CN110026068B - Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method - Google Patents

Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method Download PDF

Info

Publication number
CN110026068B
CN110026068B CN201910276056.2A CN201910276056A CN110026068B CN 110026068 B CN110026068 B CN 110026068B CN 201910276056 A CN201910276056 A CN 201910276056A CN 110026068 B CN110026068 B CN 110026068B
Authority
CN
China
Prior art keywords
control
fired power
coal
neural network
output
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN201910276056.2A
Other languages
Chinese (zh)
Other versions
CN110026068A (en
Inventor
吴啸
廖霈之
李益国
沈炯
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
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 Southeast University filed Critical Southeast University
Priority to CN201910276056.2A priority Critical patent/CN110026068B/en
Publication of CN110026068A publication Critical patent/CN110026068A/en
Application granted granted Critical
Publication of CN110026068B publication Critical patent/CN110026068B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01D—SEPARATION
    • B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
    • B01D53/34—Chemical or biological purification of waste gases
    • B01D53/46—Removing components of defined structure
    • B01D53/62—Carbon oxides
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01D—SEPARATION
    • B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
    • B01D53/34—Chemical or biological purification of waste gases
    • B01D53/74—General processes for purification of waste gases; Apparatus or devices specially adapted therefor
    • B01D53/77—Liquid phase processes
    • B01D53/78—Liquid phase processes with gas-liquid contact
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01D—SEPARATION
    • B01D2257/00—Components to be removed
    • B01D2257/50—Carbon oxides
    • B01D2257/504—Carbon dioxide
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
    • Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02C—CAPTURE, STORAGE, SEQUESTRATION OR DISPOSAL OF GREENHOUSE GASES [GHG]
    • Y02C20/00—Capture or disposal of greenhouse gases
    • Y02C20/40—Capture or disposal of greenhouse gases of CO2

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Environmental & Geological Engineering (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Analytical Chemistry (AREA)
  • General Chemical & Material Sciences (AREA)
  • Oil, Petroleum & Natural Gas (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses a large-scale coal-fired power plant CO based on neural network inverse control2A capture system and a feedforward control method for CO in a coal-fired power plant2The trapping system is regarded as a multivariable system with five inputs and five outputs, and main steam pressure, steam-water separator outlet enthalpy value, unit generating capacity and CO are selected2The trapping rate and the reboiler temperature are main controlled variables, and the coal feeding amount, the water feeding amount, the main steam valve, the lean solution flow and the reboiler steam flow of the unit are selected as corresponding control variables. The invention adopts BP neural network technology to establish CO of a large-scale coal-fired power plant2The inverse model of the system is captured, so that the required control variable can be calculated according to the given value, advanced control is realized, the large delay characteristic of the whole system can be effectively processed, and the dynamic adjustment quality of the output side is improved; in addition, the PID control compensator is added to realize the correction of the neural network inverse model, so that the disturbance resistance and uncertainty resistance of the neural network inverse model are enhanced, and the control system is suitable for the industrial field requirements.

