JP2007206830A - Servo control device for moving body and laser processing device - Google Patents

Abstract

To provide a servo control device and a laser processing device for a moving body which are excellent in compensation effect, and to provide a sufficient compensation effect even when there is a word length limitation in the operation of additional input.
In a digital servo controller for a moving body that includes a feedback loop and positions the moving body based on position command data, a transfer function relating to a state quantity of the feedback loop at the time when the position command data is received. , Digital filters 10a and 10b (additional input means) for adding zeros that cancel out the poles unique to the feedback loop are provided, and in addition to the position command data, the outputs of the digital filters 10a and 10b are fed to the feedback loop. Input to position the moving body. In addition, the digital filter has a cascade structure, and the transfer function constituting this structure has poles and zeros that are close to each other on the complex plane, so that a sufficient compensation effect can be obtained even when the word length is limited in the operation.
[Selection] Figure 1

Description

  The present invention relates to a servo control device that detects the position of a moving body and controls the position of the moving body to follow a specified target value, and a laser that controls a moving unit by the servo control device of such a moving body. The present invention relates to a processing apparatus.

  For example, a laser processing apparatus that performs drilling in a printed wiring board manufacturing process requires a positioning control mechanism for irradiating a plurality of processing positions of a workpiece one after another with high processing throughput and high accuracy. In order to realize the processing, a galvanometer mirror control device is often used.

  The laser processing apparatus is usually a numerical control (NC) apparatus having a hierarchical control structure, and the galvanometer mirror control apparatus is included in the lowest hierarchy. In the control device of the upper layer (hereinafter referred to as “upper control”), two-dimensional hole position coordinates are described in the NC program in the order in which the two-dimensional hole position coordinates are processed based on CAM (Computer Aided Manufacturing) data of the printed wiring board. . Here, the order of drilling is optimized to achieve high machining throughput. Such NC programs are created in advance. When machining starts, the host control converts the hole position coordinates in the program one after another and sends time-series angle command data to the galvanometer mirror control device. To do. In order to process the hole into a perfect circle, it is necessary to irradiate the laser beam after the galvano mirror is stopped at an angle specified by the angle command data. For this reason, transmission of angle command data and laser light irradiation control are performed in synchronization within the host control.

  The main components of the galvanometer mirror control device include a galvanometer mirror, a rotary shaft that fixes the mirror, an actuator that changes the mirror angle by rotating the rotary shaft (rotary actuator), and feedback control of the mirror angle Control circuit. As the rotary actuator, an electromagnetic rotary actuator that generates a drive torque on an electromagnetic principle is often used. Furthermore, this rotary actuator has a built-in sensor for detecting the angle of the mirror, and angle detection data from this sensor is fed back to the control circuit. This control circuit may be realized by analog control including an operational amplifier, or may be realized by digital control firmware combining a microprocessor and a program. In particular, in recent years, the latter configuration has made it possible to improve the performance of a control program flexibly in a short period of time, thereby improving work efficiency.

  There are variations of angular stroke in positioning of the galvanometer mirror, and the positioning time varies depending on the angular stroke. Corresponding to one positioning, one angle command data is received as a step input signal. In response to this, when the mirror starts to rotate, integral compensation is performed to match the mirror angle with the angle command data without error. In this compensation, the value obtained by subtracting the angle detection data from the angle command data, that is, the tracking error signal is time-integrated. Furthermore, in order for the feedback loop of the galvanomirror control device to operate stably, it is necessary to sufficiently increase the phase margin and gain margin of the one-round transfer function. For this reason, stabilization compensation or phase advance compensation by the angular velocity signal acts by differentiating the angle detection data or using a so-called state observer. These control methods are well known as the basis of feedback control theory (Non-Patent Document 1).

  In order to shorten the positioning time of the galvanometer mirror, a technique for widening the feedback loop is used. In the case of the above-described electromagnetic rotary actuator, in addition to the mirror, a sensor and a moving coil or moving magnet are attached to the rotating shaft, and these act as inertia loads. May generate vibration. Since there are a plurality of torsional vibration modes in a region of several kHz or more, in the conventional galvanometer mirror control device, the feedback loop is broadened by a vibration mode stabilization compensator. This stabilization compensator estimates and feeds back the torsional vibration mode state quantity (Patent Documents 1 and 2).

