JP2558968B2 - Actuator access method and access control device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an access method and an access control device for an actuator used in a disk storage device or the like, and in particular, access for moving the actuator from a certain stop position to an arbitrary selected position at high speed. A method and an access control device.

[0002]

2. Description of the Related Art In recent years, recording capacities of information recording / reproducing devices such as magnetic disk devices and optical disk devices have been remarkably improved, and accordingly, high speed access to a target information track of a recording or reproducing data transducer has been demanded. As one of the access control methods of the positioning actuator of the information recording / reproducing apparatus, Bang
-Bang) drive method. This method is one of the shortest time control methods of supplying the maximum acceleration / deceleration command to the actuator by the open loop control.

FIG. 23 is a block diagram of an actuator drive circuit in a conventional access control device. In FIG. 23, 101 is an operational amplifier, 102 is a power transistor, 103 is an electromagnetic coil of an actuator, 104 is an inductance component of the electromagnetic coil 103, 105 is a pure resistance component of the electromagnetic coil 103, and 106 is a current value flowing in the electromagnetic coil 103. A current detection resistor 107 is a counter electromotive voltage generated in the electromagnetic coil 103 as the actuator moves. Rectangular wave-shaped bang-bang command signal 100
Is applied to the non-inverting input of operational amplifier 101. The output of the operational amplifier 101 is amplified by the power transistor 102 and applied to one end of the electromagnetic coil 103.

This actuator drive circuit feeds back one end of the electromagnetic coil 103 to the inverting input of the operational amplifier 101, so that the current flowing through the electromagnetic coil 103 is completely proportional to the non-inverting input voltage of the operational amplifier 101. , Constant current operation. However, this is the case where the drive circuit is operating in the linear region, and as shown in the figure, the bang-bang command signal 100 having a sufficiently large amplitude.
When is added to the non-inverting input of the operational amplifier 101, the operational amplifier 101 is saturated and the drive circuit is in an open loop state to directly drive the power transistor 102. As a result, the power transistors 102A and 102B
Is a saturation operation alternately according to the positive / negative of the bang-bang command signal,
That is, it is turned on, and power supply voltages + Ve and -Ve are applied to both ends of the electromagnetic coil of the actuator. Therefore, in this case, the constant voltage operation is performed. However, in the following explanation,
For simplicity, the influence of the inductance component 104 of the electromagnetic coil 103 and the influence of the counter electromotive force 107 generated in the electromagnetic coil 103 due to the movement of the actuator are extremely small.

FIG. 24 shows a side view of a DC drive type actuator generally called a voice coil motor. In FIG. 24, 111 is an electromagnetic coil and 112 is an electromagnetic coil 1.
Magnets for supplying a bias magnetic field to 11, 113, 114
Are a center yoke and an outer yoke for receiving a magnetic flux generated by the magnet 112 to form a magnetic circuit, and are made of a material having a high relative magnetic permeability such as iron.
The operating principle of the actuator shown in the same figure is based on Fleming's left-hand rule, in which the magnetic flux generated by the magnet 112 is linked to the current flowing through the electromagnetic coil 111 so that the electromagnetic coil 111 moves in the direction indicated by the arrow. It is about moving.

This voice coil motor has a yoke portion 11
The generated force may vary depending on the position and the moving direction of the electromagnetic coil 111 due to machining tolerances of the magnets 3 and 114 and the magnet 112, uneven magnetization of the magnet 112, and the like.

FIG. 25 is a distribution diagram showing an example of variations in the generated force of the actuator depending on the position, that is, force unevenness. In FIG. 25, the horizontal axis represents the movable part of the actuator,
That is, the position of the electromagnetic coil 111 with respect to a certain reference point is shown, and the vertical axis shows the generated force per unit drive current at that position, that is, the force constant of the actuator. Electromagnetic coil 1
The position of 11 is 5 mm = Xu1 <X <Xu2 = 20 mm
It can be seen that the generated force is substantially uniform in the range of 1), but in the other range, the generated force becomes gradually weaker as the electromagnetic coil 111 moves to the end of the magnetic circuit, and there is uneven force.

FIG. 26 (a) is a waveform diagram of the current I flowing through the electromagnetic coil 103, FIG. 26 (b) is a waveform diagram of the moving speed V of the actuator, and FIG. 26 (c) is a waveform diagram of the moving distance X of the actuator. And the horizontal axis represents time. In the figure, when there is no influence of the force unevenness of the actuator, for example, as shown in FIG. 25, the case of accessing the range of 5 mm <X <10 mm is indicated by a broken line.
The case of accessing the range of 0 mm <X <5 mm is shown by a solid line.

[0009]

However, when there is no influence of the force unevenness of the actuator, the moving speed V and the moving distance X of the actuator can be estimated by simply integrating the current. The effect of unevenness cannot be easily estimated. That is, when the actuator is accessed by giving a bang-bang command, it is necessary to estimate in advance the acceleration time and deceleration time corresponding to the target movement distance at that time. Further, in order to perform smooth positioning toward the target, it is desirable that the moving speed when approaching the target position becomes substantially zero. However, as described above, in the case where the influence of the force unevenness of the actuator is exerted, it is not practical to accurately estimate the acceleration time and the deceleration time such that the moving speed becomes zero at the target position in a short time. This is quite difficult, which makes it difficult to realize an access control device using a bang-bang drive system.

As described above, the access control device using the conventional method has a problem that it is difficult to estimate an appropriate access command due to the influence of the uneven force of the actuator.

An object of the present invention is to solve the above-mentioned problems, and to access and estimate an access command in consideration of the influence of force unevenness of an actuator in an extremely short time, easily and accurately. It is to provide a control device.

[0012]

To solve the above-mentioned problems, the present invention uses the following method or has the following structure.

