大型远程客机涡扇发动机过热检测方法研究

Large Aircraft Turbofan Engine Overheat Detection Method Research

研究大型远程客机涡扇发动机过热检测优化问题;传统的大型远程客机涡扇发动机过热检测方法不能有效过滤外界干扰因素,飞机发动机过热检测的准确性较差;为实现大型远程客机涡扇发动机过热的准确检测,提出一种基于随机集与结果残差修正的融合先验知识的大型远程客机涡扇发动机过热诊断方法。对测量数据进行处理获取发动机过热的特征值,得到发动机过热证据可信度分配,并将其表示为证据的随机集形式,利用残差修正理论对计算误差进行调整,计算调整后的证据与先验信息之间的倾斜度,并依据决策规则得到发动机过热的诊断决策,提高检测准确率;结果表明,该方法的诊断正确率达到90%以上,具有很强的实际应用价值。

The large aircraft turbofan engine overheat detection optimization problem. The traditional large aircraft turbofan engine overheat detection method cannot effectively filtering interference factors, aircraft engine overheat detection accuracy is poor. To realize the accurate detection of large aircraft turbofan engine overheating, this paper proposes a fusion based on random set and the residual correction prior knowledge of large aircraft turbofan engine overheating diagnosis method. Processes the measurement data to obtain the eigenvalues of the engine overheating, get the engine overheating reliability allocation of evidence, and it is expressed as a random set of evidence form, using the theory of residual error correction to adjust calculation error, calculated the adjusted gradient between evidence and a priori information, and on the basis of the decision--making rules engine overheating diagnosis decision--making, improve the detection accuracy. Results show that the method of the diagnostic accuracy of 90% or more has the very strong practical application value.