WO2021075515A1 - Dispositif de réglage de configuration de dispositif de commande, procédé de réglage de configuration et programme - Google Patents

Dispositif de réglage de configuration de dispositif de commande, procédé de réglage de configuration et programme Download PDF

Info

Publication number
WO2021075515A1
WO2021075515A1 PCT/JP2020/038993 JP2020038993W WO2021075515A1 WO 2021075515 A1 WO2021075515 A1 WO 2021075515A1 JP 2020038993 W JP2020038993 W JP 2020038993W WO 2021075515 A1 WO2021075515 A1 WO 2021075515A1
Authority
WO
WIPO (PCT)
Prior art keywords
value
setting
controller
current
new
Prior art date
Application number
PCT/JP2020/038993
Other languages
English (en)
Japanese (ja)
Inventor
通孝 中村
寿宏 馬場
諒 大西
雅樹 小山
Original Assignee
三菱パワー株式会社
Ａｄａｐｔｅｘ株式会社
株式会社Ｍｈｉパワーコントロールシステムズ
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 三菱パワー株式会社, Ａｄａｐｔｅｘ株式会社, 株式会社Ｍｈｉパワーコントロールシステムズ filed Critical 三菱パワー株式会社
Publication of WO2021075515A1 publication Critical patent/WO2021075515A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric

Definitions

  • This disclosure relates to automatic adjustment of controller setting parameters.
  • Patent Document 1 discloses a device that automatically optimizes control parameters such as the gain of a proportional integration controller according to an object and constraints while considering variations in plant output. Based on the simulated measurement signal obtained by simulation analysis using a controlled model that can reproduce the variation in plant output, the optimum control parameters are searched for so that the plant operating characteristics have the desired performance. Then, the operation signal is calculated based on the adjusted control parameter and the measurement signal, and transmitted to the controlled target plant via the input / output interface unit. As a result, the controlled device such as a valve operates. Further, Non-Patent Document 1 discloses a method for adjusting the PID gain of a PID controller using a cerebellar calculation model (CMAC) which is a kind of neural network.
  • CMAC cerebellar calculation model
  • the optimum setting values of setting parameters such as proportional gain, integral gain, and differential gain in PID control differ depending on the load.
  • the process amount is not always constant and changes, the more the control is performed to lower the tolerance of the fluctuation centering on the target value of the process amount (the control intensity is increased), the more the operation end is adjusted according to the above-mentioned fluctuation of the process amount.
  • Great control is needed. For example, when operating a boiler in which the control strength of the process amount is increased in order to prioritize the dynamic characteristics when the load changes, and the control strength is lowered in order to suppress fuel fluctuation when the load becomes constant.
  • the control is performed by the controller by changing the control strength, but the optimum setting value of the setting parameter differs depending on the control strength.
  • the control strength is an index of the priority of the target value (process value) fluctuation and the operation end fluctuation in each control, and the strong control strength means that the target value (process value) fluctuation is preferentially suppressed. It means to allow the operation end fluctuation.
  • At least one embodiment of the present invention aims to provide a setting adjustment device capable of more appropriate parameter setting adjustment of a controller that feedback-controls a process amount.
  • the setting adjustment device is It is a setting adjustment device for setting parameters set in the controller that feedback-controls the process amount in the plant with a boiler.
  • the current load acquisition unit configured to acquire the current load value of the plant
  • the current setting acquisition unit configured to acquire the current setting value of the setting parameter, and The deviation between the measured value of the process amount and the target value when the current load value is acquired, and the operation amount output from the controller to the operation end according to the deviation under the current set value are acquired.
  • Input / output acquisition unit configured as A new setting determination unit configured to determine a new setting value of the setting parameter based on the acquired current load value, the deviation, the current setting value, and the operation amount. It includes a transmitter configured to transmit the determined new set value to the controller.
  • the setting adjustment method is It is a setting adjustment method of the setting parameter set in the controller that feedback-controls the process amount in the plant having a boiler.
  • the current load acquisition step for acquiring the current load value of the plant, and The current setting acquisition step for acquiring the current setting value of the setting parameter, and The deviation between the measured value of the process amount and the target value when the current load value is acquired, and the operation amount output from the controller to the operation end according to the deviation under the current set value are acquired.
  • Input / output acquisition step and A new setting determination step for determining a new setting value of the setting parameter based on the acquired current load value, the deviation, the current set value, and the manipulated variable. It includes a transmission step of transmitting the determined new set value to the controller.
  • the setting adjustment program is It is a setting adjustment program of setting parameters set in the controller that feedback-controls the process amount in the plant with a boiler.
  • the current load acquisition unit that acquires the current load value of the plant, and The current setting acquisition unit that acquires the current setting value of the setting parameter, and The deviation between the measured value of the process amount and the target value when the current load value is acquired, and the operation amount output from the controller to the operation end according to the deviation under the current set value are acquired.
  • Input / output acquisition unit and A new setting determination unit that determines a new setting value of the setting parameter based on the acquired current load value, the deviation, the current setting value, and the operation amount.
  • a setting adjustment device capable of more appropriate parameter setting adjustment of a controller that feedback-controls the process amount.
  • expressions such as “same”, “equal”, and “homogeneous” that indicate that things are in the same state not only represent exactly the same state, but also have tolerances or differences to the extent that the same function can be obtained. It shall also represent the existing state.
  • the expression representing a shape such as a quadrangular shape or a cylindrical shape not only represents a shape such as a quadrangular shape or a cylindrical shape in a geometrically strict sense, but also an uneven portion or chamfering within a range where the same effect can be obtained.
  • the shape including the part and the like shall also be represented.
  • the expressions “equipped”, “equipped”, “equipped”, “included”, or “have” one component are not exclusive expressions that exclude the existence of other components.
  • FIG. 1 is a diagram schematically showing a control system 7 of a plant including a setting adjustment device 1 according to an embodiment of the present invention.
  • FIG. 2 is a diagram schematically showing a controller 8 according to an embodiment of the present invention.
  • the plant control system 7 is a system for controlling a plant having a boiler, such as a power generation plant (thermal power plant).
  • a thermal power plant the generator 93 connected to the rotating shaft of the steam turbine 92 is rotationally driven through the rotational drive of the steam turbine 92 by supplying the superheated steam (main steam) generated by the boiler 91 to the steam turbine 92.
  • the control system 7 that controls such a plant is connected to the control device 71 so as to be communicable with the control device 71 or by a hard wire (wiring), and the above-mentioned control device 71 includes the control device 71.
  • the plant control system 7 will be described using a thermal power plant as an example.
  • the control device 71 is a device configured to control the operation of the plant, such as a DCS that executes automatic control of the plant and input / output processing with the site.
  • the control device 71 is subjected to the above-mentioned main steam pressure, steam temperature, and air inside the boiler from various sensors 73 such as a pressure gauge, a thermometer, and an oxygen meter installed in the plant. (O 2 ) Acquire the measured value PV of various process amounts required for controlling the flow rate and the like.
  • the control device 71 operates the operation end 74 of the valve, fan, motor, pump, etc. installed in the plant so that the measured value PV of the acquired process amount becomes the target value SV. Performs various controls required for operation.
  • control device 71 includes one or more (usually a plurality of) controllers 8 (spoofing in FIG. 2) configured to perform feedback control and the like as described above.
  • the controller 8 is a PI controller for controlling the main steam pressure, a PI controller for controlling the steam temperature, a PI controller for controlling the air flow rate (O 2 ), and the like, and various process amounts can be applied. It has a controller to control. Note that these controllers 8 may be PID controllers.
  • the controller 8 that executes the feedback control calculates the manipulated variable MV of the operating end 74 so that the deviation D of the target value SV of the process quantity and the measured value PV becomes 0 as shown in FIG. Then, the operation amount MV is transmitted to the operation end 74, and the actuator possessed by the operation end 74 is controlled by the operation amount MV.
  • the measured value PV of the process amount also changes with a change in the valve opening degree of the operating end 74 caused by this control, but the measured value PV of the process amount after this change is input to the controller 8 again, as described above. Feedback control is performed by repeating the control.
  • the calculated value by the calculation by the controller 8 is the proportional gain K P that is set in the controller 8, the integral gain K I, and a value corresponding to the set parameters C, such as differential gain K D Become. That is, for example, when the controller 8 performs PID control, the manipulated variable MV is obtained by a calculation in which P control, I control, and D control are weighted and added by proportional gain, integral gain, and differential gain, respectively. Depending on the values of these setting parameters C, the calculation result will differ even if the same deviation D is input.
  • the P control is a control that outputs a correction output proportional to the magnitude of the input deviation D.
  • the I control is a control that outputs a correction output proportional to the magnitude of the cumulative value (integral) of the past deviation D.
