Search

Agent based modelling (ABM) offers a powerful abstraction for scientific study in a broad range of domains. The use of agent based simulators encourages good software engineering design such as separation of concerns, that is, the uncoupling of the model description from its implementation detail. A major limitation in current approaches to ABM simulation is that of the trade off between simulator flexibility and performance. It is common that highly optimised simulations, such as those which target graphics processing units (GPU) hardware, are implemented as standalone software. This work presents a software framework (FLAME GPU 2) which balances flexibility with performance for general purpose ABM. Methods for ensuring high computational efficacy are demonstrated by, minimising data movement, and ensuring high device utilisation by exploiting opportunities for concurrent code execution within a model and through the use of ensembles of simulations. A novel hierarchical sub‐modelling approach is also presented which can be used to model certain types of recursive behaviours. This feature is shown to be essential in providing a mechanism to resolve competition for resources between agents within a parallel environment which would otherwise introduce race conditions. To understand the performance characteristics of the software, a benchmark model with millions of agents is used to explore the use of simulation ensembles and to parametrically investigate concurrent code execution within a model. Performance speedups are demonstrated of 3.5×$$ \times $$ and 10×$$ \times $$ respectively over a baseline GPU implementation. Our hierarchical sub‐modelling approach is used to demonstrate the implementation of a recursive algorithm to resolve competition of agent movement which occurs as a result of agent desire to simultaneously occupy discrete areas high in a 'resource'. The algorithm is used to implement a classical socio‐economics model, Sugarscape, with populations of up to 16M agents. [ABSTRACT FROM AUTHOR]

TreeSim is an open-source, user-extendable framework that offers new opportunities for users to model, simulate, visualize, and animate the dynamics of forest. It provides a general environment for modellers to implement and evaluate forest growth models. In this paper, we introduce the object-oriented architecture for our proposed simulator and examine its performance to model and simulate forest growth, management, and dynamics. The simulator uses various models to predict the dynamics and attributes of forest (e.g., regeneration, increment, mortality, biodiversity, biomass, carbon, dead wood). Finally, TreeSim can be integrated into a simulation–optimization framework for decision-making support.

An open challenge on the road to unraveling the brain's multilevel organization is establishing techniques to research connectivity and dynamics at different scales in time and space, as well as the links between them. This work focuses on the design of a framework that facilitates the generation of multiscale connectivity in large neural networks using a symbolic visual language capable of representing the model at different structural levels—ConGen. This symbolic language allows researchers to create and visually analyze the generated networks independently of the simulator to be used, since the visual model is translated into a simulator-independent language. The simplicity of the front end visual representation, together with the simulator independence provided by the back end translation, combine into a framework to enhance collaboration among scientists with expertise at different scales of abstraction and from different fields. On the basis of two use cases, we introduce the features and possibilities of our proposed visual language and associated workflow. We demonstrate that ConGen enables the creation, editing, and visualization of multiscale biological neural networks and provides a whole workflow to produce simulation scripts from the visual representation of the model. [ABSTRACT FROM AUTHOR]

An open challenge on the road to unraveling the brain's multilevel organization is establishing techniques to research connectivity and dynamics at different scales in time and space, as well as the links between them. This work focuses on the design of a framework that facilitates the generation of multiscale connectivity in large neural networks using a symbolic visual language capable of representing the model at different structural levels—ConGen. This symbolic language allows researchers to create and visually analyze the generated networks independently of the simulator to be used, since the visual model is translated into a simulator-independent language. The simplicity of the front end visual representation, together with the simulator independence provided by the back end translation, combine into a framework to enhance collaboration among scientists with expertise at different scales of abstraction and from different fields. On the basis of two use cases, we introduce the features and possibilities of our proposed visual language and associated workflow. We demonstrate that ConGen enables the creation, editing, and visualization of multiscale biological neural networks and provides a whole workflow to produce simulation scripts from the visual representation of the model.