Description

Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method
Technical Field
The invention relates to the field of automatic control of thermal engineering, in particular to a large-scale coal-fired power plant CO based on neural network inverse control2An entrapment system and a feed-forward control method.
Background
Thermal power generating unit is currently CO2The most important emission source of gases has a great influence on the greenhouse effect. Post combustion CO based on chemisorption2The trapping technology is to realize CO2An important measure for trapping and reducing greenhouse gas emission. Post combustion CO with MEA as adsorption solvent2Trapping technology, with its high efficiency, high economy, technical maturity andeasy to adjust, and become the commercial CO in the world2The mainstream of the trapping technology; at the same time, post-combustion CO2The trapping technology does not need to change the operation structure of the existing thermal power generating unit, and the trapping equipment is added behind the tail flue to effectively operate, so that the investment cost is reduced.
Thermal power generating unit and CO after combustion2The trapping system has strong coupling characteristics. According to a power grid load instruction, the thermal power generating unit needs to participate in load peak regulation, tail flue gas fluctuates along with the load of the thermal power generating unit, and downstream CO is influenced by the fluctuation of the flue gas2The trapping system has great influence on key variables such as trapping rate, reboiler temperature and the like; on the other hand, post-combustion CO2Reboiler steam in the entrapment system is provided by the steam turbine bleed, and this bleed can reduce unit generated energy, influences unit peak regulation. Considering the coupling characteristics between the thermal power generating unit and the capture system, it is necessary to optimize and control the thermal power generating unit and the capture system as a whole by comprehensively considering both the thermal power generating unit and the capture system. Meanwhile, researches show that the large coal-fired power plant trapping system has large inertia and delay, disturbance, measurement noise and uncertainty also have certain interference effect on the controller, and good control quality is difficult to obtain. At present, aiming at CO of large-scale coal-fired power plant2The trapping system generally adopts a conventional PID control scheme, and is difficult to effectively deal with the large delay and strong coupling characteristics of a controlled object.
Disclosure of Invention
The invention aims to solve the technical problem of providing a large-scale coal-fired power plant CO based on neural network inverse control2The trapping system and the feedforward control method can reduce the dynamic deviation caused by large inertia, carry out control in advance and improve the control quality.
In order to solve the technical problem, the invention provides a large-scale coal-fired power plant CO based on neural network inverse control2A capture system, comprising: target value setting unit 1, neural network inverse controller 2, PID control compensator 3, and large-scale coal-fired power plant CO2A trapping whole system model 4, a first delay unit 5 and a second delay unit 6; the target value setting unit 1 has two paths of outputs which are respectively connected with a neural network inverse controller 2 and a PID compensation controller 3Connecting; target value setting unit 1 outputs r (k +1) and large coal-fired power plant CO2The deviation e (k) of the output y (k +1) of the overall system model 4 is collected and used as the input of the PID control compensator 3, and the compensation input variable u is solvedPID(k) (ii) a CO of large coal-fired power station2Capturing the input u (k) of the overall system model 4 as the output u (k) of the PID control compensator 3PID(k) And the output u of the neural network inverse controller 2NN(k) Summing; CO of large coal-fired power station2Capturing an input variable u (k) and an output variable y (k +1) of the overall system model 4, and respectively passing through a first delay unit 5 and a second delay unit 6 to obtain a delay variable u (k-1) and a delay variable y (k); the first delay unit 5 and the second delay unit 6 output variables u (k-1) and y (k) and the target value setting unit 1 output r (k +1) as the input of the neural network inverse controller 2, and calculate the output uNN(k)。
Correspondingly, the large-scale coal-fired power plant CO based on neural network inverse control2The feedforward control method of the trapping system comprises the following steps:
(1) selecting main steam pressure, steam-water separator outlet enthalpy value, unit generating capacity and CO2The capture rate and the reboiler temperature are CO of a large-scale coal-fired power plant2Collecting controlled variables of the system model 4, and selecting the coal feeding amount, the water feeding amount, the main steam valve, the barren solution flow and the reboiler steam flow of the unit as corresponding control variables;
