摘要
根据车削过程中工件直径误差的特点,提出了用自适应模糊系统预测由弹性变形等因素引起的工件直径误差的思路,通过梯度下降算法训练M amdan i型模糊系统,以确定合理的系统参数。根据工件直径误差与切削深度、进给量等的关系,进行车削实验,得到训练数据和测试数据,用训练数据训练模糊系统,进而用测试数据进行测试,得到结果合理,从而验证了利用自适应模糊系统进行工件直径误差预测的可行性。和回归分析的预测值进行比较,比较结果显示了自适应模糊系统在车削工件直径误差预测方面的应用具有优势。
According to the characteristics of diameter error of work-piece in the turning process, a train of thought of using the serf-adaptive fuzzy system to predict the work-piece diameter error caused by factors of elastic deformation etc. has been put forward. By means of gradient descent algorithm to train the Mamdani typed fuzzy system so as to ascertain the rational systematic parameters. On the basis of the relationship between the work-piece diameter error and the cutting depth, feed-rate etc. to carrying out turning test so as to acquire the training data and the testing data. Use the training data to train the fuzzy system, and then use the testing data to carrying out the test and obtained reasonable result, thus verified the feasibility for carrying out error prediction of work-piece diameter by the use of serf-adaptive fuzzy system. Carrying out comparison with the predicted value of regression analysis, and the result of comparison showed that the self-adaptive fuzzy system possesses superiority on the application on error prediction aspect for turning work-piece diameter.
出处
《机械设计》
CSCD
北大核心
2010年第1期26-30,共5页
Journal of Machine Design
基金
航空科学基金资助项目(2008ZE53038)
国家科技支撑计划资助项目(2008BAF32B04)
西北工业大学基础研究基金资助项目(2008JC10)
关键词
工件直径误差
自适应模糊系统
弹性变形
梯度下降算法
work-piece diameter error
self-adaptive fuzzy system
elastic deformation
gradient descent algorithm