Multi-variable collaborative thermal power generating unit rapid variable load control system and method

Abstract

The invention discloses a multivariable collaborative thermal power generating unit rapid variable load control system comprising an acquisition module used for acquiring and receiving an AGC instruction; the variable load instruction preprocessing module is used for calculating a load change rate according to the AGC instruction, and calling a dynamic model according to the load change rate to predict an energy gap in the future 30 seconds; the dynamic load feed-forward compensation module is used for generating a feed-forward control quantity through an energy gap and distributing a feed-forward instruction to the fuel quantity, the water supply quantity and the steam turbine control valve of the thermal power generating unit; the multivariable coordination control module is used for establishing a steam side-combustion side decoupling control matrix, substituting the unit state quantity acquired in real time into the steam side-combustion side decoupling control matrix, calculating to obtain a feedback correction quantity, and synthesizing to obtain a total control quantity; the parameter self-optimization engine module adopts a double-layer optimization algorithm to respectively trigger a genetic algorithm to reset parameters and an application model to predict, control and optimize local parameters in a rolling manner; and the safety constraint decision module is used for monitoring the key parameter out-of-limit risk in real time and starting load rate limitation when the key parameter out-of-limit risk is out-of-limit.

Description

Technical Field

[0001] The present invention belongs to the field of thermal power plant control systems, and in particular, relates to a multi-variable coordinated rapid load change control system and method for a thermal power plant. Background Art

[0002] With the large-scale integration of renewable energy, thermal power units are increasingly required to participate in deep peak and frequency regulation of the power grid, increasing the load change rate from the traditional 1% / min to 3-5% / min. However, existing control technologies suffer from shortcomings: insufficient response speed. Traditional feedforward control only compensates for steady-state deviations and cannot predict dynamic energy gaps. Control accuracy also decreases. When the load change rate exceeds 2% / min, the actual load tracking error exceeds ±1.5% of the rated power. Summary of the Invention

[0003] In response to the problems of poor response speed and decreased control accuracy of existing thermal power unit load control systems, the present invention provides a multi-variable coordinated thermal power unit rapid load change control system and method.

[0004] In order to achieve the above technical objectives, the technical solutions adopted by the present invention are as follows:

[0005] A multi-variable coordinated rapid load change control system for thermal power units, comprising an acquisition module, a load change instruction preprocessing module, a dynamic load feedforward compensation module, a multi-variable coordinated control module, a parameter self-optimization engine module, and a safety constraint decision module;

[0006] Acquisition module, used for real-time acquisition and reception of AGC instructions;

[0007] The variable load instruction preprocessing module is used to calculate the load change rate according to the received AGC instruction, and call the thermal storage dynamic model based on the load change rate to predict the energy gap in the next 30 seconds;

[0008] The dynamic load feedforward compensation module generates feedforward control quantities based on the energy gap and distributes feedforward instructions to the fuel quantity, water supply quantity, and steam turbine regulating valve of the thermal power unit;

[0009] The multivariable coordinated control module establishes a steam-side-combustion-side decoupling control matrix, incorporates the real-time collected unit state variables into the steam-side-combustion-side decoupling control matrix, calculates the feedback correction variables, and synthesizes them to obtain the total control variable.

[0010] The parameter self-optimization engine module uses a two-layer optimization algorithm to trigger the genetic algorithm to re-tune parameters and apply model predictive control to rolling optimize local parameters;

[0011] The safety constraint decision module is used to monitor the risk of key parameters exceeding the limit in real time and initiate load rate limiting when the limit is exceeded.

[0012] Furthermore, the dynamic transfer function of boiler heat storage and steam turbine work is constructed in the thermal storage dynamic model:

[0013]

[0014] Where G(s) is the work done by the steam turbine, Q(s) is the energy stored in the boiler, and the time constants T1 and T2 change dynamically with the load rate. The model parameters are corrected online through real-time coal quality analysis.

[0015] Load change rate R(t) = dP set / dt, where P set (t) is the AGC instruction;

[0016] When |R(t)|>2%Pn / min, the rapid load change mode is activated and the thermal storage dynamic model is called to predict the energy gap in the next 30 seconds.

[0017] Furthermore, the calculation formula for the energy gap generation feedforward control quantity is:

[0018]

[0019] Where Kff is taken from the dynamic parameter library, and λ is the differential compensation coefficient;

[0020] The relationship between the fuel quantity, water supply quantity and turbine valve feedforward instruction is:

[0021] u ff-fuel =0.6u ff ,u ff-water =0.3u ff ,u ff-valve =0.1u ff

[0022] where f uel is the relationship between the fuel quantity and the feedforward control quantity, water is the relationship between the water supply quantity and the feedforward control quantity, and valve is the relationship between the turbine regulating valve and the feedforward control quantity.