(2) under the closed-loop condition, changing the given values of the controlled variables such as the flue gas and the trapping rate, and carrying out a closed-loop response test; setting a sampling period T to obtain CO of a large coal-fired power station under different flue gas and capture rate loads2Capturing the steady-state and dynamic parameters of the control quantity and the controlled quantity of the system model 4;
(3) CO of large coal-fired power station2The control quantity data of the capture system model 4 is used as output, and the CO of the large-scale coal-fired power plant is used2The controlled quantity data of the capture system model 4 is used as input, and offline training is carried out by using a BP neural network to determine CO of the large-scale coal-fired power plant2An inverse system model of the trapping system model 4, as in formula (1):
uNN(k)＝f(y(k+1),y(k),…,y(k-n1),u(k-1),…,u(k-n2)) (1)
(4) setting a control loop, controlling main steam pressure by using unit coal supply quantity, controlling steam-water separator outlet enthalpy value by using water supply quantity, controlling unit generating capacity by using main steam valve, and controlling CO by using barren liquor flow2The capture rate, the reboiler steam flow control reboiler temperature;
(5) setting the relevant parameters of the PID control compensator 3, including the proportional gain kPIntegral time constant TiDifferential gain kdDifferential time constant Td；
(6) The output u (k +1) of the neural network inverse controller 2 at the time k is calculated by the formula (1) using the output r (k +1) of the target value setting means 1 and the outputs u (k-1) and y (k) of the first delay means 5 and the second delay means 6 as input variables, respectivelyNN(k)；
(7) The target value setting unit 1 outputs r (k +1) and CO of the large-scale coal-fired power station2The output y (k +1) of the trapping system model 4 is compared, and the output error e (k) is calculated; the compensation input u is calculated by using the output error as the input of the PID control compensator 3PID(k) (ii) a Using equation (2):
Figure BDA0002020049440000031
(8) large coal-fired power plant CO for calculating k time2Actual output of the capture system model 4; using equation (3):
u(k)＝uNN(k)+uPID(k) (3)
(9) and (4) repeatedly executing the steps (6) to (8) in the following period to obtain corresponding control quantity and realize the no-difference control.
Preferably, in step (2), the sampling time T is selected according to the rule T95(ii) 5-15, wherein T95The unit step response process for the subject rises to 95% of the conditioning time.
Preferably, in step (5), the proportional gain kPIntegral time constant TiDifferential gain kdDifferential time constant TdThe selection rule is Ziegler-Nichols engineering alignment method.
The invention has the beneficial effects that: the invention can improve the dynamic regulation quality by adopting a feedforward control method based on neural network inverse control; meanwhile, by introducing the PID compensation controller, the influence caused by mismatch, disturbance and the like of the prediction model can be effectively processed, so that the control quality of the combined supply system is ensured.
Drawings
Fig. 1 is a schematic structural diagram of a control system of the present invention.
FIG. 2 is a large coal-fired power plant CO of the present invention2The trapping system is a schematic flow chart.
FIG. 3(a) is a schematic diagram comparing the control effect of the main steam pressure at the output side of the coal-fired unit when the given value is changed in steps compared with the traditional PID controller.
Fig. 3(b) is a schematic diagram comparing the control effect of the enthalpy value of the steam-water separator outlet at the output side of the coal-fired unit when the given value changes in steps compared with the traditional PID controller.
Fig. 3(c) is a schematic diagram comparing the power generation amount control effect of the unit at the output side of the coal-fired unit when the set value is changed in a step shape with the conventional PID controller.
Fig. 4(a) is a schematic diagram comparing the coal feeding amount control effect on the input side of the coal-fired unit when the set value is changed in a step shape with the traditional PID controller.
FIG. 4(b) is a schematic diagram showing the comparison of the input side feed water control effect of the coal-fired unit when the set value is changed in a step mode with that of the traditional PID controller.
Fig. 4(c) is a schematic diagram comparing the opening control effect of the main steam valve on the input side of the coal-fired unit when the given value is changed in a step mode with the traditional PID controller.
FIG. 5(a) is a diagram of CO during a step change of a set value between the present invention and a conventional PID controller2CO at the output side of the capture system2Comparative schematic of trapping rate control effect.
