JP2002258905A - Control rule deciding method for automatic controller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the control rule deciding method of an automatic controller capable of simultaneously realizing robust stability and desired response characteristics. SOLUTION: This device is provided with a controller 26 for controlling an object 28 to be controlled so that a follow-up error based on a difference between a norm output from a norm model 23 corresponding to a command value u and an observed output from the object 28 to be controlled can be reduced, and an object to be enlarged and an enlargement controller are formulated based on the object 28 to be controlled and the controller 26, and the parameter of the enlargement controller capable of maximizing a robust stable margin is decided for the object to be enlarged so that the control rule of the controller 26 can be decided. Thus, it is possible to simultaneously realize the robust stability and desired response characteristics, and to realize the control rule of the controller in a simple constitution with a few parameters.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control rule determining method for simultaneously realizing robust stability and desired response characteristics of an automatic control device for controlling an aircraft, a ship, a robot, and the like.

[0002]

2. Description of the Related Art FIG. 8 is a block diagram showing a configuration of a control device 1 according to the prior art for realizing both robust stability and desired response characteristics. The control device 1 includes a feedforward controller 2, a robust feedback controller 3, and an adder 4. The feedforward controller 2 processes the command value and outputs the command value to the adder 4 so that the control target 5 has a desirable response to the command value. The robust feedback controller 3 receives an observation output from the control target 5 and outputs a signal that suppresses disturbance acting on the system and achieves robust stabilization of the system. The adder 4 inputs to the control target 5 an operation amount obtained by adding the command value processed by the feedforward controller 2 and the value of the signal from the robust feedback controller 3. H∞
The robust feedback controller 3 is designed using the control theory, and the feed forward controller 2 is adjusted while confirming the response characteristics, and is separately designed. In this way, robust stability and desired response characteristics can be simultaneously realized.

FIG. 9 is a block diagram showing a configuration of a robust flight control device 11 disclosed in Japanese Patent Application Laid-Open No. 5-147589. The robust flight control device 11 includes a first constant gain unit 12, a second constant gain unit 13, a state estimator 14,
And an adder 15. Constant gain unit 12, 1
3 is designed by the “ExactModel Matching method” so that the transfer characteristic of the closed loop system by the state feedback of the controlled object 16 matches the transfer function that becomes the desired response characteristic. The first constant gain unit 12 processes the command value and outputs it to the adder 15. The state estimator 14 is designed using H∞ control theory, and determines the state of the controlled object 16 and the disturbance applied to the controlled object 16 based on the operation amount input to the controlled object 16 and the observation output of the controlled object 16. Simultaneously estimate the second
Output to the constant gain unit 13. The adder 15 adds the value of the signal output from the first constant gain unit 12 and the value of the signal output from the second constant gain unit 13 and outputs the result to the control target 16 and the state estimator 14. I do.

In this way, the state estimator 14 outputs a signal equivalent to the state feedback to the second constant gain unit 13 to realize a desired response characteristic, and cancels disturbance to improve robust stability. it can.

[0005]

The robust feedback controller 3 in the control device 1 described above is designed based on the H∞ control theory, so that the control law is very complicated, and the feedforward controller 2 includes the control target 5 and the robust Since the design must be made in consideration of both of the feedback controller 3, the control law is further complicated. Also, the state estimator 14 in the above-described robust flight control device 11 is designed based on the H∞ control theory, so that the control law is very complicated.

As described above, in the prior art, when the robust stability and the desired response characteristic are to be realized at the same time, the control law becomes complicated. Therefore, when the controller is mounted on the control device, a high degree of control is required. Not only is computational capability required, but also it is difficult to associate control law parameters with control performance, making it difficult to fine-tune parameters in the final process of design.

In the case where the characteristics of the controlled object change depending on the operating conditions of the controlled object, such a conventional technique requires H∞.
Although the controller's robust stability guarantees a certain degree of stability against characteristic fluctuations, it is difficult to design the performance clearly for multiple characteristics, and post-design analysis of control performance at points other than the nominal point. And increase the number of trial and error of the design.

Accordingly, an object of the present invention is to provide a method for determining a control law of an automatic control device which can simultaneously realize robust stability and desired response characteristics for a plurality of characteristics using a simple control law. It is to be.

