JP6165585B2 - Control gain optimization system for plant controller - Google Patents

Description

  The present invention relates to a PID controller that constitutes a control device of a plant, a plant having a function of evaluating an optimum value of a control gain so as to obtain a required control characteristic and setting the evaluated value in the control device. The target is a control gain optimization system of a control device.

  In thermal power plants, in particular, boiler plants, attempts have been made to suppress fluctuations in the main steam temperature, which is boiler outlet steam, and increase the temperature set value. The set value of the main steam temperature is evaluated based on the heat resistance of the material at the time of design, but is set lower on the safe side in consideration of the temperature fluctuation generated in the plant. Therefore, if the temperature fluctuation can be reduced, a margin is generated with respect to the design temperature evaluated from the heat resistance, and the temperature set value can be increased. Increasing the main steam temperature also increases plant efficiency, reducing fuel consumption and providing economic benefits.

  As a method for reducing the main steam temperature fluctuation, optimization of PID control gain can be mentioned. Many power plants use a control device composed of a PID controller, and the response of the plant can be changed by adjusting the proportional gain, integral gain, and differential gain of the PID controller. However, gain adjustment is often performed by trial and error while looking at the response of the plant, and it is not always possible to set an optimum gain.

  In order to evaluate the optimal control gain for such a problem, there is a method of determining the control gain based on the reference model. The reference model is a mathematical expression of the plant response required by the designer. As a control gain evaluation method using a reference model, there are a model reference adaptive control (MRAC), a FRIT method, and the like. The MRAC is a method for adjusting the control gain sequentially mainly online, and the FRIT method is a method for determining the control gain offline. By using these methods, if a reference model is given, the control gain can be automatically calculated so that the plant response matches the reference model. Accordingly, it is possible to determine the gain more efficiently than sequentially adjusting the proportional, integral and derivative gains while looking at the response of the plant.

  In terms of control gain calculation processing, the reference model can be given by an arbitrary function. However, if a reference model with a fast response that exceeds the control performance of the actual plant is given, if the control gain obtained by calculation is actually set in the plant controller, there will be problems such as vibration in the plant response. Shows stable behavior. Therefore, an index for determining what reference model is optimal in terms of control characteristics is desired.

For such a problem, for example, in the method described in Patent Document 1, a parameter corresponding to a time constant is provided for the reference model in order to adjust the control characteristics (particularly the rising characteristics). The reference model is optimized by automatically adjusting the time constant while feeding back the response obtained in the plant. If the time constant is reduced, the control characteristics will have a quick response, and the start-up can be quickened by changing the set value. On the other hand, when the time constant is increased, the response changes slowly.
The adjustment process of the time constant of the reference model in this known example is such that the control response becomes gentle by increasing the time constant when the following phenomenon occurs. The criterion for increasing the time constant is that the vibration of the measured value of the manipulated variable increases beyond the threshold, the deviation between the measured value of the control variable and the set value increases beyond the threshold, In the case where the deviation between the measured value and the output of the reference model increases beyond the threshold, the case where the deviation between the measured value of the control variable and the predicted value increases beyond the threshold is cited.

JP-A-1-88706

  As described above, in the above-described known example, a time constant is provided as a parameter for adjusting the control characteristics in the reference model, and the time constant is set according to some judgment criteria while feeding back the response actually obtained from the plant. The function of adjusting is added to optimize the reference model. However, in the above-described known example, a method for defining a function that is a base of a reference model for setting a time constant is not specified. Further, in the above-described known example, each time when the response actually obtained in the plant exceeds a threshold value using a plurality of judgment criteria such as vibration of the manipulated variable and set value deviation of the control variable, each time constant Take a way to adjust. Therefore, although the normative model can be sequentially corrected, it does not solve the problem of defining a normative model with optimal control characteristics.

An object of the present invention is to achieve stable operation such as optimization of control gain, improvement of plant control performance, and reduction of fluctuations.

In view of the above-described problems, the present invention provides a control response based on a condition change unit that changes a condition, a data acquisition unit that acquires response data of a plant for the condition changed by the condition change unit, and the response data. A reference model determining unit that determines a reference model by introducing parameters for adjusting the control gain, and a control gain setting unit that determines the control gain based on the reference model.

