JP4585346B2 - Control method of ultrasonic motor - Google Patents

Description

  The present invention relates to an ultrasonic motor control method for moving a moving body with high accuracy while reducing a positional deviation.

  For example, a positioning system that drives an ultrasonic motor to move a moving body to a predetermined position is known. Here, since the ultrasonic motor moves the moving body by using the frictional force generated between the ultrasonic motor and the ultrasonic motor, a drive signal (input voltage) is sent to the ultrasonic motor in order to start the ultrasonic motor in a stationary state. Even if it is turned on, there is a so-called dead zone where the moving body does not move until the input voltage rises to a certain value or until the resonance is stabilized.

  Therefore, as a method of improving the starting characteristics of the moving body in the positioning system, for example, a method of sweeping the driving frequency when starting the ultrasonic motor, a method of increasing the driving voltage of the ultrasonic motor, or a dead zone of the ultrasonic motor is fed. Methods for incorporating forward control have been proposed (see, for example, Patent Documents 1 and 2).

  FIG. 6 shows a control block diagram according to an example of such a conventional control method. A profile generator (PG) 50 generates a trajectory for moving the moving body to a command position based on a movement instruction value from a host or the like. The detected value of the encoder (ENCODER) 52 that measures the actual position of the moving object is subtracted from the position indication amount sp created by the profile generator (PG) 50, thereby obtaining a position error. A position feedback gain is obtained by multiplying this position error by a coefficient Kep for position feedback. Further, the speed instruction value sv is obtained by differentiating the position instruction quantity sp with a differential operator s, and the speed feedforward gain is obtained by multiplying the speed instruction value sv and a coefficient Kvf for speed feedforward. Further, the differential value of the detected value of the encoder (ENCODER) 52 is multiplied by a coefficient Ks for speed feedback to obtain a speed feedback gain.

  These position feedback gain, speed feed forward gain, and speed feedback gain are added together and input to the PID calculation unit G (PID) 51. A voltage obtained by combining the voltage value obtained by processing by the PID calculation unit G (PID) 51 and the offset voltage value (offset) for enhancing the starting characteristics in consideration of the dead zone is the ultrasonic motor 10 (in FIG. 6). Applied to "USM"). Then, the position of the moving body is detected by the encoder 52, and this detection value is used for feedback control as described above. In the feedforward control shown in FIG. 6, various coefficients Kvf, Kep, and Ks for obtaining the speed feedforward gain are constant values.

  However, as described above, the ultrasonic motor uses a frictional force acting between the moving body and the surface state of the portion that contacts the ultrasonic motor in the moving body is the same over the entire range. In addition, the surface state of the contact portion between the moving body and the ultrasonic motor changes every moment due to frictional wear between the moving body and the ultrasonic motor. Even if the vehicle is started under certain conditions, there arises a problem that the starting characteristic of the moving body changes depending on the stop position of the moving body.

In addition, as described above, the surface state of the portion that contacts the ultrasonic motor in the moving body is not always the same over the entire range, and changes every moment due to wear. In the feedback control, the actual speed of the moving body deviates from the planned speed (command speed) and position (command position), that is, the speed deviation and the position deviation become large and the control accuracy is lowered. .
Japanese Patent Laid-Open No. 04-322179 Japanese Patent No. 3241713

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide an ultrasonic motor control method capable of moving a moving body with high accuracy while reducing a positional deviation.

That is, according to the present invention, in an ultrasonic motor control method using both feedforward control and feedback control,
In the feedforward control, fluctuations in the driving force of the ultrasonic motor in the moving range of the moving body that moves by driving the ultrasonic motor are converted into data in advance and used as a feedforward value when the actual ultrasonic motor is driven. An ultrasonic motor control method is provided in which the data stored as table data is sequentially updated based on position information of a moving body and drive output of the ultrasonic motor at the position. The

In this control method of the ultrasonic motor, the variable feedforward value, the speed feedforward value is suitably chosen.

  According to the present invention, the feedforward value used for the feedforward control is preliminarily stored as fluctuations in the driving force of the ultrasonic motor due to changes in the frictional force between the ultrasonic motor and the moving body in the moving range of the moving body. When the actual ultrasonic motor is driven, the feedforward value is changed based on this data in accordance with the position information of the moving body. Therefore, an appropriate control value corresponding to the position of the moving body can be set. As a result, it is possible to realize highly accurate drive control with reduced speed deviation and position deviation.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Here, an ultrasonic motor configured by connecting two Langevin type ultrasonic transducers in a substantially V shape will be described as an example. FIG. 1 is a sectional view showing the schematic structure of such an ultrasonic motor 10.

