KR100652423B1 - Method for designing the controller of hard disk drives and apparatus thereof - Google Patents

Abstract

A method and an apparatus for designing a controller of an HDD(Hard Disc Drive) are provided to allow the HDD to have the optimum performance and have a stable characteristic for a characteristic deviation according to the HDD and a head by calculating a coefficient of the controller to minimize a position following error obtained from a plant model. A plurality of plant models for a notch filter of an HDD and a plurality of HSAs(Head Stack Assemblies) are obtained(900,910). An equivalent position disturbance is obtained by using a plurality of the plant models(920). A controller parameter value for satisfying certain stability standard for a plurality of the plant models and simultaneously minimizing an average position following error for the equivalent position disturbance of the HDD is calculated(930).

Description

Method for designing the controller of hard disk drives and Apparatus

Figure 1 shows the mathematical model of the conventional servo system and the frequency response characteristics of the actual plant.

2 shows a configuration of a hard disk drive to which the present invention is applied.

3 is a block diagram of a servo system to which the present invention is applied.

4 is a block diagram of a controller design apparatus according to the present invention.

5 is a board plot showing a process of obtaining a plant model including a notch filter in a discrete system.

6 is a board plot showing the frequency response characteristics of a system actually measured and the frequency response characteristics of a mathematical model according to the present invention.

7 is a block diagram of a simplified servo system in a discrete time system.

8 shows the frequency spectrum of the equivalent position disturbance (d *) estimated according to the present invention.

9 is a flow chart of the present invention.

FIG. 10 is a detailed flowchart illustrating a process of calculating and storing a controller parameter value in a drive in FIG. 9.

11 is a conceptual diagram of a process of obtaining an optimal controller parameter according to the present invention.

12 is a table showing the results according to the invention.

Figure 13 shows the change in the average frequency spectrum according to the present invention.

The present invention relates to a hard disk drive, and more particularly, to a controller design method and apparatus for a hard disk drive.

The servo controller used in the disk drive system performs various operations according to the operation mode of the drive. The most important operation of the servo controller is the track following control operation. This operation ensures the reliability of the read / write operation of the head while following the given track position command within the allowable error range. As is well known, there are various disturbances in a disk drive system, such as disk vibration, wind caused by disk rotation, measurement noise, etc., which disturb the head from following the target track. The goal of the track following controller is to have a robust characteristic against this disturbance.

In servo systems, a head disk assembly (HDA) comprising a voice coil motor (VCM) and a disk is so complex that it is almost impossible to obtain an accurate model for it. Due to this complexity, servo systems for conventional disk drive systems are designed using simplified mathematical models. As such a simplified model is a model that considers the first resonance in the bi-branch or bi-branch.

Figure 1 shows the mathematical model of the conventional servo system and the frequency response characteristics of the actual plant.

As shown in Figure 1, the above simplified model is very different from the actual plant. The resonant mode of HDA, which contributes to these differences, is not only a major factor in making the system unstable, but also unique to each drive or head, which is the biggest limiting factor in designing an optimized controller.

Because of these limitations, conventional controller design methods and apparatus design the servo system to have a sufficient design margin (design margin), but rather there is a problem that the drive cannot realize optimal performance.

SUMMARY OF THE INVENTION The present invention has been made in an effort to provide a controller design method that enables a drive to realize optimal performance while being robust to characteristic variations of each drive and head.

Another object of the present invention is to provide a controller design apparatus to which the above method is applied.

In order to solve the above technical problem, the present invention provides a plurality of plant models (P) for a plurality of head stack assembly (HSA) including a notch filter of the drive in the controller design method of the hard disk drive. Obtaining the equivalent position disturbance d * using the plurality of plant models P and satisfying a predetermined stability standard for the plurality of plant models P and simultaneously calculating a controller parameter value such that the average position tracking error for d *) is minimized.

