In this paper, we propose a sampled-data control scheme for current sharing in DC microgrids with average voltage regulation and demonstrate its effectiveness through laboratory experiments. The current sharing problem in DC microgrids can be formulated as an output consensus control problem for port-Hamiltonian systems under unknown constant disturbances, for which a control scheme has been developed. However, in practical settings, control schemes are implemented in a sampled-data manner. As a theoretical contribution, we in this paper show that a suitably discretized continuous-time controller achieves output consensus for appropriately discretized port-Hamiltonian systems. Furthermore, for DC microgrids, average voltage regulation in the sampled-data setting can be achieved by a simple modification of the control scheme. The proposed method is validated through laboratory experiments.

This paper presents a new online controller parameter tuning method based on Gaussian process regression (GPR) with experimental data collected during the control process. The proposed method adaptively adjusts parameters in response to modeling errors and unknown disturbances, and provides robust tuning against learning uncertainties by utilizing the posterior variance obtained from the GPR. In particular, our method deals with a simplified approximate model for control targets that are difficult to obtain a linear model, such as multi-rotor Unmanned Aerial Vehicles (UAVs), and tunes parameters to approach an ideal controller without elucidating the specific characteristics of the target system. To confirm the effectiveness of the proposed method, we conduct a real-machine verification of PID tuning using a quadrotor UAV.

Nets are portable due to their flexibility, and if they can be used as a climbing target, the robot can descend by hanging a net in an inaccessible area, thereby expanding its field of movement. Furthermore, a robot that climbs with manipulators can reduce its weight by eliminating extra manipulators for the work. However, this method has a problem when the climbing target is a net, because the net is flexible and the grasping position is uncertain. To solve the problem of indefinite grasping position, a grasp confirmation method that judges whether the grasp is successful or not has been proposed. However, the existing methods are limited when the climbing target is a rigid object such as a fence. In this research, the gripper imitates the shape of a carabiner to realize grasp confirmation with a flexible net by detecting movement in or out of the annulus of the climbing target by the degree of rotation of the variable parts. In addition, by attaching servo motor to the variable component, it realizes grasping objects in a similar way to the existing grippers. As the result, the same gripper to perform grasp confirmation to determine the success or failure of grasping, and to grasp an object. By using these, the same machine can perform up-down, left-right movement on the net surface and grasping of an object by manipulation.

Our previous work has demonstrated that operability improves when the predicted state is displayed to the human operator in an inverted pendulum stabilization task. To elucidate the underlying mechanism, we analyzed the stability of the linear model with respect to prediction periods and feedback gains.