International Journal of Humanoid Robotics

This paper proposes a method of giving humanoid robots a natural humanlike walk, which we call the extended-knee walk. Unlike the bent-knee walk of most humanoid robots to date, this walk includes a period in which the knee is fully extended. A parallel mechanism is used in the legs and a method of calculating the walk trajectory copes with the difficulty of the singularity in achieving a humanlike walk. The advantages of this walk were verified from two aspects: good visual appearance and good energy efficiency. An experiment comparing the trajectories of the knee angle during walking showed that the walking style produced by the proposed method is more humanlike than the usual walking style of other humanoid robots. The energy efficiency was verified through power consumption and motor temperature measurements and the possibilities for practical use of this method are discussed with reference to the results of the worldwide soccer competition RoboCup 2008.

Central Pattern Generator (CPG) is used in bipedal locomotion control to provide the basic rhythm signal for actuators. Generally, the CPG is composed of many neural oscillators coupled together. In this paper, the coordination between neural oscillators in CPG is studied to achieve robust rhythm motions. By using the entrainment property of the neural oscillator, we develop a method which uses the difference between oscillator's output and desired output to adjust the inner states of neural oscillators. In the simulation, a CPG structure with coordination between neural oscillators is used to control a 2D bipedal robot. The robot can walk continuously when several external forces are applied on the robot during walking. The method is also implemented on our humanoid robot NUSBIP-III ASLAN for the test of walking forward. With the coordination between neural oscillators, the CPG generated rhythmic and robust control signals which enable the robot to walk forward stably.

This paper proposes a method of energy-efficient trajectory planning for bipedal walking robots. In this study, we plan hip and foot trajectories in Cartesian space using polynomial interpolation. The seven key parameters which define the hip and foot trajectories are optimized by genetic algorithm (GA). Since the hip trajectory is crucial to the stability and walking performance of bipedal robot, we introduce a way to increase hip trajectory's variation by extending the order of the interpolated polynomial and using a set of key parameters. To ensure stable walking motion, we employ the zero-moment-point (ZMP) as the stability criterion. The effectiveness of our proposed method is verified by two simulation examples (flat terrain walking and slope walking) of a humanoid robot named NUSBIP-II.

This paper discusses the generation of a running pattern for a humanoid biped and verifies the validity of the proposed method of running pattern generation via experiments. Two running patterns are generated independently in the sagittal plane and in the frontal plane and the two patterns are then combined. When a running pattern is created with resolved momentum control in the sagittal plane, the angular momentum of the robot about the Center of Mass (COM) is set to zero, as the angular momentum causes the robot to rotate. However, this also induces unnatural motion of the upper body of the robot. To solve this problem, the biped was set as a virtual under-actuated robot with a free joint at its support ankle, and a fixed point for a virtual under-actuated system was determined. Following this, a periodic running pattern in the sagittal plane was formulated using the fixed point. The fixed point is easily determined in a numerical approach. In this way, a running pattern in the frontal plane was also generated. In an experiment, a humanoid biped known as KHR-2 ran forward using the proposed running pattern generation method. Its maximum velocity was 2.88 km/h.

The moving torso plays an important role in the dynamics of bipeds like human beings, the exploitation of which is the essential focus of this paper. A design is presented where the torso is actuated to make the biped walk with the required step-length while allowing the legs to move passively. A periodic response excitation is achieved and the motion of the torso is optimized with respect to the external energy input. A working model of the biped is designed and built in which only the torso is actuated and the legs are passive. This laboratory model is used to test and validate the analytical solutions.

Biped robots possess higher capabilities than other mobile robots for moving on uneven environments. However, due to natural postural instability of these robots, their motion planning and control become a more important and challenging task. This article presents a Cartesian approach for gait planning and control of biped robots without the need to use the inverse kinematics and the joint space trajectories, thus the proposed approach could substantially reduce the processing time in both simulation studies and online implementations. It is based on constraining four main points of the robot in Cartesian space. This approach exploits the concept of Transpose Jacobian control as a virtual spring and damper between each of these points and the corresponding desired trajectory, which leads to overcome the redundancy problem. These four points include the tip of right and left foot, the hip joint, and the total center of mass (CM). Furthermore, in controlling biped robots based on desired trajectories in the task space, the system may track the desired trajectory while the knee is broken. This problem is solved here using a PD controller which will be called the Knee Stopper. Similarly, another PD controller is proposed as the Trunk Stopper to limit the trunk motion. Obtained simulation results show that the proposed Cartesian approach can be successfully used in tracking desired trajectories on various surfaces with lower computational effort.

The paper focuses on design and control of a new anthropomorphic robot arm enabling the torque measurement in each joint to ensure safety while performing tasks of physical interaction with human and environment. When the contact of the robot arm with an object occurs, local admittance algorithm provides active compliance of corresponding robot arm joint. Thus, the whole structure of the manipulator can safely interact with an unstructured environment. The detailed design procedure of the 4-DOF robot arm and optical torque sensors is described in the paper. The experimental results of joint admittance control revealed the feasibility of the proposed approach to provide safe interaction of entire structure of robot arm with a person. The control system with load angle position feedback and lead compensator is proposed to improve dynamic behavior of flexible joint arm. The experimental results show high performance of the developed controller in terms of successful damping of vibrations.