Magnetic disk device, control method and program

Abstract

To provide a magnetic disk device, a control method, and a program to improve faster seek control and settling accuracy.SOLUTION: A magnetic disk device includes a magnetic disk, an arm; a support member provided at a tip of the arm; a magnetic head that is provided to the support member and accesses to the magnetic disk; a first actuator that rotates the arm; a second actuator that drives the support member; and a control unit that performs seek control of the magnetic head. The control unit includes a feedback control unit and a repeated learning control unit. The feedback control unit controls a first command value for driving the first actuator so as to reduce errors between the position of the magnetic head and a target position. The repeated learning control unit inputs an input value based on the errors and outputs a second command value for driving the second actuator by repeated learning.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] An embodiment of the present invention relates to a magnetic disk device, a control method, and a program. [Background technology]

[0002] In magnetic disk drives, for example, a two-stage actuator consisting of a voice coil motor (VCM) and a microactuator (MA) controls the positioning of the magnetic head that reads and writes data to the target track (seek control). To improve sequential read / write (R / W) performance, it is necessary to increase the speed of adjacent track seek, which moves the magnetic head one track at a time. Furthermore, to accommodate the recent trend toward narrower track pitches, there is also a demand for higher accuracy in positioning during seek settling.

[0003] To increase speed, a method using the advance operation of MA has been proposed. Also, to improve settling, a method using iterative learning control has been proposed. In conventional seek control using iterative learning control, the control input and target trajectory of the VCM are generally updated for each seek. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] U.S. Patent No. 7,626,782 [Non-patent literature]

[0005] [Non-Patent Document 1] J. Ishikawa, et al, “A Robust Stability Analysis on Learning Control for Hard Disk Drives”, Advances in Information Storage Systems, pp.49-61 (1999) [Non-patent document 2] M. Kobayashi, R. Horowitz, “Track Seek Control for Hard Disk Dual-Stage Servo Systems”, IEEE Trans. on Magnetics, Vol.37, No.2, (2001) Summary of the Invention [Problem to be solved by the invention]

[0006] An object of the present invention is to provide a magnetic disk device, a control method, and a program that can achieve high-speed seek control and high-precision settling. [Means for solving the problem]

[0007] A magnetic disk drive according to an embodiment includes a magnetic disk, an arm, a support member attached to the tip of the arm, a magnetic head attached to the support member and configured to access the magnetic disk, a first actuator for rotating the arm, a second actuator for driving the support member, and a control unit for performing seek control of the magnetic head. The control unit includes a feedback control unit and an iterative learning control unit. The feedback control unit controls a first command value for driving the first actuator so as to reduce an error between the position of the magnetic head and a target position. The iterative learning control unit receives an input value based on the error and outputs a second command value for driving the second actuator through iterative learning. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a block diagram of a magnetic disk device according to an embodiment. [Figure 2] FIG. [Figure 3] 5A and 5B are schematic diagrams showing the relationship between operations during seek control and a seek target trajectory. [Figure 4] 5A and 5B are diagrams showing examples of time changes of signals during seek control. [Figure 5] FIG. 2 is a block diagram of a seek control unit according to the embodiment. [Figure 6] FIG. 1 is a diagram showing the internal structure of an iterative learning controller. [Figure 7] FIG. 2 is a diagram schematically showing frequency characteristics. [Figure 8] 10 is a flowchart of a seek control process. [Figure 9] FIG. 10 is a diagram showing an example of the frequency response of a model that simulates the transfer characteristics of an actuator. [Figure 10] FIG. 10 is a diagram showing an example of the results of simulating adjacent track seek. [Figure 11] FIG. 10 is a diagram showing an example of the results of simulating adjacent track seek. DETAILED DESCRIPTION OF THE INVENTION

[0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of a magnetic disk drive according to the present invention will now be described in detail with reference to the accompanying drawings.

[0010] This embodiment provides a high-speed, high-precision adjacent track seek control method for a magnetic disk drive having two or more stages of actuators for head positioning. For example, in this embodiment, iterative learning control is applied to the control input related to the leading operation of MA during adjacent track seek. This realizes a control method that achieves both high speed and high-precision settling.

