Akio Yamano

Adaptive control method for an undulatory robot inspired from a Leech's nervous system

Many mobile robots using undulatory locomotion have been developed for search or rescue operations in narrow or dangerous places where people cannot enter, because undulatory locomotion can be employed for locomotion in various environments. However, an environmental change will require a mobile robot to change the pattern of undulatory locomotion. Recent studies showed that a leech changes its undulatory locomotion pattern depending on changes in the environment, and this is derived from sensory feedback from the muscle strain to the central nervous system. We propose an adaptive control method for a mobile robot that is inspired by the leechs adaptive locomotion. First, we make clear the role of this adaptive behavior of a leech by using the numerical leech model proposed in the preceding study. After that, we construct the simplified adaptive control method that simulates the leechs adaptive nervous system. Finally, we evaluate the effectiveness of our proposed control method with a numerical model. Our study showed that the proposed adaptive control method achieves regulation of body load due to the fluid despite a change in the fluid viscosity. Moreover, our control method reproduced the adaptive behavior of a living leech to changes in the fluid viscosity.

Basic study of power reduction for shape memory alloy actuator drive

Use of shape memory alloy (SMA) for a moving body has some merits against use of the other conventional actuators in view of high specific power, downsizing and simplification of actuators. On the other hand, large power consumption of SMA prevents to put to practical use. This paper shows optimization of input voltage into SMA actuators. First of all, a numerical actuator model is identified from measured data. After that, input voltage waves are optimized under three evaluation functions by using multi-objective optimization; maximum input voltage, time averaged input power and time averaged bending moment as output from SMA actuator. Obtained numerical results showed that high input maximum voltage reduced power consumption.