An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme

Jinglei Guo; Miaomiao Shao; Shouyong Jiang; Shengxiang Yang

doi:10.1145/3377929.3389958

An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme

Jinglei Guo, Miaomiao Shao, Shouyong Jiang, Shengxiang Yang^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Published conference contribution

2 Citations (Scopus)

Abstract

The multiobjective evolutionary algorithm based on decomposition (MOEA/D) decomposes a multiobjective optimization problem (MOP) into a number of single-objective subproblems. Penalty boundary intersection (PBI) in MOEA/D is one of the most popular decomposition approaches and has attracted significant attention. In this paper, we investigate two recent improvements on PBI, i.e. adaptive penalty scheme (APS) and subproblem-based penalty scheme (SPS), and demonstrate their strengths and weaknesses. Based on the observations, we further propose a hybrid penalty sheme (HPS), which adjusts the PBI penalty factor for each subproblem in two phases, to ensure the diversity of boundary solutions and good distribution of intermediate solutions. HPS specifies a distinct penalty value for each subproblem according to its weight vector. All the penalty values of suboroblems increase with the same gradient during the first phase, and they are kept unchanged during the second phase.

Original language	English
Title of host publication	GECCO 2020 companion
Subtitle of host publication	Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion
Publisher	Association for Computing Machinery, Inc
Pages	165-166
Number of pages	2
ISBN (Electronic)	9781450371278
DOIs	https://doi.org/10.1145/3377929.3389958
Publication status	Published - 8 Jul 2020
Event	2020 Genetic and Evolutionary Computation Conference, GECCO 2020 - Cancun, Mexico Duration: 8 Jul 2020 → 12 Jul 2020

Conference

Conference	2020 Genetic and Evolutionary Computation Conference, GECCO 2020
Country/Territory	Mexico
City	Cancun
Period	8/07/20 → 12/07/20

Bibliographical note

Funding Information:
This work is part funded by the National Natural Science Foundation of China (No.61673331), the open fund from Key Lab of Digital Signal and Image Processing of Guangdong Province (No.2019GDDS IPL-04) and the Fundamental Research Funds for the Central Universities (No.CCNU20TS026).

Keywords

Adaptive penalty scheme
Decomposition
Hybrid penalty scheme
Multiobjective evolutionary algorithm
Penalty boundary intersection
Subproblem-based penalty scheme

Access to Document

10.1145/3377929.3389958Licence: Unspecified

Cite this

An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme. / Guo, Jinglei; Shao, Miaomiao; Jiang, Shouyong et al.
GECCO 2020 companion: Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion. Association for Computing Machinery, Inc, 2020. p. 165-166.

Research output: Chapter in Book/Report/Conference proceeding › Published conference contribution

Guo, J, Shao, M, Jiang, S & Yang, S 2020, An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme. in GECCO 2020 companion: Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion. Association for Computing Machinery, Inc, pp. 165-166, 2020 Genetic and Evolutionary Computation Conference, GECCO 2020, Cancun, Mexico, 8/07/20. https://doi.org/10.1145/3377929.3389958

@inproceedings{97dbc42a09c8435497b9c980535bc38e,

title = "An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme",

abstract = "The multiobjective evolutionary algorithm based on decomposition (MOEA/D) decomposes a multiobjective optimization problem (MOP) into a number of single-objective subproblems. Penalty boundary intersection (PBI) in MOEA/D is one of the most popular decomposition approaches and has attracted significant attention. In this paper, we investigate two recent improvements on PBI, i.e. adaptive penalty scheme (APS) and subproblem-based penalty scheme (SPS), and demonstrate their strengths and weaknesses. Based on the observations, we further propose a hybrid penalty sheme (HPS), which adjusts the PBI penalty factor for each subproblem in two phases, to ensure the diversity of boundary solutions and good distribution of intermediate solutions. HPS specifies a distinct penalty value for each subproblem according to its weight vector. All the penalty values of suboroblems increase with the same gradient during the first phase, and they are kept unchanged during the second phase.",