[0011] A two-stage actuator refers to a configuration that includes an actuator Aa (first actuator, first-stage actuator) that moves the head by a relatively large amount, and an actuator Ab (second actuator, second-stage actuator) that moves the head by a relatively small amount. The actuator Aa is, for example, an actuator (such as a VCM) that drives an arm with a head located at its tip. The actuator Ab is, for example, an actuator (such as an MA) that is provided at the tip of the arm and drives a support member (such as a slider or suspension) that supports the head. The actuator Ab corresponds to the actuator at the tip of the two-stage actuator.

[0012] The three-stage actuator refers to a configuration including an actuator Aa and two actuators Ab (second-stage and third-stage actuators) that move the head by relatively small amounts. The actuators Ab are, for example, two actuators that respectively drive two support members that support the magnetic head. The two support members are, for example, a suspension connected to the arm and a slider connected to the suspension.

[0013] Although the following description will be given using an example in which the actuator has two stages, the same method can be applied to cases in which the actuator has three or more stages. For example, one of the actuators Ab included in an actuator with three or more stages can be configured to be controlled in the same way as the actuator Ab in a two-stage actuator.

[0014] FIG. 1 is a block diagram showing the configuration of the main part of an HDD (hard disk drive) 10 as a magnetic disk device according to this embodiment.

[0015] The host 20 is a device that uses the HDD 10 as a storage device. The host 20 is connected to the HDD 10 via a host interface IF.

[0016] The HDD 10 includes one or more disks 11 (magnetic disks), multiple magnetic heads 12, a spindle motor (SPM) 13, a microactuator (MA) 14, a support member 14a, a voice coil motor (VCM) 15, an arm 15a, a driver IC (Integrated Circuit) 16, a head IC 17, and a system LSI (Large Scale Integration) 18.

[0017] The disks 11 are magnetic recording media and are stacked at regular intervals. The disks 11 include an upper disk surface and a lower disk surface. In this embodiment, both sides of the disk 11 (the upper disk surface and the lower disk surface) are recording surfaces on which data is magnetically recorded. The disks 11 are rotated at high speed by the SPM 13. The HDD 10 may include one disk 11 or multiple disks 11 (for example, 10 or more).

[0018] The magnetic heads 12 are arranged corresponding to the recording surfaces (upper and lower disk surfaces) on both sides of each disk 11, and are used to write data to and read data written on the recording surfaces of the disks 11. The magnetic heads 12 are attached to the tip of the support member 14a.

[0019] The SPM 13 is driven by a drive current (or drive voltage) supplied from a driver IC 16 .

[0020] The support member 14a is a member that supports the magnetic head 12, and is provided at the tip of the arm 15a. The support member 14a is, for example, a slider or a suspension.

[0021] The MA 14 is attached to the base or tip of the support member 14a and drives the support member 14a. The MA 14 is, for example, a piezoelectric element such as a piezo element. The MA 14 corresponds to an actuator that drives the tip side of a two-stage actuator. The MA 14 can be driven faster and more accurately than the VCM 15. Therefore, by operating the MA 14 prior to the VCM 15, seek control can be speeded up.

[0022] The VCM 15 rotates the arm 15a. The VCM 15 corresponds to the actuator Aa, which moves the head by a relatively large amount. In contrast, the MA 14 corresponds to the actuator Ab, which moves the support member 14a provided at the tip of the arm 15a by a small amount (a relatively small amount).

[0023] The VCM 15 and MA 14 are driven by command values ​​supplied from a driver IC 16. This causes the magnetic head 12 to move in the radial direction of the disk 11. The command value is, for example, a drive current or a drive voltage.

[0024] The driver IC 16 drives the SPM 13, MA 14, and VCM 15 under the control of a CPU (Central Processing Unit) 186 (described later) in the system LSI 18.

[0025] The head IC 17 amplifies a signal (read signal) read by the magnetic head 12. The head IC 17 also converts write data transferred from a later-described R / W (read / write) channel 181 in the system LSI 18 into a write current and outputs it to the magnetic head 12.

[0026] The system LSI 18 is an LSI called an SoC (System On a Chip) in which multiple elements are integrated on a single chip. The system LSI 18 includes an R / W channel 181, a hard disk controller (HDC) 182, a buffer random access memory (RAM) 183, a flash memory 184, a program read only memory (ROM) 185, a CPU 186, and a RAM 187.

[0027] The R / W channel 181 is a signal processing device that processes signals related to reading and writing. The R / W channel 181 digitizes the read signal and decodes the read data from the digitized data. The R / W channel 181 also obtains servo data necessary for positioning the magnetic head 12 from the digital data. The R / W channel 181 also encodes the write data.

