JPH07121209A - Control unit using sliding mode control system - Google Patents

Abstract

PURPOSE:To provide the control unit which uses the sliding mode control system that can sufficiently suppress chattering of control when a controlled system is controlled by using the sliding mode control system. CONSTITUTION:This unit A1 is so constituted as to give dynamic characteristics N(s) compensating the delay of a switching time by parasitic element presents in the input and output of the controlled system 1 to a switching variable S specifying a control slide surface when the sliding mode control system which switches and outputs a control command value (u) to the controlled system 1 is used so as to converge the control state of the controlled system 1 on the control slide surface to approximates an error in controlled variable (x) to zero. The constitution can sufficiently suppress the chattering of the control.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sliding system for performing robust control by introducing a switching operation into an output to a controlled object including a non-linear element having a large parameter variation such as a motor for driving each axis of a robot. The present invention relates to a mode control system, and more particularly to a control method using a sliding mode control system capable of suppressing control chattering.

[0002]

2. Description of the Related Art For example, when driving a robot, etc.
The load inertia applied to the motor and the nonlinear term vary greatly.
Therefore, depending on the PID control, perform highly accurate control.
I can't. Therefore, the sliding mode as described above
The mode control system is applied. General as above
Conventional controller A using a sliding mode control system
₀₁, A₀₂, A₀₃Are shown in FIGS. 7, 8 and 9. Shown in FIG.
Control using a general sliding mode control system
Device A by method₀₁Then, usually the secondary system is controlled
The motor system expressed by the following equation
11 is set as a control target. Jθ ″ + C = τ + D (1 ′) where J: inertia (in the case of a robot, it changes depending on the posture)
Θ: Rotation angle of motor τ: Torque C: Non-linear term such as friction / viscosity / centrifugal / Coriolis / gravity term D: Disturbance term In the following explanation, ′ attached to the right shoulder of each symbol is
First-order time derivative of the symbol, ″ is second-order time derivative, ^
Represents the estimated value. The above conventional control device A₀₁so
Is that of the rotation angle θ and the rotation angular velocity θ ′ as the controlled variables.
Target value θ_d, Θ_dThe error e, e ′ for ′ to 0
The sliding mode control system 12 described above
Is the torque τ which is the control command value to the motor system 11.
The calculation value for calculation is switched and output.
ing. The conventional control device A₀₁Is the nonlinear compensation system 1
According to the estimated output calculated by 0, the above sliding
Control command value from the control mode 12
It is designed to further improve the degree.

The sliding mode control system 12 is
Even if the nonlinear term C and the disturbance term D change suddenly, these fluctuations are suppressed and the above errors e and e'can be made zero. Here, the error system for the motor system 11 can be expressed as Je ″ = τ−Jθ _d ″ −C + D (2 ′). Now, when the torque τ is divided into the control input ν from the sliding mode control system 12 and the control input ξ from the other control system (for example, the nonlinear compensation system 10), τ = ν + ξ (3 ') is obtained. Accordingly, the error system of the equation (2 ′) is described as Je ″ = ν + E (4 ′) E = ξ−Jθ _d ″ −C + D (5 ′). Here, the control input ξ is (−Jθ _d ″
Considering the compensation input for −C + D), the above (5 ′)
E in the equation can be considered as a compensation error that cannot be completely compensated by the control input ξ. Also, as in the following equation,
The compensation error E is the term E _P e, E _V relating to the errors e, e '.
It is possible to separate e'and the irrelevant term E _C. E = E _C + E _P e + E _V e ′ (6 ′) Substituting this formula (6 ′) into the above formula (4 ′) gives Je ″ = ν + E _C + E _P e + E _V e ′ (7 ′). The control input ν of the sliding mode control system 12 is determined so that the error system of the equation (7 ′) is asymptotically stabilized. Here, the control output to the motor system 11 is The switching variable S for switching is defined as S = e '+ λe (8') as follows, where λ is a positive number that determines the convergence characteristic of the error e. And sliding mode control System 12
The control input ν from is calculated as follows. ν = − (K _C + K _P | e | + K _V | e ′ |) · sgn (S) ... (9 ′) where sgn (S) returns a value of 1 if S ≧ 0, and S
It is a sign function that returns a value of -1 if ≤0.

