Red-Tailed Boa Optimization Algorithm

Most existing metaheuristic optimization algorithms focus on updating the positions of "candidate solution points," searching for optimal solutions in the solution space through mechanisms such as random perturbations, differential transformations, or velocity evolution. These methods commonly suffer from unstable search directions, premature convergence, and the curse of dimensionality in high-dimensional, strongly non-convex, or highly coupled problems. This paper proposes a new optimization theoretical framework—the Progressive Feasible-Space Collapse (PFSC) paradigm—and constructs a specific algorithm example under this paradigm: the Red-tailed Boa Optimization Algorithm RTBO-Ω. This algorithm no longer views the optimization process as "approaching the optimal point in the point space," but rather models it as "the dynamic compression and reconstruction of the feasible space geometry." RTBO-Ω uses the feasible space measure as the core optimization object, gradually compressing ineffective search degrees of freedom through volume-conserving anisotropic winding dynamics, and introducing a spatial re-embedding mechanism when topological degradation occurs in the feasible space, thereby achieving stable convergence and structural escape from local optima. This paper systematically presents the mathematical modeling, dynamical system description, operator definition, and theoretical property analysis of RTBO-Ω, demonstrating that this method differs from existing swarm intelligence and evolutionary algorithms at the theoretical level, and can serve as a representative of a new type of spatial structure optimization method.