Rapid development of high-fidelity, high-efficiency models of electronic products
Electrical and electronic engineers are faced with increasingly complex problems that existing tool chains struggle to solve.
New design approaches, and the tools to support them, are now emerging to address these issues.
Manufacturers of electronics-based products are facing many challenges to their business: greater demand from customers for more performance and functionality, environmental constraints in terms of electro-magnetic emissions and energy efficiency, and competitive pressures to get new products to market faster – all for less cost. Such demands are pushing existing design processes, the tools within them, and the engineers who use them, to their limit.
These pressures are pushing product designs to greater complexity. A classic case in point is the increasing use of feedback control in many consumer products where open-loop mechanisms were traditionally used.
Examples include the “smart-clean” features in dishwashers and washing machines and autofocus in cameras. In these products, electronics, mechanics, thermal, optical systems and microprocessors need to work together effectively and efficiently.
The greater complexity makes it important to consider the interaction of multiple engineering domains very early in the design process. In general, the earlier design problems can be identified and addressed, the less risk to the final design.
Traditionally, the electronics industry has been ahead of the curve with the emergence of Electronic Design Automation (EDA) tools based on SPICE and VHDL. However, the primary limitation of EDA tools is found right in their description: they are developed for electronics design only.
When it comes to gaining understanding of how an electronic circuit behaves when connected to a mechanical load via a motor, while charging and discharging a chemistry-based battery, all driven by an embedded controller, the options for these tools are very limited. While very useful in one particular aspect of the design, they are not readily scalable to the system-level in order to consider all the subsystems, which utilize different engineering domains, and how they interact with each other.
It is for this reason that multidisciplinary fields like mechatronics are emerging in many teaching and research organizations, and are now spreading into industry. These disciplines are typically control-design oriented, with a strong mix of electronics and mechanical design as well as other engineering domains.
In these situations, Model-Based Design (MBD) tools based on a signal-flow approach are most prevalent. However, MBD tools like Simulink are notoriously difficult for developing electronic system models because the system equations have to be manually derived and then rearranged into a form that can be represented using integrators, gains and feedback loops. This is an approach that has changed little since the days of the analog computer.
As various engineering domains merge in a single product design, the gaps between tools dedicated to individual engineering domains are most keenly felt. This lack has fostered a new approach to product design, and new tools to support it. In particular, a new multidomain approach has been evolving in Europe, and is starting to gain acceptance in the U.S. and elsewhere.
The approach is based on an object-oriented language called Modelica. Primarily, Modelica is a model description language that is declarative in nature, instead of the usual procedural paradigm. This means that the dynamic behavior of a component is defined simply by providing all the mathematical relationships as equations. The values of the variables do not all need to be given, but can be left as unknowns.
Using symbolic computational techniques, the equations are automatically combined and simplified to derive the overall dynamic equations for the system, then solved for the unknown variables.
As the developer of the most advanced symbolic computation technology in the world, we at Maplesoft view the combination of the equation-based structure of the Modelica language and our technologies as the perfect match to help our customers. With our system-level modeling and simulation tool, MapleSim, designers can use the wide array of built-in components, third-party libraries, equation-based custom components, and the Modelica language to build high-fidelity models of their multidomain systems in a single environment, with minimal time and effort.
In summary, a combination of the Modelica language standard and powerful symbolic technologies found in tools like MapleSim can address many of the challenges faced by electronics product manufacturers. As the technology evolves, increasingly specialized libraries will emerge.
Ultimately, other tools will provide the same level of sophistication found in the existing EDA and MBD tools inside a single, system-level, multidomain environment.
Paul Goossens is VP Application Engineering at Maplesoft.