Agent-based simulations, adaptive algorithms and solvers (ABS-AAS) Session 1

Time and Date: 10:35 - 12:15 on 12th June 2017

Room: HG D 7.1

Chair: Maciej Paszynski

-3 ICCS 2017 Workshop on Agent-Based Simulations, Adaptive Algorithms and Solvers [abstract]
Abstract: [No abstract available]
Aleksander Byrski, Maciej Paszynski, Robert Schaefer, Victor Calo and David Pardo
192 Quadrature blending for isogeometric analysis [abstract]
Abstract: We use blended quadrature rules to reduce the phase error of isogeometric analysis discretizations. To explain the observed behavior and quantify the approximation errors, we use the generalized Pythagorean eigenvalue error theorem to account for quadrature errors on the resulting weak forms. The proposed blended techniques improve the spectral accuracy of isogeometric analysis on uniform and non-uniform meshes for different polynomial orders and continuity of the basis functions. The convergence rate of the optimally blended schemes is increased by two orders with respect to the case when standard quadratures are applied. Our technique can be applied to arbitrary high-order isogeometric elements.
Victor Calo, Quanling Deng and Vladimir Puzyrev
81 Optimally refined isogeometric analysis [abstract]
Abstract: Performance of direct solvers strongly depends upon the employed discretization method. In particular, it is possible to improve the performance of Isogeometric Analysis (IGA) discretizations by introducing multiple $C^0$-continuity hyperplanes that act as separators during LU factorization [7]. In here, we further explore this venue by introducing separators of arbitrary continuity. Moreover, we develop an efficient method to obtain optimal discretizations in the sense that they minimize the time employed by the direct solver of linear equations. The search space consists of all possible discretizations obtained by enriching a given IGA mesh. Thus, the best approximation error is always reduced with respect to its IGA counterpart, while the solution time is decreased by up to a factor of 60.
Daniel Garcia, Michael Barton and David Pardo
538 Higher-Order Finite Element Electromagnetics Code for HPC environments [abstract]
Abstract: In this communication, an electromagnetic software suite developed to work in high performance computing (HPC) environments is presented. Details about the formulation used are provided, and an exhaustive flowchart is included and analyzed. Finally, results using HPC environments are shown.
Adrian Amor-Martin, Daniel Garcia-Donoro and Luis E. Garcia-Castillo
270 Coupled isogeometric Finite Element Method and Hierarchical Genetic Strategy with balanced accuracy for solving optimization inverse problem [abstract]
Abstract: The liquid fossil fuel reservoir exploitation problem (LFFEP) has not only economical signification but also strong natural environment impact. When the hydraulic fracturing technique is considered from the mathematical point of view it can be formulated as an optimization inverse problem, where we try to find optimal locations of pumps and sinks to maximize the amount of the oil extracted and to minimize the contamination of the groundwater. In the paper, we present combined solver consisting of the Hierarchical Genetic Strategy (HGS) with variable accuracy for solving optimization problem and isogeometric finite element method (IGA-FEM) with different mesh size for modeling a non-stationary flow of the non-linear fluid in heterogeneous media. The algorithm was tested and compared with the strategy using Simple Genetic Algorithm (SGA) as optimization algorithm and the same IGA-FEM solver for solving a direct problem. Additionally, a parallel algorithm for non-stationary simulations with isogeometric L2 projections is discussed and preliminarily assessed for reducing the computational cost of the solvers consisting of genetic algorithm and IGA-FEM algorithm. The theoretical asymptotic analysis which shows the correctness of algorithm and allows to estimate computational costs of the strategy is also presented.
Barbara Barabasz, Marcin Łoś, Maciej Woźniak, Leszek Siwik and Stephen Barrett