An actuator access method for accessing the actuator by an access command signal including data representing acceleration, acceleration time, deceleration and deceleration time of the actuator in order to move the actuator from a start position by a predetermined moving distance. The start position of the actuator and the acceleration, acceleration time, deceleration and deceleration time represented by the access command signal are used as input variables, and the acceleration, acceleration time, deceleration and deceleration time represented by the data of the access command signal. Of performing a fuzzy inference operation having a plurality of rules with at least one correction value as an output variable, and correcting the data of the access command signal based on the correction value obtained by the fuzzy inference operation. Have or actuate An actuator access method for accessing the actuator by an access command signal including data representing acceleration, acceleration time, deceleration and deceleration time of the actuator in order to move the actuator from the start position by a predetermined moving distance, The start position of the actuator and the movement distance to be accessed by the actuator are used as input variables, and at least one correction value of acceleration, acceleration time, deceleration and deceleration time represented by the data of the access command signal is used as an output variable. The method includes a step of performing a fuzzy inference operation having a plurality of rules, and a step of correcting the data of the access command signal based on a correction value obtained by the fuzzy inference operation.

Alternatively, an actuator access control device having the following: drive means for driving the actuator, position detection means for detecting the position of the actuator, acceleration, acceleration time, deceleration and deceleration time of the actuator Receiving an access command signal including data representing the input position, the actuator start position and the acceleration, acceleration time, deceleration and deceleration time represented by the access command signal as input variables, and the acceleration and acceleration represented by the data of the access command signal. Fuzzy inference operation means for performing fuzzy inference operation having a plurality of rules using at least one correction value of time, deceleration and deceleration as an output variable,
Acceleration, acceleration time, deceleration and deceleration of the actuator in order to correct the data of the access command signal based on the correction value obtained by the fuzzy inference calculation means and move the actuator by a predetermined moving distance. Drive signal generating means for generating a drive signal having data representing time, or driving means for driving an actuator, position detecting means for detecting a position of the actuator, acceleration of the actuator, acceleration time, An access command signal including data representing deceleration and deceleration time is received, the start position of the actuator and the moving distance to be accessed by the actuator are input variables, and the acceleration and acceleration time represented by the data of the access command signal, Outputs at least one correction value for deceleration and deceleration time A fuzzy inference calculation means for performing a fuzzy inference calculation having a plurality of rules, and the data of the access command signal is corrected based on the correction value obtained by the fuzzy inference calculation means to move the actuator in a predetermined manner. A driving signal generating means for generating a driving signal having data representing acceleration, acceleration time, deceleration and deceleration time of the actuator for moving by a distance.

Alternatively, an actuator access method comprising the following steps: an access including data representing acceleration, acceleration time, deceleration time and deceleration time of the actuator in order to move the actuator from a start position thereof by a predetermined moving distance. A method of accessing an actuator according to a command signal, comprising: an acceleration, an acceleration time, a deceleration time, and a deceleration time of an access command signal suitable for a predetermined moving distance of the actuator, and parameters for each start position of the actuator. As a result, there is a step of obtaining by trial and error over the entire movable range of the actuator.

Alternatively, an access control device for an actuator having the following: drive means for driving the actuator, position detecting means for detecting the position of the actuator, and the actual operation of the actuator based on the output of the position detecting means. An actual movement distance detecting means for detecting a movement distance, and a predetermined start position of the actuator and an acceleration, acceleration time, deceleration and deceleration time of an access command signal suitable for the movement distance, with each start position of the actuator as a parameter. An optimum access command signal creating means that is obtained by trial and error over the entire movable range of the actuator, and an acceleration and acceleration of an access command signal suitable for a predetermined start position and moving distance of the actuator found by the optimum access command signal creating means. Time, deceleration and deceleration Based on time, the acceleration of the predetermined start position and access command signal with respect to a predetermined moving distance of the actuator, acceleration time, and a interpolation calculation means for calculating the interpolation calculation the deceleration and decelerating time.

[0017]

The present invention has the following actions due to the above method or configuration. That is, the actuator is accessed on a trial basis in advance for a predetermined moving distance, and the data of the most suitable access command signal, acceleration, acceleration time, deceleration, and deceleration time based on the trial access is used, and the movable range of the actuator is set using the start position as a parameter. Estimating over the whole, using the result, the start position of the actuator, and either the data of the access command signal or the predetermined movement distance as an input variable,
By estimating the data of the access instruction signal based on a fuzzy inference operation using the correction value of the access instruction signal as an output variable, estimation of the access instruction in consideration of the influence of the force unevenness of the actuator, which has been difficult in the past. Can be performed quickly, easily and accurately.

[0018]

1 is a block diagram of an embodiment of an access control device for an actuator according to the present invention. In the figure, a recording / reproducing transducer of a recording / reproducing apparatus such as an optical disk is attached to an actuator 7 which is moved by an electromagnetic driving means 7A. The actuator 7 is moved according to the access command signal P at the time of recording or reproduction, and positions the transducer on a predetermined track of the optical disc.

The position of the actuator 7 is detected by the position encoder 8 and the position detection circuit 9, and the position signal X is output. The moving speed of the actuator 7 is detected by the speed detecting circuit 10 based on the position signal X, and the speed signal V is output. The control circuit 1 is a circuit that controls the movement of the actuator 7 according to an access command signal P from an external device (for example, a computer), and includes fuzzy inference calculation means 2, a microcomputer 3, a memory 4, and an interface, which will be described in detail later. It has a circuit 5 and a drive signal generation circuit 20. The drive signal generation circuit 20 includes an adder circuit for correcting the data of the access command signal P based on the correction value obtained by the fuzzy inference calculation means 2 and an analog switch, and the acceleration and acceleration time of the actuator 7 The drive signal U including the data of deceleration and deceleration time is applied to the drive circuit 6, and the drive current I is supplied to the actuator 7 accordingly. The drive signal U is a bipolar rectangular wave and has a waveform similar to the Bang-Bang signal of the prior art. The actuator 7 is accelerated in the first half cycle of this drive signal and decelerated in the latter half cycle. The acceleration and deceleration are determined according to the amplitude of the drive signal U.