[0009]

The present invention according to claim 1 relates to a method for obtaining a control law of a controller which simultaneously realizes a robust stability and a desired response characteristic. A controller Cw that controls the control target P so as to reduce a tracking error based on a difference between a reference response output from a reference model defining a desired response characteristic with respect to the value and a response actually output from the control target P. Formulation of the expansion control target Pw based on the control target P, and formulation of the expansion controller C based on the controller Cw,
This is a control law determination method for an automatic control device, characterized in that a parameter of the expansion controller C that maximizes a robust stability margin with respect to the expansion control target Pw is determined to obtain a control law of the controller Cw.

According to the present invention, the control target P is controlled so as to reduce a tracking error based on the difference between the reference response output from the reference model and the response actually output from the control target P with respect to the command value. Is provided, and the enlarged control object Pw is formulated based on the control object P, and the enlarged controller C is formulated based on the controller Cw,
The parameter of the expansion controller C that maximizes the robust stability margin with respect to the expansion control target Pw is determined, and the control law of the controller Cw is obtained. Accordingly, the transfer characteristic with respect to the tracking error between the command value and the output from the controller Cw is related to the robust stability margin, and the transfer characteristic can be reduced by increasing the robust stability margin. And the desired response characteristics can be realized at the same time, and the control law of the controller Cw can be simplified with a small number of parameters.

According to a second aspect of the present invention, there is provided a method for obtaining a control law of a controller which simultaneously realizes a robust stability and a desired response characteristic, wherein the controller has a command value of a control target P having a plurality of characteristics. A common controller Cw that controls the control target P so as to reduce a tracking error based on a difference between a reference response output from a reference model in which a desired response characteristic is specified and a response actually output from the control target P Formulating a plurality of extended control objects Pw based on the control object P, and formulating the extended common controller C based on the common controller Cw, so that the robust stability margin for the enlarged control object Pw is maximized. This is a control law determination method for an automatic control device, characterized in that parameters of an extended common controller C are determined to obtain a control law of the common controller Cw.

According to the present invention, the control target P is controlled so as to reduce a tracking error based on the difference between the reference response output from the reference model and the response actually output from the control target P with respect to the command value. A common controller Cw is provided, and a plurality of extended control objects Pw are formulated based on a control object P having a plurality of characteristics, and an extended common controller C is formulated based on the common controller Cw, Expansion control target Pw
, The parameter of the expanded common controller C that maximizes the robust stability margin is determined, and the control rule of the common controller Cw is obtained. As a result, the transfer characteristic with respect to the tracking error between the command value and the output from the common controller Cw is related to the robust stability margin, and the transfer characteristic can be reduced by increasing the robust stability margin. It is possible to simultaneously improve the performance and the desired response characteristics, and to make the control law of the common controller Cw a simple configuration with few parameters.

[0013]

DETAILED DESCRIPTION OF THE INVENTION Linear matrix inequalities
inequality; abbreviated name: LMI) A design procedure of a robust controller based on a loop shaping method will be described. In order to facilitate understanding, a symbol representing a transfer characteristic such as a control law of the control target and the control device is used by attaching to the control target and the control law. First, the enlargement control target Pw is represented by the following equation 1.

[0014]

(Equation 1)

The left graph of the enlargement control object Pw

(Equation 2) far. Here, G〜, M〜, and N〜 indicate a dynamic system, and Equation 2 indicates that the dynamic system G〜 is expressed by the following linear differential equations of Equations 3 and 4.

[0016]

(Equation 3)

Using the same notation, the enlargement control rule C is expressed by the following equation (5).

(Equation 4)

The right inverse graph K of the expansion control law C is

(Equation 5) far. Here, element A of right inverse graph K of enlargement control law C
_K, B _K is a given in advance. The indefinite elements in the right inverse graph K of the expansion control law C of Equation 6 are C _K1 ,
C _K2 , D _K1 , and D _K2 . Further, Q is defined as in the following Expression 7 using the normalized right inverse graph K _N of the expansion control law C.

[0019]

(Equation 6)Here, * on the right shoulder of the symbol indicates conjugate transposition. Expansion control law C
Right inverse graph K _NIs obtained as in the following Expression 8.