FIG. 5(b) is a diagram showing CO variation in step change of set value between the present invention and a conventional PID controller2Comparative schematic of the effect of reboiler temperature control at the outlet side of the capture system.
FIG. 6(a) is a drawingCO generated when the set value of the invention and the traditional PID controller changes in a step mode2Comparative schematic of lean flow control effect on the input side of the capture system.
FIG. 6(b) is a diagram showing CO variation in step change of set value between the present invention and a conventional PID controller2Comparative schematic of the effect of steam flow control in the reboiler at the input side of the capture system.
Detailed Description
As shown in figure 1, a large-scale coal-fired power plant CO based on neural network inverse control2A capture system, comprising: target value setting unit 1, neural network inverse controller 2, PID control compensator 3, and large-scale coal-fired power plant CO2A trapping whole system model 4, a first delay unit 5 and a second delay unit 6; the target value setting unit 1 has two paths of outputs which are respectively connected with the neural network inverse controller 2 and the PID compensation controller 3; target value setting unit 1 outputs r (k +1) and large coal-fired power plant CO2The deviation e (k) of the output y (k +1) of the overall system model 4 is collected and used as the input of the PID control compensator 3, and the compensation input variable u is solvedPID(k) (ii) a CO of large coal-fired power station2Capturing the input u (k) of the overall system model 4 as the output u (k) of the PID control compensator 3PID(k) And the output u of the neural network inverse controller 2NN(k) Summing; CO of large coal-fired power station2Capturing an input variable u (k) and an output variable y (k +1) of the overall system model 4, and respectively passing through a first delay unit 5 and a second delay unit 6 to obtain a delay variable u (k-1) and a delay variable y (k); the first delay unit 5 and the second delay unit 6 output variables u (k-1) and y (k) and the target value setting unit 1 output r (k +1) as the input of the neural network inverse controller 2, and calculate the output uNN(k)。
As shown in FIG. 2, the CO of a large coal-fired power plant2A capture system, comprising: main steam pressure, steam-water separator outlet enthalpy value, unit generating capacity and CO2The main variables of the trapping rate, the temperature of the reboiler, the coal feeding amount of the unit, the water feeding amount, a main steam valve, the flow of the barren solution, the flow of the reboiler steam and the like. Large-scale coal-fired power plant CO based on neural network inverse control2The feedforward control method of the trapping system comprises the following steps:
(1) selecting main steamingSteam pressure, steam-water separator outlet enthalpy value, unit generating capacity and CO2The capture rate and the reboiler temperature are CO of a large-scale coal-fired power plant2Collecting controlled variables of the system model 4, and selecting the coal feeding amount, the water feeding amount, the main steam valve, the barren solution flow and the reboiler steam flow of the unit as corresponding control variables;
(2) under the closed-loop condition, changing the given values of the controlled variables such as the flue gas and the trapping rate, and carrying out a closed-loop response test; setting a sampling period T to obtain CO of a large coal-fired power station under different flue gas and capture rate loads2Capturing the steady-state and dynamic parameters of the control quantity and the controlled quantity of the system model 4;
(3) CO of large coal-fired power station2The control quantity data of the capture system model 4 is used as output, and the CO of the large-scale coal-fired power plant is used2The controlled quantity data of the capture system model 4 is used as input, and offline training is carried out by using a BP neural network to determine CO of the large-scale coal-fired power plant2An inverse system model of the trapping system model 4, as in formula (1):
uNN(k)＝f(y(k+1),y(k),…,y(k-n1),u(k-1),…,u(k-n2)) (1)
(4) setting a control loop, controlling main steam pressure by using unit coal supply quantity, controlling steam-water separator outlet enthalpy value by using water supply quantity, controlling unit generating capacity by using main steam valve, and controlling CO by using barren liquor flow2The capture rate, the reboiler steam flow control reboiler temperature;
(5) setting the relevant parameters of the PID control compensator 3, including the proportional gain kPIntegral time constant TiDifferential gain kdDifferential time constant Td；
(6) The output u (k +1) of the neural network inverse controller 2 at the time k is calculated by the formula (1) using the output r (k +1) of the target value setting means 1 and the outputs u (k-1) and y (k) of the first delay means 5 and the second delay means 6 as input variables, respectivelyNN(k)；
