检索结果-内蒙古大学图书馆

PyTorchFire: A GPU-accelerated wildfire simulator with Differentiable Cellular Automata

ENVIRONMENTAL MODELLING & SOFTWARE 2025年 188卷

作者： Xia, Zeyu Cheng, Sibo Univ Virginia Charlottesville VA 22904 USA Inst Polytech Paris CEREA ENPC EDF R&D Palaiseau Ile De France France

Accurate and rapid prediction of wildfire trends is crucial for effective management and mitigation. However, the stochastic nature of fire propagation poses significant challenges in developing reliable simulators. In this paper, we introduce PyTorchFire, an open-access, PyTorch-based software that leverages GPU acceleration. With our redesigned differentiable wildfire Cellular Automata (CA) model, we achieve millisecond-level computational efficiency, significantly outperforming traditional CPU-based wildfire simulators on real-world- scale fires at high resolution. Real-time parameter calibration is made possible through gradient descent on our model, aligning simulations closely with observed wildfire behavior both temporally and spatially, thereby enhancing the realism of the simulations. Our PyTorchFire simulator, combined with real-world environmental data, demonstrates superior generalizability compared to supervised learning surrogate models. Its ability to predict and calibrate wildfire behavior in real-time ensures accuracy, stability, and efficiency. PyTorchFire has the potential to revolutionize wildfire simulation, serving as a powerful tool for wildfire prediction and management.

关键词： Wildfire simulation Differentiable Cellular Automata PyTorch-based software parallel computing techniques GPU-acceleration

来源：评论

学校读者我要写书评

暂无评论

parallel computing USING SEMIANALYTICAL FINITE-ELEMENT METHOD

引用

AIAA JOURNAL 1994年第5期32卷 1066-1071页

作者： KIM, JR KIM, WD KIM, SJ Seoul National University Seoul 151-742 Republic of Korea

In this paper, a parallel computing technique using a semianalytical finite element method and its implementation method are studied, and the speedup in the elapsed computing time is discussed. The structural analysis of an axisymmetric body with nonaxisymmetric boundary conditions is considered by using a semianalytical finite element method. The displacements and the boundary conditions are expressed by a Fourier series. By the orthogonality properties of the Fourier series, the equilibrium equations are fully decoupled for each mode. Therefore, these decoupled equations can be well adapted to the individual processors of a parallel computer. Two example problems are considered for the verification of the developed code and for the application of the parallel processing to practical problems. The computing time by parallel processing is compared with that by sequential processing, and a large amount of improvement in the elapsed computing time is achieved by parallel processing. From the computational experiments, a speedup of 4.6 times in elapsed time is observed with six processors of Alliant FX2812 at Seoul National University.

关键词： Finite Element Analysis parallel computing Environment parallel computing techniques parallel Computers Structural Analysis Abaqus Stress Distribution Cantilever Beam FORTRAN Composite Structures

来源：评论

学校读者我要写书评

暂无评论

Enhancement in optimal multiple-burn trajectory computation by switching function analysis

引用

JOURNAL OF SPACECRAFT AND ROCKETS 2007年第1期44卷 264-272页

作者： Xu, Yunjun Univ Oklahoma Norman OK 73019 USA

This paper describes an improvement in designing a real time optimal multiple-burn trajectory that should be rapid and adaptive. The computational cost for solving a multiple-burn trajectory problem comes from 1) trajectory propagation (or simulation of a trajectory) and 2) trajectory iterations (or computation of gradient information). Tremendous work has been done to reduce this cost, especially in the area of increasing the solution convergence or gradient calculation speed. However, in applications such as an orbit-to-orbit transfer, a long range payload delivery, or a medium or low thrust-to-weight ratio trajectory, the propagation time of the coast arc is the dominant one of the computational cost, and thus several studies have been focused on solving this problem by using parallel computing techniques. In this paper, through an oscillation frequency study of the switching function, the propagation of the coast are is avoided and "safely" replaced by a zero-finding program. Using this method, the speed of the real time trajectory design is increased. Numerical implementation issues of this algorithm are discussed in detail. To show the effectiveness of the proposed algorithm, four trajectory designs have been simulated: two Hohmann-type orbital transfers, a low Earth orbital insertion, and the McCue problem. Atmospheric drag and J(2) perturbations. have been considered in these simulations.