[0028] The HDC 182 is connected to the host 20 via the host interface IF. The HDC 182 receives commands (write commands, read commands, etc.) transferred from the host 20. The HDC 182 controls data transfer between the host 20 and the HDC 182. The HDC 182 also controls data transfer between the disk 11 and the HDC 182.

[0029] The buffer RAM 183 constitutes a buffer area for temporarily storing data to be written to the disk 11 via the head IC 17 and the R / W channel 181 and data read from the disk 11 .

[0030] The flash memory 184 is a rewritable non-volatile memory.

[0031] The program ROM 185 stores a control program (firmware). The control program may be stored in a partial area of ​​the flash memory 184. The control program is a program that is used after shipping.

[0032] The CPU 186 functions as a main controller for the HDD 10. The CPU 186 controls at least some of the elements in the HDD 10 in accordance with a control program or an adjustment program stored in a program ROM 185.

[0033] At least a portion of the RAM 187 is used as a working area for the CPU 186 .

[0034] The HDD 10 configured as described above includes a feedback control system that executes seek control of the magnetic head 12. The feedback control system is executed by, for example, the CPU 186 at regular intervals, that is, each time servo data is acquired. Hereinafter, this regular time interval will be referred to as the sample time.

[0035] 2 is a block diagram showing an example of the configuration of a seek control unit 100a (an example of a control unit) that can be used for seek control of the magnetic head 12. The seek control unit 100a includes a feedback control system 200a, calculators 311, 312, and 315, transfer characteristics 313 and 314, and an error detector 316.

[0036] The feedback control system 200a corresponds to a feedback control unit that controls a command value (first command value) for driving the VCM 15 so as to reduce (for example, set to zero) the error (position error signal PES, described later) between the actual position of the magnetic head 12 and the target position of the magnetic head 12. The feedback control system 200a includes an MA controller 201, an MA notch filter 202, a downsampler 203, a calculator 204, a state feedback observer 205a, a VCM notch filter 206, and a gain calculator 207.

[0037] The MA controller 201 controls the control input u of the MA 14. m Calculate [k], where k represents the identification information that identifies the sample time. For example, u m [k] represents the control input of the MA 14 at the k-th sample time. In the following, the notation [k] representing the sample time may be omitted.

[0038] The MA notch filter 202 receives the control input u m This is a filter that removes the mechanical resonance frequency component of the MA 14. The MA notch filter 202 is used to improve the stability of the feedback loop.

[0039] The sampling rate of the output signal of the MA notch filter 202 is determined by the control input u m The MA notch filter 202 may be configured, for example, with a control input u mThe output signal is output at a sampling rate (hereinafter referred to as SRb) that is higher than the sampling rate (hereinafter referred to as SRa). For example, SRb is a sampling rate that is twice as high as SRa. In Figure 2 and the following figures, the locations where the sampling rate is SRa are indicated by single arrows, and the locations where the sampling rate is SRb are indicated by double arrows.

[0040] The downsampler 203 downsamples the sampling rate of the output signal of the MA notch filter 202 from SRb to SRa.

[0041] The calculator 204 calculates the difference between a position error signal (PES) that is the output of the error detector 316 and the output signal of the downsampler 203. The difference output by the calculator 204 corresponds to a subtraction value obtained by subtracting a value corresponding to the amount of movement of the MA 14 (the output signal of the downsampler 203) from the position error signal PES.

[0042] The error detector 316 detects a position error signal (PES) indicating the difference between the target position r[k] of the magnetic head 12 at the kth sample time and the head position signal y[k] indicating the actual position of the magnetic head 12 (hereinafter referred to as the head position), and inputs the signal to the feedback control system 200a. The head position signal y corresponds to the observation signal. The target position r[k] is obtained as the target position at the kth sample time based on the seek target trajectory 301 of the magnetic head 12.

[0043] The calculation by the calculator 204 is performed by a non-interacting loop to calculate the displacement y m This corresponds to a calculation to obtain a signal obtained by subtracting a signal corresponding to minutes from PES. The calculator 204 inputs the calculation result to the state feedback observer 205a.