Agent-based simulations, adaptive algorithms and solvers (ABS-AAS) Session 2

Time and Date: 15:45 - 17:25 on 12th June 2017

Room: HG D 7.1

Chair: Quenling Deng

156 A wrapper around parallel MUMPS solver to reduce its memory usage and execution time for finite element method computations [abstract]
Abstract: In this paper, we present a wrapper around MUMPS solver, called Hierarchical Solver Wrapper (HSW) that allows reducing its memory usage and execution time. The wrapper is dedicated for domain decomposition based parallel finite element method computations on Linux cluster. It utilizes the identical interface as parallel MUMPS with entries provided as lists of non-zero entries distributed among multiple processors. It wraps around MUMPS solver, performing two level hierarchical calls. First, it calls multiple sequential MUMPS solvers, one per each processor. They compute Schur complements of the interior nodes on the interface nodes over each subdomain. It deallocates the sequential MUMPS solvers on every processor to minimize the memory usage. It keeps only original lists of non-zero entries as well as lists of non-zero entries obtained from the computed Schur complement. It submits all the Schur complements to one parallel MUMPS solver and asks for the skeleton problem solution. Later, it restores the upper part of the local matrix to perform local backward substitutions with Schur complements replaced by the identity matrices and right-hand-sides superseded by the local part of the global solution. The wrapper is tested with latest version of MUMPS over the three-dimensional isogeometric analysis computations. It enables for reduction of the memory usage and also the execution time, in comparison with a single parallel MUMPS call with distributed lists of non-zero entries.
Maciej Paszynski and Antonio Tadeu Gomes
48 Non-Fitting meshes for Maxwell's equations [abstract]
Abstract: In context of adaptive finite element methods, we explore the possibility of employing lowest-order non-fitting meshes for solving Maxwell´s equations. In a fitting mesh, the physical interfaces of the propagation medium must be aligned with cell faces. On the other hand, this constraint is removed for non-fitting meshes, so that a larger choice of cells is possible. As a result, non-fitting meshes can simplify the implementation and/or reduce the computational cost of the associated finite-element method. Unfortunately, non-fitting meshes can also cause an accuracy loss. Indeed, the solution of Maxwell's equations can become singular inside mesh cells due to material discontinuities. In this work, we carefully analyze the accuracy loss due to the use of non-fitting meshes. We derive error estimates that are further confirmed via numerical experiments. The main conclusion is that electric and magnetic fields exhibit different convergence rates in the L^2-norm for the case of non-fitting meshes. In particular, if the user is interested in the magnetic field, non-fitting meshes produce solutions of similar quality to those obtained using fitting meshes.
Théophile Chaumont-Frelet and David Pardo
51 Fast Isogeometric L2 Projection Solver for Tumor Growth Simulations [abstract]
Abstract: In this talk, we focus on the application of the isogeometric analysis for tumor growth simulations. We present an application of the fast algorithm for isogeometric L2 projections for simulations of the tumor growth. We utilize the of PDE describing the model of the melanoma growth, including tumor cell density, flux, pressure, extracellular and degraded extracellular matrices, previously used in finite difference simulations. We also use a discrete model of the vasculature that provides an oxygen source to the system. The system is solved using explicit scheme and fast isogeometric L2 projections, utilizing the alternating directions solver. Every 10-time steps of the simulation we couple our continuous model with the discrete vasculature model. The presentation is concluded with the two-dimensional numerical results. We show that explicit dynamics simulation needs around 30,000 times steps of the alternating direction solver to be performed in two-dimensions.
Maciej Paszynski, Witold Dzwinel, Marcin Łoś and Adrian Kłusek
86 Goal-Oriented p-Adaptivity using Unconventional Error Representations for a 1D Steady State Convection-Diffusion Problem [abstract]
Abstract: This work proposes the use of an alternative error representation for Goal-Oriented Adaptivity (GOA) in context of steady state convection dominated diffusion problems. It introduces an arbitrary operator for the computation of the error of an alternative adjoint problem. From the new representation, we derive element-wise estimators to drive the adaptive algorithm. The method is applied to a one dimensional (1D) steady state convection dominated diffusion problem with homogeneous Dirichlet boundary conditions. This problem exhibits a boundary layer that produces a loss of numerical stability. The new error representation delivers sharper error bounds. When applied to a p-GOA Finite Element Method (FEM), the alternative error representation captures earlier the boundary layer, despite the existing spurious numerical oscillations.
Vincent Darrigrand, Ángel Rodríguez-Rozas, David Pardo and Ignacio Muga
224 Algorithms for construction of Element Partition Trees for Direct Solver executed over h refined grids with B-splines and C0 separators [abstract]
Abstract: We propose a way of performing isogeometric finite element method (IGA-FEM) computations over h refined grids with B-spline basis functions. Namely, we propose to use the B-spline basis functions defined over patches of elements with C0 separators between the refinement levels. Our definition of the B-splines and C0 separators allows introduction of arbitrary order B-splines over 2D grids refined towards singularities. We also present an algorithm for construction of element partition trees (EPT) over h refined grids with modified B-splines. The EPT allows generating an ordering which gives a linear computational cost of the multi-frontal solver over 2D grids refined towards a point or an edge singularity. We present the algorithm for transforming the EPT into an ordering. We also verify the linear computational cost of the proposed method on grid with point and edge singularity. We compare our method to h-adaptive finite element method (h-FEM) computations with Lagrange polynomials.
Bartosz Janota and Maciej Paszynski

