Flying Trapeze Act Motion Planning Algorithm for Two-Link Free-Flying Acrobatic Robot
Thanapong Chuangyanyong, Panusorn Chinsakuljaroen, Worachit Ketrungsri, Thanacha Choopojcharoen · 2022 International Conference on Advanced Robotics and Mechatronics (ICARM) · 2022
Most existing pieces of research on flying trapeze robot motion planning are verified on highly abstracted models, which are susceptible to physical uncertainties and physical limitations. To make the control algorithm more versatile and physically realizable, the proposed research was designed and verified to work with a simulation of a realizable two-link free-flying acrobatic robot that mimics real-world circumstances. The resulting algorithm can adjust the posture to reduce the effect of uncertainties in the robot’s physical properties and mishaps while being airborne, by quickly recomputing its posture trajectory using the isolated posture dynamics of itself considering its drifted center-of-mass trajectory on the sagittal plane while concurrently stabilizing the posture with a time-varying LQR. The simulation results show a high success rate in catching a target when the physical parameters are uncertain.