[0044] The state feedback observer 205a estimates the position, speed, and disturbance input of the magnetic head 12 as a state, multiplies the estimated state by a state feedback gain (a gain "-F" multiplied by a gain calculator 254 described later), and outputs the control input u of the VCM 15. v The state feedback+observer 205a includes calculators 251, 253, and 257, gain calculators 252, 254, 255, 256, and 259, and a delay unit 258. Note that the configuration of the state feedback+observer 205a shown in FIG. 2 is an example, and the configuration is not limited to this.

[0045] The calculator 251 calculates the difference e between the output of the calculator 204 and the output of the gain calculator 259. The difference e corresponds to the position estimation error by the state feedback plus the observer 205a. The gain calculator 252 multiplies the output of the calculator 251 by a gain L. The calculator 253 calculates the addition of the output of the gain calculator 252 and the output of the delay device 258. The gain calculator 254 multiplies the output of the calculator 253 by a gain "-F" and outputs the multiplication result as the control input u v The gain calculator 255 multiplies the output of the calculator 253 by a gain "A". The gain calculator 256 multiplies the output of the calculator 253 by a gain "A". v The gain calculator 257 adds the output of the gain calculator 256 and the output of the gain calculator 255. The delay device 258 delays the output of the calculator 257 by one sample time. The gain calculator 259 multiplies the output of the delay device 258 by a gain "C". The output value of the gain calculator 259 is a sum of the state feedback and the position estimate by the observer 205a (y v (corresponding to the value with a bar above [k]).

[0046] The VCM notch filter 206 is a filter that outputs the control input u calculated by the state feedback observer 205a. v This is a filter that removes the mechanical resonance frequency component of the VCM 15. The VCM notch filter 206 is used to improve the stability of the feedback loop.

[0047] The gain calculator 207 multiplies the PES output from the error detector 316 by a gain of −1, and outputs the multiplication result to the MA controller 201 .

[0048] The calculator 311 performs an addition operation on the output signal of the MA notch filter 202 and the input MA advance amount command value 302, and inputs the operation result to the transfer characteristic 313. The calculator 312 performs an addition operation on the output signal of the VCM notch filter 206 and the input VCM FF (feedforward) input 303, and inputs the operation result to the transfer characteristic 314.

[0049] The MA advance amount command value 302 represents a predetermined command value for operating the MA 14 prior to driving the VCM 15 (advance operation). The VCM FF input 303 represents an input for feedforward control of the VCM 15. The VCM FF input 303 is expressed in, for example, the dimension of acceleration.

[0050] The transfer characteristic 313 represents the transfer characteristic Pm of the MA 14. The transfer characteristic 313 represents the relationship between a given control input and the displacement y m The transfer characteristic 314 represents the transfer characteristic Pv of the VCM 15. The transfer characteristic 314 represents how a given control input is dynamically reflected in the displacement y of the VCM 15. v The observed signal of the feedback control system 200a is the head position signal y, and the displacement y of the MA 14 is m and the displacement y of VCM15 v The calculator 315 calculates the displacement y of the MA 14, which corresponds to the head position signal y. m and the displacement y of VCM15 v For convenience, the sum (additive value) of

[0051] As described above, a seek target trajectory 301 corresponding to the target trajectory r of the magnetic head 12 is input to the feedback control system 200a, and a VCM FF input 303 is input to the VCM 15. Furthermore, an MA advance amount command value 302 is input to the MA 14. The seek target trajectory 301, the MA advance amount command value 302, and the VCM FF input 303 are pre-stored in the flash memory 184, for example, as time-series data. Furthermore, these time-series data are read from the flash memory 184 after the HDD 10 is started and expanded in the RAM 187, and are read from the RAM 187 and used each time seek control is executed. Each time-series data may be determined, for example, for each distance moved by seek control (for example, the distance between tracks before and after movement).

[0052] Next, an example of seek control of the magnetic head 12, including the preceding operation of the MA, will be described. Figure 3 is a schematic diagram showing the relationship between the operations of the VCM 15 and MA 14 during seek control and the seek target trajectory.

[0053] For example, when the configuration example of the feedback control system 200a shown in Fig. 2 is used, the MA advance amount command value 302 is given to the MA 14, which can be driven faster and more accurately than the VCM 15, and seek control is executed so that the MA 14 moves to the target track ahead of the VCM 15 (Fig. 3 (1)). This allows the magnetic head 12 to move from the current track to the target track at higher speed.

