PSO-Guided Recognition of Elementary Geometries for Zero-Defect Manufacturing

Automatic and robust recognition of elementary functional geometries (planes, spheres, cylinders, cones) is pivotal for Zero-Defect Manufacturing (ZDM). Curvature-based metrology provides analytic signatures for these shapes, yet practical deployments require noise-robust, minimally supervised threshold selection. We cast geometry recognition as a constrained optimization over curvature decision rules and employ Particle Swarm Optimization (PSO) as a classifier optimizer. The formulation is fully simulation-ready and theory-driven: it integrates differential-geometry estimators (Gaussian/mean curvature), ISO-consistent constraints, and discrete PSO updates with feasibility-preserving repair. We provide a rigorous problem statement, feasibility and convergence insights, and pseudocode for a discrete PSO tailored to curvature-rule optimization.