The interface circuit 5 has an A / D converter (not shown) and A / D-converts the position signal X and the speed signal V and gives them to the microcomputer.

The access command signal P has, for example, data on the start position, target position, and moving direction of the actuator 7, and is input to the microcomputer 3. The memory 4 is used as a temporary memory for data.

FIG. 2 is a flow chart showing an embodiment of an actuator access method of the present invention. The steps shown in this flow chart are performed immediately before access during normal use of this device.

The moving range of the actuator 7 is determined by the recording area of the recording medium, and the moving range is defined as the maximum moving distance.
Expressed as Xmax. Letting the target travel distance Xd be the distance when the actuator moves from a predetermined start point (for example, one end of the travel range) to the target position specified by the access command, the range of the target travel distance Xd is expressed by the formula (1) ).

0 ≦ Xd ≦ Xmax (1) The force unevenness of the actuator has a shape as shown in the distribution chart of FIG. 25, and the range and magnitude of the force unevenness represented by the equation (2). It is assumed that this is obtained in advance.

Xd ≦ Xu1 or Xd ≧ Xu2 (2) In FIG. 2, first, the target movement distance Xd to which the actuator should move is set (step 1 in the flowchart of FIG. 2). When the actuator 7 is moved by the target moving distance Xd, the acceleration time T for accelerating the actuator 7 is obtained by the known formula (3).

T = (2 · Xd / α) ^1/2 (3) where α is the acceleration of the actuator. This acceleration time T is equal to the target travel distance X after repeated trial access.
The acceleration time most suitable for moving d may be obtained, stored in the data memory in advance, and referred to as necessary (step 2). However, the target travel distance Xd and acceleration time T
When obtaining the relationship between and, the range in which there is no force unevenness of the actuator represented by the equation (4) is selected.

Xu1 <Xd <Xu2 (4) The acceleration time T obtained in the above step 2 is for the case where the actuator moves within a range where the force of the actuator is uniform.

Next, the start position Xs is set (step 3), and it is determined whether or not this is included in the range of force unevenness represented by the equation (4) (step 4).

If it is not included in the range of force unevenness, the actuator is moved to the start position Xs (step 7), and access is performed for the acceleration time T (step 8). When the acceleration time T ends, the actuator 7 is decelerated in the latter half of the drive signal U and stops at a certain position.

When it is included in the range of the force unevenness, the acceleration time T is not properly selected, so that the actuator 7 does not stop at a predetermined target position, that is, the moving speed does not become zero.
Therefore, it is necessary to select the optimum acceleration time for moving the actuator 7 to the target position. In that case, the correction amount ΔT of T is obtained based on a fuzzy inference operation described in detail later (step 5). Then, the calculation of equation (5) is executed (step 6), and access is made based on this acceleration time T.

T = T + ΔT (5) By this operation, the influence of the force unevenness of the actuator 7 on the acceleration and deceleration of the actuator 7 is added to the above acceleration time T.

Next, the fuzzy inference operation of step 5 in the flowchart of FIG. 2 will be described in detail. The basic inference rules are as follows: Rule 1: If Xs = I and Xd = S, then ΔT = S Rule 2: If Xs = I and Xd = M, then ΔT = M Rule 3: If Xs = I and Xd = B, then ΔT = B Rule 4: If Xs = C and Xd = S, then ΔT = ZR Rule 5: If Xs = O and Xd = M, then ΔT = M Rule 6: If Xs = O and Xd = B, then ΔT = B where Xs: Start position Xd: Target movement distance ΔT: Correction time I: Inner Area C: Center Area O: Outer Area S: Small M: Medium B: Big ZR: Zero However, actuator 7 In the movable range of the movable part, the side determined as the origin X = 0 is the inner area (Inner Area), the opposite side is the outer area (Outer Area), and the middle thereof is the intermediate area (C
enter Area).

In the above inference rule, "If Xs = I an
“D Xd = S” is the antecedent part, and “then ΔT = S” is the consequent part. Also, "I", "C", "O", "S", "M", "B", and "ZR" are fuzzy variables.

The application of this fuzzy inference rule will be described with respect to rule 1, for example. When the start position Xs is in the inner area and the target movement distance Xd is a small value, rule 1
Therefore, the correction time ΔT is a small value. Similarly,
The above fuzzy inference operations apply to rules 2-6.

3 to 5 are diagrams of membership functions for the above fuzzy variables. In FIG. 3, the horizontal axis represents the start position Xs and the vertical axis represents the degree of conformity. The goodness of fit takes a value from 0 to 1. In FIG. 4, the horizontal axis indicates the target movement distance Xd. The vertical axis is the degree of conformity. In FIG. 5, the horizontal axis indicates the correction time ΔT in the range of −0.2 msec to +0.8 msec. Similarly, the vertical axis is the goodness of fit. In this example, the maximum travel distance Xmax is 25 mm and the membership function is represented by a triangle.

6 to 11 are diagrams showing the process of the fuzzy inference operation in this embodiment. In these figures,
Start position Xs is 2.5 mm and target movement distance Xd is 5 mm
Is. This fuzzy inference operation is Mamdani's method or MIN
-Known as the MAX synthesis method.

In FIG. 6, the left diagram and the center diagram represent the antecedent part of rule 1, and the diagram on the right side represents the antecedent part of the same rule. Since the starting position Xs is 2.5 mm, its goodness of fit is 0.5 for the fuzzy variable I.
Also, since the moving distance Xd is 5 mm, the goodness of fit is 0.5 for the fuzzy variable S. The smaller of these values may be adopted for the "and" operations of both of them, but in this case, since they have the same value, the matching degree in the antecedent part is 0.5. As a result, in the consequent part, the figure of the membership function is cut by a line with a goodness of fit of 0.5. Therefore, the fuzzy variable S of the correction time ΔT is represented by the trapezoid Z1.

