基于神经网络的机器人关节误差研究

Research on Joint Error of Robot Based on Neural Network

机器人在日常生活和工业中拥有一定的优势。首先阐述了机器人关节误差的类型和ΔP和ΔR指标的两个特征,利用BP神经网络的应用建立机器人关节误差类型的评估模型,将机器人关节类型的区域误差分为可调节和不可调。通过对不同类型的机器人关节误差的分析提出了合理的建议。

Robot is of certain advantage in both daily life and industrial production.This paper first elaborates on the types of robot joints errors and the characteristics of two indexesΔp andΔR.Then,it proposes that robot joints errors can be divided into adjustable and non-adjustable errors,establishes evaluation model of the types of robot joints errors by applying BP neural network,and divides the areas of the types of adjustable and non-adjustable robot joints errors.Finally,this paper offers reasonable proposals for different types of robot joints errors.