[0054] VCM 15 receives VCM FF input 303 and moves to the target track so as to follow MA 14 ((2) in FIG. 3). Magnetic head 12 has already moved to the target track before VCM 15 has finished moving to the target track. This allows magnetic head 12 to read and write data more quickly than when seek control is performed solely by VCM 15. The trajectory of the head position is the sum of the trajectory of MA 14 according to (1) and the trajectory of VCM 15 according to (2).

[0055] FIG. 4 shows an example of the time variation of each signal in the above seek control. v represents the target trajectory of the VCM 15 position. m represents the MA advance amount command value 302 of the MA 14. r represents the target trajectory r v and MA advance amount command value r m represents the target trajectory of the head position, which is the sum of

[0056] The target trajectory r of the head position is calculated from the target seek time of the adjacent track seek (the time required to determine how many sample times it takes to move to the adjacent track). v is obtained by, for example, integrating the VCM FF input 303 expressed in the dimension of acceleration twice. Therefore, in general, the MA advance amount command value r m is calculated using the following formula (1). r m =rr v ···(1)

[0057] However, when the target trajectory r calculated by equation (1) is used, the actual head position (head position signal y[k]) may not follow the target trajectory r due to factors such as the following, resulting in worsening seek settling. Vibration caused by mechanical resonance of VCM15 and MA14 Delay in calculation time when the feedback control system 200a is calculated by the CPU 186 Phase delay of the current / voltage amplifier of driver IC16

[0058] Therefore, in this embodiment, the MA advance amount command value r m Instead of calculating the MA advance amount command value r m 5 is a block diagram showing an example of the configuration of a seek control unit 100b (an example of a control unit) of this embodiment. The seek control unit 100b includes a feedback control system 200b, calculators 311, 312, and 315, transfer characteristics 313 and 314, an error detector 316, and an iterative learning controller 500.

[0059] 2 in that the seek control unit 100b has a state feedback+observer 205b configuration in the feedback control system 200b and that an iterative learning controller 500 is added. Components that are common to the feedback control system 200a are given the same reference numerals and will not be described.

[0060] State feedback+observer 205b differs from state feedback+observer 205a in that it outputs difference e, which is the output of calculator 251, to iterative learning controller 500.

[0061] The iterative learning controller 500 performs iterative learning control to obtain the MA advance amount command value r m The iterative learning controller 500 generates the MA advance amount command value r using the input value, which is the state feedback plus the position estimation error (difference e) in the observer 205b, every time seek control is executed. m By repeating the seek control, the MA advance amount command value r is updated so that the difference e approaches 0 (e≒0), that is, so that the position estimate by the observer (state feedback+observer 205a) can accurately estimate the actual head position (head position signal y[k]). m will be updated little by little.

[0062] The iterative learning controller 500 receives an input value (difference e) based on the error (position error signal PES) between the position of the magnetic head and the target position, and calculates an MA advance amount command value r, which is a command value (second command value) for driving the MA 14 through iterative learning. m This corresponds to an iterative learning control unit that outputs

[0063] 6 is a diagram showing an example of the internal structure of iterative learning controller 500. Iterative learning controller 500 includes memory 501, upsampler 502, learning filter 503, low-pass filter 504, average value calculation unit 505, integrator 506, adder 507, and adder-subtractor 508. Memory 501 is realized by, for example, RAM 187 in FIG. 1. Each unit other than memory 501 is executed by, for example, CPU 186, similar to feedback control system 200b.

[0064] Memory 501 stores, in chronological order, a plurality of pieces of learning data calculated in the j-th (j is an integer equal to or greater than 1) seek control. Numerical values ​​0, 1, 2, . . . , i, . . . shown in memory 501 in FIG. 6 correspond to index i, which indicates the order in which learning data is stored. For example, learning data obtained at each sample time of the j-th seek control (sample time upsampled by upsampler 502) is stored in order at positions indicated by indexes 0, 1, 2, . . . , i, . . . in memory 501. Furthermore, at each sample time of the next (j+1)-th seek control, learning data is again stored in order starting from index 0.

[0065] The upsampler 502 upsamples the input difference e. For example, the upsampler 502 upsamples the sampling rate of the difference e from SRa to SRb. The sample values ​​at the sample times added by the upsampling are set to 0, for example.

