针对 BP网络训练过程中收敛速度慢和易陷于局部极小的缺点 ,对常规 BP算法进行改进 ,并将改进后的算法应用于凝汽器故障诊断。结果表明 ,与传统 BP算法相比 ,改进后的算法收敛速度明显加快 ,并有效地克服了 BP网络易陷于局部极小的缺点 Aiming at the shortcomings of slow convergence and easy to be trapped in local minima during BP network training, the conventional BP algorithm is improved and the improved algorithm is applied to fault diagnosis of condenser. The results show that compared with the traditional BP algorithm, the convergence speed of the improved algorithm is obviously accelerated and the BP network is easily overcome the shortcoming of local minimum