This work presents the first simulation of a large-scale, bio-physically constrained cerebellum model performed on neuromorphic hardware. A model containing 97,000 neurons and 4.2 million synapses is simulated on the SpiNNaker neuromorphic system. Results are validated against a baseline simulation of the same model executed with NEST, a popular spiking neural network simulator using generic computational resources and double precision floating point arithmetic. Individual cell and network-level spiking activity is validated in terms of average spike rates, relative lead or lag of spike times, and membrane potential dynamics of individual neurons, and SpiNNaker is shown to produce results in agreement with NEST. Once validated, the model is used to investigate how to accelerate the simulation speed of the network on the SpiNNaker system, with the future goal of creating a real-time neuromorphic cerebellum. Through detailed communication profiling, peak network activity is identified as one of the main challenges for simulation speed-up. Propagation of spiking activity through the network is measured, and will inform the future development of accelerated execution strategies for cerebellum models on neuromorphic hardware. The large ratio of granule cells to other cell types in the model results in high levels of activity converging onto few cells, with those cells having relatively larger time costs associated with the processing of communication. Organizing cells on SpiNNaker in accordance with their spatial position is shown to reduce the peak communication load by 41%. It is hoped that these insights, together with alternative parallelization strategies, will pave the way for real-time execution of large-scale, bio-physically constrained cerebellum models on SpiNNaker. This in turn will enable exploration of cerebellum-inspired controllers for neurorobotic applications, and execution of extended duration simulations over timescales that would currently be prohibitive using conventional computational platforms.

This work presents the first simulation of a large-scale, bio-physically constrained cerebellum model performed on neuromorphic hardware. A model containing 97,000 neurons and 4.2 million synapses is simulated on the SpiNNaker neuromorphic system. Results are validated against a baseline simulation of the same model executed with NEST, a popular spiking neural network simulator using generic computational resources and double precision floating point arithmetic. Individual cell and network-level spiking activity is validated in terms of average spike rates, relative lead or lag of spike times, and membrane potential dynamics of individual neurons, and SpiNNaker is shown to produce results in agreement with NEST. Once validated, the model is used to investigate how to accelerate the simulation speed of the network on the SpiNNaker system, with the future goal of creating a real-time neuromorphic cerebellum. Through detailed communication profiling, peak network activity is identified as one of the main challenges for simulation speed-up. Propagation of spiking activity through the network is measured, and will inform the future development of accelerated execution strategies for cerebellum models on neuromorphic hardware. The large ratio of granule cells to other cell types in the model results in high levels of activity converging onto few cells, with those cells having relatively larger time costs associated with the processing of communication. Organizing cells on SpiNNaker in accordance with their spatial position is shown to reduce the peak communication load by 41%. It is hoped that these insights, together with alternative parallelization strategies, will pave the way for real-time execution of large-scale, bio-physically constrained cerebellum models on SpiNNaker. This in turn will enable exploration of cerebellum-inspired controllers for neurorobotic applications, and execution of extended duration simulations over timescales that would currently be prohibitive using conventional computational platforms. [ABSTRACT FROM AUTHOR]

