A method for setting control parameters for model prediction control related to a predetermined control target, the method being executed by a control device having a first integrator to which an offset between a predetermined target command and an output of the predetermined control target corresponding to an actual target device which is an actual target of servo control is input, in order to allow an output of the actual target device tracks the predetermined target command, wherein
- the control device includes a model prediction control unit which has a prediction model that defines a correlation between a predetermined expanded state variable related to an expanded control target including at least the predetermined control target and a control input to the expanded control target in a form of a predetermined state equation and which performs the model prediction control based on the prediction model according to a predetermined evaluation function in a prediction section having a predetermined time width with respect to the predetermined target command and outputs a value of the control input at least at an initial time point of the prediction section,
- a predetermined state variable which is a portion of the predetermined expanded state variable and is related to the predetermined control target includes a predetermined integral term represented by a product of the offset and a predetermined integral gain, and
- the setting method includes:
- a time response determining step of determining a desired time response in an optimum servo control structure which includes a virtual integrator virtually corresponding to the first integrator and is related to a virtual control target virtually corresponding to the expanded control target and in which, when a virtual target command for the virtual control target is r1, a virtual control input to the virtual control target is a virtual control input u1, and a virtual output of the virtual control target is y1, the virtual control input u1 is represented by Equation 1 below including a state feedback gain KF and an integral gain KI;
- a gain calculation step of calculating the state feedback gain KF and the integral gain KI corresponding to the desired time response;
- a weighting coefficient calculation step of calculating a second weighting coefficient Q and a third weighting coefficient R among a first weighting coefficient Qf, the second weighting coefficient Q, and the third weighting coefficient R in a predetermined Riccati equation represented by Equation 3 below on the basis of the state feedback gain KF and the integral gain KI when the control input to the expanded control target is u, the output of the expanded control target is y, the offset is e, the predetermined state variable is x, and a state equation related to the predetermined control target is represented by Equation 2 below;
- and
- a setting step of setting the second weighting coefficient Q as a weighting coefficient corresponding to a state quantity cost which is a stage cost related to the predetermined state variable and setting the third weighting coefficient R as a weighting coefficient corresponding to a control input cost which is a stage cost related to the control input in the predetermined evaluation function.
- [
- Math
- .
-
- 1
- ]
- u
-
-
- 1
- =
- -
- K
- F
-
- x
- +
- K
- I
-
- ∫
- (
- r
-
-
- 1
- -
- y
-
-
- 1
- )
-
- dt
- (
- Equation
-
-
- 1
- )
- {
- x
- .
- =
- Ax
- +
- Bu
- z
- .
- =
- -
- Cx
- +
- r
- ,
- [
- x
- .
- z
- .
- ]
- =
- [
- A
- 0
- -
- C
- 0
- ]
-
- [
- x
- z
- ]
- +
- [
- B
- 0
- ]
-
- u
- +
- [
- 0
- r
- ]
-
-
- where
-
-
- e
- =
- r
- -
- y
- =
- z
- .
- (
- Equation
-
-
- 2
- )
- Q
- f
-
- A
- e
- +
- A
- e
- T
-
- Q
- f
- -
- Q
- f
-
- B
- e
-
- R
- -
- 1
-
- B
- e
- T
-
- Q
- f
- +
- Q
- =
- 0
- (
- Equation
-
-
- 3
- )
- where
-
-
- A
- e
- =
- [
- A
- 0
- C
- 0
- ]
- ,
- B
- e
- =
- [
- B
- 0
- ]