  These methods are combined with the above-described integral compensation and phase lead compensation to form a feedback loop. As for the characteristics of the feedback loop, the time required for one positioning (positioning time) satisfies the target specification, and the overshoot and residual torsional vibration included in the transient response (settling response) near the target angle are within the allowable range. To be adjusted.

  A series of angle command data (hereinafter referred to as “angle command pattern”) transmitted from the host control is based on the optimized drilling sequence described above. When the angle command pattern is Fourier-analyzed, it can be seen that each has a different frequency spectrum. If a certain spectral component matches the resonance frequency of the torsional vibration described above, it is considered that the vibration mode becomes a residual vibration and the drilling position accuracy deteriorates. In order to improve the throughput of the laser processing apparatus, the time interval of the step signal of the angle command pattern (hereinafter referred to as “command interval”) tends to be shortened. If the command interval is shortened, the frequency spectrum of the angle command pattern increases the high frequency component. For this reason, residual vibration due to the higher-order vibration mode of the rotary actuator is likely to occur, and there is a concern about deterioration of positioning accuracy. Also, if the command interval is made extremely short, after entering the allowable range of the settling response amplitude, laser light is irradiated before the galvanometer mirror completely stops, and the next positioning operation is started. At this time, if the settling response of the previous positioning remains, the state quantity (hereinafter referred to as “initial state quantity”) of the dynamic characteristic mode included in the feedback loop does not become zero at the start of the next operation. Since the angle command pattern is generally irregular, various initial state quantities can be taken. In particular, in positioning with a small angle stroke, the positioning time is short, and the vicinity of the target angle is reached before the influence of the initial state quantity is sufficiently attenuated, so that the subsequent settling response waveforms are different each time. In order to narrow the allowable range of the settling response amplitude for high-precision machining, a technique for suppressing the settling response amplitude for any angle command pattern is required. For this reason, there is a technique for inputting an additional input signal other than the angle command to the feedback loop (Patent Document 3, Non-Patent Document 2).

The technology disclosed in Patent Document 3 and Non-Patent Document 2 (hereinafter referred to as “conventional additional input type initial value compensation control”) promptly affects the influence of the initial state quantity of the dynamic characteristic mode included in the feedback loop. This technique uses an additional input signal that cancels out the pole inherent in the dynamic characteristic mode in order to attenuate the signal. Accordingly, it is possible to position the moving body with high accuracy even in a short command interval with various angle command patterns, which is effective in improving the throughput and processing accuracy of the laser processing apparatus.
JP 2002-40357 A JP 2002-40358 A Japanese Patent Application Laid-Open No. 2005-221845. Toru Katayama, “Basics of Feedback Control”, Asakura Shoten, May 20, 1987, Chapters 6-7. Noriaki Hirose, Motohiro Kawafuku, Makoto Iwasaki, Hirotake Hirai “Residual Vibration Suppression by Compensating Initial Values during Repeated Positioning Control”, IEEJ Transactions D, Vol. 125, No. 1, pp. 76-83, 2005.

  In conventional additional input type initial value compensation control, there are cases where the control input signal saturates with a higher frequency component than necessary. In this case, the high frequency resonance of the moving body is excited, and sufficient compensation as theoretically possible. It turned out that an effect was not acquired. However, a solution to this problem is not disclosed in any of the above known techniques.

  Accordingly, a first problem to be solved by the present invention is to provide a servo control device and a laser processing device for a moving body that are excellent in compensation effect.

  In addition, in the conventional additional input type initial value compensation control, an optimal implementation structure is not disclosed for a digital filter that generates an additional input signal. Therefore, a sufficient compensation effect can be obtained when double-precision floating-point arithmetic is performed on this digital filter inside the digital control microprocessor, but a theoretical compensation effect cannot be obtained by single-precision floating-point arithmetic. I found out that there was a case. However, for high-order digital filters, single-precision computation is preferable because it can shorten the computation time. However, this problem is not disclosed in any of the above-mentioned known techniques.

  Therefore, the second problem to be solved by the present invention is equivalent to double-precision floating-point arithmetic even when the digital filter operation has a word length limitation (that is, even if single-precision floating-point arithmetic is performed). It is an object of the present invention to provide a moving body servo control device and a laser processing device having a sufficient compensation effect.