Further, K _C , K _P and K _V are control gains set for the compensation error coefficients E _C , E _P and E _V , respectively. Therefore, as function V with a candidate Lyapunov function used for stability determination method of Lyapunov known as a technique for determining the stability of the system is of the ^{formula, V = S 2/2 ...} (10 ') set To be done. Further, the time derivative V'of the function V is: V '= SS' = S (e "+ λe ') = S (Je" + λJe') / J = S (-(K _C + K _P | e | + K _V | e ′ |) · sgn (S) + E _C + E _P e + E _V e ′ + λJe ′) / J ≦ | S | (− (K _C + K _P | e | + K _V | e ′ |) + | E _c | + | E _P | × | e | + | E _V + λJ | × | e ′ |) / J = | S | ((| E _C | -K _C ) + (| E _P | -K _P ) | e | + ( _{| E V + λJ | -K V} ) | e | a ') / J ... (11' ). Therefore, the control gains K _C , K _P , and K _V are _expressed as K _C ≧ | E _C | ... (12 _a ′) K _P ≧ | E _P │ ... (12 _b ′) K _V ≧ | E _V + λ × ( J + ΔJ) | (12 _c ′) and if these conditional expressions are satisfied, the time derivative V ′ of the function V of the above expression (11 ′) is V ′ ≦ | S | ((| E _C | -K _{C) + (| E P |} -K P) | e | + (| E V + λJ | -K V) | e '|) / J ≦ 0 ... (13' becomes). Therefore, it can be seen that the function V is a Lyapunov function for executing appropriate sliding mode control. The Lyapunov function V is always positive and has a minimum value of 0. If V'≤0, the Lyapunov function V converges to the minimum value (= 0). In addition, the switching variable S
(Control slip surface) always converges, and the control response is S = 0.
Is determined by the constant response function of.

That is, according to Lyapunov's theorem, the switching variable S is asymptotically stable, and if S → 0 (t → ∞) (14 ') is compensated and the switching variable S becomes 0, the above (8' From the equation, e ′ = − λe (15 ′) holds. As a result, the error e also becomes asymptotically stable, and the error e also converges to zero. In this case, it is important to be able to select the control gain that satisfies the above-mentioned conditional expressions (( _12a ') to ( _12c ')). Therefore, if the range of the error e is limited, it is possible to estimate the upper and lower limits of the compensation input ξ, the rotation angle θ, and the nonlinear term C, and thus the upper and lower limits of the compensation error E can also be determined. Furthermore, the upper and lower limits of each compensation error coefficient E _C , E _P , E _V can also be estimated. Therefore, the control gains K _C , K _P , and K _V that are sufficiently larger than the estimated values of the upper and lower limits of each compensation error coefficient may be set.

As described above, according to the sliding mode control system 12, theoretically, the influence of the fluctuation of the disturbance term D, the nonlinear term C, and the inertia J can be completely canceled out, and the optimum control can be performed. You can However, when performing actual control, it is necessary to switch the controller (not shown) according to the sign of the switching variable S, which cannot be realized by an analog control configuration. Therefore,
It is realized by a digital control configuration using a computer (CPU) or the like. Therefore, the control is executed every certain sampling cycle. As described above, according to the sliding mode control system 12, since the control input ν is discontinuously switched and changed by the switching variable S, a minute vibration called chattering is generated in the control for each sampling cycle. As described above, the reason why the sliding mode control, which is theoretically excellent, is not widely used is due to the chattering phenomenon as described above. Therefore, the present inventors have developed the following control method using a sliding mode control system (Japanese Patent Laid-Open No. 04-138579). Here, the disturbance value δ that is the difference between the output (J ^ θ ″ + C ^) of the model 13 and the torque τ that is the control command value is directly calculated, and the calculated disturbance value δ
The control gain of the sliding mode control system 12 is changed according to the above (see FIG. 8), or it is used together with the filter 14 which is the disturbance estimation observer, and the difference between the disturbance estimation value δ ^ and the disturbance value δ by the filter 14 is used. As a result, the amount of disturbance to be controlled by the sliding mode control system 12 is reduced, and the control gain is further suppressed to suppress chattering (see FIG. 9).

