Tutorials
The role of the tutorials is to provide a platform for a more intensive scientific exchange amongst researchers interested in a particular topic and as a meeting point for the community. Tutorials complement the depth-oriented technical sessions by providing participants with broad overviews of emerging fields. A tutorial can be scheduled for 1.5 or 3 hours.
Tutorial on
Functional Mock-Up Interface (FMI): Beginner’s Tutorial
Instructors
|
Christian Bertsch
Robert Bosch GmbH
Germany
|
|
|
|
Brief Bio
Christian Bertsch studied Mathematics and Physics at Heidelberg University before joining Robert Bosch GmbH in 2021 as a simulation engineer. Today he is a project manager at Bosch Research, and since 2022 project leader of the FMI project within the Modelica Association.
|
|
Maurizio Palmieri
University of Pisa
Italy
|
|
|
|
Brief Bio
Maurizio Palmieri is an assistant professor at Pisa University working on the co-simulation of platoon vehicles and machine-learning approaches for well-being. He got his PhD in Smart Computing, within a PhD program jointly offered from the Universities of Pisa, Florence, and Siena, with a thesis named "Co-simulation and verification of Cyber-Physical Systems using logic models".
|
|
Cláudio Ângelo Gonçalves Gomes
Aarhus University
Denmark
|
|
|
|
Brief Bio
Cláudio Ângelo Gonçalves Gomes is an assistant professor at Aarhus University focusing on co-simulation and its combination with numerical analysis, formal methods, and applied mathematics. He got his PhD from the University of Antwerp on the foundations for co-simulation.
|
Abstract
In this tutorial, organized by the FMI project, the basic concepts of the FMI standard are introduced and demonstrated. The first part consists of a presentation of the motivation, history, the FMI project, the basic technical idea, the different FMI versions, limitations and the current state of tool support. In the second part, hands-on exercises are given to gain first experience with creating, checking, coupling and simulating FMUs in different open source and commercial tools. This second part can also viewed as a demonstration where the exercises are done by the presenters, if the software tools are not available. At the end, an outlook is given on further material: other resources such as tutorial modules for more advanced usage of FMI.
Learning objectives:
After the tutorial, the attendees will be able to:
- Understand the motivation and history of the FMI project.
- Know how to export, import, and interact with, an FMU (a simulation model packaged according to the FMI standard), and will have a basic understanding of how a co-simulation is run.
- Have a basic grasp of the main pitfalls when interacting with FMUs.
- Have a collection of reference materials and tools that are useful when developing
Audience:
We target industrial practitioners and young researchers who have limited programming knowledge.Prerequisite knowledge of participants: while not mandatory, the following skills may make the experience better:
- Knowledge of command line software
- Limited experience with the python programming language
- Limited knowledge of jupyter notebooks
- Limited knowledge of how a drivetrain works
Tutorial on
AnyLogic: Multimethod Modelling and Simulation Platform
Instructors
|
Vladimir Koltchanov
AnyLogic
France
|
|
|
|
Brief Bio
Vladimir Koltchanov. PhD in computer sciences (Complex Systems Modelling and Simulation) from Kiev Cybernetic Institute. Chef of research team in dynamic simulation in Kiev, consultant in dynamics system’s modelling and simulation in Paris. Actual managing director of AnyLogic Europe company based in Paris Region.
|
|
Roberto Revetria
University of Genoa
Italy
|
|
|
|
Brief Bio
Roberto Revetria gets the degree with honors in 1998 and in July 1999 he is appointed Lieutenant in the Naval Army Corps. From July 1998 to July 1999 he is attached to the Control and Testing Service Head Office of the Military Maritime Arsenal of Taranto. In October 2001 received the title of PHD in Parma Univ.. On Nov. 2001 he wins the Competition for the role of Researcher at the Univ. of Genoa, he than win a second competition for the role of Associated Professor at Parma Univ., in March 2004. The personal experience achieved in the construction of computer-aided analysis models and tools brought him to develop complex software using multipurpose (C/C++, Java, Tcl/Tk, Python, Php, Fortran, Pascal, VBA) and dedicated (Automod, Simul8, ESL, GPSS/H, Arena, Witness, Plant Simulation, Powersim, iThink/Stella, VenSim, Berkeley Madonna) simulation tools. He makes use and develops models in virtual reality. On July 2010 he wins the Competition for the role of Full Professor at the Milan Polytechnic. He is currently Deputy-Director of the Italian Center of Excellence in Logistics and Transportation (CIELI). Since 2008 he is Special Appointed Consultant for the Holy See (Vatican City).