  In order to solve the above-described problem, the first means includes a feedback loop, and in the digital servo control device for a moving body that positions the moving body based on the position command data, the first means at the time when the position command data is received. An additional input means for adding a zero point that cancels a pole unique to the feedback loop to a transfer function relating to a state quantity of the feedback loop is provided, and an output of the additional input means is added to the position command data. The moving body is positioned by inputting to the feedback loop.

  The second means includes, as a laser processing apparatus, a servo control device for a moving body in which the moving body is a galvano mirror, and a control means for controlling the angle of the galvano mirror, and the laser reflected by the galvano mirror. By controlling the angle of light by the control means, the laser beam is made incident on a predetermined position of the workpiece to process a hole in the workpiece.

  Since the moving body can be positioned at high speed and with high accuracy, it is possible to improve the processing throughput and processing accuracy. Furthermore, since the word length of the digital filter operation for generating the additional input signal can be shortened, an inexpensive microprocessor can be used, and the device cost can be reduced.

  Hereinafter, a case where the present invention is applied to a galvanomirror control device will be described.

  FIG. 1 is a block diagram of a galvanometer mirror control apparatus according to the present invention. The galvanometer mirror control device is realized by digital control firmware using a microprocessor not shown, and includes an integral compensator 2, a proportional compensator 3, a speed observer 4, a torsional vibration stabilization compensator 5, Processing operations relating to the digital filters 10a and 10b and the adders 20 to 24 are described in a part of a program executed by the microprocessor. Then, the processing operation is executed at discrete times (hereinafter referred to as “discrete times”) at regular sample periods.

  One rotary galvanometer mirror (not shown) is attached to the rotary actuator 1, and the angle of the galvanometer mirror is the control amount signal 11 of this galvanometer mirror control device. The rotary actuator 1 includes a rotary encoder (not shown), and the encoder outputs angle detection data 9 for each discrete time.

  Next, a procedure for positioning the galvanometer mirror by the galvanometer mirror control device will be described.

  When the angle command data 8 is received from the host control, an adder (subtracter) 20 subtracts the angle detection data 9 from the received angle command data 8 and outputs the result to the integral compensator 2 as a tracking error. The integral compensator 2 calculates the time integration of the tracking error and outputs the calculation result to the adder 21. The adder 21 adds the outputs of the digital filters 10 a and 10 b added by the adder 22 and the output of the integral compensator 2 and outputs the result to the adder 23. Digital filters 10a and 10b are important elements of the present invention and will be described in detail later.

  The proportional compensator 3 multiplies the angle detection data 9 by a proportional gain and outputs the result to the adder 24. The velocity observer 4 calculates an estimated value of the angular velocity of the mirror from the output of the torsional vibration stabilization compensator 5 and the angle detection data 9 and outputs the result to the adder 24. The adder 24 outputs the sum of the output of the proportional compensator 3 and the speed observer 4 to the adder 23. The adder 23 subtracts the output of the adder 24 from the output of the adder 21 and outputs the result to the torsional vibration stabilization compensator 5. The torsional vibration stabilization compensator 5 stabilizes the feedback loop of the galvanomirror control device with respect to one or more torsional vibration modes of the rotary actuator 1. The DA converter 6 converts the output of the torsional vibration stabilization compensator 5 into an analog value. This analog value is a current command value, and the current control circuit 7 controls the drive current so as to follow this current command value and supplies it to the rotary actuator 1.

  Next, additional input for performing initial value compensation will be described.

  In this embodiment, the additional input is an impulse train signal generated by the digital filters 10a and 10b at discrete times. Hereinafter, for the sake of formulation, the region S surrounded by the broken line in FIG. 1 is referred to as an enlargement control target.

  The enlargement control target S is a 1-input 1-output system, and the sum of the output of the adder 21, that is, the output of the integral compensator 2 and the outputs of the digital filters 10a and 10b, is an input signal to the enlargement control target S. Detection data 9 is an output signal of the enlargement control target S.

The discrete time state equation of the expansion control object S is expressed by Equations 1 and 2.