In FIG. 7, the left diagram and the center diagram represent the antecedent part of rule 2, and the diagram on the right side represents the antecedent part of the same rule. Since the goodness of fit for the fuzzy variable I is 0.5 and the goodness of fit for the fuzzy variable M is 0.5, the goodness of fit in the antecedent is 0.5. As a result, the figure of the membership function in the consequent part is cut by a line with a goodness of fit of 0.5, and the fuzzy variable M of the correction time ΔT is represented by the trapezoid Z2.

In FIG. 8, the left diagram and the center diagram represent the antecedent part of rule 3, and the diagram on the right side represents the consequent part of the same rule. The goodness of fit for the fuzzy variable I is 0.5, but the goodness of fit for the fuzzy variable B is zero, so the goodness of fit in the antecedent is zero. As a result, the goodness of fit of the consequent part to the fuzzy variable B is zero. Therefore, Rule 3 cannot be applied in this case.

In FIG. 9, the left figure and the central figure represent the antecedent part of rule 4, and the right figure represents the consequent part of the same rule. Since the goodness of fit for the fuzzy variable C is 0.5 and the goodness of fit for the fuzzy variable S is 0.5, the goodness of fit in the antecedent part is 0.5. As a result, the figure of the membership function in the consequent part is cut by a line with a goodness of fit of 0.5, and the fuzzy variable ZR of the correction time ΔT is represented by the trapezoid Z3.

In FIG. 10, the left diagram and the center diagram represent the antecedent part of rule 5, and the diagram on the right side represents the consequent part of the same rule. Since the goodness of fit for the fuzzy variable O is zero and the goodness of fit for the fuzzy variable M is 0.5, the goodness of fit in the antecedent is zero. As a result, the goodness of fit of the consequent part to the fuzzy variable M is zero.
Therefore, rule 5 cannot be applied in this case. Similarly, rule 6 cannot be applied in this case.

In FIG. 11, the hatched figure represents the membership function of the consequent part, and the result of the fuzzy operation for all rules 1 to 6 is MI.
It is represented by the N-MAX synthesis method. In order to defuzzify this result, we find the centroid of this hatched part. In the example shown in FIG. 11, a correction time ΔT of 0.2 msec is obtained from the position of this center of gravity.

This correction time ΔT (0.2 msec in this case)
Is added to the acceleration time T (T = T + ΔT), and a new acceleration time T is calculated.

In this embodiment, the acceleration time T may be used as the variable of the antecedent part instead of the target moving distance Xd. An example of such a rule is shown below.

If Xs = I and T = S then ΔT = S FIG. 12 is a diagram of the membership function of the fuzzy variable of the acceleration time T. In Figure 12, the horizontal axis is from 0 msec to 5 msec.
The acceleration time T in the range of is graduated. The vertical axis is the degree of conformity. The process of the fuzzy inference operation of this rule is the same as that described with reference to FIG.

Further, in this embodiment, since the drive signal U is a rectangular wave having positive and negative polarities, the amplitude of the drive signal U given to the actuator 7 in the acceleration process is the same as the amplitude in the deceleration process. However, these amplitudes can be chosen arbitrarily. In such a case, the acceleration and the deceleration can be changed by changing the amplitude in the acceleration process or the deceleration process. Then, the corrected amplitude ΔD for correcting the amplitude can be used as a variable of the consequent part of the fuzzy inference operation. As an example of such a rule, the one using the correction amplitude ΔD in the consequent part is shown below.

If Xs = I and Xd = S then ΔD = S FIG. 13 is a diagram of the membership function of the fuzzy variable of the correction amplitude ΔD. In Figure 13, the horizontal axis is -400 mV to +4
The correction amplitude ΔD in the range of 00 mV is graduated. The vertical axis is the degree of conformity. The process of the fuzzy inference operation of this rule is the same as that described with reference to FIG.

Further, in this embodiment, as a variable of the consequent part, a correction time for correcting the deceleration time may be used instead of the correction time ΔT of the acceleration time T. Furthermore, in the moving operation of the actuator 7, a pause process may be interposed between the acceleration process and the deceleration process.

The contents of the above rules and FIG. 3 and FIG.
The shape of the membership function shown in FIG. 12 and FIG. 13 differs depending on the range and magnitude of force unevenness obtained in advance, that is, the conditions such as the characteristics of the actuator, and is not necessarily limited to the above-mentioned configuration. . In this embodiment, the case where the force unevenness of the actuator occurs at the end of the movable range is not limited to this, and even when the force unevenness occurs locally near the center of the movable range of the actuator, It can be corrected by adjusting the content of the above rules and the shape of the membership function.

FIG. 14 shows the MI used in this embodiment.
1 shows the configuration of a fuzzy inference operation circuit 2 that performs an operation based on the N-MAX combining method. In FIG. 14, the membership function memory 11 stores a predetermined membership function. The MIN operation circuits 12-1, 12-2, ..., 12-n are configured to include a memory for storing fuzzy inference rules, the number of which is equal to the number of rules. The MIN operation circuit 12-i calculates the antecedent and consequent parts of each rule. The output of the MIN arithmetic circuit 12-i is input to the MAX arithmetic circuit 13. The MAX arithmetic circuit 13 obtains a common set of them based on the inputs from all the MIN arithmetic circuits 12-i, and inputs the result to the centroid arithmetic circuit 14. The center of gravity calculation circuit 14 finds the center of gravity of the common set.

The specific structure of the fuzzy inference calculation means 2 is not limited to the above structure.

In the above-described actuator access method and access control device, it has been necessary to obtain the range and magnitude of force unevenness of the actuator 7 in advance. It can be obtained as follows.

FIG. 15 shows a block diagram of an embodiment of the access control device for the actuator of the present invention. FIG. 15 shows FIG.
The actuator has the same configuration as the access control device of the actuator shown in FIG. 3, and the recording-reproducing transducer of the recording / reproducing device such as an optical disk is attached to the actuator 7 moved by the electromagnetic driving means 7A. The actuator 7 is moved according to the access command signal P at the time of recording or reproduction, and positions the transducer on a predetermined track of the optical disc.