In order to solve the above other technical problem, the present invention, in the controller design apparatus of a hard disk drive, a plant model calculation unit for calculating a plant model including a notch filter using the input notch filter coefficients and the factors of the HSA, The equivalent position disturbance calculator that calculates the equivalent position disturbance (d *) of the drive using the plant model calculated by the plant model calculator, and the equivalent position disturbance (d *) calculated by the equivalent position disturbance calculator An average position tracking error calculation unit for calculating an average position tracking error by substituting an equation and an average position tracking operation calculated by the average position tracking error calculation unit while satisfying a predetermined stability standard (Ms) for the plurality of plant models (P). Optimal control that calculates controller parameter values that minimize errors and stores them as optimal controller parameters Pre-parameter calculator.

Hereinafter, the configuration of the present invention will be described with reference to the drawings.

2 shows a configuration of a hard disk drive to which the present invention is applied.

The drive 200 includes at least one magnetic disk 220 that is rotated by the spindle motor 210. The drive 200 also includes a head 230 positioned adjacent to the disk 220 surface.

The head 230 can read or write information from the rotating disk 220 by sensing and magnetizing the magnetic field of each disk 220. Typically head 230 is coupled to the surface of each disk 220. Although illustrated and described as a single head 230, it should be understood that this consists of a recording head for magnetizing the disc 220 and a separate reading head for sensing the magnetic field of the disc 220. The read head is constructed from Magneto-Resistive (MR) elements.

Head 230 may be integrated into slider 231. The slider 231 is configured to create an air bearing between the head 230 and the surface of the disk 220. The slider 231 is coupled to the head gimbal assembly 232. Head gimbal assembly 232 is attached to an actuator arm 240 having a voice coil 241. The voice coil 241 is located adjacent to the magnetic assembly 250 that specifies a voice coil motor (VCM) 242. The current supplied to the voice coil 241 generates a torque to rotate the actuator arm 240 relative to the bearing assembly 260. Rotation of the actuator arm 240 will move the head 230 across the disk 220 surface.

The information is typically stored in an annular track of the disc 220. Each track 270 generally includes a plurality of sectors. Each sector includes a data field and an identification field. The identification field is composed of a gray code identifying a sector and a track (cylinder). Head 230 is moved across the surface of disc 220 to read or write information on other tracks.

3 is a block diagram of a servo system to which the present invention is applied.

The dynamic compensator 300 corresponds to a servo controller in a servo system. The servo controller receives a track tracking command for the target position r and a track tracking error e measured by the measured head position ym and outputs a control signal.

On the other hand, a controller used in a conventional disk drive servo system has a plurality of controller coefficients, and a controller C having k coefficients X = [x1, x2, ..., xk] is represented by C (z, X). Can be.

The notch filter 310 receives a control signal and outputs a signal from which the resonance component is removed from the control signal.

The digital-to-analog converter (DAC) 320 converts the control signal into an analog signal and outputs the voice coil motor control signal u.

The power amplifier 330 amplifies the magnitude of the voice coil motor control signal u.

The head disk assembly (HDA) 340 receives the voice coil motor control signal u and moves the head on the disk.

The analog-to-digital converter (ADC) 350 converts the signal fed back from the head disk assembly 340 into a digital signal and outputs the measured head position ym.

Here, w, d and n represent run-out at each position.

4 is a block diagram of a controller design apparatus according to the present invention.

The notch filter coefficients and the HSA factor input unit 400 receive factors necessary for obtaining a plant model P including the notch filter, and transmit the factors to the plant model calculator 410. These factors include various factors related to the current amplifier gain, analog to digital converter gain, digital to analog converter gain, sensor gain, and frequency response of the HSA, including notch filter coefficients.

The plant model calculator 410 calculates a plant model including a notch filter using notch filter coefficients and various factors received from the HSA factor input unit 400. In the present invention, the plant model may be represented by an equation, and serves to simulate an actual frequency response characteristic.

The equivalent position disturbance calculator 420 calculates the equivalent position disturbance d * of the drive using the plant model calculated by the plant model calculator 410 (see Equations 3 and 4).

The average position tracking error calculator 430 calculates the average position tracking error by substituting the equivalent position disturbance d * obtained by the equivalent position disturbance calculator 420 into a predetermined equation (see Equation 7).

