基于MATLAB的切削表面质量预测软件开发

Development of Cutting Surface Quality Prediction Software Based on MATLAB

LMBP神经网络具有性能稳定、收敛快的特点,可应用于切削加工表面质量的预测。在MATLAB中设计用户界面,实现对工艺试验数据的读取和对神经网络输入输出的设置;设计了基于LMBP算法神经网络程序,完成软件的编译和验证。将软件应用在切削表面质量预测,预测结果与试验结果误差较小,具有较高精度;软件操作简单,适合现场工作人员使用。

LMBP neural network has stable performance and rapid convergence,which can be applied to the cutting surface quality prediction. The user interface was designed in MATLAB software,which could be used to read process test data from files and set the input and output of the neural network; the neural network program based on LMBP algorithm was developed,and the full software was compiled and verified. The software was used in the cutting surface quality prediction. The prediction error is smaller with the experimental results,the prediction results have high precision; the software is simple,suitable for the field staffs.