高压压气机性能老化预测和影响分析

摘要

Taking a high pressure compressor for example, efficiency and mass flow loss are determined as the main factors of performance deterioration. Variation of efficiency and mass flow loss with blade roughness is gained; their corresponding relation is established and validated by neural network. Turning two middle variables into single variable is presented by using principle ingredient analysis; the functional relation of flight circular number and blade roughness is established by introducing synthetic index of performance dete- rioration. Performance deterioration is predicted by tri-index smooth method of time series with good result. Computational model of high-pressure compressor performance deterioration is introduced into stable en- gine computational model to modify component characteristic, from which effect on high pressure turbine per- formance is gained. The results are referential to prediction and analysis of engine performance deterioration.%以高压压气机为例，确定效率和流量损失作为性能衰退分析的主要因素。得到叶片粗糙度引起的效率和流量损失变化规律，利用神经网络建立并验证其对应关系。提出运用主成分分析将两个中间变量转化成单一变量，通过引入性能衰退综合指数，建立飞行循环数与叶片粗糙度的函数关系。利用时间序列三次指数平滑方法对性能衰退进行预测，效果较好。将高压压气机性能衰退计算模型引入到发动机稳态计算模型中修正部件特性，得到对高压涡轮性能的影响。所得结论对发动机性能衰退预测和研究具有一定的参考价值。