The controller parameter variable input unit 435 generates a random controller parameter value and transmits this value to the average position tracking error calculator 430.

The optimal controller parameter calculator 440 calculates the controller parameter input from the controller parameter variable input unit 435 while satisfying a predetermined stability standard for the plurality of plant models P, and at the mean position following error calculator 430. A controller parameter value for minimizing one average position tracking error is calculated and stored as an optimal controller parameter in the memory 450. This optimal controller parameter will be used as the controller parameter in the design of the actual drive.

The disturbance measurement value input unit 460 inputs the disturbance measurement value of the drive measured by the direct measuring instrument so as to obtain an equivalent position disturbance d *. Disturbance measurement value input unit 460 may be provided according to an embodiment of the present invention.

In this case, the plant model P refers to a plurality of head stack assemblies (HSAs) in which the gains of the current (power) amplifiers of the drives, the analog-to-digital converter gains, the digital-to-analog converter gains, the sensor gains, etc. are considered along with the coefficients of the notch filter. It is a mathematical model.

The mechanical characteristics of the HSA and VCM of the hard disk drive are shown in the frequency response characteristics, which can be measured using a Lase Doppler Vibrometer (LDV), etc. These frequency responses can be measured using a mathematical model (e.g. State space model).

The input / output characteristic of the power amplifier for driving the VCM is defined as an output current according to the input residual pressure, and the frequency characteristic can be measured using a dynamic signal analyzer (DSA). Power amplifiers can also be converted into mathematical models using system identification programs. Digital-to-analog converters (DACs) and analog-to-digital converters (ADCs) have sufficient resolution and can be approximated linearly.

Such a mathematical model is represented by the following equations (1) and (2).

는 G(z)를 1/k의 주파수로 다운 샘플링(down sampling)하는 데시메이션(decimation) 연산자를, Z{ }는 z-변환 연산자를, Nm(z) 및 Ns(z)는 각각 다중 샘플링 노치 필터(Multi-rate Notch Filter) 및 1중 샘플링 노치 필터(Single-Rate Notch Filter)를 나타낸다. 또한, Td는 전체 시스템의 시간 지연을 나타낸다.In Equation 1,

Is a decimation operator that downsamples G (z) to a frequency of 1 / k, Z {} is the z-transformation operator, and Nm (z) and Ns (z) are multiple sampling, respectively. A notch filter (Multi-rate Notch Filter) and a single sampling notch filter (Single-Rate Notch Filter) are shown. Td also represents the time delay of the entire system.

In Equation 2, tpa represents the number of tracks per unit angle, Kadc represents the gain of an analog-to-digital converter, and Kdac represents the gain of a digital-to-analog converter. In addition, Pa (s) represents a power amplifier model in a continuous time system, and Ph (s) represents a model of a head stack assembly (HSA) including a VCM in a continuous time system.

This mathematical model must be secured for a sufficient number of drives and heads because the mechanical properties of the HSA are drive and head specific.

Instead of interpreting the notch filter as part of the controller, the present invention uses the plant model with the notch filter applied by interpreting the notch filter as part of the plant. The reason for this is as follows.

Due to the different resonance characteristics of the drive and the head, a multi-sampling programmable notch filter is used in the drive system. However, such a variable notch filter can only be used in a limited manner due to the limitation of the implementation and the increase of the tracking error due to the phase delay, and thus can not eliminate the resonance mode of the HSA. This multi-sampling notch filter must be different from the frequency at which the controller operates. Therefore, in order to accurately analyze the stability of the whole system and to estimate the performance of the controller, a multiple sampling analysis method is required. For this reason, notch filters need to be interpreted as part of the plant.

5 is a board plot showing a process of obtaining a plant model including a notch filter in a discrete system.

Here, Gp (s) shows the frequency response of the plant model in the continuous time system. Gp2 (z) is the plant model for double sampling in the discrete time system, Gp1 (z) is the plant model for single sampling in the discrete time system, MRF (z) is the multiple sampling notch filter, SRF ( z) shows the frequency response of the single sampling notch filter. The frequency response characteristics of applying MRF (z), SRF (z), etc. to the plant model in the discrete time system are shown in Gp (z). As a result, the new plant model including the notch filter is Gp (z).

