JP5769378B2 - Parameter determining apparatus and method, and program - Google Patents

Description

  The present invention relates to a parameter determination apparatus and method, and a program.

  Conventionally, for example, a traveling body such as a submersible has been provided with a plurality of controllers such as pitch control and depth control, and it is known to travel while switching these controllers. When switching the controller, it is necessary to give a predetermined initial value to the controller after switching. Conventionally, the initial value has been calculated using an arithmetic expression previously held by each controller.

JP 2003-121306 A

Conventionally, two arithmetic expressions are mainly known as arithmetic expressions for obtaining the initial value. When the initial value is set using one arithmetic expression, the switching operation can be performed smoothly, but it is known that followability is inferior, and the initial value is set using another arithmetic expression. In such a case, it is known that the followability is improved, but the switching operation cannot be performed smoothly.
Conventionally, the controller has either one of the arithmetic expressions, and since the initial value at the time of control switching is set using this arithmetic expression, the smoothness of the switching operation and the control followability are simultaneously achieved. There is a problem that an initial value that satisfies the condition cannot be given.

  The present invention has been made to solve the above problems, and an object of the present invention is to provide a parameter determination apparatus, method, and program capable of adjusting the characteristics of control after switching at the time of switching of a controller. .

In order to solve the above problems, the present invention employs the following means.
The present invention provides a system comprising a plurality of controllers, for switching the controller performing processing, a parameter determination device for determining an initial value of the state variables used by the controller after the switching, smooth switching operation and opposite said tracking sexual and control, along with a combination of a plurality of arithmetic expressions having the opposite characteristics, includes a parameter calculation arithmetic expression that can be changed by the user weighting characteristics against said phase, wherein The weighting of the parameter calculation formula is determined according to the state of the controller before switching, and the initial value in the controller after switching is determined using the parameter calculation formula given the determined weight. A characteristic parameter determination device is provided.

According to such a configuration, the parameter calculation arithmetic expression is combined plurality of arithmetic expression having the opposite characteristics, and are made changeable weighting conflicting features. When the controller is switched, the weights of the plurality of arithmetic expressions having the above conflicting characteristics in the parameter calculation arithmetic expression are determined according to the state of the controller before the switching, and such weighting is adopted. The parameter (for example, initial value) used in the controller after switching is determined using the calculated parameter calculation formula. Thus, since the parameter calculation calculation formula is configured by combining a plurality of calculation formulas having conflicting features, it is possible to set parameters (for example, initial values) having various features. Furthermore, by adjusting the weighting of each calculation formula constituting the parameter calculation calculation formula according to the control state of the previous controller, it is possible to adjust how to operate each feature, and parameters according to the situation (for example, Initial value) can be determined.
In addition, by combining a plurality of arithmetic expressions having conflicting characteristics in this way, a wider range of characteristics can be provided.

The parameter calculating operation equation in the parameter determination device, a plurality of arithmetic expression having the opposite characteristics respectively expressed by a polynomial, a plurality of arithmetic expression having the opposite characteristic to the linear combination, the respective terms of the polynomial It is good also as being characterized by being an arithmetic expression which provided the coefficient which can be changed.

  In this way, the weighting of features can be easily changed by changing the coefficient set for each term of the polynomial.

The present invention provides a system comprising a plurality of controllers, for switching the controller performing processing, a parameter determination method for determining the initial values of the state variables used by the controller after the switching, smooth switching operation and opposite said tracking sexual and control, along with a combination of a plurality of arithmetic expressions having the opposite characteristics, includes a parameter calculation arithmetic expression that can be changed by the user weighting characteristics against said phase, wherein Parameter determination for determining the initial value in the controller after switching by using the parameter calculation calculation formula to which the weighting of the parameter calculation calculation formula is determined according to the state of the controller before switching and the determined weight is given Provide a method.

The present invention provides a system comprising a plurality of controllers, for switching the controller performing processing, a parameter determination program for determining the initial values of the state variables used by the controller after the switching, smooth switching operation and opposite said tracking sexual and control, along with a combination of a plurality of arithmetic expressions having the opposite characteristics, includes a parameter calculation arithmetic expression that can be changed by the user weighting characteristics against said phase, wherein The weighting of the parameter calculation formula is determined according to the state of the controller before switching, and the initial value in the controller after switching is determined using the parameter calculation formula given the determined weight. A parameter determination program for causing a computer to execute is provided.

  According to the present invention, it is possible to adjust the characteristics of the control after switching when switching the controller.