[0066] The learning filter 503 filters the difference e. The learning filter 503 is, for example, a multi-rate filter for the difference e upsampled by the upsampler 502. Hereinafter, the learning filter 503 may be represented as H(z). H(z) may be a multi-rate phase lag filter that approximates the inverse characteristics of the closed-loop characteristics of the feedback control system 200b. The closed-loop characteristics of the feedback control system 200b are, for example, a multi-rate phase lag filter that approximates the inverse characteristics of the closed-loop characteristics of the feedback control system 200b when a given MA advance amount command value r m represents the characteristics of the change in the difference e with respect to

[0067] FIG. 7 is a diagram schematically illustrating the relationship between the closed-loop characteristics of feedback control system 200b and the frequency characteristics of H(z). Characteristic 701 represents the closed-loop characteristics of feedback control system 200b. As shown in FIG. 7, characteristic 701 represents a characteristic in which the gain gradually increases from a small value in the low-frequency region and becomes constant at high frequencies. In order to increase the gain by canceling out the small gain in the low-frequency region of the closed-loop characteristics, characteristic 702 representing the frequency characteristics of H(z) is determined so as to approximate the inverse characteristic of the closed-loop characteristics of feedback control system 200b.

[0068] Characteristic 703 represents the frequency characteristic of the control system, which is a combination of characteristic 701 of feedback control system 200b and characteristic 702 of iterative learning controller 500. Characteristic 703 has a constant gain over a wider frequency range, which improves the convergence characteristics of the learning control.

[0069] Returning to Fig. 6, low-pass filter 504 receives as input the learning data stored in memory 501, and outputs a signal from which high-frequency components have been removed to adder 507. As shown in Fig. 6, low-pass filter 504 may be expressed as Q(z). At the sample time corresponding to index i of the j-th seek control, the value of the learning data at the position indicated by index (i+1) (the value stored in the (j-1)-th seek control) is input to low-pass filter 504. Because the low-pass filter generates a phase delay of one sample time, zero-phase low-pass characteristics are realized by using the value at a time one sample time later.

[0070] The adder 507 adds the output of the low-pass filter 504 and the output of the learning filter 503, and stores the addition result in the location indicated by the corresponding index in the memory 501 as learning data to be used in the next (j+1)th seek control.

[0071] In parallel with the processing by the low-pass filter 504 and the adder 507 , the average value calculation unit 505 calculates the average value of the values ​​of the multiple pieces of learning data stored in the memory 501 .

[0072] Adder-subtractor 508 subtracts the average value calculated by average value calculation unit 505 from the value of the learning data, and inputs the result of the subtraction to integrator 506. When the index of memory 501 in which the output of adder 507 is stored is i (at the cycle time corresponding to index i of the jth seek control), the value at the position indicated by index (i+d) corresponding to d times later (the value stored in the (j-1)th seek control) is input to adder-subtractor 508. The value at d times later is used in order to improve the stability of the learning control.

[0073] The integrator 506 calculates the integral of the output of the adder / subtractor 508 and outputs the calculation result as the MA advance amount command value r m For example, the integrator 506 adds the output of the adder / subtractor 508 each time the index increases from the value corresponding to index 0, and outputs the sum as the integral value.

[0074] The integrator 506 integrates a value based on a plurality of learning data stored in the memory 501 to obtain the MA advance amount command value r m More specifically, the integrator 506 integrates a value obtained by subtracting the average value calculated by repeated learning in the j-th seek control from one of the multiple learning data (the value at the position indicated by the index (i+d)), and outputs the MA advance amount command value r m Output.

[0075] The repetitive learning control is executed, for example, for each of a plurality of sample times corresponding to the memory length of the memory 501. After the repetitive learning control is completed, that is, after the sample time corresponding to the memory length, the MA advance amount command value r m It is preferable that the MA advance amount command value r m If the final value of is not 0, a transient response is likely to occur, which causes a disturbance to the head positioning.

[0076] Therefore, in this embodiment, the adder-subtractor 508 outputs a signal obtained by subtracting the average value from the value of the learning data to the integrator 506, and the integrator 506 integrates this signal and calculates the value as the MA advance amount command value r m As a result, the MA advance amount command value r m The final value of is always 0, which prevents the transient response described above.

[0077] Next, a description will be given of a seek control process performed by the HDD 10. Fig. 8 is a flowchart showing an example of a seek control process according to the embodiment.