keywords = "Adaptive penalty scheme, Decomposition, Hybrid penalty scheme, Multiobjective evolutionary algorithm, Penalty boundary intersection, Subproblem-based penalty scheme",

author = "Jinglei Guo and Miaomiao Shao and Shouyong Jiang and Shengxiang Yang",

note = "Funding Information: This work is part funded by the National Natural Science Foundation of China (No.61673331), the open fund from Key Lab of Digital Signal and Image Processing of Guangdong Province (No.2019GDDS IPL-04) and the Fundamental Research Funds for the Central Universities (No.CCNU20TS026). ; 2020 Genetic and Evolutionary Computation Conference, GECCO 2020 ; Conference date: 08-07-2020 Through 12-07-2020",

year = "2020",

month = jul,

day = "8",

doi = "10.1145/3377929.3389958",

language = "English",

pages = "165--166",

booktitle = "GECCO 2020 companion",

publisher = "Association for Computing Machinery, Inc",

}

TY - GEN

T1 - An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme

AU - Guo, Jinglei

AU - Shao, Miaomiao

AU - Jiang, Shouyong

AU - Yang, Shengxiang

N1 - Funding Information: This work is part funded by the National Natural Science Foundation of China (No.61673331), the open fund from Key Lab of Digital Signal and Image Processing of Guangdong Province (No.2019GDDS IPL-04) and the Fundamental Research Funds for the Central Universities (No.CCNU20TS026).

PY - 2020/7/8

Y1 - 2020/7/8

N2 - The multiobjective evolutionary algorithm based on decomposition (MOEA/D) decomposes a multiobjective optimization problem (MOP) into a number of single-objective subproblems. Penalty boundary intersection (PBI) in MOEA/D is one of the most popular decomposition approaches and has attracted significant attention. In this paper, we investigate two recent improvements on PBI, i.e. adaptive penalty scheme (APS) and subproblem-based penalty scheme (SPS), and demonstrate their strengths and weaknesses. Based on the observations, we further propose a hybrid penalty sheme (HPS), which adjusts the PBI penalty factor for each subproblem in two phases, to ensure the diversity of boundary solutions and good distribution of intermediate solutions. HPS specifies a distinct penalty value for each subproblem according to its weight vector. All the penalty values of suboroblems increase with the same gradient during the first phase, and they are kept unchanged during the second phase.

AB - The multiobjective evolutionary algorithm based on decomposition (MOEA/D) decomposes a multiobjective optimization problem (MOP) into a number of single-objective subproblems. Penalty boundary intersection (PBI) in MOEA/D is one of the most popular decomposition approaches and has attracted significant attention. In this paper, we investigate two recent improvements on PBI, i.e. adaptive penalty scheme (APS) and subproblem-based penalty scheme (SPS), and demonstrate their strengths and weaknesses. Based on the observations, we further propose a hybrid penalty sheme (HPS), which adjusts the PBI penalty factor for each subproblem in two phases, to ensure the diversity of boundary solutions and good distribution of intermediate solutions. HPS specifies a distinct penalty value for each subproblem according to its weight vector. All the penalty values of suboroblems increase with the same gradient during the first phase, and they are kept unchanged during the second phase.

KW - Adaptive penalty scheme

KW - Decomposition

KW - Hybrid penalty scheme

KW - Multiobjective evolutionary algorithm

KW - Penalty boundary intersection

KW - Subproblem-based penalty scheme

UR - http://www.scopus.com/inward/record.url?scp=85089744631&partnerID=8YFLogxK

U2 - 10.1145/3377929.3389958

DO - 10.1145/3377929.3389958

M3 - Published conference contribution

AN - SCOPUS:85089744631

SP - 165

EP - 166

BT - GECCO 2020 companion

PB - Association for Computing Machinery, Inc

T2 - 2020 Genetic and Evolutionary Computation Conference, GECCO 2020

Y2 - 8 July 2020 through 12 July 2020

ER -

An improved multiobjective optimization evolutionary algorithm based on decomposition with hybrid penalty scheme

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this