论文部分内容阅读
论述一种基于人工神经网络（ＡＮＮ）的知识学习方法，采用改进的ＢＰ算法训练ＡＮＮ，用于船用柴油机故障诊断，以解决一般专家系统在知识获取过程中的“瓶颈”问题。还讨论了ＢＰ算法的改进与参数优化，并给出了在故障诊断中的应用实例。
A knowledge learning method based on artificial neural network (ANN) is discussed. An improved BP algorithm is used to train ANN to diagnose marine diesel engine fault, so as to solve the bottleneck problem of general expert system in knowledge acquisition. The improvement of BP algorithm and parameter optimization are also discussed. The application examples in fault diagnosis are also given.