摘要
目前的故障诊断系统中普遍存在缺乏通用性、对故障先验知识与系统动态数据利用不均衡、稳定性和可塑性难以两全等缺陷,本文针对这些问题提出了一种故障模型,运用符号学习与神经网络相结合构成的增量式混合型学习算法,成功地对故障示例集进行处理,在此基础上给出了一种通用故障诊断系统设计方法。实验表明,利用该方法实现的系统在进行故障诊断时可以取得较好的效果。
In this paper, we put forward a designing method for Fault Diagnosis Systems(FDS) by proposing a new fault model and using the incremental hybrid learning algorithm which tightly combines symbolic learning and neural networks. It's capable of overcoming several shortcomings in existing diagnosis systems, such as lack of universiality, unbalance in the use of prior fault knowledge and the use of dynamic data, and a dilemma between stability and plasticity. The experiment system implemented by this method shows a good diagnostic ability.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2000年第4期424-428,共5页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金
江苏省自然科学基金
