WO2001025861A1 - Dispositif concepteur controleur - Google Patents
Dispositif concepteur controleur Download PDFInfo
- Publication number
- WO2001025861A1 WO2001025861A1 PCT/JP2000/006898 JP0006898W WO0125861A1 WO 2001025861 A1 WO2001025861 A1 WO 2001025861A1 JP 0006898 W JP0006898 W JP 0006898W WO 0125861 A1 WO0125861 A1 WO 0125861A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- controller
- weight
- control
- controlled object
- model
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0205—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system
- G05B13/024—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B11/00—Automatic controllers
- G05B11/01—Automatic controllers electric
- G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
Definitions
- the present invention relates to a design apparatus for designing a multivariable controller based on H-infinity (H ⁇ ) control theory.
- H H control theory even if there is an error between the actual control target and the numerical model used in the design, if information on the error is obtained, the actual control target is stabilized by taking the error into account. It is possible to design a controller to be used. Also, it is said that H H control theory is easier to intuitively give control specifications when designing a control system than conventional control theory. For example, when designing a control system using conventional control theory, the design specifications were the poles of a closed loop system or a weight matrix of an evaluation function. However, the physical meaning of these values was unclear, and their setting required a lot of trial and error.
- control specifications can be specified by the frequency response of a closed-loop system consisting of the control target and the controller.
- the controller In the controlled object model used in the design of a multivariable control system, the magnitude of the error component from each manipulated variable to the controlled variable varies. In this way, the error component of the model for each manipulated variable varies, but in the H ⁇ control theory, the controller is designed based on the component with a large error gain. The response tends to be very conservative, that is, overly stable. Also, it may be necessary to adjust the control weight for each control amount in order to avoid interference between the control amounts. Therefore, it has been proposed to introduce an operation amount weight called a scaling matrix T in order to make the magnitude of the error of the control target model uniform and to weight the control amount to the control amount.
- a sensitivity weight W s it is necessary to determine a frequency weight called a sensitivity weight W s in order to determine a set value tracking characteristic of a closed loop system.
- the H ⁇ control theory is a design method in the frequency domain, it is easy to design in the control of mechanical systems, but it is used for the design of control systems that are difficult to imagine in the frequency domain such as process control. Ku rather, to choose the sensitivity weight W s properly there is a problem that it is difficult.
- the present invention has been made to solve the above-described problems, and an object of the present invention is to provide a design apparatus capable of easily realizing a controller design based on H ⁇ control theory. Target.
- the controller design apparatus of the present invention includes a storage device for storing a generalized plant, a response characteristic of a controlled object model, or a generalized plant based on a response characteristic of a closed loop system including the controlled object model and a controller.
- the generalized plant adjusts the input of the manipulated variable to the control target model and the control target model provided at the preceding stage of the control target model.
- Parameter calculating means for calculating the frequency response of the control target model; and a scaling matrix for determining the weight of the operating amount by the operating amount weight adjusting means. And a scaling matrix calculating means for calculating T based on the frequency response of the control target model so that each gain of the control target model is equalized.
- the controller calculation means includes a generalization plant of the generalized plant stored in the storage means. Applying the scaling matrix T to the manipulated variable weight adjustment means
- the generalized plant includes a first controlled object model for an operation amount, a second controlled object model for a disturbance, and a pre-stage of the first controlled object model.
- Operating amount weight adjusting means for adjusting the operating amount input to the first controlled object model provided in the first control object model; and the parameter calculation means includes the first controlled object model and the second control object.
- a frequency response calculating means for calculating the frequency response of the target model and a scaling matrix T for determining the weight of the operation amount by the operation amount weight adjusting means are calculated based on the frequency responses of the first and second control target models.
- scaling matrix calculating means for calculating each gain of the first controlled object model to be equal to the maximum value among the gains of the controlled object model, and the controller calculating means is stored in the storage means
- the controller is derived by applying the scaling matrix T to the manipulated variable weight adjustment means of the generalized plant.
- the generalized plant stored in the storage means includes the operation amount weight adjusting means and the control object model for the operation amount.
- the generalized plant stored in the storage means is a closed-loop set value follow-up characteristic of the operation amount weight adjustment means, the control object model for the operation amount, and the controller. Is provided before or after the frequency sensitivity weight adjusting means for determining the control sensitivity weight adjusting means, and the design apparatus determines the weighting of the control amount by the control amount weight adjusting means. It is provided with setting means for setting the weight matrix S.
- the parameter calculation means includes a setting means for setting a transient response characteristic of the closed loop system, and a setting of the closed loop system based on the transient response characteristic of the closed loop system.
- Frequency sensitivity weight calculation means for calculating a frequency sensitivity weight for determining the value tracking characteristic, wherein the controller calculation means applies the frequency sensitivity weight to the generalized plant stored in the storage means, Is derived.
- the frequency sensitivity weight calculating means includes a transfer function obtained by multiplying a transient response characteristic of the closed loop system and a transfer function from a set value of the closed loop system to a deviation from the set value of the closed loop system by a frequency sensitivity weight.
- the frequency sensitivity weight is calculated based on the design index that the H ⁇ norm is less than 1.
- the setting means approximates a transient response characteristic of the closed loop system by a first-order lag characteristic.
- the setting means approximates the transient response characteristic of the closed loop system with the characteristic of the secondary system.
- FIG. 1 is a block diagram showing a configuration of a controller design apparatus according to a first embodiment of the present invention.
- FIG. 2 is a block diagram illustrating a configuration of a model in which an actual control target is expressed as a mathematical expression.
- FIG. 3 is a block diagram showing the configuration of a conventional generalized plant.
- FIG. 4 is a block diagram showing the configuration of a generalized plant used in the design apparatus of the present invention. is there.
- FIG. 5 is a block diagram showing a configuration of a mouth bus control system in which a controller is added to the generalized plant of FIG.
- FIG. 6 is a block diagram showing a configuration of an actual controller including a controller designed using the designing apparatus of the first embodiment of the present invention.
- FIG. 7 is a diagram illustrating an additive error of the numerical model of the control target.
- FIG. 8 is a block diagram showing a configuration of a closed loop system used for determining the sensitivity weight.
- FIG. 9 is a diagram for explaining the function of the scaling matrix in the first embodiment of the present invention.
- FIG. 10 is a diagram for explaining the function of the scaling matrix in the second embodiment of the present invention.
- FIG. 11 is a block diagram showing a configuration of a controller design device according to a third embodiment of the present invention.
- FIG. 12 is a block diagram showing a configuration of a generalized plant according to the third embodiment of the present invention.
