Research Highlights

[IEEE TCYB] Evolutionary Large-Scale Dynamic Optimization Using Bilevel Variable Grouping (BLVG)

By EMIOn 0 Comments
[IEEE TCYB] Evolutionary Large-Scale Dynamic Optimization Using Bilevel Variable Grouping (BLVG)

Hui Bai, Ran Cheng*, Danial Yazdani, Kay Chen Tan, Yaochu Jin

 
Abstract:

Variable grouping provides an efficient approach to large-scale optimization, and multipopulation strategies are effective for both large-scale optimization and dynamic optimization. However, variable grouping is not well studied in large-scale dynamic optimization when cooperating with multipopulation strategies. Specifically, when the numbers/sizes of the variable subcomponents are large, the performance of the algorithms will be substantially degraded. To address this issue, we propose a bilevel variable grouping (BLVG)-based framework. First, the primary grouping applies a state-of-the-art variable grouping method based on variable interaction analysis to group the variables into subcomponents. Second, the secondary grouping further groups the subcomponents into variable cells, that is, combination variable cells and decomposition variable cells. We then tailor a multipopulation strategy to process the two types of variable cells efficiently in a cooperative coevolutionary (CC) way. As indicated by the empirical study on large-scale dynamic optimization problems (DOPs) of up to 300 dimensions, the proposed framework outperforms several state-of-the-art frameworks for large-scale dynamic optimization. [Source Code]

Results

Overall Comparison

 

Table 1. Comparison results of BLVG and other three frameworks using PSO optimizer on four types of large-scale DOPs. The highlighted entries are significantly better using pairwise Wilcoxon signed-rank tests with Holm-Bonferroni p-value correction with α = 0.05. The results demonstrate that PSOBLVG performs significantly better than the others.

Influence of Different Optimizers

 

Table 2. Comparison results of BLVG, CTR, RG, AND TMMO using the three optimizers, PSO, DE, and CMA-ES, on general large-scale DOPs.

 

 

Influence of Different Variable Grouping Methods

 

Table 3. Comparison results of PSOBLVG and PSOCTR using the two variable grouping methods, DG2 and ERDG, on general large-scale DOPs.

Robustness to Dynamic Changes

 

Table 4. Comparison results of PSOBLVG and PSOCTR on general large-scale DOPs with different peaks number m for each subcomponent randomized in the ranges {1, …, 10}, {1, …, 15} and {1, …, 30}.

 

Table 5. Comparison results of PSOBLVG and PSOCTR on general large-scale DOPs with different shift severity s for each peak in each subcomponent. The values are randomized in the ranges [0.5, 1], [0.5, 3], and [0.5, 5].

 

 

Table 6. Comparison results of PSOBLVG and PSOCTR on general large-scale DOPs with different change frequencies cf: 200D, 500D, and 1000D.

Citation


@ARTICLE{9772492,
  author={Bai, Hui and Cheng, Ran and Yazdani, Danial and Tan, Kay Chen and Jin, Yaochu},
  journal={IEEE Transactions on Cybernetics}, 
  title={Evolutionary Large-Scale Dynamic Optimization Using Bilevel Variable Grouping}, 
  year={2022},
  volume={},
  number={},
  pages={1-14},
  doi={10.1109/TCYB.2022.3164143}}

Acknowledgments:

This work was supported in part by the National Natural Science Foundation of China under Grant 61906081; in part by the Shenzhen Science and Technology Program under Grant RCBS20200714114817264; in part by the Guangdong Provincial Key Laboratory under Grant 2020B121201001; and in part by the Program for Guangdong Introducing Innovative and Entrepreneurial Teams under Grant 2017ZT07X386.

Categories : Evolutionary Computation

发表回复

您的电子邮箱地址不会被公开。 必填项已用*标注

Related Posts