Purpose - As main mode of modern service industry and future economy society, the research on crowd network can greatly facilitate governances of economy society and make it more efficient, humane, sustainable and at the same time avoid disorders. However, because most results cannot be observed in real world, the research of crowd network cannot follow a traditional way. Simulation is the main means to put forward related research studies. Compared with other large-scale interactive simulations, simulation for crowd network has challenges of dynamic, diversification and massive participants. Fortunately, known as the most famous and widely accepted standard, high level architecture (HLA) has been widely used in large-scale simulations. But when it comes to crowd network, HLA has shortcomings like fixed federation, limited scale and agreement outside the software system. Design/methodology/approach - This paper proposes a novel reflective memory-based framework for crowd network simulations. The proposed framework adopts a two-level federation-based architecture, which separates simulation-related environments into physical and logical aspect to enhance the flexibility of simulations. Simulation definition is introduced in this architecture to resolve the problem of outside agreements and share resources pool (constructed by reflective memory) is used to address the systemic emergence and scale problem. Findings - With reference to HLA, this paper proposes a novel reflective memory-based framework toward crowd network simulations. The proposed framework adopts a two-level federation-based architecture, system-level simulation (system federation) and application-level simulation (application federations), which separates simulation-related environments into physical and logical aspect to enhance the flexibility of simulations. Simulation definition is introduced in this architecture to resolve the problem of outside agreements and share resources pool (constructed by reflective memory) is used to address the systemic emergence and scale problem. Originality/value - Simulation syntax and semantic are all settled under this framework by templates, especially interface templates, as simulations are separated by two-level federations, physical and logical simulation environment are considered separately; the definition of simulation execution is flexible. When developing new simulations, recompile is not necessary, which can acquire much more reusability, because reflective memory is adopted as share memory within given simulation execution in this framework; population can be perceived by all federates, which greatly enhances the scalability of this kind of simulations; communication efficiency and capability has greatly improved by this share memory-based framework.

A review is conducted of the state of progress in physical modeling and computational simulations of turbulent flames as pertaining to Type Ia supernovae. The fundamental governing equations are provided, along with various assumptions made for their simplifications; and the numerical means of their discretizations. A case is made for a more systematic inclusion of turbulence, including turbulence-reaction interactions, for future investigations. [ABSTRACT FROM AUTHOR]

Highlights • A large-scale of 15,488 EnergyPlus simulations for commercial buildings in the U.S. • Nationwide energy savings for a single and combinations of multiple ECMs. • Nationwide energy savings for the ECM of fault detection and diagnostics. Abstract Buildings consumed 40% of the energy in the United States, which is more than any other sectors of the U.S. economy, including transportation and industry. The majority of this energy was used for lighting, space heating, cooling, ventilation, and water heating in the buildings. It is well known that the carbon emission is proportional to the energy consumption. Enhancing building efficiency becomes one of the easiest, most immediate and most cost-effective ways to reduce carbon emissions. This paper focuses on the simulation-based nationwide energy savings analysis of different energy conservation measures (ECMs) to enhance building energy performance. The study includes 44 different ECMs, which were simulated in 16 different climate zones for 11 different commercial building types in the U.S. Simulations were conducted using a whole building simulation program, EnergyPlus. Commercial prototype building models from the U.S. Department of Energy [1] were used as the baselines. Model input files were automatically modified to incorporate with 44 different ECMs using scripts. The baselines were based on American Society of Heating, Refrigeration, and Air-conditioning Engineers (ASHRAE) [2] Standard 90.1 and International Energy Conservation Code (IECC). A total of 15,488 simulations were conducted to study the nationwide energy savings potentials. The details of these 44 proposed ECMs and their corresponding nationwide saving potentials are presented and discussed in this paper. [ABSTRACT FROM AUTHOR]

The Korteweg–de Vries (KdV) equation is often adopted to simulate phase-resolved random waves on large scale in shallow water. It shows that the KdV equation is computationally efficient and can give sufficiently accurate results, but it is not always suitable and the error by using it cannot be predicted. This paper attempts to give the quantitative formulas for estimating the error of the statistics when simulating random waves in shallow water by using it. The formulas are obtained by fitting the errors of the KdV equation in comparison with the fully nonlinear model using the same initial condition based on the Wallops spectrum with a wide range of parameters. This paper also demonstrates how the formulas would be used, e.g., to estimate the error of the results by using the KdV model, or to justify its suitability for modelling random waves on large scale in shallow water. [ABSTRACT FROM AUTHOR]

