Architectures of Computational Verb Controllers: Towards a New Paradigm of Intelligent Control

There are at least two kinds of methods of designing (computational) verb proportional-integral-derivative(PID) controllers. The first one is to generalize fuzzy PID controllers by substituting “BE” in fuzzy control rules with non-static computational verbs. The second one is to directly construct evolving functions of verbs in human experiences of dynamic processes. These methods will lead to different architectures of verb PID controllers because computational verb theory can be used in different situations where adjectives or noun phrases have different computational interpretations. Three examples of these computational interpretations of nouns are crisp, fuzzy and probabilistic. Correspondingly, a PID controller can be designed based on crisp or fuzzy representations of human experiences. If the dynamic aspects of these human experiences are emphasized explicitly, the resulting design principle must take advantage of the computational verb theory. This is the principle behind the architectures of the verb PID controllers. The basics of fuzzy logic and fuzzy PID controllers can help the reader to understand this paper. However, the fuzzy logic is only one of at least three methods of interpreting computational nouns used together with computational verbs. Therefore, fuzzy logic is not an essential component for designing computational verb controllers though it is a very helpful component. In this paper, a detailed method of designing verb PID controllers by generalizing fuzzy PID controllers are presented. The static and local fuzzy rules are transformed into dynamic and global verb rules by representing fuzzy rule sets on phase plots. The control trajectories are chunked into a set of computational verbs based on the dynamic experiences generated by the relations among a group of fuzzy control rules. Using these computational verbs, three sets of control rules, respectively for P-, Iand D-components, based on computational verbs(verb rules for short) are generated from the dynamic experiences modelled by using mental experiments. To implement these verb rules, verb implication relations are constructed based on verb similarity functions which are generated based on a set of standard verbs for modelling dynamic experiences. The last step to implement verb PID controllers is to calculate the control verb based on the observe verb and verb rule inference called verb generalized modus ponens(GMP). The results of a verb GMP is a verb similarity which is the union of a set of verb similarity functions modified by verb rule firing adverbs. At the symbolic level, the resulting verb similarity of a verb GMP corresponding to a control verb. The control verb must be deverbified into a conventional control signal before the control loop can be closed in the physical world. Simulation results for controlling a delayed second-order process are used to demonstrate the design processes for verb P-, PDand PID controllers. Copyright c © 2004Manuscript received October 21, 2003; revised February 13, 2004; July 15, 2004. Department of Electrical Engineering and Computer Sciences, Yang’s Scientific Research Institute, 1303 East University Blvd. # 20882 Tucson, Arizona 85719-0521, USA. Email: [[email protected]] Publisher Item Identifier S 1542-5908(05)10210-3/$20.00 Copyright c ©2004-2005 Yang’s Scientific Research Institute, LLC. All rights reserved. The online version posted on June 1, 2004 at http://www.YangSky.com/ijcc32.htm 2005 Yang’s Scientific Research Institute, LLC. All rights reserved.