  In Equations 1 and 2, i is an index representing the passage of discrete time, xp is a state vector of an m-th order expansion control target, u ′ is an input signal of the expansion control target S, and y is an output signal of the expansion control target S ( Control amount signals 11), Ap, Bp, and Cp are real matrixes of m × m, m × 1, and 1 × m, respectively.

Further, the discrete time state equation of the integral compensator 2 is expressed by equations 3 and 4.

In Expressions 3 and 4, xc is a state vector of the n-th order integral compensator, r is angle command data, Ac, Bc, Cc, and Dc are real numbers of n × n, n × 1, 1 × n, and 1 × 1, respectively. It is a matrix.
From Equation 1 to Equation 4, the transfer function of the feedback loop expressed by Equation 5 can be derived.

  In Expression 5, z is a complex variable of z conversion, R is z conversion of the angle command data r, Y is z conversion of the output signal y of the enlargement control object S, and D (z) and Nr (z) are high values related to z, respectively. A first-order polynomial, Np (z) is a 1 × m-order high-order polynomial matrix for z, and Nc (z) is a 1 × n-order high-order polynomial matrix for z. Further, xp0 represents the initial state quantity xp (0) of the state variable vector xp (i) in Expressions 1 and 2, and xc0 represents the initial state quantity xc (0) of the state variable vector xc (i) in Expressions 3 and 4.

  Here, the initial time i = 0 considering the above-mentioned initial state quantity is defined as the time at which a single positioning operation is started, that is, the time at which one angle command data is received by the galvanomirror control device.

  What Equation 5 means is that the feedback loop response Y (z) is determined by a linear superposition of the response to the angle command data R (z), the response to the initial state quantity xp0, and the response to the initial state quantity xc0. That's what it means.

Since both the second term and the third term on the right side of Equation 5 are responses to the initial state quantity, these will be collectively expressed as Y0 (z). Y0 (z) can be expanded as shown in Equation 6.

  In Expression 6, Np1 (z) to Npm (z) are high-order polynomials related to z, and are elements of the high-order polynomial matrix Np (z) in Expression 5. Similarly, Nc1 (z) to Ncn (z) are high-order polynomials related to z, and are elements of the high-order polynomial matrix Nc (z) in Equation 5. Further, xp01 to xp0m are elements of the initial state vector xp0 in Expression 5, and xc01 to xc0n are elements of the initial state vector xc0 in Expression 5.

  Now, assuming that the initial value response by the q-th initial state quantity xp0q in the first term on the right side of Equation 6 has an adverse effect on the response Y (z) of the feedback loop, the digital filters 10a and 10b in FIG. Let us consider that the adverse effect is suppressed by the additional input output by.

  Hereinafter, the transfer function of one digital filter is represented as nq (z) / dq (z). Here, q is a subscript, and q = 1, 2,..., M since it is used in the same meaning as the subscript q of the q-th initial state quantity xp0q that appears in the first term on the right side of Equation 6. Nq (z) and dq (z) are high-order polynomials related to z. Further, the z-transform of the additional input output from the digital filter is expressed as Uaq (z).

As shown in Expression 7, an impulse signal having a magnitude equal to the initial state quantity xp0q is input to the transfer function nq (z) / dq (z), and the response is input to the additional input Uaq (z). Think about what to do.

As shown in FIG. 1, the additional input Uaq (z) is added to the input signal to the enlargement control target S. At this time, the term of the response to the additional input Uaq (z) is linearly superimposed on the right side of Expression 6 and developed as Expression 8.

  In Expression 8, Nu (z) is a high-order polynomial with respect to z, and the transfer function Nu (z) / D (z) is a form in which the additional input Uaq (z) is added to the input signal to the expansion control target S. Represents acting on the feedback loop.

  Regarding the third term on the right side of Equation 8, the following can be said. In Expression 6, the transfer function for the initial state quantity xp0q is Npq (z) / D (z). However, when the additional input Uaq (z) as shown in Expression 7 is used, the transfer function for the initial state quantity xp0q is expressed by 8 It can be changed as the third term on the right side.

  Therefore, if the transfer function nq (z) / dq (z) of the digital filter expressed by Equation 7 is appropriately designed, the initial value response by the initial state quantity xp0q that has adversely affected the response Y (z) can be obtained. The desired response can be changed and the response Y (z) can be improved.

  Hereinafter, a specific design example of the digital filter nq (z) / dq (z) of Expression 7 will be described.

