Optimizing Electric and Hybrid Vehicles Efficiency with Model Based DesignOptimizing Electric and Hybrid Vehicles Efficiency with Model Based Design
The need for highly efficient electric and hybrid vehicles is pushing automakers to introduce new vehicle models with entirely new architecture. The e-Tron platform from Audi or the modular electric drive matrix (MEB) from Volkswagen are ones of the most recent examples. As the automakers face extreme time to market pressure as well as permanent quest for development costs reduction in a competitive environment, there is a strong need for a holistic approach to designing new vehicle platform and, especially, efficient systems. Model based design is a pragmatic approach to achieving this.
Today’s vehicles are much more tightly packed with electronics and different electrical attributes. Optimizing the drive system with multiple electrical components and a multitude of interactions between these components, has led to a complex development process and architectural choices. The need for a strong software platform to control the different electronic components and their interactions efficiently and seamlessly has become more critical than ever. The optimization of the hardware to achieve the highest efficiency and performance while lowering the overall system costs, has increased the complexity of the software dramatically. This has a profound impact on the development times for the various car systems and the challenge for the automakers is to achieve high levels of optimization to increase the vehicle range, lower the costs and reduce time taken to charge battery while minimizing the overall development time.
Challenges with the conventional system development process
Figure 1 A V-diagram showing System Implementation stages in an Automotive design cycle1
A typical automotive system design cycle is shown in Figure 1. The vehicle manufacturers are looking for a novel approach which
- Models and simulates the interactions between the subsystems
- Analyzes the variations in systems and evaluate the sensitivity of the systems depending on various calibration parameters
- Provides a platform to develop the software controls without the actual hardware availability
- Detects and fixes errors much early in the development cycle and avoids expensive and time-consuming redesign
Model Based Design Approach – Tackling the key drawbacks with the traditional Automotive System Design
Model-based design is a holistic approach of modeling system requirements, design, analysis, verification and validation activities beginning in the conceptual design phase and continuing throughout development and later life cycle phases. The Model-Based Design paradigm is significantly different from traditional design methodology. The model-based design process starts by constructing a virtual model taking in the system requirements. The designers can formulate advanced functional characteristics by using continuous-time and discrete-time computational building blocks.
Each block has certain inputs and functions and the whole system is modelled as an interconnected building block. For example, one could be a motor, one could be a battery, and power switches, etc. For each of those functional blocks, the functional attributes representing their characteristics are modelled, expanding each block into its sub-modules, controllers etc. These models and associated simulation support tools can provide rapid prototyping, virtual functional verification, and software testing and hardware/software validation.
Figure 2 Model Based Design
Different Phases of a Model-Based Design
The various phases of a model-based design approach are illustrated in Figure 2.
Model in the Loop (MIL)
Model-in-the-loop testing (MIL) and simulation is a technique where reference virtual models are built for the different subsystems of an automotive system. These models have a high level of abstractions, but encapsulates all the critical parameters of a real-life system, thereby providing a strong platform to evaluate not only the functionality but also the efficiency and performance of the overall system. The various models of the subsystems and the interactions between them are simulated in a modelling framework without any physical hardware components. This allows for rapid testing at the beginning stages of the development cycle. It also validates the requirements used as the initial point of the development process.
Software in the Loop (SIL)
In this phase, executable code such as the control algorithms are coded in a software language and are typically auto generated by tools. This code is used in the simulation environment and tested along with the rest of the system models. This type of validation is especially useful for the software components consisting of a combination of generated code and handwritten code that need to be integrated and executed on the embedded target platform. The SIL tests typically reuse the data and model structures from MIL for examining the code behavior in simulation.
Processor in the Loop (PIL)
During this phase, the control algorithms generated from the SIL phase are ported on to the embedded target platform such as FPCU which can directly communicate with the rest of the system models. This allows execution of large number of verification and validation test suites to FPCU’s capability to run control algorithms. This step can be further used to optimize the code, as well as fine tune other parameters like memory utility, the processor performance etc. on the FPCU.
Hardware in the Loop (HIL)
The last phase involves replacing the models with an actual system. While an actual system will provide more accurate results, a lab prototype of the system or a real-time simulator of the system could be used to if the actual system is not available yet. For example, if an inverter controller is being designed then the controller code is deployed on the FPCU board that is then interfaced to an inverter by connecting the inputs and outputs at the interface points such as sensors and power electronics components. HIL testing allows rapid evaluation and debugging in a real-time environment, as well as fine tuning (calibrate) the various parameters to achieve highest performance and efficiency.
Advantages of Model-Based Design in automotive development cycles
Through virtual prototyping, system engineers can easily try out several architectural models iteratively and model several parameters of the principal design elements (e.g. e-Motor power, e-Motor position on the drivetrain, battery size, etc.). Once an architecture is finalized, the model can be further optimized for the overall efficiency, size, weight and cost of the system. By simulating the virtual models, it then makes possible to verify whether the whole system (mechanical, electrical, hydraulic, and pneumatic components, plus embedded software) will work as intended, even before the hardware is manufactured and available for testing.2
Model-based design for optimizing the efficiency of electric and hybrid vehicles
As illustrated in our previous blog post, the WLTP came into full effect for all new car registrations, since sep’18. Designing to adhere to this stringent standard, requires a strong development framework, where all the real-world driving conditions are modelled accurately for the entire automotive subsystem. The model can be simulated iteratively with ease to fine tune the various parameters along with the embedded software, before any physical design is implemented. The model-based design approach is key to achieving this.
At Silicon Mobility, we have a strong design framework enabling rapid model-based design adaption by OEMs and Tier 1s. Our OLEA® COMPOSER is a robust development framework using best of industry EDA tools, software and hardware parameters, selectable upon project needs of the customer and supports all steps of the customer ‘s V-MODEL development cycle.
For example, our OLEA® APP INVERTER HE has been comprehensively tuned within a model of a 400V Battery Electric Vehicle following the WLTP cycle. The simulations shown a 30% extension of the vehicle range when compared to a conventional control system. The vehicle model is based on a P4 architecture where only the digital control part is modified to measure the efficiency improvement. The table below presents the detailed characteristics of the vehicle model:
Stay tuned! In our next blog note, we will see how a better electric motor control increases range and can help reduce the cost of an electric vehicle.
Check out www.silicon-mobility.com for more details.
1 Figure source – Research gate
For more information on Silicon Mobility and the OLEA product suiteContact us
未分类 五月 16, 2019
Silicon Mobility and Valeo successfully finalized the GMG-EHL project on mild-hybrid electric vehicles
Silicon Mobility ann...Read more