It is the block diagram of an example which showed the function of the controller switching system which concerns on embodiment of this invention. It is the figure which showed an example of the table which set the coefficient which a parameter calculation formula has according to the last state of a controller. It is the block diagram which showed an example of the command value calculation part. It is the figure which showed the characteristic of the switching method at the time of setting an initial value using the method 1 and the method 2. FIG. It is the figure which showed the operation | movement flow of the controller switching system which concerns on embodiment of this invention. It is a figure for demonstrating the switching method of a controller at the time of applying the controller switching system which concerns on embodiment of this invention to an underwater vehicle.

  Hereinafter, an embodiment of a parameter determination device according to the present invention will be described with reference to the drawings. In the present embodiment, a parameter determination device applied to a controller switching system that controls switching of three controllers will be described as an example. In the present embodiment, there are three controllers to be controlled by the controller switching system, but the number of controllers is not particularly limited. In the present embodiment, a controller provided in a traveling body such as a submersible will be described as an example, but the present invention is not limited to this. For example, it may be provided in a vehicle such as an automobile.

FIG. 1 is a block diagram showing a schematic configuration of a controller switching system 1 that controls switching of a plurality of controllers according to the present embodiment.
The controller switching system 1 is a system that controls switching of three controllers A, B, and C, and includes an initial parameter setting unit 10, a data reading unit 11, a controller determination unit 12, a parameter determination device 13, and a command. A value calculation unit 14, a first previous value setting unit 15, a second previous value setting unit 16, and a calculation continuation determination unit 17 are provided.

The initial parameter setting unit 10 sets a pre-switching flag Cflag_old that indicates a controller that was used before the controller was switched (hereinafter referred to as “controller before switching”). For example, when the controller switching system 1 is initially activated, the controller has not been switched and there is no controller before switching, so the pre-switching flag Cflag_old indicates “0” indicating initial activation. Is set. In the case of initial start, the controller A, B, and the control amount of C is set to control the amount of _startup, and _{y aold = y a, y bold} = y b, y cold = y c.

The initial parameter setting unit 10 outputs Cflag_old, y _aold , y _bold , and y _cold to the data reading unit 11.

  The data reading unit 11 reads parameters in the controller before switching (hereinafter referred to as “parameters before switching”). This pre-switching parameter is required to calculate a command value for the controller after switching the controller (hereinafter referred to as “switched controller”). Further, the information of the controller selected as the controller after switching is read and set to the controller flag Cflag indicating the controller after switching. The pre-switching parameter acquired by the data reading unit 11 is data input from the initial parameter setting unit 10 or the calculation continuation determination unit 17.

Specifically, the parameter before switching required to calculate the command value is, for example, Cflag_old, the value u _{s of} the sensor that detects the operation amount in the controller selected as the controller after switching, the control amount Target value r _a , and control amounts y _aold , y _bold , y _cold, etc. of the controllers A, B, and C, respectively. In the data reading unit 11, the controller flag Cflag selected as the controller after switching is, for example, 1 when the controller is A, 2 when the controller is B, or the controller C. Is set to a numerical value such as 3.

  The controller pre-switch parameter and the controller flag Cflag read into the data reading unit 11 in this way are output to the controller determination unit 12.

  Based on the controller flag Cflag, the controller determination unit 12 identifies the controller selected as the controller after switching, and determines whether or not the controller has been switched. Specifically, the pre-switching flag Cflag_old acquired from the data reading unit 11 is compared with the selected controller flag Cflag to determine whether Cflag_old ≠ Cflag is satisfied. If not established, the controller is not switched, so the pre-switching parameter acquired from the data reading unit 11 is output to the command value calculation unit 14 as it is. In addition, since the controller is switched when it is established, the pre-switching parameter acquired from the data reading unit 11 and the fact that this is established are indicated by the command value calculation unit 14 and the parameter determination device 13. Output to.

  For example, Cflag_old = 0 because it is an initial activation, and Cflag = 1 when the selected controller is the controller A, Cflag_old ≠ Cflag. Thereby, the controller determination part 12 determines with the switching of a controller having been performed, Cflag_old = 0 and Cflag = 1 with respect to the command value calculation part 14 and the parameter determination apparatus 13, and the parameter before switching Output.

