基于GA-GRNN-GA的飞机发动机风扇叶片清洗参数优化

OPTIMIZATION OF CLEANING PARAMETERS FOR AIRCRAFT ENGINE BLADES BASED ON GA-GRNN-GA

以超声波清洗飞机发动机风扇叶片的清洁度为目标,在分析超声波清洁度多种影响因素基础上,建立超声波清洗工艺参数即超声清洗功率、温度、时间和清洗距离的非线性预测模型。采用基于遗传算法的广义回归神经网络(GRNN)进行清洁度预测,并以预测的清洁度为适应度函数,使用遗传算法对清洗参数进行优化。对GA-GRNN预测模型进行仿真验证,表明该模型预测清洁度较好。最后进行飞机发动机风扇叶片超声波清洗实验,结果表明,优化后的清洁度与GA-GRNN-GA模型预测优化的清洁度误差小于4.3%,为飞机发动机风扇片叶的自动化清洗提供了理论依据。

We take the cleaning degree of aero-engine fan blades as objective,and establish the nonlinear prediction model of ultrasonic cleaning process parameters such as ultrasonic cleaning power,temperature,time and cleaning distance based on the analysis of various factors affecting ultrasonic cleanliness.The generalized regression neural network(GRNN)based on genetic algorithm was adopted to predict the cleanliness,and the predicted cleanliness was used as the fitness function.We applied the genetic algorithm to optimize the cleaning parameters.The GA-GRNN prediction model was simulated and validated.The simulation results show that the predicted cleanliness of the proposed model is better.We carried out the experiment of ultrasonic cleaning for fan blades of aircraft engine.The experimental results show that the error between the optimized cleaning parameters and GA-GRNN-GA model is less than 4.3%,which provides a theoretical basis for the automatic cleaning of the fan blades of aircraft engines.