屋根型円筒ラチスシェルを対象としたシステム同定に基づくアクティブ制振

抄録

 1. Introduction
 In applying vibration control techniques to buildings, if control systems are designed based on different dynamic parameters from real buildings, a decline of control effects is expected. Therefore, an accurate grasp of dynamic parameters of the buildings is important in order to control effectively. Current researches on an active vibration control based on estimated dynamic parameters usually focus on tower type buildings. On the other hand, few researches have been reported on shell and spatial structures that have complicate dynamic parameters. We attempt to apply the active vibration control based on the estimated dynamic parameters to a cylindrical lattice shell roof as an example of the shell and spatial structures. First, using control systems based on the dynamic parameters estimated by system identification, the control effects of the active vibration control are demonstrated. Moreover, by numerical analyses, effects of sensor arrangements for control on control stability are revealed, and then are demonstrated in horizontal uni-axial vibration tests.
 2. Outline of vibration tests
 A specimen of the vibration tests is a cylindrical lattice shell roof with its edge span of 1500 mm. Voice coil motors are used as actuators. Modal control and optimum control theory are used to calculate control force. Moreover, a design method for the active vibration control based on system identification is explained.
 3. Effects of sensor arrangements on control stability
 By the numerical analyses, effects of the sensor arrangements for control on the control stability are analyzed. First, a method to evaluate the control stability is explained. Next, the control stability of all sensor arrangement patterns of the number of the sensors for control is evaluated, and characteristics of the sensor arrangements that enable stable control are showed.
 4. Estimation of dynamic parameters specimen has by system identification
 The dynamic parameters the specimen has are estimated by system identification. Natural frequencies, damping ratios and eigen-mode vectors estimated by system identification are almost same as estimated values by other estimate methods and results of eigen analysis. But, factors against actuators are estimated less than the results of eigen analysis. This may be caused by friction.
 5. Analyses of control effects in vibration tests
 Applying the active vibration control based on the dynamic parameters obtained in chapter 4, results of the vibration tests are analyzed. Seismic responses are reduced by 40% by the active vibration control, without amplifying responses of all modes. Moreover, a demonstration of differences of the control effects between the sensor arrangements, that is, an experimental validation of the numerical analyses in chapter 3 is carried out.
 6. Conclusions
 As a conclusion, the following results are obtained.
 1) By the numerical analyses for all sensor arrangement patterns of the number of sensors for control, when measurement data include less a part of control modes compared to a part of modes except for control modes, it becomes easy to select the sensor arrangements which enable the stable control.
 2) The plural dynamic parameters the cylindrical lattice shell roofs have can be estimated accurately by system identification from the responses on the same position as the sensor arrangements for control. Moreover, using the control systems based on the results of system identification, the seismic responses was reduced in the vibration tests.
 3) When the sensors are arranged to the positions that enable the stable control, the seismic response of the cylindrical lattice shell roofs can be reduced by 40% by the active vibration control, without amplifying the responses of all modes.