Agent-based simulations, adaptive algorithms and solvers (ABS-AAS) Session 3

Time and Date: 10:15 - 11:55 on 13th June 2017

Room: HG D 7.1

Chair: Maciej Paszynski

136 Memetic approach for irremediable ill-conditioned parametric inverse problems [abstract]
Abstract: The paper introduces a new taxonomy of ill-posed parametric inverse problems, formulated as global optimization ones. It systematizes irremediable problems, which appear quite often in real life but cannot be solved using the regularization method. The paper shows also a new way of solving irremediable inverse problems by a complex memetic approach including genetic computation with adoptive accuracy, random sample clustering and a sophisticated local approximation of misfit plateau region. Finally, we use a benchmark function featuring cross-shaped plateau to discuss some factors that influence the quality of plateau shape approximation.
Marcin Łoś, Jakub Sawicki, Maciej Smołka and Robert Schaefer
444 Toward hybrid platform for evolutionary computations of hard discrete problems [abstract]
Abstract: Memetic agent-based paradigm, which combines evolutionary computation and local search techniques in one of promising meta-heuristics for solving large and hard discrete problem such as Low Autocorrellation Binary Sequence (LABS) or optimal Golomb-ruler (OGR). In the paper as a follow-up of the previous research, a short concept of hybrid agent-based evolutionary systems platform, which spreads computations among CPU and GPU is shortly introduced. The main part of the paper presents an efficient parallel GPU implementation of LABS local optimization strategy. As a means for comparison, speed-up between GPU implementation and CPU sequential and parallel versions are shown. This constitutes a promising step toward building hybrid platform that combines evolutionary meta-heuristics with highly efficient local optimization of chosen discrete problems.
Dominik Żurek, Kamil Piętak, Marcin Pietroń and Marek Kisiel-Dorohinicki
335 The versatility of an entropy inequality for the robust computation of convection dominated problems [abstract]
Abstract: We present a discrete entropy inequality that exhibits versatile uses in convection dominated problems. Much like the thermodynamic entropy inequality, the sign of this so-called discrete entropy production allows us to determine unphysical regions in the numerical solution without any a-priori knowledge of the solution. Further, the sign of the discrete production also functions as an excellent indicator for mesh adaptation in convection-diffusion and other singular perturbation problems. We also show preliminary results for how the operator can be used to derive robust schemes for convection dominated problems. All the above applications i.e. (a) Detecting unphysical numerical behavior, (b) mesh adaptation and (c) stabilization, are robust in that they are achieved without any ad-hoc, user introduced, parameters making the applications robust. We show a range of numerical results that exhibit the effcacy of the operator.
Balaji Srinivasan and Vivek Kumar
282 Agent-based Decision Support System for Technology Recommendation [abstract]
Abstract: This paper presents an idea of a multi-agent decision support system. Agent-based technology allows for decentralized problem solving and creating complex decision support systems, mixing various processing techniques, such as simulation, reasoning and machine learning and allows for distributed knowledge. Our main contribution is an agent-based architecture for decision support systems which is an agent-based implementation of a labeled deductive system. Such approach allows to decompose an inference algorithm into separate modules and distribute knowledge base into parts. The system is tested on a domain of material choice support for casting.
Grzegorz Legien, Bartlomiej Sniezynski, Dorota Wilk-Kołodziejczyk, Stanisława Kluska-Nawarecka, Edward Nawarecki and Krzysztof Jaskowiec

