Research on Posture Sensing and Error Elimination for Soft Manipulator Using FBG Sensors

Fiber-optic sensors are highly promising within soft robot sensing applications, but methods based on geometry-based reconstruction limit the capability and range. In this study, a fiber-optic sensor with different deployment strategy for indirect to monitor outside posture of manipulator is presented. The internal support structure’s curvature was measured using FBG sensor, its mapping external pose then modelled modified LSTM network. error assumed follow Gaussian distribution in neural network rectified by maximum likelihood estimation address issue noise generated during deformation transfer structure. For manipulator, model’s performance demonstrated. proposed method’s average absolute 63.3% lower than before optimization, root mean square 56.9% optimization. comparison results between experiment simulation demonstrate viability measurement structure data-driven method, as well significant impact optimization method assumption.