  • the D control is a control that outputs each correction output of the D control that outputs a correction output proportional to the magnitude of the change speed of the deviation D.
  • the set values of these various setting parameters C are determined so that the optimum operation of the plant is realized before the start of operation of the plant, but after the start of operation of the plant (boiler), due to aging or the like.
  • the time constant can change.
  • the time constant may also change due to changes in the combustion state of the boiler. Therefore, when the setting parameter C is not adjusted, there may be a case where the optimum operation is not performed even if the optimum operation is performed by the setting parameter C at the initial stage of operation. Therefore, the setting parameter C of the controller 8 is adjusted (automatically adjusted) by the setting adjustment device 1.
  • FIG. 3 is a diagram showing a functional block of the setting adjustment device 1 according to the embodiment of the present invention.
  • FIG. 4 is a diagram for explaining the relationship between the dispersion of the measured value PV of the process amount, the dispersion of the manipulated variable MV, and the control intensity ⁇ according to the embodiment of the present invention.
  • the setting adjustment device 1 is a device that adjusts the setting parameter C set in the controller 8 that feedback-controls the arbitrary process amount in the plant having the boiler 91.
  • the setting adjustment device 1 has, for example, each time a periodic adjustment execution timing arrives, as described below, the operating state of the plant (current load value Lp) and the control status of the controller 8 (input information described later). Ip) is acquired, the new set value Cr of the setting parameter C is calculated, and the new set value Cr is applied to the controller 8.
  • the setting adjustment device 1 includes the current load acquisition unit 21, the current setting acquisition unit 22, the input / output acquisition unit 23, the control strength acquisition unit 24, and the new setting determination unit 3. , And a transmission unit 4. As shown in FIG. 3, the setting adjustment device 1 may further include a control strength acquisition unit 24.
  • the setting adjustment device 1 may further include a control strength acquisition unit 24.
  • the setting adjustment device 1 is composed of, for example, a computer, and includes a processor 11 (CPU or the like) and a storage device m such as a memory such as a ROM or RAM. Then, the processor 11 operates (data calculation, etc.) according to the instruction of the program (setting adjustment program) loaded in the main storage device, thereby realizing each functional unit included in the setting adjustment device 1. Further, it is assumed that the setting adjustment device 1 is provided for each target controller 8 among the plurality of controllers 8 that execute the feedback control included in the control device 71.
  • the current load acquisition unit 21 is a functional unit configured to acquire the current load value Lp of the plant (boiler).
  • the current load value Lp is the load value L (load index value) when the setting adjustment device 1 adjusts the setting parameter C.
  • This load value L may be a power value such as MW (megawatt), or may be a main steam flow rate. Further, it may be a value of a ratio to the maximum load in the specifications such as 100% or 80%. Further, the current load acquisition unit 21 acquires the current load value Lp from the outside through communication, input by an operator, or the like.
  • the current load acquisition unit 21 communicates with the current load from other devices such as the control device 71 described above and an operation terminal 76 such as an operator station (OPS) that serves as a human-machine interface (HMI) for operating and monitoring the plant. You may get the value Lp.
  • OPS operator station
  • HMI human-machine interface
  • the current load value Lp is acquired when determining the new set value Cr of the setting parameter C because the optimum value differs depending on the load value L.
  • the optimum set value each value of K P , K I , K D ) of the setting parameter C of the controller 8 is a load. It depends. Further, since the process amount is not always constant and changes, the more the control is performed to lower the tolerance of the fluctuation centering on the target value SV of the process amount (the control intensity ⁇ is increased), the more the operation is performed according to the above-mentioned fluctuation of the process amount.
  • the control intensity ⁇ of the process amount is increased to give priority to the dynamic characteristics, and when the load becomes constant, the control intensity ⁇ is decreased to suppress the fuel fluctuation.
  • the control intensity ⁇ is changed and the control is performed by the controller 8.
  • the optimum setting value of the setting parameter C differs depending on the control intensity ⁇ .
  • the current setting acquisition unit 22 is a functional unit configured to acquire the current setting value Cp of the setting parameter C in the same manner as described above.
  • the current set value Cp is a value of various gains ( each value of K P , K I , and K D ) when the above setting parameter C is adjusted by the setting adjustment device 1, and is the above current load value. It is a value of the setting parameter C when Lp is acquired.
  • current setting values Cp may include each value of the proportional gain K P and the integral gain K I.
  • current setting values Cp is a proportional gain K P, may include each value of the integral gain K I, and a derivative gain K D.
  • the current set value Cp may include only a part of a plurality of gains used for control, depending on the type of control (PI, PID, etc.).
  • the input / output acquisition unit 23 is under the deviation D between the measured value PV of the process amount and the target value SV when the current load value Lp is acquired by the current load acquisition unit 21, and the current set value Cp of the setting parameter C.
  • This is a functional unit configured to acquire the operation amount MV to the operation end 74 output from the controller 8 according to the deviation D. That is, the current load value Lp and the deviation D at the same adjustment execution timing as when the current load value Lp and the current set value Cp of the setting parameter C were acquired, the current set value Cp of the setting parameter C, and the controller under these conditions.
  • the calculation result (operation amount MV) according to 8 is acquired.
  • the deviation D may be acquired by the input / output acquisition unit 23 acquiring the measured value PV and the target value SV and calculating the deviation D, or the measured value PV and the target value on the control device 71 side.
  • the deviation D calculated from the value SV may be directly acquired (received).
  • the control strength acquisition unit 24 is a functional unit configured to acquire the control strength ⁇ that sets the adjustment target of the setting parameter C. There may be a plurality of set values (that is, combinations of gains) of the setting parameter C for causing the controller 8 to perform optimal control according to the operating state of the plant (see FIG. 4).
  • the control intensity ⁇ specifies which combination of such combinations of set values is to be selected, and directs the target of the above adjustment.
  • the value of the control intensity ⁇ may be specified in the range of 0 (0%) to (100%), for example. Further, the value of the control intensity ⁇ may be a fixed value, or may be a variable value that can be manually or automatically changed based on an instruction value by an operator or the like or a load value L (current load value Lp). Is also good. Further, since the value of the control strength ⁇ is configured to be stored in a predetermined storage area of the storage device m, the control strength acquisition unit 24 may acquire the value of the control strength ⁇ from the storage device m. ..
  • the control status by the controller 8 is a process with the variance of the manipulated variable MV as shown in FIG. 4 as the horizontal axis.
  • the variance of the measured value PV of the quantity is plotted at any point in the graph G on the vertical axis.
  • the control intensity ⁇ is used to determine which optimum point is to be targeted.
  • the position is located at the plot point p1 deviating from the optimum curve.
  • the setting adjustment device 1 sets the new setting value Cr of the setting parameter C so that the plot point p1 becomes the plot point p2. decide. In this way, the position on the target optimum curve changes according to the value of the control intensity ⁇ .
  • the higher the control intensity ⁇ the smaller the variance of the measured value PV (deviation D) of the process amount, but the larger the variance of the manipulated variable MV.
  • the lower the control intensity ⁇ the larger the variance of the measured value PV (deviation D) of the process amount, but the smaller the variance of the manipulated variable MV.
  • this control intensity ⁇ is an index of the priority of the target value fluctuation of the process amount and the operation edge fluctuation in each control, but it is also an index of determining the tolerance of the fluctuation centered on the target value of the process amount.
  • control intensity ⁇ is increased, control is performed so as to further suppress fluctuations from the target value SV of the measured value PV of the process amount.
  • the manipulated variable MV is calculated so as to cancel the constantly changing process quantity change amount more sensitively. Therefore, at the plot point p3 in FIG. 4, the fluctuation of the measured value PV of the process amount tends to be suppressed more, but the fluctuation of the manipulated variable MV tends to be larger because the manipulated variable MV is changed more greatly. Become.
  • the control performance of the process amount is improved, but the control stability is lowered.