The position of the actuator 7 is detected by the position encoder 8 and the position detection circuit 9, and the position signal X is output. The moving speed of the actuator 7 is detected by the speed detecting circuit 10 based on the position signal X, and the speed signal V is output. The control circuit 1 is a circuit that controls the movement of the actuator 7 according to an access command signal P from an external device (for example, a computer), and an interpolation inference calculation means 21, a microcomputer 3, a memory 4, an interface, which will be described in detail later. 5 and a drive signal generation circuit 20. The drive signal generation circuit 20 is, for example, an analog switch, and supplies a drive signal U including data of acceleration, acceleration time, deceleration, and deceleration time of the actuator 7 to the drive circuit 6, and the drive current I is supplied to the actuator 7 in response to the drive signal U. To be done. The drive signal U is a bipolar rectangular wave and has a waveform similar to the Bang-Bang signal of the prior art.
The actuator 7 is accelerated in the first half cycle of this drive signal and decelerated in the latter half cycle. The acceleration is determined according to the amplitude of the drive signal U.

The interface circuit 5 has an A / D converter (not shown), A / D-converts the position signal X and the speed signal V, and gives them to the microcomputer.

The access command signal P has, for example, data on the start position, target position, and moving direction of the actuator 7, and is input to the microcomputer 3. The memory 4 is used as a temporary memory for data.

FIG. 16 is a flow chart showing an embodiment of the actuator access method of the present invention. The steps shown in this flow chart are performed at the time of factory shipment of this apparatus or immediately after power-on in normal use.

In FIG. 16, first, the target moving distance Xdj to be moved by the actuator 7 is set (step 1 in the flowchart of FIG. 16).

The target moving distance Xdj is set to the actuator 7
The acceleration time Ti for accelerating the actuator 7 when moving is calculated by the known formula (6) (step 2). Ti = (2 · Xd / α) ^1/2 (6) where α is the acceleration of the actuator. The acceleration time obtained by the equation (6) is a theoretical value, and the influence of the force unevenness of the actuator is not considered. Acceleration time Ti
Is completed, the actuator 7 is decelerated in the latter half of the drive signal U and stops at a certain position. At this time, if the acceleration time Ti is not properly selected, the actuator 7 does not stop at the predetermined target position, that is, the moving speed does not become zero. Therefore, it is necessary to select the optimum acceleration time for moving the actuator 7 to the target position. In order to obtain the optimum acceleration time, the actuators are sequentially operated based on the acceleration time Ti obtained by the equation (6). First, the start position Xsi is set (step 3), and the actuator 7 is moved to the start position Xsi (step 4). Next, in order to move the actuator 7 to the position of the target moving distance Xd, the acceleration time Ti is accelerated, and then the speed is reduced until the moving speed becomes zero (step 5). As a result, the actuator stops. Then, the shift distance Xdr, which is the distance between the position where the moving speed of the actuator becomes zero and the start position, is measured. Next, the distance deviation ΔXd, which is the difference between the target movement distance Xdj and the shift distance Xdr, is calculated by the equation (7) (step 6).

ΔXd = Xdj−Xdr (7) When deviation ΔXd is sufficiently small (step 7), acceleration time Ti
And the start position Xsi is stored in the memory 4 (step 9).

On the other hand, when the deviation ΔXd is not small, the correction time ΔT of the acceleration time Ti is given as a certain value, and the correction time ΔT
And the new acceleration time Ti2 which is the sum of the acceleration time Ti and the formula (8)
(Step 8).

Ti2 = Ti + ΔT (8) Then, the actuator 7 is returned to the start position Xsi and moved again for the acceleration time Ti2. Steps 4, 5, 6, 7, above
And 8 are repeated until the deviation ΔXd becomes substantially zero, and data (Xsi, Ti) including the start position Xsi and the acceleration time Ti when the deviation ΔXd becomes zero are obtained. Required data (X
si, Ti) are stored in the memory 4 (step 9).

Further, a new start position Xs (i + 1) obtained by adding the position correction amount ΔXs to the start position Xsi is given by the equation (9).
(Step 3).

Xs (i + 1) = Xsi + ΔXs (9) The process of steps 3 ... 9 is performed at all start positions Xs0, Xs1, Xs2, ... Xsi, ..., Xsm. For the memory 4 to obtain m + 1 pieces of data (Xsi, Ti) (i = 0,1,2, ..., m).
To memory.

The start position Xsi is the maximum start position Xsmax
When the above is reached, the distance correction amount ΔX at the target movement distance Xdj
A new target movement distance Xd (j + 1) to which dd is added is calculated by the equation (10) (step 1).

Xd (j + 1) = Xdj + ΔXdd (10) All the target movement distances Xd0, Xd1, Xd2, ... Xdj ,. Xdk is executed (step 11), and m + 1 data (Xsi, Ti) (i = 0, 1, 2, ..., M) is obtained for each of them and stored in the memory 4 in the same manner as described above.

The above steps 1 to 11 constitute an optimum access command signal generating means.

If there is a variation in the generated force of the actuator depending on the moving direction, the above procedure is similarly performed for both moving directions.

FIG. 17 is a control diagram showing the relationship between the start position Xsi and the acceleration time Ti obtained by the above procedure. Figure 1
In Fig. 7, the straight line shown by the solid line is the acceleration time T for each start position Xsi when there is no actuator force variation.
Represents a straight line connecting the values of i. On the other hand, the relationship indicated by dots is that each starting position Xsi
Represents the value of the acceleration time Ti with respect to. In the figure, the broken line is a straight line connecting dots, and 31 is the target movement distance Xd.
The case where j is 1 mm is shown. Similarly, 32 and 33 are target moving distances
It shows the case where Xdj is 5 mm and 10 mm, respectively.

