We are researching the design and control of robots composed of multi-DOF linkages that can perform jumping motions. We call such a robot the "multi-DOF jumping robot". We take a 5 link, 4 joint serial link in a vertical plane as an example of a multi DOF jumping robot. We discuss the process to determine its design parameters from the viewpoint of realizing actual mechanical models. The design parameters include not only mechanical parameters but also motion planning parameters, because the motion condition of each joint strongly depends on motion planning. First, we show the type of jumping motion that we deal with and derive the conditions to achieve that. Then, we propose the motion planning method for jumping motions on the basis of a boundary state between ground and aerial phases. Next, we deliberate the limit characteristic of DC motors that are suitable for multi-DOF jumping robot design, and show the equations that express the limit characteristic. Then, we present one set of practical design parameters, and show simulation results of two types of jumping motions, normal jump and somersault.

As many biomechanists indicate, a tendon plays an important role in running or jumping motions. It stores the kinetic energy as a potential energy during stance and also absorbs the impulse at touch down. Inspired by such biomechanical studies, we propose a simple mechanical model for the hindlimb of a dog to realize a robot that imitates a dog running, and the hardware design of a one-legged running robot, "Kenken". The robot has an articulated leg and uses two hydraulic actuators as muscles and a tensile spring as a tendon. The spring being attached like the gastrocnemius or plantaris enables the robot to produce sufficient propulsion force by virtue of the "energy transfer" from the knee, even if there is no actuator at the ankle joint. Using an empirical controller based on the characteristic dynamics of the model, the robot has succeeded in planar one-legged hopping. Although the problem related to the stability at higher speeds remains, experimental results demonstrate that the proposed hindlimb mechanism is effective for legged running.

A virtual robot tracking control approach with a consideration of clearance is proposed. An algorithm which is used to avoid collision between robots is also introduced using a transformation of Cartesian to distance coordinate. Each robot is indexed by different priority numbers to indicate that which robot is prior to another. A small interval feedback control is applied to a low-priority robot to avoid collision with the higher one when it detects a possibility of collision. At the end of small-interval feedback control, the collision-free condition is established. The virtual robot tracking control is applied again to maintain the overall formation. As a result, the distribution of robots in a system is guaranteed to be in limited communication range along the motion. Numerical simulation highlights the efficiency of our control techniques.

A new digital control method for a time varying sampled data system, called a variable period deadbeat control, is proposed to control high order nonholonomic chained systems with drift terms. This control law has a potential to avoid an excessive overshoot to prevent the trajectories close to the singular manifold of the transformation needed to obtain the chained forms. The approach described is applied to the control of a three link underactuated manipulator.

In order to control gymnastic and jumping robots, we derive the complete analytical solution to the posture control problem of a two-link free flying object with initial angular momentum. We show that the solution involves singular control and derive formulae to calculate the optimal switching condition, optimal terminal time and optimal trajectories. As an application, a high diving motion is simulated.

In order to control gymnastic and jumping robots, we derive the complete analytical solution to the posture control problem of a two-link free flying object with initial angular momentum. We show that the solution involves singular control and derive formulas to calculate the optimal switching condition, optimal terminal time and optimal trajectories. As an application, a high diving motion is simulated.