Optimizing Interplanetary Trajectories using Hybrid Meta-heuristic

TL;DR

This paper introduces GMPA, a hybrid metaheuristic combining GWO and MPA features, to improve interplanetary trajectory optimization by enhancing exploration, exploitation, and solution diversity, validated on ESA's GTOPX benchmarks.

Contribution

The paper presents GMPA, a novel hybrid algorithm integrating MPA features into GWO, specifically designed for complex interplanetary trajectory optimization problems.

Findings

GMPA outperforms traditional GWO and other metaheuristics in convergence speed.

GMPA achieves higher solution quality on GTOPX benchmarks.

Enhanced exploration and exploitation balance reduces premature convergence.

Abstract

This paper proposes an advanced hybrid optimization (GMPA) algorithm to effectively address the inherent limitations of the Grey Wolf Optimizer (GWO) when applied to complex optimization scenarios. Specifically, GMPA integrates essential features from the Marine Predators Algorithm (MPA) into the GWO framework, enabling superior performance through enhanced exploration and exploitation balance. The evaluation utilizes the GTOPX benchmark dataset from the European Space Agency (ESA), encompassing highly complex interplanetary trajectory optimization problems characterized by pronounced nonlinearity and multiple conflicting objectives reflective of real-world aerospace scenarios. Central to GMPA's methodology is an elite matrix, borrowed from MPA, designed to preserve and refine high-quality solutions iteratively, thereby promoting solution diversity and minimizing premature convergence. Furthermore, GMPA incorporates a three-phase position updating mechanism combined with L\'evy flights and Brownian motion to significantly bolster exploration capabilities, effectively mitigating the risk of stagnation in local optima. GMPA dynamically retains historical information on promising search areas, leveraging the memory storage features intrinsic to MPA, facilitating targeted exploitation and refinement. Empirical evaluations demonstrate GMPA's superior effectiveness compared to traditional GWO and other advanced metaheuristic algorithms, achieving markedly improved convergence rates and solution quality across GTOPX benchmarks. Consequently, GMPA emerges as a robust, efficient, and adaptive optimization approach particularly suitable for high-dimensional and complex aerospace trajectory optimization, offering significant insights and practical advancements in hybrid metaheuristic optimization techniques.