Calibration of Piezoelectric Dynamometer based on Neural Networks

This paper presents calibration methods for force measuring piezoelectric dynamometer based on neural networks. Training and test data for neural network is obtained through experiments. Two BP-algorithms, Broyden-Fletcher-Goldfarb- Shanno (BFGS) - Quasi Newton algorithm and Leven berg- Marquardt (LM) algorithm are used to calibrate the dynamometer outputs. Then the proposed method is validated by comparing the outputs of both algorithms. The results prove that LM algorithm based trained network generates precise output compared to the network trained with BFGS algorithm.