  The parameter determination device 13 includes a parameter calculation formula. The parameter calculation calculation formula is composed of a combination of a plurality of calculation formulas having different features, and the weighting of different features can be changed. This weighting can be set by the coefficient of the parameter calculation formula. In addition, the parameter determination device 13 has a table that associates the relationship between the previous controller state (before switching) and the weighting coefficient corresponding to this state, and when switching of the controller is detected. In addition, the weighting coefficient corresponding to the state of the controller before switching is referred to from the table and output.

When the parameter determination device 13 acquires from the controller determination unit 12 that the controller has been switched (Cflag_old = 0 and Cflag = 1), the parameter determination device 13 obtains a weighting coefficient corresponding to the previous (before switching) controller state. The parameter calculation formula is constructed using this weighting coefficient. Furthermore, the initial value used for the command value of the controller A after switching calculated by the command value calculation unit 14 (for example, the controller A based on Cflag = 1) using this parameter calculation formula. Is calculated. The weighting coefficients are Kap1ini, Kad1ini, Kap2ini, and Kad2ini shown in the following equation (6).

  The parameter calculation calculation formula is a combination of a plurality of calculation formulas having conflicting characteristics. For example, a parameter calculation formula is a formula in which a plurality of formulas having different characteristics are represented by polynomials, and a plurality of formulas having different characteristics are linearly combined, and a changeable coefficient is provided for each term of the polynomial. It is a formula. In the present embodiment, it is assumed that two arithmetic expressions of method 1 and method 2 are given as a plurality of arithmetic expressions having conflicting characteristics.

Hereinafter, the polynomials and characteristics of the method 1 and the method 2 included in the parameter determination device 13 will be described.
If the initial value used in the command value calculation unit 14 (details will be described later) for calculating the command value of the controller after switching is x _aold , the initial value x _aold calculated based on the method 1 is (1). It is shown as an expression. Here, in equation (1), u _s is the amount of operation of the sensor in selected control unit as the control unit after the switching, r _a is the target value, y _a control amount, e _a control target value r _a deviation between the amount _{y a, K ap1} and _{K ap2} is a proportional _{gain, K ad1} and _{K ad2} is a derivative gain.

Thus, the initial value x _Aold calculated based on the method 1, because it is calculated on the basis of the deviation e _a between the target value r _a control amount y _a, the control in consideration of the deviation e _a is performed . That is, when the command value calculation unit 14 in the _subsequent stage calculates the command value of the controller after switching based on x _aold in the equation (1), control is performed in consideration of deviation when switching is performed based on this command value. . Specifically, the switching operation is performed smoothly, but the switching operation is performed such that the followability deteriorates.

On the other hand, if the initial value calculated based on method 2 is x _aold , x _aold is expressed as in equation (5).

Using the initial value x _aold calculated based on method 2 that is the above equation (5), the command value of the controller after switching is calculated in the command value calculation unit 14 at the subsequent stage. As a result, when the controller is switched based on this command value, the switching followability is improved, but the switching operation is not smooth.

Subsequently, the parameter determination device 13 linearly combines the above-described polynomial of the method 1 and the polynomial of the method 2 and provides a changeable coefficient in each term of the polynomial. Specifically, the above equation (1) and the above equation (5) are linearly combined, a coefficient is provided for each term, and the following equation (6) is obtained. _{Here, x Aold} initial value, _{u s} is the amount of operation of the sensor in selected control unit as the control unit after the _{switching, K ap1} and _{K ap2} is a proportional _{gain, K ad1} and _{K ad2} is a derivative gain.

Further, the changeable coefficients provided in the respective terms of the polynomial are K _ap1ini , K _ad1ini , K _ap2ini , and K _ad2ini, which are the above-described weighting coefficients. In the present embodiment, each coefficient is set to 0 or more and 1 or less, and may be a non-integer value. In addition, all the coefficients may be the same value or different values.

Here, r _a is the target value, y _a control amount, e _a deviation between the target value and the controlled variable, e _Aold deviation of the controlled variable of the controller A before the controller is switched to the controller A, DT is a sampling period, K _ap1 and K _ap2 are proportional gains, and K _ad1 and K _ad2 are differential gains.

Also, the coefficients K _ap1ini , K _ad1ini , K _ap2ini , and K _ad2ini provided in each term of the polynomial are set by the user from the outside via an input device, for example. When the weights of these coefficients are changed, the way of using the feature of the command value u obtained from x _aold calculated by the equation (6) is changed. For example, four coefficients _{K ap1ini, K ad1ini, K ap2ini} , by setting all the _{K Ad2ini} 1, when the the _{x Aold} equal _{x Aold} (in other words calculated by scheme 1, switching is performed based on the _{x Aold} is Is smooth, but the follow-up property is deteriorated).

