基于多传感器数据融合的机器人腕力测量  

Multi-sensors Data Fusion Based Measurement of Wrist Force for Robots

作　　者：徐科军[1,2] 李巧利[1] 刘扬[1] 童利标[1] 梅涛[3] 

机构地区：[1]合肥工业大学自动化研究所 [2]中科院合肥智能机械研究所合肥230031 [3]中科院合肥智能机械研究所

出　　处：《仪器仪表学报》2003年第3期286-290,294,共6页Chinese Journal of Scientific Instrument

摘　　要：提出根据手爪上的多个指力传感器的输出变化 ,通过数据融合方法得到腕力的数值。研制了标定腕力与手爪上指力之间关系的实验台 ,进行了标定实验。采用径向基函数神经元网络对实验数据进行训练 ,得到融合网络的权值。Proposes to measure the wrist force using the data fusion method according to output variance of finger force sensors on the gripper. A calibrating bench is developed, which detects the relationship between the wrist force and finger force, and the calibrating experiments are carried out. The experimental data are trained using a radial base function artificial neural network, and the weight values for data fusion are obtained.The results of data fusion are identical to the testing results, which shows the data fusion method proposed by this paper is effective.

关 键 词：机器人 腕力测量 传感器 数据融合 人工神经元网络 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TP212[自动化与计算机技术—控制科学与工程]