According to the present invention, the control gain can be optimized, and stable operation such as improvement in plant control performance and reduction in fluctuations can be realized.

The structure of the control gain optimization system of the plant control apparatus which is a 1st Example of this invention. Definition range of the normative model for evaluating control gain. The example of the plant data which a data acquisition part acquires. The graph which shows the calculation result of the control gain with respect to a response adjustment parameter. The figure which shows the relationship between a disturbance signal and the definition range of a reference | standard model. The figure which shows the kind of response which a response estimation part estimates. The figure which shows the kind of parameter | index showing the control characteristic used by evaluation of control gain. The graph which shows the relationship between a response adjustment parameter and the value of each parameter | index. The graph which shows the relationship between a response adjustment parameter and the value of an evaluation function. The figure which shows the example of a display of the control gain optimization process by an input / output device. Trend of set value and main steam temperature during control gain optimization processing and temperature set value increase operation. The figure which shows the example of a data display for judging execution of the control gain optimization process by an input / output device. The structure of the control gain optimization system of the plant control apparatus which is a 2nd Example of this invention. The structure of the control gain optimization system of the plant control apparatus which is the 3rd Example of this invention. Definition range of the normative model for evaluating control gain. The example of the plant data which a data acquisition part acquires. The graph which shows the relationship between a response adjustment parameter and the value of each parameter | index. The structure of the control gain optimization system of the plant control apparatus which is 4th Example of this invention. The example of the plant data which a data acquisition part acquires. The figure which shows the response at the time of the setting value change estimated from disturbance response data.

A configuration of a control gain optimization system for a plant control apparatus according to the present invention will be described below with reference to the drawings.

  FIG. 1 is a diagram illustrating a configuration of a plant including a control gain optimization system of a plant control apparatus according to an embodiment of the present invention. 1 is a plant. 2 is a control device. However, the plant and the control apparatus show only the parts related to the processing of the present invention. Reference numeral 21 denotes a PID controller targeted for optimization of control gain. Although omitted in the figure, the PID controller 21 is mounted with a proportional device, an integrator, and a differentiator, and gains are set, respectively, in the same manner as a general PID controller. In this embodiment, the plant is assumed to be a boiler power plant. In the boiler plant, the temperature of steam generated in the boiler, for example, the main steam temperature is controlled. One of the operating variables for controlling the main steam temperature is the fuel flow rate. In this embodiment, the feedback control system for the main steam temperature based on the fuel flow rate is targeted. As shown in the figure, the measured value of the main steam temperature is taken into the control device 2, and the difference from the set value of the main steam temperature is obtained by the subtractor 22 and becomes the input of the PID controller 21. The PID controller 21 outputs a fuel flow demand, and the plant 1 operates the fuel supply device 11 so that the fuel flow matches the demand. The feedback control system shown in the figure operates the fuel supply device to increase the fuel flow rate when the measured value of the main steam temperature is lower than the set value, and to decrease the fuel flow rate when the measured value is higher than the set value. The main steam temperature follows the set value.

  FIG. 2 shows the relationship between the feedback control system and the reference model in the system according to the present embodiment. A reference model M (s) is defined for the range indicated by 91 in the figure. Here, s is a Laplace operator of the transfer function. When the trend of the set value and the main steam temperature with respect to time t is R (t) and Y (t), respectively, the relationship with the reference model M (s) can be expressed by the following equation (1).

  In this specification, when the output when R (t) is input to the transfer function M (s) is Y (t), an expression such as Expression (1) is used. In the system according to the present embodiment, a control gain evaluation method using a reference model is used. Therefore, obtaining a reference model that optimizes the control characteristics is equivalent to optimizing the control gain.

  In the control gain optimization system 3 shown in FIG. 1, reference numeral 31 denotes a set value changing unit. In the system according to the present embodiment, plant data at the time of changing the set value is used as input data for optimizing the control gain. The set value changing unit 31 performs a set value changing operation by changing the set value of the control device 2.

