198
S. Zhan et al.
tribology, control techniques and informatics, has significant advantages in the
following aspects: controlling the lateral vibration of flexible rotating shafts; modifying bearing dynamic characteristics, as stiffness and damping properties; increasing
the rotational speed ranges by improving damping and eliminating instability problems [3]. However, modeling, control and other related problems of the active control
bearing-rotor system, especially of the active lubricated oil bearing-rotor system,
have not been well solved.
In recent years, a variety of active lubricated bearings have been put forward.
These papers mainly include two types of bearings: gas bearing and oil bearing.
Mizumoto [4, 5] has carried out further research on the active lubricated bearings,
and designed a high-speed gas bearing. Moros [6] put forward a gas bearing with
active force generated by piezoelectric actuator installed on the back of the bearing
sleeve. Li and Liu [7, 8] present an active lubricated system based on hydrostatic oil
bearing and restrictor. In this system, the restrictor can control the rotor trajectory
by changing the flow rate into the oil chamber of the hydrostatic bearing. Nicoletti
[9] researched the active lubricated oil bearings with tilting-pad, and designed a
non-linear controller for this system.
The above research explored the application potential of the active lubricated
bearing. At present, there are still some problems in controlling this kind of bearing:
(a) the dynamic equation derived by theory is too complex to be used in active control
of the system; (b) there are many unknown factors in the system, which are difficult to
describe by mathematical model. Based on the above factors, parameter identification
of active lubricated bearing has been studied. Thiesen [10] carried out modeling and
parameter identification of the gas bearing, and designed a controller to enhance the
stability of its system. Matta [11] uses the impact hammer to identify the model
parameters of the pneumatic bearing. Chuang [12] applies the state feedback control
to the system, and then identify the parameters of the system model. These papers
indicate the identification and control method of the lubricated bearings. In addition,
active magnetic bearings (AMB) also have potential applications in this system, so
the identification research of such kind of bearing is also of great significance. In
[13], a predictor-based subspace identification method was proposed to identify the
AMB system, and the obtained model was used to design robust controllers. In [14],
a frequency-based method was proposed for identification of the transfer function
matrix model of a non-rotating shaft supported by AMBs. The method consisted
of steps identifying the submodels separately and finally combining them together.
Wang [15] proposed a weighted instrumental variable (WIV) method to identify the
damping ratio of the AMB system. The experimental results show that this method
can be more accurate than the traditional methods.
In summary, modeling and parameter problems of active gas bearings and AMBs
have been studied, but there is little literature on active hydrostatic oil bearings.
The hydrostatic oil bearing is different from the other two, in terms of large bearing
capacity and self-stability. Therefore, based on the active lubricated bearing-rotor
system proposed by Liu and Li [7, 8], the equipment in this paper adopts the hydrostatic oil bearing and the piezoelectric membrane restrictor to form an active control
S. Zhan et al.
tribology, control techniques and informatics, has significant advantages in the
following aspects: controlling the lateral vibration of flexible rotating shafts; modifying bearing dynamic characteristics, as stiffness and damping properties; increasing
the rotational speed ranges by improving damping and eliminating instability problems [3]. However, modeling, control and other related problems of the active control
bearing-rotor system, especially of the active lubricated oil bearing-rotor system,
have not been well solved.
In recent years, a variety of active lubricated bearings have been put forward.
These papers mainly include two types of bearings: gas bearing and oil bearing.
Mizumoto [4, 5] has carried out further research on the active lubricated bearings,
and designed a high-speed gas bearing. Moros [6] put forward a gas bearing with
active force generated by piezoelectric actuator installed on the back of the bearing
sleeve. Li and Liu [7, 8] present an active lubricated system based on hydrostatic oil
bearing and restrictor. In this system, the restrictor can control the rotor trajectory
by changing the flow rate into the oil chamber of the hydrostatic bearing. Nicoletti
[9] researched the active lubricated oil bearings with tilting-pad, and designed a
non-linear controller for this system.
The above research explored the application potential of the active lubricated
bearing. At present, there are still some problems in controlling this kind of bearing:
(a) the dynamic equation derived by theory is too complex to be used in active control
of the system; (b) there are many unknown factors in the system, which are difficult to
describe by mathematical model. Based on the above factors, parameter identification
of active lubricated bearing has been studied. Thiesen [10] carried out modeling and
parameter identification of the gas bearing, and designed a controller to enhance the
stability of its system. Matta [11] uses the impact hammer to identify the model
parameters of the pneumatic bearing. Chuang [12] applies the state feedback control
to the system, and then identify the parameters of the system model. These papers
indicate the identification and control method of the lubricated bearings. In addition,
active magnetic bearings (AMB) also have potential applications in this system, so
the identification research of such kind of bearing is also of great significance. In
[13], a predictor-based subspace identification method was proposed to identify the
AMB system, and the obtained model was used to design robust controllers. In [14],
a frequency-based method was proposed for identification of the transfer function
matrix model of a non-rotating shaft supported by AMBs. The method consisted
of steps identifying the submodels separately and finally combining them together.
Wang [15] proposed a weighted instrumental variable (WIV) method to identify the
damping ratio of the AMB system. The experimental results show that this method
can be more accurate than the traditional methods.
In summary, modeling and parameter problems of active gas bearings and AMBs
have been studied, but there is little literature on active hydrostatic oil bearings.
The hydrostatic oil bearing is different from the other two, in terms of large bearing
capacity and self-stability. Therefore, based on the active lubricated bearing-rotor
system proposed by Liu and Li [7, 8], the equipment in this paper adopts the hydrostatic oil bearing and the piezoelectric membrane restrictor to form an active control