A Bio-Inspired Comprehensive Learning Strategy-Enhanced Parrot Optimizer: Performance Evaluation and Application to...

Area of Science:

• Computational intelligence
• Swarm intelligence
• Bio-inspired algorithms

Background:

• Swarm intelligence algorithms require a balance between exploration and exploitation for high-dimensional problems.
• The original Parrot Optimizer (PO) faces limitations like population homogenization and local optima entrapment due to single-source social learning.

Purpose of the Study:

• To introduce the Comprehensive Learning Parrot Optimizer (CL-PO), a novel bio-inspired metaheuristic.
• To enhance the Parrot Optimizer's ability to maintain population diversity and adaptive exploration.

Main Methods:

• CL-PO incorporates a dimension-wise multi-exemplar social learning mechanism inspired by avian social dynamics.
• Individuals learn from multiple superior peers to reconstruct search trajectories, sustaining diversity.
• The algorithm was benchmarked on 29 CEC 2017 test functions and applied to reservoir production optimization.

Main Results:

• CL-PO demonstrated statistically superior performance against nine state-of-the-art algorithms, achieving an average Friedman rank of 1.28.
• The algorithm consistently maximized Net Present Value (NPV) in a reservoir production optimization task, reaching 9.625×10^8 USD.
• CL-PO proved effective in handling large-scale engineering optimization problems with complex constraints.

Conclusions:

• CL-PO offers a powerful and reliable solution for complex engineering optimization challenges.
• The enhanced social learning mechanism effectively addresses limitations of previous swarm intelligence approaches.
• The algorithm shows significant potential for real-world applications requiring robust optimization.