基于遗传BP神经网络的内圆磨削ZTA陶瓷材料去除率预测  

Prediction of material removal rate for internal grinding ZTA ceramics based on genetic BP neural network

作　　者：赵明利[1] 仝攀攀 聂立新[1] 吕晓峰 ZHAO Mingli;TONG Panpan;NIE Lixin;LV Xiaofeng(School of Mechanical and Power Engineering,Henan Polytechnic University,Jiaozuo 454000,CHN)

机构地区：[1]河南理工大学机械与动力工程学院

出　　处：《制造技术与机床》2019年第9期44-47,52,共5页Manufacturing Technology & Machine Tool

基　　金：国家自然科学基金(E51175153);河南理工大学博士基金(B2012-105)

摘　　要：针对磨削加工中材料去除率(MRR)在线检测困难这一问题,构建材料去除率的预测模型显得尤为重要。考虑到单独运用BP神经网络不仅存在收敛速度较慢,而且容易坠入局部最优解等问题,故建立了遗传算法与BP神经网络相结合的模型来对给定的超声频率、砂轮速度、工件速度、磨削深度等工艺参数对材料去除率(MRR)进行预测。首先运用遗传算法的全局搜寻作用来对BP神经网络的最初权值以及阈值进行优化,而后运用L-M优化算法对网络进行多次训练,利用训练好的BP神经网络模型来对输出进行预测。结果表明:遗传算法与BP神经网络相结合的模型比单独使用BP神经网络模型预测效果要好,能够提高材料去除率的预测精度和收敛速度。Aiming at the difficulty of on-line detection of material removal rate(MRR)during grinding,it is particularly important to construct a prediction model for material removal rate.Considering that the use of BP neural network alone not only has a slow convergence rate,but also easily falls into the local optimal solution and other problem,a model based on genetic algorithm and BP neural network is established to predict the material removal rate(MRR)with certain ultrasonic frequency,grinding wheel speed,workpiece speed and grinding depth.First,the global search function of the genetic algorithm can be used to optimize the initial weights and thresholds of the BP neural network.Then,the network is trained multiple times using the L-M optimization algorithm,and the output is predicted using the trained BP neural network model.The results show that the model combining genetic algorithm with BP neural network is better than using BP neural network model alone.,and can improve the prediction accuracy and convergence speed of the material removal rate.

关 键 词：材料去除率 BP神经网络 遗传算法 L-M优化算法 

分 类 号：TH145.1[一般工业技术—材料科学与工程]