基于人工神经网络的BTA钻削表面粗糙度预测研究-机械制造及自动化专业论文.docxVIP

基于人工神经网络的 BTA 钻削表面粗糙度预测研究 摘要 随着时代的发展，深孔加工越来越受到广泛的关注，在机加工领域占有举足轻重的 地位。它的加工难度比较大，而且每次加工的工作量大，加工时间往往都比较长，是制 造加工中的一个重要加工过程。然而由于人们对产品的要求越来越高，如何获得高的表 面质量已成为当前关注的热点和关键性问题。本文对现阶段深孔加工中如何控制表面粗 糙度进行了研究，利用神经网络建立模型预测表面粗糙度，为深孔加工领域提高表面质 量和优化切削参数提供了新的思路，具有重要的理论意义和实际应用价值。 本文以 BTA 深孔钻削表面粗糙度的预测为主要研究内容，对影响表面粗糙度的因 素进行正交试验，分析 BTA 钻削的加工机理和表面粗糙度在各个影响因素作用下的结 果，得出它们对表面粗糙度的影响规律，为更好的预测表面粗糙度提供理论依据。 本文通过引入神经网络理论到深孔加工领域，利用切削速度、进给量和轴向力对表 面粗糙度的影响规律建立预测 BTA 钻削表面粗糙度的网络模型。动态调整隐含层的节 点数来确定网络结构。通过训练与验证网络模型，比较模型的预测精度和收敛能力能否 达到要求。 课题研究表明，本文所选取的 3-16-1 三层 BP 网络模型的预测误差都在 3%以内， 能够比较准确的预测表面粗糙度。这样就能够选择最优的切削参数，量化预测表面粗糙 度，解决 BTA 钻削无法在线控制表面质量的问题，具有重要的理论意义和实际应用价 值。 关键词：BTA 钻削，人工神经网络，表面粗糙度，正交试验 Surface Roughness Prediction of BTA Drilling Based on Artificial Neural Network Abstract With the development of the times, people are more and more attention on the deep processing, it occupies a pivotal position in the field of machining. It is much difficult machined and each processing has heavy workload, processing time is often long, the manufacturing process is an important process. However, people have increasingly high demand for the product, how to obtain a high surface quality has become a focus of attention and key issues. In this paper, how to control the surface roughness were studied by using neural network model, it provides a new way of thinking to predict the surface roughness and improve the surface quality and optimized cutting parameters for the deep processing field, it has important theoretical significance and practical value. In this paper, the factors of affecting the surface roughness make the orthogonal test and analyze their significant influence, which draws on the influence of surface roughness on BTA deep hole drilling for the study, it provides a better prediction of surface roughness theory. In this paper, it builds predictive network models of surface roughness on BTA drilling by introducing the neural network theory into the deep processing field. It determines the network structure by dyn