Augmented Gold Rush Optimizer Is Used for Engineering Optimization Design Problems and UAV Path Planning

Abstract

The Gold Rush Optimizer (GRO) is a recently proposed popular metaheuristic algorithm, but it has sluggish convergence speed, low accuracy, and a tendency to become stuck in local optima when dealing with actual situations. To solve these deficiencies, we present an enhanced version of the Gold Rush Optimizer known as the Augmented Gold Rush Optimizer (AGRO). In AGRO, we first introduce a good point-set population initialization to improve GRO’s global exploration capabilities, consequently increasing the algorithm’s convergence speed. Second, we implement a dynamic Lévy flight search approach to improve the algorithm’s exploratory performance and population diversity. Then we introduce a dynamic centroid reverse learning technique to update population individuals, increase population quality, and speed up the algorithm’s convergence and accuracy. Finally, we use dynamic tangent flight to increase population variety while effectively preventing the algorithm from reaching local optima. To validate AGRO’s performance, we employed the CEC2017 and CEC2022 test suites to verify its ability to solve unconstrained problems. We further validate AGRO’s ability to tackle limited optimization problems by using six engineering optimization design problems and a three-dimensional unmanned aerial vehicle path planning problem. The experimental results show that AGRO has high scalability and practicality.

Keywords

• Gold rush optimizer
• dynamic centroid reverse learning
• dynamic tangent flight
• engineering optimization design problems
• unmanned aerial vehicle path planning