Related Experiment Videos
A Hybrid Multi-Strategy Chinese Pangolin Optimization Algorithm and Its Applications
Chaochuan Jia1,2, Yaqi Yang1, Yujie Cheng1
1School of Electronic Information and Artificial Intelligence, West Anhui University, Lu'an 237012, China.
Biomimetics (Basel, Switzerland)
|July 27, 2026
Summary
This study introduces the Adaptive Cauchy-dynamic CPO (ACDCPO) algorithm to improve population distribution and escape local optima in optimization. ACDCPO demonstrates superior performance in benchmark tests and engineering tasks, including accurate moisture content prediction.
Area of Science:
- Computational Intelligence
- Optimization Algorithms
- Machine Learning
Background:
- The Chinese Pangolin Optimization (CPO) algorithm suffers from uneven population initialization and local optima convergence.
- Existing optimization methods require enhancement in population distribution and exploration-exploitation balance.
Purpose of the Study:
- To propose an enhanced CPO algorithm, termed ACDCPO, addressing its inherent limitations.
- To improve population distribution uniformity, enhance local optima escape capability, and balance exploration-exploitation.
Main Methods:
- Integration of boundary-adaptive contraction initialization.
- Incorporation of Cauchy inverse cumulative distribution mutation.
- Application of dynamic opposition-based learning strategies.
Main Results:
- ACDCPO demonstrated superior convergence precision and stability against nine algorithms on CEC2017 and CEC2022 test functions.
- The algorithm exhibited strong constraint handling and adaptability in four engineering optimization tasks.
- Optimized BP network using ACDCPO achieved 91.211% R² for *Dendrobium huoshanense* moisture content prediction.
Conclusions:
- ACDCPO effectively overcomes CPO's drawbacks, enhancing optimization performance.
- The proposed algorithm shows significant potential for practical engineering and data analysis applications.
- ACDCPO offers a robust and adaptive approach for complex optimization problems.
