JP2003280703A - Controller using estimator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To cancel much disturbance related to a controlled system without affecting disturbance in the other frequencies. <P>SOLUTION: A robot (driving device) 12 which suffers from disturbance accompanied with a weaving operation fluctuated with a frequency ω is controlled by a controller 11 using a periodic disturbance estimating observer (an estimator) 16 which estimates a weaving frequency ω to perform feedback control, wherein the periodic disturbance estimating observer 16 provides an oscillation characteristic oscillating with the disturbance fluctuation frequency ωas a disturbance characteristic, estimates the disturbance based on the disturbance characteristic and incorporates a band pass characteristic passing the disturbance fluctuation frequency ω. Alternatively, the periodic disturbance estimating observer 16 provides the oscillation characteristic oscillating with the disturbance fluctuation frequency ω as the disturbance characteristic, estimates the disturbance based on the disturbance characteristic and incorporates an attenuating fluctuation characteristic. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device using an estimation device.

[0002]

2. Description of the Related Art A periodic disturbance estimation observer is used as an estimation device for estimating a disturbance in a controlled object that is subject to a disturbance that fluctuates at a frequency ω. The conventional periodic disturbance estimation observer is applied to structures such as buildings. Here, the application of the periodic disturbance estimation observer to the structure means that the disturbance due to the natural vibration of the structure is estimated by the periodic disturbance estimation observer. At this time, since the disturbance of the structure basically only has the disturbance of the natural vibration frequency, it can be estimated by the periodic disturbance estimation observer,
The effects on disturbances at other frequencies were hardly considered.

[0003]

The above-described periodic disturbance estimation observer has hardly been applied to a drive device. The reason for this is that, in addition to the disturbance of a specific frequency, the drive device also has disturbances of various other frequencies. That is, since the conventional periodic disturbance estimation observer has a structure specialized for estimating the disturbance of a specific frequency, when it is applied to a drive device, it affects the disturbance of another frequency.

For example, FIG. 5 is a gain diagram of an estimated value with respect to a disturbance when a pole of the periodic disturbance estimation observer is given with a multiple root of 0.1ω for a specific disturbance frequency ω.
As shown in FIG. 5, the disturbance and the estimated value gain match at the frequency ω, but the gain is largely estimated at the low frequency. Further, if the pole is further lowered, the gain at low frequencies becomes large, which affects low frequency disturbance.

FIG. 6 is a gain diagram when a pole is given by a multiple root of 10ω. Similar to FIG. 5, as shown in FIG. 6, the disturbance and the estimated value gain match at the frequency ω, but the gain is largely estimated at the high frequency. Further, if the poles are further raised, the gain at high frequencies becomes large, which affects the disturbance at high frequencies.

As a specific example, when considering a welding robot that performs a weaving operation in which the torch is oscillated at a weaving frequency as a control target, the oscillation amplitude of the torch in this weaving operation affects the quality of the welding beat, and therefore high accuracy is required. It is necessary to secure it. In the past, feedforward control has attempted to ensure accuracy by using a normal period disturbance estimation observer, etc., but it is not enough to reduce accuracy due to disturbance factors such as machine error and friction, or to generate vibration due to excitation of natural vibration. The accuracy is not secured.
When the weaving frequency is high, the disturbance associated with the weaving operation is dominated by the modeling error of the inertial force, and most of the frequency components of the disturbance are occupied by the frequency of the weaving frequency. Therefore, in order to suppress the disturbance that fluctuates at a high weaving frequency, it is usually necessary to widen the control band to a high frequency, but in that case, in addition to the disturbance, the natural vibration of the mechanism part is excited to cause vibration. I will end up.

As a prior art in which a periodic disturbance estimation observer is applied to the weaving operation of a driving device which is a welding robot, there is JP-A-7-185817. However, Japanese Patent Laid-Open No. 7-185817 does not consider the influence of other frequencies on the disturbance, and is not practically sufficient.

The present invention has been made in view of the above problems, and has a small influence on disturbances of other frequencies, and a control device using an estimation device capable of canceling out most of the disturbances related to the controlled object. Is provided.

[0009]

In order to solve the above-mentioned problems, a control device using an estimation device according to claim 1 of the present invention is an estimation device for estimating a disturbance related to a controlled object and performing feedback control. In the control device using, the control target is a drive device that receives a disturbance that fluctuates at a frequency ω, and the estimation device provides a vibration characteristic that vibrates at a disturbance fluctuation frequency ω as a disturbance fluctuation characteristic, The disturbance is estimated based on the disturbance fluctuation characteristic, and the disturbance estimation characteristic of the estimation device includes a bandpass characteristic that allows the disturbance fluctuation frequency ω to pass.

