许志伟的主页

亮点成果

TCYB 2021

Kai Zhang, Zhiwei Xu(许志伟), Shengli Xie, and Gary G. Yen*. Evolution Strategy-Based Many-Objective Evolutionary Algorithm Through Vector Equilibrium. IEEE Transactions on Cybernetics , vol. 51, no. 11, pp. 5455–5467, Nov. 2021. (JCR:Q1; IF:11.8)
In this paper, we propose a novel evolution strategy for solving many-objective optimization problems, named MaOES.
The proposed algorithm uses mutation and selection for individual self-adaptation. The Precision Controllable Mutation operator is designed for individuals to explore and exploit the decision space efficiently. The Maximum Extension Distance strategy is tailored to guide the individuals to keep uniform distance among particles in the population and to facilitate the extension to approximate the entire Pareto front automatically.
[Link] [Download][Code]

TEVC 2024

Kai Zhang, Zhiwei Xu(许志伟), Gary G. Yen*, Ling Zhang. Two-Stage Multi-Objective Evolution Strategy for Constrained Multi-Objective Optimization. IEEE Transactions on Evolutionary Computation , vol. 28, no. 1, pp. 17–31, Feb. 2024 (JCR:Q1; IF:14.3)
In this paper, a novel parameter-less constraint handling technique is proposed, which divides the whole population into three mutually exclusive subsets dynamically, including FNDS, the subset dominated by FNDS, and the subset not dominated by FNDS. According to the proposed division of labor, it is not necessary to balance the convergence and constrained satisfaction in each subset.
Secondly, to avoid been trapped into local optima, the proposed algorithm adopts a two-stage strategy to solve CMOPs.
In the first stage, the proposed algorithm focuses solely on converging toward the unconstrained PF without considering the constrained satisfaction.
In the second stage, the FNDS constraint handling technique is adopted to guide the population converging towards PF effectively.
[Link] [Download][Code]

INS 2022

Zhiwei Xu (许志伟), Xiaoming Liu, Kai Zhang*, and Juanjuan He. Cultural transmission based multi-objective evolution strategy for evolutionary multitasking. Information Sciences , vol. 582, pp. 215–242, Jan. 2022. (JCR:Q1; IF：8.1)
In this paper, a novel multi-objective evolution strategy was proposed for solving multitask optimization problems. Inspired by modern cultural evolution theory, elite-guided variation strategy, and horizontal cultural transmission strategy, two evolutionary operators were proposed.
To make full use of the two transfer strategies, an adaptive information transfer strategy is proposed to adjust the probability of the information transfer adaptively according to the dominant relationship between the offspring and its parent to reasonably allocate the evolutionary resources.
[Link] [Download][Code]

INS 2022

Zhiwei Xu (许志伟), Kai Zhang, Juanjuan He*, and Xiaoming Liu. A novel membrane-inspired evolutionary framework for multi-objective multi-task optimization problems. Information Sciences , vol. 596, pp. 236–263, Jun. 2022. (JCR:Q1; IF：8.1)
In this paper, a multi-objective multi-task evolutionary framework based on membrane system (EMT-MOMIEA) is proposed to solve the multi-objective multi-task optimization (MOMTO) problems.
Inspired by the binding process of information molecules and receptors during information exchange between cells, the information molecule concentration vector (IMCV) concept is proposed. The IMCV can dynamically adjust the information transfer probability and reduce negative information transfer.
[Link] [Download]

ASOC 2021

Zhiwei Xu (许志伟) and Kai Zhang*. Multiobjective multifactorial immune algorithm for multiobjective multitask optimization problems. Applied Soft Computing , vol. 107, p. 107399, Aug. 2021. (JCR:Q1; IF：8.7)
In this paper, a novel multiobjective multifactorial immune algorithm (MOMFIA) is proposed to solve MOMTO and MOMaTO problems. The proposed MOMFIA applied a novel multipopulation framework and a novel information transfer method based on the dimensional information of solutions (DIS). The proposed multi-population framework can evenly distribute individuals to different subpopulations, each of which maintains an independent task module, can evolve independently, but is also equipped to transfer their knowledge when necessary.
[Link] [Download]

ASOC 2024

Zhiwei Xu (许志伟)*, Jia feng Xu, Kai Zhang, Xin Xu, Juanjuan He, Ni Wu, Decision Variable Classification based Multi-objective Multifactorial Memetic Algorithm for Multi-objective Multi-task Optimization Problem. Applied Soft Computing , vol. 152, p. 111232, Feb. 2024. (JCR:Q1; IF：8.7)
In this paper, a novel hybrid multi-objective multifactorial memetic algorithm is proposed to solve MOMTO problems. The proposed variable classification method will classify decision variables into convergence-related and diversity-related decision variables. Only the same type of decision variables in the source and target tasks can transfer information to avoid negative transfer.
Different evolutionary operators are adopted according to the characteristics of decision variables during individual recombination.
In addition, the proposed algorithm hybridizes the immune algorithm as the global evolutionary operator and the evolutionary gradient search algorithm as the local search operator into the multifactorial framework to enhance the searching ability.
[Link] [Download]