Simulation Runtime Research Articles

Dynamic Steam-Trap-Control Simulation Technique: Formulation, Implementation, and Performance Analysis

Summary Steam-trap subcool is a technique that is used to maintain the energy efficiency of the steam-assisted-gravity-drainage (SAGD) process by most heavy-oil producers in Canada. The concept is rather simple (i.e., create a liquid pool around the production well to prevent steam from escaping from the steam chamber into the production well). A numerical steam trap based on a thermodynamic approach was implemented by Edmunds (2000), and it has been used in simulations with different types of wellbore models in commercial codes. In this approach, a thermodynamic relationship is solved as a well-residual equation to guarantee that the bottomhole temperature (BHT) is less than the saturation temperature of water at the hottest location along the wellbore. The location of the hottest spot along the wellbore is static in time. Steam trap is a dynamic process, and inflow temperatures can vary significantly along the wellbore according to the local fluid and rock properties along the well. It is highly possible that the location of the hottest spot along the well changes frequently with time during the SAGD operation. In this study, the simulation of a dynamic steam-trap-control technique is provided. The location of the hottest spot along the wellbore is scanned at every timestep. Severe numerical instabilities are observed when the thermodynamic approach is used. A new constraint based on the total production rate at reservoir condition is introduced. Details of the mathematical formulation and the numerical behavior of the new method are discussed in this paper. Several real field models with different wellbore designs (multiple tubing strings) are simulated, and the results of the new approach are compared with the thermodynamic approach. Simulation results show that the numerical performance of this new approach is significantly more stable. We also compared the run time of simulation between cases with new well constraint and thermodynamic steam-trap control, and results show a significant improvement in simulation run time.

DRV Evaluation of 6T SRAM Cell Using Efficient Optimization Techniques

An optimization based method which uses bisection search algorithm has been proposed to evaluate the accurate value of Data Retention Voltage (DRV) of a 6T Static Random Access Memory (SRAM) cell using 45 nm technology in the presence of process parameter variations. Further, we incorporate an Artificial Neural Network (ANN) block in our proposed methodology to optimize the simulation run time. The highest values obtained from these two methods are declared as the DRV. We noted an increase in DRV with temperature (T) and process variations (PVs). The main advantage of the proposed technique is to reduce the DRV evaluation time and for our case, we observe improvement in evaluation time of DRV by ≈46, ≈27, and ≈8 times at 25°C for 3 σ, 4 σ, and 5 σ variations, respectively, using ANN block to without using ANN block.

Studying the interactions in a mammalian nerve fiber: a functional modeling approach.

Modern therapeutic interventions are increasingly favoring electrical stimulation to treat neurophysiological dis-orders. These therapies are associated with suboptimal efficacy since most neurostimulation devices operate in an open-loop manner $(i.e.$, stimulation settings like frequency, amplitude are preprogrammed). A closed-loop system can dynamically adjust stimulation parameters and may provide efficient therapies. Computational models used to design these systems vary in complexity which can adversely affect their real-time performance. In this study, we compare two models of varying degrees of complexity. We constructed two computational models of a myelinated nerve fiber (functional versus mechanistic) each receiving two inputs: the underlying physiological activity at one end of the fiber, and the external stimulus applied to the middle of the fiber. We then defined relay reliability as the percentage of physiological action potentials that make it to the other end of the nerve fiber. We applied the two inputs to the fiber at various frequencies and analyze reliability. We found that the functional model and the mechanistic model have similar reliability properties, but the functional model significantly decreases the computational complexity and simulation run time. This modeling effort is the first step towards understanding and designing closed loop, real-time neurostimulation devices.

Parallel I/O for 3-D Global FDTD Earth–Ionosphere Waveguide Models at Resolutions on the Order of ~1 km and Higher Using HDF5

3-D finite-difference time-domain (FDTD) models of the global Earth–ionosphere waveguide have reached grid resolutions on the order of $1\times 1 \times 1$ km and higher. At these high resolutions, efficient parallel input–output (I/O) schemes are required to keep the simulation run times reasonable. However, generating optimal parallel I/O schemes is challenging due to the spherical-coordinate global mesh that includes merging of cells in the east–west direction as either pole is approached. Hierarchical Data Format 5 (HDF5) is applied and optimized for a global latitude–longitude FDTD model. The performance of the parallel HDF5 approach is compared with a message passing interface-based approach and a brute-force approach. The methodology presented here may be simplified to Cartesian-based grids in a straightforward manner.

