Data-Driven Co-Optimization of Blade Selection and Sequencing for Aeroengine Rotors under Industrial Assembly Constraints

Abstract

During the rotor assembly of aeroengines, the combined effect of blade mass moment variations and fixed tenon slot constraints makes single-phase rotor unbalance optimization strategies insufficient for real-world industrial assembly scenarios. This often leads to excessive residual unbalance after assembly, resulting in engine vibrations and compromised operational stability. To address the lack of blade selection strategies and low qualification rates due to tenon slot constraints in industrial settings, this paper proposes a co-optimization method for blade selection and sequencing under industrial assembly constraints. A two-stage data-driven optimization framework is developed. In the first stage, a Dynamic Replacement Roulette Selection (DRWS) algorithm is introduced for global multi-set blade selection, improving blade utilization and avoiding selection failure caused by excessive moment dispersion. In the second stage, under fixed tenon slot constraints, blade sequencing is optimized using a Constrained Adaptive Genetic Algorithm (CAGA), effectively suppressing residual unbalance. Experimental results demonstrate that the proposed method achieves a blade utilization rate of 92.4% on 145 samples, with well-balanced group sets. Under tenon slot constraints, the residual unbalance is reduced from 58 g·mm and 94 g·mm (random assembly) to 7 g·mm and 10 g·mm, respectively. This study offers a novel solution and technical support for improving assembly precision and enabling intelligent decision-making in aeroengine rotor assembly lines.