- FIG. 13 is a block diagram showing a configuration of an actual controller including a controller designed using the design apparatus according to the third embodiment of the present invention.
- FIG. 14 is a block diagram showing a configuration of a generalized plant according to the fourth embodiment of the present invention.
- FIG. 15 is a block diagram showing a configuration of a controller design device according to a fifth embodiment of the present invention.
- FIG. 16 is a block diagram illustrating a configuration of a model in which an actual control target is expressed by a mathematical expression.
- FIG. 17 is a block diagram illustrating a configuration of a conventional generalized plant.
- FIG. 18 is a block diagram showing a configuration of a generalized plant used in the design apparatus of the present invention.
- FIG. 19 is a block diagram showing a configuration of a robust control system in which a controller is added to the generalized plant of FIG.
- FIG. 20 is a block diagram showing the configuration of an actual controller including a controller designed using the design apparatus of the present invention.
- Figure 21 is a diagram showing the additive error of the numerical model of the controlled object.
- FIG. 22 is a block diagram showing a configuration of a closed loop system used for determining the sensitivity weight.
- Figure 23 shows the time response characteristics when the closed-loop system is approximated by a second-order system.
- FIG. 1 is a block diagram showing a configuration of a controller design device according to a first embodiment of the present invention.
- the design apparatus shown in Fig. 1 includes a control target model input unit 1 for inputting parameters of the control target model, a control target model registration unit 2 for registering the model parameters in a storage unit described later, and a general control unit.
- Storage unit 3 for storing the formula of the generalized plant and the formula of the control target model which is a part of this generalized plant; a frequency response calculation unit 4 for calculating the frequency response of the control target model; A scaling matrix T for calculating a scaling matrix T for making the magnitude of the error equal to the maximum value among the gains of the control target model is stored in the storage matrix 3 and the storage matrix 3. And a controller calculation unit 6 for applying a scaling matrix T to the generalized plant thus derived to derive parameters of the controller.
- FIG. 2 is a block diagram showing a configuration of a model in which an actual control target is expressed as a mathematical expression.
- the numerical model of the controlled object shown in FIG. 2 includes a first controlled object model 11 for a manipulated variable u and a second controlled object model 12 for a disturbance w.
- P u is the transfer function of model 11
- P w is the transfer function of model 12.
- Models 11 and 12 were obtained as a result of model identification using data obtained from a step response test for an actual controlled object.
- the control amount y which is the output of the control target, is the sum of the outputs of models 11 and 12.
- Figure 3 shows the configuration of a conventional generalized plant that includes such a numerical model of the controlled object.
- the generalized plant as shown in FIG. 3, a frequency weight called a sensitivity weight W s used to determine the set value tracking characteristic, complementary used to determine the robust preparative stability sensitivity weight W t Frequency weights called, and set values r and outputs ⁇ 1 and ⁇ 2 are introduced in addition to the input (manipulation amount) u, input (disturbance) w, and output (control amount) y of the controlled object. This is to achieve both tracking characteristics and robust stability.
- 13 is a block (frequency sensitivity weight adjusting means) representing the sensitivity weight W s, and Z, is an output for evaluating the set value tracking characteristic. Further, 1 4 is a block representing the complementary sensitivity weight W t, Z 2 is an output for evaluating the robustness bets stability.
- the uncertainty of the model is estimated based on the numerical model of the control target, the complementary sensitivity weight W is determined, and the frequency response is taken into account by following the set value r.
- the sensitivity weight W s was determined by directly specifying, and the parameters of the controller were determined by the eye television.
- the controller will be designed based on the one with the larger gain due to the difference in gain with respect to the plant output for each manipulated variable. It tends to be a target, that is, overly stable.
- the controller can have an integral characteristic.
- the generalized plant is not stable, it cannot be reduced to the standard H ⁇ problem.
- a generalized plant as shown in FIG. 4 is considered.
- M is a scaling matrix for adjusting the effect of disturbance w on control variable y
- T is a scaling matrix for adjusting the magnitude of the error of the controlled model
- ⁇ This is a weight for giving the controller integral characteristics to eliminate the deviation.
- ⁇ (s) is defined as sZ (s + a). Where s is the Laplace operator And a (> 0) is any real number.
- 15 is a block representing the scaling matrix M
- 16 is a block representing the scaling matrix T (operation amount weight adjusting means)
- 17 is a block representing the weight ⁇ _′ ⁇ .
- Deviation e 2 is intended multiplied by the weight alpha-1 I to the deviation e, the input to the controller.
- Fig. 5 shows the configuration of a robust control system that adds controller II to the generalized plant described above.
- reference numeral 18 denotes a block representing a controller ⁇ .
- the control amount y which is the output of the control target, follows the set value r, the influence of the disturbance w is removed, and the control target fluctuates and the control target model
- the purpose is to determine the parameters of controller K so that they can be stabilized even if there is an error.
- the H ⁇ control problem can be considered as a problem that reduces the H ⁇ norm (gain) of the transfer function from (r, w) to (z,, z).
- each of the set value tracking characteristic, robust stability, and disturbance suppression can be considered as follows.
- (A) set value tracking characteristic H ⁇ Bruno transfer functions ranging from the set value r to deviation e (more precisely, multiplied by the frequency weight a- 'W s to the set value r, the transfer function ranging from r)
- a- 'W s is a frequency weight for restricting the bandwidth to follow (so as to follow only example if low frequency).
- D 2 D 22 can be expressed as One D a D pl MD a D p2 T
- the sensitivity weight w s and the complementary sensitivity weight W are designed, multiplied by the output of equation (11), and the eye television is performed. Is found in the state space representation.
- the output unit of the formula (1 1), the output of FIG. 4 ⁇ ⁇ ', ⁇ 2' means the portion corresponding to. Therefore, parameters Isseki C of formula (1 1),, D, D, 2 output equations to consisting of pairs as follows that the sensitivity weight W s and the complementary sensitivity weight W t and the diagonal Multiply the angular matrix Q from the left.
- the parameters of the controller K can be calculated.
- Controller K is a controller that solves the H ⁇ control problem in a generalized plant, and the actual controller implemented in a plant such as a distillation tower has weight a given to controller K as shown in Fig. 6. — 'I multiplied by the scaling matrix T.
- FIG. 7 shows the additive error of the control target model 11.
- reference numeral 19 denotes a block representing an additive error ⁇ .
- the model error due to the reduction of the model 11 is expressed as an additive error ⁇ as shown in Fig. 7.
- the controller is designed so that the controller output is stable even if the target characteristics deviate from the model 11.