Comparison between the NMPC and EOR control of wind turbines using LIDAR wind measurements

The upstream wind is known as the main source of disturbance to the wind turbine. Therefore, having the wind information before it hits the turbine, allows the wind turbine controller to take necessary actions to proper rejection of the disturbances. Several advanced control methods have been proposed to exploit the LIDAR wind measurements to enhance the wind turbine control. To date, the Nonlinear Model Predictive Control (NMPC), has been one of the most successful methods in using wind data to improve the control performance. However, due to the immense computational burden, its real-time application is still challenging. Very recently, an implementation of the Exact Output Regulation (EOR) scheme for wind turbines control has been proposed. In this paper, the performances of the two model-based controllers, namely the NMPC controller and the EOR controller, in mitigating mechanical loads on the DTU10MW wind turbine are compared against each other. The results are also compared against the classic baseline feedback controller. Simulation results indicate that both controllers show significant improvement in reducing fatigue loads on the wind turbine structure, whilst maintaining the power production at the desired rated level in comparison to the baseline controller. It is also shown that the EOR controller can closely compete with the disturbance rejection performance of the NMPC controller. The simulation running times are considerably lower for the EOR scheme, potentially making EOR more suitable for real-time applications.

Numerical and experimental investigation of H2-air and H2[sbnd]O2 detonation parameters in a 9 m long tube, introduction of a new detonation model

Experimental and numerical investigation of hydrogen-air and hydrogen-oxygen detonation parameters was performed. A new detonation model was introduced and validated against the experimental data. Experimental set-up consisted of 9 m long tube with 0.17 m in diameter, where pressure was measured with piezoelectric transducers located along the channel. Numerical simulations were performed within OpenFoam code based on progress variable equation where the detonative source term accounts for autoignition effects. Autoignition delay times were computed at a simulation run-time with the use of a multivariate regression model, where independent variables were: pressure, temperature and fuel concentration. The dependent variable was the autoignition delay time. Range of the analyzed gaseous mixture composition varied between 20% and 50% of hydrogen-air and 50%–66% of hydrogen in oxygen. Simulations were performed using LES one-equation eddy viscosity turbulence model in 2D and 3D. Calculations were validated against experimental data.

Validation of the material point method for the simulation of thin-walled tubes under lateral compression

This paper evaluates the performance of the material point method for the simulation of thin-walled tubes under lateral compression. Validation is carried out against actual experimental results for three different scenarios, namely: quasi-static loading, impact on rigid target, and wave propagation. A systematic approach is taken to gain insight on the trade-off between accuracy and computational cost at different levels of refinement of the model. Accuracy is assessed by comparing simulation results against experimental data. Computational cost is measured by the simulation runtime, or more specifically, in terms of the ratio between simulation time and execution time. Results indicate that, from highest influence to lowest, the factors affecting accuracy are: grid resolution, particle count along the thickness of the tube, and particle count along the circumference of the tube. Overall, it is demonstrated that the MPM is a reliable and accurate method to model circular thin-walled tubes under various excitation conditions.

An algorithm for fast elastic wave simulation using a vectorized finite difference operator

Modern geophysical imaging techniques exploit the full wavefield information which can be simulated numerically. These numerical simulations are computationally expensive due to several factors, such as a large number of time steps and nodes, big size of the derivative stencil and huge model size. Besides these constraints, it is also important to reformulate the numerical derivative operator for improved efficiency. In this paper, we have introduced a vectorized derivative operator over the staggered grid with shifted coordinate systems. The operator increases the efficiency of simulation by exploiting the fact that each variable can be represented in the form of a matrix. This operator allows updating all nodes of a variable defined on the staggered grid, in a manner similar to the collocated grid scheme and thereby reducing the computational run-time considerably. Here we demonstrate an application of this operator to simulate the seismic wave propagation in elastic media (Marmousi model), by discretizing the equations on a staggered grid. We have compared the performance of this operator on three programming languages, which reveals that it can increase the execution speed by a factor of at least 2–3 times for FORTRAN and MATLAB; and nearly 100 times for Python. We have further carried out various tests in MATLAB to analyze the effect of model size and the number of time steps on total simulation run-time. We find that there is an additional, though small, computational overhead for each step and it depends on total number of time steps used in the simulation. A MATLAB code package, ’FDwave’, for the proposed simulation scheme is available upon request.

