摘要
针对传统部件特性修正方法未考虑发动机多状态导致修正精度不高的问题,提出了1种基于粒子群优化和滑动最小二乘法的某型发动机部件特性修正方法。该方法利用粒子群优化算法分别求得在发动机不同状态下的修正系数,并以这些系数为基础,采用滑动最小二乘方法拟合修正系数曲面,从系数曲面上获取原有部件特性图上各点对应的修正系数,从而得到修正特性。试验结果表明:该方法克服了传统方法的不足,提高了特性修正精度,为开展单机监控和视情维修提供准确的部件数据基础。
Aiming at the problem that the conventional correction methods doesn't consider the multi-condition of engine, the correction of component characteristic based on particle swarm and Moving Least Squares (MLS) was presented. The particle swarm method was used to solve the correction coefficients under different conditions. The MLS method was used to fit the curve consisted of correction coefficients, and got the correction coefficients from the orginal characteristic. The experiment results show the proposed method outperforms the conventional methods and improves the accuracy of component characteristic correction, which provide the accurate data for single aircraft monitoring and maintenance.
出处
《航空发动机》
2015年第1期94-98,共5页
Aeroengine
关键词
部件特性修正
粒子群算法
滑动最小二乘法
航空发动机
component characteristic correction
particle swarm algorithms
Moving Least Square method
aeroengine