※一部利用できない機能があります
| 1.
論文 |
論文
|
井開, 重男 ; 樫本, 弘 ; 大沢, 恵一 ; 長屋, 幸助
概要:
application/pdf<br />Journal Article<br />This paper presents an actuator with a sensor for controlling vibrations of machines. It consists of a voice coil type electromagnetic actuator connected to piezoelectric sensor and a coil spring.
…
The actuator is used for controlling a motor laid on the beam. The disturbance force due to the centrifugal force of the motor is detected by the sensor, and the force which is equivalent to the disturbance force is applied to the motor and the beam by the actuator. This implies that the disturbance force is canceled. The control technique as just mentioned has advantages because it reduces the vibration in the low frequency region, in addition, the actuator becomes compact, and the control force becomes small. But the control is insufficient for controlling the resonance peak. The PD control is available for reducing the resonance peak, but the displacement and velocity sensors are required. In the present article, a method of vibration control of disturbance cancellation combining the PD control is presented in which both the displacement and velocity sensors are not required. The analytical results for obtaining the control current and the response of both the motor and beam have been obtained. To validate the present actuator sensor system and the control method, experimental tests are carried out.
続きを見る
|
||
| 2.
論文 |
論文
|
李, 連進 ; 長屋, 幸助
概要:
application/pdf<br />Journal Article<br />This paper presents a method for optimizing vibration absorbers on a flexible beam by means of a neural network procedure. Analytical results for the response of the beam with multi-absorbers
…
have been obtained using a transfer matrix method. To achieve optimal control, a cost function composed of second-order nonlinear forms in terms of the displacement, strain, and frequencies was used. A method of minimizing the function was presented, in which input values, weights and output function were effectively controlled on the basis of the back-propagation procedure of the neural network method. Numerical calculations were carried out for an example problem. It is ascertained that the present method gives a better result than that obtained by the previously published method. The validity of the method and the analytical results was confirmed by experimental tests.
続きを見る
|