Agent-based simulations, adaptive algorithms and solvers (ABS-AAS) Session 4

Time and Date: 14:10 - 15:50 on 13th June 2017

Room: HG D 7.1

Chair: Maciej Woźniak

427 Agent-based Evolutionary and Memetic Black-box Discrete Optimization [abstract]
Abstract: Hybridizing agent-based paradigm with evolutionary or memetic computation can enhance the field of meta-heuristics in a significant way, giving to usually passive individuals autonomy and capabilities of perception and interaction with other ones. In the article, an evolutionary multi-agent system (EMAS) is applied to solve difficult discrete benchmark problems without any domain-specific knowledge---thus they may be called ``black-box'' ones. As a means for comparison, a parallel evolutionary algorithm (constructed along with Michalewicz model) versus evolutionary and memetic versions of EMAS are used. The obtained results point out that EMAS is significantly more efficient than classical evolutionary algorithms and also finds better results in the examined problem instances.
Michal Kowol, Kamil Piętak, Marek Kisiel-Dorohinicki and Aleksander Byrski
121 Multi-agent large-scale parallel crowd simulation [abstract]
Abstract: This paper presents design, implementation and performance results of a new modular, parallel, agent-based and large scale crowd simulation environment. A parallel application, implemented with C and MPI, was implemented and run in this parallel environment for simulation and visualization of an evacuation scenario at Gdansk University of Technology, Poland and further in the area of districts of Gdansk. The application uses a parallel MPI I/O run on two different clusters and a two or three node Parallel File System (PFS) to store a current state in a file. In order to make this implementation efficient, we used our previously developed and tuned Byte-addressable Non-volatile RAM Distributed Cache - a solution that allows to access small data chunks from spread locations within a file efficiently. We have presented application execution times versus the number of agents (up to 100000), versus the number of simulation iterations (up to 25000), versus map size (up to 6km^2) and versus the number of processes (up to more than 650) showing high speed-ups.
Artur Malinowski, Pawel Czarnul, Krzysztof Czurylo, Maciej Maciejewski and Pawel Skowron
355 On the performance and scalability of an HPC enhanced Multi Agent System based evacuation simulator [abstract]
Abstract: This paper presents some of the techniques, algorithms and designs used to enable mass evacuation simulations to take advantage of high performance computing infrastructure. A brief overview of a tsunami mass evacuation simulator capable of simulating urban areas of hundreds of km2 in sub-meter detail is provided. Enhancements to the serial algorithms for path finding reducing the path finding time in 94% and a cache friendly visual boundary extraction algorithm cutting the overall simulation time in 50% are presented. Furthermore the hybrid parallel (distributed memory (MPI) + shared memory (OpenMP)) framework is described. A dynamic load balancing technique reducing the idling time from 50% of the execution time to 3% is presented. Finally measures of the thread parallel strong scalability up to 16 threads of 82.69% and distributed process strong scalability up to 2048 processes of 75.93% are presented.
Leonel Enrique Aguilar Melgar, Maddegedara Lalith, Tsuyoshi Ichimura and Muneo Hori
241 Lightweight Volunteer Computing Platform using Web Workers [abstract]
Abstract: Volunteer computing is a very appealing way of utilizing vast available resources in an efficient way. However currently available platforms supporting this computing style are either difficult to use or not available at all, being the results of e.g. finished scientific projects. In this paper a novel, lightweight volunteer computing platform is presented. In order to contribute the resources to this platform, only a web-browser is required without the need to install any additional plugins or other software. In this paper, besides general considerations and presentation of the platform structure and functionalities, selected results proving its efficiency are shown.
Pawel Chorazyk, Mateusz Godzik, Kamil Piętak, Wojciech Turek, Marek Kisiel-Dorohinicki and Aleksander Byrski
