uncertainty quantification of the inverse problem solution of the multichannel analysis of surface waves method - the inversion of the curves of the phase velocity via frequency dependence. The uncertainty estimation approach is based on the Monte Carlo sampling strategy and a multilayer fully connected artificial neural network to approximate nonlinear dependence of shear wave velocity and layers thickness via values of phase velocity surface waves. Frequency-dependent noise in the data and errors of the inverse operator are projected onto the inverse problem solution. The results of unimodal and multimodal inversion are compared on the example of synthetic data processing. The experimental results show that using of machine learning approaches makes it possible to quickly and accurately estimate the posterior probability density of the reconstructed velocity model parameters.

Deep Neural Networks for Detection and Location of Microseismic Events and Velocity Model Inversion from Microseismic Data Acquired by Distributed Acoustic Sensing Array.

Fiber-optic cables have recently gained popularity for use as Distributed Acoustic Sensing (DAS) arrays for borehole microseismic monitoring due to their physical robustness as well as high spatial and temporal resolutions. As a result, the sensors record large amounts of data, making it very difficult to process in real-/semi-real-time using the conventional processing routines. We present a novel approach, based on deep learning, for handling the large amounts of DAS data in real-/semi-real-time. The proposed neural network was trained on synthetic microseismic data contaminated with real-ambient noise from field data and was validated using field DAS microseismic data obtained from a hydraulic fracturing operation. The results indicate that the trained network is capable of detecting and locating microseismic events from DAS data and simultaneously update the velocity model to a high degree of precision. The mean absolute errors in the event locations and the velocity model parameters are 2.04, 0.72, 2.76, 4.19 and 0.97 percent for distance (x), depth (z), P-wave velocity, S-wave velocity and density, respectively. In addition to automation and computational efficiency, deep learning reduces human expert data handling during processing, thus preserving data integrity leading to more accurate and reproducible results.

Polynomial surrogates for Bayesian traveltime tomography

This paper tackles the issue of the computational load encountered in seismic imaging by Bayesian traveltime inversion. In Bayesian inference, the exploration of the posterior distribution of the velocity model parameters requires extensive sampling. The computational cost of this sampling step can be prohibitive when the first arrival traveltime prediction involves the resolution of an expensive number of forward models based on the eikonal equation. We propose to rely on polynomial chaos surrogates of the traveltimes between sources and receivers to alleviate the computational burden of solving the eikonal equation during the sampling stage. In an offline stage, the approach builds a functional approximation of the traveltimes from a set of solutions of the eikonal equation corresponding to a few values of the velocity model parameters selected in their prior range. These surrogates then substitute the eikonal-based predictions in the posterior evaluation, enabling very efficient extensive sampling of Bayesian posterior, for instance, by a Markov Chain Monte Carlo algorithm. We demonstrate the potential of the approach using numerical experiments on the inference of two-dimensional domains with layered velocity models and different acquisition geometries (microseismic and seismic refraction contexts). The results show that, in our experiments, the number of eikonal model evaluations required to construct accurate surrogates of the traveltimes is low. Further, an accurate and complete characterization of the posterior distribution of the velocity model is possible, thanks to the generation of large sample sets at a low cost. Finally, we discuss the extension of the current approach to more realistic velocity models and operational situations.

Experimental investigation and its analysis of gas holdup in a three-phase counter-current microstructured bubble column

The present work investigates the gas holdup in a three-phase counter-current microstructured bubble column with and without the presence of surfactant. The coal (specific gravity 1.4) was taken to prepare slurry (a mixture of water and coal) and Methyl Isobutyl Carbinol (MIBC) was used as a surfactant. The effects of the dispersed phase and continuous phase velocity, solid loading and solid size on gas holdup are studied. The result shows that the gas holdup is highly dependent on the operating variables and the solid loading, but only to a small extent on the size of the particles. The gas holdup is analyzed with a slip velocity model. The maximum gas holdup in the column is found to be 38.570% and 33.195% in the presence and absence of the surfactant, respectively. It has been found that increasing the concentration of solids reduces gas holdup. Moreover, increasing particle size and dispersed phase and continuous phase velocity increases gas holdup. A novel correlation has been developed to predict the gas holdup and slip velocity model parameters. Gas holdup data from the available literature and the current experimental values were compared to the recommended correlation. The results of the study show that the recommended correlations give a precise prediction of experimental values in the error range of ± 20%.