- the complementary sensitivity weight W should be determined so as to cover the additive error ⁇ .
- the general formula of this complementary sensitivity weight W t Is shown in the following equation. It should be noted that since the change in the model 12 is not related to the stability of the system, the design is performed on the assumption that only the model 11 changes.
- the maximum value Gmax of the gain of the error ⁇ is multiplied by a safety factor ⁇ 5 ( ⁇ 5 is, for example, 1) for the additive error ⁇ whose size is adjusted using the scaling matrix T.
- ⁇ 5 is, for example, 1
- the element of the complementary sensitivity weight W That is, the complementary sensitivity weight W t of elements (weight) W, ', W I 2 , W, 3, ⁇ ⁇ ' W, N ' is defined as follows.
- the complementary sensitivity weight W is an NXN matrix.
- the sensitivity weight W s is an LXL matrix.
- the element W s L of the sensitivity weight W s is the weight for the L-th control variable y L.
- a closed loop system as shown in Fig. 8, which is a simplified version of the robust control system in Fig. 5, is considered.
- 11 a is a block representing a numerical model P to be controlled
- 13 a is a block representing a frequency weight Ws ′.
- S (s) is a sensitivity function that mainly indicates control performance related to quick response, such as set value tracking and disturbance suppression
- V w means that equation (2 0) holds for all frequencies ⁇ .
- the frequency weight W SL '(s) is the product of ⁇ 1 (s) and W SL (s), and is defined as Is done.
- Equation (2 1) is a transfer function from the set value r of the closed loop system shown in FIG. This shows that the H ⁇ norm of the transfer function from r to z, obtained by multiplying the set value r by the frequency weight ⁇ 1 (s) WSL (s) is less than 1. This equation (2 1) By setting the weight W SL (s) so as to satisfy, it is possible to design the controller K in consideration of the set value tracking characteristics.
- the scaling matrix M is a JXJ matrix.
- the element M ”of the scaling matrix M is the weight for the Jth disturbance w”, and its initial value is 1.
- Each element ⁇ ” is converted to the control amount y by each disturbance w” This is an adjustment parameter that determines the disturbance suppression performance by adjusting the influence. That is, when it is desired to increase the suppression of the disturbance w of L-specified 1, the element M related to the disturbance w is added to one plate.
- the scaling matrix ⁇ is a ⁇ ⁇ ⁇ matrix.
- the element ⁇ , of the scaling matrix ⁇ is the weight for the ⁇ th manipulated variable UN.
- Each element TN ' is determined so that the magnitude of each gain of the control target model 11 is as equal as possible. More specifically, each element TN is determined as in the following equation. max G ylui G ylu2 G yluN
- G yLu is the transfer from the Nth manipulated variable u of the controlled object model 11 shown in Fig. 4 to the Lth controlled variable y.
- ma X (II G, II ⁇ II G yLu2 II « ,
- G Y LI is the H ⁇ norm II G, II II G Y LI, ⁇ , II GN II
- FIG. 9A shows the gain characteristic of the control target model 11 (frequency response characteristic of the model 11). Note that FIG.
- FIG. 9 shows only three types of gain characteristics for simplicity, but if the number of manipulated variables u is N and the number of control variables y is L, NXL types of gain characteristics are used. Inn exists.
- FIG. 9A when there is no scaling matrix T, it can be seen that the gains of the control target model 11 1 are not uniform. In general, when the gains of the controlled model are not uniform, the magnitudes of the errors of the controlled model also become uneven.
- the complementary sensitivity weight W is determined so as to cover the additive error ⁇ , the controller must be designed based on the one with a large error, and the obtained controller is very conservative, that is, excessively stable. It tends to be a typical thing. Therefore, the size of the gain is made uniform using the scaling matrix T.
- FIG. 9B shows the gain characteristic of the control target model 11 when the scaling matrix T of the present embodiment is provided. II G II ⁇ is the maximum value in each gain of model 11
- the method of determining the scaling matrix T of the present embodiment represented by the equations (24) and (25) is based on the gain maximum value II G II ⁇ ( The scheduling matrix T is determined so that each gain is aligned with the gain near the maximum value.
- the parameters of the control target model 11 are set in the control target model input unit 1 by the user of the design device.
- the control target model registration unit 2 registers the parameter input from the control target model input unit 1 in the mathematical expression of the control target model stored in the storage unit 3 in advance.
- the control target model input unit 1 and the control target model registration unit 2 provide a model setting procedure for setting the control target model. Constitutes a stage.
- the storage unit 3 stores the equations of the generalized plant of FIG. 4 described in Equations (1) to (15) and the equations of the control target model that is a part of this generalized plant.
- the frequency response calculation unit 4 converts the model 11 represented by the state equation expression registered in the storage unit 3 into a transfer function expression, and calculates a gain for each frequency from the transfer function. Subsequently, the scaling matrix calculation unit 5 calculates the scaling matrix T based on the gain calculated by the frequency response calculation unit 4 using the equations (24) and (25). Is output to the controller calculation unit 6.
- the controller calculation unit 6 calculates the parameters of the controller K by registering the scaling matrix T in the generalized plant equation stored in the storage unit 3 and performing an eye telecommunication. At this time, the complementary sensitivity weight W t and sensitivity weight W s and scaling matrix M, are preset in the generalized plant in the memory unit 3. Thus, the controller K can be designed.
- the scaling matrix T has been determined empirically.
- the scaling is performed so that each gain is aligned with the maximum gain value (more precisely, near the maximum gain value) of the control target model 11 based on the frequency response of the control target model 11.
- the scaling matrix T can be easily determined. This makes it easy to design a multivariable controller based on the H ⁇ control theory, which has excellent setpoint tracking characteristics and can be stabilized even if the control target fluctuates or the control target model 11 has an error. Become.
- a multivariable controller can be realized that takes advantage of the features of H ⁇ control that the computational load during control execution is light and can be implemented in a small-scale control system.
- the scaling matrix T is determined such that each gain is aligned with the maximum gain value (more precisely, near the maximum value) of the control target model 11.
- the gain of the model 11 The scaling matrix T may be determined such that each gain is equal to the minimum value or the average gain value.
- ma x in Equation (25) should be II G y u, II, II G y L u 2
- the above-mentioned max should be averaged in II G, II, II G II, ⁇ , 11 G y Lu , 'II ⁇ You can replace it with E to find the value.
- the disturbance w is not considered, but a controlled object model for the disturbance w may be obtained. Therefore, in the present embodiment, a method of determining the scaling matrix T in such a case while considering the influence of the disturbance w will be described. Also in the present embodiment, the general expression of the scaling matrix T can be expressed by Expression (24) as in the first embodiment.