According to the structure of claim 1, the disturbance estimation characteristic becomes the bandpass characteristic having the frequency ω as the center frequency, and it is possible to estimate not only the disturbance of the frequency ω but also the disturbance of the frequency around it. become.

A control device using the estimation device according to a second aspect of the present invention is a control device using the estimation device according to the first aspect, wherein the band-pass characteristic includes the disturbance and the disturbance. There is a frequency band centered on the disturbance fluctuation frequency ω within which the gain with the estimated value is within ± 10 db, the gain is negative at frequencies other than the frequency band, and the frequency band is different from the disturbance. It does not include the frequency of the natural vibration of the controlled mechanical part that is excited by the control.

According to the second aspect of the invention, the frequency band does not include the frequency of the natural vibration of the mechanical part to be controlled, which is excited in accordance with the control, in addition to the disturbance. With almost no impact,
Most of the disturbance related to the controlled object can be offset.

A control device using an estimation device according to a third aspect of the present invention is a control device using an estimation device for estimating a disturbance related to a controlled object and performing feedback control, wherein the controlled object is , A drive device that receives a disturbance that fluctuates at a frequency ω, wherein the estimation device provides a vibration characteristic that vibrates at a disturbance fluctuation frequency ω as a disturbance fluctuation characteristic, estimates the disturbance based on the disturbance fluctuation characteristic, and The disturbance estimation characteristic of the estimation device is characterized by including a damping vibration characteristic that allows the disturbance fluctuation frequency ω to pass.

According to the structure of the third aspect, the disturbance estimation characteristic becomes the damping vibration characteristic for estimating only the frequency ω, and it becomes possible to estimate only the disturbance of the frequency ω.

A control device using the estimation device according to claim 4 of the present invention is a control device using the estimation device according to claim 3, wherein the damping vibration characteristic is the disturbance and the disturbance. There is an inverse notch type frequency band centered on the disturbance fluctuation frequency ω within which the gain with the estimated value is within ± 10 db, and the gain at frequencies other than the frequency band is negative, and the frequency band is Apart from the disturbance, it does not include the frequency of the natural vibration of the mechanical part to be controlled, which is excited by the control.
It is characterized by

According to the structure of the fourth aspect, the frequency band does not include the frequency of the natural vibration of the mechanical part of the controlled object which is excited in accordance with the control, in addition to the disturbance. Therefore, by estimating only the disturbance of the frequency ω, it is possible to deal with the case where the disturbance frequency is near the frequency of the mechanism natural vibration.

A control device using the estimation device according to claim 5 of the present invention is a control device using the estimation device according to any one of claims 1 to 4, wherein the estimation device has a frequency n. A feature is that a disturbance vibrating at × ω is estimated.

According to the configuration of the fifth aspect, the disturbance of the frequency n × ω is estimated with respect to the frequency ω of the disturbance, so that it is possible to realize more accurate control.

A control device using the estimation device according to claim 6 of the present invention is a control device using the estimation device according to any one of claims 1 to 5, wherein the drive device is a speed reducer. Is a driving device in which the driving force of the actuator is transmitted to the driven body via the actuator, and the speed reducer acts as a spring element. The estimating device includes an actuator, an inertial part associated with the actuator, the driven body, and the driven body. A feature is that a two-inertia system model composed of an inertial part attached to the body is used for estimation as a dynamic characteristic model of a controlled object.

According to the sixth aspect of the present invention, the influence of elastic deformation of the spring element such as the speed reducer can be incorporated as a model, and the disturbance can be estimated and compensated with higher accuracy.

A control device using the estimation device according to claim 7 of the present invention is a control device using the estimation device according to any one of claims 1 to 6, wherein the drive device is a welding robot. And the disturbance is a disturbance associated with the weaving operation of the welding robot, and the estimation device gives the weaving frequency as the disturbance fluctuation frequency ω.