Kinematic migration for determination of velocity model of the medium while solving practical problems of seismic exploration

Seismic migration allows solving different problems of seismic exploration and depending on the properties of the wave field which take part during the processing is subdivided into two types: kinematic and dynamic. Kinematic migration allows to determine seismic velocities and interfaces in geological medium for further implementation of dynamic migration can be enabled while solving different problems of seismic exploration where fast information on a priori parameters of velocity model of the study medium is needed. Kinematic migration for both fields of refracted and reflected waves is based on the continuation of temporal field which is implemented by finite-differential solution of eikonal equation. The values of temporal field are determined by distinct scheme which has quadratic degree of approximation and conditional stability that is proved during the studies. In this case the algorithm of kinematic migration of the field of refracted waves envisages two reverse continuations of temporal fields observed from two opposite sources while algorithm of kinematic migration of the field of reflected waves includes both direct continuation of temporal field from dotted source and reverse continuation from seismic receivers. Particularly urgent problem in this direction is the study of feasibility of finite-differential kinematic migration in case of determination of velocity model of multiple-structured geological medium while processing seismic data observed in different range of distances from the source. Possibilities of application of elaborated finite-differential kinematic migration for solving practical problems of seismic exploration have been considered in the paper. Effectiveness of elaborated techniques is demonstrated by model and real examples.

Full waveform inversion with angle-dependent gradient preconditioning using wavefield decomposition

The main objective of the full waveform inversion (FWI) is to deliver a velocity model which produces accurate and high resolution depth images. An integral part of FWI is the iterative optimization process aiming at minimizing an objective function describing the misfit between the observed and modelled data with respect to the velocity model parameters. However, FWI commonly suffers from local minima due to errors in data fitting and the starting velocity model. In this study, we develop a methodology to enhance convergence and mitigate the risk of having solutions falling into local minima. We employ optical flow to decompose the source and residual receiver wavefields and map them into the subsurface angle domain. Having evaluated the propagation vectors of the respective cleaner wavefields, subsurface reflection and azimuth angles can be robustly computed. Using such information we can apply angle-dependent preconditioning within the inversion algorithm. During a conventional FWI workflow, larger angles (i.e., transmission wavepaths) are mainly responsible for long wavelength velocity anomalies whereas narrow angles (i.e., reflection wavepaths) influence short wavelength velocity anomalies. In the proposed FWI methodology, we apply an optimized weighting function based on these estimated subsurface reflection angles in order to weight certain angle directions needed for gradient estimation using in parameter updating. Following a hierarchical approach, allowing for larger angles at the early stage of the FWI application, we initially solve for the smooth and longer wavelength velocity updates. Relaxing the opening angle from large to mid to narrow, the gradient direction will first estimate long wavelength velocity model updates (transmission-based wavepaths), following by short wavelength velocity model updates (reflection-based wavepaths). Moreover, angle-dependent-based gradient preconditioning can reduce cycle skips and improve convergence during FWI iterations.

Autonomous Vehicle Dynamic Model-Based Navigation for Small UAVs

This paper presents a novel approach to autonomous navigation for small UAVs, with no extra sensor added to the conventional INS/GNSS setup. The proposed method significantly increases the accuracy and reliability of autonomous navigation, especially for small UAVs with low-cost IMUs. This improvement is of special interest in the case of GNSS outages, where inertial coasting drifts very quickly. In the proposed architecture, the VDM provides the estimate of position, velocity, and attitude, which is updated within a filter based on available observations, such as IMU data or when available, GNSS measurements. The filter is capable of estimating wind velocity and dynamic model parameters, in addition to navigation states and IMU sensor errors. Monte Carlo simulations reveal major improvements in navigation accuracy compared to conventional INS/GNSS systems during GNSS outages of 5 min. A discussion on the observability is also presented at the end. Copyright (C) 2016 Institute of Navigation.