  • the control intensity ⁇ is lowered, the control is performed so as to allow the fluctuation of the measured value PV of the process amount from the target value SV. That is, the manipulated variable MV is calculated so that the constantly changing process quantity can be more tolerated. Therefore, at the plot point p4 in FIG. 4, the fluctuation of the measured value PV of the process amount tends to be larger, but the fluctuation of the manipulated variable MV tends to be more suppressed. That is, the control stability is high, but the control performance of the process amount is deteriorated.
  • the new setting determination unit 3 sets the new setting value Cr of the setting parameter C based on the current load value Lp, the deviation D, the current setting value Cp, the manipulated variable MV, and the control intensity ⁇ acquired by each of the above-mentioned functional units. It is a functional unit configured to determine. In other words, the new setting determination unit 3 uses the above-mentioned current load value Lp, deviation D, current set value Cp, manipulated variable MV, and control intensity ⁇ as input information Ip, and new setting parameter C according to the input information Ip. Output the set value Cr.
  • the new setting determination unit 3 has an arbitrary load value L of the plant, which is an arbitrary load value L, obtained by executing feedback control by the controller 8 described above. Operate the above target value SV and the measured value PV or the above deviation D, the set value of the setting parameter C, the operation amount MV according to the above deviation D under this set value, and the operation end 74.
  • the new set value Cr of the setting parameter C may be determined based on the learning model M (learned model) in which the relationship between the change amount of the measured value PV of the process amount after the operation with the quantity MV is learned. ..
  • the control intensity ⁇ may be included in the above relationship.
  • the learning model M has the above-mentioned load value L, the target value SV, the measured value PV, or both obtained at a plurality of time points past the current adjustment execution timing such as when the current load value Lp is acquired.
  • a plurality of data in which the deviation D calculated from the above, the set value of the setting parameter C, the operation amount MV, and the change amount of the measured value PV of the process amount under this condition are associated with each other are learned as teacher data (machine learning).
  • This is a model in which the above relationships are defined by learning (machine learning).
  • the new set value Cr (output information) of the setting parameter C for the above input information Ip is determined.
  • a learning period may be provided during the operation of the control device 71 to collect the above data, and a learning model M may be created based on the collected data.
  • Re-learning may be executed by starting a learning period or the like based on the evaluation index value E described later.
  • This learning model M may be a neural network such as a cerebellar arithmetic model (CMAC), or may be a model in which the above relationships are learned by another learning method.
  • the setting adjustment device 1 may be provided with the learning model M by storing it in the storage device m or the like, or when the learning model M is provided by another device (not shown), the new setting determination unit 3 inputs. The above determination may be made by transmitting the information Ip to another device and receiving the calculation result by the learning model M.
  • CMAC cerebellar arithmetic model
  • the manipulated variable MV of the operating end 74 with respect to the arbitrary deviation D output from the controller 8 differs depending on the set value of the setting parameter C. Further, if the manipulated variable MV output from the controller 8 is different, the amount of change in the measured value PV of the process quantity is also different. In other words, the control result of the process amount for the same deviation D changes according to the setting parameter C of the controller 8. Therefore, it is possible to determine the optimum new setting value Cr of the setting parameter C by using the relationship such as the learning model M created based on the above relationship.
  • the transmission unit 4 is a functional unit configured to transmit the new setting value Cr determined by the new setting determination unit 3 described above to the controller 8.
  • the controller 8 sets the new setting value Cr by updating (overwriting, etc.) the current setting value Cp of the setting parameter C with the received new setting value Cr, and thus under the setting value of the new setting parameter C. Feedback control will be executed.
  • the current load acquisition unit 21, the current setting acquisition unit 22, the input / output acquisition unit 23, and the control strength acquisition unit 24 are connected to and acquired by the new setting determination unit 3, respectively.
  • the information is output to the new setting determination unit 3.
  • the new setting determination unit 3 is connected to the transmission unit 4 and outputs the new setting value Cr of the setting parameter C determined based on the input information Ip to the transmission unit 4.
  • the transmission unit 4 transmits the input new set value Cr to the corresponding controller 8.
  • the above-mentioned functional units may operate on the same computer, or may be distributed on different computers.
  • the control system 7 has a control network N1 to which a plurality of field devices (sensor 73, operation end 74, etc.) installed in the plant or the like are connected, and a plant such as the OPS76 described above. It is provided with a control information network N2 to which various computer devices for controlling, monitoring, and managing the above are connected.
  • a plurality of field devices (sensor 73, operation end 74, etc., which will be described later) installed in a plant or the like are usually connected to the control network N1 via a PLC (Programmable Logic Controller) or the like (not shown).
  • PLC Programmable Logic Controller
  • the setting adjustment device 1 and the control device 71 are connected to the control network N1, and bidirectional communication between the two is possible. Then, the setting adjustment device 1 acquires the above input information Ip from the control device 71 by communication via the control network N1 and transmits the new setting value Cr of the setting parameter C to the control device 71. It has become.
  • the setting adjustment device 1 may acquire the measured value PV of the process amount directly from the field device. Further, the setting adjustment device 1 communicates with the control device 71 for adjusting the setting parameter C via another communication network such as via the control information network N2 by being connected to the control information network N2. Is also good.
  • the setting adjustment device 1 may be connected to the control device 71 by hard wiring (wiring).
  • the setting parameters C such as the proportional gain K P , the integral gain K I , and the differential gain K D of the PI, the PID controller 8, etc. are taken into consideration, for example, the current load value Lp of the plant such as the boiler is taken into consideration.
  • the degree of agreement of the measured value PV of the process amount by the controller 8 with the target value SV is determined according to the control intensity ⁇ , but the optimum value of the setting parameter C according to the arbitrary control intensity ⁇ depends on the load value L of the plant. Change. Therefore, the set value of the setting parameter C of the controller 8 as described above is determined in consideration of the current load value Lp of the plant, and the controller 8 controls according to the new set value Cr of the determined setting parameter C. , The control of the process amount can be executed more appropriately.
  • FIG. 5 is a diagram showing a functional block of the new setting determination unit 3 according to the embodiment of the present invention.
  • FIG. 6 is a diagram for explaining linear interpolation according to an embodiment of the present invention.
  • any load value L (hereinafter, a reference load value L B) of the plurality of mutually different a predetermined plurality of learned each of the above relationship under The learning model M may be included.
  • a new set value Cr configuration parameters C which corresponds to the current load value Lp, at least two reference load value L B, and based on at least two new setpoint Cr was calculated using the learning model M created under the reference load value L B of the respective, interpolating the current load value Lp It may be decided by.
  • the above-mentioned learning model M is a first learning model Ma that has learned the above relationship when the load value L of the plant (boiler 91) is the first value L1, and the load value L is the second value L2.
  • the first learning model Ma which has learned the above-mentioned relationship in the case of, is included.
  • the new setting determination unit 3 acquires the first new setting value C1 and the second new setting value C2, which are the new setting values Cr calculated by the first learning model Ma and the second learning model Mb, respectively. And, by performing interpolation based on the first value L1 and the first new set value C1 and the second value L2 and the second new set value C2, the values different from the first value L1 and the second value L2 are obtained.
  • the new setting determination unit 3 also includes a learning model M (CMAC).
  • the new setting determination unit 3 is created in accordance with a plurality of reference load value L B Using the calculation results of the plurality of learning models M, a new set value Cr (combination of proportional gain K P , integral gain K I , and differential gain K D ) of the setting parameter C with respect to the current load value Lp is determined.
  • the acquired two closest reference load value L B to the current load value Lp was may be the first value L1 and second value L2 above.
  • the reference load value L B value is larger than the acquired current load value Lp side, and, from among the current load value reference load value side than smaller Lp L B, most value is near each select reference load value L B may be the first value L1 and second value L2 above.
