基于数据驱动的轴流压气机特性图规律及预测

为探索轴流压气机特性线的内在规律,结合数据分析方法和热力学原理,对五十余幅轴流压气机特性数据进行了研究。通过将多种热力学参数引入特性图,并提出了压气机特性描述参数,从各转速峰值效率点连线、近喘振区和近堵塞区3个方面分析了压气机特性中物理量间的规律,发现了压气机特性参数与设计指标间的关联,并依据前述规律发展了基于设计指标和性能参数的压气机特性预测算法。研究发现:在工作转速范围内,多个热力学参数之间存在明确的线性相关性,各轴流压气机的特性参数与设计点总压比有显著关联,而与设计点效率关联不明显。提出的预测算法可以在没有叶片几何信息的条件下,仅依靠设计指标和性能参数,较好地预测出总压比小于8的轴流压气机性能。 

Laws and prediction of axial compressor performance map based on data-driven

To explore the inherent laws of axial compressors’ performance maps, more than 50 axial compressors’ characteristic data were studied based on data analysis and thermodynamic principle. By introducing a variety of thermodynamic parameters into the performance curves and novel performance parameters, the research of laws between physics quantities of compressors was performed from three aspects which were peak efficiency lines, near-surge zones and near-stall zones. The relationship between compressor characteristic parameters and design indicators was found and the prediction algorithm was developed to predict the performance of compressors based on design indicators and performance parameters. It showed that there was a linear relationship between thermodynamics parameters within working speed range. The performance parameters of compressors were obviously correlated with design total pressure ratio and little correlated with design efficiency. The prediction algorithm can predict the performance map of compressors (total pressure ratio less than 8) well just relying on the design indicators and performance parameters without blades geometry information.