[0020]

(Equation 7)

Here,

(Equation 8) It is. T at the right shoulder of the symbol indicates transposition of the matrix. Equation 1
X at 0 is a symmetric positive definite solution of the Riccati equation of the following equation (11).

[0022]

(Equation 9)

Where R in equation 11 is

(Equation 10) It is.

Using the elements of Equations 2, 6, and 7 above,
Matrix A of the following Expressions 13 to 16_GK, B _GK, C_GK, D_GKSet
Justify.

[0025]

[Equation 11]

Using the matrices A _GK , B _GK , C _GK , and D _GK ,
A convex optimization problem expressed by a linear matrix inequality of the following equation 17;
Q and K are obtained by solving the algebraic expression of the following expression 18 alternately,
Indefinite elements C _K1 , C of right inverse graph K of optimal expansion control law C
_K2, it is possible to obtain the D _K1, D _K2.

[0027]

(Equation 12)

Minimizing γ in equation (17) is equivalent to a closed loop robust stability margin b _PwC including the designed expansion control law C.
Is known to be maximized, and the indefinite elements C _K1 , C _K2 , D of the right inverse graph K of the extended control law C that optimizes robust stability by the LMI loop shaping method described above. _K1 and _DK2 can be obtained. The robust stability margin b _PwC is defined by the following equation (19).

[0029]

(Equation 13)

The enlargement control law C actually used is given in advance.
Element A_K , B_K And determined by the above method
Element C of right inverse graph K of expanded control law C_K1, C_K2,
D_K1, D _K2And can be obtained by the following equation (20).

[0031]

[Equation 14]

FIG. 1 is a flowchart showing a procedure for determining parameters of the expansion control law C by the above-described LMI loop shaping method. The procedure is started in step s0, and proceeds to step s1. In step s1, initial values of the enlargement control target Pw and the enlargement control rule C are prepared, and the process proceeds to step s2. In step s2, the enlargement control target Pw
And the right inverse graph K of the enlargement control law C is obtained, and the process proceeds to step s3. In step s3, a normalized right inverse graph K _N of the enlargement control law C is obtained from Expressions 8 to 12, and
Q is obtained from Expression 7, and the process proceeds to Step s4.

In step s4, Q is fixed, and
Equation 17 is solved using Equations (16) to (16) to determine optimal indefinite elements C _K1 , C _K2 , D _K1 , D _K2, and γ of _K, and obtains step s5
Proceed to. In step s5, K is fixed, and equations 8 to 1 are set.
2. Q is obtained by solving Expression 18 using Expression 20, and the process proceeds to Step s6. In step s6, it is determined whether or not γ has converged. If it is determined that γ has converged, the process proceeds to step s7. If it is determined that γ has not converged, the process returns to step s4. In step s7, an element of the expansion control law C is obtained from Expression 20 using the element of K obtained when γ has converged, the process proceeds to step s8, and all procedures are ended.

The above-described LMI loop shaping method is also effective for designing a common control law for a plurality of controlled objects and a controlled object having a plurality of characteristics. For example, a case will be described in which an enlarged common control rule C for the first enlarged controlled object Pw1 and the second enlarged controlled object Pw2 is designed.
The left graph of these two enlarged control objects Pw1 and Pw2 is
As in Equation 2,

(Equation 15) far. The right inverse graph K of the enlarged common control rules C and C for the two enlarged controlled objects Pw1 and Pw2 is expressed by Expressions 5 and 6 in the same manner as described above. Also, using the left graphs G ₁ to G ₂の of the enlarged control objects Pw ₁ and Pw ₂ and the normalized right inverse graph K _N of the enlarged common control law C, Q ₁ and Q ₂ are calculated by 23 and Equation 24.

[0035]

(Equation 16)

Using the elements of Equation 6 and Equations 21 to 24, the matrices A _G1K , B _G1K , C of the following Equations 25 to 32 are used.
_{G1K, D G1K, A G2K,} B G2K, C G2K, defines a D _G2K.

[0037]

[Equation 17]

Matrices A _G1K , B _G1K , C _G1K , D _G1K , A
_G2K, B _G2K, C _G2K, the use of D _G2K, by solving the algebraic expression of the formula 33 of LMI convex optimization problems and the following equation 34 is represented by alternately to Q _1, Q _2, K Thus, the optimum indefinite elements C _K1 , C _K2 , D _K1 , and D _K2 of the common expansion control law C can be obtained.