_{Incidentally,} K _Ap1ini, the _{K Ad1ini} set to _0, setting K _Ap2ini, the _{K Ad2ini} to 1 and is the _{x Aold} equal _{x Aold} (in other words calculated by scheme 2, switching is performed based on the _{x Aold} is , Followability is improved, but not smooth). Further, when each coefficient is set to a non-integer value of 0 or more and 1 or less (for example, all four coefficients are set to 0.5), a feature obtained by mixing the features of the method 1 and the method 2 (that is, It is smooth to some extent and has good followability to some extent). In this way, by adjusting the weights of the coefficients K _ap1ini , K _ad1ini , K _ap2ini , and K _ad2ini given in the above equation (6), it is possible to adjust how to use the characteristics when switching the controller.

In addition, the coefficients K _ap1ini , K _ad1ini , K _ap2ini , and K _ad2ini of such parameter calculation formulas are set according to the previous controller state, and are given by, for example, a table as shown in FIG. . FIG. 2 shows that in the case where the underwater vehicle is switched from the controller A to the controller B, the coefficient is set according to whether the underwater vehicle is infiltrated or ascended. Specifically, as shown in FIG. 2, when switching from the controller A to the controller B, the coefficients K _ap1ini , K _ad1ini , K _ap2ini , K _{ad2ini at} the time of infiltration are respectively α ₁ , α ₂ , α ₃ and α _{4 are} assumed, and the coefficients K _ap1ini , K _ad1ini , K _ap2ini , and K _ad2ini when _rising are β ₁ , β ₂ , β ₃ , and β ₄ , respectively.
Further, when the controller A is a controller that performs pitch control and the controller B is a controller that performs depth control, α ₁ <β ₁ , α ₂ <β ₂ , α ₃ > β ₃ , More preferably, α ₄ > β ₄ .

  Note that the arithmetic expressions having a plurality of features included in the parameter determination device 13 according to the present embodiment are linearly combined, and the coefficients provided in the terms of the linearly combined arithmetic expressions are set by the user. However, it is not limited to this. For example, the coefficient provided in each term may be set depending on the magnitude of a predetermined index.

As described above, the parameter determination device 13 calculates the initial value x _aold and outputs it to the command value calculation unit 14.

The command value calculation unit 14 calculates a command value u for the controller after switching. For example, the calculation of the command value as shown in FIG. 3 is shown in block diagram, the command value u is calculated from the basis of the IP-PD control with the target value r _a control amount y _a.

As shown in FIG. 3, the difference e _a between the target value r _a and the controlled variable y _a is integrated by the PID controller, given a proportional gain, and is differentiated by giving a differential gain, are respectively added . Further, from this result, _a command value u is obtained by subtracting a value obtained by giving _a proportional gain to the control amount ya and a value obtained by giving _a differential gain to the control amount ya.

More specifically, the command value u by the command value calculation unit 14 is calculated by the equation (15) based on the following equations (11) to (14).

In the above equation (15), the initial value x _aold calculated based on the above-described equation (6) is used. In this way, the command value u is calculated by the command value calculator 14 and the command value u is output to the controller A after switching. Further, the calculation result is output to the first previous value setting unit 15.

  Next, the controller switching system according to the present embodiment will be described more specifically. Here, FIG. 4 shows a case where control in the operation of the underwater vehicle is switched from pitch angle control (for example, controller B that performs pitch angle control) to depth control (for example, controller A that performs depth control). Will be described. For example, a case where the pitch angle is controlled to an angle of −10 degrees and the pitch angle controller is switched to the depth controller at the timing shown in FIG. 4D will be described.

When the controller is switched, the initial value x _aold of depth control (controller A), which is the controller after switching, is calculated by the above equation (6). For example, in FIG. 4, the controller is switched based on the initial value x _aold calculated by setting all the coefficients K _ap1ini , K _ad1ini , K _ap2ini , and K _ad2ini in equation (6) to “1”. and if, (6) the coefficients a is _{K Ap1ini,} "0" is set to the _{K Ad1ini,} coefficient _{K Ap2ini,} controlled on the basis of the _{K Ad2ini} set to the calculated initial value _{x Aold} to "1" The case where the device is switched will be described.

