Deep peak regulation control method and system for thermal power generating unit

Abstract

The application provides a thermal power unit deep peak regulation control method and system, and relates to the field of power plant thermal power engineering. The peak regulation control method comprises the following steps: obtaining a dynamic characteristic model at N load points of a thermal power unit coordination system; generating N multivariable predictive controllers based on the dynamic characteristic model and a variable structure objective function; taking the current actual load as a switching condition, selecting a multivariable predictive controller of the load point closest to the current load as a controller currently actually executed to perform control action output. The control method can realize automatic input of the unit coordination control system in the full load range, improve the automation level in the deep peak regulation state of the unit, enhance the flexible peak regulation capacity of the unit under the condition of wide range new energy access, and solve the problems of narrow application range, poor stability and poor control precision of the traditional thermal power unit coordination system.

Description

Technical Field

[0001] This invention relates to the field of thermal power engineering in power plants, and in particular to a deep peak-shaving control method and system for thermal power units. Background Technology

[0002] Traditional supercritical thermal power unit coordinated control systems are only applicable to loads above 50% of rated load. In lower load ranges, the coordinated control system cannot be effectively put into operation for two reasons: first, the parameters of the coordinated control system cannot be extended to low loads; second, below 30% of rated load, the steam-water separator will undergo a dry-wet state transition, which causes the control structure of the coordinated control system to change from intermediate point temperature control under dry operating conditions to separator liquid level control.

[0003] With the increasing proportion of new energy power generation, deep peak shaving of thermal power units has become the norm, and the development and design of ultra-low load control systems for thermal power units has become a hot topic.

[0004] Current deep peak shaving control strategies are all based on the low-load extension of traditional coordinated control systems and combined with sequential control. Traditional coordinated control of supercritical thermal power units cannot achieve deep peak shaving low-load automatic control and dry-wet state automatic switching control. The controlled object itself has large inertia, large lag, nonlinearity, etc., which causes the coordinated system to have a low load tracking rate.

[0005] No improvement measures have yet been proposed within the industry to adopt the aforementioned advanced control strategies. Summary of the Invention

[0006] Purpose of the invention: To address the problems of low load tracking rate in the coordination control of traditional supercritical thermal power units, which cannot achieve deep peak shaving and low load automatic control as well as dry-wet state automatic switching control, and the large inertia, large lag, and nonlinearity of the controlled object itself, a deep peak shaving control method for thermal power units is proposed. Furthermore, a system for implementing the above method is proposed, thereby effectively solving the above-mentioned problems in the existing technology.

[0007] Firstly, a deep peak-shaving control method for thermal power units is proposed, including the following steps:

[0008] S1. Obtain the dynamic characteristic model of N load points in the thermal power unit coordination system;

[0009] S2. Based on the N dynamic characteristic models and the variable structure objective function, generate N multivariable predictive controllers accordingly;

[0010] S3. Using the current actual load as the switching condition, select the multivariate predictive controller that is closest to the current load point as the controller that is currently executing the control action output.

[0011] In a further embodiment of the first aspect, step S1 further includes:

[0012] The dynamic characteristic models of the thermal power unit coordination system under the conditions of 20%, 30%, 50%, 70%, and 90% of the rated load were obtained and named Model 1, Model 2, Model 3, Model 4, and Model 5, respectively.

[0013] In a further embodiment of the first aspect, the expression of the dynamic characteristic model is as follows:

[0014]

[0015] In the formula, x(k+1) is the state variable at time k+1; x(k) is the state variable at time k; u(k) is the control variable at time k; y(k) is the controlled variable at time k; and A, B, and C are all coefficient matrices.

[0016] In a further embodiment of the first aspect, the expression for the variable structure objective function J in step S2 is as follows:

[0017]

[0018] In the formula, np is the prediction step size, which is generated by the fuzzy function; r1, r2, r3, and r4 are the unit load setpoint, main steam pressure setpoint, intermediate point temperature setpoint, and steam-water separator water level setpoint, respectively. These are the predicted values ​​of unit load, main steam pressure, intermediate point temperature, and steam-water separator water level at the i-th sampling time, respectively; f and g are both functions of the intermediate point superheat and satisfy f+g=1; nu is the control step size, set to a constant value of 50; λ(P) is the actual generated power function, which limits the amplitude of the control increment; Δu1(k+j), Δu2(k+j), and Δu3(k+j) are the incremental commands for coal feed, water feed, and turbine valve main control at the j-th sampling time, respectively.