  The order m of the expansion control object S described in Equations 1 and 2 depends on the torsional vibration mode inherent to the rotary actuator 1, the speed observer 4, the torsional vibration stabilization compensator 5, and the magnitude of the phase delay of the feedback loop. Determined.

  Now, suppose that it is desired to improve the initial value response resulting from the primary mode of torsional vibration. Since each vibration mode has two state quantities, angular displacement and angular velocity, the initial state quantities of the angular displacement and angular velocity of the torsional vibration primary mode are formalized as q = 1, 2 in the first term on the right side of Equation 6. To do. In the following, the digital filter for compensating the initial state quantity (q = 1) of the angular displacement is denoted by 10a in FIG. 1, and the digital filter for compensating the initial state quantity (q = 2) of the angular velocity is denoted by 10b in FIG. To do.

  First, for q = 1, the digital filter n1 (z) / d1 (z) is designed as follows. The basic idea here is the arrangement of poles and zeros in the transfer function.

  FIG. 9 is a diagram showing an arrangement of poles and zeros relating to the transfer function Np1 (z) / D (z) related to the initial state quantity xp01 of the angular displacement of the first-order torsional vibration mode in the first term on the right side of Equation 6. .

  FIG. 9 is a complex plane, where the horizontal axis is the real axis representing the real part of the complex number, and the vertical axis is the imaginary axis representing the imaginary part. A circle indicated by a broken line in the figure is a unit circle having a radius of 1 centered on the coordinate origin, a symbol x represents a pole, and a symbol ◯ represents a zero point.

  Since poles and zeros are generally complex numbers, they are represented by two-dimensional coordinates on the complex plane. The arrangement of these poles and zeros determines the response characteristics of the dynamic characteristic mode of the feedback loop.

  There are m + n = 16 poles unique to the feedback loop, and these are arranged at symbol x in FIG. 9 when the characteristic equation D (z) = 0 is solved by numerical calculation. Because the feedback loop is stable, all poles are in the unit circle. On the other hand, the zero point is the root of the equation Np1 (z) = 0, and when this equation is solved by numerical calculation, it is arranged at symbol O in FIG.

  Next, consider the transfer function relating to the initial state quantity xp01 after adding the additional input Uaq (z), that is, the pole and zero of the transfer function in the third term on the right side of Equation 8.

  Since the pole of this transfer function is the root of the equation with the denominator polynomial = 0, it coincides with the combination of the root of the characteristic equation D (z) = 0 and the root of the equation d1 (z) = 0. This polynomial d1 (z) is a denominator polynomial of the transfer function n1 (z) / d1 (z) of the digital filter to be designed here. On the other hand, the zero of the transfer function in the third term on the right side of Equation 8 is the root of the equation where the numerator polynomial = 0, and the arrangement of the zeros can be determined arbitrarily. This is because the numerator polynomial includes a denominator polynomial d1 (z) and a numerator polynomial n1 (z) of the transfer function of the digital filter, which are design parameters.

  Therefore, the zeros are arranged so as to cancel the 16 poles shown in FIG. Further, in order to prevent the control input from including a higher frequency component than necessary, the pole of the digital filter that generates the additional input is arranged in a region below the Nyquist frequency.

  When the numerator polynomial n1 (z) and the denominator polynomial d1 (z) of the transfer function of the digital filter are determined so that the arrangement of the poles and zeros is as described above, the pole of the transfer function in the third term on the right side of Equation 8 The zero points are arranged as shown in FIG. Here, FIG. 10 is a complex plane similar to FIG.

  Since the 16 poles included in FIG. 9 are offset by the newly arranged 16 zeros, the influence of these poles on the initial value response is suppressed. Further, the newly added pole is a region surrounded by a broken-line square in FIG.

Moreover, in order to realize the transfer function of the digital filter designed based on the above idea by the program of the microprocessor, it is necessary to make it less susceptible to the word length restriction. As a configuration suitable for this, a cascade type shown in FIG. 11 is used in the present invention. This is a method in which a second-order / second-order transfer function is connected in series. Note that −b ₀₁ to −b _L2 and a _{01 to} a _L2 in FIG. 11 are coefficients, and z ⁻¹ is a shift operator. X is an impulse input, and Y is an additional input signal.