  The ultrasonic motor 10 includes two ultrasonic transducers 11a and 11b having a Langevin type structure, a holding member 12 that holds the ultrasonic transducers 11a and 11b at a predetermined angle (90 degrees in FIG. 1), And a head 13 having a substantially V-shape in contact with the moving body 15. For example, a pressing mechanism 14 such as an air cylinder, a hydraulic cylinder, or a spring coil is attached to the holding member 12, and the head 13 is pressed against the moving body 15 with a predetermined force.

  The ultrasonic transducer 11a includes a bolt 21 having both ends threaded, a cap nut 22 having a screw hole that fits into a screw groove of the bolt 21, and two ring-shaped piezoelectric plates through which the bolt 21 can pass. 23a and 23b, and ring-shaped electrode plates 24a to 24c through which the bolts 21 can pass.

  Similar to the ultrasonic transducer 11a, the ultrasonic transducer 11b includes a bolt 21 ', a cap nut 22', two ring-shaped piezoelectric plates 23a 'and 23b', and ring-shaped electrode plates 24a 'to 24a'. 24c '. Electrodes (not shown) are formed on the front and back surfaces of the piezoelectric plates 23a, 23b, 23a ', and 23b'. The number of piezoelectric plates provided in one ultrasonic transducer is arbitrary and is not limited to two.

  The holding member 12 is provided with a hole for allowing the bolt 21 to pass therethrough. The head 13 includes a contact part 13a in contact with the moving body 15, a connection part 13b / 13b 'connected to the ultrasonic transducers 11a / 11b, and a neck connecting the contact part 13a and the connection part 13b / 13b'. It is comprised from the part 13c * 13c '.

  Screw holes that fit into the thread grooves of the bolts 21 and 21 'are formed in the connecting portions 13b and 13b' of the head 13, respectively. The head 13 is made of a material having excellent wear resistance, for example, a metal material such as stainless steel or cemented carbide, or a ceramic such as alumina, silicon nitride, or silicon carbide.

  As shown in FIG. 1, the piezoelectric plates 23 a and 23 b are arranged so as to be sandwiched between the electrode plates 24 a to 24 c, and bolts 21 are inserted into the holes of the piezoelectric plates 23 a and 23 b, the electrode plates 24 a to 24 c and the holding member 12. The head 13 and the cap nut 22 are attached to the end portions of the bolts 21, respectively. As a result, the piezoelectric plates 23a and 23b are fastened with a predetermined force, and the ultrasonic vibrator 11a is configured. In the same manner, the ultrasonic transducer 11b is configured.

  Normally, a metal material is used for the bolt 21, the cap nut 22, and the holding member 12. In this case, the electrode plates 24 a and 24 c are electrically connected to the cap nut 22 through the holding member 12. Therefore, the holding member 12 or the cap nut 22 of the ultrasonic transducer 11a can be used as a ground electrode for driving the piezoelectric bodies 23a and 23b. At this time, the piezoelectric plate 23a ′ included in the ultrasonic transducer 11b. -The ground for driving 23b 'can be taken simultaneously.

  For the piezoelectric plates 23a and 23b, piezoelectric ceramics such as PZT are preferably used. The directions of polarization of the piezoelectric plates 23a and 23b are symmetric with respect to the electrode plate 24b sandwiched between the piezoelectric plates 23a and 23b. The electrode plates 24a and 24c are electrically connected to each other. Therefore, when a voltage is applied between the electrode plate 24b and the electrode plate 24c, the piezoelectric plates 23a and 23b are displaced (vibrated) in the same phase. That is, the piezoelectric plates 23a and 23b extend in the thickness direction or contract together.

  By applying a voltage signal having a resonance frequency to the piezoelectric plates 23a and 23b to vibrate the ultrasonic transducer 11a, this vibration is magnified by the neck portion 13c and transmitted to the contact portion 13a. At the same time, when a voltage signal having a resonance frequency that is 90 degrees out of phase with this voltage signal is input to the piezoelectric plates 23a 'and 23b' to vibrate the ultrasonic transducer 11b, this vibration is also caused by the neck portion 13c '. It is enlarged and transmitted to the contact part 13a.