[0019] In a further embodiment of the first aspect, the expression for the actual power function λ(P) is as follows:

[0020]

[0021] In the formula, P is the actual generated power and P0 is the rated power; the purpose of this actual generated power function is to reduce large fluctuations in the control quantity under low load conditions, with the main objective of unit stability.

[0022] In a further embodiment of the first aspect, the prediction step size np employs the following fuzzy rule:

[0023] When the actual power feedback value deviates significantly from the load setpoint, the prediction step size is reduced to achieve rapid load adjustment and improve the load change rate.

[0024] When the deviation between the actual generated power feedback value and the load setpoint is small, the prediction step size is increased to achieve stable load regulation and prevent over-adjustment oscillation.

[0025] Based on the above performance indicators and prediction step size, and based on the dynamic characteristic models corresponding to 20%, 30%, 50%, 70%, and 90% of the rated load, five predictive controllers, mpc1, mpc2, mpc3, mpc4, and mpc5, are generated respectively.

[0026] In a further embodiment of the first aspect, step S3 further includes:

[0027] Under operating conditions with a load range of 80%-100% of rated load, multivariable predictive control based on Model 5 is used to output MPC. 5 ;

[0028] Under the condition of a load range of 60%-80% of the rated load, multivariate predictive control based on Model 4 is used to output MPC4;

[0029] Under the load range of 40%-60% of the rated load, multivariate predictive control based on Model 3 is used to output MPC3;

[0030] Under the condition of a load range of 25%-40% of the rated load, the multivariate predictive control output mpc2 based on Model 2 is adopted;

[0031] Under the load range of 20%-25% of the rated load, multivariable predictive control based on Model 1 is used to output mpc1;

[0032] The expression is as follows:

[0033]

[0034] In the formula, P is the actual generated power and P0 is the rated power.

[0035] In a second aspect, the present invention provides a deep peak shaving control system for thermal power units, which includes a dynamic characteristic model acquisition module, a predictive controller generation module, and a control output module.

[0036] The dynamic characteristic model acquisition module is used to acquire the dynamic characteristic models of N load points in the thermal power unit coordination system.

[0037] The predictive controller generation module generates N multivariable predictive controllers based on the N dynamic characteristic models and the variable structure objective function.

[0038] The control output module is used to select the multivariable predictive controller that is closest to the current load point as the controller that is currently executing the control action output, based on the current actual load as the switching condition.

[0039] A third aspect of the present invention provides an electronic device comprising a processor and a memory storing computer program instructions; wherein the processor, when executing the computer program instructions, implements the deep peak shaving control method for thermal power units as described in the first aspect.

[0040] In a fourth aspect, the present invention provides a computer-readable storage medium storing at least one executable instruction that, when executed on an electronic device, causes the electronic device to perform the deep peak shaving control method for thermal power units as described in the first aspect.

[0041] Beneficial Effects: This invention discloses a deep peak-shaving control method and system for thermal power units. The method employs an improved multivariable model predictive controller (MMC). The predictive model in the improved MMC adopts a multi-model structure, and the controller's objective function uses a novel variable-structure objective function with a unified structure, generated by a fuzzy function. This control method can achieve automatic activation of the unit's coordinated control system across the entire load range, improve the automation level of the unit under deep peak-shaving conditions, enhance the unit's flexible peak-shaving capability under large-scale renewable energy access conditions, and address the unit's safety issues under deep peak-shaving conditions. It also effectively improves the problem of low load tracking rate in the coordinated system caused by the large inertia, large lag, and nonlinearity of the controlled object itself, solving the problems of narrow applicability, poor stability, and poor control accuracy of traditional thermal power unit coordinated systems. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the membership function curve in an embodiment of the present invention.

