Methods based on ordered fuzzy numbers used in fuzzy control

摘要

The model of the ordered fuzzy numbers (OFN) has been defined and explored in the recent publications of the author and his co-workers (2003 and 2004). It provides a quite simple representation of non-precise information and algebraic operations on them. If one performs many operations on OFN's then the fuzziness of their results is limited in contrast to the operations on standard convex fuzzy numbers. The implementation of OFN in a fuzzy controller is made. The controller makes use of multi-conditional fuzzy rules. New methods of determining their levels of activities are proposed as well inference methods based on special properties of OFNs. Fuzzy control is applied here to a hypothetical process of moving device (can be related to car, robot, etc.). Proposed control uses the readings from two devices: actual speed and distance to a critical (boundary) point, which cannot be achieved. The control variable is the applying acceleration. Goal of the control is to keep the device in a safe distance from the critical point.