.png){kind=icon}
.png){kind=icon}
Return
A q-learning assisted particle swarm optimization algorithm with dynamic niching selection for global optimization
.png){kind=icon}
PRE
AI
颜
W
X
梁
DOI：10.1007/s12293-026-00517-2.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Finding the global optimal solution in complex global optimization problems (GOPs) poses a formidable challenge. This paper presents Q-learning-assisted Particle Swarm Optimization with Dynamic Niching Selection (PSO-DNQC), an innovative memetic algorithm designed for addressing intricate global optimization problems. In this method, a dynamic multi-method strategy assisted by Q-learning is designed to select effective niching method from a candidate pool during the evolutionary process. To support this strategy, a performance metric is devised based on the Q-learning rewards to evaluate all the candidate niching methods. Additionally, to improve solution quality, a convergence-enhancement strategy is incorporated into PSO-DNQC acting as a local search operator in memetic algorithm. The effectiveness of the proposed method is verified on 22 benchmark functions from CEC 2019 and CEC 2022 test suites. Experimental results demonstrate the advantages of PSO-DNQC in robustness and convergence compared with eight state-of-the-art algorithms as well as six variants.
Keywords:
Global optimization
Evolutionary algorithm
Particle swarm optimization
Niching techniques
Q-learning
Adaptive selection
Local search
Diversity
Journal
IF:
2.3
Papers:
447
Citations:
718
Organization
Cited Papers