[0007]

In the control method using the conventional sliding mode control system as described above, the dead time and the shaft torsional vibration caused by the sensor system, control calculation, control cycle, sample hold, etc. It does not consider the influence of parasitic elements at all. Therefore, when there is a delay time L due to the phase delay of the parasitic element, the chattering amplitude A
Becomes A≈J ⁻¹ λLK _c (16 ′), the chattering increases due to the delay time L, and the control accuracy may decrease. By the way, in the prior art, the control gain _Kc is suppressed to a small value in the above equation (16 '). However, as described above, unless the control gain K _c is set to a sufficiently large value, the conditional expression (12 _a ′) for the control gain cannot be satisfied. Therefore, it cannot be said that the conventional technique is sufficient to suppress chattering. The present invention has been made in view of the above circumstances, and provides a control device using a sliding mode control system capable of sufficiently suppressing control chattering when controlling a controlled object using the sliding mode control system. It is intended to be provided.

[0008]

In order to achieve the above object, a first aspect of the present invention provides the above control so that the control state of a control target converges on a control slip surface for making an error of a control amount close to zero. In a control device using a sliding mode control system that outputs a control command value to an object by switching, a switching variable that defines the control slip surface causes a delay in switching time due to a parasitic element existing during input / output of the control object. It is configured as a control device using a sliding mode control system characterized by giving a dynamic characteristic for compensating for A second aspect of the invention is a sliding mode control system that switches and outputs a control command value to the control target so that the control state of the control target converges on a control slip surface for making the error of the control amount close to zero. In a control device using the above-described control slip, a switching variable that defines the control slip surface is given a dynamic characteristic that compensates for a switching time delay due to a parasitic element existing in the input / output of the control target, and Sliding characterized in that a first gain changing unit for directly calculating a disturbance amount applied to the controlled object based on an input / output value and changing a control gain of the sliding mode control system based on the calculated value is provided. It is configured as a control device using a mode control system.

A third aspect of the present invention is a sliding device that switches and outputs the control command value to the control target so that the control state of the control target converges on the control slip surface for making the error of the control amount close to zero. In a control device using a mode control system, a switching variable that defines the control slip surface is given a dynamic characteristic that compensates for a switching time delay due to a parasitic element existing in the input / output of the controlled object, and A disturbance estimator that estimates the amount of disturbance applied to the controlled object based on the input / output value of the controlled object, and a second gain change that changes the control gain of the sliding mode control system based on the estimated value by the disturbance estimator Is provided as a control device using a sliding mode control system. Furthermore, it is a control device using a sliding mode control system in which the parasitic element is a higher-order mode. Further, it is a control device using a sliding mode control system in which the parasitic element has a dead time. Furthermore, the parasitic element is a control device using a sliding mode control system, which is a delay element of the lower control loop.

[0010]

According to the first aspect of the present invention, the sliding which switches and outputs the control command value to the control target so that the control state of the control target of the control slip surface for making the error of the control amount close to zero converges. When a mode control system is used, the switching variable that specifies the control slip surface is given a dynamic characteristic that compensates for the switching time delay due to parasitic elements existing in the input and output of the controlled object. Thus, by giving the switching variable dynamic characteristics according to the parasitic element, the delay time can be reduced to zero or shortened, and chattering can be reduced. According to the second invention,
Based on the input / output value of the control target, the switching time designating the control slip surface is given a dynamic characteristic that compensates for a delay in the switching time due to a parasitic element existing in the input / output of the control target. The amount of disturbance applied to the controlled object is directly calculated, and the control gain of the sliding mode control system is changed by the first gain changing unit based on the calculated value. In this way, the control gain can be reduced by directly calculating the disturbance amount and changing the gain according to the calculated disturbance amount, so that chattering can be further reduced.

