This paper investigate vibration suppression by the active variable stiffness system (AVS system). AVS system can change its stiffness by special elements called on-off elements. The control logic deciding on-off states of on-off element is based on the method of dividing phase plane. A phase plane is composed of displacement and velocity axis. This control logic is easily applicable to both single and multi degree of system, because it is local control logic that determines on-off states according to the relative displacement and velocity between the floors in which each on-off element is installed. By this control logic, on-off elements can absorb and dissipate the earthquake energy. On-off element is on state when the sign of displacement multiplied by velocity is positive, because on-off element absorb the system vibration energy. Conversely, on-off element is off state when the sign of displacement multiplied by velocity is negative, because on-off element dissipate the absorbed system vibration energy. The effectiveness of this control logic can be proved in the second way through the active control experiment using the active mass damper (AMD).

In the case of conventional soft robots, the basic stiffness is small due to the use of flexible materials. Therefore, there is a limitation that the load that can bear is limited. In order to overcome these limitations, a study on a variable stiffness method has been conducted. And it can be seen that the jamming mechanism is most effective in increasing the stiffness of the soft robot. However, the jamming mechanism as a method in which a large number of variable act together is not even theoretically analyzed, and there is no study on intrinsic principle. In this paper, a study was carried out to increase the stability of the force chain to increase the stiffness due to the jamming transition phenomenon. Particle size variables, backbone mechanisms were used to analyze the stability of the force chains. We choose a jamming mechanism as a variable stiffness method of a soft robot, and improve the effect of stiffness based on theoretical analysis, modeling FEM simulation, prototyping and experiment.

In this study, we developed an FSEA(Force-sensing Series Elastic Actuator) composed of a spring and an actuator has been developed to compensate for external disturbance forced. The FSEA has a simple structure in which the spring and the actuator are connected in series, and the external force can be easily measured through the displacement of the spring. And the characteristic of the spring absorbs the shock to the small disturbance and increases the sense of stability. It is designed and constructed to control the stiffness of such springs more flexibly according to the situation. The conventional FSEA uses a fixed stiffness spring and the actuator is not compensated properly when it receives large or small external force. Through this experiment, it is confirmed that FSEA compensates the external force through the proposed algorithm that the variable stiffness compensates well for large and small external forces.

Twisted string actuators (TSAs) are tendon-driven actuators that provide high transmission ratios. Twisting a string reduces the length of the string and generates a linear motion of the actuators. In particular, TSAs have characteristic properties (compliance) that are advantageous for operations that need to interact with the external environment. This compliance has the advantage of being robust to disturbance in force control, but it is disadvantageous for precise control because the modeling is inaccurate. In fact, many previous studies have covered the TSA model, but the model is still inadequate to be applied to actual robot control. In this paper, we introduce a modified variable radius model of TASs and experimentally demonstrate that the modified variable radius model is correct compared to the conventional variable radius string model. In addition, the elastic characteristics of the TSAs are discussed along with the experimental results.