Research of Pareto Set by Genetic Algorithm, Application to Multicriteria Optimization of Fuzzy Controler

نویسنده

  • L. Gacôgne
چکیده

In a multiobjective problem of decision, it is very often difficult to say which solution we want to keep as aggregation for the different goals we have. We present here an adaptation of an evolution strategy which is able to give a set of solutions. First, the user must define all goals as a list of numeric functions to optimize, like cost, quality... During evolution, it is possible to keep some fixed parameters and make a concentration of the research among other parameters. This is possible thanks to a familly of genetic operators defined in relation with the problem. The user may order the different functions of fitness in relation with their importance for him, but it is possible to avoid a too homogeneous population if there is not at all order in it with a randomly mixing of the whole population at each generation. In this setting, user has to choose the most convenient solution, or to direct the research in an other way by changing a part of the list of fitness functions or their order. We make a trial with fuzzy directed robots which must keep a smooth road inside rings and we give some remarks after a lot of running. ZUSAMMENFASSUNG Wenn wir eine Entscheidung treffen haben, wo es eine Vielfalt von Pielen gibt, ist es sehr oft schwierig zu sagen welche Lösung wir behalten wollen als eine Zusammensetzung für die verschiederen Ziele, die wir haben. Wir stellen hier eine Anpassung an einer evolutiven Strategie dar, die eine Reihe von Lösungen ermöglicht. Zuerst einmal, muβ der Verbraucher alle Ziele richtig klarstellen und einreihen und diese Ziele als numerische Functionen ordnen die zu optimisieren sind, wie beisplielweise Kosten und Qualität. Im evolutiven Prozeβ, bleiben einige Variabel unverändert und das Forschen wird auf andere Parameter konzentriert. Dies wird durch eine Familie von genetischen operatoren in Verbindung mit den Problem ermöglicht. Der Verbrauscher mag die Verschidenen Leistung Fonctionen in Verhältnis mit der Bedeutung die er darauf legt, einrein, doch kann er eine zu homogene Bevölkerung verneiden wenn er dieses Bewölkerung durch zufall vermischt. Der Verbraucher hat in diesem Fall, die beste Lösung zu wählen oder das Forschen in eine andere Richtung zu orientieren, indem er einen Teil der Liste umstellt oder aber die Reihe der Leistungfunction. Wir machen einen Versuch mit Robotten die eine weiche Richtung innererhalb eines Ringes behalten und wir verzeichnen einige Bemerkungen nach etlichen Abläufen.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Application of Genetic Algorithm to Determine Kinetic Parameters of Free Radical Polymerization of Vinyl Acetate by Multi-objective Optimization Technique

A Multi-objective optimization procedure has been developed to determine some kinetic parameters of free radical polymerization of vinyl acetate based on genetic algorithm. For this purpose, mathematical modeling of free radical polymerization of vinyl acetate is carried out first and then selected kinetic parameters are optimized by minimizing objective functions defined from comparing exp...

متن کامل

Load Frequency Control in Power Systems Using Multi Objective Genetic Algorithm & Fuzzy Sliding Mode Control

This study proposes a combination of a fuzzy sliding mode controller (FSMC) with integral-proportion-Derivative switching surface based superconducting magnetic energy storage (SMES) and PID tuned by a multi-objective optimization algorithm to solve the load frequency control in power systems. The goal of design is to improve the dynamic response of power systems after load demand changes. In t...

متن کامل

Decision Strategies in Evolutionary Optimization

This paper describes a sequel of the previous experiments on real number genetic algorithm behavior [2], [3]. A particular example of multicriteria optimization is discussed. The behavior of the two previously explored genetic algorithms is compared with a simple evolutionary algorithm. The main idea of the experiments is to stimulate the algorithm to find the Pareto set without measuring domin...

متن کامل

Fuzzy PID Tuned by a Multi-Objective Algorithm to Solve Load Frequency Control Problem

Abstract In this paper, a fuzzy PID with new structure is proposed to solve the load frequency control in interconnected power systems. in this study, a new structure and effective of the fuzzy PID-type Load frequency control (LFC) is proposed to solve the load frequency control in interconnected power systems. The main objective is to eliminate the deviations in the frequency of different area...

متن کامل

Optimization of Heat Transfer Enhancement of a Domestic Gas Burner Based on Pareto Genetic Algorithm: Experimental and Numerical Approach

The present study attempts to improve heat transfer efficiency of a domestic gas burner by enhancing heat transfer from flue gases. Heat transfer can be augmented using the obstacles that are inserted into the flow field near the heated wall of the domestic gas burner. First, to achive the maximum efficiency, the insert geometry is optimized by the multi-objective genetic algorithm so that heat...

متن کامل

Thermal-Economic Optimization of Shell and Tube Heat Exchanger by using a new Multi-Objective optimization method

Many studies are performed by researchers about Shell and Tube Heat Exchanger but the Multi-Objective Big Bang-Big Crunch algorithm (MOBBA) technique has never been used in such studies. This paper presents application of Thermal-Economic Multi-Objective Optimization of Shell and Tube Heat Exchanger Using MOBBA. For optimal design of a shell and tube heat exchanger, it was first thermally model...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2007