  • the proportional gain K P for any load value L, the value of the integral gain K I, and a derivative gain K D can be calculated as a function f (L), new for both reference load value load value is not a L B L or reference load value L B and the reference load value L is not a B load value L,
  • the set value Cr may be obtained. Note that if the same as the obtained current load value Lp is the reference load value L B, using the calculated result of the learning model M created under the reference load value L B, definitive for the current load value Lp set The new set value Cr of the parameter C may be calculated.
  • each of the plurality of learning models M is created by a cerebellar arithmetic model (CMAC) which is a kind of neural network.
  • Learning model M for example 100%, 80%, 60%, four learning model M total created respectively for the four reference load value L B selected one each from each load band at 40%, etc. Includes.
  • the current load value Lp obtained, for example, a plurality of reference load value L most value selects two reference load value L B close, for example to the current load value Lp of B, the two reference calculating a new set value Cr output from the learning model M corresponding to the load value L B, by linear interpolation between them, and interpolating the current load value Lp.
  • the current load value Lp may be calculated by obtaining the function f (L) for the above linear interpolation and substituting the current load value Lp into this function f (L). At this time, the new setting value Cr corresponding to the reference load value L B that are not selected may not be calculated.
  • the new set value Cr corresponding to the current load value Lp is calculated.
  • the new set value Cr of the setting parameter C corresponding to every load value L of the plant can be appropriately calculated.
  • FIG. 7 is a diagram illustrating a screen display according to an embodiment of the present invention.
  • the setting adjustment device 1 has a measured value PV, a target value SV, a deviation D, a manipulated variable MV, a control intensity ⁇ , a current set value Cp of the setting parameter C, and a setting parameter.
  • the latest value or time transition of at least one of a plurality of numerical information including the new set value Cr of C and the evaluation index value E indicating the evaluation for the adjustment target determined based on the control intensity ⁇ described above.
  • At least one of the graph G showing the relationship between the variance index value of the measured value PV and the variance index value of the manipulated variable MV, and the information on the execution status of the adjustment of the setting parameter C (control usage status S).
  • An output unit 5 configured to output one to the display device 14 is further provided.
  • the display device 14 may be, for example, a display included in the OPS 76.
  • the above-mentioned evaluation index value E is a value that correlates with the distance d from the current state (p1) to the optimum point (p2) in FIG. 4 described above, and is an evaluation value such as a score for evaluating controllability. is there.
  • the evaluation index value E may be the control performance evaluation index (CPI).
  • the above-mentioned variance index value may be any index such as a standard deviation value or a variance value that shows the degree of variation in the measured value PV and the manipulated variable MV.
  • the control usage status S is information indicating whether or not the setting adjustment device 1 (optimization function) is used.
  • the setting adjustment device 1 uses the output unit 5 to obtain the latest values and time transitions of all the above numerical information, the above graph G, and the setting parameter C. Whether or not adjustment is executed is output at the same time, for example, displayed collectively on one screen.
  • the specified time minutes for example, 1 hour, 2 hours, etc.
  • the latest value is displayed by clicking (tapping) the current value button, and the past value and the past transition are displayed by clicking (tapping) the past value button.
  • the evaluation index value E is the control performance evaluation index (CPI), and the variance index value is the standard deviation.
  • the main process amount is displayed (in FIG.
  • the controller 8 is PI controller, the latest values and time course are displayed only for the proportional gain K P and the integral gain K I.
  • N shown in FIG. 7 is an evaluation interval
  • M is a noise order
  • Border is a PID update threshold value
  • Deadtime is a waste time, which are parameters for internal processing of PID OPTIMIZER.
  • P, I respectively, the proportional gain K P that is currently set for the controller 8, a respective value of the integral gain K I.
  • the output unit 5 outputs the time transition of the numerical information so as to be displayed as a graph in which the horizontal axis is the time and the vertical axis is the value of the numerical information. Further, as shown in FIG. 7, the output unit 5 associates the latest value of the numerical information with the graphic F representing at least one of the new setting determination unit 3 (setting adjustment device 1) or the controller 8. It may be output so as to be displayed on the display device 14 together with the display device 14. More specifically, the output unit 5 is a graphic representing the input / output relationship between the graphic F (first graphic Fa) representing at least one of the new setting determination unit 3 or the controller 8 and the first graphic Fa of the above numerical information. It is configured to further output F (second figure Fb). Then, the output unit 5 displays the numerical information, the first graphic Fa, and the second graphic so that the numerical information is displayed on the display device 14 in a state of being associated with the first graphic Fa and the second graphic Fb. Output Fb.
  • the display device 14 is provided with the first graphic Fa (PID OPTIMIZER) representing the new setting determination unit 3 (setting adjustment device 1) and the controller 8.
  • Both figures F of the first figure Fa (PID) to be represented are displayed.
  • a plurality of second figures Fb represented by arrows are displayed adjacent to (contacting) each other.
  • the second figure Fb which is the arrow, has a different direction depending on the numerical information input to the target (3, 8) represented by the first figure Fa or the numerical information output from the target.
  • the label display indicating the type of the numerical information input to the target (3, 8) indicated by each first figure Fa or the numerical information output from the target and the latest value thereof are adjacent to the second figure Fb. It is displayed. This makes it possible to easily grasp the type of each displayed numerical information, whether it is input value information or output information to the target represented by the first figure Fa, and the latest value thereof. Is. Further, with respect to the figure F representing the new setting determination unit 3 (setting adjustment device 1), whether the operating state is normal or abnormal is displayed (the display in FIG. 7 is normal).
  • the setting adjustment device 1 has an automatic mode in which the optimization function is switched to ON (execution start) when the license condition is satisfied, and is switched to OFF when the license condition is not satisfied, and an operation of an operator or the like. It has a manual mode to turn on or off the optimization function.
  • the license condition may be a case where the evaluation index value E is equal to or less than the threshold value. Therefore, the control usage status S includes whether or not the optimization function is executed, whether or not the usage permission condition is satisfied, and the mode in which the optimization function is being executed, and these are displayed.
  • both the automatic mode and the permission condition are ON, and the optimization function (boiler master optimization mode) is ON.
  • FIG. 8 is a diagram showing a setting adjustment method according to an embodiment of the present invention.
  • the setting adjustment method is a method of adjusting the setting parameter C set in the controller 8 that feedback-controls an arbitrary process amount in the plant having the boiler 91.
  • the setting adjustment method includes a current load acquisition step, a current setting acquisition step, an input / output acquisition step, a new setting determination step, and a transmission step.
  • the setting adjustment method may further include a control strength acquisition step. Each of these steps will be described in the order of the steps shown in FIG. Note that this setting adjustment method may be executed each time any periodic adjustment execution timing arrives.
  • step S1 of FIG. 8 the current load acquisition step, the current setting acquisition step, the input / output acquisition step, and the control strength acquisition step are executed. That is, in step S1, the input information Ip (already mentioned) is acquired.
  • the current load acquisition step, the current setting acquisition step, the input / output acquisition step, and the control strength acquisition step are the current load acquisition unit 21, the current setting acquisition unit 22, the input / output acquisition unit 23, and the control strength, which have already been described. Since it is the same as the processing content executed by the acquisition unit 24, the details will be omitted.
  • step S2 the new setting determination step is executed.
  • the new setting determination step is a step of determining the new setting value Cr of the setting parameter C based on the current load value Lp, the deviation D, the current setting value Cp, the manipulated variable MV, and the control intensity ⁇ acquired by executing step S1. Is.
  • the new setting value Cr of the setting parameter C may be determined based on the learning model M described above. Since this new setting determination step is the same as the processing content executed by the new setting determination unit 3 already described, the details will be omitted.
  • step S3 the transmission step is executed.
