Novel Calibration Methodologies for Compliant, Multiaxis, Fiber-Optic-Based Force/Torque Sensors

Abstract

In this article, we present a number of novel calibration methodologies for compliant, multiaxis, fiber-optic-based force/torque sensors and evaluate the performance of these methods in real-time experiments with our custom-designed sensors. These methods address the challenges arising from the complex dynamic behavior of a compliant sensor and the nonlinearities in the force–deflection relationship. This article also investigates compliant sensor performance against its response characteristics, such as the impact of compliance on the sensor’s bandwidth using dynamic modeling and identification process. A powerful hysteresis compensation solution using a dynamic estimation model is also presented. Furthermore, we propose and apply a new calibration strategy building on the combination of a linear dynamic model combined with a static nonlinear model. This includes a state space (SS) model with Gaussian Process Regression (GPR) model named SSGPR. The results achieved from the proposed calibration method have revealed an improvement from an <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}$ </tex-math></inline-formula> -squared value of 93.86%–100% when compared to data obtained using a linear dynamic model.