Research on Static Characteristics of Six-Dimension Force Sensor

The static coupling of six-dimension force sensor is one major factor of limiting its measuring accuracy. This paper analyzed the fundamental principle of the linear decoupling on six-dimension force sensor, and compared two different linear decoupling algorithms with the index of isotropy, which are based on the Cramer theorem and least square method. Two nonlinear decoupling methods of data base and Radial Basis Function neural network were proposed. The research result indicates that the isotropy of calibration matrix gained by least square method is superior to the one by Cramer theorem and that the linear error and dimensional interference are greatly reduced by using the RBF neural network model to approach the functional interrelation of generalized force vector and output voltage, whose overall error is less than 1% full scale.