AbstractComplex technological artefacts are often decomposed into smaller components to keep their design manageable. The resulting challenge is to coordinate decisions that involve multiple components and to design components such that high-level targets are met. Analytical Target Cascading (ATC) is an analytical coordination method for the optimization of decomposed systems, which we aim to incorporate in systems engineering design process. To this extent, we relate the domain of engineering optimization to the domain of requirements engineering, and propose a method that constructs an ATC problem from functional specifications and requirements written in the newly developed Elephant Specification Language. The proposed method is demonstrated in the two-level design of an automotive powertrain. This contribution is a step towards design automation and is expected to increase the usability of decomposed optimization techniques.
Abstract
Gilead et al. offer a thoughtful and much-needed treatment of abstraction. However, it fails to build on an extensive literature on abstraction, representational diversity, neurocognition, and psychopathology that provides important constraints and alternative evidence-based conceptions. We draw on conceptions in software engineering, socio-technical systems engineering, and a neurocognitive theory with abstract representations of gist at its core, fuzzy-trace theory.
Multi-level Decomposed Systems Design: Converting a Requirement Specification into an Optimization Problem
