Session Description: Fuzzy control system methods have suggested an approach for characterizing uncertainty in the performance of physical processes when actual test data are very limited. Fuzzy control system applications will be reviewed, and the suggested approach will be considered via examples related to probabilistic risk analysis, probabilistic safety assessment, and reliability analysis. The relative merits of a probabilistic approach and a fuzzy logic approach to the general uncertainty-characterization problem will be considered.
Theme Session: (yes or no) No
Applied Session: (yes or no) Yes
Session Chair & Affiliation: Jane M. Booker, Los Alamos National Laboratory
Mailing Address: Mail Stop F600, Los Alamos National Laboratory 87545 (505)667-1479 jmb@lanl.gov FAX (505)667-4470
Session Organizer & Affiliation: Thomas R. Bement, Los Alamos National Laboratory
Mailing Address: Mail Stop F600, Los Alamos National Laboratory 87545 (505)667-7578 trb@lanl.gov FAX (505)667-4470
Participant No. 1 & Affiliation: Mohammed Jamshidi NASA Center for Autonomous Control Engineering University of New Mexico at Albuquerque
Co-Authors & Affiliations (for papers only):
Title of Paper OR Role in Session: FUZZY LOGIC : From a concept to an indusrial revolution
Key Words (for papers only):Fuzzy Logic, fuzzy sets, fuzzy logic control. Abstract (for papers only): One of the biggest challenges of any control paradigm is being able to handle complex systems. A system may be called complex here if its dimension (order) is too high and its model (if available) is nonlinear, interconnected, and uncertain such that classical techniques can not easily handle the system. Examples of complex systems are electric power entwork of USA, US Economy, National Air Traffic network, etc. In this seminar fuzzy sets and fuzzy logic are first introduced and then the impact of fuzzy logic control on the indsutreis in Japan, Europe, and US are discussed. A videotape of the experimental implementation in Japan, Europe and USA will be shown.
Mailing Address (include zip), Phone, Fax, E-mail for Participant No. 1: Mohammed Jamshidi NASA Center for Autonomous Control Engineering University of New Mexico at Albuquerque Albuquerque, New Mexico 87131. FAX : +1 505 277 4681 Tel. : + 1 505 277 5538 Email : jamshidi@unm.edu http://ace.unm.edu
Participant No. 2 & Affiliation: Thomas R. Bement, Los Alamos National Laboratory Co-Authors & Affiliations (for papers only): William Parkinson, Ronald E. Smith, Mary A. Meyer, and Fred N. Mortensen, all of Los Alamos National Laboratory.
Title of Paper OR Role in Session: The Use of Fuzzy Control System Techniques to Develop Uncertainty Distributions Key Words (for papers only): Uncertainty distributions; Fuzzy logic; Control systems Abstract (for papers only): A control system maps observed "plant" output parameter values into required control actions, or plant inputs. In a fuzzy control system, these observed plant outputs are transformed into degrees of membership in fuzzy plant-output sets. "IF-THEN" rules transform these degrees of membership into fuzzy control-action (or plant-input) sets. The precise control action is determined via a defuzzification process such as selecting the centroid of the fuzzy control-action set. A similar process can be applied to the development of uncertainty distributions in applications such as probabilistic risk assessment, probabilistic safety assessment, and reliability analysis. In a safety application for instance, the plant-output parameters used by the control system may become component condition and accident-scenario parameters, and the control-action may become the predicted component response to the accident conditions. With proper formulation, the fuzzy control-action set can be considered to be an uncertainty distribution in the form of a probability density function. We consider and compare various methods of formulating the process just described, with an application in reliability analysis used as an example.
Mailing Address (include zip), Phone, Fax, E-mail for Participant No. 2: Thomas R. Bement Mail Stop F600, Los Alamos National Laboratory 87545 (505)667-7578 trb@lanl.gov FAX (505)667-4470
Participant No. 3 & Affiliation: William H. Woodall University of Alabama - Tuscaloosa
Co-Authors & Affiliations (for papers only):
Title of Paper OR Role in Session: "Comparisons of Fuzzy and Statistical Methods" Key Words: Fuzzy logic, control theory, linear regression, quality control. Abstract (for papers only): Fuzzy methods have been proposed as alternatives to traditional statistical methods in areas such as linear regression, reliability, forecasting, statistical process control, and capability analysis. In addition, fuzzy control theory is widely applied although there are some advantages in using a probabilistic approach. The alleged deficiencies of statistics and probability are examined in several applications. It is shown that the justification given for fuzzy methods is often quite weak. Also, statistical or probabilistic methods typically have advantages over the corresponding fuzzy methods, either in simplicity, performance, or ease of interpretation.
Mailing Address (include zip), Phone, Fax, E-mail for Participant No. 3: William H. Woodall Professor of Management Science and Statistics University of Alabama - Tuscaloosa Tuscacaloosa, AL 35487-0226 (205)348-8992 wwoodall@ua1vm.ua.edu FAX: (205)348-0560
Participant No. 4 & Affiliation: Ronald L. Iman Soutwest Technology Consultants
Co-Authors & Affiliations (for papers only):
Title of Paper OR Role in Session: Discussant Key Words (for papers only): Abstract (for papers only):
Mailing Address (include zip), Phone, Fax, E-mail for Participant No. 4: Ronald L. Iman Soutwest Technology Consultants 1065 Tramway Lane NE Albuquerque, NM 87122 (505) 856-6500 73664.266@compuserve.com
Thomas R. Bement Group TSA-1 MS F600 Los Alamos National Laboratory Los Alamos, NM 87545 Ph: (505)667-7578 FAX: (505)667-4470