  • the transmission step is a step of transmitting the new setting value Cr determined by executing the new setting determination step (S2) to the controller 8. Since this transmission step is the same as the processing content executed by the transmission unit 4 described above, the details will be omitted.
  • the setting adjustment method further comprises the above-mentioned measured value PV, target value SV, deviation D, manipulated variable MV, control intensity ⁇ , and current set value Cp of the setting parameter C.
  • the latest value of at least one of a plurality of numerical information including the new set value Cr of the setting parameter C and the evaluation index value E indicating the evaluation for the adjustment target determined based on the control intensity ⁇ described above.
  • the graph G showing the relationship between the time transition, the variance index value of the measured value PV and the variance index value of the manipulated variable MV, and the information on the execution status of the adjustment of the setting parameter C (control usage status S).
  • It further includes an output step that outputs at least one of the above to the display device 14. Since this output step is the same as the processing content executed by the output unit 5 already described, the details will be omitted. In FIG. 8, the output step is executed in step S4.
  • the present invention is not limited to the above-described embodiment, and includes a modified form of the above-described embodiment and a combination of these embodiments as appropriate. ⁇ Additional notes>
  • the setting adjustment device (1) is It is a setting adjustment device (1) of the setting parameter (C) set in the controller (8) that feedback-controls the process amount in the plant having the boiler (91).
  • the current load acquisition unit (21) configured to acquire the current load value (Lp) of the plant
  • the current setting acquisition unit (22) configured to acquire the current setting value (Cp) of the setting parameter (C), and
  • the deviation (D) between the measured value (PV) and the target value (SV) of the process amount when the current load value (Lp) is acquired, and the deviation (D) under the current set value (Cp).
  • an input / output acquisition unit (23) configured to acquire the operation amount (MV) output from the controller (8).
  • a new setting determination unit (3) configured to determine, and It includes a transmission unit (4) configured to transmit the determined new set value (Cr) to the controller (8).
  • proportional gain such as a PID controller (8) (K P), integral gain (K I), the derivative gain (K D) such as to become setting parameters (C)
  • K P PID controller (8)
  • integral gain K I
  • K D derivative gain
  • C setting parameters
  • the set value of the setting parameter (C) of the controller (8) as described above is determined in consideration of the current load value (Lp) of the plant, and the new set value (Cr) of the determined setting parameter (C) is determined.
  • the controller (8) can control the process amount more appropriately.
  • the new setting determination unit (3) An arbitrary load value (L) of the plant, a target value (SV) under the arbitrary load value (L), and the measured value (PV) obtained by executing the feedback control by the controller (8). ) Or the deviation (D), the set value of the setting parameter (C), the operation amount (MV) according to the deviation (D) under the set value, and the operation end (74).
  • the new set value (Cr) of the set parameter (C) is set based on the learning model (M) that learns the relationship between the change amount of the measured value (PV) after the operation with the quantity (MV). decide.
  • learning model (M) the relationships between the various types of information obtained by learning (machine learning) the actual data of the amount of change (control amount) in the process amount
  • the setting parameter (C) Determine the new set value (Cr).
  • the manipulated variable (MV) of the operating end (74) with respect to the arbitrary deviation (D) output from the controller (8) differs depending on the set value of the setting parameter (C).
  • the optimum new setting value (Cr) of the setting parameter (C) can be determined by using the learning model (M) created based on the above relationship.
  • the learning model (M) is The load value and the first learning model (L) has learned the relationship when a first value (L1. E.g. L B 1 in FIG. 6) (M), Wherein wherein the load value (L) is the second value and the second learning model trained the relationship when it is (L2. E.g. L B 2 in FIG. 6) (M), and
  • the new setting determination unit (3) The first new set value (C1; for example, the proportional gain K P , which is the new set value (Cr) calculated by the first learning model (M) and the second learning model (M), respectively, is shown in FIG.
  • Interpolation for example, the straight line l in FIG. 6 is performed based on the first value (L1), the first new set value (C1), the second value (L2), and the second new set value (C2).
  • the interpolation unit that calculates the new set value (Cr) according to the load value (for example, Lp shown in FIG. 6) having a value different from the first value (L1) and the second value (L2). (32) and
  • the interpolation unit (32) When the current load value (Lp) acquired by the current load acquisition unit (21) has a value different from the first value (L1) and the second value (L2), the interpolation unit (32) ), It has an interpolated value output unit (33) that outputs the new set value (Cr).
  • the new set value (Cr) corresponding to the current load value (Lp) is determined based on at least two new set values (Cr) calculated using the learning model (M). Thereby, the new set value (Cr) of the setting parameter (C) corresponding to any load value (L) of the plant can be appropriately determined.
  • the learning model (M) is further provided.
  • the setting adjustment device (1) includes a learning model (M).
  • the new set value (Cr) can be determined quickly and stably without being affected by the communication status or the like.
  • a control strength ( ⁇ ) acquisition unit (24) configured to acquire the control strength ( ⁇ ) that sets the adjustment target of the setting parameter (C) is provided.
  • the relationship further includes said control intensity ( ⁇ ).
  • the new setting determination unit (3) has the current load value (Lp), the deviation (D), and the current setting value (Cp) acquired by each of the above-mentioned functional units.
  • the new set value (Cr) of the setting parameter (C) is set by using a learning model (M) that learns the relationship between the manipulated amount (MV) and the control intensity ( ⁇ ). decide.
  • M learning model
  • the degree of agreement of the measured value (PV) of the process amount by the controller (8) with the target value (SV) is determined according to the control intensity ( ⁇ ). Therefore, the optimum value of the setting parameter (C) can be obtained according to the control intensity ( ⁇ ).
  • the latest value of the numerical information for example, xx.x in FIG. 7) or the time transition (for example, CPI in FIG. 7), the variance index value of the measured value (PV), and the variance index value of the manipulated variable (MV).
  • At least one of a graph showing the relationship (for example, graph G in FIG. 7) and information on the execution status of the adjustment of the setting parameter (C) is displayed on a display device (14, for example, the display of OPS76 in FIG. 1).
  • An output unit (5) configured to output is further provided.
  • the output unit (5) includes a first graphic (Fa, for example, the PID OPTIMIZER in FIG. 7 or a block graphic represented by the PID) representing at least one of the new setting determination unit (3) or the controller (8).
  • a second figure (Fb; for example, the line of the arrow in FIG. 7) representing the input / output relationship of the numerical information with the first figure (Fa) is further output.
  • the numerical information, the first graphic (Fa), and the numerical information so as to be displayed on the display device (14) in a state associated with the first graphic (Fa) and the second graphic (Fb).
  • the second figure (Fb) is output.
  • the controller (8) is a controller (8) configured to execute at least one of P control, I control, and D control.
  • the setting parameter (C), the P control gain (K P), the I control of the gain (K I), include parameters defining at least one of the D control gain (K D).
  • the setting adjustment method is It is a setting adjustment method of the setting parameter (C) set in the controller (8) that feedback-controls the process amount in the plant having the boiler (91).
  • the current load acquisition step (S1 in FIG. 7) for acquiring the current load value (Lp) of the plant, and
  • the current setting acquisition step (S1 in FIG. 7) for acquiring the current setting value (Cp) of the setting parameter (C), and
  • the new setting determination step (S2 in FIG. 7) to be determined and A transmission step (S3 in FIG. 7) for transmitting the determined new set value (Cr) to the controller (8) is provided. According to the configuration of the above (9), the same effect as the above (1) is obtained.
  • the setting adjustment program is It is a setting adjustment program of the setting parameter (C) set in the controller (8) that feedback-controls the process amount in the plant having the boiler (91).
  • the current load acquisition unit (21) that acquires the current load value (Lp) of the plant, and The current setting acquisition unit (22) for acquiring the current setting value (Cp) of the setting parameter (C), and The deviation (D) between the measured value (PV) and the target value (SV) of the process amount when the current load value (Lp) is acquired, and the deviation (D) under the current set value (Cp).
  • the input / output acquisition unit (23) that acquires the operation amount (MV) output from the controller (8) to the operation end (74).
  • New setting decision unit (3) to decide and This is a program that realizes a transmission unit (4) that transmits the determined new set value (Cr) to the controller (8). According to the configuration of the above (10), the same effect as the above (1) is obtained.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Feedback Control In General (AREA)