- the element TN ′ of the scaling matrix T is determined as in the following equation.
- G yLw is a transfer function from the J-th disturbance WJ to the L-th control variable y of the controlled object model 12 shown in FIG. 4, and II G yLwj II »is the transfer function ⁇ norm (gain).
- G y I) is the H ⁇ norm II G y ,., II, II G y L w 2 II » ⁇ ⁇ ⁇ , II G, L wj II Means to select the maximum value among ⁇ .
- y wL is the output of the control target model 12 with respect to the disturbance w.
- H ⁇ norm II G y II ⁇ disturbance w, Ki de be determined in each control amount y, it can be determined element T 'of scaling matrix T from the formula (2 7).
- the operation of the scaling matrix T will be described with reference to FIG. Fig. 10A shows the gain characteristics of the controlled object model 12 (the frequency response characteristics of the model 12). Note that in Fig.
- FIG. 10B shows the gain characteristic of the control target model 11. As shown in FIG. 10B, it can be seen that the maximum gain value of the controlled object model 12 and each gain of the controlled object model 11 are not uniform.
- FIG. 10C shows the gain characteristic of the control target model 11 when the scaling matrix T of the present embodiment is provided.
- the method of determining the scaling matrix T of the present embodiment represented by Equations (24) and (27) is based on the maximum gain II Gywmax II ⁇ (more accurate The scaling matrix T is determined so that each of the gains of the model 11 is aligned with the gain near the maximum value.
- the scaling matrix T for the manipulated variable u is included in the closed-loop system when implementing the controller. Therefore, it is meaningful to make the magnitudes of the gains of the model 11 equal, and it is not always important where the gains from the operation amount u to the control amount y are made equal.
- the first embodiment described above shows one example of where the gains are aligned.
- the disturbance input is taken into account, from the viewpoint of suppressing the disturbance w, the influence of the input disturbance w needs to be suppressed by the manipulated variable u. There is. Therefore, in this embodiment, the worst case can be dealt with.
- the scaling matrix T is determined so that each gain of the model 11 is aligned with the maximum gain II G y wmax ax II (more precisely, near the maximum gain) of the model 12. Also in the present embodiment, the configuration as a design device is almost the same as in the first embodiment. Therefore, the operation of the design apparatus of the present embodiment will be described with reference to FIG.
- the parameters of the control target model are set in the control target model input unit 1 by the user of the design device.
- the control target model registration unit 2 registers the parameter input from the control target model input unit 1 in the mathematical expression of the control target model stored in the storage unit 3 in advance.
- the frequency response calculation unit 4 converts the models 11 and 12 represented by the state equation expression registered in the storage unit 3 into a transfer function expression, and calculates a gain for each frequency from the transfer function.
- the scaling matrix calculation unit 5 calculates the scaling matrix T based on the gain calculated by the frequency response calculation unit 3 using equations (24) and (27), and Output to calculation unit 6.
- the operation of the controller calculator 6 is exactly the same as in the first embodiment.
- the controller K can be designed.
- the first control is performed to set the maximum value in each gain of the second control target model 12.
- the scaling matrix T can be easily determined. This makes it possible to design a multivariable controller based on H ⁇ control theory that has excellent setpoint tracking characteristics and disturbance suppression, and can be stabilized even if the control target fluctuates or the control target model has errors. It becomes possible.
- the scaling matrix T is used to equalize the magnitudes of the model gains, and try to make the control weights for each control variable equal.
- the control for each control amount y may interfere with each other, causing problems such as instability of the control, and it may be necessary to adjust the control weight for each control amount. . Therefore, in this embodiment, a weight matrix S for directly weighting each control amount y is introduced.
- FIG. 11 shows a configuration of a controller designing apparatus according to a third embodiment of the present invention.
- FIG. 12 is a block diagram showing a configuration of the generalized plant in the present embodiment.
- the design device shown in FIG. 11 includes a control amount weight input unit 7 for inputting a weight for the control amount y and a control amount weight added to the design device of the first or second embodiment shown in FIG. And a weight matrix calculating section 9 for calculating a weight matrix S based on the weight of the control quantity.
- the generalized plant shown in FIG. 12 is obtained by adding a block (control amount weight adjusting means) 20 representing the weight matrix S to the generalized plant of the first embodiment or the second embodiment shown in FIG. .
- control amount weight adjustment means 20 (weight matrix S) is replaced by an operation amount weight adjustment means 16 (scaling matrix T), a controlled object model 11 and a controller K inside a closed loop system. Is provided.
- the general formula of the weight matrix S is shown below.
- the weight matrix S is an LXL matrix.
- the element SL of the weight matrix S is a weight for the L-th control amount y L.
- Each element SL is determined as follows.
- Equations (8), (9), and (10) are rewritten as
- Equation (34), Equation (35), and Equation (36) are represented by Doyle's notation as Equation (11), the parameter A in Equation (11) must be represented as follows: Can be.
- the sensitivity weight w s and the complementary sensitivity weight W t are designed, and the equation ( Multiplying by the output section of 1) and performing eye telemetry, the controller K is found in the state space representation.
- the actual controller implemented in a plant such as a distillation tower is
- the controller K is obtained by multiplying the weight matrix S, the weight a ⁇ ′ I and the scaling matrix T.
- the storage unit 3a includes the equations of the generalized plant of FIG. 12 described in Equations (1) to (5), Equation (11), Equations (32) to Equation (40).
- the equation of the control target model, which is a part of this generalized plant, is stored.
- the control amount weight W yL for the L-th control amount y L is set in the control amount weight input unit 7 by the user of the design apparatus.
- the setting of the control amount weight W yL is performed for each control amount y.
- the control amount weight registration unit 8 outputs the control amount weight W y input from the control amount weight input unit 7 to the weight matrix calculation unit 9.
- the weight matrix calculating unit 9 calculates the weight matrix S using the equations ( 29 ) and (30) based on the control weight WyL. This is calculated and output to the controller calculation unit 6a.
- the controller calculation unit 6a registers the scaling matrix T and the weight matrix S in the generalized plant equation stored in the storage unit 3a, and performs an iterative operation to obtain the parameters of the controller K. calculate.
- the complementary sensitivity weight W t and sensitivity weight W s and the scaling matrix M are preset in the generalized plant in the memory unit 3 a.
- the controller K can be designed.
- weighting can be directly performed for each control amount y by introducing the weight matrix S.
- a controller with higher control performance and higher stability can be designed.
