Iterative refinement algorithm for signal parameter estimation in optical interference systems based on differential evolution

This paper presents an iterative refinement framework for signal parameter estimation in optical interference systems through differential evolution optimization and adaptive model selection. The proposed methodology employs interference pattern analysis capturing spectral signatures across wavelength ranges, achieving coefficient of determination R-squared of 0.238 through nonlinear regression with statistical significance below 0.001. The framework integrates three computational stages comprising spectral preprocessing via Savitzky-Golay filtering, differential evolution optimization with population size 50 across 200 iterations through iterative population-based refinement converging to optimal parameter configurations, and adaptive correction mechanisms. addressing systematic errors through Fabry-Perot interference models. The iterative refinement process progressively improves parameter estimates through successive generations, maintaining solution diversity while converging toward global optima. Experimental validation demonstrates parameter estimation accuracy within two percent tolerance under measurement perturbations, with computational efficiency achieving millisecond-level processing suitable for real-time signal analysis. Key contributions include evolutionary optimization strategies navigating complex parameter spaces, adaptive model switching between single-beam and multiple-beam interference regimes, and robust error correction protocols maintaining estimation stability across heterogeneous operating conditions.