Non-residential buildings are accountable for 11% of global energy-related CO2 emissions (United Nations Environment Programme 2018). To increase the performance in this sector, Building Energy Performance Simulation (BEPS) is one feasible approach. Therefore, there is need for reliable and fast simulation models. One feasible approach are so called Reduced Order Models (ROMs). Thus in this paper, a comparison between the results of the established BEPS tool EnergyPlus and a ROM in Modelica with a reduced number of resistances and capacities is applied at the use case of a non-residential building. A self-developed toolchain was used to create equal models for ROM and EnergyPlus based on the same Building Information Modeling (BIM) model. The comparison shows that the reduced model deviates by 10%in annual heating and cooling. To increase accuracy and decrease computational effort the zoning strategy of non-residential buildings is investigated. The investigation shows that using a suitable zoning approach can reduce the computational effort by up to 97 %.
This paper introduces the last evolution of Daccosim cosimulation environment, with Daccosim NG developed in 2018.Main features of Daccosim NG are described: enhanced Graphic User Interface and Command-Line Interface, algorithm and mechanism of co-simulation, coexecution shell, software architecture designed for both centralised and distributed architectures, aggregation of a co-simulation graph into a Matryoshka FMU, and declarative language to design large scale co-simulation graphs.A new industrial use case in simulation of energetic systems is also introduced, and first performances of Daccosim NG on multi-core architectures are analysed.
The goal is to have a flexible platform based on open standards that can be used to efficiently solve current and future engineering problems for distributed energy sources and storage systems. By basing it on a unified architecture, collaboration and efficient work flows are enabled.
Exploiting several precursor papers on specific railway modeling topics, their models have been collected and reorganized in order to propose a sound modeling framework dedicated to railway dynamics.
This paper presents a concept how a hybrid DAE of a vapor compression cycle can be initialized in steady state using the homotopy method.A simplified equation system for a vapor compression cycle is described and its computational causality explained.It is discussed how additional boundary conditions can be applied to the simplified equation system, which do not apply to the actual equation system.The robustness and CPU time for different cases is examined and discussed based on transition plots.
A method that designers can intuitively and easily design a bush is proposed. It combines the results of the developed bushing design tool with a Modelica system model to evaluate NVH, R&H, and durability performance. Designers can create a new bushing using the bushing design tool and check the vehicle performance at an early stage of development process. The developed bushing design tool allows designers to evaluate vehicle performance by reflecting bushing characteristics without actual products.
In this paper a multi-physics model of a proton exchange membrane electrolyzer with selectable physics submodels is developed in Modelica R .It will be included in the open-source TransiEnt Library for future studies on the efficiency of energy storage for intermittent renewable sources and the coupling of power, gas, and heat grids.The model is derived almost explicitly from a previous research paper by (Espinosa-López et al., 2018) but uses different models for cooling system power and anode/cathode gas pressures.The model is then demonstrated in an application with wind speed records and corresponding power generation over the course of one year at a wind farm in northern Germany.It produces results similar to experimental results in other papers for use in general applications of further study.
The constantly increasing computing power enables the implementation of complex simulation models.Therefore, it is possible to create more detailed models to predict system behavior more accurately.Modelica, for example, has proven great suitability in modelling complex systems, because of its high degree of reusability.However, understanding these models is quite difficult and many simulation models are poorly documented.Consequently, it is very time-consuming to retrace given model structures especially for novice.The Unified Modeling Language (UML) provides a user-friendly and graphical structure for documentation to simplify working with existing simulation models.Hence, an algorithm (ADoCSM) is developed to automatically present the structure of a Modelica simulation model in UML.This algorithm is exemplarily applied to a refrigerant circuit expansion valve model.Thereby, we contribute to an increase of simulation model quality as well as simplifying the entry in the world of Modelica.ADoCSM and the expansion valve model are freely available on GitHub:
The latest extension of the DLR FlexibleBodies Library concerns the field of automotive applications, namely the anti-roll bar.For the particular purposes of NVH and vehicle dynamics, the anti-roll bar module provides two appropriate levels of detail, both being based upon the beam preprocessor.In this paper, the procedure on preparing the models and their application for particular automotive related analyses is presented.
Complex fluid systems created with Modelica have often been difficult to simulate under hard real-time constraints since they typically involve non-linear equation systems that are difficult to solve especially within a predictable finite time frame. This paper explores the usage of a new approach that avoids non-linear equation systems and its suitability for real-time simulation. A model of an aircraft environmental control system is taken as a use case for this study.
In this paper, a modeling approach for comparing two heat-transfer-stations (HTS) is presented.By comparing a usual HTS with a modified HTS, where the return temperature on primary side of the district heating network (DHN) is used for heating the domestic warm water (DWW), it can be shown that utilizing the return flow of the heating positively contributes to a reduction of temperatures within a DHN and in this way saves exergy.The simulation model is implemented in Modelica.
The cascade refrigeration system and cascade air source heat pump have both been a well-known advanced technology for providing cooling and heating under certain circumstances. The intermediate temperature has been proven as a key factor in affecting the operating performance for both the refrigeration and heat pump system. However, the optimal intermediate temperature varies along with the change of the ambient temperature, cooling/heating demand. So achieving the real-time optimal intermediate temperature in a realistic operating circumstance is critical for improving the performance of cascade refrigeration or heat pump system. In this paper, the extreme seeking control strategy is proposed to optimize the coefficient of performance via seeking the real-time optimal intermediate temperature. A cascade refrigeration system model and a cascade air source heat pump water heater model are built based on Modelica, and the ASHP water heater model is validated by experimental data. The simulations are conducted under fixed and variable operating conditions to validate the effectiveness of the proposed strategy.