Abstract

L'invention concerne un dispositif de réglage de configuration qui est destiné à un dispositif de commande effectuant une commande de rétroaction sur un volume de traitement et qui peut effectuer un ajustement plus approprié de configuration de paramètre.　L'invention concerne également un dispositif (1) de réglage de configuration qui est destiné à un paramètre de configuration défini pour un dispositif de commande effectuant une commande de rétroaction sur un volume de traitement dans une installation comprenant une chaudière, le dispositif de réglage de configuration comprenant : une unité d'acquisition de charge actuelle (21) pour acquérir une valeur de charge actuelle (Lp) de l'installation ; une unité d'acquisition de configuration actuelle (22) pour acquérir une valeur de configuration actuelle (Cp) du paramètre de configuration ; une unité d'acquisition d'entrée/sortie (23) pour acquérir un écart (D) entre la valeur de sujet (SV)) et une valeur de mesure de volume de traitement (PV) à un moment où la valeur de charge actuelle a été acquise et acquérir un volume de manipulation (MV) délivré du dispositif de commande à une extrémité de manipulation en fonction de l'écart sous la valeur de configuration actuelle ; une unité de détermination de nouvelle configuration (3) qui détermine une valeur de nouvelle configuration (Cr) pour le paramètre de configuration sur la base de la valeur de charge actuelle, de l'écart, de la valeur de configuration actuelle et du volume de manipulation ayant été acquis ; et une unité de transmission (4) pour transmettre au dispositif de commande la valeur de nouvelle configuration déterminée.
PCT/JP2020/038993 2019-10-17 2020-10-15 Dispositif de réglage de configuration de dispositif de commande, procédé de réglage de configuration et programme WO2021075515A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2019-189921 2019-10-17
JP2019189921A JP7475125B2 (ja) 2019-10-17 2019-10-17 コントローラの設定調整装置、設定調整方法およびプログラム