[0078] For example, when the HDD 10 is started, the seek control unit 100b reads the seek target trajectory 301 and the VCM FF input 303 from the flash memory 184 and stores them in the RAM 187 (step S101).

[0079] The seek control unit 100b determines whether or not seek control has been instructed based on, for example, a command from the host 20 (step S102). If seek control has not been instructed (step S102: No), the process is repeated until it is instructed.

[0080] When the seek control is instructed (step S102: Yes), the seek control unit 100b controls the control input u of the MA 14 and the VCM 15 by feedback control. m and u v is calculated, and the MA advance amount command value r m (Step S103). After that, the process returns to Step S102 for the next seek control, and the process is repeated.

[0081] Next, the results of a simulation of the seek control of this embodiment will be described. Fig. 9 is a diagram showing an example of the frequency response of a model that simulates the transfer characteristics of the MA 14 and VCM 15. In general, in a magnetic disk drive equipped with a two-stage actuator, the main mechanical resonance mode of the MA 14 is designed to be in a frequency band about 10 times higher than the main mechanical resonance mode of the VCM 15. The model shown in Fig. 9 represents a model designed using a similar method.

[0082] Fig. 10 is a diagram showing an example of the results of simulating an adjacent track seek using the model of Fig. 9 and seek control according to equation (1) as in Fig. 2. Fig. 11 is a diagram showing an example of the results of simulating an adjacent track seek using the model of Fig. 9 and the seek control of this embodiment (Fig. 5). Figs. 10 and 11 show the time response of the position of the VCM 15 (top graph), the time response of the position of the MA 14 (center graph), and the time response of the head position, which is the sum of the movement amounts of both (bottom graph), during seek control. The horizontal axis represents time.

[0083] As shown in the area 1001 of FIG. 10, r m In the generation method of (1), for example, due to the influence of mechanical resonance included in the MA 14 and VCM 15, a deviation occurs between the target trajectory and the actual position, causing vibration during seek settling.

[0084] On the other hand, as shown in area 1101 of FIG. 11, the iterative learning control m It can be seen that in this embodiment, which generates the above equation, the vibration during seek settling can be suppressed by the effect of iterative learning.

[0085] As described above, in this embodiment, a second- or third-stage actuator (such as an MA) precedes the operation of a first-stage actuator (such as a VCM) to move ahead of the target track, thereby achieving high-speed adjacent track seek. At this time, the control input of the MA is updated for each seek by iterative learning control. This improves seek settling. In other words, this embodiment can achieve faster seek control and more accurate settling.

[0086] Each unit shown in FIG. 5 is realized, for example, by one or more processors. For example, each unit may be realized by having a processor such as CPU 186 execute a program, i.e., by software. Each unit may be realized by a processor such as a dedicated IC (Integrated Circuit), i.e., by hardware. Each unit may be realized by a combination of software and hardware. When multiple processors are used, each processor may realize one of the units, or two or more of the units.

[0087] The programs executed by the magnetic disk device of the embodiment are provided in advance in the program ROM 185 or the like.

[0088] The program executed by the magnetic disk device of the embodiment may be configured to be provided as a computer program product by being recorded in an installable or executable file format on a computer-readable recording medium such as a CD-ROM (Compact Disk Read Only Memory), a flexible disk (FD), a CD-R (Compact Disk Recordable), or a DVD (Digital Versatile Disk).

[0089] Furthermore, the program executed by the magnetic disk device according to the embodiment may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. Also, the program executed by the magnetic disk device according to the embodiment may be provided or distributed via a network such as the Internet.

[0090] The program executed by the magnetic disk device according to the embodiment can cause a computer to function as each part of the magnetic disk device described above. In this computer, the CPU 186 can read the program from a computer-readable storage medium onto the main storage device and execute it.