According to the third aspect of the invention, the switching variable designating the control sliding surface is given a dynamic characteristic for compensating for a delay in switching time due to a parasitic element existing in the input / output of the controlled object. At the same time, the disturbance amount applied to the controlled object is estimated based on the input / output value of the controlled object by the disturbance estimation unit, and the control gain of the sliding mode control system is set to the second based on the estimated value by the disturbance estimation unit. It is changed by the gain changing unit. in this way,
By using the disturbance estimator together, the amount of unknown disturbance can be suppressed and the control gain can be reduced, which further reduces chattering. Furthermore, the designated element is said to be in the higher mode. Furthermore, it is said that the above parasitic element is a dead time. Furthermore, it is said that the parasitic element is a delay of the lower control loop. As described above, the sliding mode control can be applied to a control object having a higher-order mode such as torsional vibration, a dead time of the sensor system or the controller unit, or a delay of a lower control loop such as a current loop, thereby reducing chattering. In particular, when the state estimation unit is used, the state estimation can be performed stably, so that chattering can be further reduced. As a result, it is possible to obtain a control device using a sliding mode control system capable of sufficiently suppressing control chattering when controlling a controlled object using the sliding mode control system.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention (first to third embodiments) will be described below with reference to the accompanying drawings.
The invention will be described to provide an understanding of the present invention. It should be noted that the following embodiments are merely examples embodying the present invention and do not limit the technical scope of the present invention. 1 is a block diagram showing a schematic structure of a control device A ₁ using a sliding mode control system according to an embodiment of the first invention, and FIG. 2 is a sliding mode according to an embodiment of the second invention. FIG. 3 is a block diagram showing a schematic configuration of a control device A ₂ using a control system, and FIG. 3 is a block diagram showing a schematic configuration of a control device A ₃ using a sliding mode control system according to an embodiment of the third invention. 4 is a partial block diagram showing the case where a parasitic element exists in the input and output for the nominal mathematical model of the controlled object, FIG. 5 is an equivalent conversion block diagram of the controlled object, and FIG. 6 adopts a motor drive system as the controlled object. It is a schematic block diagram of the said drive system at the time of doing. Figure 1
As shown in (a) and (b), the control device A _{1 according} to the first embodiment of the present invention converges the control state of the controlled object 1 on the control slip surface for bringing the error of the control amount x close to zero. This is similar to the conventional example in that the sliding mode control system 2 that switches and outputs the control command value u to the controlled object 1 is used. However, in the present embodiment, a dynamic characteristic N (s) is provided to the switching variable S designating the controlled slip surface so as to compensate for the switching time delay due to the parasitic element existing in the input / output of the controlled object 1. The point is different from the conventional example.

Here, the basic principle and the like of the device A ₁ will be briefly described. As shown in FIG. 4, for the nominal mathematical model x (s) = J ⁻¹ (u (s) + d (s)) / s ² (1) of the controlled object 1, it exists in the input / output. Parasitic elements P _I (s) _, P _O
(S) exists, and the transfer function of the controlled object 1 is represented by x (s) = P _o (s) J ⁻¹ (P (s) u (s) + d (s)) / s ² (2) Shall be done. Where J is the inertia, u is the input that is the control command value, x is the output that is the control amount, d is the disturbance, s is the Laplace operator, and the parasitic element P _I (s),
It is assumed that P _o (s) does not have an unstable zero point. Here, if a variable δ of d (s) = P _I (s) δ (s) (3) is used as the disturbance instead of the disturbance d, the transmission of the controlled object 1 expressed by the above equation (2) is transmitted. The function is χ (s) = J ⁻¹ (u (s) + δ (s)) / s ² (4) x (s) = P _o (s) P _I (s) χ (s) (5) It is represented by. FIG. 5 shows an equivalent conversion block diagram at this time. Here, what is actually observed is the output x and the control input u.
Χ and δ are not observed.