[0023] Furthermore, the unit state quantity: X=[P act ,T main ,P reheat ,O2%], where P act is the main steam pressure, T main is the main steam temperature, P reheat is the reheat steam pressure, O2% is the oxygen content;

[0024] Feedback correction amount: u fb =G -1 (s)·(X set -X act ), where X set is the set value of the unit state quantity, Xact is the actual value of the unit state quantity;

[0025] Total control amount: u total =u ff +u fb .

[0026] Furthermore, the two-layer optimization algorithm includes an outer layer and an inner layer, where the outer layer uses an improved genetic algorithm (adaptive crossover rate 0.6-0.9) to globally optimize the control parameter set;

[0027] The inner layer applies model predictive control (MPC) to optimize local parameters in a rolling manner, updating the PID parameter K every 10 seconds. p , T i , T d ;

[0028] Parameter evaluation formula:

[0029]

[0030] e p is the load deviation, e T is the steam temperature deviation, Δu is the control variable change rate;

[0031] If J>J threshold , triggering the genetic algorithm to re-tune parameters, which takes <3 seconds.

[0032] Furthermore, the real-time monitoring key parameter out-of-limit risk calculation formula is:

[0033]

[0034] w i is the weight coefficient, X lim is the safety limit;

[0035] When Risk>0.8, load rate limiting is enabled:

[0036] R new =min(R set ,0.5R max ).

[0037] A multi-variable coordinated rapid load change control method for a thermal power unit comprises the following steps:

[0038] Collect and receive AGC instructions;

[0039] Calculate the load change rate based on the received AGC command, and call the thermal storage dynamic model based on the load change rate to predict the energy gap in the next 30 seconds;

[0040] Generate feedforward control quantities based on the energy gap and assign feedforward instructions to the fuel quantity, water supply quantity, and turbine regulating valve of the thermal power unit;

[0041] Establish a steam-side-combustion-side decoupling control matrix, bring the real-time collected unit state variables into the steam-side-combustion-side decoupling control matrix, calculate the feedback correction variables, and synthesize them to obtain the total control variable;

[0042] A two-layer optimization algorithm is used to trigger the genetic algorithm to re-tune parameters and to apply model predictive control to optimize local parameters in a rolling manner.

[0043] Monitor the risk of key parameters exceeding the limit in real time, and initiate load rate limiting when the limit is exceeded.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] By collecting and receiving AGC instructions for preprocessing, and calling the heat storage dynamic model based on the load change rate to predict the energy gap in the next 30 seconds to generate a feedforward control variable, feedforward instructions are distributed to the fuel quantity, water supply and steam turbine regulating valve of the thermal power unit; the response speed of the thermal power unit load control system is improved; multi-layer parameter optimization algorithm and multi-variable coordinated control are used to calculate the feedback correction quantity and synthesize it to obtain the total control quantity, thereby improving control accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The figure is a block diagram of the overall structure of a multi-variable coordinated rapid load change control system for a thermal power plant in an embodiment of the present invention.

[0047] Explanation of the marks in the figure: 10 - acquisition module, 20 - variable load instruction preprocessing module, 30 - dynamic load feedforward compensation module, 40 - multivariable coordinated control module, 50 - parameter self-optimization engine module, 60 - safety constraint decision module. DETAILED DESCRIPTION

[0048] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and drawings. The contents mentioned in the embodiments are not intended to limit the present invention.

[0049] like Figure 1 As shown, this embodiment provides a multi-variable coordinated rapid load change control system for thermal power units, including an acquisition module 10, a load change instruction preprocessing module 20, a dynamic load feedforward compensation module 30, a multi-variable coordinated control module 40, a parameter self-optimization engine module 50, and a safety constraint decision module 60;

[0050] Acquisition module 10, used for real-time acquisition and reception of AGC instructions;

[0051] The variable load instruction preprocessing module 20 is used to calculate the load change rate according to the received AGC instruction, and call the thermal storage dynamic model to predict the energy gap in the next 30 seconds according to the load change rate;

[0052] The dynamic load feedforward compensation module 30 generates feedforward control variables based on the energy gap and distributes feedforward instructions to the fuel quantity, water supply quantity and turbine regulating valve of the thermal power unit;

[0053] The multivariable coordinated control module 40 establishes a steam-side-combustion-side decoupling control matrix, incorporates the real-time collected unit state variables into the steam-side-combustion-side decoupling control matrix, calculates the feedback correction variables, and synthesizes them to obtain the total control variable;

[0054] The parameter self-optimization engine module 50 uses a two-layer optimization algorithm to trigger the genetic algorithm to re-tune parameters and apply model predictive control to optimize local parameters in a rolling manner;

[0055] The safety constraint decision module 60 is used to monitor the risk of key parameters exceeding the limit in real time, and start load rate limiting when the limit is exceeded.