[0039]

(Equation 18)

Although the above-described method for obtaining the common control rule is for the case where there are two control targets, the common control rule can be obtained in exactly the same manner when there are three or more control targets.
As described above, by using the LMI loop shaping method, a control law for optimizing robust stability can be obtained, but the desired response characteristic, which is another object of control law design, cannot be realized as it is.

A method for optimizing robust stability and obtaining a control law for realizing desired response characteristics by using the LMI loop shaping method will be described.

FIG. 2 is a block diagram showing a configuration of a control device 21 to which the control law determining method of the present invention is applied. The control device 21 includes an input unit 22, a reference model 23, a subtractor 2
4. It comprises an adjustment gain unit 25, a controller 26 and an adder 27, and controls a control target 28.

The input means 22 is operated by an operator of the control device 21 to output a signal having a command value u. The reference model 23 is a signal processor that outputs a reference output that is a desired observation output of the control target 28 for a certain command value u. For example, in an aircraft, the input means 22 is a pilot device, the command value u is an input to the pilot device of the pilot, the command value u is added to the controller output v and input to the control object 28, and the reference model 23 is also input. The reference output output from the reference model 23 is a desirable value for a command value such as an attitude angle change and an attitude angle measured by a sensor of an airframe in an aircraft, for example.

The subtractor 24 is a computing unit that outputs a difference between the reference output output from the reference model 23 and the observation output output from the control object 28. The adjustment gain unit 25 is a gain for adjusting the degree of realization of a desired response characteristic, and outputs a tracking error e obtained by multiplying a difference between the reference output output from the subtractor 24 and the observation output by a predetermined constant. The controller 26 is a controller to which the control law determination method of the present invention is applied. The controller 26 robustly stabilizes the control device 21 based on the tracking error e output from the adjustment gain unit 25, and sets the observation output as a standard. A correction signal having a controller output v which is a command value to be matched with the output is output. The adder 27 is an arithmetic unit that adds the command value u and the controller output v and outputs the result to the control target 28.

The closed loop transfer characteristics from the command value u and the controller output v to the following error e are as follows: P is the transfer characteristic of the controlled object 28, H is the transfer characteristic of the reference model 23, and W is the value of the adjustment gain unit 25. , 35.

[0046]

[Equation 19]

If the transfer characteristic of the controller 26 is Cw, the following equation 36 is established.

(Equation 20)

The right sides of Equations 35 and 36 are respectively

(Equation 21) In other words, the closed loop transfer characteristic from the command value u and the controller output v to the following error e is expressed by the following equation 39.

[0049]

(Equation 22)

The transfer characteristic (I-PwC) expressed by equation 39
Reducing ^-1 Pw is to reduce the tracking error e. That is, the response characteristic represented by the reference model 23 can be realized by reducing the transfer characteristic (I-PwC) ^-1 Pw. Further, since the transfer characteristic (I-PwC) ^-1 Pw expressed by Expression 39 is included in the upper right element on the right side of Expression 19, the robust stability margin b _{PwC of} Expression 19
Is maximized, the H∞ norm on the right-hand side of Equation 19 becomes smaller, so that the transfer characteristic (I-PwC) ^{-1 of} Equation 39 is obtained.
Pw can be reduced. That is, the control target 28
Transfer characteristic P of the controller 26 to be designed
By defining the transfer characteristic Pw of the enlargement control object and the transfer characteristic C of the enlargement controller as Expressions 37 and 38 using w, the robust stability is optimized by applying the LMI loop shaping method. The controller 26 that achieves the desired response characteristics can be determined.

FIG. 3 is a block diagram showing a configuration of a control device 31 to which the control law determining method of the present invention is applied. The control device 31 includes an input unit 32, a reference model 33, a subtractor 3
4. It is composed of an adjustment gain unit 35 and a controller 36, and controls a control target 37. The input unit 32, the reference model 33, the subtractor 34, the adjustment gain unit 35, and the controller 36 are the input unit 2 in the control device 21 described above.
2, the same as the reference model 23, the subtractor 24, the adjustment gain unit 25, and the controller 26, respectively. In control device 31, command value u output from input means 32
Is input only to the reference model 33, and only the controller output v from the controller 36 is input to the control target 37.

