Energy Management by Constraint Solving Techniques (EMCON)

Keywords: unit commitment, electric power systems, optimization, fuzzy logic

Objectives and approach. The objective in EMCON is to develop a software platform for solving the type of problems, that can be described by constraints. The approach relies on array based logic, a theory developed at the Department of Electric Power Engineering by Ole I. Franksen. A sub-project of EMCON is to extend the logical basis into fuzzy logic. The test example is a unit commitment problem, that is, finding an optimal mix of power production units that will meet the consumer demand. Traditionally this is solved by the power utilities using mathematical programming. The constraint solver uses discrete variables and soft constraints, for example 'approximately equal' instead of equality, 'much greater than' instead of greater than, and 'a little' instead of zero. So far the applications are within power systems, but any configuration problem is a potential application. No other research groups work on the same approach, but there is a lot of activity within constraint programming. The project is headed by Bang & Olufsen, Struer, and Sydkraft AB, Malmö. The project has Eureka status.

Results. The problem of defining suitable fuzzy operations (and, or, nand, nor, if-then, etc.) is partly solved (Jantzen, 1995), and remains to be implemented and tested. An article describes how to apply soft constraints (Jantzen & Eliasson, 1996). The array approach has been compared with linear programming in an Msc project by Jan Mandrup Johansen (1996 in Danish).

Partners. Bo Eliasson, Sydkraft AB, Malmö, Sweden; Gert Møller, Beologic A/S, Cph; O.I.Franksen and P.Falster, Technical University of Denmark, Department of Electric Power Engineering

Contact point. Jan Jantzen.

Array Approach To Fuzzy Logic

J. Jantzen

Technical University of Denmark

Keywords: Approximate Reasoning, Operators, Process Control

Abstract: Promising results from applying an array-based approach to two-valued logic suggests its application to fuzzy logic. The idea is to limit the domain of truth-values to a discrete, finite domain, such that a logical relationship can be evaluated by an exhaustive test of all possible combinations of truth-values. The paper presents a study of the topic from an engineer's viewpoint. As an example 31 logical sentences valid in two-valued logic were tested in three-valued logic using the Nested Interactive Array Language, Nial. Out of these, 24 turned out to be valid in a three-valued extension based on the well known S* implication operator, also called "Gödel's implication operator". Applications to automated approximate reasoning and fuzzy control are also illustrated.

Fuzzy Sets And Systems, 1995, 70, 359-370.

Fuzzy Array Approach to Unit Commitment

J. Jantzen and B. Eliasson

Technical University of Denmark and Sydkraft AB, Sweden

Abstract: The paper investigates the unit commitment problem as a constraint satisfaction problem. The power system is a simplified system with nuclear, thermal, and hydro generators as well as power interchange. The purpose is to accommodate uncertainty by means of soft constraints.

Proc. 4th IEE Int. Conf. on Power System Control And Management, London, 1996, Conf.publ. no 421, 47-51.

Optimering af energiproduktion med array-baseret logik

J.M.Johansen

Keywords: unit commitment, linear programming, constraints

Resume - EMCON er en ny programpakke til analyse og løsning af constraintsproblemer. Programmellets funktioner er baseret på teorien for array-baseret logik. Formålet med projektet er at vurdere, hvorvidt og hvordan programmellet kan anvendes til optimeringsproblemer. Dette undersøges ved analyse og løsning af lineære programmeringsproblemer. Konklusionen er, at optimeringsfunktionen i EMCON bør videreudvikles, da denne er meget primitiv og yderst begrænset i ydeevne. Dog er små LP-problemer løst med en god præcision. Såfremt ydeevnen af optimeringsfunktionen forbedres, kan EMCON blive et særdeles kraftfuldt værktøj til behandling af især ulineære optimeringsproblemer.

DTU: Elteknik, 1996 (eks.arbejde).


Mail to Jan Jantzen