Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design
The multi-objective optimization (MOO) of complex systems remains a challenging task in engineering domains. The methodological approach of applying MOO algorithms to simulation-enabled models has established itself as a standard. Despite increasing in computational power, the effectiveness and effi...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Information |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2078-2489/14/5/299 |
| _version_ | 1797599766370058240 |
|---|---|
| author | Herbert Palm Lorin Arndt |
| author_facet | Herbert Palm Lorin Arndt |
| author_sort | Herbert Palm |
| collection | DOAJ |
| description | The multi-objective optimization (MOO) of complex systems remains a challenging task in engineering domains. The methodological approach of applying MOO algorithms to simulation-enabled models has established itself as a standard. Despite increasing in computational power, the effectiveness and efficiency of such algorithms, i.e., their ability to identify as many Pareto-optimal solutions as possible with as few simulation samples as possible, plays a decisive role. However, the question of which class of MOO algorithms is most effective or efficient with respect to which class of problems has not yet been resolved. To tackle this performance problem, hybrid optimization algorithms that combine multiple elementary search strategies have been proposed. Despite their potential, no systematic approach for selecting and combining elementary Pareto search strategies has yet been suggested. In this paper, we propose an approach for designing hybrid MOO algorithms that uses reinforcement learning (RL) techniques to train an intelligent agent for dynamically selecting and combining elementary MOO search strategies. We present both the fundamental RL-Based Hybrid MOO (RLhybMOO) methodology and an exemplary implementation applied to mathematical test functions. The results indicate a significant performance gain of intelligent agents over elementary and static hybrid search strategies, highlighting their ability to effectively and efficiently select algorithms. |
| first_indexed | 2024-03-11T03:38:57Z |
| format | Article |
| id | doaj.art-cdb0d6470f02423a9a61322269a90897 |
| institution | Directory Open Access Journal |
| issn | 2078-2489 |
| language | English |
| last_indexed | 2024-03-11T03:38:57Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Information |
| spelling | doaj.art-cdb0d6470f02423a9a61322269a908972023-11-18T01:48:19ZengMDPI AGInformation2078-24892023-05-0114529910.3390/info14050299Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm DesignHerbert Palm0Lorin Arndt1Systems Engineering Laboratory, University of Applied Sciences, Lothstrasse 64, 80335 München, GermanySystems Engineering Laboratory, University of Applied Sciences, Lothstrasse 64, 80335 München, GermanyThe multi-objective optimization (MOO) of complex systems remains a challenging task in engineering domains. The methodological approach of applying MOO algorithms to simulation-enabled models has established itself as a standard. Despite increasing in computational power, the effectiveness and efficiency of such algorithms, i.e., their ability to identify as many Pareto-optimal solutions as possible with as few simulation samples as possible, plays a decisive role. However, the question of which class of MOO algorithms is most effective or efficient with respect to which class of problems has not yet been resolved. To tackle this performance problem, hybrid optimization algorithms that combine multiple elementary search strategies have been proposed. Despite their potential, no systematic approach for selecting and combining elementary Pareto search strategies has yet been suggested. In this paper, we propose an approach for designing hybrid MOO algorithms that uses reinforcement learning (RL) techniques to train an intelligent agent for dynamically selecting and combining elementary MOO search strategies. We present both the fundamental RL-Based Hybrid MOO (RLhybMOO) methodology and an exemplary implementation applied to mathematical test functions. The results indicate a significant performance gain of intelligent agents over elementary and static hybrid search strategies, highlighting their ability to effectively and efficiently select algorithms.https://www.mdpi.com/2078-2489/14/5/299multi-objective optimizationcomplex systemsPareto fronthybrid search algorithmsreinforcement learningintelligent agent |
| spellingShingle | Herbert Palm Lorin Arndt Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design Information multi-objective optimization complex systems Pareto front hybrid search algorithms reinforcement learning intelligent agent |
| title | Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design |
| title_full | Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design |
| title_fullStr | Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design |
| title_full_unstemmed | Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design |
| title_short | Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design |
| title_sort | reinforcement learning based hybrid multi objective optimization algorithm design |
| topic | multi-objective optimization complex systems Pareto front hybrid search algorithms reinforcement learning intelligent agent |
| url | https://www.mdpi.com/2078-2489/14/5/299 |
| work_keys_str_mv | AT herbertpalm reinforcementlearningbasedhybridmultiobjectiveoptimizationalgorithmdesign AT lorinarndt reinforcementlearningbasedhybridmultiobjectiveoptimizationalgorithmdesign |