  The following concept is used for determining the coefficients of the second-order / second-order transfer function.

  The poles and zeros of the digital filter are arranged as shown in FIG. As in the case of FIG. 9, the symbol x in the figure is a pole, the symbol O is a zero point, and the dashed circle is a unit circle of radius 1 centered on the coordinate origin.

  Here, for poles and zeros close to the unit circle, the closest poles and zeros are set as a pair. This operation is called pairing. Since poles and zeros that are not real numbers are always conjugate complex numbers, a second-order / second-order transfer function having a certain pair and its conjugate complex number at the poles and zeros can be created. In FIG. 12, regions a, b, c, and d surrounded by solid lines are the results of pairing. A general digital filter implementation method based on such pairing is described in Signal Processing Tool User's Guide, MathWorks, Inc. (2000).

  As described above, by implementing a digital filter that generates an additional input by pairing with a cascade structure, it is possible to realize the effect of initial value compensation as expected even in a control program with word length limitation.

  As described above, the transfer function n1 (z) / d1 (z) of the digital filter 10a for compensating the initial state quantity xp01 of the angular displacement of the torsional vibration primary mode is designed.

  Similarly, the transfer function n2 (z) / d2 (z) of the digital filter 10b is designed for q = 2, that is, the initial state quantity xp02 of the angular velocity of the torsional vibration primary mode.

  Next, an initial value response compensation method using the digital filter will be described.

  First, initial state quantities xp01 and xp02 of the angular displacement and angular velocity of the torsional vibration primary mode are detected at the initial time. Usually, it is difficult to directly detect these state quantities with a sensor or the like. Therefore, if a state equation model of a control target including a plurality of torsional vibration modes is constructed, a state observer is designed based on the model, and the initial state quantity can be estimated and calculated. In particular, if a state observer taking account of unknown disturbances is used, the initial state quantity can be estimated with high accuracy. The state observer design method considering unknown disturbance is described in Iwai, Inoue, Kawaji: “Observer”, Chapter 6, Corona (1988). These initial state quantities are impulse-inputted to the respective digital filters 10a and 10b with q = 1, 2 as described in the explanation regarding Expression 7. An additional input that compensates for the initial value response of the torsional vibration primary mode is the addition of the impulse responses of the digital filters 10a and 10b. This is added to the output of the integral compensator 2 as shown in FIG. Further, in order to continuously perform positioning with respect to the angle command pattern one after another, the following is performed. That is, the internal state variables of the digital filters 10a and 10b are cleared to zero at each initial time. The reason is that when the command interval is very short, that is, when the next angle command data is received before the impulse responses of the digital filters 10a and 10b are steadily attenuated to 0, and a new initial time is reached. However, this is to generate a normal additional input.

  Next, operations of the galvanomirror control device and the laser processing device using the additional input type initial value compensation control of the present invention will be described. FIG. 2 is a schematic diagram showing a hole processing position of a printed wiring board being manufactured by a circle (◯), and it is assumed that the laser processing apparatus performs the hole processing of FIG.

  The laser processing apparatus of the present invention is equipped with a galvano mirror that handles each of the vertical axis direction and the horizontal axis direction, and the galvano mirror control apparatus shown in FIG. 1 is used for each mirror. Different angle command patterns are transmitted from the host control to the galvanometer mirror control devices on the vertical axis and the horizontal axis. Since the mirror in the vertical axis direction reciprocates at an angular stroke corresponding to the hole interval L, an angle command pattern as shown in FIG. 3 is obtained. Since the angular stroke is constant, the command interval is also constant, and this value is hereinafter expressed as τ (where τ is a dimensionless time). On the other hand, the mirror in the horizontal axis direction performs a feed operation in a fixed direction with an angular stroke corresponding to the hole interval L, so that an angle command pattern as shown in FIG. 4 is obtained. The command interval is τ. In this explanation, for the sake of simplicity, the number of holes is 9 in FIG. 2, and the angle command patterns in FIGS. 3 and 4 are step-like patterns based on eight angle command data. In the machining apparatus, the number of holes to be machined is enormous, and the angle command pattern for each axis also becomes a step-like pattern continuous in a longer time.

  Next, the response waveform of the galvanometer mirror control device will be described.