316 Using timescale realignment to construct and validate agent-based escape simulations [abstract]
Abstract: Agent-based modelling (ABM) is a widely recognized simulation technique that is particularly relevant for economic, environmental, security and humanitarian scenarios [1,2]. Within this work we focus particularly on escape scenarios. On this topic, ABMs have been extensively applied to model events such as fire escapes [2-4], evacuations from large scale disasters [5-7], or the movements of refugees in times of conflict [8-9]. Unfortunately, many of these ABM simulation and model validation studies are constrained by limitations in the available empirical data. Most commonly, there is insufficient empirical information to validate ABMs in any complete sense [10], a phenomenon which is further exacerbated by ABMs containing a relatively rich set of adjustable parameters. In some cases, these limitations are so extreme that they hinder the construction of these models. For example, in the case of refugee modelling, people are only registered by the UNHCR once they have reached a safe location (i.e., a refugee camp, see data.unhcr.org), yet the modellers wish to know how many people reside in the area of danger, and at what times. Similar mismatches occur in simulations of fire escape or disaster relief, when registrations and recordings are only made in safe locations. Using the registration data from safe locations directly to determine the number of people in areas of danger results in a structural underestimation of the number of occupants in safe locations, as the agents in the simulation require time to travel from the area of danger to a safe haven [9]. One common method to reduce these kind of errors is by permitting a warmup time in the simulation (see e.g., [11]). However, in escape scenarios such warmup times only eliminate these errors altogether if the time required for agents to arrive at the safe location is constant across all agents, and known in advance. In this talk I will present a timescale realignment technique that allows researchers to use safe location data as an input parameter for refugee escape simulations. Using this technique it is possible to fully eliminate the validation error in total safe location registrations, given that the number of safe location registrations remains positive, and at the expense of introducing an additional error in travel speeds. References: [1] J Doyne Farmer and Duncan Foley. The economy needs agent-based mod- elling. Nature, 460(7256):685–686, 2009. [2] Eric Bonabeau. Agent-based modeling: Methods and techniques for sim- ulating human systems. Proceedings of the National Academy of Sciences, 99(suppl 3):7280–7287, 2002. [3] Timo Korhonen, Simo Hostikka, Simo Heli ̈ovaara, and Harri Ehtamo. Fds+ evac: an agent based fire evacuation model. In Pedestrian and Evacuation Dynamics 2008, pages 109–120. Springer, 2010. [4] Fangqin Tang and Aizhu Ren. Agent-based evacuation model incorpo- rating fire scene and building geometry. Tsinghua Science & Technology, 13(5):708–714, 2008. [5] Xuwei Chen, John W Meaker, and F Benjamin Zhan. Agent-based mod- eling and analysis of hurricane evacuation procedures for the florida keys. Natural Hazards, 38(3):321–338, 2006. [6] Jianyong Shi, Aizhu Ren, and Chi Chen. Agent-based evacuation model of large public buildings under fire conditions. Automation in Construction, 18(3):338–347, 2009. [7] AS Mordvintsev, VV Krzhizhanovskaya, MH Lees, and PMA Sloot. Sim- ulation of city evacuation coupled to flood dynamics. In Pedestrian and Evacuation Dynamics 2012, pages 485–499. Springer International Pub- lishing, 2014. [8] J. A. Sokolowski, C. M. Banks, and R. L. Hayes. Modeling population displacement in the syrian city of aleppo. In Proceedings of the 2014 Winter Simulation Conference, pages 252–263. IEEE Press, 2014. [9] D. Groen. Simulating refugee movements: Where would you go? Procedia Computer Science, 80:2251–2255, 2016. [10] Andrew Crooks, Christian Castle, and Michael Batty. Key challenges in agent-based modelling for geo-spatial simulation. Computers, Environment and Urban Systems, 32(6):417–430, 2008. [11] Marek Laskowski and Shamir Mukhi. Agent-based simulation of emer- gency departments with patient diversion. In International Conference on Electronic Healthcare, pages 25–37. Springer, 2008.
Derek Groen