  The data acquisition unit 32 acquires signals of a set value, a fuel flow demand, and a main steam temperature measurement value when the set value is changed. FIG. 3 is an example of acquired data. (a) is the set value, (b) is the main steam temperature measurement value, and (c) is the fuel flow demand. In this embodiment, since the set value is changed stepwise, a step response of the main steam temperature with respect to the fuel flow rate can be obtained.

  Next, the normative model determination unit 33 illustrated in FIG. 1 determines a normative model using the data acquired by the data acquisition unit 32. This process will be described below.

  First, the trend of the set value shown in FIG. 3A is input, the main steam temperature measurement value is output, and the fitting process is performed in the function form represented by the following equation (2).

  As shown in Equation (2), the system according to the present embodiment uses a second-order transfer function, but the function form is preferably changed in accordance with the characteristics of the target plant. The coefficients a1 and a2 of Equation (2) are obtained by the fitting process, and the transfer function M0 (s) becomes the base function of the reference model. Next, a parameter r for adjusting the control response is introduced into the base function, and the reference model M (s, r) is defined by the following equation (3).

  With the processing described above, in the system according to the present invention, the reference model M (s, r) for evaluating the control gain is defined based on the plant data acquired at the time of changing the set value of the plant. However, since the parameter r is not determined at this stage, the final reference model cannot be defined. The parameter r is determined so as to be optimal from the viewpoint of control characteristics by a process described later.

  When the reference model determining unit 33 defines the reference model M (s, r), the response estimation unit 34 evaluates the control gain when the parameter r of the reference model M (s, r) is changed, and at the same time Estimate the response of the plant. Further, an index representing control characteristics is extracted from the estimated plant response. This process will be described below.

  If the reference model is determined, the control gain can be calculated by an evaluation method such as MRAC or FRIT described above. Using these evaluation methods, the response estimation unit 34 calculates a control gain when the response adjustment parameter r is changed. FIG. 4 shows an example of calculation results of proportional gain, integral gain, and differential gain. In the figure, the horizontal axis represents the response adjustment parameter r. When the response adjustment parameter r is decreased and the control response is accelerated, the control gain value generally tends to increase.

  Furthermore, the response estimation unit 34 estimates various responses of the plant using the reference model for the response adjustment parameter r and the calculated control gain. Details of this process will be described below.

  The transfer function C (s, r) of the PID controller can be expressed by the following equation.

  Here, Kp is a control gain, Ki is an integral gain, Kd is a differential gain, and τ is a parameter for reducing the influence of differentiator noise. If the transfer function of the PID controller at the control gain calculated with the reference model M (s, r) and the control gain is C (s, r), the main steam temperature Y (t, r) with respect to the set value R (t) Can be expressed by the following equation (5).

  The set value R (t) is an arbitrary function. If a step signal is set, the step response of the main steam temperature Y (t, r) can be calculated.

  The response estimation unit 34 also estimates a disturbance response when a disturbance is applied to the system. The concept of disturbance response is shown using FIG. In FIG. 5, a disturbance signal D (t) is added as an input to the feedback control system shown in FIG. When the response from the set value R (t) to the main steam temperature Y (t) is expressed by the transfer function M (s), the response G (s) from the disturbance signal D (t) to the main steam temperature Y (t) ) Can be expressed by the following equation (6):

  Based on this relationship, if the reference model with the response adjustment parameter r introduced is M (s, r) and the transfer function of the control gain calculated with the reference model is C (s, r), the main signal for the disturbance signal D (t) The response of the steam temperature Y (t, r) can be expressed by the following equation (7).

  FIG. 6 schematically shows the types of responses estimated by the response estimation unit 24. As shown in the figure, according to the reference model M (s, r), the trend of fuel flow when the set value is changed in steps, the step disturbance response when a step signal is applied to the disturbance, the vibration signal to the disturbance Estimate the vibration disturbance response when. In Equation (7), a step disturbance signal is obtained by applying a step signal to the disturbance signal D (t), and a vibration disturbance response is obtained by applying a vibration signal to the disturbance signal D (t). Here, when the response adjustment parameter r is 1.0, it is equivalent to the response of the actual plant at the present time (before gain adjustment). Increasing r results in a gradual change compared to the actual machine response, and decreasing r can speed up the response.

