The use of external sensors can guarantee higher stability and reduced tracking errors on the expense of increased overall system’s complexity and cost. For speed control of ultrasonic motors, an isolated electrode can be embedded within the piezo ring instead of direct speed measurement. The vibration amplitude and rotor speed can be estimated directly through the feedback voltage (FBV) signal. The speed/FBV relationship can be fitted using a linear regressor, or a nonlinear function approximator, and then further integrated with other controllers. However, this complicates controller design due to the decoupling of the speed estimator and the controller. Another control approach that simulatenously perfroms estimation and control is Deep Reinforcement Learning (DRL) which directly learns a model-free optimal nonlinear control policy by iterative trial and error. To apply DRL to sensorless speed control of USM, an NN-based policy learns the mapping from an input state (A Markovian state includes the driving frequency, motor temperature, target speed, and FBV signal) to an optimal action (Driving frequency update step) by maximizing some reward function (a function of the speed error and the control effort). Using the Soft Actor-Critic (SAC) algorithm, optimal sensorless speed tracking was realized both in simulation and experiment.

Tool temperature measurement in cutting is important for optimizing tools, cutting conditions, cutting fluid, etc. However, it is currently difficult to measure the tool temperature, especially the cutting edge temperature, of a machining machine in which the tool rotates. Therefore, we will introduce an example of examining the application of the heat transfer analysis method by CAE for predicting the temperature of the cutting edge.

I examined a method for determining the arrangement and posture of robots that can operate with a low load when multiple industrial robots perform work. When deciding the placement of a robot, the work area of ​​the robot is generally used as a reference, but this criterion alone may impose an unnecessarily large load. Therefore, by establishing a placement determination method focusing on the torque and operating force ellipsoid acting on the motor, it was possible to lead to highly accurate movement with a small load on the joints.

In recent years, there has been an increase in the use of drones for a various of tasks. Using autonomy drones to work safely in high places has been researched in this laboratory. In order to work safely, it is necessary to estimate the position of the autonomy drone with high accuracy and always grasp it. In this study, the Kalman filter is used to improve the self-positioning accuracy of the drone that retrieves lost objects left at high altitude. The results are verified by simulations and actual experiments.