6 is a board plot showing the frequency response characteristics of a system actually measured and the frequency response characteristics of a mathematical model according to the present invention.

In FIG. 6, the frequency response characteristic of the actual system is the total frequency response characteristic of the plant unit reconstructed using the frequency response characteristic measured from the system. In addition, the frequency response characteristic of the mathematical model (Fitted Model) is the frequency response of the mathematical model of the entire plant part derived from the serial response of the frequency response characteristics of the notch filter, the analog-digital converter, etc. using a system identification program. Characteristic. In Figure 6, the mathematical model according to the present invention depicts the actual system well.

As described above, if a part of the servo system is replaced with a plant model (the mathematical model of the present invention), the servo system of FIG. 3 becomes as shown in FIG. 7.

7 is a block diagram of a simplified servo system in a discrete time system.

The controller 700 receives the measured track following error e by the target position r and the measured head position ym, and receives the control signal as a result of processing the signal according to the controller function C (z). Output

The plant 710 receives the control signal, processes the operation of the notch filter, the current (power) amplifier, the analog-to-digital converter, and the digital-to-analog converter according to the plant function Gp (z), and outputs the result value.

At this time, d * is an equivalent position disturbance, and reflects the influence of the disturbances w, d and n of FIG. 1 to the servo system of FIG. 6 as it is. d * = Gp * w + d + n y represents the actual head position. ym is the measured head position, expressed as ym = y + n.

In this case, in the conventional disk drive servo system, the target position r = 0 may be assumed, and the tracking error e or the measured head position ym may be expressed by Equation 3 below.

Here, S (z) represents a sensitivity function. According to the linear system theory, d * can be calculated from the measured position error (measured head position) ym and the sensitivity function S (z). That is, the equivalent position disturbance d * estimated from the measured position error (measured head position) ym may be expressed by Equation 4 below.

Since the disturbance d and the plant model Gp (z) are different for the plurality of drives and the heads, the equivalent position disturbance d * estimated using the position tracking error (measured head position) ym is also different for each drive and head.

8 shows the frequency spectrum of the equivalent position disturbance d * estimated according to the present invention.

In FIG. 8, among equivalent position disturbances, NRRO represents a non-repetitive disturbance component, and RRO represents a repetitive disturbance component. FIG. 8 shows the mean values of the spectra of equivalent position disturbances (d *) different for each drive and head. As such, the equivalent position disturbance (d *) for obtaining the average position tracking error may be calculated by an equation, or may be calculated using a disturbance measuring device currently used in the art.

9 is a flow chart of the present invention.

First, a frequency response is obtained for a plurality of HSAs, a resonance frequency is detected using the frequency response, and then coefficients of respective notch filters for the plurality of HSAs are calculated to correspond to the detected resonance frequency (step 900).

In operation 910, a plurality of plant models P for a plurality of HSAs are obtained based on the obtained coefficients of the notch filter. At this time, the plurality of plant models (P) is a mathematical model reflecting the characteristics of the HSA different for each drive and head, it can be briefly represented as a set P.

The equivalent position disturbance d * is calculated using the obtained plurality of plant models P (step 920). The procedure for obtaining the plant model P is as described above.

Finally, among the controller parameters satisfying the predetermined stability standard (Ms), a controller parameter value for minimizing the average position tracking error for the equivalent position disturbance (d *) of the drive is calculated and stored in the drive (step 930). ).

In this case, the predetermined stability standard Ms may be determined by a gain margin GM, a phase margin PM, a high frequency vector margin SPH, and the like.

In this case, the high frequency vector margin (SPH) is introduced because the resonance mode of the HSA existing in the high frequency region has a lot of variation for each drive and head. In particular, in a disk drive system, a large amount of gain error and phase error in a low frequency region can be compensated for through various calibration techniques, so it is necessary to consider high frequency vector margin separately.

This high frequency vector margin (SPH) can be represented by the following equation (5).