- by introducing the weight matrix S it is not necessary to give the scaling matrix T a role of weighting the control amount y.
- the control amount weight adjusting means 20 (weight matrix S) is provided inside the closed loop system, but may be provided outside the closed loop system.
- FIG. 14 is a block diagram showing a configuration of a generalized plant according to the fourth embodiment of the present invention.
- the control amount weight adjusting means 20 (weight matrix S) is provided before the frequency sensitivity weight adjusting means 13.
- the method of determining the weight matrix S is exactly the same as the method of determining the weight matrix S of the third embodiment described in the equations (29) and (30).
- equation (4) is rewritten as the following equation.
- the night time DH, D, 2, D2D can be expressed as
- the configuration as a design device is almost the same as that of the third embodiment. Therefore, the operation of the design apparatus of the present embodiment will be described with reference to FIG.
- the operations of the control target model input unit 1, the control target model registration unit 2, the frequency response calculation unit 4, and the scaling matrix calculation unit 5 are exactly the same as in the first embodiment or the second embodiment.
- the storage unit 3a includes equations (1) to (3), equations (5) to (8), equations (10) to (13), equations (41) to (44) ), The equations of the generalized plant of Fig. 14 and the equations of the controlled object model that is part of this generalized plant are stored.
- the operations of the control amount weight input unit 7, the control amount weight registration unit 8, and the weight matrix calculation unit 9 are exactly the same as in the third embodiment.
- the controller calculation unit 6a registers the scaling matrix T and the weight matrix S in the generalized plant equation stored in the storage unit 3a, and performs a satellite telemetry to obtain the controller K. Is calculated. In this way, the controller K can be designed.
- control amount weight adjusting means 20 (weight matrix S) is provided before the frequency sensitivity weight adjusting means 13, but may be provided after the frequency sensitivity weight adjusting means 13. You may.
- the weight matrix S can be easily adjusted. Therefore, in the third embodiment and the fourth embodiment, the first embodiment or the fourth embodiment can be used. Explained in the second embodiment. It is a prerequisite to use the method of determining the scaling matrix T.
- FIG. 15 is a block diagram showing a configuration of a controller design device according to a fifth embodiment of the present invention.
- the design device shown in Fig. 15 includes a transient response parameter input unit 101 for inputting transient response parameters representing the transient response characteristics of a closed loop system consisting of a control target and a controller.
- Transient response parameters to be registered in the temporary storage unit 102 and a closed-loop transfer function calculation unit to calculate the transient response characteristics of the closed-loop system based on the transient response parameters input from the transient response parameter storage unit 102 103, a frequency sensitivity weight calculator 104 for calculating a frequency sensitivity weight for determining a set value follow-up characteristic of the closed loop system based on a transient response characteristic of the closed loop system, and a preset generalization
- a controller calculation unit 105 for applying a frequency sensitivity weight to the plant to derive parameters of the controller.
- the transient response parameter input section 101, the transient response parameter register section 102, and the closed loop transfer function calculation section 103 constitute setting means for setting the transient response characteristic of the closed loop system.
- FIG. 16 is a block diagram showing the configuration of a model in which the actual control target is expressed as a mathematical expression.
- the numerical model of the controlled object shown in Fig. 16 consists of a model 1 1 1 for the manipulated variable u and a model 1 1 2 for the disturbance w.
- P u is the transfer function of model 1 1 1
- P w is the transfer function of model 1 1 2.
- the models 111 and 112 are obtained as a result of performing model identification using data obtained from a step response test for an actual controlled object.
- the control amount y which is the output of the control target, is the sum of the outputs of the models 111 and 112.
- Figure 17 shows the configuration of a conventional generalized plant that includes such a numerical model of the controlled object.
- the generalized plant as shown in FIG. 1 7, a frequency weight called a sensitivity weight W s used to determine the set value tracking characteristic, and the complementary sensitivity weight that is used to determine the robust bets stability
- a frequency weight called, and introducing the set value r, output zl, and z2 in addition to the input (operation amount) u, input (disturbance) w, and output (control amount) y of the control target
- 1 13 is a block representing the sensitivity weight W s , and Z, is an output for evaluating the set value tracking characteristic. Further, 1 1 4 is a block representing the complementary sensitivity weight W,, Z 2 is Ru output der in order to evaluate the robustness bets stability.
- the uncertainty of the model is estimated based on the numerical model of the control target, the complementary sensitivity weight W is determined, and the frequency is calculated in consideration of the followability to the set value r.
- the characteristics were directly specified, the sensitivity weight W s was determined, and the parameters of the controller were determined by an iterative method.
- the controller will be designed based on the one with the largest gain due to the difference in gain with respect to the plant output for each manipulated variable. It tends to be very conservative, that is, overly stable.
- the controller can have an integral characteristic, but since the generalized plant is not stable, it cannot be reduced to the standard H ⁇ problem. .
- a generalized plant as shown in FIG. 18 is considered.
- M is a scaling matrix for adjusting the influence of disturbance w on the control variable y
- T is a scaling matrix for adjusting the magnitude of the error of the control target
- a'I eliminates the steady-state error.
- s is the Laplace operator, and a (> 0) is any real number.
- 1 1 6 is a block representing the scaling matrix T
- 1 1 7 is a block representing the weight alpha-1 I.
- the deviation e 2 is obtained by multiplying the deviation e by the weight ⁇ —'1, and is an input to the controller.
- Figure 19 shows the configuration of a robust control system that adds a controller ⁇ to the generalized plant described above.
- reference numeral 118 denotes a block representing the controller ⁇ .
- the control amount y which is the output of the control target, follows the set value r, the influence of the disturbance w is removed, and the control target fluctuates or the model of the control target is changed.
- the purpose is to determine the parameters of controller K so that it can be stabilized even if there is an error.
- the H ⁇ control problem can be considered as a problem that reduces the H ⁇ norm (gain) of the transfer function from (r, w) to (z,, z).
- H ⁇ norm gain
- each of the set value tracking characteristic, robust stability, and disturbance suppression can be considered as follows.
- d is a frequency weight for limiting the band to be followed (for example, to follow only the low band).
- a a Parameter B ,, B 2 can be expressed as Also, parameters Isseki C ,, C 2 can be expressed by the following equation L D a C p C a. (1 14)
- Controller K is a controller that solves the H ⁇ control problem using a generalized plant.
- the actual controller implemented in a plant such as a distillation tower is shown in Fig. 20
- the controller K is multiplied by the weight ⁇ -'1 and the scaling matrix ⁇ .
- ⁇ control theory is a design method in the frequency domain.
