Experimental realization of dynamic walking of the biped humanoid robot KHR-2 using zero moment point feedback and inertial measurement

This paper describes a novel control algorithm for dynamic walking of biped humanoid robots. For the test platform, we developed KHR-2 (KAIST Humanoid Robot-2) according to our design philosophy. KHR-2 has many sensory devices analogous to human sensory organs which are particularly useful for biped walking control. First, for the biped walking motion, the motion control architecture is built and then an appropriate standard walking pattern is designed for the humanoid robots by observing the human walking process. Second, we define walking stages by dividing the walking cycle according to the characteristics of motions. Third, as a walking control strategy, three kinds of control schemes are established. The first scheme is a walking pattern control that modifies the walking pattern periodically based on the sensory information during each walking cycle. The second scheme is a real-time balance control using the sensory feedback. The third scheme is a predicted motion control based on a fast decision from the previous experimental data. In each control scheme, we design online controllers that are capable of maintaining the walking stability with the control objective by using force/torque sensors and an inertial sensor. Finally, we plan the application schedule of online controllers during a walking cycle according to the walking stages, accomplish the walking control algorithm and prove its effectiveness through experiments with KHR-2.     

Gait transition for a quadrupedal robot by replacing the gait matrix of a central pattern generator model

The pattern-generator-based approach for legged robot control is inspired by biological neural mechanisms of locomotion, in which a special challenge is gait transition. In this paper we build a holosymmetric central pattern generator model and propose parameter-setting principles for a gait matrix capable of producing typical quadrupedal gaits, and based on them present an approach of directly replacing the gait matrix for gait transition, with a focus on three problems emerging during transition: breakpoint, phase-lock and oscillation-stop. Breakpoints are smoothed by remaining the current outputs during transition, similar to a zero-order holder. Breaking the phase-lock is accomplished by adding a perturbation to the state matrix at the transiting point. An oscillation-stop of less than one period can be ignored. With such treatments, it is proved that gait transitions between any two gaits on a quadrupedal robot can be achieved at arbitrary phase locations in a walk cycle, theoretically and experimentally in part.     

Autonomously generating efficient running of a quadruped robot using delayed feedback control

We report on the design and stability analysis of a simple quadruped running controller that can autonomously generate steady running of a quadruped with good energy efficiency and suppress such disturbances as irregularities of terrain. In this paper, we first consider the fixed point of quasi-passive running based on a sagittal plane model of a quadruped robot. Next, we regard friction and collision as disturbances around the fixed point of quasi-passive running, and propose an original control method to suppress these disturbances. Since it is difficult to accurately measure the total energy of the system in a practical application, we use a delayed feedback control (DFC) method based on the stance phase period measured by contact sensors on the robot's feet with practical accuracy. The DFC method not only stabilizes running around a fixed point, but also results in the transition from standing to steady running and stabilization in running up a small step. The effectiveness of the proposed control method is validated by simulations. MPEG footage of these simulations can be viewed at: http://www.kimura.is.uec.ac.jp/running.