  Next, the evaluation function determination unit 24 extracts index data representing control characteristics from the various responses shown in FIG.

  FIG. 7 shows parameters used as an index of control characteristics for evaluating the control gain in the system according to the present embodiment. (a) shows the fluctuation of the main steam temperature when the set value is changed in steps, 101 is an overshoot amount with respect to the set value of the main steam temperature, and 102 is when the fluctuation of the main steam temperature is attenuated. Shows the settling time. The settling time is expressed as the time until the decay curve falls below a predetermined threshold. (b) shows the fluctuation of the fuel flow rate when the set value is changed in steps, 103 shows the maximum value of the fluctuation width, and 104 shows the maximum value of the change rate. (c) shows the fluctuation of the main steam temperature when a step signal is applied as a disturbance signal, 105 is the maximum value of the swing width, and 106 is the settling time. (d) shows the fluctuation of the main steam temperature when a vibration signal is applied as a disturbance signal, and 107 shows the fluctuation width.

  These indices are calculated according to the response adjustment parameter r. FIG. 8 is a diagram schematically showing calculation results of each index with respect to the response adjustment parameter r. The horizontal axis is the response adjustment parameter r, and the vertical axis is the value of each index. (a) shows each index when the set value is changed in steps. As shown in the figure, for the main steam temperature, the smaller the response adjustment parameter r, that is, the faster the control response, the smaller the overshoot amount and the shorter the settling time. That is, the control performance can be improved. On the other hand, with respect to the fuel flow rate, the smaller the response adjustment parameter r, the larger the fluctuation width and the rate of change. When the control response is made faster, the set value followability of the main steam temperature is improved, but the fuel flow rate changes rapidly in order to make the main steam temperature follow the set value. An increase in the fuel flow rate means that the operation of the fuel supply device also changes greatly, which causes a failure of the device and a progress of deterioration. Therefore, it is appropriate to determine the control gain by a trade-off between both the set value followability of the main steam temperature and the magnitude of the fluctuation of the fuel flow rate. Further, (b) shows each index when a step signal is applied as a disturbance. Similarly, (c) shows each index when a vibration signal is applied as a disturbance. Regarding the disturbance response, the smaller the response adjustment parameter r, that is, the faster the control response, the smaller the fluctuation of the main steam temperature. In other words, the concept of control gain in disturbance response is the same as that for changing the set value, and the control gain is determined by the trade-off between both the main steam temperature fluctuation suppression effect and the fuel flow fluctuation magnitude when disturbance is applied. Is appropriate.

  Based on the above points, the evaluation function determination unit 35 determines the evaluation function F (r) expressed by the following equation (8) in order to comprehensively determine the control characteristics using the data of each index.

  Here, fi (r) is the value of the index i shown in FIG. 8 in the response adjustment parameter r. Wi is a load coefficient for the index i. By setting the load coefficient for a highly important index to a large value, a control gain emphasizing the performance for that index can be obtained.

  FIG. 9 shows the evaluation function F (r) represented by the equation (8). The smaller the value of the evaluation function, the closer the control characteristic is to the required state. Therefore, the optimum value of the response adjustment parameter r can be determined from the evaluation function.

  The control gain determination unit 36 determines the optimum value of the response adjustment parameter r based on the evaluation function shown in FIG. 9, and then uses the calculation result of the control gain shown in FIG. 4 to obtain the optimum value of r. Find the corresponding proportional gain, integral gain, and derivative gain.

  Finally, the control gain setting unit 37 updates the control gain of the PID controller 21 of the control device 2 to the value obtained by the above processing.

  FIG. 10 is an example of a display screen of the input / output device 4 attached to the system according to the present embodiment. Various information used in the control gain optimization process is displayed on the input / output device. On the screen shown in (a), the calculation result of the control gain and the reference model used in the calculation are displayed. The operator confirms the calculation result of the control gain based on these data, and presses the “gain correction” button to correct the gain of the PID control device with respect to the control device 2. In addition, an input window for setting the load coefficient of the evaluation function is displayed on the screen shown in (b). In the figure, load factors can be set for the indices from f1 to f7, but the operator can set the load coefficients according to the importance of these indices. The graph of the evaluation function at this time is also displayed.

