采用BP神经网络的叶片电解加工精度预测

工件成型精度的预测是实际电解加工的重要研究课题,快速、准确地选取加工参数并预测出工件的形状精度可以减少试验次数,缩短试制周期,降低生产成本。本文以某型发动机叶片为研究对象,对影响电解加工精度的主要加工参数进行了分析,结合工艺试验的数据建立了BP网络模型,并采用该模型进行了不同加工参数组合下叶片型面的预测。结果表明,该模型的预测精度比较高,具有一定的工程实用性。

Accuracy Prediction of Blade Electrochemical Machining Based on BP Neural Network

Predicting the accuracy of a machined profile is a great concern for electrochemical machining(ECM).Quick and accurate selection of machining parameters and prediction of the accuracy of the machined profile may reduce times of experiment,shorten the cycle of trial machining and lower down production costs.The paper studied the blades of a model of aero-engine,analyzed the main machining parameters that affect the ECM accuracy and established its back-propagation(BP) neural network model with experimental data taken into account.The model was used to predict the machining profile of the blade in various machining parameter combinations.Its prediction proves to be highly accurate.