When “1” is set for all the coefficients in the equation (6), the calculation formula is the same as the above equation (1), so the graph in FIG. The coefficient _{K Ap1ini,} "0" is set to the _{K Ad1ini,} coefficient _{K Ap2ini,} if you set the "1" to the _{K Ad2ini,} since the same calculation formula as the equation (5), the graph of FIG. 4 Is shown as scheme 2 line.

  FIG. 4A shows a change in depth due to control based on the initial value calculated from method 1 and method 2. FIG. FIG. 4B shows a change in the pitch angle by the control based on the initial value calculated from the method 1 and the method 2. FIG. 4C shows a change in the steering angle by the control based on the initial value calculated from the method 1 and the method 2.

As shown in FIGS. 4A and 4C, the method 2 line has better followability than the method 1 line, but the fluctuation of the horizontal rudder angle is large, and the switching is not smooth. FIG. 4E is a diagram showing the magnitude of the depth deviation between the method 1 line and the method 2 line in time series. As shown in FIG. 4 (e), the method 2 follows the target value faster than the method 1. In this way, the initial value x _aold can be adjusted by changing the weighting of the coefficients, so that it is possible to easily adjust how the features of each method are used according to the situation. In FIG. 4, an integer of 0 or 1 is given to the coefficient. However, by setting a numerical value of 0 or more and 1 or less, the degree of advantage of the feature can be adjusted. Switching can be performed with little mixing.

The first previous value setting unit 15 sets the previous value (pre-switching parameter) when the next calculation is performed in the selected controller as the second previous time value x _aold , which is an integral value of the deviation e _aold and the deviation e _aold. Output to the value setting unit 16. For example, when the controller after switching is the controller A, e _aold which is the deviation before switching of the controller A is set to e _a for the next calculation, and the integral x _a of the deviation e _a X _aold = x _a is set as a value before switching, and is output to the second previous value setting unit 16.

The second previous value setting unit 16 sets the control amount, the pre-switching flag Cflag_old, and the pre-switching command value u _old when the next calculation is performed regardless of the selected controller. Specifically, the second previous value setting unit 16 sets y _old = y _a , y _bold = y _b , y _cold = y _c , Cflag_old = Cflag, and u _old = u, and these pre-switch parameters Is output to the calculation continuation determination unit 17.

  The calculation continuation determination unit 17 determines whether or not to continue the calculation. If the calculation is continued, the information output from the first previous value setting unit 15 and the second previous value setting unit 16 is sent to the data reading unit 11. Output. If the calculation is not continued, the process is terminated.

  In the controller switching system according to the present embodiment, when the controller determination unit 12 determines that the controller is switched, the coefficient of the parameter calculation arithmetic expression that the parameter setting device 13 has is: Although it was supposed to change according to the state of the last controller, it is not limited to this. For example, in the case of switching from the controller A to the controller B, the coefficient given to the controller B is varied according to the state of the controller A, and in the case of switching from the controller B to the controller A. Depending on the combination of controllers to be switched, such as a unique coefficient, it may be determined whether to vary the coefficient according to the previous controller state.

Next, the operation of the controller switching system 1 according to the present embodiment will be described with reference to FIG. First, a case where the controller A is selected after the controller switching system 1 is initially activated will be described.
When the controller switching system 1 is activated, the controller A, controller B, and controller C included in the controller switching system 1 are activated, and the pre-switching flag Cflag_old is set to “0” in the initial parameter setting unit 10. Is done. In addition, the control amount y _aold = y _a of the controller A before switching is set, and y _bold = y _b and y _cold = y _c are similarly set for the controller B and the controller C (step SA1). This setting way been _{Cflag_old = 0, y aold = y} a, y bold = y b, y cold = y c is output to the data reading unit 11.

In the data reading unit 11, Cflag_old = 0, y _aold , y _bold , and y _cold are acquired. Further, when the data reading unit 11 recognizes that the controller A is selected, the controller flag Cflag is set to Cflag = 1. Pre-switching parameters such as Cflag_old = 0, y _aold , y _bold , y _cold , and Cflag = 1 are output to the controller determination unit 12 (step SA2).

The controller determination unit 12 determines whether or not Cflag_old ≠ Cflag. If Cflag_old ≠ Cflag, it is determined that the controller has been switched (step SA3). Further, based on the fact that Cflag = 1, the selected controller is identified as the controller A. When it is determined that the controller has been switched, the parameter determination device 13 obtains a weighting coefficient based on the state of the controller before switching (before the controller A is selected) (step SA4). The initial value x _aold used for the command value of the controller A is determined using a parameter calculation formula that employs the acquired weighting coefficient. The initial value x _aold calculated here is output to the command value calculation unit 14 (step SA5). If the controller is not switched, the pre-switching parameter is output as it is to the command value calculation unit 14 (step SA3).

