論文

© 2018 IEEE. This study proposes a novel prosthetic hand that can artificially replace human sensation using haptic transplant technology. Conventional motorized prostheses are unable to precisely control the grasping force of the hand due to the lack of force sensation. In this paper, a haptic prosthetic hand is developed that allows for intuitive operation and precise force adjustment. These functions are realized by artificially transplanting the tactile sensation of a healthy part of the user's body to the amputated part of their body. The haptic transplant technology transmits the force sensation of the prosthesis to the master interface attached to the amputee's healthy part of body using bilateral control without a force sensor. Furthermore, variable compliance control is proposed for flexible adaptation to the shape of the object being grasped. The effectiveness of the proposed system was experimentally verified by comparing the controllability of the proposed system with that of a myoelectric prosthesis. In summary, an artificial replacement for human sensation was developed, and the intuitive operability and flexibility of the proposed system were confirmed.

© 2017 IEEE. Hand recovery process is a major issue in the rehabilitation field as hands are vital to perform most daily activities. This paper presents the design and evaluation of a one actuated degree-of-freedom exoskeleton finger using flexible actuator to assist patients in rehabilitation process and more specifically in tendon recovery protocols. The control strategy based on the motion-copying system is used to be able to take advantages of the sensation of the patient. The wearability and adaptability of the device are improved, compared to other specific devices, mainly by the use of remote actuation and light-weight materials. Evaluation of the device in terms of wearability, adaptability, repeatability and accuracy of the position estimation are conducted. All these criteria are confirmed. The next step is to test the device in actual recovery protocols to evaluate its efficiency.

This paper proposes a bilateral control system implementing an admittance control system. Bilateral control system realizes teleoperation of a remote robot with a sense of touch. We focus on an admittance-based bilateral control system, which needs only two communication channels and has almost the same control performance as an acceleration-based bilateral control system, which is one of the highest transparent teleoperation systems. However, the conventional admittance-based system assumes an ideal communication situation where there is no communication time delay and no packet loss and has possible to increase position error as time goes on in practical use. The proposed admittance-based system achieves the same control performance as the conventional structure without increasing position error even if there is a communication delay. The validity of the proposed system is confirmed through simulations and experiments.

Admittance-based bilateral control is reported to have the same performance as 4ch bilateral control with only force communication. However, admittance-based bilateral control was easy to vibrate when in contact with the hard environment. Previous studies did not assume the contact vibration. Damping injection is one of the effective methods to suppress the vibration, however it generates the extra operational force proportional to velocity. Thus the operability deteriorates. To suppress the contact vibration, this paper proposed admittance separation method. This method used different admittance values for position control and acceleration feedforward. Making a smaller values of admittance for position control is effective in suppressing vibration when in contact with the environment. Moreover, the extra operational force is inertial force, so it is generated only at the beginning of the movement. Thus proposal has advantages in terms of operability. The validity of the proposal was confirmed by simulation of contact with the environment.

A corona discharge is a type of self-sustaining discharge, which is often observed between pin-to-plate electrodes under high-voltage application. Force generation using the corona discharge has no moving parts and is robust; therefore it can be used as a new actuator. However, the relationship between the input voltage and output force caused by the corona discharge is highly nonlinear, so it is essential to deal with this issue through control. In particular, it is important to consider the two types of limitations induced by the corona discharge. First, the corona discharge has a lower voltage limit at which the discharge is initiated, thus without the consideration of the corona inception voltage, the response speed deteriorates, and the overshoot increases. Second, excessive applied voltage causes spark discharge, which causes safety problems. Therefore, this paper proposes voltage reference generation and an angle control method based on differential mode voltage that considers the aforementioned voltage range limitations. Simulations confirm that the proposed method improves the response speed by 13.2% and suppresses the overshoot by 28.0% when torque disturbances are added while the applied voltage fit within the desired range.

