论文部分内容阅读
针对航空发动机管理“安全关口前移”的新要求,提出从核心单元体层面进行性能评估。依据发动机工作原理提出显著影响性能的核心单元体及可测参数;以BP神经网络建立核心单元体可测参数与Cycles、DEGT性能参数的非线性网络,验证可测参数的有效性;以TOPSIS信息熵构建核心单元体性能评估模型。以PW4077D发动机1441-1751Cycles信息进行模型验证,并与PCA模型比较,结果显示:两种模型都得出核心单元体性能随着Cycles增加而下降,但TOPSIS信息熵模型能显示HPC和HPT性能下降过程复杂,比PCA模型评估精度高。故为选取重点管理单元体提供依据,实现从整机到单元体的前移,提高预警裕度。
In response to the new requirements of aeroengine management and “safety gate advancement”, a performance evaluation from the core unit level is proposed. Based on the working principle of the engine, the core unit and the measurable parameters that significantly affect the performance are put forward. The nonlinear network of the measurable parameters of the core unit and the Cycles and DEGT performance parameters is established by BP neural network to verify the validity of the measurable parameters. The TOPSIS information Entropy to build core unit performance evaluation model. The PW4077D engine 1441-1751Cycles information is used to verify the model and compared with the PCA model. The results show that the core unit performance decreases with the increase of Cycles in both models, but the TOPSIS entropy model can show the performance degradation of HPC and HPT Complicated, more accurate than PCA models. Therefore, the key management unit for the selection of the basis to provide to achieve from the machine to the unit body forward, improve warning margin.