Modeling sit to stand motion of humanoid robot using telescopic inverted pendulum for predicting stable motion

Chew , Xiao Lin (2014) Modeling sit to stand motion of humanoid robot using telescopic inverted pendulum for predicting stable motion. Project Report. UTeM. (Submitted)

Text MODELING SIT TO STAND MOTION OF HUMANOID ROBOT USING TELESCOPIC INVERTED PENDULUM FOR PREDICTING STABLE MOTION 24pages.pdf Download (587kB)

Abstract

Sit to Stand (STS) motion is a challenging motion for any humanoid robot. Hence, development in humanoid robotics system is essential. In biomechanical fields, several models have been developed through observation of STS motion from human subjects. One of the models developed is based on telescopic inverted pendulum (TIP) concept which is an inverse kinematics concept. TIP model is the most suitable for STS trajectory generation since that TIP model focuses on motion of center of mass of human body in Cartesian space. However, the suitability of TIP model for representing the STS motion of humanoid robot is unknown. Furthermore, the resulting torque (Nm) from motion generated by TIP model has not been validated hence the accuracy of the robot when implementing the STS trajectory from the model is unknown. Therefore, the research on modeling STS motion of humanoid robot using TIP for predicting stable motion is proposed for the contribution to the development of humanoid robotics field as well as rehabilitation, prosthetic and exoskeleton robots. The objectives of this project is to model and validate sit to stand behavior using TIP model by simulation prove and to validate accuracy of TIP model in representing STS motion by comparing the output torque with three-link multi-segment robot. This project is carried out by MATLAB simulation in terms of the TIP model output and three-link multi-segment robot torque. The theoretical (based on three-link multi-segment robot) and actual (based on TIP model) torque (Nm) acted on COM are compared and analyze using statistical technique in terms of mean, percentage error and RMSE. Based on the analysis, there is high accuracy exist for TIP model in representing seat-off movement since the RMSE value is only 0.4207 and there is inaccurate for the seat-unloading phase. The deviation from actual value may be due to the model doesn’t taking account of momentum components. Therefore, in future, these all parameters must be taken into account in order to increase the accuracy of TIP model in representing STS motion.

Actions (login required)

View Item