AUTONOMOUS NAVIGATION OF SMALL UAVS BASED ON VEHICLE DYNAMIC MODEL

This paper presents a novel approach to autonomous navigation for small UAVs, in which the vehicle dynamic model (VDM) serves as the main process model within the navigation filter. The proposed method significantly increases the accuracy and reliability of autonomous navigation, especially for small UAVs with low-cost IMUs on-board. This is achieved with no extra sensor added to the conventional INS/GNSS setup. This improvement is of special interest in case of GNSS outages, where inertial coasting drifts very quickly. In the proposed architecture, the solution to VDM equations provides the estimate of position, velocity, and attitude, which is updated within the navigation filter based on available observations, such as IMU data or GNSS measurements. The VDM is also fed with the control input to the UAV, which is available within the control/autopilot system. The filter is capable of estimating wind velocity and dynamic model parameters, in addition to navigation states and IMU sensor errors. Monte Carlo simulations reveal major improvements in navigation accuracy compared to conventional INS/GNSS navigation system during the autonomous phase, when satellite signals are not available due to physical obstruction or electromagnetic interference for example. In case of GNSS outages of a few minutes, position and attitude accuracy experiences improvements of orders of magnitude compared to inertial coasting. It means that during such scenario, the position-velocity-attitude (PVA) determination is sufficiently accurate to navigate the UAV to a home position without any signal that depends on vehicle environment.

AUTONOMOUS NAVIGATION OF SMALL UAVS BASED ON VEHICLE DYNAMIC MODEL

Abstract. This paper presents a novel approach to autonomous navigation for small UAVs, in which the vehicle dynamic model (VDM) serves as the main process model within the navigation filter. The proposed method significantly increases the accuracy and reliability of autonomous navigation, especially for small UAVs with low-cost IMUs on-board. This is achieved with no extra sensor added to the conventional INS/GNSS setup. This improvement is of special interest in case of GNSS outages, where inertial coasting drifts very quickly. In the proposed architecture, the solution to VDM equations provides the estimate of position, velocity, and attitude, which is updated within the navigation filter based on available observations, such as IMU data or GNSS measurements. The VDM is also fed with the control input to the UAV, which is available within the control/autopilot system. The filter is capable of estimating wind velocity and dynamic model parameters, in addition to navigation states and IMU sensor errors. Monte Carlo simulations reveal major improvements in navigation accuracy compared to conventional INS/GNSS navigation system during the autonomous phase, when satellite signals are not available due to physical obstruction or electromagnetic interference for example. In case of GNSS outages of a few minutes, position and attitude accuracy experiences improvements of orders of magnitude compared to inertial coasting. It means that during such scenario, the position-velocity-attitude (PVA) determination is sufficiently accurate to navigate the UAV to a home position without any signal that depends on vehicle environment.

Microthrix parvicella abundance associates with activated sludge settling velocity and rheology – Quantifying and modelling filamentous bulking

The objective of this work is to identify relevant settling velocity and rheology model parameters and to assess the underlying filamentous microbial community characteristics that can influence the solids mixing and transport in secondary settling tanks. Parameter values for hindered, transient and compression settling velocity functions were estimated by carrying out biweekly batch settling tests using a novel column setup through a four-month long measurement campaign. To estimate viscosity model parameters, rheological experiments were carried out on the same sludge sample using a rotational viscometer. Quantitative fluorescence in-situ hybridisation (qFISH) analysis, targeting Microthrix parvicella and phylum Chloroflexi, was used. This study finds that M. parvicella – predominantly residing inside the microbial flocs in our samples – can significantly influence secondary settling through altering the hindered settling velocity and yield stress parameter. Strikingly, this is not the case for Chloroflexi, occurring in more than double the abundance of M. parvicella, and forming filaments primarily protruding from the flocs. The transient and compression settling parameters show a comparably high variability, and no significant association with filamentous abundance. A two-dimensional, axi-symmetrical computational fluid dynamics (CFD) model was used to assess calibration scenarios to model filamentous bulking. Our results suggest that model predictions can significantly benefit from explicitly accounting for filamentous bulking by calibrating the hindered settling velocity function. Furthermore, accounting for the transient and compression settling velocity in the computational domain is crucial to improve model accuracy when modelling filamentous bulking. However, the case-specific calibration of transient and compression settling parameters as well as yield stress is not necessary, and an average parameter set – obtained under bulking and good settling conditions – can be used.

