Kazuyoshi Hatada

Energy-efficient power assisting methods for periodic motions and its experimental verification

This paper considers power assist control methods for periodic motions in a steady-state. In most power assist applications, additional force is applied in proportion to the instantaneous value of force generated by human. However, human persistent tasks are sometimes periodic and it can be shown that the proportional assist is not optimal for such motions. Therefore, we propose power assisting methods to suppress the velocity fluctuation similar to the repetitive control. By monitoring an energy consumption, we compare the effectiveness and the energy efficiency of each method through experiments.

Power assisting control for electric bicycles using an adaptive filter

This paper considers controller synthesis problem of power assisting system for periodic motions. Most commercial electric power assisting bicycles are using proportional assist to human pedaling torque. However, human pedaling motion is almost periodic and it can be shown that such a strategy is not optimal in the sense of energy efficiency. In our previous work, we observed that pedaling frequency of human fluctuates to some extent. Thus, an adaptive compensation is introduced here. In particular, we propose a power assisting control with adaptive notch filter to suppress the velocity deviation. The effectiveness of the proposed method is demonstrated through numerical simulations.

Robust controller synthesis for a class of uncertain systems and application to visual feedback control

In this paper, we consider the robust synthesis problem of dynamic output feedback controller for a class of polytopic uncertain systems. The motivation dealing with this special class arises from visual feedback stabilization problems under image distortion. The effectiveness of the proposed matrix inequality condition is verified by the experiments of visual feedback stabilization with a cart and pendulum system.

Energy-efficient power assist control with periodic disturbance observer and frequency estimator

In this paper, energy-efficient power assist control problem for almost-periodic motions is considered. Specifically, we focus on the electric power assisted bicycles. Previously, it was shown that the energy efficiency improves by flattening the pedaling torque pattern. To cope with the frequency fluctuation of our motion, we introduce the periodic disturbance observer and a frequency estimator and suppress the pulsation of the human pedaling torque based on a disturbance observer framework. The effectiveness of the proposed method is demonstrated through experiments with an actual bicycle.

Power assisting method for almost-periodic motions based on motion prediction

In this paper, we consider a synthesis procedure of power assisting systems for almost-periodic motions. In our previous study, we showed that the energy efficiency can be improved when a flat torque pattern is realized by removing the input pulsation. In this study, we apply a motion prediction technique based on the Time-Varying Seasonal AR model to the design problem of power assisting systems. The effectiveness of the proposed method is demonstrated through numerical simulations.