Title :
Decoupling Strategy of Multi-dimensional Force Sensor Based on LS-SVM and αth-order Inverse System Method
Author :
Jingzhu, Zhang ; Kai, Guo ; Cheng, Xu
Author_Institution :
Nanjing Univ. of Sci. & Technol., Nanjing
Abstract :
The coupling of multi-component force sensor is a key factor to limiting its measuring precision. It has become a popular direction of study. The alpha th-order inverse system decoupling strategy is newly proposed in the paper. One strong coupling MIMO system is decoupled into a number of 2nd-order integral subsystems. The dynamic calibration experiments should be processed by applying loads on the force sensor. In this way, the dynamic character of force sensor would be caught. The dynamic relations among inverse system could be indicated with integrators. Static nonlinear function´s approximation of correspondence of inverse system is implemented via training learning network of LS-SVM. Finally, the decoupling of force sensor was realized. Simulation examples verified the availability of the method.
Keywords :
MIMO systems; calibration; force sensors; inverse problems; learning (artificial intelligence); least squares approximations; support vector machines; 2nd-order integral subsystems; LS-SVM; alpha th-order inverse system method; coupling MIMO system; decoupling strategy; dynamic calibration; multidimensional force sensor; training learning network; Calibration; Force measurement; Force sensors; Function approximation; Instruments; MIMO; Machine learning algorithms; Nonlinear dynamical systems; Sensor systems; Support vector machines; LS-SVM; Multi-dimensional force sensor; decoupling; inverse system method;
Conference_Titel :
Electronic Measurement and Instruments, 2007. ICEMI '07. 8th International Conference on
Conference_Location :
Xi´an
Print_ISBN :
978-1-4244-1136-8
Electronic_ISBN :
978-1-4244-1136-8
DOI :
10.1109/ICEMI.2007.4351163
