Virtual Prototyping of Self-optimizing Mechatronic Systems

1. Design challenges of self-optimizing mechatronic systems The term mechatronics expresses the mergence of technologies from mechanical engineering, electronics and control design in modern industrial products. These innovative products are based on the combined effect of mechatronic system elements (also referred to as "solution elements" [GL00]). Future systems will comprise configurations of solution elements with an inherent partial intelligence. These components are relying on mathematical models of optimization as well as behavioral optimization, e.g., cognitive capabilities and case-based reasoning. We call this new class of systems “self-optimizing systems” and define the self-optimization of a technical system as an endogenous modification of a goal vector based on changing environmental settings which result from a goalcompliant autonomous adaptation of the structure, the behavior and the parameters of the system. For this reason, self-optimization reaches far beyond conventional strategies for rules and adaptations. Self-optimization allows for manageable systems with inherent intelligence that are able to react autonomously and are flexible to changing environmental settings. The structure of self-optimizing systems is based on the structure of mechatronic systems, in that self-optimizing information processing is superimposed on the controlling mechatronic information processing. This applies at each level of the hierarchical structure that makes up a complex mechatronic system: e.g., mechatronic function modules (MFM) like an intelligent suspension strut, autonomous mechatronic systems (AMS) like a car, and networked mechatronic systems (VMS) like a vehicle convoy. The complexity of self-optimizing systems and the necessity to efficiently analyze and explore a large number of potential behavior patterns requires the creation of new development methods and tools. In this paper we introduce a new concept for the solution element based design and analysis of selfoptimizing systems. Our tool uses virtual reality, simulation and visualization techniques to facilitate a more intuitive approach to virtual prototyping in the conceptual design phase. It also supports the communication within interdisciplinary design teams which are typically required in the development of self-optimizing mechatronic systems. For the analysis of self-optimizing systems discreet and continuous simulation models of different levels of abstraction must be assigned to the solution elements. Our approach builds on the VDI design guideline for mechatronic systems (VDI recommendation 2206 [VDI03]) that defines a procedural model consisting of two main elements: The V-model on the macro level to structure the process, a general problem solution cycle on the micro level, and predefined process components for recurring design activities [GM03].