Integrando conceitos cosmológicos e quânticos em algoritmos de otimização para planejamento florestal

Abstract

This work explores the intersection of optimization and physics concepts to address complex problems, applying these ideas to strategic forest planning. The research focuses on maximizing the Net Present Value (NPV) over a 16-year planning horizon, tackling the challenge of selecting eucalyptus management prescriptions that meet production demands within minimum and maximum limits. Inspired by cosmological phenomena, such as the motion of photons in extreme environments and the gravitational influence of black holes, a hybrid cosmological operator was developed. This operator is designed to adjust neighborhood structures in optimization algorithms like Variable Neighborhood Search (VNS) and Tabu Search, resulting in the hybrid variants named Hybrid Cosmological Variable Neighborhood Search (VNSHC) and Hybrid Cosmological Tabu Search (CHT). The method distinguishes itself by treating promising solutions as regions near black holes, enabling detailed exploration, while distant solutions undergo broader adjustments to search for new areas in the solution space, thereby reducing the likelihood of convergence to local minima. The mathematical model implementation was carried out in MATLAB, with statistical validations conducted in Python using tests such as Friedman and Wilcoxon to ensure result robustness. Furthermore, the hybrid algorithm was compared to classical approaches, including Genetic Algorithm (GA), classical VNS, and Tabu Search, to evaluate its efficiency and performance. The results demonstrate that integrating cosmological concepts with computational heuristics significantly enhances the quality of solutions, achieving an effective balance between exploration and exploitation in the search space.