[0043] Figure 2 This is a flowchart of the deep peak shaving control method for thermal power units in an embodiment of the present invention.

[0044] Figure 3 This is a schematic diagram of the deep peak shaving control system for thermal power units in an embodiment of the present invention. Detailed Implementation

[0045] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0046] Example 1:

[0047] This embodiment discloses an improved predictive coordinated control method applicable to deep-tunneling units, which is implemented through the following technical solution:

[0048] First, dynamic characteristic models at five load points of the thermal power unit coordination system are obtained using the conventional step dynamic characteristic measurement method. Second, based on the five dynamic characteristic models and a novel variable structure objective function with a unified structure, five multivariable predictive controllers are generated. Finally, using the current actual load as the switching condition, the multivariable predictive controller at the load point closest to the current load is selected as the controller currently in operation to output control action.

[0049] This method uses the conventional step dynamic characteristic measurement method to obtain dynamic characteristic models at five load points of the thermal power unit coordination system, namely, dynamic characteristic models under the conditions of 20%, 30%, 50%, 70%, and 90% of rated load, named Model 1-Model 5 respectively. The model structure adopts the traditional state-space model, and the specific model structure is as follows:

[0050]

[0051] In the formula, x(k+1) is the state variable at time k+1; x(k) is the state variable at time k; u(k) is the control variable at time k; y(k) is the controlled variable at time k; and A, B, and C are all coefficient matrices.

[0052] This method then generates five multivariate predictive controllers based on five dynamic characteristic models and a novel variable structure objective function with a unified structure. The novel variable structure objective function J with a unified structure is expressed as follows:

[0053]

[0054] In the formula, np is the prediction step size, which is generated by the fuzzy function; r1, r2, r3, and r4 are the unit load setpoint, main steam pressure setpoint, intermediate point temperature setpoint, and steam-water separator water level setpoint, respectively. These are the predicted values ​​of unit load, main steam pressure, intermediate point temperature, and steam-water separator water level at the i-th sampling time, respectively; f and g are both functions of the intermediate point superheat and satisfy f+g=1.

[0055] In this embodiment, f is set to 1 if the superheat is greater than 3°C, and 0 otherwise; nu is the control step size, which is set to a constant value of 50; λ(P) is a function of the actual generated power, which limits the magnitude of the control increment. In this embodiment, the following function is used:

[0056]

[0057] The purpose of this function is to reduce large fluctuations in control quantities under low load conditions, with unit stability as the primary objective; Δu1(k+j), Δu2(k+j), and Δu3(k+j) are the incremental instructions for coal feed, water feed, and turbine valve main control at the j-th sampling time in the future, respectively.

[0058] The prediction step size np adopts the following fuzzy rule:

[0059] When the actual power feedback value deviates significantly from the load setpoint, the prediction step size is reduced to achieve rapid load adjustment and improve the load change rate.

[0060] When the deviation between the actual generated power feedback value and the load setpoint is small, the prediction step size is increased to achieve stable load regulation and prevent over-adjustment oscillation.

[0061] Taking a specific 600MW unit setting in this embodiment as an example, the main regulator's calculation cycle is 1 second, and the specific setting function for the fuzzy rule is as follows:

[0062] Let the unit load deviation e = unit load setpoint - actual generated power feedback value; the prediction step size np is calculated based on the unit load deviation e and the set fuzzy membership function. In this embodiment, triangular and trapezoidal membership functions are used, and their function curves are shown below. Figure 1 As shown in the figure, the graph includes membership function curves 1, 2, and 3. The vertical axis represents the membership value, and the horizontal axis represents the unit load deviation e. Using triangular and trapezoidal membership functions, the fuzzy membership function curves are shown in the attached figure. Figure 1 As shown:

[0063] For any point, the unit load deviation value e is:

[0064] np = integer operation (200 * membership degree 1 + 100 * membership degree 2 + 100 * membership degree 3).