If the target movement distance Xdj is 1 mm, then 5 m
In the range of m = Xu1 <Xs <Xu2 = 20 mm, the point (Xsi, Ti) is almost on the solid line. Here, Xu1 and Xu2 represent the positions of the boundaries between the region with force unevenness and the region without force unevenness. This is because when the target movement distance Xdj is relatively small,
It shows that the position of the boundary of the force unevenness of the actuator can be estimated by obtaining the relationship between the start position Xsi and the acceleration time Ti in the above procedure.

FIG. 17 shows that the acceleration time Ti must be larger than the value on the solid line in the region where the actuator has uneven force (Xs <Xu1 and Xs> Xu2). Further, the greater the unevenness of force, the larger the acceleration time Ti must be larger than the value on the solid line.

As described above, by using the above procedure, it is possible to estimate the acceleration time of the access command suitable for the access by accessing the predetermined travel distance using the start position Xs as a parameter. .

The relationship between the start position Xs and the acceleration time Ti obtained by the above procedure is a discrete relationship as shown in FIG. Therefore, the acceleration time T with respect to the arbitrary target movement distance Xd and the start position Xs is obtained by the interpolation calculation as described below. This step is performed just prior to access during normal use of the device.

18 to 20 are partially enlarged views of the graph shown in FIG. FIG. 18 is an enlarged view of a portion where the start position Xs is Xs <Xu1 when the target movement distance Xd is 10 mm. In this case, the acceleration time T with respect to the arbitrary start position Xs can be calculated by the equation (11).

Tj = [Ti-T (i-1)] · [Xs-Xs (i-1)] / [Xsi-Xs (i-1)] + T (i-1) (11) Therefore, in the example shown in FIG. 18, Xs = 2 mm, Xsi = 2.5 in equation (11).
The acceleration time Tj = 2.6 mm can be obtained by substituting mm, Xs (i-1) = 1.25 mm, Ti = 2.5 msec, T (i-1) = 2.75 msec.

Next, the acceleration time T with respect to the arbitrary target movement distance Xdc can be calculated as follows based on the relationship between the start position Xsi and the acceleration time Ti which is obtained in advance and shown in FIG.

FIG. 19 is an enlarged view of a portion where the start position Xs is Xs <Xu1 when the target movement distance Xd is 5 mm and 10 mm. Certain start position Xs (Xs = 2mm in this case)
In, the target movement distance Xdj and the acceleration time Tj are regarded as having a linear relationship, and interpolation is performed based on the equation (12).

T = [Tj-T (j-1)] · [Xdc-Xd (j-1)] / [Xdj-Xd (j-1)] + T (j-1) (12) Therefore, in the example shown in FIG. 19, first, the acceleration time Tj at the start position Xs = 2 mm is calculated based on the equation (11) for both Xdj = 10 mm and Xd (j-1) = 5 mm. Tj = 2.6m
We get m, T (j-1) = 2.3. Then in equation (12) these values and X
By substituting dc = 8mm, the acceleration time T = 2.48mm can be obtained.

FIG. 20 is an enlarged view of a portion where the start position Xs is Xs> Xu2 when the target movement distance Xd is 5 mm and 10 mm. Predetermined target movement distance Xd (i-1), Xd
A straight line obtained by interpolating the straight lines 32 and 33 corresponding to i with an arbitrary target movement distance Xsc is shown as a straight line 34. First with this straight line 34
The value of the start position Xsc at the intersection A with the straight line of T = 2 msec is interpolated based on the equation (13).

Xsc = [Xsi-Xs (i-1)] / [Xdc-Xd (i-1)] / [Xdi-Xd (i-1)] + Xs (i-1) (13) In the example shown in Fig. 20, Xdc = 8mm, Xs (i-1) = 17.5m in equation (13).
The interpolated start position Xsc = 16 mm is obtained by substituting m, Xsi = 15 mm, Xd (i-1) = 5 mm, and Xdi = 10 mm.

Next, the slopes K (i-1) and Ki of the straight lines 32 and 33 are calculated by the formula (1
Interpolate based on 4). Kc = [Ki-K (i-1)] ・ [Xdc-Xd (i-1)] / [Xdi-Xd (i-1)] + K (i-1) (14) Further Fig. 20 In the example shown in, the equation (14) has Xdc = 8mm, K (i-1) = 0.2
By substituting sec / m, Ki = 0.28 sec / m, Xd (i-1) = 5 mm, and Xdi = 10 mm, the slope Kc = 0.248 sec / m of the interpolation line 34 can be obtained. Further, using the relationship of the obtained interpolation straight line, the acceleration time T with respect to the start position Xs is calculated based on the equation (15).

T = Kc · [Xs−Xsc] + Tc (15) Further, in the example shown in FIG. 20, K = 0.24 sec / m, Xs = 18.
By substituting 75 mm, Xsc = 16 mm, Tc = 2 msec, the acceleration time T = 2.682 msec can be obtained.

As described above, using the above procedure, based on the discrete relationship between the start position Xs and the acceleration time Ti,
The acceleration time T for an arbitrary target movement distance Xd and start position Xs can be obtained by interpolation calculation.

21 to 22 show the construction of the interpolation calculation means 21 used in this embodiment. FIG. 21 shows the above equation (1
It is configured to be able to execute interpolation calculation based on 1) and (12). Further, FIG. 22 is configured to be able to execute the interpolation calculation based on the above equations (13), (14) and (15).
The configuration and operation will be described below.

In FIG. 21, the interpolation calculation means 21 is composed of an interface circuit 41, a first interpolation calculation circuit 42, a register 43, and a second interpolation calculation circuit 44. The interface circuit 41 receives the digital data from the microcomputer 3 and sends it to the first interpolation calculation circuit 42 and the second interpolation calculation circuit 44.

The first interpolation calculation circuit 42 uses the above equation (11).
It is provided so that the interpolation calculation represented by can be executed, and the acceleration time Tj with respect to an arbitrary start position Xs is obtained based on the relationship of (target movement distance Xdj, start position Xsi, acceleration time Ti) given in advance. The register 43 stores two or more pieces of this calculation data Tj, and sets the data to the second interpolation calculation circuit 44 according to the timing (not shown) from the outside.
Send out one by one.

