Abstract
The reliability of electrogoniometers as wearable sensors depends on the biomechanical characteristics of the human body. The major causes of electrogoniometric measurement errors are the complexity of movements and the anatomical features of joints. Accordingly, the main objective of this study is to enhance goniometric measurements of body kinematics as a dynamical form of calibration method. In this case study, an experiment was designed in the reachable workspace for gathering joint angles and modeling the related information among them, so a nonlinear neural network was established to compensate the goniometric measurement errors with mapping angular variations simultaneously. The results can be confirmed that the proposed method compensated measurement errors. According to the mean absolute error (MAE) criterion, the results indicated up to 4.5 cm difference between the desired and estimated values of target positions in a reachable workspace and elucidated the significant effect between the desired and estimated 3D target positions before and after using the proposed method. Consequently, since the electrogoniometer has been limited to measuring variations of joint angles separately, the most striking result to emerge from the present study is that the proposed method copes with the finding relationship among measurements of upper extremity kinematics to compensate the goniometric measurement errors.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call