大量程预紧式六维力传感器及静态标定研究

Large range prestressed six-axis force sensor and study on static calibration

提出一种新型的基于6/3-3Stewart结构的大量程预紧式六维力传感器,介绍了传感器的结构特点和测量原理。假设传感器为线性系统的情况下,研究超静定受力状态下的静态标定原理,并给出两种求解标定矩阵的标定方案。为了改善维间耦合和制造误差等因素带来的非线性影响,利用误差反传算法和径向基函数神经网络,研究传感器的非线性静态标定方法。最后对传感器样机进行静态标定实验,给出了4种标定方案的实验结果并进行比较。标定结果表明该六维力传感器具有较好的测试精度,4种方法有效地解决了预紧式传感器超静定受力而带来的标定难题。

A novel large range prestressed six-axis force sensor based on a modified 6/3-3 Stewart platform architecture is proposed, the structure characteristic and the measurement principle of the sensor is introduced. Provided that the sensor is a linear system, the static calibration theory for the parallel six-axis force sensor based on Stewart plat- form under hyperstatic state is studied, and two calibration schemes to derive the calibration matrix are given. In order to overcome the nonlinear influence of manufacturing error and coupling between different dimensions, the non- linear calibration methods based on back-propagation and radial basis function neural network are also introduced. Finally, the static calibration experiment is carried out, the comparison and discussion of the calibration results using the four calibration methods are presented. The results show that the sensor has better precision, the four methods effectively solve the calibration problem caused by the statically indeterminate structure.