Integration of Arrival-Time Datasets for Consistent Quality Control: A Case Study of Amphibious Experiments along the Middle America Trench

We have integrated waveform and arrival‐onset data collected in Costa Rica as part of the National Science Foundation (NSF)‐sponsored Costa Rica Seismogenic Zone Experiment (CRSEIZE) and along central Costa Rica and Nicaragua as part of the German SFB 574 program. The five arrays, composed of different sensor types (one‐ and three‐component land and ocean bottom seismometers and hydrophones), were archived using different software packages (Antelope and SEISAN) and were automatically and manually picked using various quality criteria resulting in a disparate set of pick weights. We evaluate pick quality using automated arrival detection and picking algorithm based on the wavelet transform and Akaike information criterion picker. The consistency of the arrival information over various scales provides a basis for assigning a quality to the analyst pick. Approximately 31% of P arrival times and 26% of S times have been classified as high‐quality picks (quality 0–1). An additional 21% of P times and 27% of S arrivals are good quality (quality 2–3). The revised quality picks are mapped directly into new pick weights for inversion studies. We explore the effect of new weighting and removal of poor data by relocating hypocenters through a minimum 1D velocity model and conducting double‐difference local earthquake tomography (LET). Analysis of the hypocenter relocation and seismic velocity tomography results suggest that using the improved quality determinations have a greater effect on improving sharpness in the velocity images than on the magnitude of hypocentral movement. Online Material: Figures of waveforms, event statistics, and tomography; and tables of station and event parameters, station qualities, velocity model, and hypocentral parameters.

Determination and uncertainty of moment tensors for microearthquakes at Okmok Volcano, Alaska

Efforts to determine general moment tensors (MTs) for microearthquakes in volcanic areas are often hampered by small seismic networks, which can lead to poorly constrained hypocentres and inadequate modelling of seismic velocity heterogeneity. In addition, noisy seismic signals can make it difficult to identify phase arrivals correctly for small magnitude events. However, small volcanic earthquakes can have source mechanisms that deviate from brittle double-couple shear failure due to magmatic and/or hydrothermal processes. Thus, determining reliable MTs in such conditions is a challenging but potentially rewarding pursuit. We pursued such a goal at Okmok Volcano, Alaska, which erupted recently in 1997 and in 2008. The Alaska Volcano Observatory operates a seismic network of 12 stations at Okmok and routinely catalogues recorded seismicity. Using these data, we have determined general MTs for seven microearthquakes recorded between 2004 and 2007 by inverting peak amplitude measurements of P and S phases. We computed Green's functions using precisely relocated hypocentres and a 3-D velocity model. We thoroughly assessed the quality of the solutions by computing formal uncertainty estimates, conducting a variety of synthetic and sensitivity tests, and by comparing the MTs to solutions obtained using alternative methods. The results show that MTs are sensitive to station distribution and errors in the data, velocity model and hypocentral parameters. Although each of the seven MTs contains a significant non-shear component, we judge several of the solutions to be unreliable. However, several reliable MTs are obtained for a group of previously identified repeating events, and are interpreted as compensated linear-vector dipole events.

Regularized inversion of controlled source and earthquake data

Estimation of the seismic velocity structure of the Earth’s crust and upper mantle from travel-time data has advanced greatly in recent years. Forward modelling trial-and-error methods have been superseded by tomographic methods which allow more objective analysis of large two-dimensional and three-dimensional refraction and/or reflection data sets. The fundamental purpose of travel-time tomography is to determine the velocity structure of a medium by analysing the time it takes for a wave generated at a source point within the medium to arrive at a distribution of receiver points. Tomographic inversion of first-arrival travel-time data is a nonlinear problem since both the velocity of the medium and ray paths in the medium are unknown. The solution for such a problem is typically obtained by repeated application of linearized inversion. Regularization of the nonlinear problem reduces the ill posedness inherent in the tomographic inversion due to the under-determined nature of the problem and the inconsistencies in the observed data. This paper discusses the theory of regularized inversion for joint inversion of controlled source and earthquake data, and results from synthetic data testing and application to real data. The results obtained from tomographic inversion of synthetic data and real data from the northern Cascadia subduction zone show that the velocity model and hypocentral parameters can be efficiently estimated using this approach.