关键词： Hohmann Transfer Earth Atmospheric Drag Thrust to Weight Ratio parallel computing techniques Pontryagin's Maximum Principle Orbital Eccentricity Orbital Elements MATLAB Hamiltonian Functions

来源：评论

学校读者我要写书评

暂无评论

Implementing a Collaborative Online Course to Extend Access to HPC Skills

引用

computing IN SCIENCE & ENGINEERING 2016年第1期18卷 73-79页

作者： Gordon, Steven I. Demmel, James Destefano, Lizanne Rivera, Lorna Ohio Supercomp Ctr Columbus OH 43212 USA Univ Calif Berkeley EECS & Math Berkeley CA 94720 USA Georgia Inst Technol Psychol Atlanta GA 30332 USA Univ Illinois Chicago IL 60680 USA

Progress in science and engineering is increasingly tied to the effective use of computational modeling on high-performance computing systems. Expertise in parallel computing techniques is therefore required, given the current architecture of those systems. Many campuses don't offer such a course because of course load limits, a lack of faculty expertise, and/or lack of access to appropriate computing resources. As part of the XSEDE project, an initial MOOC-style course on parallel computing proved to be ineffective at resolving this problem. A subsequent blended online course with collaborating faculty from multiple institutions where students receive academic credit has proven to be a much more effective approach to creating a STEM workforce with the requisite parallel computing knowledge.

关键词： computer aided instruction educational courses groupware parallel processing HPC skills blended online course collaborative online course computational modeling high-performance computing systems parallel computing techniques Collaboration Computational modeling Online services parallel processing Programming profession Resource management Education HPC STEM high-performance computing scientific computing

来源：评论

学校读者我要写书评

暂无评论

Galerkin Variational Integrators for Orbit Propagation with Applications to Small Bodies

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2019年第2期42卷 347-363页

作者： Bolatti, Dante A. de Ruiter, Anton H. J. Ryerson Univ Dept Aerosp Engn Toronto ON M5B 2K3 Canada Ryerson Univ Spacecraft Dynam & Control Dept Aerosp Engn Toronto ON M5B 2K3 Canada

The development of fixed-step higher-order variational integrators based on geometric integration methods is presented. These methods are symplectic, preserve energy in Hamiltonian conservative systems, and are highly accurate for long-term integration. By properly selecting a quadrature rule and control points in the polynomial approximation of the trajectory, a fast algorithm that reduces the number of computations of the gravitational force terms is constructed. Numerical simulations are presented for different scenarios, comparing Galerkin variational integrators against traditional fixed-step propagators used in astrodynamics applications. An error analysis is then performed, focusing on the implications of energy conservation and the accuracy of the methods. It is found that, under certain conditions, Galerkin methods are superior to nonsymplectic integrators for long-term orbital propagation, reducing simulation times significantly for long integration periods. This makes the methods suitable for long-term orbital dynamics applications about small bodies, such as asteroids.

