An Entropy-Balanced Orthogonal Learning Bamboo Forest Growth Optimization Algorithm with Quasi-Affine Transformation...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Heuristic Computing

Background:

• The bamboo forest growth optimization (BFGO) algorithm demonstrates effectiveness but lacks strong exploitation capabilities.
• Existing optimization algorithms often face challenges in balancing exploration and exploitation for complex problems.

Purpose of the Study:

• To introduce the orthogonal learning quasi-affine transformation evolutionary bamboo forest growth optimization (OQBFGO) algorithm.
• To enhance the searchability and global development capabilities of the BFGO algorithm.
• To evaluate the performance of OQBFGO against other established algorithms and its application to the capacitated vehicle routing problem.

Main Methods:

• The OQBFGO algorithm integrates quasi-affine transformation evolution to broaden particle distribution and improve searchability via entropy increase.
• An orthogonal learning strategy is employed to refine particle aggregation from chaotic states, facilitating global development through entropy reduction.
• Performance evaluation involved testing OQBFGO, BFGO, QBFGO, OBFGO, and three other mature algorithms on the CEC2017 benchmark function.

Main Results:

• The OQBFGO algorithm demonstrated superior performance compared to BFGO, QBFGO, OBFGO, and three other benchmark algorithms on the CEC2017 test suite.
• Application of OQBFGO to the capacitated vehicle routing problem yielded better results than competing algorithms.
• The proposed enhancements significantly improved particle searchability and global development capabilities.

Conclusions:

• The OQBFGO algorithm offers a significant advancement in optimization techniques, particularly for problems requiring robust exploration and exploitation.
• The integration of quasi-affine transformation and orthogonal learning provides a synergistic approach to enhance heuristic search.
• OQBFGO shows strong potential for solving complex real-world optimization challenges, including the capacitated vehicle routing problem.