Golden eagle optimizer: A nature-inspired metaheuristic algorithm

This paper proposes a nature-inspired swarm-based metaheuristic for solving global optimization problems called Golden Eagle Optimizer (GEO). The core inspiration of GEO is the intelligence of golden eagles in tuning speed at different stages of their spiral trajectory for hunting. They show more propensity to cruise around and search for prey in the initial stages of hunting and more propensity to attack in the final stages. A golden eagle adjusts these two components to catch the best possible prey in feasible region the shortest possible time. This behavior is mathematically modeled to highlight exploration and exploitation for a global optimization method. The performance of the proposed algorithm is tested and confirmed using 33 benchmark test functions and a scalability test. Results were compared to that of six other well-known algorithms, which revealed GEO's superiority, which indicates that it can find the global optimum and avoid local optima effectively. The Multi-Objective Golden Eagle Optimizer (MOGEO) is also proposed to solve multi-objective problems. The performance of MOGEO is also tested and verified on ten multi-objective benchmark functions. Results were compared to that of two other multi-objective algorithms, which showed that it can approximate true Pareto optimal solutions better than the other two algorithms. The software (toolbox) and source code for GEO and MOGEO are also provided, which are publicly available.

Automated summary

This paper introduces a nature-inspired global optimization algorithm, GEO, based on the hunting behavior of golden eagles, and a multi-objective algorithm, MOGEO. Testing on benchmark functions and multi-objective benchmark functions show that GEO and MOGEO outperform other algorithms in optimization performance.