| 基于BP神经网络的磨损微粒智能识别 |
| |
| 引用本文： | 李忠,曾昭翔,陈大融.基于BP神经网络的磨损微粒智能识别[J].北京交通大学学报(自然科学版),1998,22(1). |
| |
| 作者姓名： | 李忠 曾昭翔 陈大融 |
| |
| 作者单位： | 北方交通大学机械与电气工程学院，清华大学精密仪器与机械学系 |
| |
| 摘 要： | 磨粒图象自动分析与识别问题不但是铁谱技术发展需要解决的关键问题，同时也是针对机械设备磨损工况的状态监测与故障诊断后续信息流程实现自动化、智能化和系统化时遇到的关键技术难点和重点．本文在分析综述近年有关研究方法和结果的基础上，提出解决这一问题的新的研究思路，及相应的处理流程、模型和软件实现，并给出了具体的实验结果．本文提出的流程可用于其它微粒图象（如血液细胞分析）自动识别领域．
|
| 关 键 词： | 磨损 故障诊断 铁谱 神经网络 磨粒识别 |
Intelligent Recognition of Wear Particles Based on BP Neural Network |
| |
| Abstract: | Automatic recognition and analysis of wear particles is one of the most important problems to be resolved in ferrography technique but it is also the difficult technical key to realize the machine wear monitoring and fault diagnosis analysis process in automatization, intelleetualization and systematization. This paper will introduce our ANN based approach to this problem, a completely new method different from ordinary methods in the past several years, and give in detail its procedure, model, software realization, expeimental results and application in practical industrial site. Our idea and models here can also be used in other fields of micro particle image auto analysis. |
| |
| Keywords: | wear\ fault diagnosis\ ferrography\ ANN\ pattern recognition |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