The second interpolation calculation circuit 44 uses the above equation (12).
Is provided so as to be able to execute the interpolation calculation, and the acceleration time T for an arbitrary target movement distance Xd is obtained based on the data T (j-1), Tj obtained by the first interpolation calculation circuit 42. . The obtained data T is sent to the microcomputer via the interface circuit 41.

In FIG. 22, the interpolation calculation means 21 is composed of an interface circuit 41, a third interpolation calculation circuit 45, a fourth interpolation calculation circuit 46, and a fifth interpolation calculation circuit 47. The interface circuit 41 receives the digital data from the microcomputer 3 and receives the digital data from the third interpolation calculation circuit 45 and the fourth interpolation calculation circuit 45.
To the interpolation calculation circuit 46 and the fifth interpolation calculation circuit 47.

The third interpolation calculation circuit 45 uses the above equation (13).
It is provided so that the interpolation calculation represented by can be executed, and based on the values of the start positions Xs (i-1) and Xsi for the target movement distances Xd (i-1) and Xdi given in advance, any target movement can be performed. The data of the start position Xsc at the intersection A with respect to the distance Xdc is obtained.

The fourth interpolation calculation circuit 46 uses the above equation (14).
It is provided so that the interpolation calculation represented by can be executed, and the inclination K (i for the target movement distances Xd (i-1) and Xdi given in advance
-1), based on the value of KXi, the data of the slope Kc with respect to an arbitrary target moving distance Xdc is obtained.

The fifth interpolation calculation circuit 47 uses the above equation (15).
Is provided so as to be able to execute the interpolation calculation represented by, and the start position Xsc, the inclination Kc data obtained by the third interpolation calculation circuit 45 and the fourth interpolation calculation circuit 46, and any start position Xs, acceleration The acceleration time T is calculated based on each input data of the time Tc. The obtained data T is sent to the microcomputer via the interface circuit 41.

The first interpolation calculation circuit 42, the second interpolation calculation circuit 44, the third interpolation calculation circuit 45, the fourth interpolation calculation circuit 46, and the fifth interpolation calculation circuit 47 are each plural in number. The adder and the multiplier are used to form the circuit using a known circuit configuration method.

[0093]

As described above, according to the present invention, the actuator is preliminarily accessed for a predetermined movement distance, and the most suitable access command signal based on the trial access is used, and the movable range of the actuator is set using the start position as a parameter. Estimate over the whole and use the result,
A method of correcting the access command signal based on a fuzzy inference operation in which the start position of the actuator and either the access command signal or a predetermined moving distance are used as input variables and the correction value of the access command signal is used as an output variable As a result, it has an excellent effect that it is possible to easily and accurately estimate an access command in a short time in consideration of the influence of the force unevenness of the actuator, which has been difficult in the past.

[Brief description of drawings]

FIG. 1 is a block diagram of an access control device that realizes an actuator access method according to an embodiment of the present invention.

FIG. 2 is a flowchart showing an actuator access method according to an embodiment of the present invention.

FIG. 3 is a waveform diagram showing a membership function in one embodiment of the present invention.

FIG. 4 is a waveform diagram showing a membership function in one embodiment of the present invention.

FIG. 5 is a waveform diagram showing a membership function in one embodiment of the present invention.

FIG. 6 is a waveform diagram showing an example of a calculation process of a fuzzy inference rule.

FIG. 7 is a waveform diagram showing an example of a calculation process of a fuzzy inference rule.

FIG. 8 is a waveform diagram showing an example of a calculation process of a fuzzy inference rule.

FIG. 9 is a waveform diagram showing an example of a calculation process of a fuzzy inference rule.

FIG. 10 is a waveform diagram showing an example of a calculation process of a fuzzy inference rule.

FIG. 11 is a waveform diagram showing an example of a calculation process of a fuzzy inference rule.

FIG. 12 is a waveform diagram showing a membership function in another example of the present invention.

FIG. 13 is a waveform diagram showing a membership function in another example of the present invention.

FIG. 14 is a block diagram showing a specific configuration of a fuzzy inference operation circuit according to an embodiment of the present invention.

FIG. 15 is a block diagram of an access control device that realizes an actuator access method according to an embodiment of the present invention.

FIG. 16 is a flowchart showing an actuator access method according to an embodiment of the present invention.

FIG. 17 is a control diagram showing the relationship between the start position of the actuator and the acceleration time in the embodiment of the present invention.

[Fig. 18] Fig. 18 is a contrast diagram in which a part of the relationship of (Fig. 17) is enlarged.

[Fig. 19] A comparative view in which a part of the relationship of (Fig. 17) is enlarged.

[Fig. 20] Fig. 17 is a contrast diagram showing an enlarged part of the relationship.

FIG. 21 is a block diagram showing a specific configuration of an interpolation calculation circuit according to an embodiment of the present invention.

FIG. 22 is a block diagram showing a specific configuration of an interpolation calculation circuit according to an embodiment of the present invention.

FIG. 23 is a block diagram of an actuator drive circuit in a conventional access control device.

