Final Report Summary - ISSE (Improvement of numerical models for JTI/GRA Shared Simulation Environment)
Executive Summary:
In the iSSE project, TWT has set up an advanced Shared Simulation Environment (SSE) to model the electrical power generation, consumption and regulation within an All-Electric Aircraft (AEA). The SSE architecture developed in iSSE is based on the “TWT co-simulation framework”. This tool is an FMI-compliant (see http://www.modelisar.com/) backbone for the synchronization and communication between simulation models provided by third parties and running in heterogeneous simulation tools (e.g. Matlab/Simulink, Dymola Modelica) on different machines. It is written in Java using modern, efficient libraries. It includes a GUI for control and monitoring as well as Connectors for several common simulation tools. The individual models are provided by their respective suppliers and are made ready for inclusion into the co-simulation by TWT by adding the functional wrappers to them. The software requirements for co-simulation installation and the expected folder structure for successful execution of the software have been documented in the comprehensive user manual. Within the project, the system models for flight data, energy management, cabin, environmental control, power generation, ice protection, landing gear and flight control have been integrated into the SSE and can now be co-simulated via GUI and batch operations. Models are easily exchangeable, offering high user value for the system simulation.
The GUI has been specifically designed to meet the requirements of the proponent. It allows the user to choose available models, model versions, variants, settings, parameter files, values, and sync rates. Moreover, dedicated post-processing routines have been implemented to analyse and visualize co-simulation results. Specifically, the signals exchanged during the co-simulation can be plotted using a ‘Visualizer’ panel. Also, signal communication between different models of the co-simulation can be depicted by way of a Signal Routing Tree Graph.
Project Context and Objectives:
Project Context: Green Regional Aircraft
The advancement of low noise, low emission air transport is one of the most urgent requirements in the area of transportation. As personal and economical mobility increase, so does the demand on transportation. Especially regional transportation faces new challenges, as the demands for fast regional air traffic increase, both as end-to-end transportation and as feeders for long-haul flights from main airports. A remarkable number of airports is served by regional aircrafts only. At the same time environmental awareness and demands for air traffic to become less intrusive in terms of emission and noise pollution increase. Therefore, regional aircraft need to be developed that can cater to the need of silent and efficient regional air transport.
In the Clean Sky ITD Green Regional Aircraft (GRA), the WP 3.2 “Technology for Systems” addresses the All Electrical Domain (AEA) for aircraft which directly aims at reducing the environmental footprint of aviation. While the vast majority of today’s aircraft use the propulsion system as power source to drive the main aircraft on-board systems, powering these electrically is considered a promising option to provide answers to the environmental challenges.
However, this approach causes power absorption issues for electrical generators which call for adequate components and energy management control logics. The energy management system (EMS) has to guarantee that all critical systems are provided with enough power in all flight conditions, respecting the maximum allowed power. Thus it may be necessary to actively reduce the power supply for non-critical system under certain conditions. The final goal therefore lies in developing an EMS which acts as power supplier to the other systems and offers an optimized sharing and allocation of the total electrical power. Realizing this technical complexity, a call for a holistic solution is obvious.
Concept and Objectives
It is evident that the integration of a large number of novel electro-mechanical systems into an architecture as complex and long-standing as that of an aircraft is no task for trial and error. Concise sub-system level simulations are already being carried out to predict power consumption and operational parameters for these new systems. However, there are many different simulation tools in use, and combining the results of sub-system tests into a holistic picture is difficult.
In this context, the iSSE project aims at developing a Shared Simulation Environment (SSE) able to simultaneously co-simulate the static and dynamic performance of on-board aircraft systems with particular focus on electrical power absorption and thermal energy production. The SSE thereby has to be designed and implemented as a comprehensive solution that provides means to master all essential aspects of the dimensioning for the energy management properties. Given the novelty in the concept of an AEA and the fact that no ready-to-use market solution for a SSE is available, the research need is evident. The final scope is to permit the testing and the validation of the logics used by the Energy Management System (EMS) and the internal environmental condition.
In the iSSE project, TWT has set up an SSE architecture that enhances the “TWT co-simulation Framework”. This tool is a FMI-compliant (see http://www.modelisar.com/) backbone for the synchronization and communication between simulation models provided by third parties and running in heterogeneous simulation tools (e.g. Matlab/Simulink, Dymola Modelica, etc.) on different machines. The co-simulation Framework already provides many of the features expected by the proponent. In order to fully meet the proponent’s needs as to the SSE, the project iSSE has had the following main objectivs:
• The requirements of the iSSE, approximately specified in the CfP, are structured, quantified, and documented. To this end the requirements will be broken into an array of sub-requirements.