The closed loop transfer characteristics from the command value u and the controller output v to the following error e are as follows: P is the transfer characteristic of the controlled object 37, H is the transfer characteristic of the reference model 33, and W is the value of the adjustment gain unit 35. , Expressed by the following equation (40).

[0053]

(Equation 23) Assuming that the transfer characteristic of the controller 36 is Cw, Expression 36 is established.

The right side of Equation 40 is

(Equation 24) If the right side of Expression 36 is expressed by Expression 38, the controlled object 37
Characteristics P of the controller 36 to be designed
By defining the transfer characteristic Pw of the enlargement control object and the transfer characteristic C of the enlargement controller using Expression w and Expression 41 and Expression 38, respectively, the LMI loop shaping method is applied to optimize the robust stability. In addition, it is possible to find a controller 36 that achieves a desired response characteristic.

In the control devices 21 and 31 described above, the expression 3
7. When the value W of the adjustment gain devices 25 and 35 included in the enlargement control target Pw of Expression 41 is increased, the controller 26,
The tracking error e input to 36 is amplified, and the controllers 26 and 36 designed in this way are controllers that place importance on tracking performance, but the robust stability is degraded.
Further, when the value W of the adjustment gain units 25 and 35 is reduced, the tracking performance is degraded, but the controller is focused on robust stability.

When a common control law is designed for a plurality of control objects in the above-described control devices 21 and 31, the transfer characteristic Pw of the enlarged control object for each control object is calculated by the following equation (37).
Alternatively, the controller can be determined using the LMI loop shaping method, as defined in Equation 40.

FIG. 4 is a block diagram showing a configuration of a flight control device 41 including a controller to which the control law determination method of the present invention is applied. The control purpose of the flight control device 41 is a high-speed design point (altitude 3048 m (10000 f
t), the ratio of flight speed to sound speed, Mach M = 0.
9) and a low speed design point (altitude 6096m (20,000f)
t) and Mach M = 0.45) to stabilize the body Pa, and make the roll rate response to the roll operation and the skid angle response to the yaw operation follow the reference response. Also, the sideslip angle response to roll maneuver and the roll rate response to yaw maneuver shall be minimized.

The flight control device 41 includes an input means 42, a first
Reference model H _roll , second reference model H _yaw , first filter L _p , second filter L _β , washout filter L _r , first to fifth adjustment gain units W ₁ , W ₂ , W ₃ , W _4.
, W _5, the roll axis controller C _p, yaw controller C _beta, aileron servo controller C _.delta.a, yaw rate controller C _r, rudder servo controller C
_δr , first subtractor 43, second subtractor 44, first adder 4
5 and a second adder 46. Input means 42
Is operated by a pilot who is an operator of the flight control device 41 to output a signal having a command value.

[0059] The first reference model H _roll, the input unit 42 is operated by the pilot, for the role maneuver u _p which is output from the input unit 42, the roll rate nominal response is desirable roll rate response of the airframe Pa Output I do. The transfer characteristic of the first reference model H _roll is defined by the following equation 42.

[0060]

(Equation 25) Here, s indicates a Laplace operator.

The first subtractor 43 has a first reference model H _roll
Of the roll rate reference error e _p , which is the difference between the roll rate reference response output from the controller and the value of the roll rate signal p output from the sensor attached to the body Pa, and outputs the result to the first filter L _p I do. First filter L _p
Improves the phase characteristics of the closed loop system. The first adjustment gain unit W ₁ multiplies the roll rate following error e _p passed through the first filter L _p by a constant W ₁ and outputs the result to the roll axis controller C _p .

The second reference model H _yaw outputs a side slip angle reference response, which is a desired side slip angle response of the body Pa, to the yaw maneuver u _β output from the input means 42 when the input means 42 is operated by the pilot. I do. Transfer characteristic of the second reference model H _yaw is defined by the following equation 43.

