Evolution strategy: A new Time-Variant Mutation for fine local tuning

M. M.A. Hashem; Keigo Watanabe; Kiyotaka Izumi

Evolution strategy: A new Time-Variant Mutation for fine local tuning

M. M.A. Hashem, Keigo Watanabe, Kiyotaka Izumi

Research output: Contribution to conference › Paper › peer-review

Abstract

The inherent strength of the Evolution Strategy (ES) algorithm lies in the mutation operation. In this paper, a new Time-Variant Mutation (TVM) operator is proposed for fine local tuning and high precision solutions. Its action depends upon the age of the populations and its performance is quite different from the Uniform Mutation (UM) operation. The proposed approach makes it possible that the size of mutation steps (i.e., standard deviations) can be dynamically adjusted to the so called `Evolution Window'. The efficacy of the proposed TVM operator is illustrated by solving the dynamic linear-quadratic control problem (a discrete time optimal control model).

Original language	English
Pages	1099-1104
Number of pages	6
Publication status	Published - Dec 1 1997
Externally published	Yes
Event	Proceedings of the 1997 36th SICE Annual Conference - Tokushima, Jpn Duration: Jul 29 1997 → Jul 31 1997

Other

Other	Proceedings of the 1997 36th SICE Annual Conference
City	Tokushima, Jpn
Period	7/29/97 → 7/31/97

ASJC Scopus subject areas

Control and Systems Engineering
Computer Science Applications
Electrical and Electronic Engineering

Cite this

@conference{c5c76c0a39944dd7bc14abf711fee8dd,

title = "Evolution strategy: A new Time-Variant Mutation for fine local tuning",

abstract = "The inherent strength of the Evolution Strategy (ES) algorithm lies in the mutation operation. In this paper, a new Time-Variant Mutation (TVM) operator is proposed for fine local tuning and high precision solutions. Its action depends upon the age of the populations and its performance is quite different from the Uniform Mutation (UM) operation. The proposed approach makes it possible that the size of mutation steps (i.e., standard deviations) can be dynamically adjusted to the so called `Evolution Window'. The efficacy of the proposed TVM operator is illustrated by solving the dynamic linear-quadratic control problem (a discrete time optimal control model).",

author = "Hashem, {M. M.A.} and Keigo Watanabe and Kiyotaka Izumi",

year = "1997",

month = dec,

day = "1",

language = "English",

pages = "1099--1104",

note = "Proceedings of the 1997 36th SICE Annual Conference ; Conference date: 29-07-1997 Through 31-07-1997",

}

TY - CONF

T1 - Evolution strategy

T2 - Proceedings of the 1997 36th SICE Annual Conference

AU - Hashem, M. M.A.

AU - Watanabe, Keigo

AU - Izumi, Kiyotaka

PY - 1997/12/1

Y1 - 1997/12/1

N2 - The inherent strength of the Evolution Strategy (ES) algorithm lies in the mutation operation. In this paper, a new Time-Variant Mutation (TVM) operator is proposed for fine local tuning and high precision solutions. Its action depends upon the age of the populations and its performance is quite different from the Uniform Mutation (UM) operation. The proposed approach makes it possible that the size of mutation steps (i.e., standard deviations) can be dynamically adjusted to the so called `Evolution Window'. The efficacy of the proposed TVM operator is illustrated by solving the dynamic linear-quadratic control problem (a discrete time optimal control model).

AB - The inherent strength of the Evolution Strategy (ES) algorithm lies in the mutation operation. In this paper, a new Time-Variant Mutation (TVM) operator is proposed for fine local tuning and high precision solutions. Its action depends upon the age of the populations and its performance is quite different from the Uniform Mutation (UM) operation. The proposed approach makes it possible that the size of mutation steps (i.e., standard deviations) can be dynamically adjusted to the so called `Evolution Window'. The efficacy of the proposed TVM operator is illustrated by solving the dynamic linear-quadratic control problem (a discrete time optimal control model).

UR - http://www.scopus.com/inward/record.url?scp=0031354583&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031354583&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0031354583

SP - 1099

EP - 1104

Y2 - 29 July 1997 through 31 July 1997

ER -

Evolution strategy: A new Time-Variant Mutation for fine local tuning

Abstract

Other

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this