(7) The target value setting unit 1 outputs r (k +1) and CO of the large-scale coal-fired power station2The output y (k +1) of the trapping system model 4 is compared, and the output error e (k) is calculated; using output error as PID controlInput to the compensator 3, calculating a compensation input uPID(k) (ii) a Using equation (2):
Figure BDA0002020049440000051
(8) large coal-fired power plant CO for calculating k time2Actual output of the capture system model 4; using equation (3):
u(k)＝uNN(k)+uPID(k) (3)
(9) and (4) repeatedly executing the steps (6) to (8) in the following period to obtain corresponding control quantity and realize the no-difference control.
Example (b):
(1) determining CO of large coal-fired power plant2The capture system control loop and the corresponding controlled and controlled quantities are shown in table 1:
TABLE 1
Figure BDA0002020049440000052
Figure BDA0002020049440000061
(2) Setting sampling time T as 30s, using controlled quantity data as neural network input, using controller data as neural network data, using BP neural network tool box to establish CO of coal-fired power station2And (4) capturing a system inverse model. The neural network comprises two hidden layers, the number of neurons is 20 and 5 respectively, and a training function is thingdm;
(3) calculating the output u of the neural network inverse controller according to the given value r (k +1), the past input data u (k-1) and the output data y (k)NN(k)；
(4) Setting relevant parameters of the PID control compensator as shown in the formula (4):
Figure BDA0002020049440000062
(5) and calculating the deviation. e (k) ═ r (k +1) -y (k + 1);
(6) calculating the PID control compensation output according to the deviation e (k) and the formula (5):
Figure BDA0002020049440000063
(7) calculating the coal feeding amount, water feeding amount, main steam valve, barren liquor flow and reboiler steam flow u (k) u of the unit at the next momentNN(k)+uPID(k)；
(8) Outputting the optimal control quantity u (k), calculating and updating the inverse input u (k) of the neural network at the next moment according to the measurement signalNN(k) In that respect And then, repeatedly executing the steps (3) to (8) in each sampling period.
The comparison of the control effect of the feedforward control method based on the neural network inverse control and the traditional PID control effect is shown in the attached figures 3(a) to 6 (b). At initial steady state condition u1＝60.4620kg/s、u2＝425.2630kg/s、u3＝92.31％、u4＝513.4947kg/s、u5＝135.874kg/s、y1＝21.3693MPa、y2＝2722.1325kJ/kg、y3＝432.9270MWe、y4＝90％、y5At 392.2k, the output target values were changed to 24.8430MPa, 2674.4886kJ/kg, 540MWe, 90%, 392.2k, respectively, at 600 seconds, and after a certain period of operation, the output target values were changed to 24.01MPa, 2702.4781kJ/kg, 506.896MWe, 90%, 392.2k, again, at 10500 seconds. The system runs for 20400 seconds in total, and points are taken and plotted by taking 30 seconds as a sampling period for convenient observation and comparison. As shown in fig. 3(a) -6 (b), the feedforward controller based on the neural network inverse control has better control effect, small fluctuation and fast response speed; meanwhile, due to the effect of the PID compensation controller, the actual output has no deviation from the set value.
The invention uses CO from large coal-fired power station2The trapping system is used as a multivariable object with five inputs and five outputs, and a feedforward control technology based on neural network inverse control is adopted to select the coal feeding amount, the water feeding amount and the main steam valve of the unitThe lean solution flow and the reboiler steam flow are control variables, and the main steam pressure, the steam-water separator outlet enthalpy value, the unit generating capacity and the CO are respectively controlled2Capture rate and reboiler temperature. On one hand, the input quantity of the system can be estimated, the control is carried out in advance, and the large delay characteristic of the whole system can be effectively responded; in addition, by introducing the PID compensation controller, the influence caused by adaptation, disturbance and the like of a prediction model can be effectively processed, so that the control quality of the whole system is ensured.
By using the feedforward control method based on the neural network inverse controller, the control variables required by the whole system can be estimated in advance, the coordination control of the whole system can be better realized, and the dynamic adjustment quality of the system is improved; meanwhile, a PID compensation controller is added to correct errors of the neural network simulation model, so that error-free control is achieved, the capacity of the system for resisting external disturbance and uncertain disturbance is improved, the control system can be better adapted to an industrial field, and the control quality is improved.