[Fig. 24] Side view of a conventional actuator

[Fig.25] Distribution diagram of the force constant of the actuator

FIG. 26 (a) Waveform diagram of current I flowing in the electromagnetic coil (b) Waveform diagram of actuator movement speed V (c) Waveform diagram of actuator movement distance X

[Explanation of symbols]

 1 control circuit 2 fuzzy inference operation means 3 microcomputer 4 memory 5 interface circuit 6 drive circuit 7 actuator 8 position encoder 9 position detection circuit 10 speed detection circuit 11 membership function memory 12 MIN operation circuit 13 MAX operation circuit 14 centroid operation circuit 20 drive Signal generation circuit 21 Interpolation calculation means 41 Interface circuit 42 First interpolation calculation circuit 43 Register 44 Second interpolation calculation circuit 45 Third interpolation calculation circuit 46 Fourth interpolation calculation circuit 47 Fifth interpolation calculation circuit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-63-77388 (JP, A) JP-A-50-62010 (JP, A) JP-A-61-77905 (JP, A) JP-A-1- 293401 (JP, A)

Claims (8)

(57) [Claims] 1. An access to an actuator for accessing the actuator by an access command signal including data representing acceleration, acceleration time, deceleration and deceleration time of the actuator for moving the actuator from a start position by a predetermined movement distance. In the method, the start position of the actuator and the acceleration, acceleration time, deceleration and deceleration time represented by the access command signal are input variables, and the acceleration, acceleration time, deceleration and the data represented by the access command signal are used. Performing a fuzzy inference operation having a plurality of rules in which at least one correction value of deceleration time is used as an output variable, and correcting the data of the access command signal based on the correction value obtained by the fuzzy inference operation Of actuator with Method. 2. Access to the actuator by an access command signal including data representing acceleration, acceleration time, deceleration and deceleration time of the actuator in order to move the actuator from the start position by a predetermined movement distance. A method, wherein the start position of the actuator and a moving distance to be accessed by the actuator are input variables, and at least one correction value of acceleration, acceleration time, deceleration, and deceleration time represented by the data of the access command signal is used. The method of accessing an actuator, comprising: performing a fuzzy inference operation having a plurality of rules using the output variable as an output variable; and correcting the data of the access command signal based on a correction value obtained by the fuzzy inference operation. 3. A drive unit for driving an actuator, a position detection unit for detecting a position of the actuator, an access command signal including data representing acceleration, acceleration time, deceleration time and deceleration time of the actuator, The start position of the actuator and the acceleration, acceleration time, deceleration and deceleration time represented by the access command signal are used as input variables, and at least one of the acceleration, acceleration time, deceleration and deceleration time represented by the data of the access command signal. A fuzzy inference operation means for performing a fuzzy inference operation having a plurality of rules using a correction value as an output variable, and the actuator for correcting the data of the access command signal based on the correction value obtained by the fuzzy inference operation means. Acceleration of the actuator in order to move the And a drive signal generating means for generating a drive signal having data representing the acceleration time, the acceleration time, the deceleration time, and the deceleration time. 4. A driving means for driving an actuator, a position detecting means for detecting a position of the actuator, an access command signal including data representing acceleration, acceleration time, deceleration time and deceleration time of the actuator, A plurality of actuators having start positions and movement distances to be accessed by the actuators as input variables, and having at least one correction value of acceleration, acceleration time, deceleration, and deceleration time represented by the data of the access command signal as an output variable. And a fuzzy inference calculation means for performing a fuzzy inference calculation having the above rule, and based on a correction value obtained by the fuzzy inference calculation means, corrects the data of the access command signal and moves the actuator by a predetermined movement distance. For the acceleration, acceleration time, deceleration of the actuator And a drive signal generating means for generating a drive signal having data representing deceleration time. 5. An actuator that accesses the actuator by an access command signal including data representing acceleration, acceleration time, deceleration and deceleration time of the actuator in order to move the actuator by a predetermined moving distance from its start position. An access method, wherein an acceleration, an acceleration time, a deceleration and a deceleration time of an access command signal suitable for a predetermined moving distance of the actuator,
Using each start position of the actuator as a parameter, repeat it over the entire movable range of the actuator.
A method for accessing an actuator, the method including the step of obtaining by returning access . 6. An acceleration of the access command signal for a predetermined start position and a predetermined movement distance of the actuator based on the acceleration, acceleration time, deceleration and deceleration time of the access command signal suitable for the predetermined movement distance of the actuator. 6. The actuator access method according to claim 5, further comprising a step of obtaining the acceleration time, the deceleration time, and the deceleration time by interpolation calculation. 7. A step of setting a target moving distance corresponding to an access command signal of an actuator, a step of calculating an acceleration time of the access command signal by a predetermined calculation formula based on the target moving distance, and the actuator. Setting a start position of the actuator, moving the actuator to a start position, and actually moving the actuator with an access command signal having the calculated acceleration time; and measuring an actual moving distance of the actuator. A step of calculating a deviation between the actual movement distance and the target movement distance, a step of determining whether or not the deviation is smaller than a predetermined threshold value, and the deviation being smaller than the threshold value , in addition a predetermined correction amount to said acceleration time, obtains the corrected acceleration time step , In the case where the deviation is greater than the threshold value, the Aku
Move the chute to the start position and perform the corrected acceleration
A step of actually moving the actuator with an access command signal having a gap, a step of measuring an actual moving distance of the actuator, a step of calculating a deviation between the actual moving distance of the actuator and a target moving distance, and the deviation being Storing the target movement distance, the start position and the corrected acceleration time in a memory when repeating the step of judging whether the deviation is smaller than a predetermined threshold and the step of judging whether the deviation is smaller than the threshold. 6. A step of changing the starting position and repeating the steps for all the starting positions, and a step of changing the target moving distance and repeating the steps for all the target moving distances. Actuator access method described in. 8. Driving means for driving an actuator, position detecting means for detecting a position of the actuator, and actual moving distance detecting means for detecting an actual moving distance of the actuator based on an output of the position detecting means. , Repeatedly accessing the acceleration, acceleration time, deceleration, and deceleration time of an access command signal suitable for a predetermined start position and movement distance of the actuator over the entire movable range of the actuator using each start position of the actuator as a parameter. To seek by
The optimum access command signal creating means, based on the acceleration, acceleration time, deceleration and deceleration time of the access command signal suitable for the predetermined start position and movement distance of the actuator obtained by the optimum access command signal creating means, An access control device comprising: an interpolating calculation means for obtaining an acceleration, an acceleration time, a deceleration and a deceleration time of an access command signal for a predetermined start position and a predetermined movement distance of the actuator.