  FIG. 11 is a diagram showing trends in the main steam temperature and the set value when an operation for reducing the main steam temperature variation and increasing the temperature set value is performed by optimizing the control gain. The system acquires data when the main steam temperature set value is increased or decreased (recovered to the current value), and evaluates the optimum value of the control gain by the above-described processing. When the gain is corrected to the optimum value, the fluctuation of the main steam temperature is reduced. Next, the operator performs an operation for increasing the temperature set value. Thereby, the plant can be operated in a state where the main steam temperature is higher than before the gain correction, the plant efficiency is increased, and the fuel consumption can be reduced.

  Optimizing the control gain using the system according to the present embodiment is effective not only at the time of trial operation before the start of operation of the plant or at the time of periodic inspection, but also when the operating conditions change. The operating conditions include when the plant operating load changes, the fuel properties change (if the coal boiler changes the coal type), or the atmospheric conditions (temperature, humidity, etc.) change. . It is also effective to perform optimization when the fluctuation range of the main steam temperature becomes large due to factors such as aging. FIG. 12 is an example of a display screen that provides various types of information in order to determine whether the system is performing optimization processing. Information is displayed by comparing the current plant state with the previous control gain correction. Examples of the display information include fluctuation range of main steam temperature, gain correction date, atmospheric temperature and humidity, operational load, and fuel type (coal coal type). The operator determines the necessity of control gain optimization based on this information, and when performing optimization, presses the “optimization processing screen display” button on the display screen, and is shown in FIG. Display the optimization screen.

  As described above, in the system according to the present invention, data at the time of a plant setting value change operation is acquired, and a reference model used for control gain optimization processing is defined. At this time, the normative model is based on items that serve as indices of control characteristics, that is, set value followability (overshoot amount, settling time), manipulated variable fluctuation, disturbance suppression effect (runout width, settling time). It is decided from the viewpoint. This makes it possible to evaluate an optimal control gain that matches the actual control characteristics of the plant.

Although the present embodiment has been described with respect to a boiler power plant, the present invention can be applied not only to a boiler power plant but also to a combined thermal power plant, a nuclear power plant, or a chemical plant.

  FIG. 13 is a diagram showing a configuration of a plant including a control gain optimizing system for a plant control apparatus according to an embodiment of the present invention. The difference from the first embodiment shown in FIG. 1 is that there is a PID controller 23 that is not subject to control gain optimization in addition to the PID controller 21 that is subject to control gain optimization. Hereinafter, only processing different from the first embodiment will be described.

  In the boiler plant to be applied in this embodiment, the main steam temperature is controlled by both the fuel flow rate and the spray flow rate. If the main steam temperature deviates from the set value, the spray flow rate can act on the main steam temperature faster. However, when the deviation is large, the spray flow rate cannot sufficiently cope with it, and the main steam temperature is made to follow the set value by adjusting the fuel flow rate. That is, a minute and fast change in temperature is handled by a spray flow rate, and a large and slowly changing temperature change is handled by a fuel flow rate.

  The control device 2 shown in FIG. 13 is equipped with a PID controller 21 for operating the fuel flow rate according to the set value deviation of the main steam temperature and a PID controller 23 for operating the spray flow rate. The fuel supply device 11 operates the fuel flow rate according to the fuel flow rate demand output from the PID controller 21, and the spray flow rate is operated by changing the spray valve opening according to the spray flow rate demand output from the PID controller 23. .

  In this embodiment, the PID controller 21 for manipulating the fuel flow rate is targeted for gain optimization. Similar to the processing described in the first embodiment, the set value changing unit 31 operates the set value in order to acquire the plant data used for gain evaluation. At this time, in the system according to the present embodiment, the setting value unit 31 stops the function of the PID controller 23 for operating the spray flow rate so that the spray flow demand becomes a constant value. This is because the influence of the PID controller 23 is excluded from the plant data at the time of the setting value changing operation so that only the influence of the PID controller 21 that is the target of gain optimization appears. In normal control, when a set deviation occurs in the main steam temperature, both the fuel flow rate and the spray flow rate are adjusted to cause the main steam temperature to follow the set value. However, in this embodiment, gain optimization is performed on the PID controller 21 that adjusts the fuel flow rate, so that the spray flow rate does not affect the main steam temperature. By this processing, the plant data acquired by the acquisition unit 32 is only the influence of the main steam temperature due to the fuel flow rate due to the feedback control system, and the gain of the PID controller 21 can be accurately evaluated.