Claims (4)

1. Large-scale coal-fired power plant CO based on neural network inverse control2An entrapment system, comprising: target value setting unit (1), neural network inverse controller (2), PID control compensator (3), and large-scale coal-fired power plant CO2A capture overall system model (4), a first delay unit (5) and a second delay unit (6); the target value setting unit (1) is provided with two paths of outputs which are respectively connected with the neural network inverse controller (2) and the PID compensation controller (3); target value setting unit (1) outputs r (k +1) and CO of large-scale coal-fired power station2The deviation e (k) of the output y (k +1) of the overall system model (4) is collected and used as the input of a PID control compensator (3), and a compensation input variable u is solvedPID(k) (ii) a CO of large coal-fired power station2Capturing the input u (k) of the overall system model (4) as the output u (k) of the PID control compensator (3)PID(k) And the output u of the neural network inverse controller (2)NN(k) Summing; CO of large coal-fired power station2Capturing an input variable u (k) and an output variable y (k +1) of the whole system model (4) and respectively passing through a first delay unit (5) and a second delay unit (6) to obtain a delay variable u (k-1) and a delay variable y (k); first delay sheetThe output variables u (k-1) and y (k) of the element (5) and the second delay unit (6) and the output r (k +1) of the target value setting unit (1) are used as the input of the neural network inverse controller (2), and the output u is calculatedNN(k)。
2. Large-scale coal-fired power plant CO based on neural network inverse control2The feedforward control method for the trapping system is characterized by comprising the following steps of:
(1) selecting main steam pressure, steam-water separator outlet enthalpy value, unit generating capacity and CO2The capture rate and the reboiler temperature are CO of a large-scale coal-fired power plant2Collecting controlled variables of the system model (4), and selecting the coal feeding amount, the water feeding amount, the main steam valve, the barren solution flow and the reboiler steam flow of the unit as corresponding control variables;
(2) under the closed-loop condition, changing the given values of the flue gas and the capture rate controlled variable, and performing a closed-loop response test; setting a sampling period T to obtain CO of a large coal-fired power station under different flue gas and capture rate loads2Capturing the steady-state and dynamic parameters of the control quantity and the controlled quantity of the system model (4);
(3) CO of large coal-fired power station2The control quantity data of the capture system model (4) is used as output, and CO of the large-scale coal-fired power plant is used2The controlled quantity data of the capture system model (4) is used as input, and offline training is carried out by using a BP neural network to determine CO of the large coal-fired power plant2An inverse system model of the trapping system model (4), as in equation (1):
uNN(k)＝f(y(k+1),y(k),…,y(k-n1),u(k-1),…,u(k-n2)) (1)
(4) setting a control loop, controlling main steam pressure by using unit coal supply quantity, controlling steam-water separator outlet enthalpy value by using water supply quantity, controlling unit generating capacity by using main steam valve, and controlling CO by using barren liquor flow2The capture rate, the reboiler steam flow control reboiler temperature;
(5) setting relevant parameters of the PID control compensator (3) including a proportional gain kPIntegral time constant TiDifferential gain kdDifferential time constant Td；
(6) The output u (k-1) of the neural network inverse controller (2) at the time k is calculated by using the formula (1) with the output r (k +1) of the target value setting unit (1) and the outputs u (k-1) and y (k) of the first delay unit (5) and the second delay unit (6) as input variables respectivelyNN(k)；
(7) The target value setting unit (1) outputs r (k +1) and the CO of the large-scale coal-fired power station2The output y (k +1) of the trapping system model (4) is compared, and the output error e (k) is calculated; the output error is used as the input of a PID control compensator (3) to calculate a compensation input quantity uPID(k) (ii) a Using equation (2):
Figure FDA0002020049430000021
(8) large coal-fired power plant CO for calculating k time2Actual output of the trapping system model (4); using equation (3):
u(k)＝uNN(k)+uPID(k) (3)
(9) and (4) repeatedly executing the steps (6) to (8) in the following period to obtain corresponding control quantity and realize the no-difference control.
3. Large-scale coal-fired power plant CO based on neural network inverse control as claimed in claim 22The feedforward control method of the trapping system is characterized in that in the step (2), the sampling time T is selected according to the rule of T95(ii) 5-15, wherein T95The unit step response process for the subject rises to 95% of the conditioning time.
4. Large-scale coal-fired power plant CO based on neural network inverse control as claimed in claim 22The feedforward control method for the trapping system is characterized in that, in the step (5), the proportional gain kPIntegral time constant TiDifferential gain kdDifferential time constant TdThe selection rule is a Ziegler-Nichols engineering integral method.
CN201910276056.2A 2019-04-08 2019-04-08 Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method Active CN110026068B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910276056.2A CN110026068B (en) 2019-04-08 2019-04-08 Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910276056.2A CN110026068B (en) 2019-04-08 2019-04-08 Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method

