A memory-saving metaheuristic algorithm for onboard optimization: Solitary Inchworm Foraging Optimizer

Abstract

This paper proposes a memory-saving metaheuristic algorithm, the Solitary Inchworm Foraging Optimizer, designed for onboard optimization problems under memory resource limitations. The proposed algorithm employs a unique single-agent search mechanism that mathematically models the behaviors of an inchworm. Parallel communication strategies are developed to enable information exchange among parallel agents, enhancing solution quality while preserving computational efficiency. As a result, Solitary Inchworm Foraging Optimizer is not only effective for global optimization but also efficient enough for onboard optimization. Theoretical analyses provide computational complexity evaluations and a proof of global convergence. Comparative numerical experiments on three well-known benchmark test suites demonstrate the significant superiority of the proposed algorithm over eight state-of-the-art metaheuristic algorithms. Additionally, hardware-in-the-loop simulations of two onboard application case studies are carried out. The simulation results further validate the efficiency of the proposed algorithm, consuming less computation time while achieving better solution quality compared with five baseline algorithms.