关键词： 216 Kleopatra Numerical Integration Energy Conservation Highly Elliptical Orbit Numerical Simulation parallel computing techniques Spacecraft Dynamics FORTRAN Banach Fixed Point Theorem Kepler's Equation

来源：评论

学校读者我要写书评

暂无评论

General framework for dynamic substructuring: History, review, and classification of techniques

引用

AIAA JOURNAL 2008年第5期46卷 1169-1181页

作者： de Klerk, D. Rixen, D. J. Voormeeren, S. N. Delft Univ Technol Fac Mech Maritime & Mat Engn Sect Engn Dynam NL-2628 CD Delft Netherlands Muller BBM VibroAkustik Syst D-82152 Planegg Martinsried Germany

Four decades after the development of the first dynamic substructuring techniques, there is a necessity to classify the different methods in a general framework that outlines the relations between them. In this paper, a certain vision on substructuring methods is proposed, by recalling important historical milestones that allow us to understand substructuring as a domain decomposition concept. Thereafter, based on the dual and primal assembly of substructures, a general framework for the classification of the methods is presented. This framework allows us to indicate how the various classes of methods, proposed along the years, can be derived from a clear mathematical description of substructured problems. Current bottlenecks in experimental dynamic substructuring, as well as solution;found in literature, will also be briefly discussed.

关键词： Lightly Damped Structures Finite Element Modeling Frequency Response Functions Rayleigh Ritz Approach Singular Value Decomposition Structural Dynamics Finite Element Tearing and Interconnect Mathematical Models Modal Analysis parallel computing techniques

来源：评论

学校读者我要写书评

暂无评论

Dual Quaternion-Based Powered Descent Guidance with State-Triggered Constraints

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2020年第9期43卷 1584-1599页

作者： Reynolds, Taylor P. Szmuk, Michael Malyuta, Danylo Mesbahi, Mehran Acikmese, Behcet Carson, John M., III Univ Washington Aeronaut & Astronaut Seattle WA 98195 USA NASA Johnson Space Ctr Houston TX 77058 USA

This paper presents a numerical algorithm for computing six-degree-of-freedom free-final-time powered descent guidance trajectories. The trajectory generation problem is formulated using a unit dual quaternion representation of the rigid body dynamics and several standard path constraints. Our formulation also includes a special line-of-sight constraint that is enforced only within a specified band of slant ranges relative to the landing site: a novel feature that is especially relevant to terrain and hazard relative navigation. We use the newly introduced state-triggered constraints to formulate these range constraints in a manner that is amenable to real-time implementations. The resulting nonconvex optimal control problem is solved iteratively as a sequence of convex second-order cone programs that locally approximate the nonconvex problem. Each second-order cone program is solved using a customizable interior point method solver. Also introduced are a scaling method and a new heuristic technique that guide the convergence process toward dynamic feasibility. To demonstrate the capabilities of our algorithm, two numerical case studies are presented. The first studies the effect of including a slant-range-triggered line-of-sight constraint on the resulting trajectories. The second study performs a Monte Carlo analysis to assess the algorithm's robustness to initial conditions and real-time performance.

关键词： Powered Descent Guidance Quaternions Three Degrees of Freedom Pontryagin's Maximum Principle Monte Carlo Analysis parallel computing techniques Numerical Algorithms Model Predictive Control Nonlinear Programming Sensor Technology

来源：评论

学校读者我要写书评

暂无评论

Development of a parallel Multi-dimensional Grid-based Vlasov Solver for Plasma Plume Simulation

Development of a Parallel Multi-dimensional Grid-based Vlaso...

引用

AIAA Propulsion and Energy Forum / IEEE Electric Aircraft Technologies Symposium (EATS)

作者： Cui, Chen Wang, Joseph Univ Southern Calif Dept Astronaut Engn Los Angeles CA 90089 USA

ISBN: (数字)9781624106118

ISBN: (纸本)9781624106118

We present a parallel, multi-dimensional grid-based Vlasov solver (Vlasolver) for plasma plume simulations. The Vlasov-Poisson system is solved using the third order Positive Flux Conservation (PFC) scheme and the parallelization of the code is implemented using domain decomposition and MPI. We find that the grid-based Vlasov method is advantageous over particle-based methods in the low-density plume region because it eliminates the inherent statistical noise in particle-based methods and allows us to extend the solution beyond the self-similar expansion region.