[0063]

(Equation 26)

The second subtractor 44 _{generates a} second reference model H _yaw
Outputs the sideslip angle standard response outputted, the second filter L _beta to calculate the sideslip angle tracking error e _beta which is the difference between the value included in the side slip signal beta output from the sensor attached to the machine body Pa from . The second filter L _beta, to improve the phase characteristic of the closed loop system. The second adjustment gain unit W ₂ multiplies the side slip angle tracking error e _β passed through the second filter L _β by a constant W ₂ and outputs the result to the yaw axis controller C _β .

The washout filter _Lr is connected to the body P
The stationary component of the yaw rate signal r output from the sensor attached to a is removed. Third adjustment gain device W ₃ being a yaw rate signal r output from the sensor attached to the airframe Pa passing through the washout filter L _r, and outputs the yaw rate controller C _r is multiplied by a constant W _3.

The fourth adjustment gain device W ₄ multiplies a signal δ _a for detecting displacement of the aileron control surface output from a sensor attached to the body Pa by a constant W ₄ and outputs the result to the aileron servo controller C _δa . I do. Fifth adjustment gain unit W ₅
Is the signal [delta] _r of the displacement detected in the ladder control surface which is output from the sensor attached to the body Pa, and outputs the ladder servo controller C _{[delta] r} is multiplied by a constant W _5.

The first adder 45 outputs the output from the roll axis controller C _p and the aileron servo controller C _δa
And outputs the aileron operation command value _δac to the body Pa. The second adder 46 outputs the output from the yaw axis controller C _beta, the output from the yaw rate controller C _r, the rudder servo controller C ladder operation command value [delta] _rc of an output obtained by adding from _δr to the fuselage Pa. The aileron operation command value δ _ac is an operation command value for an actuator that operates the aileron control surface, and the rudder operation command value δ _rc is an operation command value for an actuator that operates the rudder control surface.

The body Pa moves according to the movement of each control surface operated based on the operation command values δ _ac and δ _rc , and the body Pa
A sensor attached to the sensor detects the movement.

The first filter L _p , the second filter L _β and the washout filter L _r for improving the phase characteristics
Is determined as in the following Expressions 44 to 46.

[0070]

[Equation 27]

Enlarged control object P in flight control device 41
w and the expansion controller C are given by
Equation 48 is obtained.

[0072]

[Equation 28]

In Equation 47, W, H, and P shown on the lower side of the right side correspond to W, H, and P on the left side of Equation 41. The P ₁ of the right side of expression 47 is an element relating to the roll rate and the side slip angle of the motion characteristics of the aircraft Pa, P
Reference numeral ₂ denotes an element relating to the yaw rate and the response of each actuator in the motion characteristics of the body Pa.

After the formula 47 is applied to both the high-speed design point and the low-speed design point of the airframe Pa, the common controllers C _p and C of the high-speed design point and the low-speed design point are formed according to the LMI loop shaping method. _{β, C r, C δa,} C
_{When δr} is calculated, it is obtained as in the following Expressions 49 to 53.

[0075]

(Equation 29) Here, j represents an imaginary unit.

FIG. 5 shows a flight control device 51 for a small aircraft including a controller to which a conventional robust design method is applied.
FIG. 3 is a block diagram showing the configuration of FIG. The calculation result of the control law by the conventional method is described in the Transactions of the Society of Instrument and Control Engineers, Vol. 30, No. 7,
Nos. 54 to 67 described on pages 767 to 775
It is. In this conventional robust design method, a controller is designed with the body Pa at the high-speed design point as a control target,
The low-speed design point is covered by the robust stability of the controller.

[0077]

[Equation 30]

[0078]

(Equation 31)

[0079]

(Equation 32)

In such a conventional robust design method,
As shown in FIG. 5, it is difficult to provide the flight control device 51 with a clear configuration, and the transfer characteristics of the controller shown in Expressions 54 to 67 use the present invention shown in Expressions 49 to 53. Higher order and more complex than controllers. When the transfer characteristics of the controller are high-order and complicated, not only does the flight control computer for calculating the control law of the controller mounted on the aircraft require high computational power, but also there are many parameters to be adjusted. That means
Tuning the control law by fine-tuning the parameters becomes difficult. Since the flight control device 41 using the present invention has a clear configuration and a small number of parameters as compared with the conventional flight control device 51, calculations can be performed using a flight control computer having a low calculation capability. At the same time, tuning by fine adjustment of parameters is easy.