[0056] The dynamic transfer function of boiler heat storage and steam turbine work is constructed in the thermal storage dynamic model:

[0057]

[0058] Where G(s) is the work done by the steam turbine, Q(s) is the energy stored in the boiler, and the time constants T1 and T2 change dynamically with the load rate. The model parameters are corrected online through real-time coal quality analysis.

[0059] Load change rate R(t) = dP set / dt, where P set (t) is the AGC instruction;

[0060] When |R(t)|>2%Pn / min, the rapid load change mode is activated and the thermal storage dynamic model is called to predict the energy gap in the next 30 seconds.

[0061] The calculation formula of the energy gap generation feedforward control quantity is:

[0062]

[0063] Where Kff is taken from the dynamic parameter library, and λ is the differential compensation coefficient;

[0064] The relationship between the fuel quantity, water supply quantity and turbine valve feedforward instruction is:

[0065] u ff-fuel =0.6u ff ,u ff-water =0.3u ff ,u ff-valve =0.1u ff where f uel is the fuel quantity, water is the water supply quantity, and valve is the steam turbine regulating valve.

[0066] Unit state quantity: X=[P act ,T main ,P reheat ,O2%]; where Pact is the main steam pressure, Tmain is the main steam temperature, Preheat is the reheat steam pressure, and O2% is the oxygen content.

[0067] Feedback correction amount: u fb =G -1 (s)·(X set -X act ); where Xset is the set value of the unit state quantity, and Xact is the actual value of the unit state quantity.

[0068] Total control amount: u total =u ff +u fb .

[0069] The two-layer optimization algorithm includes an outer layer and an inner layer. The outer layer uses an improved genetic algorithm (adaptive crossover rate 0.6-0.9)

[0070] Global optimization control parameter set;

[0071] The inner layer applies model predictive control (MPC) to optimize local parameters in a rolling manner, updating the PID parameter K every 10 seconds. p , T i , T d ;

[0072] Parameter evaluation formula:

[0073]

[0074] e p is the load deviation, e T is the steam temperature deviation, Δu is the control variable change rate;

[0075] If J>J threshold , triggering the genetic algorithm to re-tune parameters, which takes <3 seconds.

[0076] The calculation formula for the risk of real-time monitoring key parameters exceeding the limit is:

[0077]

[0078] w i is the weight coefficient, X lim is the safety limit;

[0079] When Risk>0.8, load rate limiting is enabled:

[0080] R new =min(Rset ,0.5R max ).

[0081] A multi-variable coordinated rapid load change control method for a thermal power unit comprises the following steps:

[0082] Collect and receive AGC instructions;

[0083] Calculate the load change rate based on the received AGC command, and call the thermal storage dynamic model based on the load change rate to predict the energy gap in the next 30 seconds;

[0084] Generate feedforward control quantities based on the energy gap and assign feedforward instructions to the fuel quantity, water supply quantity, and turbine regulating valve of the thermal power unit;

[0085] Establish a steam-side-combustion-side decoupling control matrix, bring the real-time collected unit state variables into the steam-side-combustion-side decoupling control matrix, calculate the feedback correction variables, and synthesize them to obtain the total control variable;

[0086] A two-layer optimization algorithm is used to trigger the genetic algorithm to re-tune parameters and to apply model predictive control to optimize local parameters in a rolling manner.

[0087] Monitor the risk of key parameters exceeding the limit in real time, and initiate load rate limiting when the limit is exceeded.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] By collecting and receiving AGC instructions for preprocessing, and calling the heat storage dynamic model based on the load change rate to predict the energy gap in the next 30 seconds to generate a feedforward control variable, feedforward instructions are distributed to the fuel quantity, water supply and steam turbine regulating valve of the thermal power unit; the response speed of the thermal power unit load control system is improved; multi-layer parameter optimization algorithm and multi-variable coordinated control are used to calculate the feedback correction quantity and synthesize it to obtain the total control quantity, thereby improving control accuracy.

[0090] The above describes in detail the multivariable coordinated rapid load change control system and method for thermal power units provided by this application. The description of the specific embodiments is intended only to facilitate understanding of the method and core concepts of this application. It should be noted that those skilled in the art may make various improvements and modifications to this application without departing from the principles of this application, and such improvements and modifications also fall within the scope of protection of the claims of this application.

