Optimal design of a hyperstatic Stewart platform-based force/torque sensor with genetic algorithms

This paper proposes a novel hyperstatic six-component force/torque sensor structure based on the Stewart platform, and presents the parameter optimization of the sensor structure with genetic algorithms. The structure characteristic of the hyperstatic sensor is analyzed compared with the traditional Stewart platform-based sensor. The mathematical expression of the sensor’s force mapping matrix is built by using the screw theory. In this paper, the condition number and generalized amplifying index are used to evaluate the isotropy and sensitivity of the sensor, respectively. Based on genetic algorithms, the optimal design of the sensor structure is performed with the objective of achieving high measurement sensitivity and well-conditioned transformation between the input and output forces. The computed results prove the better performance of the proposed sensor structure and the effectiveness of the optimization algorithm.