The command value calculation unit 14 calculates a command value u using the initial value x _aold based on a predetermined parameter calculation formula and outputs it to the controller A after switching (step SA6). In the first preceding value setting unit 15, when the controller A is selected, the deviation e _a between the target value r _a and the controlled variable y _a is set to e _Aold for the next operation, the deviation the integral _{x a,} is set to _{x Aold} for the next operation (step SA7). In the second previous value setting unit 16, the control amount in each controller is set for the next calculation, and y _aold = y _a , y _bold = y _b , y _cold = y _c , Cflag_old = Cflag, u _old = u is set (step SA8).

When such calculation is completed, the calculation continuation determination unit 17 determines whether or not to continue the calculation. If the calculation is continued, the calculation is set by the first previous value setting unit 15 and the second previous value setting unit 16. The previous value is output to the data reading unit 11. If the calculation is to be stopped, this process is terminated (step SA9). If the controller determination unit 12 determines that the controller has not been switched, the process proceeds to the command value calculation unit 14 without calculating the initial value xa. Process.
Then, by repeating the above-described series of processing, when the controller is switched, the switching is detected, and the initial value in the controller after switching is calculated by the same method.

  Next, the switching method of the controller of the underwater vehicle according to the present embodiment will be described in more detail with a specific example. In the following explanation, when the underwater vehicle is navigating at a depth of 100 meters, there is a command for the operator to infiltrate to a depth of 150 meters, and then a command to ascend from a depth of 150 meters to 100 meters. A method of switching the controller of the underwater vehicle that is sometimes performed will be described with reference to FIG.

First, the controller switching method when the underwater vehicle intrudes from a depth of 100 meters to 150 meters will be described.
When the underwater vehicle is in the vicinity of a depth of 100 meters, the depth controller (for example, the controller B) is controlled to maintain the depth in water at 100 meters, and the data reading unit 11 before switching The flag is set as Cflag_old = 2.

At time t ₁ in FIG. 6, when the target value r _a to infiltrate the underwater vehicle on the depth 150 m is input by the operator, the target value r _a is input to the data reading unit 11. The current value and the depth difference is a predetermined value between the target value r _a (e.g., 20 meters) when it is higher, the pitch controller from the depth controller (controller B) by a control unit (not shown) (e.g., controller A ) Is switched to. Then, in the data reading unit 11, a controller flag Cflag = 1 indicating that the pitch controller (controller A) has been selected as the controller after switching is input.

The controller determination unit 12 determines that the controller has been switched because Cflag_old ≠ Cflag. When the controller is switched, the parameter determination device 13 determines the initial value of the pitch controller (controller A) from the table held by itself to determine the initial value of the depth controller. The weighting coefficients γ ₁ , γ ₂ , γ ₃ , γ ₄ of the pitch controller (controller A) corresponding to the state of (controller B) are acquired. The parameter determination device 13 determines the initial value x _aold of the pitch controller (controller A) based on the parameter calculation formula owned by itself and the coefficients γ ₁ , γ ₂ , γ ₃ , γ ₄ acquired from the table. calculate. Here, since the depth control state is considered as a case where the depth is constant, the coefficients are uniquely γ ₁ , γ ₂ , γ ₃ , γ ₄ , but are not limited thereto. For example, it is good also as changing a coefficient according to whether it is beyond a predetermined depth.

In the command value calculating section 14, based on the initial value x _Aold and the target value r _a like command value u is calculated with respect to the pitch controller (controller A) for causing infiltrate the underwater vehicle to a depth 150 meters The A control unit (not shown) controls the pitch controller (controller A) to make the depth 150 meters based on the calculated command value u. When the control is continued, the current value such as the calculated command value u is output to the data reading unit 11 as the previous value, and the data reading unit 11 waits for the controller instruction information.

At time t ₂ in FIG. 6, when the underwater vehicle is infiltrated 20 meters before the depth 130 m of the target depth 150 m, the pitch controller by a control unit (not shown) (controller A) from the depth controller (control Controller B) is switched over. As a result, the controller flag Cflag = 2 indicating that the selected controller is the controller B is input to the data reading unit 11. When the controller determination unit 12 determines that Cflag_old ≠ Cflag, the parameter determination device 13 calculates the initial value x _bold of the depth controller (controller B) in order to calculate the pitch controller (controller A). The coefficient corresponding to the state is acquired from its own table.