According to the structure of claim 7, when the weaving frequency is high, the disturbance frequency component of the weaving frequency ω occupies most of the disturbance frequency component, so that an estimating device for estimating the disturbance of the frequency ω is constructed. As a result, it is possible to completely cancel each other without adversely affecting other frequency components, and it is possible to realize a high weaving amplitude. When the weaving frequency is slow, the frequency component of the disturbance is the weaving frequency ω
In addition to many disturbance components of frequency n × ω,
By configuring an estimation device that estimates a disturbance of frequency n × ω, it is possible to realize more accurate control.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION A control device using an estimation device according to the present invention will be described below with reference to the drawings.
It should be noted that the embodiment of the control device using the estimation device according to the present invention is not limited to the following embodiment, and various design changes are possible within the scope of the claims.

[Embodiment 1] A first embodiment of the present invention will be described below. In the first embodiment, a rigid robot is assumed as the control target, and an application example to the control target represented by the following equation (Equation 1) is shown as the dynamic characteristic model of the control target.

[0025]

[Equation 1]

Here, J is the inertia of the robot, τ is the input torque, θ is the position of the robot, and d is the disturbance.

As a vibration characteristic that fluctuates at the disturbance fluctuation frequency ω, an observer (estimation device) that estimates the disturbance d by assuming a dynamic characteristic model represented by the following equation (Equation 2) as the disturbance d
Make up.

[0028]

[Equation 2]

Here, the observer constructed on the basis of the dynamic characteristic model of the disturbance d is an estimating device for estimating the disturbance. Specifically, the observer is designed so as to satisfy the observer configuration condition by using the controlled object model (Equation 1) and the disturbance dynamic characteristic model (Equation 2). For example, the observer poles are specified by the following equation (Equation 3).

[0030]

[Equation 3]

That is, the dynamic characteristic of the observer is calculated by the following equation (Equation 4).
Give in. Here, Ω is a parameter for changing the dynamic characteristics of the disturbance.

[0032]

[Equation 4]

From the above, the actual disturbance d of this observer
To the estimated disturbance value d ′ is given by the following equation (Equation 5).

[0034]

[Equation 5]

Further, a gain diagram from the actual disturbance d of this observer to the disturbance estimated value d'is shown in FIG. In addition,
In FIG. 1, the x line shows the case where the parameter Ω is given as 20ω, the ◯ line shows the case where the parameter Ω is given as 2ω, and the − line shows the case where the parameter Ω is given as 0.2ω.

Therefore, in the above dynamic characteristic (Equation 5),
When Ω ≧ ω, the gain is substantially zero (gain is within ± 10 db), the band to be passed is the frequency band from 2ω-Ω to Ω, and band pass (band pass) characteristics are obtained. For example, when the parameter Ω is given as 20ω, the gain diagram is as shown by the x line in FIG. 1, and the disturbance estimated dynamic characteristic (dynamic characteristic from the disturbance to the estimated value) is the frequency ω.
It is obtained with a bandpass characteristic whose center frequency is. Also,
As shown in FIG. 1, the gain is negative at frequencies other than the frequency band where the gain is substantially zero. Furthermore, FIG.
As shown in, the frequency band of substantially zero gain does not include the frequency of the natural vibration of the mechanical section. Here, the mechanical section natural vibration is different from the disturbance, and refers to the vibration generated by exciting the natural vibration of the mechanical section when the control band is expanded to a high frequency during high frequency weaving operation. Say.

In the above dynamic characteristics (Equation 5), Ω
In the case of <ω, the damping vibration characteristic allows only the frequency ω to pass, and as Ω becomes smaller, the gain becomes substantially zero (gain is within ± 10 db), the band to pass becomes a narrow notch type, and other than ω The cutoff characteristic of the frequency band of is improved. For example, if the parameter Ω is given as 0.2ω, the gain diagram will be as shown by the-line in FIG. 1, and the disturbance estimation dynamic characteristic can be obtained by the damping vibration characteristic capable of estimating only the frequency ω. Further, as shown in FIG. 1, the gain is negative at frequencies other than the frequency band (frequency ω) where the gain is substantially zero. Furthermore, as shown in FIG.
The frequency band (frequency ω) where the gain is substantially zero does not include the frequency of the natural vibration of the mechanical section.

Although the observer is designed by designating the observer poles in the above embodiment, the observer design is not limited to this. There are various methods for designing a general bandpass filter. For example, as another method, the following equation (Equation 6) is obtained from the dynamic characteristic model of the controlled object.

[0039]

[Equation 6]

D is calculated by the equation 6, and an estimated value d'is obtained by using a general bandpass filter Bd (z) in the discrete system or a general bandpass filter Bc (s) in the continuous system. Can be estimated by the following Equation 7 or Equation 8.