Claims

1. A multi-variable coordinated rapid load change control system for thermal power generation units, characterized in that: It includes an acquisition module (10), a variable load instruction preprocessing module (20), a dynamic load feedforward compensation module (30), a multivariable coordinated control module (40), a parameter self-optimization engine module (50) and a safety constraint decision module (60); An acquisition module (10) is used for acquiring and receiving AGC instructions in real time; A variable load instruction preprocessing module (20) is used to calculate the load change rate according to the received AGC instruction, and to call the thermal storage dynamic model to predict the energy gap in the next 30 seconds according to the load change rate; A dynamic load feedforward compensation module (30) generates a feedforward control quantity according to the energy gap and distributes feedforward instructions to the fuel quantity, water supply quantity and steam turbine regulating valve of the thermal power unit; A multivariable coordinated control module (40) establishes a steam side-combustion side decoupling control matrix, brings the unit state quantity collected in real time into the steam side-combustion side decoupling control matrix, calculates the feedback correction quantity, and synthesizes it to obtain the total control quantity; The parameter self-optimization engine module (50) uses a two-layer optimization algorithm to trigger the genetic algorithm to re-adjust parameters and applies model predictive control to optimize local parameters in a rolling manner; The safety constraint decision module (60) is used to monitor the risk of key parameters exceeding the limit in real time, and start load rate limitation when the limit is exceeded.

2. A multi-variable coordinated rapid load change control system for thermal power plants according to claim 1, characterized in that: The dynamic transfer function of boiler heat storage and steam turbine work is constructed in the thermal storage dynamic model: Where G(s) is the work done by the steam turbine, Q(s) is the energy stored in the boiler, and the time constants T1 and T2 change dynamically with the load rate. The model parameters are corrected online through real-time coal quality analysis. Load change rate R(t) = dP set / dt, where P set (t) is the AGC instruction; When |R(t)|>2%Pn / min, the rapid load change mode is activated and the thermal storage dynamic model is called to predict the energy gap in the next 30 seconds.

3. A multi-variable coordinated rapid load change control system for thermal power plants according to claim 2, characterized in that: The calculation formula of the energy gap generation feedforward control quantity is: where K ff Taken from the dynamic parameter library, λ is the differential compensation coefficient, ΔQ(t) is the energy gap; The relationship between the fuel quantity, water supply quantity and turbine valve feedforward instruction is: and ff-fuel =0.6u ff ,and ff-water =0.3u ff ,and ff-valve =0.1u ff where f uel is the relationship between the fuel quantity and the feedforward control quantity, water is the relationship between the water supply quantity and the feedforward control quantity, and valve is the relationship between the turbine regulating valve and the feedforward control quantity.

4. A multi-variable coordinated rapid load change control system for thermal power plants according to claim 3, characterized in that: Unit state quantity: X=[P act ,T main ,P reheat ,O2%]; where Pact is the main steam pressure, Tmain is the main steam temperature, Preheat is the reheat steam pressure, and O2% is the oxygen content. Feedback correction amount: u fb =G -1 (s)·(X set -X act ); where Xset is the set value of the unit state quantity, and Xact is the actual value of the unit state quantity. Total control amount: u total =u ff +u fb .

5. A multi-variable coordinated rapid load change control system for thermal power plants according to claim 4, characterized in that: The two-layer optimization algorithm includes an outer layer and an inner layer. The outer layer uses an improved genetic algorithm to globally optimize the control parameter set. The inner layer applies model predictive control to rolling optimize local parameters, and updates the PID parameter K every 10 seconds. p , T i , T d ; Parameter evaluation formula: e p is the load deviation, e T is the steam temperature deviation, Δu is the control variable change rate; If J>J threshold , triggering the genetic algorithm to re-tune parameters, which takes <3 seconds.

6. A multi-variable coordinated rapid load change control system for thermal power plants according to claim 5, characterized in that: The calculation formula for the risk of real-time monitoring key parameters exceeding the limit is: w i is the weight coefficient, X lim is the safety limit; When Risk>0.8, load rate limiting is enabled: R new =min(R set ,0.5R max )。 7. A multi-variable coordinated rapid load control method for thermal power units, characterized in that: Including steps: Collect and receive AGC instructions; Calculate the load change rate based on the received AGC command, and call the thermal storage dynamic model based on the load change rate to predict the energy gap in the next 30 seconds; Generate feedforward control quantities based on the energy gap and assign feedforward instructions to the fuel quantity, water supply quantity, and turbine regulating valve of the thermal power unit; Establish a steam-side-combustion-side decoupling control matrix, bring the real-time collected unit state variables into the steam-side-combustion-side decoupling control matrix, calculate the feedback correction variables, and synthesize them to obtain the total control variable; A two-layer optimization algorithm is used to trigger the genetic algorithm to re-tune parameters and to apply model predictive control to optimize local parameters in a rolling manner. Monitor the risk of key parameters exceeding the limit in real time, and initiate load rate limiting when the limit is exceeded.

Images