Optimal design strategy for gear pairs integrating macro and micro geometry optimization

Abstract

An optimal design methodology for involute gear pairs integrating macro- and micro geometry optimization is proposed. A two-stage multi-objective optimization using NSGA-III minimizes gear mesh power loss and peak-to-peak static transmission error under strength and geometric constraints. Macro geometry optimization identifies diverse candidate designs, and micro geometry optimization refines tooth profile modifications to further enhance performance. A random forest regression model analyzes the optimized results and extracts key macro geometry features associated with superior outcomes. Based on these insights, a macro geometry design strategy is formulated and validated through re-optimization. The proposed methodology suggests a systematic approach to establishing generalized gear design guidelines, it should be noted that the findings are derived under specific design conditions and may require further verification for broader applications.