Effects of the number of events, their depth distributions, and the number of layers in a model on traveltime inversion

This study introduces the effects of the number of events, their source distributions, and the number of layers in the model on 1-D traveltime velocity inversion. The synthetic data generated from the known velocity model and hypocentral parameters allow us to test the accuracy of traveltime inversion. One dimensional traveltime inversion is performed using the hypocentral parameters determined by GA-MHYPO. The accuracy of traveltime velocity inversion with the synthetic data is tested by changing the number of events, the source positions, and the number of layers in a model. The estimated velocities in each layer are almost the same as the true velocity if the true depths of layer boundaries are known, accurate hypocentral parameters are used, and each layer has more than one event. The estimated weighted average velocity is close to the true one independent of the number of layers in the model and the number of events. The velocity errors in each layer decrease as the numbers of events in each layer increase in general. The estimated velocities using the noisy data are similar to those from the error-free data. However, the former has slightly larger velocity errors than the latter. Traveltime inversion may yield artificial low-velocity layers depending on used conditions. These results of synthetic studies are applied to the traveltime inversion of the Himalaya Nepal-Tibet Seismic Experiment earthquake data to estimate the detailed 1-D velocity structure beneath Nepal region.

Robust control strategy for suppression of regenerative chatter in turning

Chatter suppression is of great importance in machining processes for achieving more material removal rate, high precision and surface quality. In this paper, an H ∞ control algorithm is proposed for chatter suppression in the presence of tool wear and parameter uncertainties. Orthogonal turning process is modelled as a single degree of freedom model that includes the effect of tool flank wear. Control input of the system is the force provided by a piezo-actuator. The turning process model includes the uncertainties in cutting velocity, tool wear and dynamic model parameters. Using the μ -synthesis technique, an H ∞ optimal controller is designed based on a DK-iteration algorithm. The performance of this optimal controller is compared with a sliding mode controller. Both optimal robust and sliding mode controllers guarantee robust performance of the system against the uncertainties and result in the desired time responses of the vibration amplitude. Sliding mode control is able to achieve a faster response with less control effort. However, H ∞ robust control results in a more continuous and less oscillatory time response.

A simple analytical theory of localisation in 2D foam rheology

A very simple derivation is given for the dependence of localisation length on boundary velocity and various model parameters, in the continuum theory of 2D foam shear localisation. It is pointed out that the existence of distinct yield and limit stresses can complicate this theory for low boundary velocities, by introducing another mechanism for localisation, which does not depend on wall drag.

Kinematically equivalent velocity distributions

For a layered medium, the seismic velocity model can be vertically heterogeneous within the layers. The traveltime parameters estimated from each reflection must be converted into layer traveltime parameters by using the layer-stripping method. The layer traveltime parameters must be inverted into layer velocity model parameters. Interpretation or inversion of layer traveltime parameters depends on the chosen velocity model within the layer. Different or kinematically equivalent velocity distributions can result in the same traveltime parameters. The inversion problem for traveltime parameters is strongly nonunique even if they are estimated accurately. To evaluate the accuracy of a velocity model, one can choose the phase for the two-way propagator. The discrepancy in this phase factor between the kinematically equivalent velocity models depends on the number of traveltime parameters estimated and increases with spatial frequency. By estimating two traveltime parameters, we approximately preserve the average velocity, regardless of the complexity of the vertically heterogeneous model. By estimating three traveltime parameters, we approximately preserve the average velocity gradient.

Convex constrained optimization for the seismic reflection tomography problem

The nonlinear seismic reflection tomography problem consists on minimizing the function g( p) = || T − f( p)|| 2 2, where p is a vector containing the velocity model parameters and depth position of the reflectors, T contains the travel time of the rays, and f( p) is a nonlinear function that depends on the velocity of the subsurfaces. Recently a new approach, based on the spectral gradient method, was applied to find unconstrained local minimizers of g( p). This new idea requires very low storage and computational cost, but it presents an erratic behavior when applied to ill-conditioned problems. In this work, we combine the new low cost iterative technique with the recently proposed spectral projected method for convex constrained optimization to improve the convergence of the process, and to avoid the erratic behavior by imposing regularity into the optimization problem. Preliminary numerical experiments are also presented on synthetic data sets to illustrate the advantages of the new combined scheme.