For this controlled object 1, Lyapunov function candidates V V (t) = σ (t) ² (6) σ (t) = χ (t) + λdχ (t) / dt (7) Think However, λ is a positive number (time constant) that determines the convergence property of χ. In the conventional sliding mode control, the control input u is expressed by u (t) =-(K _P | x (t) | + K _V | dx (t) / dt | + K _c ) × sgn (S (t)) (8) ) S (t) = x (t) + λdx (t) / dt (9) is given (corresponding to the equations (8 ′) and (9 ′) above),
In this embodiment, the control input u is given by the following equation. However, K
_P , K _V , and K _c are control gains, S is a switching variable for the conventional sliding mode control, and R is a switching variable newly given in this embodiment (see FIG. 1 (b)). u (t) =-(K _P | x (t) | + K _V | dx (t) / dt | + K _c ) × sgn (R (t)) (10) R (s) = N (s) S (S) (11) S (t) = x (t) + λdx (t) / dt (9) Here, sgn (σ (t)) = sgn (R (t)) (12) Thus, given the dynamic characteristic N (s) of the above equation (11),
The respective control gains of the equation (10) are expressed as K _P ≧ 0, K _V ≧ Jλ ⁻¹ , K _c ≧ │δ (s) │ + Jλ ^-1 │dχ / dt-dx (t) / dt│ (13) Then, in the above equations (9) to (12), the time differential value of the Lyapunov function candidate V is dV / dt = 2σdσ / dt = 2σ (dχ / dt + λd ² χ / dt ² ) = 2σ (dχ / dt + λJ ^{− 1} (u (s) + δ (s)) = 2σλJ ⁻¹ (Jλ ⁻¹ dχ / dt + u (s) + δ (s)) = 2σλJ ⁻¹ (Jλ ⁻¹ dχ / dt− (K _P | x (t ) | + K _V | dx (t) / dt | + K _c ) × sgn (R (t)) + δ (s)) = 2σλJ ⁻¹ (Jλ ⁻¹ dχ / dt− (K _P | x (t) | + K _V | dx (t) / dt | + K _c ) × sgn (σ (t)) + δ (s)) ≦ 2 | σ | λ J ⁻¹ (Jλ ⁻¹ | dχ / dt | −K _P | x (t) │-K _V │ dx (t) / d t | −K _c + | δ (s) |) ≦ −2 | σ | λJ ⁻¹ × {K _P | x (t) | + (K _V −Jλ ⁻¹ ) | dx (t) / dt | + a} ... (14), and the candidate V is Lyapunov function _{- (K c | δ (s} ) | -Jλ -1 | | dχ / dt-dx (t) / dt). Therefore, since the Lyapunov function V converges to 0, from the above equation (6), σ
Is asymptotically stable and converges to zero. That is, from the equation (7), χ also converges to 0 with the time constant λ, and the output x also becomes 0 from the equation (5). Thus, by giving the dynamic characteristic N (s) to the switching variable R as shown in the above equations (9) to (11), the delay time between the switching variable R and σ is eliminated, and ) Can be satisfied. As a result, the above expression (14) can be satisfied, and theoretically chattering can be completely eliminated.

However, in the prior art, the above (8),
As shown in the equation (9) (or the equations (8 ') and (9')), the switching variable S does not have a dynamic characteristic.
If a parasitic element exists, a delay time exists between the switching variable S and σ, and the above equation (14) is not satisfied. That is,
In the prior art, the maximum value of the delay time of the switching variable with respect to σ is L L = ω ^max (2π / ω × ∠P (jω)) (15) P (s) = P _o (s) P _I (s) ( 16), the chattering amplitude A is given by A≈J ⁻¹ λLK _c (17) as described above. However, j is an imaginary unit. In this case, in the present embodiment, since the delay time L = 0, the chattering amplitude A becomes 0 and chattering does not occur.