  Since the subsequent processing is the same as that of the first embodiment, description thereof is omitted in this embodiment. In this embodiment, in order to optimize the gain of the PID controller for adjusting the fuel flow rate, the PID controller for adjusting the spray flow rate is stopped, but conversely, the gain of the PID controller for adjusting the spray flow rate is optimized. To do so, the PID controller that adjusts the fuel flow rate may be stopped.

As described above, in the system according to the present invention, even when a plurality of PID controllers act in the setting value changing operation, it is possible to extract the control characteristics of only the PID controller that is the target of control gain optimization. Therefore, gain evaluation can be performed with high accuracy.

  FIG. 14 is a diagram showing a configuration of a plant including a control gain optimizing system for a plant control apparatus according to an embodiment of the present invention. The difference from the first embodiment shown in FIG. 1 is that there are two PID controllers that are targets for control gain optimization. Hereinafter, only processing different from the first embodiment will be described.

  In this embodiment, the main steam temperature is controlled by the fuel flow rate and the spray flow rate. Therefore, a PID controller 21 for operating the fuel flow rate and a PID controller 23 for operating the spray flow rate are mounted on the control device 2 in accordance with the set value deviation of the main steam temperature. The fuel supply device 11 operates the fuel flow rate according to the fuel flow rate demand output from the PID controller 21, and the spray flow rate is operated by changing the spray valve opening according to the spray flow rate demand output from the PID controller 23. .

  FIG. 15 shows the relationship between the feedback control system and the reference model in the system according to the present embodiment. A reference model M (s) is defined for the range indicated by 92 in the figure. Although it is comprised by two PID controllers, it is the same as that of Example 1 at the point which defines a reference | standard model with respect to the response of the main steam temperature Y (t) with respect to setting value R (t).

  The data acquisition unit 32, which is one of the components of the control gain optimization system 3 shown in FIG. 14, acquires both the fuel flow demand and the spray flow demand as operation variables. FIG. 16 shows an example of data acquired by the data acquisition unit 32. (a) is a set value, (b) is a main steam temperature measurement value, (c) is a fuel flow demand, and (d) is a spray flow demand. As for the spray flow rate, when the flow rate is increased, the main steam temperature changes in a decreasing direction.

  FIG. 17 is a diagram schematically showing index data representing control performance obtained from a response estimated using a reference model. The horizontal axis is the response adjustment parameter r, and the vertical axis is the value of each index. In the present embodiment, as the data different from the first embodiment, only each index when the set value is changed stepwise is shown. In this embodiment, in addition to the fuel flow rate, the fluctuation range and rate of change of the spray flow rate are also used for evaluating the control gain. The definition of the evaluation function is the same as in the first embodiment and can be expressed by the above formula (8). Since the process for determining the control gain using the evaluation function is the same as that in the first embodiment, description thereof is omitted in this embodiment.

As described above, in the system according to the present invention, even when there are a plurality of PID controllers targeted for control gain optimization, the optimum gain in terms of control characteristics can be evaluated. In this embodiment, a system with multiple inputs and one output is targeted, but a system with multiple inputs and multiple outputs can be handled by the same processing. Therefore, the system according to the present invention can be applied to any control device that uses a PID controller without limiting the target.

  FIG. 18 is a diagram illustrating a configuration of a plant including a control gain optimization system of a plant control apparatus according to an embodiment of the present invention. The difference from the first embodiment shown in FIG. 1 is that a disturbance signal applying unit 38 is mounted instead of the set value changing unit. Hereinafter, only processing different from the first embodiment will be described.