Agent-based simulations, adaptive algorithms and solvers (ABS-AAS) Session 5

Time and Date: 16:20 - 18:00 on 13th June 2017

Room: HG D 7.1

Chair: Kamil Piętak

472 Declarative Representation and Solution of Vehicle Routing with Pickup and Delivery Problem [abstract]
Abstract: Recently we have proposed a multi-agent system that provides an intelligent logistics brokerage service focusing on the transport activity for the efficient allocation of transport resources (vehicles or trucks) to the transport applications. The freight broker agent has a major role to coordinate transportation arrangements of transport customers (usually shippers and consignees) with transport resource providers or carriers, following the freight broker business model. We focus on the fundamental function of this business that aims to find available trucks and to define their feasible routes for transporting requested customer loads. The main contribution of this paper is on formulating our scheduling problem as a special type of vehicle routing with pickup and delivery problem. We propose a new set partitioning model of our specific problem. Vehicle routes are defined on the graph of cities, rather than on the graph of customer orders, as typically proposed by set partitioning formulations. This approach is particularly useful when a large number of customer orders sharing a significantly lower number of pickup and delivery points must be scheduled. Our achievement is the declarative representation and solution of the model using ECLiPSe state-of-the-art constraint logic programming system.
Amelia Badica, Costin Badica, Florin Leon and Lucian Luncean
153 A multi-world agent-based model working at several spatial and temporal scales for simulating complex geographic systems [abstract]
Abstract: Interest in the modelling and simulation of complex systems with processes occurring at several spatial and temporal scales is increasing, particularly in biological, historical and geographic studies. In this multi-scale modelling study, we propose a generic model to account for processes operating at several scales. In this approach, a ‘world’ corresponds to a complete and self-sufficient submodel with its own places, agents, spatial resolution and temporal scale. Represented worlds can be nested: a world (with new scales) may have a greater level of detail than the model at the next level up, making it possible to study phenomena with greater precision. This process can be reiterated, to create many additional scales, with no formal limit. Worlds’ simulations can be triggered simultaneously or in cascade. Within a world, agents can choose destinations in other worlds, to which they can travel using routes and inter-world ‘gates’. Once they arrive in a destination world, the agents ‘fit’ the new scale. An agent in a given world can also perceive and interact with other agents, regardless of the world to which they belong, provided they are encompassed by its perception disc. We present an application of this model to the issue of the spread of black rats by means of commercial transportation in Senegal (West Africa).
Pape Adama Mboup, Karim Konaté and Jean Le Fur
466 Role of Behavioral Heterogeneity in Aggregate Financial Market Behavior: An Agent-Based Approach [abstract]
Abstract: In this paper, an agent-based model of stock market is proposed to study the effects of cognitive processes and behaviors of the traders (e.g. decision-making, interpretation of public information and learning) on the emergent phenomena of financial markets. In financial markets, psychology and sociology of the traders play a critical role in giving rise to unique and unexpected (emergent) macroscopic properties. This study suggests that local interactions, rational and irrational decision-making approaches and heterogeneity, which has been incorporated into different aspects of agent design, are among the key elements in modeling financial markets. When heterogeneity of the strategies used by the agents increases, volatility clustering and excess kurtosis arises in the model, which is in agreement with real market fluctuations. To evaluate the effectiveness and validity of the approach, a series of statistical analysis was conducted to test the artificial data with respect to a benchmark provided by the Bank of America (BAC) stock over a sufficiently long period of time. The results revealed that the model was able to reproduce and explain some of the most important stylized facts observed in actual financial time series and was consistent with empirical observations.
Yasaman Kamyab Hessary and Mirsad Hadzikadic
233 A case based reasoning based multi-agent system for the reactive container stacking in seaport terminals [abstract]
Abstract: With the continuous development of seaports, problems related to the storage of containers in terminals have emerged. Unfortunately, existing systems suffer limitations related to the distributed monitoring and control, real-time stacking strategies efficiency and their ability to handle dangerous containers. In this paper, we suggest a multi-agent architecture based on a set of knowledge models and learning mechanisms for disturbance and reactive decision making management. The suggested system is able to capture, store and reuse knowledge in order to detect disturbances and select the most appropriate container location by using a Case Based Reasoning (CBR) approach. The proposed system takes into account the storage of dangerous containers and combines Multi-Agent Systems (MAS) and case based reasoning to handle different types of containers.
Ines Rekik, Sabeur Elkosantini and Habib Chabchoub