• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.33 no.10
• /
• pp.1014-1022
• /
• 2009

A biped robot in locomotion can be regarded to be kinetically redundant in that the link-chain from its foot on the ground to its swing foot has more degrees of freedom that needed to realize stable bipedal locomotion. This paper proposes a new method to generate a trajectory for bipedal locomotion based on this redundancy, which directly generates a locomotion trajectory at the joint level unlike some other methods such as LIPM (linear inverted-pendulum mode) and GCIPM (gravity-compensated inverted-pendulum mode), each of which generates a trajectory of the center of gravity or the hip link under the assumption of the dominance of the hip-link inertia before generating the trajectory of the whole links at the joint level. For the stability of the trajectory generated in the proposed method, a stability condition based on the ZMP (zero-moment point) is used as a constraint as well as other kinetic constraints for bipedal motions. A 6-DOF biped robot is used to show how a stable locomotion trajectory can be generated in the sagittal plane by the proposed method and to demonstrate the feasibility of the proposed method.

• Proceedings of the KSME Conference
• /
• 2008.11a
• /
• pp.876-881
• /
• 2008

When biped robots speed up to run and reduce speed to walk after running, it needs stable speed translation. This paper proposed simple speed translation using the modified LIPM (Linear Inverted Pendulum Mode). We can change stride and period time of a biped robot in some bounded sets with this propose algorithm. This method is simple and effective in simulation.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.33 no.1
• /
• pp.27-33
• /
• 2009

Human postural responses appeared to have stereotyped modality, such as ankle mode, knee mode and hip mode in response to various perturbations. We examined whether human postural control gain of full-state feedback could be decoupled along with the eigenvector. To verify the model, postural responses subjected to fast backward perturbation were used. Upright posture was modeled as 3-segment inverted pendulum incorporated with feedback control, and joint torques were calculated using inverse dynamics. Postural modalities such as ankle, knee and hip mode were obtained from eigenvectors of biomechanical model. As oppose to the full-state feedback control, independent eigenvector control assumes that modal control input is determined by the linear combination of corresponding modality. We used optimization method to obtain and compare the feedback gains for both independent eigenvector control and full-state feedback control. As a result, we found that simulation result of eigenvector feedback was not competitive in comparison with that of full-state feedback control. This implies that the CNS would make use of full-state body information to generate compensative joint torques.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.8
• /
• pp.124-133
• /
• 1996

This paper presents the implementation of variable structure control system for a linear or nonlinear system using neural networks. The overall control system consists of neural network controller and a reaching mode controller. While the former approximates the equivalent control input on the sliding surface, the latter is used to bring the entire system trajectories toward the sliding surface. No supervised learning procedures are needed and the weights of the neural network are tuned on-line automatically. The neural netowrk-based variable structure control system is applied to a nonlinare unstable inverted pendulum system through computer simulations, and implemented using a microcomputer (80486-50MHz) and applied to the DC servomotor position control system. Simulation and experimental results show the expected approximation sliding property is occurred. The proposed controller is compared with a PID controller and shows better performance than the PID controller in abrupt plant parameter change.

• Proceedings of the KSME Conference
• /
• 2004.04a
• /
• pp.810-815
• /
• 2004

In this paper, we propose new parallel mechanism of a 3 dimensional biped robot whose each leg is composed of two 3-dof parallel platforms linked serially. This proposed parallel mechanism is able to move freely in the man-made environment and is applied to various fields, such as medical, welfare, and so on. And a total weight of each leg is expected to be lighter than serial linked leg. One side leg consists of a 3-dof orientation platform and 3-dof asymmetric parallel platform. The former consists of three active linear actuators and seven passive joints, and the latter of two active linear actuators, one active rotational actuator and eight passive joints. Thus, there are two kinds of parallel platforms each chain's elements and active joint's positions are different for the biped robot to move freely like a serial link without the kinematics constraints. The effectiveness and the performance of the proposed parallel mechanism and locomotion trajectory are shown in computer simulations with a 12-DOF parallel biped robot.