|
Abstract
Most simulation software tools for business applications are single-method (Discrete, Event Simulation DES, Agent – Based Modelling ABM and System Dynamics SD) and each of them serves a particular level of abstraction. Most real-world business systems are very complex. The ability to capture business systems with their real complexity and interactions can be seriously limited by using only one modelling method. So, single-method software tools are not ideal for handling complicated tasks. With multimethod software tools you can build models at any abstraction level you need and find a solution for almost any problem. With multimethod modelling software tools, you can create efficient and manageable models without using workarounds. You can choose the best representation of your system, not limited by methods or tools. Using multimethod simulation software gives you the flexibility needed to successfully solve any problem without excluding important elements or having to develop workarounds. With multimethod modelling, you can always choose the most efficient combination of approaches and develop the best solutions for your problems. On this tutorial session we’ll show how to build the simulation model based on three major methodologies commonly used for building dynamic business simulation models: discrete event modelling, agent-based modelling, and system dynamics. The most important, we’ll show how it is easy with AnyLogic platform combine these tree approach in the same model. We’ll illustrate multimethod modelling approach by examples of real business projects.
Aims and learning objectives:
The tutorial will demonstrate how to build a model of a highly automated picking process in which the basic layout will be build in a parametric way starting from an existing modular layout. The model will use both discrete event modelling with the help of the built-in Material Handling Library, human/vehicle interaction using Agent Based modelling and the help of the built-in Pedestrian Library and external database integration using API REST. The overall implementation will be done using minimal or no coding. The model will be deployed in the AnyLogic Cloud and accessed from mobile devices.
Prerequisite knowledge of participants:
Basic material handling and internal logistics principles, basic java programming, principles of modelling and simulation are a plus.
Tutorial on
Equation-Based Object-Oriented Modelling Using Modelica: An Introduction
Instructor
|
Hans Vangheluwe
University of Antwerp
Belgium
|
|
|
|
Brief Bio
Hans Vangheluwe is a Professor in the Antwerp Systems and Software Modelling (AnSyMo) group within the Department of Mathematics and Computer Science at the University of Antwerp in Belgium, where he is a founding member of the NEXOR Consortium on Cyber-Physical Systems (CPS). AnSyMo is a Core Research Lab of Flanders Make, the strategic research centre for the Flemish manufacturing industry. He heads the Modelling, Simulation and Design Lab (MSDL), distributed over the University of Antwerp and McGill University in Montreal, Canada.
His fundamental work covers the foundations of modelling and simulation, of model management, model transformation, and domain-specific (visual) modelling environments. This work is always accompanied by prototype tools such as PythonPDEVS, the Modelverse, T-Core, AToM3 and AToMPM.
|
Abstract
Abstract
Modelica https://www.modelica.org/ is an Equation-based Object-Oriented Language (EOOL) for hierarchical, lumped-parameter modelling of physical systems. Its Object-Oriented nature makes it highly suited for the modular design of re-usable components such as those found in the Modelica Standard Library (MSL), covering a variety of physical domains. Thanks to Modelica's mathematical semantics, models from different physical domains can be elegantly combined by simply connecting power ports (to denote energy exchange). Modelica models are (computationally) a-causal which means that constituent laws of physics can be written in their natural mathematical form: Differential Algebraic constraint Equations over state variables (functions of time). A Modelica compiler then assigns computational causality, performs symbolic simplifications and generates efficient simulation code, if needed, conforming to the Functional Mockup Interface (FMI) standard, to be used in co-simulation. Modelica also allows for the expression of discontinuities, based on an underlying notion of superdense time. As a result, hybrid (continuous-discrete) models are supported. These hybrid models are typically time-scale or parameter abstractions of the continuous dynamics of physical systems. Modelica is also an Object-Oriented programming language for modelling computational logic. The hybrid constructs (in particular, the temporal "when" operator), combined with the programming constructs, allow modelling of software and network components at a discrete-event (or discrete-time) level of abstraction. These can be seemlessly combined with models of the physics. It if for this reason that Modelica is often the language of choice for modelling and simulating Cyber-Physical Systems. Modelica also has basic support for dynamic structure models. EOOLs (such as Modelica, Simscape, Maplesim, and Ecosim Pro) will first be put in context, placing them between Bond Graphs and computationally causal languages such as Simulink. The language syntax and semantics will then be introduced incrementally, through a small example in the Electrical domain. The OpenModelica (https://openmodelica.org/) open source tool suite will be used.