摘要
在分析讨论了金刚石工具的摩擦学性能及其磨损机理之后,采用人工神经网络计算机技术编制了金刚石工具摩擦学设计的计算机程序,并且依据这种程序设计制造了一组铣磨轮,进而将其装机作了大理石材的磨削加工现场应用考察.结果表明,通过本程序设计可以使铣磨轮的平均消耗数减少11%,磨削数率提高28%.这是在采用人工神经网络计算机程序进行设计时,将实际工作数据作为神经元程序的学习和检查样本,它充分引入了现场使用状况对铣磨工具的综合影响,从而使设计制造的工具具有较长使用寿命和较高磨削效率良好组合的结果.此外,利用人工神经网络计算机程序进行摩擦学设计,还可以大大减少传统设计方法所必须进行的大量模拟工况试验,也不是主要凭靠设计人员的经验进行,容易实现设计的规范化和标准化.
The computer program of tribological design for diamond tools has been made by means of artificial neural network (ANN) technique after analyzing tribological properties and wear mechanism of the diamond tools. Then a set of diamond drilling wheels has been designed and manufactured based on the output data from this ANN computer program.These diamond drilling wheels have been used for marble grinding in a marble factory, Beijing,China.It is proved that the ANN design program not only decreases substantial number of the simulation tests for industrial practice,but also improves the applied properties of the wheels, for example, the mean consumptive number of the wheels decreases about 11% and the grinding efficiency increases about 28%. The ANN tribological design of the diamond tools is a nonlinear problem in substance.The input data of learning samples and checking samples can be obtained from practice in factory. The synthetic influences of different applied factors in industrial practice for the diamond drilling wheels can be considered carefully.Then the output data of manufacturing parameters for this kind of diamond tools give much better combination of longer service life and higher grinding efficiency. It can be concluded that the tribological design for machinery parts using the ANN computer program can decreases simulation tests in the factory a lot and much easier to realize standardization and uniformity for tribological computer aided design (CAD) of machinery parts, this design doesn,t mainly depend on experiences of designers. It is concluded that this method can be spread into other kinds of tribological design for many machinery parts.
出处
《摩擦学学报》
EI
CAS
CSCD
北大核心
1997年第2期171-177,共7页
Tribology
关键词
人工神经网络
计算机技术
金刚石刀具
artificial neural network computer technigue diamond tool tribological design