Publications (2)

Publication Number Publication Date
CN110026068A CN110026068A (en) 2019-07-19
CN110026068B true CN110026068B (en) 2021-07-09

Family

ID=67237735

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910276056.2A Active CN110026068B (en) 2019-04-08 2019-04-08 Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method

Country Status (1)

Country Link
CN (1) CN110026068B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110737198B (en) * 2019-10-09 2022-11-18 东南大学 Large-scale coal-fired power plant CO based on BP neural network 2 Capture system prediction control method
CN110687792B (en) * 2019-11-04 2022-04-26 东南大学 Anti-smoke disturbance fuzzy control method for carbon dioxide capture system after chemical adsorption combustion
CN110764419B (en) * 2019-11-15 2022-06-10 江苏方天电力技术有限公司 CO of large coal-fired power plant2Capture global scheduling and predictive control system and method
CN111871164B (en) * 2020-06-15 2023-03-21 浙江工业大学 Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs
CN113393905B (en) * 2021-06-03 2024-03-22 东南大学 Chemical absorption CO 2 Dynamic robust soft measurement system and method for trapping system
CN113467237B (en) * 2021-06-22 2024-05-28 东南大学 Dynamic modeling method of main steam temperature based on deep learning
CN114397922B (en) * 2021-09-29 2023-03-14 北京百利时能源技术股份有限公司 Temperature control system of carbon dioxide capture reboiler of coal-fired power plant
CN116679572B (en) * 2023-08-03 2023-09-29 北京绿能碳宝科技发展有限公司 Carbon dioxide trapping self-learning method based on deep Q learning network

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0710901B1 (en) * 1994-11-01 2000-09-27 The Foxboro Company Multivariable nonlinear process controller
CN1828656A (en) * 2006-04-13 2006-09-06 东南大学 Electricity-generating terminal contest price decision information obtaining and processing method based on Markov chain
CN101856590A (en) * 2010-06-11 2010-10-13 清华大学 Carbon capturing system and method for controlling electric carbon coordination in carbon capturing power plant after combustion
CN201711089U (en) * 2010-07-08 2011-01-19 陕西正大环保科技有限公司 pH regulating control system of wet process desulfurization equipment of small and middle sized boilers
CN103339441A (en) * 2010-12-07 2013-10-02 阿尔斯通技术有限公司 Optimized integrated controls for oxy-fuel combustion power plant
CN105498497A (en) * 2016-01-05 2016-04-20 中国科学院自动化研究所 Flue gas desulfurization and denitration integrated equipment controlled through multiple variables and control method thereof
CN107694337A (en) * 2017-11-03 2018-02-16 吉林省电力科学研究院有限公司 Coal unit SCR denitrating flue gas control methods based on network response surface

