COMPUTATIONAL FOUNDATIONS FOR COMPUTER AIDED CONCEPTUAL DESIGN OF MULTIPLE INTERACTION-STATE MECHATRONIC DEVICES
Files
Publication or External Link
Date
Authors
Advisor
Citation
DRUM DOI
Abstract
Increasing autonomy and intelligence in mechatronic devices requires them to be multiple interaction-state devices. Different modes of operations and different types of interactions with the use-environment require the device to have multiple interaction-states, each state capable of producing a different behavior to meet its intended requirements. For multiple interaction-state mechatronic devices, a satisfactory framework does not exist for representing, evaluating, and synthesizing design concepts. Hence, majority of mechatronic designers currently use informal methods for representing and evaluating design concepts during the conceptual design. This leads to the following problems. First, informal representation of design concepts hinders information exchange and reuse. Second, in absence of a validation methodology, it is not clear how to determine if a proposed design concept is consistent with the requirements. Finally, designers cannot perform computer aided evaluation during the conceptual design stage.
This dissertation focuses in the area of computational foundations for representing, validating, evaluating, and synthesizing design concepts of multiple interaction-state mechatronic devices. A modeling and simulation framework has been developed for representing design concepts behind multiple interaction-state mechatronic devices. The problem of consistency-checking of interaction-states has been studied and an algorithm has been developed for solving the interaction consistency-checking problem. The problem of determining the presence of unsafe parameter values has been studied and an algorithm has been developed to determine whether an interaction-state in the proposed design concept can attain unsafe parameter values. Algorithms have been developed for evaluating design concepts based on the maximum power consumption and sharability of components. Finally, algorithms have been developed for automatically synthesizing transition diagrams for meeting the desired behavior specifications, given a components library.
We believe that the results reported in this dissertation will provide the underlying foundations for constructing the next generation computer aided design tools for conceptual design of mechatronic devices. We expect that these tools would streamline the product development process, facilitate information reuse, and reduce product development time.