涡轴发动机试飞辅助监控模型研究  

Study of Turboshaft Engine Flight Test Auxillary Monitoring Model

作　　者：任智勇[1] 王俊琦[1] 李志鹏[1] Ren Zhiyong;Wang Junqi;Li Zhipeng(Chinese Flight Test Establishment, Xi＇an 710089, China)

机构地区：[1]中国飞行试验研究院,陕西西安710089

出　　处：《航空科学技术》2018年第4期52-56,共5页Aeronautical Science & Technology

摘　　要：为了完成飞行试验中的涡轴发动机重要参数监控,利用飞行试验数据结合神经网络方法建立了某涡轴发动机的重要参数模型,输出参数包括发动机轴功率、耗油率、燃气涡轮后温度。通过对比模型输出和试验状态点,评估了模型的准确度,确定了模型应用的置信区间。将神经网络模型应用于飞行试验中不同高度、不同温度条件下的数据,模型输出结果与试验结果吻合较好,证明了模型的有效性。In order to monitor the important parameters of in-flight turboshaft engine,a turboshaft engine monitoring parameter model was built using flight test datas combined with neural networks.The output parameters included the shaft power,fuel consumption,temperature after gas turbine.By comparing the model outputs and test points,the precision of model was assessed and the confidence interval of model application was confirmed.The neural network model was applied to flight test datas in different altitudes and temperatures.The model outputs coincide with the test datas well,thus the validity is proved.

关 键 词：涡轴发动机 飞行试验 监控模型 神经网络 

分 类 号：V235.12[航空宇航科学与技术—航空宇航推进理论与工程]