Evaluation of rotating components by combining individual classification with adaptive updating searching method in aero engine

Abstract

Abstract In order to evaluate the roundness error of rotating components in aero engine with increased speed and more accurate estimation, this paper presents an enhanced method that integrates individual classification with adaptive updating searching (ICAUS) strategy. The individual classification method simulates the renewal regulations of various candidate points as iterations increase, subsequently generating new solutions for the next iteration. Additionally, an adaptive updating search zone method that varies in accordance with the increment of iterations is incorporated. These newly generated solutions are distributed within the updated search zone to ensure that the global optimal solution is not overlooked and the search zone is constrained to prevent excessive divergence, thereby enhancing computational efficiency. To validate the efficacy and accuracy of the proposed method, we utilized datasets and conducted comparative experiments using experimental methods previously studied by predecessors. The experimental results and comparisons demonstrate that the proposed ICAUS method exhibits superior convergence characteristics compared to previous studies. Across five distinct datasets, the average computation time satisfying the terminal conditions ranged from 0.0013s to 0.0018s, the standard deviation of the roundness was less than 6.2051×10–9 mm and the optimal solutions were obtained within 10 to 20 iterations, achieving the solution accuracy of ε = 10–8 mm.