Where G represents an open loop system and wH represents a critical frequency. In other words, the high frequency vector margin SPH represents the distance between the complex representation value of the frequency response characteristic of the open-loop system G and -1 on the complex plane in the region where the frequency w is greater than wH. This value, according to classical control theory, corresponds to a vector margin in the region where w> wH. The higher this value, the more unstable the system.

The meaning of "controller C (z) is Ms strong and stable with respect to set P of plant models" means set controller P (z, X), set P of plant models consisting of a number of plant models (Gp) and predetermined stability Open-loop system G (z) = Gp (z) * C (z) consisting of any plant model Gp and C (z) belonging to the set P of plant models, for margin standard Ms = [Gm, Pm, Sh] Means that the following Equation 6 is satisfied.

GM (G (z))> Gm

PM (G (z))> Pm

-SPH (G (z)) <-Sh

Where GM is a gain margin, PM is a phase margin, and SPH is a high frequency vector margin. In addition, Gm means gain margin of stability margin standard, Pm means phase margin of stability margin standard, and Sh means high frequency vector margin of stability margin standard.

On the other hand, according to the classical control theory (Parseval's Theorem), given the controller C (z, x), the average position tracking error F (x, P) for the set P of the given plant model and the equivalent position disturbance d is It can be expressed as Equation 7.

Where S (*, x, p) is a sensitivity function for an open-loop system consisting of individual plant model p belonging to P and controller C (z, x), where F (x, P) is the optimal controller coefficient. It is an evaluation function (here, average position tracking error).

FIG. 10 is a detailed flowchart illustrating a process of calculating and storing a controller parameter value in a drive in FIG. 9.

First, the count value k, the value of the optimal parameter xopt, the value of the target position r, and the minimum position tracking error yopt are all initialized (step 1000). As an example, k = 0, the value of the optimal parameter (xopt) is an arbitrary value, r is any value less than zero, and the minimum position tracking error (yopt) is a value very large (greater than 1) (>> 1). Can be set.

After the initialization, the count value k is increased by 1 and the controller parameter x of the k th count is set to an arbitrary value (step 1010). As an example, x = xopt + r * rand () may be obtained, where rand () represents any vector R of the same order as the controller parameter x. That is, the controller parameter may increase in part as the target position increases, but the increase characteristic (amount of increase, etc.) varies depending on an arbitrary vector R.

After setting the controller parameter x to an arbitrary value, it is determined whether the controller C (z, x) satisfies the system stability specification Ms for the plant model P (step 1020). The method of determining whether the controller C (z, x) satisfies the system stability standard Ms is as described above in Equation 6.

At this time, if the system stability standard Ms is not satisfied, the target position r is increased and the process proceeds to step 1060 (step 1040). At this time, if the system stability standard Ms is satisfied, the process proceeds to the following process (1030 process).

The minimum position tracking error (yopt) obtained by the average position tracking error F (x, P) of the kth count to the k-1th count, for the controller parameter that the controller C (z, x) satisfies the system stability standard Ms. Determine whether it is smaller (step 1030).

At this time, if the average position tracking error F (x, P) of the k th count is not smaller than the minimum position tracking error yopt obtained until the k-1 th count, the target position r is increased and the process proceeds to step 1060 (step 1040). . At this time, if the average position tracking error F (x, P) of the kth count is smaller than the minimum position tracking error yopt obtained up to the k-1th count, the controller parameter x of the kth count is optimized parameter (xopt). In this case, the average position tracking error F (x, P) is set to the minimum position tracking error yopt, and the target position r is decreased (step 1050). This is to check whether the minimum position tracking error (yopt) obtained so far corresponds to the minimum position tracking error that the system can have.

Next, it is determined whether the count k is greater than a predetermined number of times (step 1060).

At this time, if the count k is not greater than a predetermined number of times, the process returns to step 1010 and repeats the processes up to now (steps 1060 and 1010). At this time, if the count k is greater than a predetermined number of times, the above-described repetition is completed, and the minimum position tracking error yopt and the optimal parameter value xopt thus far obtained are stored in the drive (step 1070).

Through this process, it is possible to determine a controller parameter for minimizing the position tracking error while satisfying a predetermined stability margin.