  In this manner, in the contact portion 13a, the vibrations transmitted from the two ultrasonic transducers 11a and 11b via the neck portions 13c and 13c ′ are combined to generate an elliptical motion. The moving body 15 can be moved in a predetermined direction according to the rotation direction of the elliptical motion. Further, when the phase of the voltage signal for driving the two ultrasonic transducers 11a and 11b 'is reversed, the direction of rotation of the elliptical motion generated in the contact portion 13a is reversed, so that the moving direction of the moving body 15 is also reversed. be able to.

  Next, a method for driving the ultrasonic motor 10 will be described. FIG. 2 shows a schematic configuration of a control block of the ultrasonic motor 10. Various operations described below can be performed by one or a plurality of processors (CPUs), and the processor (CPU) is connected to a storage device such as a semiconductor memory storing a control program, a control data table described later, and the like. It is assumed that information is transmitted and received between them.

  Based on the movement instruction value from the host or the like, the profile generator (PG) 50 generates a trajectory for moving the moving body 15 to a predetermined position. The detected value of the encoder (ENCODER) 52 that measures the actual position of the moving body 15 is subtracted from the position indication amount sp created by the profile generator (PG) 50, thereby obtaining a position error.

  A position feedback gain is obtained by multiplying this position error by a coefficient Kep for position feedback. Further, based on this position error, a calculation amount g for the coefficient Kvf for determining the speed feedforward gain is obtained, and a coefficient Kvf corresponding to the position of the moving body 15 is determined by this calculation amount g. This is because the surface state of the portion that contacts the ultrasonic motor 10 in the moving body 15 is not always the same over the entire range. That is, the coefficient Kvf is handled as a variable parameter.

In order to determine the coefficient Kvf according to the position of the moving body 15, for example, the ultrasonic motor 10 is driven in advance under a predetermined condition to collect moving speed data of the moving body 15 and analyze it. As shown in FIG. 3, a control data table in which the position (P _{1 to} P _n ) of the moving body 15 is associated with the coefficient Kvf _{1 to} Kvf _n of the speed feedforward gain at that position is created. This control data table is stored in a readable / writable storage medium such as a semiconductor memory, for example.

Then, it is assumed that the position P _k (k = 1 to n−1) of the moving body 15 is recognized by the position error obtained at a certain timing, and the coefficient Kvf _k is determined at that time. Timing after a predetermined time has elapsed from the timing, for example, at the timing after one sampling time course of the encoder (ENCODER) 52, when the position of the moving body 15 has been recognized that changes from P _k to P _{k + 1} is The control data table is referred to and the coefficient Kvf _k is changed to the coefficient Kvf _{k + 1} .

When the position of the moving body 15 recognized by the encoder (ENCODER) 52 is recognized as being in the middle of the adjacent positions shown in Table 3, for example, the coefficient Kvf _k is used as the coefficient Kvf, or the coefficient considering the proportions shown with Kvf _{k + 1,} or using the average value of the coefficient KVF _k and the coefficient _{Kvf k + 1,} or whether the difference in coefficient KVF _k and the coefficient _{KVF k + 1} values that closer to either the position _{P k} and _{P k + 1} The coefficient Kvf can be determined using various approximation methods such as obtaining a new coefficient Kvf by proportional distribution.

  The speed feedforward gain is obtained by multiplying the coefficient Kvf thus obtained by the speed instruction value sv. The speed instruction value sv is obtained by differentiating the position instruction amount sp created by the profile generator (PG) 50 with a differential operator s. Further, the differential value of the detected value of the encoder (ENCODER) 52 is multiplied by a coefficient Ks for speed feedback to obtain a speed feedback gain. Further, the acceleration instruction value sa is obtained by differentiating the speed instruction value sv with the differential operator s, and the acceleration feedforward gain is obtained by multiplying this by the coefficient Kaf for determining the acceleration feedforward.

  The position feedback gain, speed feed forward gain, and speed feedback gain obtained in this way are added together and input to the PID calculation unit G (PID) 51. Considering the voltage value obtained by processing by the PID calculation unit G (PID) 51 and the dead zone of the ultrasonic motor 10 (that is, the voltage value that cannot move the moving body 15 even when a voltage is applied). A voltage obtained by adding the offset voltage value (offset) for enhancing the starting characteristic and the acceleration feedforward gain is applied to the ultrasonic motor 10 (indicated by “USM” in FIG. 2). The position of the moving body 15 is detected by an encoder (ENCODER) 52, and as described above, this detected value is used for feedback control.