  In the first and second embodiments, the reference model was defined using the response data of the plant when the set value was changed, and the control gain was evaluated. On the other hand, in the system according to the present embodiment, the response data of the plant when a disturbance signal is applied is used. This assumes a case where the output variable of the control system (main steam temperature in the present embodiment) is controlled at a constant value and the plant is difficult to change the set value. The disturbance signal applying unit 38 shown in FIG. 18 applies a disturbance through the adder 24 to the fuel flow demand that is the output of the PID controller 21.

  The data acquisition unit 32 acquires plant response data when the disturbance signal application unit 38 applies a disturbance signal. FIG. 19 is an example of acquired data. (a) is the applied disturbance signal, (b) is the main steam temperature measurement, and (c) is the fuel flow demand. The disturbance signal shown in (a) is given as a step signal. At this time, the fuel flow demand shown in (c) rises stepwise by applying the disturbance signal, but the feedback control system works and converges to the original value. As for the main steam temperature measurement value shown in (b), the temperature temporarily rises according to the change in the fuel flow demand, but converges to the original value.

  In the system according to the present embodiment, the reference model is defined using these disturbance response data. As described above with reference to FIG. 5, the reference model M (s) is represented by a response from the set value R (t) to the main steam temperature Y (t). However, the plant response data obtained in this embodiment is a response from the disturbance signal D (t) to the main steam temperature Y (t). Below, the process for defining a normative model using a disturbance response is demonstrated.

  The response from the disturbance signal D (t) to the main steam temperature Y (t) can be expressed by the above equation (6) using the reference model M (s) and the transfer function C (s) of the controller. Therefore, the disturbance response function G0 (s) is expressed by the following equation (9).

  Here, by substituting Equation (2) and Equation (4) into Equation (9), the following Equation (10) can be derived.

  In Expression (10), the proportional gain Kp, the integral gain Ki, the differential gain Kd, and the noise reduction parameter τ of the differentiator are values set in the controller. Therefore, the unknown coefficients in the transfer function represented by Equation (10) are only a1 and a2. If the disturbance signal is input to the plant response data shown in FIG. 19 and the main steam temperature is output, the coefficient a1 and a2 can be obtained by performing the fitting process using equation (10). it can. The coefficients a1 and a2 obtained by the fitting process are equivalent to the coefficients of the base function M0 (s) of the reference model as shown in the above equation (2). That is, the base function of the reference model can be obtained by fitting the disturbance response data acquired in the plant with the equation (10).

  FIG. 20 shows a reference model base function obtained from disturbance response data obtained in the plant. As shown in the figure, it is possible to estimate the main steam temperature response data when the set value is changed from the disturbance response data.

  With the above processing, the base function M0 (s) of the norm model can be determined, so the response adjustment parameter r is introduced to this and the norm model M (s, r) expressed by the above equation (3) is defined. To do. The subsequent processing up to the determination of the control gain is the same as that in the first embodiment, and thus the description thereof is omitted in this embodiment.

  As described above, in the system of the present embodiment, the control gain optimization process can be performed by applying the disturbance signal without performing the setting value changing operation. As a result, the system according to the present embodiment can be applied to a plant in which it is not preferable to change the set value of the output variable (main steam temperature in the present embodiment).

In the plant control apparatus that determines the control gain of PID control using the reference model, the control apparatus according to each embodiment described above includes a set value changing unit 31 and a disturbance signal applying unit 38 that are condition changing units that change conditions. The data acquisition unit 32 for acquiring the response data of the plant with respect to the condition changed by the condition changing unit, and the reference model determination for determining the reference model by introducing parameters for adjusting the control response based on the response data And a control gain setting unit 37 for setting a control gain based on the reference model. According to this control device, the response data of the plant when the condition is changed is obtained, the response data is transformed, the parameters for adjusting the control response are introduced, the reference model is determined, and the reference model is The control gain can be set on the basis of the control variable and the response of the control variable when the set value is changed, or the response of the control variable when a disturbance signal is applied. The model can be determined. As a result, it is possible to optimize the control gain, improve the control performance of the plant, and realize stable operation such as reduction of fluctuations.

The system according to the present invention can be used for all plants having a PID controller such as a power plant and a chemical plant.