A corona discharge is a kind of self-sustaining discharge that is generated by applying high voltage to asymmetric electrodes. The thrust by the corona discharge is expected as a new actuator since it has no moving parts, is quiet, and is robust. However, the relationship between the applied voltage and the thrust is highly nonlinear, which means that this problem must be solved by control. Furthermore, a control system that compensates for disturbances such as wind and modeling errors is also required. Therefore, this paper proposes a control system with a voltage range limitation that takes into account the characteristics of corona discharges and the disturbance observer to improve disturbance suppression characteristics. Two experiments were conducted to confirm the effectiveness of the proposed method. The first experiment shows that the proposed method improves the response speed by 70.0% in step response. The second experiment indicates that the proposed method enhances the response speed by 60.0% in disturbance response and suppresses angle change by 42.3%. Through experiments, it is confirmed that the proposed method succeeded in attitude control on the centimeter order and improved control performance.

It is important to understand how humans gen-erate tracking forces when considering the increase of human- robot interactions. Previous research has mostly focused on po-sitional target tasks. A recent study was conducted to investigate the basic characteristics of the force tracking task. That study suggested predictive compensatory behavior in humans during fast rhythmic movements. In this paper, we investigated the relationship between the frequency of rhythmic force generation and predictive compensation in humans. Here we conducted experiments using four square waves with different periods as target values with four healthy young subjects. Subjects were instructed to control the force exerted at the tip of their fingers during a force pursuit tracking task. The results suggest the following: the faster the tracking period, the faster the delay time converges to a steady value; the relationship between the motion period and the rate of convergence is not linear; individual differences in delay time are large when the force generation is transitioning from slow to very fast, which corresponds to the rhythmic task; when the period is extremely short (the period is 1 sec), the ascent and descent force characteristics are different.

In micro-manipulations such as cell manipulation, if the operator can feel the tactile sensation of the object, those operations are expected to be more efficient. Scale bilateral control, in which micro-manipulation is performed remotely and position and force information is transmitted bidirectionally between the two systems, could be one of the solutions for tactile transmission. In this control, the use of a linear motor as a human-operating system and a stacked piezoelectric actuator as a micro-manipulating system has been proposed. In this context, a linear motor is an inertial system, and a piezoelectric actuator is a stiff system. However, the control scheme and theory of bilateral control have been established only for inertial systems. In previous studies, piezoelectric actuators were modeled as inertial systems, and the construction of bilateral control systems according to the scheme of inertial systems was attempted. This paper attempts to construct bilateral control by modeling piezoelectric actuators as a second-order linear system close to the actuator's physical properties. Experiments revealed that the proposed bilateral control scheme performs well as the conventional scheme. This study contributes to constructing bilateral control between two systems with different structures.

Electroaerodynamics (EAD) is a phenomenon that generates a gas motion by applying high voltage between electrodes. One method of creating thrust with EAD is the use of the corona discharge, which is a self-sustaining discharge. Since it does not require any moving parts, the corona discharge produces thrust robustly and quietly. Therefore, this thrust system is expected to be utilized as the thrust source for drones. However, since the relationship between thrust and applied voltage is nonlinear, dealing with this problem in control is indispensable. Although some studies have been conducted on control methods with the corona discharge, few have been performed on multidegrees of freedom (DOF) control based on theoretical modeling, which is essential for achieving attitude control during flight. Thus, this paper proposes a control system based on theoretical modeling to realize 2-DOF height and angle control. This study utilizes the idea of differential mode and voltage range limitations and introduces a disturbance observer (DOB) to construct a control system. Two simulations are conducted to verify the effectiveness of the proposal. The first simulation shows that the proposal can follow the command value with small errors, indicating the possibility of multi-DOF control. The second simulation revealed that the proposal significantly suppressed attitude change under disturbance and contributed to improved responsiveness.

Motion generation technology contributes to the automation of the robots, especially in the industrial field. Imitation learning and motion reproduction system (MRS) are one of the motion generation methods that widely applied. However, machine learning-based methods require a large amount of data and are weak against extrapolation. Although MRS has merit in terms of the data number, it is weak regarding environmental variation. In particular, the environmental impedance variation have occurred, both position and force tracking are impossible. As several conventional studies have been done to reduce only position or force tracking errors, they can not adapt force or position priority tasks, respectively. Therefore, this paper proposes a novel compensation law based on designed control stiffness using the hybrid angle. The validity of this proposal is confirmed by simulation. This paper contributes to the MRS expansion of the environmental impedance variation with consideration of the task's priority physical value.