[0041]

[Equation 7]

[0042]

[Equation 8]

As described above, according to the present embodiment, when Ω is given large, the estimated characteristic becomes the bandpass characteristic,
It is possible to estimate not only the disturbance of the frequency ω but also the disturbance of the frequencies around it. Further, when Ω is given small, the estimation characteristic becomes the damping vibration characteristic, and it becomes possible to estimate only the disturbance of the frequency ω.

On the other hand, focusing on the disturbance of the frequency ω, the deviation of the amplitude of the estimated value with respect to the amplitude of the actual disturbance, that is, the time required to estimate the amplitude of the disturbance is increased by increasing Ω, It becomes slower by making Ω smaller. That is, when Ω is increased, the frequency band in which the gain is substantially zero becomes wider, and it is possible to cope with a plurality of disturbances, and the amplitude can be estimated in a short time. Further, if Ω is made small, the frequency band where the gain is substantially zero becomes narrow, and it is possible to cope with the case where there is a disturbance frequency in the vicinity of the frequency of the natural vibration of the mechanism section, but it takes time to estimate the amplitude.

[Second Embodiment] A second embodiment of the present invention will be described below. The description of the same mathematical expressions as those in the first embodiment will be omitted.

In the second embodiment, as an object to be controlled, an actuator unit (motor) 2 as shown in FIG. 2 is used.
And a driven device (robot) 1 in which a driven body part (arm) 3 is connected to a spring element such as a speed reducer (spring) 4.
The dynamic characteristic model of the controlled object is composed of two inertias: the inertia of the actuator section (inertia associated with the actuator and the actuator) and the inertia of the driven body section (inertia associated with the driven body and the driven body). An example of application to a controlled object represented by the following equation (Equation 9) modeled by a two-inertia system will be shown.

[0047]

[Equation 9]

Here, Jl is the inertia of the driven body part (arm), and Jm is the inertia of the actuator part (motor). Further, θl is the position of the driven body portion, and θm is the position of the actuator portion. Furthermore, k is the spring constant of the spring element. c is a force that acts on the driven body portion from the outside, and indicates interference on each axis of the robot, a gravitational term, and the like, which is changed / deleted according to the target control target.

Based on the equations (9) and (2), as in the first embodiment, the dynamic characteristics of the observer poles of the equation (3), that is, the observer of the equation (4) are given to design the observer so as to satisfy the observer constituent conditions. .

As described above, according to this embodiment, by replacing the dynamic characteristic model of the two-inertia system with the dynamic characteristic model of the two-inertia system, the influence of elastic deformation of the spring element such as the speed reducer can be taken in as a model, and the disturbance can be more accurately performed. And then
It becomes possible to compensate.

[Embodiment 3] A third embodiment of the present invention will be described below. The description of the same mathematical expressions as those in the first embodiment will be omitted.

In the third embodiment, the control target is
Assume a welding robot that performs a weaving operation in which a torch is swung at a weaving frequency. FIG. 3 is a block diagram of a control device using the estimation device. Also, FIG.
3 is a flowchart of a control device using the estimation device.

First, a block diagram of a control device using the estimation device will be described with reference to FIG. As shown in FIG. 3, the control device 11 controls the robot 1 to be controlled.
2, which is a device for controlling 2, and includes a position / speed control unit 13, a dynamics compensation unit 14, an elastic deformation compensation unit 15, and a periodic disturbance estimation observer (estimation device) 16.

In the position / speed control unit 13, for example, PID
The control is performed, and the torque as a target value is calculated based on the position / speed error. The dynamic compensator 14 is for compensating for the delay in the response of the position / speed controller 13, and calculates the torque as a necessary value based on the dynamic model. Further, the elastic deformation compensator 15 calculates the torque applied to the spring when there is a spring. The periodic disturbance estimation observer 16 is designed by the method described in the first or second embodiment, and the robot 1
The torque as the estimated disturbance value is calculated from the velocity of 2, the weaving cycle ω that is the disturbance fluctuation frequency, and the parameter Ω calculated from the weaving cycle ω. In addition,
The periodic disturbance estimation observer 16 gives the poles of the observer of the equation 3 or the dynamic characteristics of the observer of the equation 4 based on the dynamic characteristic model of the controlled object and the equation 2 as in the first or second embodiment. Design to meet the observer configuration conditions.

Next, the operation of the control device using the estimation device will be described with reference to the flowchart of FIG.