[Study 1] In the above embodiment, the parasitic elements P _I (s) and P _o (s) of the controlled object 1 are assumed to have no unstable zero point. Therefore, the case of having an unstable zero is considered below. At this time, the dynamic characteristic N (s) satisfying the above equation (12) becomes unstable and cannot be adopted. Therefore, the above formula (14) generally does not hold. In this case, there is a delay time between the switching variable S and σ, and the maximum value of the switching variable delay time with respect to σ is L = ω ^max (2π / ω × ∠N (jω) P (jω) ) (18), the chattering amplitude A is the same as in (17) above.
Given by the formula. Therefore, it is considered to reduce the maximum delay time L in the equation (18). That is, the parasitic element P
Even if an unstable zero exists in (s) (= P _I (s) P _o (s)), a stable motion that satisfies ∠M (jω) P (jω) = 0 for ∀ω (19) Since the characteristic M (s) always exists, if the dynamic characteristic M (s) is N (s), then the above (18)
From the equation, the maximum delay time L becomes 0, and from the equation (17), the amplitude A also becomes 0, so that chattering can be suppressed.

[Study 2] Further, an unstable zero point exists in the parasitic element P (s), and the dynamic characteristic M (s) of the equation (19) is set to N.
If (s), the transfer function from the control input u to the switching variable R may not be truly proper. in this case,
It is generally impossible to directly change the dynamic characteristic M (s) of the equation (19) to N (s). Therefore, a low-pass filter F (s) having a sufficiently high cutoff frequency is used as a dynamic characteristic M
In addition to (s), a new dynamic characteristic F (s) M (s) in which a delay element is provided in the high range of this dynamic characteristic M (s) is set as N (s) to switch from the control input u. While making the transfer function to the variable R truly proper, ∠N (jω) P (j
ω) can be reduced. That is, the maximum delay time L is suppressed from the above equation (18), and the chattering amplitude A is suppressed from the above equation (17).

Next, the second invention will be described.
As shown in FIG. 2, a control device A according to an embodiment of the second invention.
_In addition to the device A ₁ of the _first embodiment of the present invention, ₂ directly calculates the disturbance amount applied to the controlled object 1 based on the input / output value of the controlled object 1, and the sliding mode control system based on the calculated value. Gain changing unit 3 for changing the control gain of 2
_This is different from the conventional example in that _a (corresponding to the first gain changing unit) is provided. The basic principle of the device A ₂ of the _second embodiment of the present invention will be briefly described below. In the embodiment of the first aspect of the present invention, the control gain K _{c is set} to K _c (t) = | Δ (t) | + k (20) Δ (s) = F _G (s) P (s) δ (s) = F _G (s) (Js ² x (s) −P (s) u (s)) (21) k = | δ (t) −Δ (t) | + Lλ ⁻¹ | dχ / dt−dx ( t) / dt | ... (22), and change. However, the dynamic characteristics F in this case
_G (s) is a design parameter given arbitrarily. At this time, K _c (t) = | Δ (t) | + K ≧ | Δ (t) | + | δ (t) -Δ (t) | + Jλ ⁻¹ | dχ / dt-dx (t) / dt | ≧ | δ (t) | + Jλ ⁻¹ | dχ / dt−dx (t) / dt | (23), the above equation (13) is satisfied. Therefore, the same can be said for the first embodiment of the invention. In this embodiment, further, the gain changer 3 using the equation (20) is used.
_The control gain K _c is changed by a, and this control gain K
_{Since c} can be suppressed to a small value, the chattering amplitude A can be suppressed to a smaller value from the equation (18). Therefore, chattering can be further reduced as compared with the first embodiment of the invention.