Since there are a lot of problems of the traditional multi-resolution model and simulation technique such as the chained disaggregation problem, difficult consistence maintenance problem, high complexity and poor practicability, this paper propose a cycle drive dynamic aggregation and disaggregation (AD) multi-resolution (CADMR) simulation model. The AD operation is not controlled by the entities' complex interaction process but is by the fixed cycle, besides, cross resolution communications are allowed and the model aims to achieve the macro consistence instead of the micro consistence. Introduces the simulation and schedule mechanism, builds the formalized equation, and then the effect of this model to the characters of the system of systems is analyzed. Besides, the controllable character (error and computation complexity) of this model is proved theoretically and the existence of the macro consistence solution is also proved, at the same time, the unity-gain factor is existent and is only one, for which the solving process is simplified. Lastly, taking the weapon SoS as example and based on its multi resolution framework, the feasibility and validation of the model is verified with experimental confirmation. The model has an important promotion meaning for the popular application of the multi-resolution model and simulation technology. [ABSTRACT FROM AUTHOR]

Phase field (PF) models are one of the most popular methods for simulating solidification microstructures due to their fundamental connections to the physics of phase transformations. However, these methods are numerically very stiff due to the multiple length scales in a solidifying material, from the nanoscopic solid-liquid interface, to dendritic structures on the order of hundreds of microns. While this problem can be greatly alleviated by thin-interface analytical treatments of the PF equations, additional numerical methods are required to explore experimentally relevant sample sizes and times scales. It was shown about 18 years ago that the use of dynamic adaptive mesh refinement (AMR) can alleviate this problem by exploiting the simple fact that the majority of the solidification kinetics occur at the solid-liquid interface, which scales with a lower dimensionality than the embedding system itself. AMR methods, together with asymptotic analysis, nowadays provide one of the most efficient numerical strategies for self-consistent quantitative PF modelling of solidification microstructure processes. This paper highlights the latest developments in the AMR technique for 3D modelling of solidification using classical phase field equations. This includes a move away from finite element techniques to faster finite differencing through the use of dynamic mini-meshes which are each associated with each node of a 3D Octree data structure, and distributed MPI parallelism that uses a new communication algorithm to decompose a 3D domain into multiple adaptive meshes that are spawned on separate cores. The numerical technique is discussed, followed by demonstrations of the new AMR algorithm on select benchmark solidification problems, as well as some illustrations of multi-phase modelling using a recently developed multi-order parameter phase field model. [ABSTRACT FROM AUTHOR]

This brief presents a reconfigurable and efficient 2-D neuron model capable of extending to higher dimensions. The model is applied to the Izhikevich and FitzHugh-Nagumo neuron models as 2-D case studies and to the Hindmarsh-Rose model as a 3-D case study. Hardware synthesis and physical implementations show that the resulting circuits can reproduce neural dynamics with acceptable precision and considerably low hardware overhead compared to previously published piecewise linear models. [ABSTRACT FROM PUBLISHER]

A method for the large-scale simulation with the lattice Boltzmann method (LBM) is proposed for liquid water movement in a gas diffusion layer (GDL) of polymer electrolyte membrane fuel cells. The LBM is able to analyze two-phase flows in complex structures, however the simulation domain is limited due to heavy computational loads. This study investigates a variety means to reduce computational loads and increase the simulation areas. One is applying an LBM treating two-phases as having the same density, together with keeping numerical stability with large time steps. The applicability of this approach is confirmed by comparing the results with rigorous simulations using actual density. The second is establishing the maximum limit of the Capillary number that maintains flow patterns similar to the precise simulation; this is attempted as the computational load is inversely proportional to the Capillary number. The results show that the Capillary number can be increased to 3.0 × 10 −3 , where the actual operation corresponds to Ca = 10 −5 ∼10 −8 . The limit is also investigated experimentally using an enlarged scale model satisfying similarity conditions for the flow. Finally, a demonstration is made of the effects of pore uniformity in GDL as an example of a large-scale simulation covering a channel. [ABSTRACT FROM AUTHOR]