As shown in FIG. 2, when the state of the pitch controller (controller A) before being switched to the depth controller (controller B) is infiltration, α ₁ , α ₂ , α ₃ , α ₄ is obtained. A command value u of the depth controller (controller B) is calculated by the command value calculation unit 14 based on the initial value x _boldα calculated by these coefficients. Accordingly, underwater vehicles, based on a command value u, is the depth controlled to depth of 150 meters the target r _a.

Next, a controller switching method when the underwater vehicle is lifted from a depth of 150 meters to 100 meters will be described.
At time t ₃ in FIG. 6, when the target r _a ′ for ascending the underwater vehicle to a depth of 100 meters is input by the operator, the target r _a ′ is input to the data reading unit 11. Depth difference is a predetermined value of the current depth and the target r _{a '(e.g.,} 20 meters) when it is higher, the switching of the control unit depth Control from (controller B) Pitch control to (controller A) Is done. As a result, the controller flag Cflag = 1 indicating the pitch controller (controller A) is input to the data reading unit 11. The initial value x _aold of the pitch controller (controller A) is calculated based on the coefficients γ ₁ , γ ₂ , γ ₃ , γ ₄ according to the state of the depth controller (controller B). Based on the initial value x _aold calculated based on these coefficients γ ₁ , γ ₂ , γ ₃ , γ ₄ and the parameter calculation formula, a command value u that rises to a depth of 100 meters is calculated. As a result, a controller (not shown) controls the pitch controller (controller A) based on the command value u.

At time t ₄ in FIG. 6, when the underwater vehicle emerged to 120 meters 20 meters before the target depth 100m depth controller from a pitch controller by a control unit (not shown) (controller A) (controller In B), the controller is switched. Thereby, the controller flag Cflag = 2 indicating that the depth controller (controller B) has been selected is input to the data reading unit 11. If the controller determination unit 12 determines that the controller has been switched because Cflag_old ≠ Cflag, the parameter determination device 13 determines its own value x _bβ for calculating the initial value x _bβ of the depth controller (controller B). A coefficient corresponding to the state of the pitch controller (controller A) is acquired from the table. As shown in FIG. 2, when the pitch controller (controller A) is floating, β ₁ , β ₂ , β ₃ , and β ₄ are acquired as coefficients. The parameter determination device 13 calculates an initial value x _boldβ based on the coefficients β ₁ , β ₂ , β ₃ , and β ₄ , and the command value calculation unit 14 determines the depth controller (controller B) based on the initial value x _boldβ. ) Is calculated. Accordingly, underwater vehicles, based on a command value u, is the depth controlled to depth of 100 meters the target r _{a '.}

As described above, the parameter calculation arithmetic expression of the parameter determination device 13 in the controller switching system 1 according to the present embodiment is a combination of arithmetic expressions of a plurality of polynomials having different characteristics, The coefficient provided in the term is determined according to the previous controller state, and the initial value x _iold (i is a code indicating the controller) of the controller after switching is set from the parameter calculation formula. In this way, by adjusting the coefficient according to the state of the previous controller, the calculated initial value x _iold of the controller can be changed, so the characteristics at the time of switching obtained when switching the controller are It is possible to easily adjust according to the state of the controller, and it is possible to set a switching method having various characteristics.

  As a result, for example, conventionally, when a switching method that prioritizes smooth operation is selected, the responsiveness deteriorates, and conversely, when a switching method that prioritizes responsiveness is selected, the operation cannot be performed smoothly. When there is a switching method having different characteristics, one of them is selected, and a common switching method is applied regardless of the state of the previous controller. The parameter determination device 13 in the controller switching system 1 according to the present embodiment includes a switching method having a plurality of features as described above, and can set weights according to how to give coefficients. It is possible to adjust the switching method according to the state (for example, whether it has entered or has surfaced). Further, instead of selecting only a switching method having any one feature as in the prior art, it is possible to perform switching by blending a plurality of features.

[Modification 1]
In addition, although the parameter determination apparatus which concerns on this embodiment was supposed to be used for control of an underwater vehicle, it is not limited to this. For example, it may be used for controlling a transmission of a vehicle such as a tank.
For example, in the case of switching between a swash plate angle controller that controls the swash plate angle and a speed ratio control that controls the speed ratio, when starting (for example, switching the gear from stopping (P) to moving forward (D), etc.) ) And the coefficient at the time of shifting (for example, the gear is switched from the first speed to the second speed). In addition, for example, when the gear is set to the second speed, at the time of shifting to increase the speed (for example, switching from the first speed to the second speed) and at the time of shifting to decrease the speed (for example, switching from the third speed to the second speed). In this case, the weighting coefficient may be made different.