Publications (1)

Publication Number Publication Date
WO2021075515A1 true WO2021075515A1 (fr) 2021-04-22

Family

ID=75486365

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2020/038993 WO2021075515A1 (fr) 2019-10-17 2020-10-15 Dispositif de réglage de configuration de dispositif de commande, procédé de réglage de configuration et programme

Country Status (2)

Country Link
JP (1) JP7475125B2 (fr)
WO (1) WO2021075515A1 (fr)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111744969B (zh) * 2020-06-02 2022-04-19 北京科技大学设计研究院有限公司 可设定零调压力偏差的零调方法
JPWO2022215657A1 (fr) 2021-04-05 2022-10-13

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS59188701A (ja) * 1983-04-08 1984-10-26 Mitsubishi Heavy Ind Ltd ボイラマスタ制御装置
JPH028905A (ja) * 1988-06-28 1990-01-12 Toshiba Corp 燃焼制御装置
JPH0251702A (ja) * 1988-08-16 1990-02-21 Toshiba Corp 最適化制御装置
JPH03179502A (ja) * 1989-12-08 1991-08-05 Hitachi Ltd 適応制御方法
JP6400511B2 (ja) * 2015-03-16 2018-10-03 株式会社日立製作所 プラント制御用のデータ処理装置およびプラント

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR2744132B1 (fr) 1996-01-31 1998-04-24 Rhone Poulenc Chimie Systeme de base d'un agent tensioactif non-ionique et d'un silicate de metal alcalin, sous forme d'une dispersion ou de granules et son utilisation en detergence

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS59188701A (ja) * 1983-04-08 1984-10-26 Mitsubishi Heavy Ind Ltd ボイラマスタ制御装置
JPH028905A (ja) * 1988-06-28 1990-01-12 Toshiba Corp 燃焼制御装置
JPH0251702A (ja) * 1988-08-16 1990-02-21 Toshiba Corp 最適化制御装置
JPH03179502A (ja) * 1989-12-08 1991-08-05 Hitachi Ltd 適応制御方法
JP6400511B2 (ja) * 2015-03-16 2018-10-03 株式会社日立製作所 プラント制御用のデータ処理装置およびプラント

Also Published As

Publication number Publication date
JP7475125B2 (ja) 2024-04-26
JP2021064310A (ja) 2021-04-22

Similar Documents

Publication Publication Date Title
WO2021075515A1 (fr) Dispositif de réglage de configuration de dispositif de commande, procédé de réglage de configuration et programme
JP6067812B2 (ja) コンピュータ実施方法、処理モデル展開システム、処理監視システム
JP4786072B2 (ja) 多入力／多出力論理装置及び該多入力／多出力論理装置を使用してプロセスを制御する方法
CN103485897B (zh) 基于模型和仿真的涡轮控制
JP2013096418A (ja) エンジンおよび後処理システムの統合された最適化および制御
US20220052633A1 (en) Torque Control Based on Rotor Resistance Modeling in Induction Motors
CN106884724B (zh) 对功率输出-排放参数的燃气轮机调节中的概率控制建模
US20070135938A1 (en) Methods and systems for predictive modeling using a committee of models
CN106907246B (zh) 在燃气涡轮调节中对功率输出-排放参数的概率控制
CN114280927B (zh) 模糊神经网络pid火驱智能点火控制方法
CN106884726B (zh) 用于燃气涡轮的调节和控制的系统及方法
CN106884727B (zh) 在燃气涡轮调节中对功率输出-排放参数的概率控制
JP4272234B2 (ja) 石炭焚きボイラのガス濃度推定方法及び装置
US9771877B2 (en) Power output and fuel flow based probabilistic control in part load gas turbine tuning, related control systems, computer program products and methods
JP2016098818A (ja) 劣化したガスタービンのチューニング及び制御システム、コンピュータプログラム製品、及び関連する方法
Wetter et al. Verification of control sequences within OpenBuildingControl
US9599025B2 (en) Application of probabilistic control in gas turbine tuning for power output-exhaust energy parameters, related control systems, computer program products and methods
US9599026B2 (en) Application of probabilistic control in gas turbine tuning for exhaust energy-power output parameters, related control systems, computer program products and methods
US9771875B2 (en) Application of probabilistic control in gas turbine tuning, related control systems, computer program products and methods
WO2013034171A1 (fr) Agencement et procédé d'identification de système d'un établissement ou processus industriel
US20160305344A1 (en) Application of probabilistic control in gas turbine tuning for fuel flow-exhaust energy parameters, related control systems, computer program products and methods
CN106050434B (zh) 用于燃气涡轮的调整的计算系统、计算机可读介质和方法
JP2005010915A (ja) 並列型コントローラ及び並列型コントローラシステムのシミュレータ
JP6666122B2 (ja) 測定誤差に応じたガスタービンのチューニングにおける確率的制御の適用に関連する制御システム、コンピュータプログラム製品、及びその方法
US9771874B2 (en) Power output and fuel flow based probabilistic control in gas turbine tuning, related control systems, computer program products and methods

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20876088

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20876088

Country of ref document: EP

Kind code of ref document: A1