Conventional interatomic potential methods for Li-ion battery cathode materials normally use fixed-charge models, which are not accurate enough to model the dynamical oxidation state change of transition metals during electrochemical reactions. In order to enable more accurate large-scale simulations for battery cathode materials, here we report a semiempirical interatomic potential of the Li-Ni-O ternary system based on an advanced dynamic charge method: Charge-Transfer Modified Embedded-Atom Method (CT-MEAM). The potential is parameterized by fitting the atomic Bader charges and energy-strain curves of the LiNiO 2 cathode material under uniaxial, biaxial and hydrostatic strains to results obtained with ab initio density-functional theory (DFT) calculations. A variety of structural, electrochemical and dynamical properties derived from the fitted CT-MEAM potential are observed to be in excellent agreement with previous DFT and experimental data. The transferability of the potential is validated by comparing relative phase stabilities and charge distribution states within the ternary Li-Ni-O systems. Our results support the capability of the present CT-MEAM to model complex ternary transition metal oxides. This method will facilitate not only the optimal design of Lithium-ion battery cathode materials, but also other transition metal oxide-based applications involving electrochemical reactions. [ABSTRACT FROM AUTHOR]

A slender body's impact and water entry generate a violent flow with the free surface, time-dependent hydrodynamic loads, and motion or structural responses. The strong three-dimensional effects of the two-phase flow amplify the computational cost of predicting the hydrodynamic flows accurately with the free surface. This paper proposed a high-efficiency Smoothed Particle Hydrodynamics (SPH) model to simulate the oblique water entry of a cylinder. The number of particles is almost 50 million. The numerical model is validated by comparison with the experimental data and the convergence study of particle resolution. The results show that the cavity may shift transversely due to the unbalanced pressure gradients. The transverse shift is stronger with a smaller entry angle. The results are compared with Logvinovich's model, which shows that Logvinovich's model can accurately predict the shape of the cavity away from the splash, whereas the SPH model can accurately predict violent deformation, sealing, and breaking of the free surface in all stages of cavity evolution. The splash dynamics are quantified, benefiting from the sharp interface computed by the SPH method of high resolution. The splash can obtain more kinetic energy when the entry angle is small. Besides, The pinch-off pattern is more complicated than the vertical water entry due to the transverse shift of the cavity. Two columnar bubbles are separated from the cavity due to the transverse shift of the cavity when the entry angle is small. [Display omitted] • The oblique water entry is computed by multi-resolution two-phase SPH method. • The pressure gradient generated transveral shift of the cavity is found. • The changes of the potential and kinetic energy of splashes are presented. [ABSTRACT FROM AUTHOR]

This paper demonstrates a new process that has been specifically designed for the support of the U.S. Department of Transportation's (DOT's) Corporate Average Fuel Economy (CAFE) standards. In developing the standards, DOT's National Highway Traffic Safety Administration made use of the CAFE Compliance and Effects Modeling System (the 'Volpe model' or the 'CAFE model'), which was developed by DOT's Volpe National Transportation Systems Center for the 2005-2007 CAFE rulemaking and has been continuously updated since. The model is the primary tool used by the agency to evaluate potential CAFE stringency levels by applying technologies incrementally to each manufacturer's fleet until the requirements under consideration are met. The Volpe model relies on numerous technology-related and economic inputs, such as market forecasts, technology costs, and effectiveness estimates; these inputs are categorized by vehicle classification, technology synergies, phase-in rates, cost learning curve adjustments, and technology 'decision trees'. Part of the model's function is to estimate CAFE improvements that a given manufacturer could achieve by applying additional technology to specific vehicles in its product line. A significant number of inputs to the Volpe decision-tree model are related to the effectiveness (fuel consumption reduction) of each fuel-saving technology. Argonne National Laboratory has developed a fullvehicle simulation tool named Autonomie, which has become one of the industry's standard tools for analyzing vehicle energy consumption and technology effectiveness. Full-vehicle simulation tools use physics-based mathematical equations, engineering characteristics (e.g., engine maps, transmission shift points, and hybrid vehicle control strategies), and explicit drive cycles to predict the effectiveness of individual and combined fuel-saving technologies. The Large-Scale Simulation Process accelerates and facilitates the assessment of individual technological impacts on vehicle fuel economy. This paper will show how Argonne efficiently simulates hundreds of thousands of vehicles to model anticipated future vehicle technologies. [ABSTRACT FROM AUTHOR]