• The models for different on board systems which are supplied by the proponent are integrated in the TWT Co-Simulation Framework and first simulation runs are executed. Results and insights gathered from these runs are documented.
• The existing co-simulation framework is extended to support additional simulation tools, such as AMESim.
• A customised graphical user interface is implemented for the SSE, allowing convenient selection of the flight scenario, available models, model versions, variants, settings, parameter files, values, and sync rates.
• A batch process for configuring the simulation environment, i.e. selecting the models and their initial parameterization is developed.
• Post-processing capabilities are implemented to analyse and visualize co-simulation results. Specifically, the signals exchanged during the co-simulation are to be plotted using a ‘Visualizer’ panel. Also, signal communication between different models of the co-simulation should be depicted by way of a Signal Routing Tree Graph.
• Support is provided to the proponent to ensure correct installation and usage of the developed iSSE software.
Project Results:
The major science and technology achievements in the iSSE project are the extension of the TWT co-simulation framework in terms of supported modelling tools, graphical user interface, batch processing, and analysis and post-processing capabilities. In the following, the TWT co-simulation technology and the aforementioned extensions are discussed in greater detail.
TWT Co-Simulation Framework
The TWT co-simulation framework manages signal exchange between
• multiple simulations, running in
• different tools, possibly located on
• multiple hosts
It is written in Java using modern, efficient libraries. It includes a GUI for control and monitoring as well as connectors for several common simulation tools.
The TWT co-simulation framework uses FMI model description standards to describe
• simulations
• their signals
The framework generates these FMI model descriptions for several tools. Functional mock-up units (FMUs) can be included
• directly via the framework FMU connector
• via the TWT Matlab/Simulink FMU interface
Architecture
The principal architecture of the TWT co-simulation framework is a star network. At the centre of the star is the router. The router communicates with all simulations via their connectors and forwards all sent messages from the simulations to the specified recipients. All co-simulation participants register with the router. The router’s abilities are limited; its main purpose is to enable a star-like network topology.
The simulation master provides a GUI and accompanying functionality that allows the user to control and monitor the co-simulation.
Every connector adjoins a simulation to the framework and contains most of the logic. Multiple specialized connectors are available. For this project, three connector types are relevant:
• Simulink connector: this connector uses the Java API to directly connect a Simulink model to the framework.
• Dymola connector: this connector consists of two parts (and processes) – router-sided and Dymola-sided. The Router-sided process is a socket server written in Java, which communicates with the Dymola-sided process. The latter is a socket client written partly in C and partly in Modelica, using TCP.
• FMU connector: this connector consists of the same socket server as the Dymola connector and a specialized socket client that is capable of simulating an FMU.
Communication
Messages
The connectors and the master exchange messages during the setup and co-simulation phases. These messages are either broadcasted or addressed to a specific recipient. The payload of these messages comprises:
• Type of message
• Simulation time
• Variable number of string, double and Boolean values
There are five message types:
• State messages: a simulation communicates its current state (initialized, synchronized, running, paused, finished)
• Command messages: the master sends a command to one or multiple simulations
• Information requests: (new) simulations request information about the other co-simulation members
• Information messages: a co-simulation member sends various information including its FMI model description
• Signal messages: simulations send the current values of their output signals
The payload is generated with Google Protocol Buffers, which is a widely used and well developed library. This implementation features a fast generation of small messages with minimal overhead, which is important as many messages are exchanged during co-simulation.
Signals
Signals are assigned to namespaces, not models. This enables a clearer model structure and allows for the re-use of signal names and simulations in different namespaces.
Step size
The chosen communication step size is
• fixed (not adaptive)
• the same for all simulations (a subset of simulations may not communicate with a different step size with each other than with other simulations)
The communication step size is computed as the least number which is
• a multiple of at least two step sizes and
• a divisor (or multiple) of all other step sizes
Therefore, if all step sizes are identical, that step size is chosen.
Software
In order to run a co-simulation, the following software must be installed:
• Dymola 2012.FD01
• Matlab/Simulink 2011b
• Java SE Runtime Environment Version 7 or 8
• Microsoft Visual Studio C++ 2010 Express
• psexec (freeware), e.g. from the following source:
http://www.heise.de/download/psexec-1157761.html
The psexec installation folder must be added to the Windows system path variable.