[0081] Figure 6, the flight control system 41 designed using the present invention, a graph illustrating a simulation result of the response when a roll maneuver u _p and the yaw steering u _beta inputted to the step input by pilot is there. 7, the flight control device 51 designed by the conventional robust design method, which is a graph showing simulation results of response when the roll maneuver u _p and the yaw steering u _beta was step input. The flight control device 41 designed by using the present invention and the flight control device 51 designed by the conventional robust design method follow the desired response characteristics, and the control law designed by using the present invention is comparable in performance. It turns out there is no.

[0082]

According to the present invention, the improvement of robust stability and the desired response characteristic can be realized at the same time, and the control law of the controller Cw has a simple configuration with few parameters. be able to. By making the control law of the controller Cw a simple configuration with a small number of parameters, the amount of calculation when controlling the control target P can be reduced, so that a computer with low calculation capability can be used, and Tuning of the control law by fine adjustment can be easily performed, and the number of trial and error of design can be reduced.

According to the second aspect of the present invention, it is possible to simultaneously improve the robust stability and the desired response characteristics, and to make the control law of the common controller Cw a simple configuration with few parameters. Can be. Common controller C
By making the control law of w a simple configuration with few parameters, the amount of calculation when controlling the control target P can be reduced, so that a computer with low calculation capability can be used, and fine adjustment of the parameters can be performed. Can easily tune the control law, and the number of trial and error of the design can be reduced.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a procedure for determining a parameter of a control law C by an LMI loop shaping method.

FIG. 2 is a block diagram showing a configuration of a control device 21 including a controller 26 to which a control law determining method of the present invention is applied.

FIG. 3 is a block diagram showing a configuration of a control device 31 including a controller 36 to which the control law determination method of the present invention is applied.

FIG. 4 is a block diagram showing a configuration of a flight control device 41 including a controller to which the control law determination method of the present invention is applied.

FIG. 5 is a block diagram showing a configuration of a small aircraft flight control device 51 including a controller to which a conventional robust design method is applied.

In flight control apparatus 41 designed using the [6] The present invention, it is a graph illustrating a simulation result of the response when a roll maneuver u _p and the yaw steering u _beta was step input.

FIG. 7 is a flight control device designed by a conventional robust design method.
In the position 51, the roll steering u _p And yaw control u_β 
Simulation of the response when
It is a graph which shows a result.

FIG. 8 is a block diagram showing a configuration of a control device 1 according to the related art that simultaneously realizes robust stability and desired response characteristics.

FIG. 9 is a block diagram showing a configuration of a robust flight control device 11 disclosed in Japanese Patent Application Laid-Open No. 5-147589.

[Explanation of symbols]

21 and 31 the control device 23, 33 the reference model 24 and 34 the subtractor 26 and 36 a controller 28 and 37 controlled object 41 flight control system 43,44 subtractor H _roll, H _yaw reference model _{C p, C β, C r} , C _δa , C _δr controller Pa body

Claims (2)

[Claims] 1. A method for determining a control law of a controller that simultaneously realizes robust stability and a desired response characteristic, wherein the controller outputs an output from a reference model in which a desired response characteristic to a command value of a control target P is defined. Normative response to be
A controller Cw for controlling the controlled object P so as to reduce a tracking error based on a difference between a response actually output from the controlled object P and a control object Pw based on the controlled object P. The expansion controller C is formulated based on the controller Cw, and the parameters of the expansion controller C that maximize the robust stability margin with respect to the expansion control target Pw are determined.
A control rule determination method for an automatic control device, wherein the control rule is determined. 2. A method for obtaining a control law of a controller which simultaneously realizes robust stability and a desired response characteristic, wherein the controller defines a desired response characteristic for a command value of a control target P having a plurality of characteristics. A common controller Cw for controlling the control target P so as to reduce a tracking error based on a difference between a reference response output from the reference model and a response actually output from the control target P; A plurality of extended control targets Pw are formulated using the common controller Cw, and the extended common controller C is formulated based on the common controller Cw. A method for determining a control law of an automatic control device, wherein a parameter is determined to obtain a control law of a common controller Cw.