摘要
基于BP神经网络原理,利用轴承空载和负载运行下的各参数建立轴承剩余寿命的预测模型。在MATLAB中对轴承数据样本进行学习与训练,获得较精确的BP神经网络结构的权值和阈值,根据BP神经网络算法编写m函数文件,在MATLAB中生成COM组件。用Visual Basic 6.0编写的系统软件界面,在界面中调用COM组件中的DLL文件。解决算法优化和模块间调用等问题,成功开发出轴承寿命预测系统,用于机电产品的质量控制管理。
Based on BP neural network theory, this paper establishes a bearing residual life prediction model by using parameters of bearing under unload and load operation. Beating data samples are learned and trained in MATLAB to obtain more accurate weights and threshold of BP neural network and to obtain m function file through BP neural network algorithm. A COM component is generated in MATLAB. DLL flies in the COM component are called in the software interface made by Visual Basic 6.0. Solving the problem of algorithm optimization and module calls, this paper successfully develops a bearing life prediction software for the management of quality control of mechanical and electrical products.
出处
《现代机械》
2014年第3期7-10,共4页
Modern Machinery
作者简介
宁武龙（1988-），西南交通大学硕士研究生,研究方向为机电液一体化。