[0065] The specific fuzzy rule table is set as follows:

[0066] Table 1: Specific Fuzzy Rules Table

[0067] np 100 200 200 100

[0068] Based on the above performance indicators and prediction step size, and based on the dynamic characteristic models corresponding to 20%, 30%, 50%, 70%, and 90% of the rated load, five predictive controllers are generated respectively: mpc1, mpc2, mpc3, mpc4, and mpc.

[0069] The calculation process for generating the predictive controller is as follows:

[0070] Define the output of the multivariate predictive controller at time k as follows: (coal feed increment command Δu1(k), water feed increment command Δu2(k), turbine valve main control increment command Δu3(k)). Then the final MPC i The control command calculation expression is:

[0071]

[0072] In the formula, u(k-1) is the control quantity at time k-1.

[0073] This method ultimately uses the current actual load as the switching condition, selecting the multivariate predictive controller (MPC) of the load point closest to the current load as the controller currently executing the control action. Specifically: under load conditions of 80%-100% rated load, MPC5 is output based on Model 5; under load conditions of 60%-80% rated load, MPC4 is output based on Model 4; under load conditions of 40%-60% rated load, MPC3 is output based on Model 3; under load conditions of 25%-40% rated load, MPC2 is output based on Model 2; and under load conditions of 20%-25% rated load, MPC1 is output based on Model 1. The specific expressions are as follows:

[0074]

[0075] P represents the actual generated power, and P0 represents the rated power.

[0076] Example 2:

[0077] See Figure 3 As shown, this embodiment discloses a deep peak-shaving control system 400 for thermal power units. This deep peak-shaving control system 400 can be used to implement the deep peak-shaving control method for thermal power units disclosed in Embodiment 1 above. The deep peak-shaving control system 400 includes a dynamic characteristic model acquisition module 401, a predictive controller generation module 402, and a control output module 403. The dynamic characteristic model acquisition module 401 is used to acquire dynamic characteristic models at N load points of the thermal power unit coordination system. The predictive controller generation module 402 generates N multivariable predictive controllers based on the N dynamic characteristic models and a variable structure objective function. The control output module 403 is used to select the multivariable predictive controller closest to the current load as the currently executing controller, using the current actual load as the switching condition, and outputs the control action.

[0078] The deep peak shaving control system 400 for thermal power units can realize the automatic activation of the unit's coordinated control system over a wide load range, improve the automation level of the unit under deep peak shaving conditions, enhance the unit's flexible peak shaving capability under large-scale new energy access conditions, take into account the unit's safety issues under deep peak shaving conditions, and effectively improve the problem of low load tracking rate of the coordinated system caused by the large inertia, large lag, and nonlinearity of the controlled object itself. It also solves the problems of narrow applicability, poor stability and control accuracy of traditional thermal power unit coordinated systems.

[0079] As a preferred embodiment, an electronic device is disclosed, comprising a processor, a memory, a communication interface, and a communication bus. The processor, memory, and communication interface communicate with each other via the communication bus. The memory stores at least one executable instruction, which causes the processor to execute the deep peak-shaving control method for thermal power units disclosed in the above embodiment.

[0080] As a preferred embodiment, a computer-readable storage medium is disclosed, storing at least one executable instruction. When executed on an electronic device, the executable instruction causes the electronic device to perform the operation of the deep peak-shaving control method for thermal power units as described in the above embodiments. More specific examples of the computer-readable storage medium in this disclosure may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. The computer-readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable signal medium may also be any readable medium other than a readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0081] In summary, the present invention can be configured using the DCS system of a current power plant control platform. By adopting the method of the present invention, fully automatic control of the unit within the 20%-100% rated load range can be achieved, with a load change rate of not less than 2%, and fully automatic operation during the dry-wet state transition process, with a maximum load tracking accuracy deviation of no more than 1MW.

