Search Results

It is not news anymore when somebody talks about increasing software content in today's vehicles, transportation systems and machinery. The software content and complexity has grown so tremendously and rapidly that even the most advanced product/software development techniques leave more to desire in view of evolving product life-cycles, feature content and need for development efficiency. Model-Based Design (MBD) techniques and V-Cycle based development processes address the significant need for managing complexity, and to some extent, efficiency in product development. Further efficiency in the development process can be achieved by enabling virtual validation of software components. The virtual validation environment for software not only has the ability to run the software component as a standalone unit for performance validation, but is also extended to the validation of the performance of the entire embedded software of an ECU, multiple ECUs and the entire system.

Data design is a very critical step in the model-based software development process. This paper discusses the importance of a data dictionary in modeling, automatic code generation and in creating model dependency graphs.

Manual ‘porting” of source code is often required in order to “reuse” control software in different applications with different target hardware. This process is not cost effective. Maintaining multiple “versions” of the same software also causes problems. This paper describes a way in which multiple target source code can be generated from a single model. A custom data class is developed so that it can be used to define both signal and parameter data types necessary for data dictionary-driven models. This capability allows a single model to be used to generate code for multiple target hardware architectures. A software development process using a generic model to support multiple hardware targets is compared with the hand porting process (e.g. floating-point to/from fixed-point). Auto code generation from a sample multi-target feature model will be presented. The efficiency of the auto code will also be discussed.

Hybrid electric vehicles (HEV) typically have complex interaction between different powertrain devices and incorporates complex electronic control unit (ECU) network. For the electrified vehicles of the future, comprehensive ECU tests are more necessary than ever before, as the complexity and amount of embedded software increases at a breathtaking speed. The V-cycle is a widely recognized approach in the development of ECUs spanning from offline controller development to the final implementation on production hardware. This paper describes a case study with a focus on electric drive components and electric drive controller development. The V-cycle for motor controller development involving phases control design, rapid control prototyping and target implementation is explained in detail. The model components needed for HEV simulation were selected from dSPACE Automotive Simulation Models (ASM).

This presentation focuses on several examples of partnerships between tool suppliers and embedded software developers in which state-of-the-art tools are used to optimize a variety of electric and hybrid vehicle architectures. Projects with Automotive OEMs, Tier One Suppliers as well as with academic institutions will be described. Due to the growing complexity in multiple electronic control units (“ECUs”) inter-communicating over numerous network bus systems, combined with the challenge of controlling and maintaining charges for electric motors, vehicle development would be impossible without use of increasingly sophisticated tools. Hybrid drive trains are much more complex than conventional ones, they have at least one degree of freedom more.

The Aerospace and Defense industry is currently challenged in multiple ways - cost cutting and sequestration on the defense side, and spurt of growth on the commercial aviation side of business. While these are opposing trends, both will impose severe challenges to the management of product development process for both the Air framers and the suppliers. The challenge becomes severe as the innovation expectations become rapid with increases in embedded software content in avionics and the advent of a new category of autonomous ground, marine, and air systems. Clearly, the industry need is to have a product development process that allows for reducing costs, while increasing embedded software quality and thereby product quality even in an iterative development process.