关键词： Velocity Distribution Function Poisson's Equation Advection parallel computing techniques Finite Volume Method Vortices Cartesian Coordinate System Electromagnetic Field Space Communication Electric Propulsion

来源：评论

学校读者我要写书评

暂无评论

Direct Methods Comparison for the Active Target Defense Scenario

Direct Methods Comparison for the Active Target Defense Scen...

引用

AIAA Science and Technology Forum and Exposition (AIAA SciTech)

作者： Weintraub, Isaac E. Cobb, Richard Baker, William Pachter, Meir US Air Force Res Lab Wright Patterson AFB OH 45433 USA US Air Force Res Lab Dept Aerosp Engn Wright Patterson AFB OH 45433 USA US Air Force Res Lab Dept Mathemat Wright Patterson AFB OH 45433 USA US Air Force Res Lab Dept Elect Engn Wright Patterson AFB OH 45433 USA

ISBN: (数字)9781624105951

ISBN: (纸本)9781624105951

The goal of optimal control is to obtain an admissible function which minimizes an objective functional subject to specified constraints and boundary conditions. Four popular methods of computing the optimal control directly are the Single Shooting Method, Multiple Shootings Method, Even Collocation Method, and Pseudospectral Method. In this paper, the four direct methods of computing the optimal control for the Active Target Defense Scenario are presented and compared. The Active Target Defense Scenario is a three-agent engagement where an evading agent tries to escape a pursuing agent with the assistance of a defending agent. The goal of the Evader is to maneuver in such a way as to assist the Defender in capturing the Pursuer while maximizing his range from the Pursuer. This work begins by describing each direct method and summarizing advantages and disadvantages of each. Next, each method is used to solve the optimal control for the Active Target Defense Scenario;results are presented and a comparison of the performance of each method is made. Finally, concluding remarks describing the results, performance, and real-time implementation are presented.

关键词： Optimal Control Theory Gaussian Quadrature Pseudospectral Methods parallel computing techniques Nonlinear Programming Sequential Quadratic Programming Heading Angle Proportional Navigation Runge Kutta Methods MATLAB

来源：评论

学校读者我要写书评

暂无评论

Optimizing energy expenditure in agricultural autonomous ground vehicles through a GPU-accelerated particle swarm optimization-artificial neural network framework

引用

Cleaner Energy Systems 2024年 9卷

作者： Ambuj Machavaram, Rajendra Indian Institute of Technology Kharagpur

The accurate energy consumption prediction in Agricultural Ground Vehicles (AGVs) holds immense potential for optimizing operational efficiency and minimizing environmental impact. However, existing optimization methods for such prediction tasks often suffer from high computational demands, hindering their practical implementation. This paper introduces a ground-breaking approach that overcomes this limitation by leveraging the potent computational power of Graphics Processing Units (GPUs) to accelerate the optimization process dramatically. We propose a novel adaptation of the Particle Swarm Optimization (PSO) algorithm, specifically tailored to the intricate multi-objective challenges of AGV energy prediction. This framework harnesses the strengths of a multi-objective approach, enabling the simultaneous optimization of prediction accuracy and model complexity. To further enhance efficiency, we seamlessly integrate GPU parallelization techniques, significantly expediting both the optimization process and the training of Artificial Neural Networks (ANNs) employed for prediction. Preliminary results demonstrate a remarkable improvement in the accuracy of AGV energy consumption predictions, directly attributed to the synergistic effect of optimizing the ANN architecture and parameters through our proposed PSO framework. This tailored PSO adaptation distinguishes itself by its ability to tackle the complex multi-objective nature of AGV energy prediction with enhanced efficiency and precision. It thus emerges as a compelling and novel solution within the realm of Machine Learning and heuristic methods for agricultural robotics, paving the way for sustainable and optimal AGV operations. © 2024

关键词： Agricultural AGVs GPU-accelerated computing parallel computing techniques Pareto optimization PSO- ANN framework

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：