In this work we address the issue of sustainable cities by focusing on one of their very central components: daily mobility. Indeed, if cities can be interpreted as spatial organizations allowing social interactions, the number of daily movements needed to reach this goal is continuously increasing. Therefore, improving urban accessibility merely results in increasing traffic and its negative externalities (congestion, accidents, pollution, noise, etc.), while eventually reducing the quality of life of people in the city. This is why several urban-transport policies are implemented in order to reduce individual mobility impacts while maintaining equitable access to the city. This challenge is however non-trivial and therefore we propose to investigate this issue from the complex systems point of view. The real spatial-temporal urban accessibility of citizens cannot be approximated just by focusing on space and implies taking into account the space-time activity patterns of individuals, in a more dynamic way. Thus, given the importance of local interactions in such a perspective, an agent based approach seems to be a relevant solution. This kind of individual based and “interactionist” approach allows us to explore the possible impact of individual behaviors on the overall dynamics of the city but also the possible impact of global measures on individual behaviors. In this paper, we give an overview of the Miro Project and then focus on the GaMiroD model design from real data analysis to model exploration tuned by transportation-oriented scenarios. Among them, we start with the the impact of a LEZ (Low Emission Zone) in the city center.

[Display omitted] In this work, the Quasi-particle Approach (QA) is applied to qualitatively reproduce the underlying mechanisms of the displacive fcc (γ) → bcc (α) transformation. At the microstructural scale, we demonstrate that the QA is able to predict the growth of a bcc nucleus in a fcc matrix, and the eventual formation of an internally twinned structure consisting in two variants with Kurdjumov-Sachs orientation relationship. At the atomic level, the defect structure of twinning boundaries and fcc/bcc interfaces is identified, and the main mechanism for the propagation of the fcc/bcc interface is analyzed. In detail, it is confirmed that twin boundaries are propagated by the glide of pairs of partial twin dislocations, while the propagation of fcc screw dislocations along coherent terrace edges is the pivotal vector of the fcc/bcc transformation. The simulation results are compared qualitatively with our TEM and HRTEM observations of Fe-rich bcc twinned particle embedded in the fcc Cu-rich matrix in the Cu-Fe-Co system. [ABSTRACT FROM AUTHOR]

• SIMREG is a stochastic tree-level distance-independent forest model. • Model the impact of silviculture and management on forest development. • Calibrated with widely available national forest inventory data. • Generic growth, thinning and recruitment models for 17 species groups. • Current management of Walloon spruce forests is unsustainable. SIMREG is a non-deterministic tree-level distance independent forest model that can simulate forest growth, yield and management on a regional scale while representing the wide diversity of composition, structure and management found in forest stands. It is composed of several sub-models to represent the main forest dynamics (growth, recruitment, removal, clearcut and reforestation) and to account for species composition, stand density, tree size and social status, forest ownership type and some sites characteristics. We used the data collected by the permanent forest inventory of Wallonia (IPRFW) between 1994 and 2015 to calibrate SIMREG and forecast the development of Wallonia's 479 500 ha of productive forest (465 million simulated trees) until 2050. According to our simulation, the harvesting rate of Norway spruce (the main production species) is currently unsustainable and it is gradually being replaced by other species such as Douglas-fir, larch and various hardwoods. It appears that in terms of total softwood volume production, the higher production level of Douglas-fir and larch should eventually compensate for the decline in spruce. In contrast, the harvest rate in hardwood stands is around 75% of the annual yield, resulting in a steady increase in the total hardwood stock of about 600 000 m³ per year. Our methodology is easily replicable and the data required for sub-model calibration are consistent with those measured by most permanent NFIs, so our forest simulation model could be adapted to other regions and countries. [ABSTRACT FROM AUTHOR]