By storing the optimum controller parameters as described above in each drive, the optimum controller parameters can be applied during operation of the drive.

11 is a conceptual diagram of a process of obtaining an optimal controller parameter according to the present invention.

In FIG. 11, a process of searching for a controller parameter according to a random vector space R starting from an initial value r0 of the target position and an initial value x0 of the controller parameter and having a minimum position tracking error is illustrated. x1 is a controller parameter after one repetition, x2 is a controller parameter after two repetitions, and xm is a controller parameter after repeating a predetermined number of times (m), and xm is an optimal controller parameter. Of course, such controller parameters are the controller parameters for the plant model that satisfy a certain stability margin, Ms, among the set P of plant models.

12 is a table showing the results according to the invention.

12 shows that the stability margin Ms is set to Gm = 5 dB, Pm = 33 deg, Sph = 2 dB, the predetermined number of repetitions is 200, and the initial value of the target position is set to 0.5 to obtain simulation values and experimental values. The comparison is shown. In particular, it is assumed that the components of the vector rand () are uniformly distributed between -0.5 and 0.5 intervals. In FIG. 12, the position tracking error PES is given as an average% track width.

The simulation results are in close agreement with the actual experimental results. That is, it means that the optimal controller parameter obtained using the equation in the present invention is almost identical to the optimal controller parameter obtained experimentally. Note that the standard deviation of total position tracking error (PES) was reduced by 9% while satisfying all stability margins (GM, PM, SPH, etc.).

Figure 13 shows the change in the average frequency spectrum according to the present invention.

In FIG. 13, looking at the frequency spectrum of the position tracking error (PES) after optimization, the size is reduced in particular in the low frequency region than before optimization, mainly due to the increase in bandwidth. Of course, stability margins have been reduced to a negligible extent.

In the present invention, the controller parameter is obtained using the average position tracking error, but the controller parameter may be obtained using another evaluation function.

In the present invention, a controller parameter for minimizing the average position tracking error is obtained, but a controller parameter for varying the controller parameters for a plurality of plant models and not generating a maximum value of the position tracking error can be calculated as an optimal controller parameter. .

Although the present invention has been described with reference to one embodiment shown in the drawings, this is merely exemplary and will be understood by those of ordinary skill in the art that various modifications and variations can be made therefrom. However, such modifications should be considered to be within the technical protection scope of the present invention. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

As described above, according to the present invention, a set P of a plant model including a notch filter is obtained for a plurality of HSAs, and a position obtained from the plant model among controller coefficients satisfying a predetermined stability margin Ms in the set P. By calculating a controller coefficient that minimizes the tracking error, there is an effect that the drive has stable characteristics against drive and head-specific characteristic variations while having optimal performance.

Claims (16)

In the controller design method of a hard disk drive, Obtaining a plurality of plant models (P) for a plurality of head stack assemblies (HSAs) including a notch filter of the drive; Obtaining an equivalent position disturbance d * using the plurality of plant models P; And Calculating a controller parameter value that satisfies a predetermined stability standard for the plurality of plant models (P) and at the same time minimizes an average position tracking error for an equivalent position disturbance (d *) of the drive. Controller design method characterized by. The method of claim 1, Obtaining the plurality of plant models (P) Calculating a plurality of plant models (P) for the plurality of head stack assemblies (HSAs) by considering the gain of the current amplifier, the analog-to-digital converter gain, and the digital-to-analog converter gain of the drive together with the coefficients of the notch filter. Controller design method characterized by. The method of claim 1, A controller parameter that meets a certain stability margin Each of the gain margins, phase margins, and high frequency vector margins among the plurality of plant models P is included in a range of a predetermined gain margin, a predetermined phase margin, and a predetermined high frequency vector margin, which is the stability standard Ms of the drive. Controller design method, characterized in that. The method of claim 3, wherein The high frequency vector margin is

And G is the product of the plant model Gp (z) and the controller C (z). The method of claim 1, The average position tracking error is

Represented by the formula

Is a sensitivity function of the open loop system consisting of the individual plant model p belonging to the set P of the plurality of plant models P and the controller C (z, x),

