共轭斜量法在发动机故障诊断中的应用

Application of Conjugate Gradient Method to Engine Fault Diagnosis

提出了基于主因子模型的发动机故障诊断共轭斜量算法,该算法可以有效地解决发动机故障诊断算法中由于故障方程存在多重共线性所引起的误诊与多解问题。某型发动机的故障诊断结果表明,共轭斜量法能够有效地反映出实际的故障部件,相比于最小二乘法,其诊断精度有大幅度提高,具有一定的工程应用价值。

The Conjugate gradient Algorithm on the primary factor model is presented. This meth- od is able to solve the problems such as mistaken and multiple solutions in the engine fault diag-nosis algorithm based on fault equation, which result from the multi-collinearity of the fault equa-tion. Simulation conclusion of a certain engine fault diagnosis shows that this method can find out the actual fault component effectively. Compare to least square solution, its diagnosis effect is improved greatly. The method is of some engineering value.