UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Knowledge based dynamic restructuring of flexible production systems Gu, Jianhua

Abstract

This thesis addresses high level automation in flexible production systems (FPS). In a flexible production system, for example in a dedicated FPS for fish processing, the process components are typically organized as distinct workcells where each workcell is responsible for a special class of production activity. Changes in production demand, variation in raw material supply, and malfunction or failure of workcell components can change the workload on some components in the workcell. A technique that is termed dynamic restructuring, which enables an FPS to change its configuration automatically to suit new situations, thereby achieving a near-optimal load distribution among system components is developed in the thesis. Characteristics of the restructuring problem are known to be fuzzy (both restructuring goal and FPS status), complex, and non-analytic. A knowledge-based method is employed for solving this problem rather than the conventional mathematical methods. Different types of knowledge sources are organized in a three-level hierarchy of decision making: • The first (highest) level determines whether there should be a restructuring oper-ation according to the current FPS status. • The second level determines a restructuring method, in which, due to the fuzziness of the restructuring goal, general heuristics are used. • The third level selects an optimal action based on the current situation, employing fuzzy logic to handle system uncertainty. Various approaches of knowledge representation and reasoning are used, as appro-priate in different levels of the hierarchy. A blackboard architecture is implemented to coordinate associated knowledge sources, which makes the entire system easily ex- pandable. Heuristics are organized in the order of their priorities, in the planning of restructuring actions. Techniques of fuzzy associative memory are developed to evaluate a set of possible restructuring actions from which a decision maker could pick an optimal one. Therefore, the system is intelligent in the sense that the decision maker always selects a proper action based on the current system status. The FPS restructuring system is implemented in Prolog. The approach is applied to an automated fish processing plant, in computer simulation. Several special cases are studied through simulation experiments, to demonstrate the practical use of the approach.

Item Media

Item Citations and Data

Rights

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.