The controller design method, characterized in that the equivalent position disturbance (d *) of the drive. The method of claim 1, Designing a controller by calculating the controller parameter A predetermined number of times of calculating a controller parameter value that satisfies a predetermined stability standard for the plurality of plant models (P) and minimizes an average position tracking error for the equivalent position disturbance (d *) of the drive. Iteratively repeating as much as the controller design method characterized in that the process of calculating the controller parameter value that generates the minimum average position tracking error as the optimal controller parameter. The method of claim 6, The process of calculating the optimal controller parameter And repeating the process of calculating an average position tracking error by a predetermined number of times while randomly changing the controller parameter. In the controller design method of a hard disk drive, Obtaining a plurality of plant models (P) for a plurality of head stack assemblies (HSAs) including a notch filter of the drive; Obtaining an equivalent position disturbance d * of the drive by measuring disturbance d for each position of the drive; And Calculating a controller parameter value that satisfies a predetermined stability standard for the plurality of plant models (P) and at the same time minimizes an average position tracking error for an equivalent position disturbance (d *) of the drive. Controller design method characterized by. In the controller design of the hard disk drive, A plant model calculator configured to calculate a plant model including a notch filter using the input notch filter coefficients and factors of the HSA; An equivalent position disturbance calculator for calculating an equivalent position disturbance (d *) of the drive by using the plant model calculated by the plant model calculator; An average position tracking error calculator which calculates an average position tracking error by substituting an equivalent position disturbance (d *) calculated by the equivalent position disturbance calculator into a predetermined equation; And A controller parameter value that satisfies a predetermined stability standard (Ms) for the plurality of plant models (P) and minimizes an average position tracking error calculated by the average position tracking error calculator is stored as an optimal controller parameter. And a controller controller for calculating an optimum controller parameter. The method of claim 9, The plant model calculation unit A plurality of plant models P for the plurality of head stack assemblies HSA are obtained by considering the gain of the current amplifier of the drive, the analog-to-digital converter gain, and the digital-to-analog converter gain together with the coefficients of the notch filter. Controller design device. The method of claim 9, The optimal controller parameter calculator Each of the gain margins, phase margins, and high frequency vector margins for the plurality of plant models P includes a range of predetermined gain margins, predetermined phase margins, and predetermined high frequency vector margins, the stability specification of the drive being Ms. And a controller parameter value for minimizing the average position following error calculated by the average position following error calculating unit among the controller parameters. The method of claim 11, The high frequency vector margin is

Wherein the G (open loop system) is represented by the product of the plant model Gp (z) and the controller C (z). The method of claim 9, The average position tracking error is

Represented by the formula

Is a sensitivity function of the open-loop system consisting of the individual plant model P belonging to the set P of the plurality of plant models P and the controller C (z, x).

Controller design apparatus, characterized in that the equivalent position disturbance of the drive. The method of claim 9, The optimal controller parameter calculator While satisfying a predetermined stability standard among the plurality of plant models (P), the process of calculating the average position tracking error for the equivalent position disturbance of the drive is repeated a predetermined number of times to generate the minimum average position tracking error. Controller design apparatus, characterized in that for calculating the controller parameter value as the optimal controller parameter. The method of claim 14, The controller design unit And calculating the average position tracking error by a predetermined number of times while randomly changing the controller parameter. In the controller design of the hard disk drive, A plant model calculator configured to calculate a plant model including a notch filter using the input notch filter coefficients and factors of the HSA; An equivalent position disturbance calculator for calculating an equivalent position disturbance d * by receiving a disturbance measurement value d directly measuring the disturbance of the drive; An average position tracking error calculator which calculates an average position tracking error by substituting an equivalent position disturbance (d *) calculated by the equivalent position disturbance calculator into a predetermined equation; And A controller parameter value that satisfies a predetermined stability standard (Ms) for the plurality of plant models (P) and minimizes an average position tracking error calculated by the average position tracking error calculator is stored as an optimal controller parameter. And a controller controller for calculating an optimum controller parameter.

Images