Wei Junhao

Mr. Wei Junhao

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 April 2025 in PLOS ONE
Verified
Journal Article
Open access
PDF file
WJWei Junhao
Wei Junhao
YGYanzhao Gu
Yanzhao Gu
BLBaili Lu
Baili Lu
NCNgai Cheong
Ngai Cheong
Whale Optimization Algorithm (WOA) is a biologically inspired metaheuristic algorithm with a simple structure and ease of implementation. However, WOA suffers from issues such as slow convergence speed, low convergence accuracy, reduced population diversity in the later stages of iteration, and an imbalance between exploration and exploitation. To address these drawbacks, this paper proposed an enhanced Whale Optimization Algorithm (RWOA). RWOA utilized Good Nodes Set method to generate evenly distributed whale individuals and incorporated Hybrid Collaborative Exploration strategy, Spiral Encircling Prey strategy, and an Enhanced Spiral Updating strategy integrated with Levy flight. Additionally, an Enhanced Cauchy Mutation based on Differential Evolution was employed. Furthermore, we redesigned the update method for parameter a to better balance exploration and exploitation. The proposed RWOA was evaluated using 23 classical benchmark functions and the impact of six improvement strategies was analyzed. We also conducted a quantitative analysis of RWOA and compared its performance with other state-of-the-art (SOTA) metaheuristic algorithms. Finally, RWOA was applied to nine engineering design optimization problems to validate its ability to solve real-world optimization challenges. The experimental results demonstrated that RWOA outperformed other algorithms and effectively addressed the shortcomings of the canonical WOA.

25 March 2025 in Sensors
Verified
Journal Article
Open access
PDF file
WJWei Junhao
Wei Junhao
YGYanzhao Gu
Yanzhao Gu
YYYuzheng Yan
Yuzheng Yan
ZLZikun Li
Zikun Li
LBLu Baili
Lu Baili
NCNgai Cheong
Ngai Cheong
The Whale Optimization Algorithm (WOA) is a bio-inspired metaheuristic algorithm known for its simple structure and ease of implementation. However, WOA suffers from issues such as premature convergence, low population diversity in the later stages of iteration, slow convergence rate, low convergence accuracy, and an imbalance between exploration and exploitation. In this paper, we proposed an enhanced whale optimization algorithm with multi-strategy (LSEWOA). LSEWOA employs Good Nodes Set Initialization to generate uniformly distributed whale individuals, a newly designed Leader-Followers Search-for-Prey Strategy, a Spiral-based Encircling Prey strategy inspired by the concept of Spiral flight, and an Enhanced Spiral Updating Strategy. Additionally, we redesigned the update mechanism for convergence factor a to better balance exploration and exploitation. The effectiveness of the proposed LSEWOA was evaluated using CEC2005, and the impact of each improvement strategy was analyzed. We also performed a quantitative analysis of LSEWOA and compare it with other state-of-art metaheuristic algorithms in 30/50/100 dimensions. Finally, we applied LSEWOA to nine engineering design optimization problems to verify its capability in solving real-world optimization challenges. Experimental results demonstrate that LSEWOA outperformed better than other algorithms and successfully addressed the shortcomings of the classic WOA.

3 September 2025 in PLOS ONE
Verified
Journal Article
Open access
PDF file
WJWei Junhao
Wei Junhao
YGYanzhao Gu
Yanzhao Gu
ZXZhanxi Xie
Zhanxi Xie
YYYuzheng Yan
Yuzheng Yan
BLBaili Lu
Baili Lu
SZSong Zhang
Song Zhang
Whale Optimization Algorithm (WOA) suffers from issues such as premature convergence, low population diversity in the later stages of iteration, slow convergence rate, low convergence accuracy, and an imbalance between exploration and exploitation. Thus, an enhanced Whale Optimization Algorithm (LSWOA) based on multiple strategies is proposed, aiming to overcome the limitations of the canonical WOA. The performance of the canonical WOA is improved through innovative strategies: first, an initialization process using Good Nodes Set is introduced to ensure that the search starts from a higher-quality baseline; second, a distance-based guided search strategy is employed to adjust the search direction and intensity by calculating the distance to the optimal solution, which enhances the algorithm’s ability to escape local optima; and lastly, LSWOA introduces an enhanced spiral updating strategy, while the enhanced spiral-enveloping prey strategy effectively balances exploration and exploitation by dynamically adjusting the spiral shape parameters to adapt to different stages of the search, thereby more accurately updating the positions of individuals and improving convergence speed. In the experimental section, we validate the efficiency and superiority of LSWOA by comparing it with outstanding metaheuristic algorithms and excellent WOA variants. The experimental results show that LSWOA exhibits significant optimization performance on the benchmark functions with various dimensions. Additionally, LSWOA is tested on seven engineering design optimization problems, and the results demonstrate that it performs excellently in these application scenarios, effectively solving complex optimization problems in different dimensions and showing its potential for a wide range of applications in real-world engineering challenges.