We present an original tool for analyzing the dynamics of a wide range of thermal energy systems using Modelica, developed by the Central Research Institute of Electric Power Industry in Japan.This tool was originally developed to analyze thermal power generation systems and has been validated against several sets of operational data so far.This paper reports the extension of the tool to the calculation of the thermophysical properties of not only water/steam and air/gas but also various refrigerants implemented by the ExternalMedia library.This addition expands the range of energy systems that can be analyzed with the tool.To validate the simulation data, a comparison between experimental and simulation data targeting a CO2 heatpump loop facility was drawn.
This paper presents a new planar wheel model with bore friction, a control strategy to avoid locking conditions of floor vehicles with caster wheels, and the new FMI-Adapter software package, which connects the Functional Mock-up Interface (FMI) standard with the Robot Operating System (ROS).It is demonstrated how this technology enables a convenient model-based control design workflow.The approach is applied to the ActiveShuttle, a self-driving vehicle (SDV) for industrial logistics.After modeling the wheel friction characteristics of the Ac-tiveShuttle, a feed forward controller to avoid high friction torques at the caster wheels in critical operation scenarios is designed and validated by model-in-the-loop simulations.The control function is exported as Functional Mock-up Unit (FMU) for co-simulation.With help of the FMI-Adapter package, the FMU is integrated as ROS node into the service-oriented robot control architecture, enhancing the existing motion controller.The functionality and performance is tested and successfully verified on the ActiveShuttle Dev Kit prototype.
This paper presents a practical application of computer aided control systems design using a new OpenModelica API (OMJulia) which allows to conveniently operate on Modelica models from the Julia language.Julia is a rather young language (Julia 1.0 was released in August 2018) designed to address the needs of numerical analysis and computational science, in particular it already has decent support for the control community.The magnetic levitation application at hand demonstrates how control system design can benefit from a suitable integration between Julia and Modelica.It is based on a commercially available control education kit in which the original controller is replaced by our own digital controller developed in this work.There exists an accompanying but independent paper which introduces the complete OMJulia API.
Electro-hydrostatic actuator (EHA) is a new trend in the more electric aircraft related research works and engineering applications.As a high-performance mechatronics product, however, the physical effects of actuator behavior are multidisciplinary, coupled and strongly nonlinear.Although many commercialized multi-domain and system-level simulation packages exist, they are rarely considered and analyzed as a whole, lacking of a unified model architecture, efficient modeling forms, and comprehensive simulation verification.In this paper, Modelica is used to build a multi-domain virtual prototype of the dual redundant electro-hydrostatic actuation system (DREHAS) that consists of two EHAs in parallel, which supports multiview modeling and interdisciplinary application of the system.Finally, a simulation application case of the elevator actuation system is presented to demonstrate the effective role of Modelica models in system modeling and evaluation.
The German national metrological institute, PhysikalischTechnische Bundesanstalt, is developing a new concept for volumetric primary standard to calibrate high pressure gas flow meters. The TUHH is supporting these R&D activities with its competence to elaborate computational models for detailed analysis of complex mechanical systems including fluid flow aspects. The new primary standard is based on a actively driven piston prover to measure the gas flow rate using the time the piston needs to displace a defined enclosed volume of gas in a cylinder. A computational model written in Modelica R © is developed to investigate the Flow Comparator’s dynamic behavior. Validation of the model shows good compliance of the piston velocity and differential pressure at the piston in the model with measured data. With this model the control voltage trajectory can be optimized to increase the available measuring time and it allows to gather detailed information about pressure and temperature development at arbitrary chosen locations in the system with high time resolution.
The interdisciplinary modelling language Modelica is increasingly used in the design and evaluation of energy systems. Heat supply represents a considerable share of the global energy supply. Especially in European cities, district heating grids are often used and implemented for heat coverage. The increasing integration of renewable energies and the extension of existing grids require engineers to be able to analyze and evaluate the behavior of such grids, not only statically in certain operating conditions, but also dynamically to enable the representation of complex system interaction.
This paper demonstrates using component oriented modeling and acausal semantics to create a basic library of behavioral components to model supply and demand. The models presented are each steady state models. While some examples include shifting economic conditions that cause the equilibrium points to change during the simulation, none of the models feature dynamic states. The main purpose of this paper is to demonstrate to people unfamiliar with Modelica (Modelica Association 2017) how Modelica can be used to model non-engineering systems and how it makes such modeling faster, easier and less errorprone compared to other approaches (e.g., using spreadsheets).
The present paper introduces a recently developed tool for building web-based simulators called Bodylight.js. Simulators are applications composed of a mathematical model and a graphical user interface that allows the user to easily interact with the model and visualize the results. A modelica model is first exported to FMI with sources, transcompiled into JavaScript and WebAssembly and connected to a GUI, comprised of graphical animations created in Adobe Animate and elements that allow to control the input model such as sliders, buttons, etc. A physiological e-learning application explaining the function of a nephron – the basic functional unit of kidneys – is presented later as a use-case. The model was developed primarily as a teaching aid for use in courses of physiology for medical students at our university. Purpose of this work is to describe the new Bodylight.js tool and to prove its usability by building the medium-complex e-learning kidney simulator. The simulator helps medical students to better understand renal function at the very basic level.