  FIGS. 5-8 is a figure which shows the response waveform of a galvanometer mirror at the time of continuing reciprocation like FIG. The waveforms shown in these figures are tracking errors, that is, time waveforms having values obtained by subtracting the angle detection data 9 from the angle command data 8 in FIG. 5A to 8A sequentially show all the tracking error signals of the four reciprocating operations performed from the time when the galvano mirror starts the reciprocating operation (time 0) to the time 36. The lower part (b) is the result of overlaying the follow-up and return errors of the four reciprocating motions (that is, when the discontinuous rising edge of the follow-up error signal at the initial time of the forward motion is used as a trigger) Of the waveform drawn by the oscilloscope) in the vicinity of the zero tracking error. In addition, a region between two broken lines in the figure is a range (hereinafter referred to as “settling range”) set to show the difference in settling waveform between the present invention and the prior art.

  FIG. 5 shows the waveform of the tracking error according to the present invention when the command interval is set to τ = 4.5. FIG. 6 shows a tracking error when the same reciprocating operation is performed in a conventional galvanometer mirror control device that does not use the initial value compensation control, for comparison with FIG.

  In the case of the conventional technique (FIG. 6), the settling range is temporarily entered by one positioning, but the subsequent settling response is oscillatory and falls outside the settling range. Further, the response in the vicinity of the zero tracking error varies with time.

  On the other hand, in the case of the present invention (FIG. 5), a good settling response with little variation is realized within the settling range.

  FIG. 7 shows the waveform of the tracking error according to the present invention when the command interval is set to τ = 3.5. For comparison with FIG. 7, FIG. 8 shows a tracking error when the same reciprocating operation is performed in a conventional galvanometer mirror control apparatus that does not use the initial value compensation control.

  In the present invention, even under these conditions, a good settling response that is suppressed within the settling range and hardly fluctuates is realized.

  Although the galvanometer mirror control device has been described above, the present invention can also be applied to servo control of a table that holds and moves a workpiece in a printed circuit board processing device or the like.

It is a block diagram of the galvanometer mirror control device concerning the present invention. It is a schematic diagram which shows the hole processing position of a printed wiring board. It is a figure which shows the angle command pattern with respect to a galvanometer mirror. It is a figure which shows the angle command pattern with respect to a galvanometer mirror. It is a figure which shows the response waveform of the tracking error in this invention. It is a figure which shows the response waveform of the conventional tracking error. It is a figure which shows the response waveform of the tracking error in this invention. It is a figure which shows the response waveform of the conventional tracking error. It is a figure which shows arrangement | positioning of the pole and zero of a transfer function before adding an additional input. It is a figure which shows arrangement | positioning of the pole and zero of a transfer function in this invention. It is a block diagram of the digital filter in this invention. It is a figure which shows arrangement | positioning of the pole and zero of a digital filter in this invention.

Explanation of symbols

2 Integral compensator 8 Angle command data 9 Angle detection data 10a, 10b Digital filter

Claims (6)

In a digital servo controller for a moving body that includes a feedback loop and positions the moving body based on position command data,
An additional input means is provided for adding a zero that cancels out a pole unique to the feedback loop, with respect to a transfer function related to the state quantity of the feedback loop at the time when the position command data is received,
A digital servo controller for a moving body, wherein the moving body is positioned by inputting an output of the additional input means to the feedback loop in addition to the position command data.   2. The digital servo control apparatus for a moving body according to claim 1, wherein the pole of the additional input means is a pole not higher than the Nyquist frequency of the digital servo control.   The movement according to claim 2, wherein the additional input means is a digital filter having an impulse signal equal to the state quantity at the time of receiving the position command data as an input signal, and has a cascade structure. Servo control device of the body.   4. The digital servo controller for a moving body according to claim 3, wherein the transfer function constituting the cascade structure of the digital filter has a pole and a zero that are close to each other on a complex plane.   2. The digital servo controller for a moving body according to claim 1, wherein the moving body is a galvanometer mirror. A digital servo control device according to claim 5;
Control means for controlling the angle of the galvanometer mirror;
With
By controlling the angle of the laser beam reflected by the galvanometer mirror by the control means, the laser beam is incident on a predetermined position of the workpiece to process a hole in the workpiece. Laser processing equipment.

Images