Next, the third invention will be described.
As shown in FIG. 3, an apparatus A _{3 according} to an embodiment of the third invention is shown.
Is a disturbance estimation observer 4 (corresponding to a disturbance estimation unit) that estimates the amount of disturbance applied to the controlled object 1 based on the input / output value of the controlled object 1, and the disturbance estimation observer 4. This is different from the conventional example in that a gain changing unit 3 _b (corresponding to a second gain changing unit) that changes the control gain of the sliding mode control system 2 based on the estimated disturbance value is provided. The basic principle of the device A ₃ of the _third embodiment of the present invention will be briefly described below. Here, the estimated disturbance value δ _O is estimated by the disturbance estimation observer 4 given by the following equation. _{δ O (s) = F G} (s) P (s) δ (s) = F G (s) (Js 2 x (s) -P (s) u (s)) ... (24) i.e., the control input u and the control gain K _c are expressed as follows: u (t) = − (K _P | x (t) | + K _V | x (t) / dt | + K _C ) × sgn (S (t)) − δ _O (t) (25) K _C (t) = | Δ (t) −δ _O (t) | + k (26) If the control gain K _C is further changed by the gain changer 3 _b than the equation (20), The chattering width A can be further reduced to a smaller value according to the equation (18). Thus, by using the disturbance estimation observer 4 together, the amount of unknown disturbance can be suppressed, the control gain can be reduced, and chattering can be further reduced.

Next, a specific example of the parasitic element will be described.
For example, consider a motor drive system having a torsional vibration of a shaft as a parasitic element in the controlled object 1 as shown in FIG.
Now, output the rotation angle θ _{m of the} motor as x, load side inertia as J _L , motor inertia as J _M , shaft spring constant as K,
If a model that does not consider torsional vibration as in the following equation is a nominal mathematical model P _N (s), then P _N (s) = (J _M + J _L ) ⁻¹ / s ² (27) Then, the parasitic element P _O (s) can be expressed as P _O (s) = (s ² J _L + K) / (s ² J _L J _M / (J _M + J _L ) + K) (28). Alternatively, when the load-side rotation angle θ _L is selected as the output x, the parasitic element P _O (s) is P _O (s) = K / (s ² J _L J _M / (J _M + J _L ) + K )… (29) Furthermore, if there is a dead time in the sensor system or the controller as a parasitic element of the controlled object 1, the input dead time is L _I , the output dead time is L _O , and the control input u changes from the input observed value u _s . If the dead time is L _U , the parasitic elements P _I (s), P _O (s) and P _U (s) are P _I (s) = e ^−Lis , P _O (s) = e ^−Los , It can be expressed as P _U (s) = e ^−Lus (30)

Further, let us consider a case where the controlled object 1 is a motor drive system in which a delay of the current loop exists as a parasitic element. At this time, if the time constant of the current loop is T, the parasitic element P _I (s), P _U (s) can be expressed as P _I (s) = 1 / (Ts + 1), P _U (s) = 1 / (Ts + 1) ... (31). The above considers the case where various types of parasitic elements exist independently, but further considers the case where the above parasitic elements exist across multiple types. In this case, each parasitic element is expressed as the product of the parasitic elements of the above-mentioned embodiment. For example, when the dead time L _U and the delay T of the current loop exist at the same time, the parasitic element P _U (s) is expressed by P _U (s) = e ^−Lus / (Ts + 1) (32). As described above, the devices A _{1 to} A _1
When the sliding mode control in ₃ is applied to the controlled object 1 which has a higher mode such as torsional vibration or a delay of a lower control loop such as a dead time of a sensor system or a controller or a current loop, chattering can be all reduced. it can. In particular, when applied to the device A ₃ of the _third embodiment of the present invention, the state can be estimated stably. As a result, in any case, when controlling the controlled object using the sliding mode control system, a control system using the sliding mode control system that can sufficiently suppress control chattering can be obtained.
Each of the above-mentioned embodiments A _{1 to} A ₃ may have a non-linear compensation system as in the conventional example. In that case, the control command value u from the sliding mode control system 2 is compensated. It is possible to further improve the control accuracy.