- the idea hardly control system design in the frequency domain, such as process control rather difficulty utilizing appropriately choosing the complementary sensitivity weight W t and sensitivity weight W s It is difficult.
- control design is performed based on a single model.However, in robust control design, fluctuations in the control target and the magnitude of modeling errors are added to the control design in advance, and these fluctuations and errors are considered. Even if there is a difference, design so that it is stable and the control performance does not deteriorate much.
- Figure 21 shows the additive error of the control target model 1 1 1.
- 1 19 is a block representing an additive error ⁇ .
- characteristic fluctuations of the controlled object due to operating conditions and the like, and model errors due to the reduced dimensions of the model 111 are expressed as an additive error ⁇ as shown in Fig. 21.
- the design is such that the controller output is stable even if the characteristics of the control target deviate from the model 111.
- the complementary sensitivity weights should be determined so as to cover the additive error ⁇ .
- the general formula of this complementary sensitivity weight is shown below. It should be noted that only the model 1 1 1 fluctuates because the change of the model 1 1 2 is not related to the stability of the system. And design
- the maximum value Gmax of the gain of the error ⁇ is multiplied by a safety factor ⁇ 5 ( ⁇ is, for example, 1) for the additive error ⁇ whose size is adjusted using the scaling matrix T.
- ⁇ 5 ⁇ is, for example, 1
- W , is the weight for the Nth manipulated variable u
- the scaling matrix ⁇ is a ⁇ X ⁇ matrix.
- the element ⁇ of the scaling matrix ⁇ is the weight for the ⁇ th manipulated variable u,.
- Each element T is determined such that the magnitude of each component of the additive error ⁇ is as equal as possible.
- the scaling matrix M is a JXJ matrix.
- the elements of the scaling matrix M are the weights for the Jth disturbance WJ.
- Each element ⁇ ” is an adjustment parameter that determines the disturbance suppression performance by adjusting the effect of each disturbance WJ on the control amount y.
- the sensitivity weight Ws is an LXL matrix.
- the element W s L of the sensitivity weight W s is a weight for the L-th control amount y L.
- Fig. 22 which is a simplified version of the robust control system in Fig. 19.
- reference numeral 111a denotes a block representing a numerical model P to be controlled
- reference numeral 113a denotes a block representing a frequency weight W s ′.
- S (s) is the sensitivity function that mainly shows the control performance related to quick response such as set value tracking and disturbance suppression
- a transfer function G yr (s) from the set value r of the closed-loop system shown in FIG. 22 to the control amount y is obtained as follows.
- Gy r — spec (s) c + 1 ⁇ ⁇ ⁇ (1 2 6)
- T s L is a time constant for the L-th control amount y.
- the frequency weight W SL ′ (s) for the L-th control amount y L is set as in the following equation. T sL s +
- the frequency weight W s L '(s) is obtained by multiplying a- 1 (s) by W SL (s) and is defined by the following equation.
- Equation (13 1) is the transfer function from the set value r of the closed-loop system shown in Fig. 22 to the deviation e. It shows that the H ⁇ norm of the weight a 1 (s) multiplied by WSL (s), the transfer function from r to z!
- This equation (13 1) is a design index of the controller K in consideration of the set value tracking characteristic. Therefore, by setting the frequency weights W s L '(s) as in equation (128), equation (131) is satisfied, and the design of controller K in consideration of the set-point tracking characteristic It is possible.
- equation (128) By transforming equation (128), the following equation is obtained.
- the first term on the right side of the equation (1 3 2) is ⁇ - 1 (s). Therefore, the element W s L (s) of the sensitivity weight W s can be calculated as follows.
- W sL (s) -3 ⁇ 4 133 (133)
- equation (12 1) can be expressed as follows.
- the transient response parameter that is, the time constant T s L is set in the transient response parameter input unit 101 by the user of the design equipment.
- the setting of the time constant T s L is performed for each control amount y.
- Transient response parameter Isseki registering unit 1 0 2 outputs a constant T SL when input from transient response parameter Isseki input unit 1 0 1 directly to the closed loop transduction function calculating section 1 0 3.
- the closed-loop transfer function calculation unit 103 substitutes the input time constant T s L into the equation (1 26) to obtain a transfer function G from the set value r of the closed-loop system shown in FIG. yr (s) is calculated and output to the frequency sensitivity weight calculator 104. Subsequently, the frequency sensitivity weight calculation unit 104 calculates the equation (125), the equation (127), the equation (128), and the equation (132) based on the transfer function G > r (s). ) To calculate the sensitivity weight W s using the equation (1 34), and output this to the controller calculation section 105.
- the storage unit 106 stores the equations of the generalized plant of FIG. 18 described in the equations (101) to (115).
- the controller calculation unit 105 registers the sensitivity weight W s in the generalized plant equation stored in the storage unit 106 and performs an iterative operation to determine the parameters of the controller K. calculate. At this time, the complementary sensitivity weight W, and the scaling matrices T, M are set in advance in the generalized plant of the storage unit 106. Have been. Thus, the controller K can be designed.
- the frequency sensitivity weight W s can be calculated based on the transient response characteristic.
- This makes it possible to design controllers based on the H ⁇ control theory even in fields where it is difficult to provide frequency response characteristics as control specifications, such as process control. As a result, it becomes easy to design a multivariable control system that takes into account the fluctuations of the control target and the uncertainty of the numerical model.
- a controller can be realized that makes use of the features of H ⁇ control, in which the computational load at the time of control execution is light and can be implemented in a small-scale control system. Also, by approximating the transient response characteristics of the closed-loop system with the first-order lag characteristics, the parameters for control design can be intuitively understood by the designer, realizing a design device that is easy to understand and use by the designer. can do. In addition, since the parameters can be intuitively understood by the designer intuitively, even if the design is changed once after designing, it is possible to realize a design device that can be easily changed.
- the sensitivity weight W s was determined by approximating the transient response characteristic of the closed-loop system with a first-order lag and specifying the closed-loop time constant for each control amount.
- the design is performed by approximating the characteristics of a closed-loop system with a commonly used secondary system and designating its transient response characteristic parameters.
- FIG. 23 shows an example of the time response of the closed-loop system represented by equation (140).
- FIG. 23 shows the state of the control amount y when a step-like set value r of 100% at time 0 is given.
- the time required for the controlled variable y to reach the same value as the set value r (here 100%) is It is the time until the overshoot 0S, which is the first extreme value of the transient deviation taken after the certain rise time tr and the control amount y exceeds the set value r, and the time until the control amount y reaches the overshoot amount 0S
- the overshoot time tp and the settling time t until the controlled variable y falls within the range of 5% of the set value r.