  In such drive control of the ultrasonic motor 10, feedforward control and feedback control are performed using an optimum feedforward value according to the position of the moving body 15, so the speed and position of the moving body 15 can be set with high accuracy. Can be controlled. FIG. 4 is a graph showing command speeds when the linear stage is moved using two ultrasonic wave motors using the ultrasonic motor 10 having a predetermined shape, and the actually obtained speed deviation (that is, the command speed). The graph showing the difference between the value and the actually measured speed).

  The command speed curve in FIG. 4A and the command speed curve in FIG. 4B are the same. FIG. 4 (a) shows a result of using the driving method described above with reference to FIG. 2 and having a constant coefficient for determining the speed feed forward, and the speed deviation at the start. Variation is extremely large, and the speed deviation when moving at a constant speed is also large. That is, it cannot be said that the position controllability and the speed controllability are high.

  On the other hand, FIG. 4 (b) causes the correction table to be learned by operating about 5 reciprocations using the driving method shown in FIG. 2 in which the coefficient for determining the speed feed forward is changed according to the position of the moving body. The speed deviation at the time of starting and at the time of moving at a constant speed is always constant, and the speed deviation is smaller than that in FIG. It was confirmed that highly accurate drive control was realized.

  As described above, when the moving body 15 is driven using the driving method shown in FIG. 2, the surface state of the contact portion between the moving body 15 and the ultrasonic motor 10 changes every moment due to friction. The value of the coefficient Kvf determined for each position shown also changes. Further, the frictional force generated between the moving body 15 and the ultrasonic motor 10 also changes when the temperature, humidity, or the like of the environment used changes. Thus, when the change of each coefficient Kvf becomes large, the actual speed of the moving body 15 deviates from the command speed and the command position, which causes a problem that the control accuracy is lowered.

  In order to avoid this problem, each value of the coefficient Kvf shown in FIG. 3 is sequentially updated based on the position error (position information) of the moving body 15 and the drive output of the ultrasonic motor 10 at that position ( That is, it is preferable to have a configuration in which data is rewritten and stored in a storage device such as a semiconductor memory. FIG. 5 shows a schematic control block for performing such control for updating the coefficient Kvf.

  In the control method shown in FIG. 5, a calculation amount g for correcting the coefficient Kvf is calculated based on the output obtained by the PID calculation unit G (PID) 51, and a control data table is calculated based on the calculation amount g. Each Kvf value stored in is sequentially rewritten. In this way, since the latest coefficient Kvf is selected according to the position of the moving body 15, the reliability of the speed feedforward gain is increased, thereby improving the controllability.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to such a form. For example, in the above description, the ultrasonic transducers 11a and 11b shown in FIG. 1 are arranged at a predetermined angle. However, the ultrasonic motor is not limited to this and is in contact with a moving body. The structure is not limited as long as it is an ultrasonic motor that moves the moving body in a predetermined direction by moving the part to draw an elliptical orbit.

  The method for driving an ultrasonic motor according to the present invention is suitable for controlling a feed mechanism of an XY stage mounted on a semiconductor manufacturing apparatus such as an exposure apparatus (stepper) that requires precise positioning.

Sectional drawing which shows schematic structure of an ultrasonic motor. The figure which shows schematic structure of the control block of an ultrasonic motor. A control data table in which the position of the moving body is associated with the coefficient of the speed feedforward gain at that position. The graph which shows the command speed and speed deviation at the time of driving a stage with an ultrasonic motor. The figure which shows another schematic structure of the control block of an ultrasonic motor. The figure showing schematic structure of the conventional control block of an ultrasonic motor.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10; Ultrasonic motor 11a * 11b; Ultrasonic vibrator 12; Holding member 13; Head 13a; Contact part 13b * 13b '; Connection part 13c * 13c'; Neck part 14; Pressing mechanism 15; Bolts 22 and 22 '; cap nuts 23a, 23b, 23a' and 23b '; piezoelectric plates 24a to 24c; 24a' to 24c ';

Claims (2)

In an ultrasonic motor control method using both feedforward control and feedback control,
In the feedforward control, fluctuations in the driving force of the ultrasonic motor in the moving range of the moving body that moves by driving the ultrasonic motor are converted into data in advance and used as a feedforward value when the actual ultrasonic motor is driven. An ultrasonic motor control method comprising : sequentially updating the data stored as table data based on position information of a moving body and drive output of the ultrasonic motor at the position . The method of controlling an ultrasonic motor according to claim 1 , wherein the feedforward value is a velocity feedforward value.

Images