[Modification 2]
Moreover, in the controller used in the production line of the variable nozzle turbo (VG turbo) that changes the speed of the gas flowing to the turbine, it may be used when switching between the opening control of the variable nozzle turbo and the turbo outlet pressure control. . For example, the weighting coefficient is switched between when the opening degree deviation and the pressure deviation are large and when the deviation is small.

[Modification 3]
Moreover, although the parameter determination apparatus which concerns on this embodiment was supposed to initialize a controller in order to change a control amount, when switching a controller, it is not limited to this. For example, when the controller is switched from the off state to the on state, the weighting coefficient may be changed according to the magnitude of the deviation to be controlled.
Specifically, in the steam temperature control, flow rate control, pressure control, etc. of the power plant, when the control is switched from the off state (manual time) to the on state, the coefficient depends on the power generation load condition, temperature deviation, etc. Is provided.

Further, in the above-described embodiment, the parameter determination device applied to the controller switching system is premised on processing by hardware, but it is not necessary to be limited to such a configuration. For example, a configuration in which processing is performed separately by software is also possible. In this case, the parameter determination device includes a main storage device such as a CPU and a RAM, and a computer-readable recording medium on which a program for realizing all or part of the above processing is recorded. Then, the CPU reads out the program recorded in the storage medium and executes information processing / calculation processing, thereby realizing processing similar to that of the parameter determination device described above.
Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

  In the parameter calculation calculation formula according to the present embodiment, a plurality of calculation formulas are two types, and the respective formulas have characteristics that conflict with each other. However, the present invention is not limited to this. For example, the number of expressions in the plurality of arithmetic expressions is not particularly limited, and the characteristics of the respective methods may be different from each other even if they are not contradictory.

  Moreover, although each coefficient of each term of the arithmetic expression obtained by linearly combining the arithmetic expressions of the plurality of methods included in the parameter determination device 13 according to the present embodiment is set to 0 or more and 1 or less, the present invention is not limited to this. . For example, the value may be larger than 1.

DESCRIPTION OF SYMBOLS 1 Controller switching system 10 Initial parameter setting part 11 Data reading part 12 Controller determination part 13 Parameter determination apparatus 14 Command value calculation part

Claims (4)

In a system including a plurality of controllers, a parameter determination device that determines an initial value of a state variable used in the controller after switching when switching a controller that performs processing,
A parameter calculation calculation formula that has a contradictory characteristic of switching operation smoothness and control followability, and that is composed of a combination of a plurality of calculation formulas having the conflicting characteristics, and that allows the user to change the weight of the conflicting characteristics. With
The weighting of the parameter calculation formula is determined according to the state of the controller before switching, and the initial value in the controller after switching is determined using the parameter calculation formula given the determined weight. A parameter determination device characterized by the above.   The parameter calculation arithmetic expression represents a plurality of arithmetic expressions having the contradictory characteristics by a polynomial, linearly combines the arithmetic expressions having the contradictory characteristics, and a changeable coefficient for each term of the polynomial. The parameter determination device according to claim 1, wherein the parameter determination device is an arithmetic expression provided. In a system including a plurality of controllers, when switching a controller for processing, a parameter determination method for determining an initial value of a state variable used in the controller after switching,
A parameter calculation calculation formula that has a contradictory characteristic of switching operation smoothness and control followability, and that is composed of a combination of a plurality of calculation formulas having the conflicting characteristics, and that allows the user to change the weight of the conflicting characteristics. With
A parameter that determines the weighting of the parameter calculation formula according to the state of the controller before switching, and determines the initial value in the controller after switching using the parameter calculation formula given the determined weight Decision method. In a system including a plurality of controllers, a parameter determination program for determining an initial value of a state variable used in the controller after switching when switching a controller for processing,
A parameter calculation calculation formula that has a contradictory characteristic of switching operation smoothness and control followability, and that is composed of a combination of a plurality of calculation formulas having the conflicting characteristics, and that allows the user to change the weight of the conflicting characteristics. With
The weighting of the parameter calculation formula is determined according to the state of the controller before switching, and the initial value in the controller after switching is determined using the parameter calculation formula given the determined weight. Parameter determination program for causing a computer to execute.

Images