- There is a reduced ⁇ DR is the ratio of a 2 a, a shown in FIG 3.
- Rise time t r can be expressed by the following equation by using the ⁇ damping factor and natural frequency [omega [pi. 1
- the overshoot amount 0 S can be obtained by the following equation using the attenuation coefficient ⁇ . Then, the damping ratio DR can be obtained as follows:
- G yr (s) (1 4 6) s 2 + 2 ⁇ o n s + ⁇ ⁇ as in the fifth embodiment, setting the frequency weight W s' (s) is as follows, formula (1 3 1) is satisfied, and the controller K can be designed in consideration of the setpoint tracking characteristics.
- Equation (147) can be transformed into the following equation.
- the first term on the right side of the equation (148) is a- 1 (s). Therefore, the element W s L (s) of the sensitivity weight W s can be calculated as follows. By substituting the equation (149), the equation (122) can be expressed as follows.
- OS L is the overshoot amount with respect to the L-th control amount y L.
- natural frequency w nL than the rise time t r shown can be obtained at the cormorants good of the following equation. ⁇ one cos 1
- Transient response parameter Isseki i.e. the rise and the time t r L overshoot OS
- Isseki input unit 1 0 1. Setting the overshoot OS L and rise time t r L is performed for each controlled variable y.
- the transient response parameter registering unit 102 in the present embodiment outputs the overshoot amount OS and the rise time t ⁇ input from the transient response parameter input unit 101 directly to the closed-loop transfer function calculating unit 103. I do.
- the degree of freedom of design can be increased by approximating the transient response characteristics of the closed loop system with the characteristics of the secondary system, and the application range of the controller obtained by the design device can be increased. Can be spread.
- the first to sixth embodiments are design apparatuses for designing a multivariable controller. Further, the design apparatuses of the first to sixth embodiments can be realized on a computer. That is, the computer includes an arithmetic device, a storage device, and an input / output device, and operates as the above-described design device according to a program.
- the present invention is suitable for designing a multivariable controller.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Feedback Control In General (AREA)
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP00964640A EP1139188A4 (en) | 1999-10-05 | 2000-10-04 | CONTROLLER DESIGN DEVICE |
KR1020017007057A KR20010086084A (ko) | 1999-10-05 | 2000-10-04 | 컨트롤러의 설계장치 |
US09/857,482 US7092856B1 (en) | 1999-10-05 | 2000-10-04 | H-infinity controller design using control object models |
AU75558/00A AU7555800A (en) | 1999-10-05 | 2000-10-04 | Design device of controller |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP28397899A JP3779511B2 (ja) | 1999-10-05 | 1999-10-05 | コントローラの設計装置 |
JP11/283978 | 1999-10-05 | ||
JP11/283981 | 1999-10-05 | ||
JP28398199A JP3779512B2 (ja) | 1999-10-05 | 1999-10-05 | コントローラの設計装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2001025861A1 true WO2001025861A1 (fr) | 2001-04-12 |
Family
ID=26555266
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2000/006898 WO2001025861A1 (fr) | 1999-10-05 | 2000-10-04 | Dispositif concepteur controleur |
Country Status (6)
Country | Link |
---|---|
US (1) | US7092856B1 (ja) |
EP (1) | EP1139188A4 (ja) |
KR (1) | KR20010086084A (ja) |
CN (1) | CN1267795C (ja) |
AU (1) | AU7555800A (ja) |
WO (1) | WO2001025861A1 (ja) |
Families Citing this family (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7499763B2 (en) * | 2005-07-20 | 2009-03-03 | Fuel And Furnace Consulting, Inc. | Perturbation test method for measuring output responses to controlled process inputs |
CN100517135C (zh) * | 2005-11-18 | 2009-07-22 | 鸿富锦精密工业(深圳)有限公司 | 自动控制模拟系统及自动控制模拟方法 |
US7436617B2 (en) * | 2006-06-21 | 2008-10-14 | Broadcom Corporation | Controller, state estimator and methods for use therewith |
KR101044078B1 (ko) * | 2008-12-11 | 2011-06-23 | (주)현대공업 | 골조체 고정 및 배기형 금형 |
DE102010007560B4 (de) * | 2010-02-10 | 2014-05-15 | Deutsches Zentrum für Luft- und Raumfahrt e.V. | Regelungsverfahren und Regelungseinrichtung |
US8682453B2 (en) * | 2010-06-04 | 2014-03-25 | The Mathworks, Inc. | Interactive system for controlling multiple input multiple output control (MIMO) structures |
US8606375B2 (en) | 2010-06-04 | 2013-12-10 | The Mathworks, Inc. | Interactive control of multiple input multiple output control structures |
US9157950B2 (en) | 2011-04-18 | 2015-10-13 | International Business Machines Corporation | Loop parameter sensor using repetitive phase errors |
US8493113B2 (en) | 2011-09-12 | 2013-07-23 | International Business Machines Corporation | PLL bandwidth correction with offset compensation |
CN102455660A (zh) * | 2011-12-26 | 2012-05-16 | 浙江工业大学 | 基于数字h∞pid控制器的的连续时滞系统控制方法 |
CN106647252A (zh) * | 2016-09-26 | 2017-05-10 | 华南理工大学 | 电磁驱动微镜的h∞控制方法及系统 |
CN110850712B (zh) * | 2018-08-20 | 2023-08-22 | 富士电机株式会社 | 控制装置的设计装置及设计方法 |
US11407467B2 (en) * | 2018-09-14 | 2022-08-09 | Honda Motor Co., Ltd. | Stable balance controller |
US11421653B2 (en) | 2020-11-13 | 2022-08-23 | General Electric Renovables Espana, S.L. | Systems and methods for multivariable control of a power generating system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07253804A (ja) * | 1994-03-14 | 1995-10-03 | Canon Inc | 制御装置 |
JPH0854906A (ja) * | 1994-08-09 | 1996-02-27 | Toshiba Corp | 非干渉pid制御系の設計方法および設計装置 |
WO1996031811A1 (en) * | 1995-04-03 | 1996-10-10 | Honeywell Inc. | A method of optimal scaling of variables in a multivariable predictive controller utilizing range control |
JPH08328605A (ja) * | 1995-06-05 | 1996-12-13 | Toyota Motor Corp | 制御装置及びその製造方法、並びにクラッチのスリップ制御装置およびその製造方法 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB8727602D0 (en) * | 1987-11-25 | 1987-12-31 | Nat Res Dev | Industrial control systems |