DESCRIPTION OF SYMBOLS 1 Plant 2 Control apparatus 3 Control gain optimization system 4 Input / output apparatus 11 Fuel supply apparatus 12 Spray valve 21 PID controller 23 PID controller 31 Setting value change part 32 Data acquisition part 33 Reference model determination part 34 Response estimation part 35 Evaluation Function determining unit 36 Control gain determining unit 37 Control gain setting unit 38 Disturbance signal applying unit

Claims (11)

In a plant control apparatus that determines a control gain of PID control using a reference model,
A condition changing section for changing the condition;
A data acquisition unit for acquiring response data of the plant with respect to the conditions changed by the condition change unit;
A norm model determining unit that defines a basic function of a norm model by introducing a parameter that adjusts the speed of a control response with respect to a transfer function obtained by fitting the response data;
An evaluation function determination unit that defines an evaluation function for evaluating the control characteristics;
A control gain setting unit that determines a reference model by determining the parameter based on the evaluation function, and sets a control gain based on the determined reference model;
The evaluation function determination unit evaluates the fluctuation suppression effect of the manipulated variable when changing the set value of the plant from the fluctuation width or rate of change of the manipulated variable, or the fluctuation of the output variable when the disturbance is applied by applying a disturbance signal. A plant control device characterized in that a suppression effect is evaluated from a fluctuation width or settling time of an output variable . The plant control device according to claim 1,
A plant control apparatus, wherein the control gain is set in a plant control apparatus. The plant control device according to claim 1 ,
The said condition change part is a setting change part which changes the setting value of a plant, or a disturbance signal application part which applies a disturbance signal, The plant control apparatus characterized by the above-mentioned. The plant control device according to claim 3 ,
The evaluation function determining unit is based on at least one of the set value followability of the output variable when the set value is changed, the effect of suppressing the fluctuation of the manipulated variable when the set value is changed, and the effect of suppressing the change of the output variable when applying the disturbance. A plant controller characterized by defining an evaluation function. The plant control apparatus according to claim 4 , wherein
The said evaluation function determination part evaluates the set value followability of an output variable from the set value deviation or settling time of the output variable at the time of a set value change, The plant control apparatus characterized by the above-mentioned. The plant control apparatus according to claim 4 , wherein
The evaluation function determination unit includes at least a set value deviation or settling time of the output variable when the set value is changed, a swing width or change rate of the manipulated variable when the set value is changed, and a swing width or settling time of the output variable when the disturbance is applied. An evaluation function is defined by a total value obtained by multiplying one by a load coefficient. The plant control apparatus according to claim 6 , wherein
A function of outputting a signal for displaying on the screen at least one item to be multiplied by the load factor;
A function of receiving an input of a load coefficient for the item;
A plant control apparatus having an output function of a signal for displaying an evaluation function reflecting the result of the input. The plant control device according to claim 1,
In order to compare the control characteristics after the control gain correction and the current control characteristics, the plant control device has a function of displaying the fluctuation range of the output variable, the previous control gain correction date, the atmospheric condition, the load, and the fuel type . The plant control device according to claim 1,
A plant control device for a boiler plant control device, wherein an output variable is a boiler main steam temperature and an operation variable is a fuel flow rate or a spray flow rate. The plant control apparatus according to claim 9 , wherein
When optimizing the control gain to adjust the fuel flow rate, evaluate the control gain with the plant data when the operation to adjust the spray flow rate is stopped,
When optimizing the control gain for adjusting the spray flow rate, a plant control apparatus characterized by evaluating the control gain with plant data when the operation for adjusting the fuel flow rate is stopped. In a plant control method for determining a control gain of PID control using a reference model,
Get the response data of the plant when the conditions are changed,
Define the basic function of the reference model by introducing a parameter that adjusts the speed of the control response to the transfer function obtained by fitting the response data,
Define an evaluation function to evaluate the control characteristics,
Determine a normative model by determining the parameter based on the evaluation function, set a control gain based on the determined norm model,
The evaluation function evaluates the fluctuation suppression effect of the manipulated variable when changing the set value of the plant from the fluctuation width or rate of change of the manipulated variable, or the fluctuation suppression effect of the output variable when the disturbance is applied by applying a disturbance signal. Is evaluated from the amplitude or settling time of the output variable .