Folder Structure
The TWT co-simulation expects a specific folder structure, which needs to be observed in order to guarantee correct program execution. The main folders are:
• ‘com.twt.cosim.core’ contains the co-simulation core (do not alter)
• ‘cosim’ contains the co-simulation executable and configurations (do not alter)
• ‘models’ contains the model files (e.g. Simulink or Dymola); the folder structure and names should not be changed; every model folder contains folders for the different level models (e.g. ‘L2’ or ‘L3’)
Co-Simulation architecture
Model directory structure
All models must be located in subdirectories of a general “..\models” directory. A first layer of subdirectories groups the models by the modelled system; a second layer of subdirectories groups models of the same system by their detail level (architectural, functional, and behavioural). In the final version of the iSSE, the following models can be connected to a whole systems simulation:
Model Directory Level (Tool)
Flight data ..\flightdata\default
Energy Management System ..\EMS\L0 Dummy (Simulink)
..\EMS\L2 Level 2 (Dymola)
El. Power Generation System ..\EPGS\L2 Level 2 (Simulink)
Environmental Control System ..\ECS\L2 Level 2 (Simulink)
..\ECS\L3 Level 3 (Simulink)
Cabin Thermal Model ..\CT\L0 Dummy (Simulink)
..\CT\L2 Level 2 (AMESim/Simulink)
Ice Protection System ..\IPS\L0 Dummy (Simulink)
..\IPS\L1 Level 1 (Dymola)
Flight Control System ..\FCS\L2 Level 2 (Simulink)
..\FCS\L3 Level 3 (Dymola)
Landing Gear System ..\LGS\L2 Level 2 (Simulink)
..\LGS\L3 Level 3 (Dymola)
Fuel System ..\FS\L0 Dummy (Simulink)
Other Systems ..\other\default Default (Simulink)
AMESim Support
It should be noted that the above list includes an AMESim model. The support for this simulation tool was implemented as part of the iSSE project activities. In order to enable AMESim models to be integrated in the co-simulation, a corresponding connector had to be developed, that enables the required message exchange.
Co-Simulation GUI, Batch File
The iSSE co-simulation can conveniently be set up and run through the co-simulation GUI, which was specifically developed for the purposes of the project:
..\cosimUI\TWTCosimulationFramework.exe
The GUI will automatically create a batch file depending on the simulations the user chooses to run. First, this batch file starts the router process and then the master process, which registers with the router. Then, one by one, the simulators are started and these register with the router as well.
A simulator registers with the master only after the simulator has initialized, i.e. has solved its initialization problem (at this point, the connector – part of the model – halts the running simulation until the master issues the start command). To cause a simulator to open, compile, and start running, the batch file opens a new instance of the proper tool (e.g. Matlab, Dymola) and provides the tool with a customized start script as command line argument. The tool executes this start script, which contains the commands needed to load, compile and run the desired model. These start scripts are
• a start_model.m script file interpreted by Matlab,
• a start.mos script file interpreted by Dymola, or
• a.xml file interpreted by the FMU connector.
Distributed Computing
The TWT Co-Simulation Framework technology inherently supports distributed computing in a LAN because all interprocess communication is via TCP/IP. At first, a manual test and configuration of the technology is recommended. Then, the process can be automated via the Co-Simulation GUI.
Visualization of the Results
For the iSSE project, the TWT Co-Simulation framework has been extended to include results visualization. The signals exchanged during the co-simulation can be plotted in the ‘Visualizer’ panel. To plot a signal, click on a signal from the ‘signals’ list box and select an axis; then click the Right-Arrow (‘-->’) button. The plotted signal will now appear in the ‘selected’ list box and as a plot vs time in the graph on the right. To remove a signal from the graph, click on a signal in the ‘selected’ list box; then click the Left-Arrow (‘<--’) button. The signal graph panel offers different visualization tools via buttons (hover with the mouse over those buttons to see their functionalities).
Signal Routing Tree Graph
The Signal Routing Tree Graph depicts how signals are communicated between the different models in a co-simulation.
1. Click “Engines” to see all simulation models that are connected to the co-simulation.
2. If you click on any model you will see an input and an output branch. Under input all signals are listed which the current model receives. Under output all signals are listed which the current model sends to other simulators.
3. By going a level deeper it is possible to see which models the current model communicates with.
a. for an input signal the one model is shown where this signal originates
b. for an output signal all models are listed which receive this signal from the current model
Potential Impact:
Impact
The project iSSE is expected to create a strong impact on the European Regional Aircraft sector, by not only contributing to a concept for a more environment aware A/C configuration, but moreover by creating an efficient methodology for the exploitation of the All Electric Aircraft concept for Green Regional Aircraft. Thus the activity focused on building up a strong European position for the given challenges of exploiting the green aviation domain. The holistic approach of the project—a high-quality shared simulation environment—will strengthen the European leadership in cross-cutting technologies. The integrated use of advanced and combined simulation techniques in the aeronautical domain moreover will leverage the overall application of digital design towards more and effective solutions in the frame of optimising air transport.
Addressing the overall mission of the Green Region Aircraft, the activity has derived a beneficial combination of techniques for reaching the overall GRA goals. The straight-forward approach of iSSE addresses all relevant aspects of a high-quality simulation environment for illuminating the energy management logics within the scope of sharing the total on-board electric power. TWT has taken care to implement each step of the project optimally; in order to achieve the overall intended reduced fuel consumption and lower CO2 emissions.
In addition to helping to achieve CSJUs overall GRA targets, this activity has strengthend the position of European Aircraft development in an ever more competitive world market. As Europe is among the leaders in striving for CO2 emission reduction and cleaner transport, this competitive advantage will become even more beneficial when other markets turn their gaze upon Green Aviation.
To summarize, in the iSSE project, critical aspects of the all electric aircrafts (AEA) concept have been investigated. The problems associated with the supply of electrical power are very complex, since they involve the electrical and thermal behaviour of a wide variety of critical and non-critical onboard systems under varying flight conditions. This undertaking has addressed this complexity through a Shared Simulation Environment. This approach sets the stage for the following array of impacts:
• The simulation models provide a quantitative basis for optimizing the energy management system and thermal architecture of an AEA. Since the simulation includes a wide variety of numerical models, this process offers a comprehensive and more accurate method than other solutions currently available.
• Due to the FMI-conformity of the approach including the co-simulation Framework, and due to the strict use of standards, the simulation models used for the EMS-simulation, can readily be integrated into a larger simulation of the whole aircraft. This modularity and adaptability makes it possible to cross-optimize multiple aircraft systems.
• More generally, this approach reveals, at the design stage, whether and in which particular form the AEA concept is feasible within the Green Region Aircraft domain (GRA).
Hence, this activity contributes to building towards a strong European case for the upcoming challenges of exploiting the green aviation domain. The iSSE project therefore has the potential to become an important brick of the overall CleanSky approach, and in this way the project contributes to the overall reduction of fuel consumption and to and lower CO2 emissions.
Dissemination and exploitation of project results
Innovation is TWT's key driver of competitive advantage, growth, and profitability. Being an established and well-known business partner for companies out of automotive, aerospace and medical domain, research and development is an indispensable part of our business mission. The focus of the company is the development of new and marketable products, processes and services with high customer benefit.
Therefore, three parts are inherent to TWT’s exploitation and dissemination strategy. All approaches will consider confidentiality aspects as required by the CSJU.
• The project and its results are intended to be used to extending and strengthening TWT’s business portfolio in the aviation domain. TWT heads for offering its experience and transfer its knowledge into the aviation domain, therefore improving Europe’s competitiveness by identifying and leveraging meaningful areas of technology transfer, in particular between the strong automotive and strong aviation disciplines.
• As the results of the iSSE project can be directly applied to its daily business processes, TWT has a strong interest in disseminating and exploiting the project’s results in its business portfolio. Since a large number of TWT customers are located within the automotive domain, an impact to enrich the portfolio of TWT, its customers and therefore the economically critical automotive domain is expected. Specifically, the extension of the TWT Co-Simulation in terms of supported modelling tools, batch capability and visualization techniques can directly be ported to the automotive domain. Moreover, the greening of any transport systems is not only a field for European market leadership but also for the society at large.
• TWT disseminates projects results on several channels to general public audience. TWT frequently attends and presents at international conferences, at which the iSSE activity and its results will and already have been demonstrated. Moreover, TWT organizes so-called “TWT forums”, where project results are presented to its staff and customers. Finally, the iSSE project has been brought to the attention of a wider (also non-technical) public audience by inclusion on TWT’s webpage, both in English and German language.
The iSSE project has been presented at the following conferences:
• C. Radermacher, F. Burger, O. Kotte, C. Kübler, U. Odefey, M. Pfeil, V. Fäßler: Design, Optimization and validation of all-electric aircraft systems via co-simulation, Greener Aviation, Brüssel, 2014 (talk und conference paper)
• M. Pfeil.,F. Burger, I. Bausch-Gall, F. Schettini, E. Denti, G. Di Rito, R. Galatolo, L. Gall, U. Odefey, O. Kotte, C. Kübler, V. Fäßler.: Co-Simulation for Design, Optimization and Analysis of All-Electric Aircraft Systems, AIRTEC - Supply on the Wings, Frankfurt/Main, 2013 (tortrag und conference paper)
• C. Radermacher: The TWT Co-Simulation framework in all-electric aircraft applications, CEAS-SCAD Symposium Toulouse 2014 (talk)
• M. Pfeil, C. Knörzer, V. Fäßler: Cooperative Engineering using Digital Functional Mock-Up, AIRTEC München 2015 (talk and conference paper)
List of Websites:
No dedicated website has been created for the iSSE project. However, the project is listed among the TWT research activities on the company website:
https://www.twt-gmbh.de/en/research/projects/projects.html
In the iSSE project, TWT has set up an advanced Shared Simulation Environment (SSE) to model the electrical power generation, consumption and regulation within an All-Electric Aircraft (AEA). The SSE architecture developed in iSSE is based on the “TWT co-simulation framework”. This tool is an FMI-compliant (see http://www.modelisar.com/) backbone for the synchronization and communication between simulation models provided by third parties and running in heterogeneous simulation tools (e.g. Matlab/Simulink, Dymola Modelica) on different machines. It is written in Java using modern, efficient libraries. It includes a GUI for control and monitoring as well as Connectors for several common simulation tools. The individual models are provided by their respective suppliers and are made ready for inclusion into the co-simulation by TWT by adding the functional wrappers to them. The software requirements for co-simulation installation and the expected folder structure for successful execution of the software have been documented in the comprehensive user manual. Within the project, the system models for flight data, energy management, cabin, environmental control, power generation, ice protection, landing gear and flight control have been integrated into the SSE and can now be co-simulated via GUI and batch operations. Models are easily exchangeable, offering high user value for the system simulation.
The GUI has been specifically designed to meet the requirements of the proponent. It allows the user to choose available models, model versions, variants, settings, parameter files, values, and sync rates. Moreover, dedicated post-processing routines have been implemented to analyse and visualize co-simulation results. Specifically, the signals exchanged during the co-simulation can be plotted using a ‘Visualizer’ panel. Also, signal communication between different models of the co-simulation can be depicted by way of a Signal Routing Tree Graph.
Project Context and Objectives:
Project Context: Green Regional Aircraft
The advancement of low noise, low emission air transport is one of the most urgent requirements in the area of transportation. As personal and economical mobility increase, so does the demand on transportation. Especially regional transportation faces new challenges, as the demands for fast regional air traffic increase, both as end-to-end transportation and as feeders for long-haul flights from main airports. A remarkable number of airports is served by regional aircrafts only. At the same time environmental awareness and demands for air traffic to become less intrusive in terms of emission and noise pollution increase. Therefore, regional aircraft need to be developed that can cater to the need of silent and efficient regional air transport.
In the Clean Sky ITD Green Regional Aircraft (GRA), the WP 3.2 “Technology for Systems” addresses the All Electrical Domain (AEA) for aircraft which directly aims at reducing the environmental footprint of aviation. While the vast majority of today’s aircraft use the propulsion system as power source to drive the main aircraft on-board systems, powering these electrically is considered a promising option to provide answers to the environmental challenges.
However, this approach causes power absorption issues for electrical generators which call for adequate components and energy management control logics. The energy management system (EMS) has to guarantee that all critical systems are provided with enough power in all flight conditions, respecting the maximum allowed power. Thus it may be necessary to actively reduce the power supply for non-critical system under certain conditions. The final goal therefore lies in developing an EMS which acts as power supplier to the other systems and offers an optimized sharing and allocation of the total electrical power. Realizing this technical complexity, a call for a holistic solution is obvious.
Concept and Objectives
It is evident that the integration of a large number of novel electro-mechanical systems into an architecture as complex and long-standing as that of an aircraft is no task for trial and error. Concise sub-system level simulations are already being carried out to predict power consumption and operational parameters for these new systems. However, there are many different simulation tools in use, and combining the results of sub-system tests into a holistic picture is difficult.
In this context, the iSSE project aims at developing a Shared Simulation Environment (SSE) able to simultaneously co-simulate the static and dynamic performance of on-board aircraft systems with particular focus on electrical power absorption and thermal energy production. The SSE thereby has to be designed and implemented as a comprehensive solution that provides means to master all essential aspects of the dimensioning for the energy management properties. Given the novelty in the concept of an AEA and the fact that no ready-to-use market solution for a SSE is available, the research need is evident. The final scope is to permit the testing and the validation of the logics used by the Energy Management System (EMS) and the internal environmental condition.
In the iSSE project, TWT has set up an SSE architecture that enhances the “TWT co-simulation Framework”. This tool is a FMI-compliant (see http://www.modelisar.com/) backbone for the synchronization and communication between simulation models provided by third parties and running in heterogeneous simulation tools (e.g. Matlab/Simulink, Dymola Modelica, etc.) on different machines. The co-simulation Framework already provides many of the features expected by the proponent. In order to fully meet the proponent’s needs as to the SSE, the project iSSE has had the following main objectivs:
• The requirements of the iSSE, approximately specified in the CfP, are structured, quantified, and documented. To this end the requirements will be broken into an array of sub-requirements.
• The models for different on board systems which are supplied by the proponent are integrated in the TWT Co-Simulation Framework and first simulation runs are executed. Results and insights gathered from these runs are documented.
• The existing co-simulation framework is extended to support additional simulation tools, such as AMESim.
• A customised graphical user interface is implemented for the SSE, allowing convenient selection of the flight scenario, available models, model versions, variants, settings, parameter files, values, and sync rates.
• A batch process for configuring the simulation environment, i.e. selecting the models and their initial parameterization is developed.
• Post-processing capabilities are implemented to analyse and visualize co-simulation results. Specifically, the signals exchanged during the co-simulation are to be plotted using a ‘Visualizer’ panel. Also, signal communication between different models of the co-simulation should be depicted by way of a Signal Routing Tree Graph.
• Support is provided to the proponent to ensure correct installation and usage of the developed iSSE software.
Project Results:
The major science and technology achievements in the iSSE project are the extension of the TWT co-simulation framework in terms of supported modelling tools, graphical user interface, batch processing, and analysis and post-processing capabilities. In the following, the TWT co-simulation technology and the aforementioned extensions are discussed in greater detail.
TWT Co-Simulation Framework
The TWT co-simulation framework manages signal exchange between
• multiple simulations, running in
• different tools, possibly located on
• multiple hosts
It is written in Java using modern, efficient libraries. It includes a GUI for control and monitoring as well as connectors for several common simulation tools.
The TWT co-simulation framework uses FMI model description standards to describe
• simulations
• their signals
The framework generates these FMI model descriptions for several tools. Functional mock-up units (FMUs) can be included
• directly via the framework FMU connector
• via the TWT Matlab/Simulink FMU interface
Architecture
The principal architecture of the TWT co-simulation framework is a star network. At the centre of the star is the router. The router communicates with all simulations via their connectors and forwards all sent messages from the simulations to the specified recipients. All co-simulation participants register with the router. The router’s abilities are limited; its main purpose is to enable a star-like network topology.
The simulation master provides a GUI and accompanying functionality that allows the user to control and monitor the co-simulation.
Every connector adjoins a simulation to the framework and contains most of the logic. Multiple specialized connectors are available. For this project, three connector types are relevant:
• Simulink connector: this connector uses the Java API to directly connect a Simulink model to the framework.
• Dymola connector: this connector consists of two parts (and processes) – router-sided and Dymola-sided. The Router-sided process is a socket server written in Java, which communicates with the Dymola-sided process. The latter is a socket client written partly in C and partly in Modelica, using TCP.
• FMU connector: this connector consists of the same socket server as the Dymola connector and a specialized socket client that is capable of simulating an FMU.
Communication
Messages
The connectors and the master exchange messages during the setup and co-simulation phases. These messages are either broadcasted or addressed to a specific recipient. The payload of these messages comprises:
• Type of message
• Simulation time
• Variable number of string, double and Boolean values
There are five message types:
• State messages: a simulation communicates its current state (initialized, synchronized, running, paused, finished)
• Command messages: the master sends a command to one or multiple simulations
• Information requests: (new) simulations request information about the other co-simulation members
• Information messages: a co-simulation member sends various information including its FMI model description
• Signal messages: simulations send the current values of their output signals
The payload is generated with Google Protocol Buffers, which is a widely used and well developed library. This implementation features a fast generation of small messages with minimal overhead, which is important as many messages are exchanged during co-simulation.
Signals
Signals are assigned to namespaces, not models. This enables a clearer model structure and allows for the re-use of signal names and simulations in different namespaces.
Step size
The chosen communication step size is
• fixed (not adaptive)
• the same for all simulations (a subset of simulations may not communicate with a different step size with each other than with other simulations)
The communication step size is computed as the least number which is
• a multiple of at least two step sizes and
• a divisor (or multiple) of all other step sizes
Therefore, if all step sizes are identical, that step size is chosen.
Software
In order to run a co-simulation, the following software must be installed:
• Dymola 2012.FD01
• Matlab/Simulink 2011b
• Java SE Runtime Environment Version 7 or 8
• Microsoft Visual Studio C++ 2010 Express
• psexec (freeware), e.g. from the following source:
http://www.heise.de/download/psexec-1157761.html
The psexec installation folder must be added to the Windows system path variable.
Folder Structure
The TWT co-simulation expects a specific folder structure, which needs to be observed in order to guarantee correct program execution. The main folders are:
• ‘com.twt.cosim.core’ contains the co-simulation core (do not alter)
• ‘cosim’ contains the co-simulation executable and configurations (do not alter)
• ‘models’ contains the model files (e.g. Simulink or Dymola); the folder structure and names should not be changed; every model folder contains folders for the different level models (e.g. ‘L2’ or ‘L3’)
Co-Simulation architecture
Model directory structure
All models must be located in subdirectories of a general “..\models” directory. A first layer of subdirectories groups the models by the modelled system; a second layer of subdirectories groups models of the same system by their detail level (architectural, functional, and behavioural). In the final version of the iSSE, the following models can be connected to a whole systems simulation:
Model Directory Level (Tool)
Flight data ..\flightdata\default
Energy Management System ..\EMS\L0 Dummy (Simulink)
..\EMS\L2 Level 2 (Dymola)
El. Power Generation System ..\EPGS\L2 Level 2 (Simulink)
Environmental Control System ..\ECS\L2 Level 2 (Simulink)
..\ECS\L3 Level 3 (Simulink)
Cabin Thermal Model ..\CT\L0 Dummy (Simulink)
..\CT\L2 Level 2 (AMESim/Simulink)
Ice Protection System ..\IPS\L0 Dummy (Simulink)
..\IPS\L1 Level 1 (Dymola)
Flight Control System ..\FCS\L2 Level 2 (Simulink)
..\FCS\L3 Level 3 (Dymola)
Landing Gear System ..\LGS\L2 Level 2 (Simulink)
..\LGS\L3 Level 3 (Dymola)
Fuel System ..\FS\L0 Dummy (Simulink)
Other Systems ..\other\default Default (Simulink)
AMESim Support
It should be noted that the above list includes an AMESim model. The support for this simulation tool was implemented as part of the iSSE project activities. In order to enable AMESim models to be integrated in the co-simulation, a corresponding connector had to be developed, that enables the required message exchange.
Co-Simulation GUI, Batch File
The iSSE co-simulation can conveniently be set up and run through the co-simulation GUI, which was specifically developed for the purposes of the project:
..\cosimUI\TWTCosimulationFramework.exe
The GUI will automatically create a batch file depending on the simulations the user chooses to run. First, this batch file starts the router process and then the master process, which registers with the router. Then, one by one, the simulators are started and these register with the router as well.
A simulator registers with the master only after the simulator has initialized, i.e. has solved its initialization problem (at this point, the connector – part of the model – halts the running simulation until the master issues the start command). To cause a simulator to open, compile, and start running, the batch file opens a new instance of the proper tool (e.g. Matlab, Dymola) and provides the tool with a customized start script as command line argument. The tool executes this start script, which contains the commands needed to load, compile and run the desired model. These start scripts are
• a start_model.m script file interpreted by Matlab,
• a start.mos script file interpreted by Dymola, or
• a
Distributed Computing
The TWT Co-Simulation Framework technology inherently supports distributed computing in a LAN because all interprocess communication is via TCP/IP. At first, a manual test and configuration of the technology is recommended. Then, the process can be automated via the Co-Simulation GUI.
Visualization of the Results
For the iSSE project, the TWT Co-Simulation framework has been extended to include results visualization. The signals exchanged during the co-simulation can be plotted in the ‘Visualizer’ panel. To plot a signal, click on a signal from the ‘signals’ list box and select an axis; then click the Right-Arrow (‘-->’) button. The plotted signal will now appear in the ‘selected’ list box and as a plot vs time in the graph on the right. To remove a signal from the graph, click on a signal in the ‘selected’ list box; then click the Left-Arrow (‘<--’) button. The signal graph panel offers different visualization tools via buttons (hover with the mouse over those buttons to see their functionalities).
Signal Routing Tree Graph
The Signal Routing Tree Graph depicts how signals are communicated between the different models in a co-simulation.
1. Click “Engines” to see all simulation models that are connected to the co-simulation.
2. If you click on any model you will see an input and an output branch. Under input all signals are listed which the current model receives. Under output all signals are listed which the current model sends to other simulators.
3. By going a level deeper it is possible to see which models the current model communicates with.
a. for an input signal the one model is shown where this signal originates
b. for an output signal all models are listed which receive this signal from the current model
Potential Impact:
Impact
The project iSSE is expected to create a strong impact on the European Regional Aircraft sector, by not only contributing to a concept for a more environment aware A/C configuration, but moreover by creating an efficient methodology for the exploitation of the All Electric Aircraft concept for Green Regional Aircraft. Thus the activity focused on building up a strong European position for the given challenges of exploiting the green aviation domain. The holistic approach of the project—a high-quality shared simulation environment—will strengthen the European leadership in cross-cutting technologies. The integrated use of advanced and combined simulation techniques in the aeronautical domain moreover will leverage the overall application of digital design towards more and effective solutions in the frame of optimising air transport.
Addressing the overall mission of the Green Region Aircraft, the activity has derived a beneficial combination of techniques for reaching the overall GRA goals. The straight-forward approach of iSSE addresses all relevant aspects of a high-quality simulation environment for illuminating the energy management logics within the scope of sharing the total on-board electric power. TWT has taken care to implement each step of the project optimally; in order to achieve the overall intended reduced fuel consumption and lower CO2 emissions.
In addition to helping to achieve CSJUs overall GRA targets, this activity has strengthend the position of European Aircraft development in an ever more competitive world market. As Europe is among the leaders in striving for CO2 emission reduction and cleaner transport, this competitive advantage will become even more beneficial when other markets turn their gaze upon Green Aviation.
To summarize, in the iSSE project, critical aspects of the all electric aircrafts (AEA) concept have been investigated. The problems associated with the supply of electrical power are very complex, since they involve the electrical and thermal behaviour of a wide variety of critical and non-critical onboard systems under varying flight conditions. This undertaking has addressed this complexity through a Shared Simulation Environment. This approach sets the stage for the following array of impacts:
• The simulation models provide a quantitative basis for optimizing the energy management system and thermal architecture of an AEA. Since the simulation includes a wide variety of numerical models, this process offers a comprehensive and more accurate method than other solutions currently available.
• Due to the FMI-conformity of the approach including the co-simulation Framework, and due to the strict use of standards, the simulation models used for the EMS-simulation, can readily be integrated into a larger simulation of the whole aircraft. This modularity and adaptability makes it possible to cross-optimize multiple aircraft systems.
• More generally, this approach reveals, at the design stage, whether and in which particular form the AEA concept is feasible within the Green Region Aircraft domain (GRA).
Hence, this activity contributes to building towards a strong European case for the upcoming challenges of exploiting the green aviation domain. The iSSE project therefore has the potential to become an important brick of the overall CleanSky approach, and in this way the project contributes to the overall reduction of fuel consumption and to and lower CO2 emissions.
Dissemination and exploitation of project results
Innovation is TWT's key driver of competitive advantage, growth, and profitability. Being an established and well-known business partner for companies out of automotive, aerospace and medical domain, research and development is an indispensable part of our business mission. The focus of the company is the development of new and marketable products, processes and services with high customer benefit.
Therefore, three parts are inherent to TWT’s exploitation and dissemination strategy. All approaches will consider confidentiality aspects as required by the CSJU.
• The project and its results are intended to be used to extending and strengthening TWT’s business portfolio in the aviation domain. TWT heads for offering its experience and transfer its knowledge into the aviation domain, therefore improving Europe’s competitiveness by identifying and leveraging meaningful areas of technology transfer, in particular between the strong automotive and strong aviation disciplines.
• As the results of the iSSE project can be directly applied to its daily business processes, TWT has a strong interest in disseminating and exploiting the project’s results in its business portfolio. Since a large number of TWT customers are located within the automotive domain, an impact to enrich the portfolio of TWT, its customers and therefore the economically critical automotive domain is expected. Specifically, the extension of the TWT Co-Simulation in terms of supported modelling tools, batch capability and visualization techniques can directly be ported to the automotive domain. Moreover, the greening of any transport systems is not only a field for European market leadership but also for the society at large.
• TWT disseminates projects results on several channels to general public audience. TWT frequently attends and presents at international conferences, at which the iSSE activity and its results will and already have been demonstrated. Moreover, TWT organizes so-called “TWT forums”, where project results are presented to its staff and customers. Finally, the iSSE project has been brought to the attention of a wider (also non-technical) public audience by inclusion on TWT’s webpage, both in English and German language.
The iSSE project has been presented at the following conferences:
• C. Radermacher, F. Burger, O. Kotte, C. Kübler, U. Odefey, M. Pfeil, V. Fäßler: Design, Optimization and validation of all-electric aircraft systems via co-simulation, Greener Aviation, Brüssel, 2014 (talk und conference paper)
• M. Pfeil.,F. Burger, I. Bausch-Gall, F. Schettini, E. Denti, G. Di Rito, R. Galatolo, L. Gall, U. Odefey, O. Kotte, C. Kübler, V. Fäßler.: Co-Simulation for Design, Optimization and Analysis of All-Electric Aircraft Systems, AIRTEC - Supply on the Wings, Frankfurt/Main, 2013 (tortrag und conference paper)
• C. Radermacher: The TWT Co-Simulation framework in all-electric aircraft applications, CEAS-SCAD Symposium Toulouse 2014 (talk)
• M. Pfeil, C. Knörzer, V. Fäßler: Cooperative Engineering using Digital Functional Mock-Up, AIRTEC München 2015 (talk and conference paper)
List of Websites:
No dedicated website has been created for the iSSE project. However, the project is listed among the TWT research activities on the company website:
https://www.twt-gmbh.de/en/research/projects/projects.html