[0082] As described above, although the invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A method for deep peak shaving control of thermal power units, characterized in that, Includes the following steps: S1. Obtain the dynamic characteristic model of N load points in the thermal power unit coordination system; S2. Based on the N dynamic characteristic models and the variable structure objective function, N multivariable predictive controllers are generated accordingly; the expression of the variable structure objective function J is as follows: In the formula, np is the prediction step size, which is generated by the fuzzy function; r1, r2, r3, and r4 are the unit load setpoint, main steam pressure setpoint, intermediate point temperature setpoint, and steam-water separator water level setpoint, respectively. , , , These are the predicted values ​​of unit load, main steam pressure, intermediate point temperature, and steam-water separator water level at the i-th sampling time, respectively; f and g are both functions of the intermediate point superheat and satisfy f+g=1; nu is the control step size, which is set to a constant value of 50. This is the actual power generation function, used to limit the magnitude of the control increment; , , These are, respectively, the incremental instructions for coal feed rate, water feed rate, and turbine valve main control at the j-th sampling time in the future; S3. Using the current actual load as the switching condition, select the multivariate predictive controller that is closest to the current load point as the controller that is currently executing the control action output.

2. The deep peak-shaving control method for thermal power units according to claim 1, characterized in that, Step S1 further includes: The dynamic characteristic models of the thermal power unit coordination system under the conditions of 20%, 30%, 50%, 70%, and 90% of the rated load were obtained and named Model 1, Model 2, Model 3, Model 4, and Model 5, respectively.

3. The deep peak-shaving control method for thermal power units according to claim 2, characterized in that, The expression for the dynamic characteristic model is as follows: In the formula, Let be the state variable at time k+1; Let k be the state variable at time k; The control quantity at time k; Let be the controlled variable at time k; A, B, and C are all coefficient matrices.

4. The deep peak-shaving control method for thermal power units according to claim 1, characterized in that, The actual power function The expression is as follows: In the formula, P is the actual generated power and P0 is the rated power; the purpose of this actual generated power function is to reduce large fluctuations in the control quantity under low load conditions, with the main objective of unit stability.

5. The deep peak-shaving control method for thermal power units according to claim 1, characterized in that, The prediction step size np adopts the following fuzzy rule: When the deviation between the actual generated power feedback value and the load setpoint is greater than or equal to the preset value, the prediction step size is reduced to achieve rapid load adjustment and improve the load change rate. When the deviation between the actual generated power feedback value and the load setpoint is less than the preset value, the prediction step size is increased to achieve stable load regulation and prevent over-adjustment oscillation. Based on the deviation between the actual generated power feedback value and the load setpoint and the prediction step size, and based on the dynamic characteristic models corresponding to 20%, 30%, 50%, 70%, and 90% of the rated load, five predictive controllers, mpc1, mpc2, mpc3, mpc4, and mpc5, are generated respectively.

6. The deep peak-shaving control method for thermal power units according to claim 5, characterized in that, Step S3 further includes: Under the condition of 80%-100% rated load, multivariable predictive control based on Model 5 is used to output MPC5; Under the condition of 60%-80% rated load, multivariable predictive control based on Model 4 is used to output MPC4; Under the load range of 40%-60% of the rated load, multivariable predictive control based on Model 3 is used to output MPC3; Under the condition of 25%-40% rated load, multivariable predictive control based on Model 2 is used to output MPC2; Under the condition of 20%-25% rated load, multivariable predictive control based on Model 1 is used to output mpc1; The expression is as follows: In the formula, P is the actual generated power and P0 is the rated power.

7. A deep peak-shaving control system for thermal power units, characterized in that, The thermal power unit deep peak shaving control system is used to execute the thermal power unit deep peak shaving control method as described in any one of claims 1 to 6; The thermal power unit deep peak shaving control system includes: The dynamic characteristic model acquisition module is used to acquire the dynamic characteristic models of N load points in the thermal power unit coordination system. The predictive controller generation module generates N multivariable predictive controllers based on the N dynamic characteristic models and the variable structure objective function. The control output module is used to select the multivariate predictive controller that is closest to the current load point as the controller that is currently executing the control action output, based on the current actual load as the switching condition.

8. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the deep peak shaving control method for thermal power units as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The storage medium stores at least one executable instruction, which, when executed on an electronic device, causes the electronic device to perform the deep peak shaving control method for thermal power units as described in any one of claims 1 to 6.

Images