Accelerating cycle-accurate system-level simulations through behavioral templates

This work presents a method to accelerate the running time of cycle-accurate system-level simulations. The proposed method substitutes the computational units specified at the RT-level or behavioral level in the system with fast templates of the exact latency of its original design and thus, preserving the performance measurement accuracy. The method has been extended to deal with control dependencies inside loops in order to maintain high modelling accuracies under any condition. Experimental results show that our proposed method works well speeding up individual accelerator kernels by up to 8 and 15×. Moreover when used to explore entire SoC configurations it achieves similar result as using the exact models while achieving an average speedup of 4.7×.

Reproducing Quantum Probability Distributions at the Speed of Classical Dynamics: A New Approach for Developing Force-Field Functors.

Modeling nuclear quantum effects is required for accurate molecular dynamics (MD) simulations of molecules. The community has paid special attention to water and other biomolecules that show hydrogen bonding. Standard methods of modeling nuclear quantum effects like Ring Polymer Molecular Dynamics (RPMD) are computationally costlier than running classical trajectories. A force-field functor (FFF) is an alternative method that computes an effective force field that replicates quantum properties of the original force field. In this work, we propose an efficient method of computing FFF using the Wigner-Kirkwood expansion. As a test case, we calculate a range of thermodynamic properties of Neon, obtaining the same level of accuracy as RPMD, but with the shorter runtime of classical simulations. By modifying existing MD programs, the proposed method could be used in the future to increase the efficiency and accuracy of MD simulations involving water and proteins.

Exploring the feasibility of lossy compression for PDE simulations

Checkpoint restart plays an important role in high-performance computing (HPC) applications, allowing simulation runtime to extend beyond a single job allocation and facilitating recovery from hardware failure. Yet, as machines grow in size and in complexity, traditional approaches to checkpoint restart are becoming prohibitive. Current methods store a subset of the application’s state and exploit the memory hierarchy in the machine. However, as the energy cost of data movement continues to dominate, further reductions in checkpoint size are needed. Lossy compression, which can significantly reduce checkpoint sizes, offers a potential to reduce computational cost in checkpoint restart. This article investigates the use of numerical properties of partial differential equation (PDE) simulations, such as bounds on the truncation error, to evaluate the feasibility of using lossy compression in checkpointing PDE simulations. Restart from a checkpoint with lossy compression is considered for a fail-stop error in two time-dependent HPC application codes: PlasComCM and Nek5000. Results show that error in application variables due to a restart from a lossy compressed checkpoint can be masked by the numerical error in the discretization, leading to increased efficiency in checkpoint restart without influencing overall accuracy in the simulation.

The parallel multi-physics phase-field framework Pace3D

The phase-field method has been established for the numerical investigation of various microstructure evolution processes. The accurate description of these complex processes requires large domains and suitable models, allowing to couple several physical fields in statistical representative volume elements. To simplify the implementation of new models and to reduce the simulation run time, different frameworks have been developed. In this work, the parallel multi-physics phase-field framework Pace3D is introduced. The general structure of the solver, its modules and the parallelization are described. For increasing the performance of the implemented phase-field models, various optimization techniques are outlined. To efficiently store the simulation results, different data formats and parallel writing mechanisms are presented. The performance of an optimized implementation for a specific phase-field model is analyzed on a single core, showing a good peak performance. For a single node, the memory bandwidth is analyzed and ruled out as possible bottleneck. In addition, a proper weak scaling behavior is demonstrated on the three supercomputers ForHLR I, ForHLR II and Hazel Hen, for up to 96 100 cores.

Evaluation of bogie centre bowl friction models in the context of safety against derailment simulation predictions

This paper investigates the influence of the models of dry friction suspension component used in the numerical simulation of freight car dynamics on the possible predictions related to safety against derailment. Moreover, the performance of the models is compared on the basis of simulation runtime. Three groups of static and dynamic models of friction are selected as three different approaches which are often utilized in popular multibody simulation tools. The models are used for the centre bowl friction joint between the car body and bogies. The friction torque in the centre bowl affects the rotational resistance of the bogie in motion relative to the car body and consequently influences the interaction forces between the wheels and rails. Simulation results of safety against derailment index obtained from the vehicle curving scenario suggest noticeable differences in predictions of the selected models of friction.

Accelerated and Reliable Analog Circuits Yield Analysis Using SMT Solving Techniques

Existing yield analysis methods are computationally expensive and generally encounter challenges with high-dimensional process parameters space. In this paper, we propose a new method for accelerated and reliable computation of parametric yield that combines the advantages of sparse regression and satisfiability modulo theory (SMT) solving techniques, and avoids issues in both. The key idea is to characterize the failure regions as a collection of hyperrectangles in the parameters space. Toward this goal, the method constructs sparse polynomial models based on adaptive least absolute shrinkage and selection operator to find low degree approximations of the circuit performances. A procedure inspired by statistical model checking is then introduced to assess the model accuracy. Given the constructed models, an SMT-based solving algorithm is employed to locate the failure hyperrectangles in the parameters space. The yield estimation is based on a geometric calculation of probabilistic volumes subtended by the located hyperrectangles. We demonstrate the effectiveness of our method using circuits that require expensive run-time simulation during yield evaluation. They include: an integrated ring oscillator, a 6T static RAM cell and a multistage fully-differential amplifier. Experimental results show that the proposed method is suitable for handling problems with tens of process parameters. Meanwhile, it can provide $5{\times }{-}2000{\times}$ speed-up over Monte Carlo methods, when a high prediction accuracy is required.

Impacts of building geometry modeling methods on the simulation results of urban building energy models

Urban-scale building energy modeling (UBEM)—using building modeling to understand how a group of buildings will perform together—is attracting increasing attention in the energy modeling field. Unlike modeling a single building, which will use detailed information, UBEM generally uses existing building stock data consisting of high-level building information. This study evaluated the impacts of three zoning methods and the use of floor multipliers on the simulated energy use of 940 office and retail buildings in three climate zones using City Building Energy Saver. The first zoning method, OneZone, creates one thermal zone per floor using the target building’s footprint. The second zoning method, AutoZone, splits the building’s footprint into perimeter and core zones. A novel, pixel-based automatic zoning algorithm is developed for the AutoZone method. The third zoning method, Prototype, uses the U.S. Department of Energy’s reference building prototype shapes. Results show that simulated source energy use of buildings with the floor multiplier are marginally higher by up to 2.6% than those modeling each floor explicitly, which take two to three times longer to run. Compared with the AutoZone method, the OneZone method results in decreased thermal loads and less equipment capacities: 15.2% smaller fan capacity, 11.1% smaller cooling capacity, 11.0% smaller heating capacity, 16.9% less heating loads, and 7.5% less cooling loads. Source energy use differences range from -7.6% to 5.1%. When comparing the Prototype method with the AutoZone method, source energy use differences range from -12.1% to 19.0%, and larger ranges of differences are found for the thermal loads and equipment capacities. This study demonstrated that zoning methods have a significant impact on the simulated energy use of UBEM. One recommendation resulting from this study is to use the AutoZone method with floor multiplier to obtain accurate results while balancing the simulation run time for UBEM.

A fast algorithm for calculating isothermal phase behavior using machine learning

Compositional models are frequently used to describe fluids in petroleum reservoir simulation, particularly for simulations of enhanced oil recovery. While compositional models are more accurate than black oil models, they incur a larger computational cost, in part, due to more complex phase-equilibrium calculations and can result in longer run times. Here, we develop an algorithm to reduce the cost of phase-equilibrium calculations for compositional models by applying two machine learning techniques: relevance vector machines and artificial neural networks. We test the algorithm on three fluid data sets and find a speedup of over 20% with an error of 0.01%, and a speedup of over 90% with a maximum error of 5%. These results suggest that the algorithm can be used to reduce the overall run time of compositional reservoir simulations with a small impact on accuracy.