First, when the control is started, step S1
At, the weaving period ω and the parameter Ω are set. Also, the state variables of the periodic disturbance estimation observer are initialized.

Next, in step S2, the estimated disturbance value (torque) is calculated by the periodic disturbance estimation observer, and in step S3, the estimated disturbance value is added to the torque.
Then, in step S4, it is confirmed whether or not the control is stopped.

To continue the control (step S4: N
O), and in step S5, it is confirmed whether or not the disturbance cycle is changed. When changing the disturbance period (step S
5: YES), returning to step S1, the weaving period ω and the parameter Ω are set. Also, the state variables of the periodic disturbance estimation observer are initialized. On the other hand, when the disturbance cycle is not changed (step S5: NO), step S
Returning to 2, the estimated disturbance value is calculated by the periodic disturbance estimation observer.

To stop the control (step S4: Y
ES), in step S6, the setting of the weaving period ω and the parameter Ω is cleared. Then, the state variable of the periodic disturbance estimation observer is initialized and the control is stopped.

As described above, according to the present embodiment, the dynamic characteristic of the disturbance is changed according to the weaving frequency, and if necessary, the dynamic characteristic of the periodic disturbance estimation observer is also changed according to the weaving frequency to estimate the disturbance. In addition, by feeding back the estimated value, it is possible to completely cancel the disturbance of the frequency ω without adversely affecting the disturbance of other frequencies.

When the weaving frequency is high, the disturbance associated with the weaving operation is dominated by the modeling error of the inertial force, and most of the frequency components of the disturbance are occupied by the weaving frequency. Therefore, the disturbance of the weaving frequency that occupies most of the frequency component of the disturbance can be completely canceled without adversely affecting other frequencies, and a high weaving amplitude can be realized.

When the weaving frequency is slow, the disturbance due to the weaving operation is dominated by the influence of friction, etc., and the frequency component of the disturbance has a frequency n × ω (n = 2, 3) in addition to the weaving frequency ω. There are many disturbance components (...). Therefore, if the periodic disturbance estimation observer 16 that estimates the disturbance having the frequency n × ω with respect to the disturbance fluctuation frequency ω is configured in the same manner as described above, it is possible to realize more accurate control.

[0063]

According to the control device using the estimation device according to the first aspect of the present invention, by including the bandpass characteristic in the dynamic characteristic of the estimation device, not only the disturbance of the frequency ω but also the frequency of the surrounding frequency It is possible to estimate the disturbance as well. According to the control device using the estimation device according to claim 2 of the present invention, the frequency band does not include the frequency of the natural vibration of the mechanical part of the control target that is excited in accordance with the control, in addition to the disturbance. It is possible to cancel many disturbances related to the controlled object with almost no influence on the frequency of the natural vibration of the part.

According to the control device using the estimating device according to the third aspect of the present invention, by including the damping vibration characteristic in the dynamic characteristic of the estimating device, it is possible to estimate only the disturbance of the frequency ω, Even if there is a disturbance frequency near the natural vibration frequency, it is possible to deal with it. Claim 4 of the present invention
According to the control device using the estimation device described in (1), the frequency band does not include the frequency of the natural vibration of the mechanical part to be controlled, which is excited in accordance with the control, in addition to the disturbance. Therefore, by estimating only the disturbance of the frequency ω, it is possible to deal with the case where the disturbance frequency is near the frequency of the mechanism natural vibration.

According to the control device using the estimation device of the fifth aspect of the present invention, the estimation device is configured to estimate the disturbance of frequency n × ω with respect to the frequency ω of the disturbance. Highly precise control can be realized.

According to the controller using the estimating apparatus of the sixth aspect of the present invention, the driving force of the actuator is transmitted to the driven body through the reduction gear, and the reduction gear acts as a spring element. On the other hand, a two-inertia system model including an actuator, an inertial part associated with the actuator, and a driven body and an inertial part associated with the driven body should be used as a dynamic characteristic model of the controlled object for estimation of the estimation device. As a result, the influence of elastic deformation of the spring element such as the speed reducer can be incorporated as a model, and the disturbance can be estimated and compensated with higher accuracy.