USH1410H (en) * | 1993-05-17 | 1995-01-03 | The United States Of America As Represented By The Secretary Of The Navy | H-infinity controller for an electro-mechanical actuator |
JPH09269804A (ja) | 1996-03-29 | 1997-10-14 | Aisin Seiki Co Ltd | 自動制御系の安定制御装置 |
US6230062B1 (en) * | 1999-01-08 | 2001-05-08 | Voyan Technology | Adaptation to unmeasured variables |
-
2000
- 2000-10-04 AU AU75558/00A patent/AU7555800A/en not_active Abandoned
- 2000-10-04 WO PCT/JP2000/006898 patent/WO2001025861A1/ja active Application Filing
- 2000-10-04 KR KR1020017007057A patent/KR20010086084A/ko active Search and Examination
- 2000-10-04 US US09/857,482 patent/US7092856B1/en not_active Expired - Fee Related
- 2000-10-04 CN CNB008021678A patent/CN1267795C/zh not_active Expired - Fee Related
- 2000-10-04 EP EP00964640A patent/EP1139188A4/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07253804A (ja) * | 1994-03-14 | 1995-10-03 | Canon Inc | 制御装置 |
JPH0854906A (ja) * | 1994-08-09 | 1996-02-27 | Toshiba Corp | 非干渉pid制御系の設計方法および設計装置 |
WO1996031811A1 (en) * | 1995-04-03 | 1996-10-10 | Honeywell Inc. | A method of optimal scaling of variables in a multivariable predictive controller utilizing range control |
JPH08328605A (ja) * | 1995-06-05 | 1996-12-13 | Toyota Motor Corp | 制御装置及びその製造方法、並びにクラッチのスリップ制御装置およびその製造方法 |
Non-Patent Citations (5)
Title |
---|
HOKUJO, SHIMIZU, SONE, MIHIRA, FURUTA: "Kodo Seigyo no Joryu Process no Tekiyou; Benchmark Study", SAVEMATION REVIEW, YAMATAKE HONEYWELL K.K., vol. 15, no. 2, 1 August 1997 (1997-08-01), pages 2 - 13, XP002935816 * |
MIHIRA, KOGA, MORINAGA, HOKUJO: "Advanced Seigyo wo Ikani Jitsugen suruka; Seigyo shuhou no Sentei to Seinou Hikaku: H00 Robust Seigyo", KEISOU, QUGEN KAISHA KOGYO GIJUTSUSHA, vol. 41, no. 4, 1 April 1998 (1998-04-01), pages 53 - 58, XP002935815 * |
MOHAMMED DAHLEH, ALBERTO TESI, ANTONIO VICINO: "Robust Stability/Performance of Interconnected Interval Plants with Structured Norm Bounded Perturbations", PROC. OF THE 31ST CONFERENCE ON DECISION AND CONTROL, IEEE, December 1992 (1992-12-01), pages 3169 - 3174, XP002935818 * |
See also references of EP1139188A4 * |
T. D. MORPHOPOULOS, M. G. SAFONOV: "Convexity of diagonally-scaled infinity norm optimal control problems - the square, one-sided case", PROC. OF THE 27TH CONFERENCE ON DECISION AND CONTROL IEEE, December 1988 (1988-12-01), pages 1341 - 1342, XP002935817 * |
Also Published As
Publication number | Publication date |
---|---|
CN1327551A (zh) | 2001-12-19 |
EP1139188A1 (en) | 2001-10-04 |
CN1267795C (zh) | 2006-08-02 |
KR20010086084A (ko) | 2001-09-07 |
EP1139188A4 (en) | 2006-02-01 |
AU7555800A (en) | 2001-05-10 |
US7092856B1 (en) | 2006-08-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2001025861A1 (fr) | Dispositif concepteur controleur | |
Vu et al. | Analytical design of fractional-order proportional-integral controllers for time-delay processes | |
Shi et al. | Robust design of integrated feedback and iterative learning control of a batch process based on a 2D Roesser system | |
Söderlind | Automatic control and adaptive time-stepping | |
Xia et al. | Adaptive output feedback dynamic surface control of nonlinear systems with unmodeled dynamics and unknown high-frequency gain sign | |
Alhajeri et al. | Tuning guidelines for model-predictive control | |
Kevickzy et al. | Two-degree-of-freedom control systems: The Youla parameterization approach | |
Jafari et al. | Adaptive type-2 fuzzy backstepping control of uncertain fractional-order nonlinear systems with unknown dead-zone | |
Cheng et al. | Adaptive finite-time command-filtered control for switched nonlinear systems with input quantization and output constraints | |
Mandić et al. | A new optimisation method of PIDC controller under constraints on robustness and sensitivity to measurement noise using amplitude optimum principle | |
Hoseini et al. | Robust adaptive control of nonlinear non-minimum phase systems with uncertainties | |
Kokunko et al. | Cascade synthesis of differentiators with piecewise linear correction signals | |
Ghogare et al. | Design and Analysis of Decentralized Dynamic Sliding Mode Controller for TITO Process. | |
Maciejowski | Reverse-engineering existing controllers for MPC design | |
Pekař | On Simple Algebraic Control Design and Possible Controller Tuning for Linear Systems with Delays | |
JP3779511B2 (ja) | コントローラの設計装置 | |
Gaikwad et al. | Multivariable frequency-response curve fitting with application to control-relevant parameter estimation | |
RU2510956C2 (ru) | Способ адаптивного управления по пид-закону и система для его реализации | |
Osinuga et al. | An iterative algorithm for maximizing robust performance in loop‐shaping design | |
JP2001109505A (ja) | コントローラの設計装置 | |
JPH11506553A (ja) | ダイナミックプロセスのモデル化装置 | |
Dehnert et al. | Design of Robust Output Feedback Controllers for Discrete-Time Systems with Polynomial Uncertainties | |
Chen | Galerkin approximationwith Legendre polynomials for a continuous-time nonlinear optimal control problem | |
Engell | Design of robust control systems with time-domain specifications | |
Béthoux | Linear Controllers: PID controller design |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 00802167.8 Country of ref document: CN |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CR CU CZ DE DK DM DZ EE ES FI GB GD GE GH GM HR HU ID IL IN IS KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2000964640 Country of ref document: EP Ref document number: 1020017007057 Country of ref document: KR |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 09857482 Country of ref document: US |
|
WWP | Wipo information: published in national office |
Ref document number: 2000964640 Country of ref document: EP |
|
REG | Reference to national code |
Ref country code: DE Ref legal event code: 8642 |