[0022]

Since the control device using the sliding mode control system according to the first aspect of the invention is configured as described above, the delay time can be reduced by providing the switching variable with the dynamic characteristic according to the parasitic element. Zeroing or shortening can reduce chattering. According to the second invention,
Since the control gain can be reduced by directly calculating the disturbance amount and changing the gain according to the calculated disturbance amount, chattering can be further reduced. Also,
According to the third aspect of the present invention, the amount of unknown disturbance can be suppressed and the control gain can be reduced by using the disturbance estimation unit together, so that chattering can be further reduced. Further, the above sliding mode control can be applied to a control object having a higher order mode such as torsional vibration or a dead time of the sensor system or the controller section or a delay of a lower control loop such as a current loop, thereby reducing chattering. In particular, when the state estimation unit is used, the state estimation can be performed stably, so that chattering can be further reduced. As a result, it is possible to obtain a control device using a sliding mode control system capable of sufficiently suppressing control chattering when controlling a controlled object using the sliding mode control system.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a control device A ₁ using a sliding mode control system according to an embodiment of the first invention.

FIG. 2 is a block diagram showing a schematic configuration of a control device A ₂ using a sliding mode control system according to an embodiment of the second invention.

FIG. 3 is a block diagram showing a schematic configuration of a control device A ₃ using a sliding mode control system according to an embodiment of the third invention.

FIG. 4 is a partial block diagram showing a case where an input / output element exists for a nominal mathematical model to be controlled.

FIG. 5 is an equivalent conversion block diagram of the controlled object.

FIG. 6 is a schematic configuration diagram of a drive system when a motor drive system is adopted as the control target.

FIG. 7 is a block diagram showing a schematic configuration of a first example A ₀₁ of an apparatus according to a control method using a conventional sliding mode control system.

FIG. 8 is a block diagram showing a schematic configuration of a second example A ₀₂ of an apparatus according to a control method using a conventional sliding mode control system.

FIG. 9 is a block diagram showing a schematic configuration of a third example A ₀₃ of an apparatus according to a control method using a conventional sliding mode control system.

[Explanation of symbols]

A _1, A _2, A ₃ ... controller 1 ... controlled target 2 ... sliding mode control system 3 _a, 3 _b ... gain modifier (first, corresponds to a second gain changing unit) 4 ... disturbance estimation observer (disturbance Equivalent to the estimation section)

Claims (6)

[Claims] 1. A sliding mode control system for switching and outputting a control command value to the control target so that the control state of the control target converges on a control slip surface for bringing the error of the control amount close to zero is used. In a control device, a sliding mode control characterized by providing a switching variable that defines the control slip surface with a dynamic characteristic that compensates for a switching time delay due to a parasitic element existing in the input / output of the controlled object. Control device using a system. 2. A sliding mode control system for switching and outputting a control command value to the control target so that the control state of the control target converges on a control slip surface for bringing the error of the control amount close to zero is used. In the control device, a switching variable that defines the control slip surface is given a dynamic characteristic that compensates for a delay in switching time due to a parasitic element existing in the input / output of the control target, and the input / output value of the control target is set. The sliding mode control system is characterized in that a first gain changing unit for directly calculating the amount of disturbance applied to the controlled object based on the calculated value and changing the control gain of the sliding mode control system based on the calculated value is provided. Control device using. 3. A sliding mode control system for switching and outputting a control command value to the control target so that the control state of the control target converges on a control slip surface for making the error of the control amount close to zero is used. In the control device, a switching variable that defines the control slip surface is given a dynamic characteristic that compensates for a delay in switching time due to a parasitic element existing in the input / output of the control target, and the input / output value of the control target is set. And a second gain changing unit for changing the control gain of the sliding mode control system based on the estimated value by the disturbance estimating unit. And a control device using a sliding mode control system. 4. The control device using the sliding mode control system according to claim 1, wherein the parasitic element is a higher order mode. 5. The control device using the sliding mode control system according to claim 1, wherein the parasitic element is dead time. 6. A controller using a sliding mode control system according to claim 1, wherein the parasitic element is a delay element of a lower control loop.