According to the control device using the estimating device according to the seventh aspect of the present invention, the weaving frequency is used as the disturbance fluctuation frequency ω for estimating the estimating device with respect to the disturbance caused by the weaving operation of the welding robot as the driving device. As a result, when the weaving frequency is fast, the disturbance frequency component of the weaving frequency ω occupies most of the disturbance frequency component, so that by constructing an estimation device that estimates the disturbance of the frequency ω, it adversely affects other frequency components. It is possible to completely cancel each other without giving, and a high weaving amplitude can be realized. Further, when the weaving frequency is slow, the disturbance frequency component has a large number of disturbance components of frequency n × ω in addition to the weaving frequency ω. Therefore, an estimation device for estimating the disturbance of frequency n × ω should be configured. With this, it is possible to realize more accurate control.

[Brief description of drawings]

FIG. 1 is an observer (estimation device) according to a first embodiment.
FIG. 6 is a gain diagram from the actual disturbance d to the estimated disturbance value d ′.

FIG. 2 is a schematic diagram of a drive device (robot) in which an actuator unit (motor) and a driven body unit (arm) according to a second embodiment are connected to a spring element such as a speed reducer (spring).

FIG. 3 is a block diagram of a control device using an estimation device according to a third embodiment.

FIG. 4 is a flowchart of a control device using the estimation device according to the third embodiment.

FIG. 5 is a gain diagram of an estimated value for a disturbance when a pole of the periodic disturbance estimation observer is given by a multiple root of 0.1ω in the conventional periodic disturbance estimation observer.

FIG. 6 is a gain diagram when a pole of the periodic disturbance estimation observer is given by a multiple root of 10Ω in the conventional periodic disturbance estimation observer.

[Explanation of symbols]

1 drive 2 Actuator part 3 Driven body part 4 reducer 11 Control device 12 Robot (driving device) 16 Periodic disturbance estimation observer (estimator)

Continued front page    F-term (reference) 3C007 AS11 KS21 KS22 KV01 LT13                       LV23 LW05 MT05                 5H004 GA07 GB16 HA07 HB07 JB22                       KB02 KB04 KB06 LA13 MA14                       MA15                 5H303 AA10 BB03 BB07 CC01 DD01                       EE03 EE07 FF03 JJ02 KK02                       KK03 KK04 KK11 MM05

Claims (7)

[Claims] 1. A control device using an estimation device that estimates a disturbance related to a controlled object and performs feedback control, wherein the controlled object is a drive device that receives a disturbance that fluctuates at a frequency ω, The estimation device uses the disturbance fluctuation frequency ω as the disturbance characteristic.
A disturbance fluctuation frequency ω as a disturbance estimation characteristic of the estimation device.
A control device using an estimation device, characterized in that it includes a band-pass characteristic for passing the. 2. The bandpass characteristic has a gain of ± 1 between the disturbance and an estimated value of the disturbance.
There is a frequency band centered on the disturbance fluctuation frequency ω of 0 db or less, and the gain at frequencies other than the frequency band is negative, and the frequency band is a control target that is excited in accordance with the control separately from the disturbance. The control device using the estimation device according to claim 1, wherein the control device does not include the frequency of the natural vibration of the mechanism section. 3. A control device using an estimation device that estimates a disturbance related to a controlled object and performs feedback control, wherein the controlled object is a driving device that receives a disturbance that fluctuates at a frequency ω, The estimation device uses the disturbance fluctuation frequency ω as the disturbance characteristic.
A disturbance fluctuation frequency ω as a disturbance estimation characteristic of the estimation device.
A control device using an estimation device, characterized in that it includes a damped vibration characteristic for passing through. 4. The damping vibration characteristic has an inverse notch type frequency band centered on a disturbance fluctuation frequency ω within which the gain of the disturbance and an estimated value of the disturbance is within ± 10 db. The gain at frequencies other than is negative, and the frequency band does not include the frequency of the mechanism-specific natural vibration of the control target that is excited along with the control, in addition to the disturbance.
A control device using the estimation device according to claim 3. 5. The control device using the estimation device according to claim 1, wherein the estimation device estimates a disturbance vibrating at a frequency of n × ω. 6. The driving device is a driving device in which a driving force of an actuator is transmitted to a driven body through a speed reducer, and the speed reducer acts as a spring element, and the estimating device includes an actuator and an actuator. 6. A two-inertia system model composed of an associated inertial part and a driven body and an inertial part associated with the driven body is used for estimation as a dynamic characteristic model of a controlled object. A control device using the estimation device according to claim 1. 7. The driving device is a welding robot, the disturbance is a disturbance associated with a weaving operation of the welding robot, and the estimation device gives a weaving frequency as a disturbance fluctuation frequency ω. A control device using the estimation device according to any one of items 1 to 6.