[0091] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0092] 10 HDD 11 discs 12 Magnetic head 13 SPM 14 MA 14a Support member 15 VCM 15a Arm 16 Driver IC 17 Head IC 18 System LSI 20 hosts 100a, 100b Seek control section 181 R / W channels 182 HDC 183 Buffer RAM 184 Flash Memory 185 Program ROM 186 CPU 187 RAM 200a, 200b Feedback control system 201 MA controller 202 MA notch filter 203 Down Sampler 204 Arithmetic unit 205a, 205b State feedback + observer 206 VCM notch filter 207 Gain calculator 251, 253, 257 arithmetic unit 252, 254, 255, 256, 259 Gain calculator 258 Delay 301 Seek target trajectory 302 MA advance amount command value 303 VCM FF input 311, 312 Arithmetic unit 313, 314 Transfer characteristics 316 Error Detector 500 Learning Controller 501 memory 502 Upsampler 503 Learning Filter 504 Low-pass filter 505 Average value calculation unit 506 Integrator 507 Adder 508 Adder / Subtractor

Claims

1. A magnetic disk, Arm and a support member provided at the tip of the arm; a magnetic head provided on the support member and configured to access the magnetic disk; a first actuator that rotates the arm; a second actuator that drives the support member; a control unit that performs seek control of the magnetic head; Equipped with The control unit a feedback control unit that controls a first command value for driving the first actuator so as to reduce an error between a position of the magnetic head and a target position; an iterative learning control unit that receives an input value based on the error and outputs a second command value for driving the second actuator through iterative learning, The iterative learning control unit a learning filter that is a filter for the input value and is a multi-rate phase lag filter that approximates the inverse characteristics of the closed loop characteristics of the feedback control unit; a memory for storing a plurality of learning data output from the learning filter at a plurality of sampling times of the seek control; an integrator that integrates a value based on the plurality of learning data stored in the memory and outputs the second command value; Equipped with the input value is a difference between a subtraction value obtained by subtracting a value corresponding to the movement amount of the second actuator from the error and an estimated value of the position of the magnetic head; Magnetic disk device.

2. the iterative learning control unit includes an average value calculation unit that calculates an average value of a plurality of the learning data, the integrator integrates a value obtained by subtracting the average value calculated by the repeated learning in the seek control for a j-th time (j is an integer equal to or greater than 1) from one of the plurality of learning data, and outputs the second command value for the seek control for a (j+1)-th time.

2. The magnetic disk drive according to claim 1.

3. The feedback control unit an observer that estimates the position of the magnetic head as a state and outputs the estimated value; 2. The magnetic disk drive according to claim 1.

4. the support member is one or both of a slider and a suspension; 2. The magnetic disk drive according to claim 1.

5. A control method executed by a magnetic disk device, comprising: The magnetic disk device A magnetic disk, Arm and a support member provided at the tip of the arm; a magnetic head provided on the support member and configured to access the magnetic disk; a first actuator that rotates the arm; a second actuator that drives the support member; a control unit that performs seek control of the magnetic head; Equipped with The control unit a feedback control step of controlling a first command value for driving the first actuator so as to reduce an error between the position of the magnetic head and a target position; an iterative learning control step of inputting an input value based on the error and outputting a second command value for driving the second actuator through iterative learning; Including, The iterative learning control step a filter step for performing processing on the input value, which is a multi-rate phase lag filter that approximates the inverse characteristics of the closed loop characteristics of the feedback control step; a storage step of storing a plurality of learning data output from the filtering step at a plurality of sampling times of the seek control in a memory; an integration step of integrating a value based on the plurality of learning data stored in the memory and outputting the second command value; Including, the input value is a difference between a subtraction value obtained by subtracting a value corresponding to the movement amount of the second actuator from the error and an estimated value of the position of the magnetic head; Control method.

6. A program to be executed by a computer included in a magnetic disk device, The magnetic disk device A magnetic disk, Arm and a support member provided at the tip of the arm; a magnetic head provided on the support member and configured to access the magnetic disk; a first actuator that rotates the arm; a second actuator that drives the support member; a control unit that performs seek control of the magnetic head; Equipped with The computer, a feedback control step of controlling a first command value for driving the first actuator so as to reduce an error between the position of the magnetic head and a target position; an iterative learning control step of inputting an input value based on the error and outputting a second command value for driving the second actuator through iterative learning; Execute The iterative learning control step a filter step for performing processing on the input value, which is a multi-rate phase lag filter that approximates the inverse characteristics of the closed loop characteristics of the feedback control step; a storage step of storing a plurality of learning data output from the filtering step at a plurality of sampling times of the seek control in a memory; an integration step of integrating a value based on the plurality of learning data stored in the memory and outputting the second command value; Including, the input value is a difference between a subtraction value obtained by subtracting a value corresponding to the movement amount of the second actuator from the error and an estimated value of the position of the magnetic head; program.

Images