A comparative study of model-based capacity estimation algorithms in dual estimation frameworks for lithium-ion batteries under an accelerated aging test

The actual capacity of a battery is an essential indicator for calculating both the state of health and the remaining electric driving range. Numerous model-based techniques employing adaptive filters have been proposed for the online capacity estimation. However, in these filter-based methods, the impacts of filter configurations and the algorithm effectiveness at various aging stages have not yet been fully investigated. To address this gap and to evaluate the performance of three most popular algorithms, i.e. the extended Kalman filter, the particle filter, and the least-squares-based filter, they are coupled with an SOC estimator in dual frameworks. The characterization and accelerated aging tests have been carried out on a lithium-ion battery. After investigating the possible impacts from the configurations, the tracking accuracy, the robustness against the uncertainty of the initial capacity and the long-term performance of the three algorithms are compared. Furthermore, their computational efforts are extensively assessed regarding complexity, simulation runtime as well as compiled code size utilizing an automotive prototype hardware. The results show that the extended Kalman filter is the least sensitive to model degradation with the lowest computational effort; the particle filter shows the fastest convergence speed but has the highest computational effort; and the least-squares-based filter has an intermediate behavior in both long-term performance and computational effort.

Direct Probabilistic Load Flow in Radial Distribution Systems Including Wind Farms: An Approach Based on Data Clustering

The ongoing study aims to establish a direct probabilistic load flow (PLF) for the analysis of wind integrated radial distribution systems. Because of the stochastic output power of wind farms, it is very important to find a method which can reduce the calculation burden significantly, without having compromising the accuracy of results. In the proposed approach, a K-means based data clustering algorithm is employed, in which all data points are bunched into desired clusters. In this regard, probable agents are selected to run the PLF algorithm. The clustered data are used to employ the Monte Carlo simulation (MCS) method. In this paper, the analysis is performed in terms of simulation run-time. Also, this research follows a two-fold aim. In the first stage, the superiority of data clustering-based MCS over the unsorted data MCS is demonstrated properly. Moreover, the impact of data clustering-based MCS and unsorted data-based MCS is investigated using an indirect probabilistic forward/backward sweep (PFBS) method. Thus, in the second stage, the simulation run-time comparison is carried out rigorously between the proposed direct PLF and the indirect PFBS method to examine the computational burden effects. Simulation results are exhibited on the IEEE 33-bus and 69-bus radial distribution systems.

Automatic Correction of Dynamic Power Management Architecture in Modern Processors

The increasing demand for lower power forces designers to use sophisticated power management strategies such as multivoltage and power gating which are often accompanied with many design bugs. Correcting such bugs can be a time-consuming process that requires considerable manual efforts. In this paper, we propose a scalable automated method for correcting dynamic power management architectures by an incremental SAT-based mechanism. First, an initial counterexample (CEX) is generated by checking the equivalency between the specification model of the processor and its buggy implementation model. Then, we find two candidate solutions instead of one to satisfy this CEX. If two solutions are not equivalent, we generate new CEX in an iterative process which effectively converges into the final solution. The proposed method enables designers to correct multiple bugs such as missing isolation cells between two power domains, disordering in the sequence of control signals, error in the data restoring or saving, and powering off in always-on domains which are not addressed by existing methods. We have shown the effectiveness of our method on modern processors supporting complex power management mechanisms. The results confirm that our proposed method, respectively, reduces symbolic simulation steps and runtime by 2.33× and 47.93× compared to the state-of-the-art methods.

Scalable Cloning on Large-Scale GPU Platforms with Application to Time-Stepped Simulations on Grids

Cloning is a technique to efficiently simulate a tree of multiple what-if scenarios that are unraveled during the course of a base simulation. However, cloned execution is highly challenging to realize on large, distributed memory computing platforms, due to the dynamic nature of the computational load across clones, and due to the complex dependencies spanning the clone tree. We present the conceptual simulation framework, algorithmic foundations, and runtime interface of C lone X, a new system we designed for scalable simulation cloning. It efficiently and dynamically creates whole logical copies of a dynamic tree of simulations across a large parallel system without full physical duplication of computation and memory. The performance of a prototype implementation executed on up to 1,024 graphical processing units of a supercomputing system has been evaluated with three benchmarks—heat diffusion, forest fire, and disease propagation models—delivering a speed up of over two orders of magnitude compared to replicated runs. The results demonstrate a significantly faster and scalable way to execute many what-if scenario ensembles of large simulations via cloning using the C lone X interface.