11 August 2025 in Symmetry
Verified
Journal Article
Open access
PDF file
LBLu Baili
Lu Baili
ZXZhanxi Xie
Zhanxi Xie
WJWei Junhao
Wei Junhao
YGYanzhao Gu
Yanzhao Gu
YYYuzheng Yan
Yuzheng Yan
RZRuishen Zhou
Ruishen Zhou
To address the limitations of the Red-billed Blue Magpie Optimization algorithm (RBMO), such as its tendency to get trapped in local optima and its slow convergence rate, an enhanced version called MRBMO was proposed. MRBMO was improved by integrating Good Nodes Set Initialization, an Enhanced Search-for-food Strategy, a newly designed Siege-style Attacking-prey Strategy, and Lens-Imaging Opposition-Based Learning (LIOBL). The experimental results showed that MRBMO demonstrated strong competitiveness on the CEC2005 benchmark. Among a series of advanced metaheuristic algorithms, MRBMO exhibited significant advantages in terms of convergence speed and solution accuracy. On benchmark functions with 30, 50, and 100 dimensions, the average Friedman values of MRBMO were 1.6029, 1.6601, and 1.8775, respectively, significantly outperforming other algorithms. The overall effectiveness of MRBMO on benchmark functions with 30, 50, and 100 dimensions was 95.65%, which confirmed the effectiveness of MRBMO in handling problems of different dimensions. This paper designed two types of simulation experiments to test the practicability of MRBMO. First, MRBMO was used along with other heuristic algorithms to solve four engineering design optimization problems, aiming to verify the applicability of MRBMO in engineering design optimization. Then, to overcome the shortcomings of metaheuristic algorithms in antenna S-parameter optimization problems—such as time-consuming verification processes, cumbersome operations, and complex modes—this paper adopted a test suite specifically designed for antenna S-parameter optimization, with the goal of efficiently validating the effectiveness of metaheuristic algorithms in this domain. The results demonstrated that MRBMO had significant advantages in both engineering design optimization and antenna S-parameter optimization.

3 September 2025 in PLOS ONE
Verified
Journal Article
Open access
PDF file
YGYanzhao Gu
Yanzhao Gu
WJWei Junhao
Wei Junhao
ZLZikun Li
Zikun Li
BLBaili Lu
Baili Lu
SPShirou Pan
Shirou Pan
NCNgai Cheong
Ngai Cheong
This paper analyzes the shortcomings of the traditional Whale Optimization Algorithm (WOA), mainly including the tendency to fall into local optima, slow convergence speed, and insufficient global search ability for high-dimensional and complex optimization problems. An improved Whale Optimization Algorithm (GWOA) is proposed to overcome these issues. By integrating several improvement strategies, such as adaptive parameter adjustment, enhanced prey encircling, and sine-cosine search strategies, GWOA significantly enhances global search ability and convergence efficiency. However, GWOA increases computational complexity, which may lead to longer computation times when handling large-scale problems. It may also fall into local optima in high-dimensional cases. Several experiments were conducted to verify the effectiveness of GWOA. First, 23 classic benchmark functions were tested, covering unimodal, multimodal, and compositional optimization problems. GWOA was compared with other basic metaheuristic algorithms, excellent WOA variants, and the latest algorithms. Then, a comparative scalability experiment is performed on GWOA. The experimental results showed that GWOA achieved better convergence speed and solution accuracy than other algorithms in most test functions, especially in multimodal and compositional optimization problems, with an Overall Efficiency (OE) value of 74.46%. In engineering optimization problems, such as pressure vessel design and spring design, GWOA effectively reduced costs and met constraints, demonstrating stronger stability and optimization ability. In conclusion, GWOA significantly improves the global search ability, convergence speed, and solution stability through multi-strategy integration. It shows great potential in solving complex optimization problems and provides an efficient tool for engineering optimization applications.

18 June 2025 in PLOS ONE
Verified
Journal Article
Open access
PDF file
WLWanyan Li
Wanyan Li
BLBaili Lu
Baili Lu
SPShirou Pan
Shirou Pan
SBShuaifei Bai
Shuaifei Bai
JHJiayu He
Jiayu He
DXDanning Xu
Danning Xu
The protective effects of the polysaccharide of Atractylodes macrocephala Koidz (PAMK) against lipopolysaccharide (LPS)-induced intestinal injury in goslings was determined using 16S rRNA analysis of cecal contents and serum metabolomics analysis. PAMK was administered to goslings following LPS-induced intestinal injury, and its effects were assessed. PAMK significantly reduced the serum levels of inflammatory factors including interleukin (IL)-6 and C-reactive protein and decreased the expression of pro-inflammatory cytokines including interleukin(IL)-1 β , IL-6, and toll-like receptor 2 in jejunal tissues. Moreover, PAMK significantly upregulated the relative mRNA expression levels of the tight junction proteins Zonula occludens-1, Occludin, Claudin, and Mucin-2, enhancing the integrity of the intestinal barrier and alleviating LPS-induced intestinal injury. 16S rRNA sequencing revealed that PAMK could alleviate LPS-induced disruption of the intestinal microbiota structure and improve microbial diversity. Metabolomics analysis revealed that PAMK could influence key metabolic pathways, including the mTOR, PI3K-Akt, and FoxO signaling pathways, and regulate metabolites such as L-aspartic acid and S-adenosylmethionine. Integrated analysis indicated that PAMK could promote the enrichment of beneficial bacteria (e.g., Allobaculum and Peptococcus) while alleviating LPS-induced microbial dysbiosis by modulating the correlation between key metabolites and specific microbial populations. Overall, PAMK could alleviate LPS induced intestinal injury by enhancing intestinal barrier function, optimizing gut microbiota composition, and regulating metabolic signaling pathways. Our findings provide a novel strategy for maintaining the intestinal health of poultry and preventing intestinal diseases.