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0710901B1 (en) * 1994-11-01 2000-09-27 The Foxboro Company Multivariable nonlinear process controller
CN1828656A (en) * 2006-04-13 2006-09-06 东南大学 Electricity-generating terminal contest price decision information obtaining and processing method based on Markov chain
CN101856590A (en) * 2010-06-11 2010-10-13 清华大学 Carbon capturing system and method for controlling electric carbon coordination in carbon capturing power plant after combustion
CN201711089U (en) * 2010-07-08 2011-01-19 陕西正大环保科技有限公司 pH regulating control system of wet process desulfurization equipment of small and middle sized boilers
CN103339441A (en) * 2010-12-07 2013-10-02 阿尔斯通技术有限公司 Optimized integrated controls for oxy-fuel combustion power plant
CN105498497A (en) * 2016-01-05 2016-04-20 中国科学院自动化研究所 Flue gas desulfurization and denitration integrated equipment controlled through multiple variables and control method thereof
CN107694337A (en) * 2017-11-03 2018-02-16 吉林省电力科学研究院有限公司 Coal unit SCR denitrating flue gas control methods based on network response surface

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Modeling, simulation and control of a methanol synthesis fixed-bed reactor;Shahrokhi, M;《CHEMICAL ENGINEERING SCIENCE》;20050418;第60卷(第15期);4275-4286 *
基于神经网络的电厂烟气二氧化碳捕集过程建模;揭超;《计算机与应用化学》;20150628;第32卷(第6期);757-760 *

Also Published As

Publication number Publication date
CN110026068A (en) 2019-07-19

Similar Documents

Publication Publication Date Title
CN110026068B (en) Large-scale coal-fired power plant CO based on neural network inverse control2Trapping system and feedforward control method
Fang et al. Backstepping-based nonlinear adaptive control for coal-fired utility boiler–turbine units
CN102563599B (en) Coordinative control method for supercritical unit boilers quickly adaptive to change of heat value of fire coal
CN110764419B (en) CO of large coal-fired power plant2Capture global scheduling and predictive control system and method
CN107023825A (en) Fluidized-bed combustion boiler is controlled and combustion optimizing system
CN108227488B (en) Sliding mode prediction control-based ultra-supercritical thermal power generating unit coordination control method
CN104482525B (en) The control method of extra-supercritical unit reheat steam temperature and system
CN110737198B (en) Large-scale coal-fired power plant CO based on BP neural network 2 Capture system prediction control method
CN104089270A (en) Optimization and adjustment testing method for load control of generator set boiler
CN106842955B (en) CO after burning with exhaust gas volumn Disturbance Rejection2Trapping system forecast Control Algorithm
CN106292277B (en) Subcritical thermal power generating unit coordination control method based on global sliding mode control
CN108958031B (en) Post combustion CO2Coordinated prediction control method for trapping coal-fired power generation system
CN101286045A (en) Coal-burning boiler system mixing control method
Wu et al. Power-carbon coordinated control of BFG-fired CCGT power plant integrated with solvent-based post-combustion CO2 capture
CN109188911A (en) One kind is for CO after burning2The improvement INA feed forward control method of trapping system
CN111765447B (en) Power generation boiler main steam temperature control method and system based on multivariable decoupling
CN110345506A (en) A kind of coal-burning boiler unit fast frequency hopping method and apparatus
CN116594435A (en) Pulverized coal flow control method, device, equipment, system and storage medium
CN112965386A (en) Coal water slurry gasification and coal-fired boiler ultra-low NOx emission integrated cooperative control method
CN102676713A (en) Feedforward feedback simulation method for TRT (Top Gas Pressure Recovery Turbine unit) blast furnace top pressure control stamping process and system therefor
CN114397813B (en) Power generation boiler combustion continuous sliding mode control method based on slow time-varying disturbance observer
CN113110030B (en) CO (carbon monoxide)2Trapped DMC-PID cascading system and control method thereof
CN113189871B (en) Thermal power generating unit flexible operation-oriented coordination control strategy
Ma et al. Intelligent Compensation for the Set Values of PID Controllers to Improve Boiler Superheated Steam Temperature Control
CN113341717A (en) Large-scale coal-fired power plant CO2Method for suppressing disturbance of overall trapping system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant