Analyses as early as possible
Virtual prototyping minimizes development times and is therefore becoming increasingly important. With the ETAS INTECRIO-VP add-on, system models can be analyzed without the need for complex prototyping hardware. This means that already in the early stage of development it is possible to use open-loop validation on the basis of existing test data or Model-in-the-Loop (MiL) technologies in order to validate and pre-calibrate new functions.
INTECRIO-VP brings virtual prototyping to the PC workstation. It achieves especially informative results through the use of a virtual real-time operating system and thanks to the ECU-compliant transmission of signals between software components. In addition to the function models, it is possible to integrate vehicle or environment models, which enables MiL applications. This allows not only the validation of functions, but also, for example, their pre-calibration.
The simulation time can be scaled in INTECRIO-VP. Virtual experiments can then be carried out quicker using fast-motion timing, which offers significant advantages, for example, for automated processes. By contrast, the individual simulation steps become easier to understand when virtual experiments are performed using slow-motion timing.
The stimuli are generated by the ETAS Experiment Environment’s signal generator.
Its wide range of capabilities make INTECRIO-VP an ideal tool for virtual prototyping.
- Virtual prototyping with different development artifacts
- System models can be analyzed without complex prototyping hardware
- Use of MiL simulations in the early stage of development to validate and pre-calibrate new functions
- Additional integration of vehicle and environmental models alongside the function models
- The simulation time can be scaled in INTECRIO-VP using the slow-motion/fast-motion function in processes
Use Case: Fuel cell development at GM
The ECU software development and the plant modeling teams at GM Fuel Cell Activities (FCA) used Simulink® and Stateflow by The Mathworks for the representation of both the ECU fuel cell system controls model and the fuel cell system plant model. Since there were no legacy fuel cell algorithms, the controls software development environment had to be completely model-based. Models that had been validated in previous studies with the aid of rapid prototyping hardware were able to be used and the implementation of the control for the ECU derived from them. A modified ECU integration environment for engine controls was used to embed the automatically generated control algorithm code into an existing ECU software infrastructure.
The physical system must be replaced by a comprehensive, high-fidelity plant model of the entire fuel cell system to enable control algorithm testing in Model-in-the-Loop (MiL) and Hardware-in-the-Loop (HiL) environments.
INTECRIO replaces the ECU with an efficient MiL simulation. The MiL environment is used both by developers of the control algorithms and plant models and by software testers as a joint platform. With the aid of a configuration management system, the individual groups can change the MiL environment independently of one another.
GM-FCA uses INTECRIO to integrate the separate control and plant models. At the beginning of the INTECRIO evaluation, the controls model was based on MATLAB® version R14SP2 but the plant model on version 2006b. A simple MATLAB® script brings the inports and outports of the control and plant models into conformity with INTECRIO to prepare the models for connection as INTECRIO modules. If the names of the inport and outport combinations are the same, the plant and controls can be automatically connected with one another in INTECRIO.
INTECRIO supports the integration of code that was generated by Real Time Workshop (RTW) or Embedded Coder (EC) based on several submodels with different MATLAB®/Simulink® versions. The behavior of the integrated model can then be tested in a simulation on the PC.
Developments and tests can be carried out at an early stage in the development process and results reused once they have been generated.
As the powerful INCA and LABCAR tools can be easily grouped around INTECRIO, the same tools can be used for measurement and calibration as well as for testing and test automation in the various development stages.
In addition, INTECRIO accelerates the simulation and reduces the turnaround times relating to the change and the subsequent retesting of the models.
More details: Virtual Prototyping at GM
Use Case: Software-in-the-Loop through co-simulation of the EDC model and GT model using a MAN in-line six-cylinder engine of the D2676LF 25 series as an example
Today’s engine controls, such as the Bosch EDC Electronic Diesel Control system, contain hundreds of functions, the purpose of which also consists in ensuring compliance with the statutory exhaust emission limits. As an ideal test vehicle for Software-in-the-Loop applications, a MAN in-line six-cylinder engine of the D2676LF 25 series (piston displacement: 12.4 l, power: 353 kW at 1900 min-1) was used on account of its high complexity. Engines of this type are used in the TGS and TGX heavy truck series. Due to the two-stage turbocharging with charge-air cooling and intercooler, high-pressure common-rail fuel injection system, lambda-controlled cooled exhaust gas recirculation (EGR), and exhaust gas treatment consisting of an oxidation catalyst, a diesel particle filter (DPF), and a SCR catalyst (Selective Catalytic Reduction), this engine falls below the EURO 6 exhaust emission limits.
The development and parametrization of the functions is usually carried out on the physical ECU in a highly complex and time-consuming procedure that sometimes requires the integration of further hardware and drive components on real test benches.
A virtual test bench made the connection to existing components and hardware superfluous. This flexibility could be used to arrange the development process more efficiently.
With the aid of the ETAS INTECRIO-IP integration and configuration platform and the INTECRIO-VP add-on, a virtual development environment was created (also as a virtual test bench) for the application scenarios of calibration, function development, and optimization. To this end, plant models (e.g. GT-Suite, Matlab®/Simulink®) on the one side were connected with software models (Bosch EDC Electronic Diesel Control) on the other side and executable code generated that could ultimately be controlled with INCA.
The virtual test bench used consisted of simulatable engine and exhaust gas treatment models coupled with software that is also PC-simulable.
The solution opens up a broad range of applications both in the development of new functions and in the application of new or existing functions, while at the same time maintaining high reproducibility and adaptability of the tests. The effort required for the parametrization of time-consuming processes such as the loading of the diesel particle filter (DPF) can be greatly reduced, as test cycles could be calculated in a virtual environment in a fraction of the real time.
The simulation of the software with or without appropriate plant models on a virtual test bench allowed genuine front-loading in the various phases of the development process. Functions can be parametrized and/or tested offline and without special hardware in a very early phase without it being necessary to use the real target. The virtual test bench is operated in a usual environment with the INCA application tool; this substantially increases user-friendliness and thus user acceptance.
As the PC-compatible software modules used are identical to the real target software components, and as they are executed on a real-time operating system, the realistic behavior of the software models is ensured. The results that were able to be achieved on the virtual test bench crucially depend on the quality of the plant models used. In this case, use of plant models from engine development guaranteed good model availability with a suitably high model quality.
There are various other application scenarios for the virtual development environment. Besides the classic calibration of functions, automated optimization can be carried out with this tool with the aid of optimization tools. In addition, individual functions can be exchanged at any time, thanks to the modular structure of the virtual environment. This means the development environment can also be used in function development.