传感器测量偏差下的航空发动机智能性能诊断

Intelligent performance diagnosis of aeroengine with sensor measurement bias

以某型涡扇发动机为研究对象,构建了基于神经网络的航空发动机智能性能诊断方法,讨论了测量噪声及测量偏差对诊断结果的影响及其处理方法.建立一簇并行的神经网络组和发动机模型,通过比较各模型输出与发动机测量参数之间的误差,判断传感器是否存在测量偏差.仿真结果表明,该方法能有效消除测量噪声,准确判断并隔离有测量偏差的传感器,得出正确的发动机性能诊断结果.

An Intelligent aeroengine performance diagnostic method, based on neural networks, was investigated. Sensor measurement deviations from the nominal condition are the only information for the estimation of engine health parameters. They are often distorted by noise and bias, thereby mask the true engine condition and lead to incorrect diagnostic resuits. A bank of neural networks and engine models was developed. The errors between the model outputs and sensor measurements can be used to detect and isolate the sensor, which has measurement bias. Then the engine performance can be estimated by the set of measurements without biases. The simulation results show that this approach is promising for reliable aero-engine diagnosis.