基于相似应力特征分布的挖掘机动臂应力预测模型

New Stress Prediction Model of Excavator Boom Based on Distribution of Similar Stress Characteristics

针对现有挖掘机动臂结构的应力神经网络预测模型通用性低的问题,提出一种基于动臂的相似应力特征分类的方法,对分类的样本分别建立神经网络预测模型。以中小型挖掘机双液压缸鹅颈式动臂为例,采用分层抽样的方法得到分析样本,构建基于样本分类的应力特征相似分类模型,实现样本相似应力分布的分类,并对各类分别建立通用性更强的动臂应力预测模型,实现优化约束处理体系中快速准确预测应力,提高动臂结构优化设计效率。

The low universality exists in the neural network prediction model of the excavator boom structure stress. This paper pres- ents a method which is used to classify the similar characteristic stress of the boom, establishes the neural network prediction model, takes small and medium-sized excavator hydraulic cylinder with double gooseneck boom as an example, uses the stratified sampling method to obtain the analysis samples, and establishes the similar stress feature classification model to classify the sample, then es- tablishes the stress prediction models to achieve more rapid and accurate prediction and improve the efficiency of the boom structure optimization.