Thermal power generating unit reheating steam temperature control method based on deep learning

Abstract

The invention discloses a thermal power generating unit reheating steam temperature control method based on deep learning, and the method comprises the steps: S1, collecting the historical operation data of a unit under a wide load, and selecting the steady-state working condition data in the historical operation data of the unit under the wide load; s2, inputting the steady-state working condition data of the unit into an LSTM neural network, and performing offline training to obtain a neural network model 1; s3, collecting dynamic historical data under wide-load operation of the unit, inputting the dynamic historical data into the LSTM neural network, and performing offline training to obtain a neural network model 2; and S4, according to output data of the neural network model 1 and the neural network model 2, combining real-time operation data of the unit to construct the reheating steam temperature control system under the wide load of the unit. The method solves the problems that according to an existing reheating steam temperature control method, thermal power generating units all require wide-load operation capacity, and high-load and low-load operation characteristics are difficult to consider in a control strategy.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent control of power generation processes, and specifically, relates to a reheat steam temperature control method for a thermal power unit based on deep learning. Background Art

[0002] The reheat steam temperature of a thermal power unit is an important parameter that directly affects the thermal efficiency of the unit. Improving the accuracy of automatic control of reheat steam temperature is an important means to improve the economy of the unit. The reheat steam temperature control object is a process with large inertia, large delay, and strong nonlinearity. The characteristics of this process object are mainly related to the unit load, and are also affected by changes in coal quality, cooling water volume, temperature and pressure of inlet steam, flow rate and temperature of inlet flue gas, etc. It is seriously disturbed and presents nonlinear and time-varying characteristics. The current adjustment means are difficult to play a direct role, so the reheat steam temperature damper control system faces problems such as difficulty in automatic input and poor adjustment effect.

[0003] In the Chinese invention patent with application number CN202111340430.4, a reheat steam temperature control method of a secondary reheat boiler based on flue gas damper adjustment is disclosed. In the Chinese invention patent with application number CN202311407747.4, a reheat steam temperature optimization control method, device, storage medium and equipment of a secondary reheat unit are disclosed. In the Chinese invention patent with application number CN202311044755.7, a reheat steam temperature control method, system and equipment based on a double closed-loop control structure are disclosed. In the Chinese invention patent with application number CN201911311466.2, a reheat steam temperature control method based on a neural network inverse model is disclosed.

[0004] The defect of the above patents is that they all adopt the reheat steam temperature control method based on the cascade control structure. However, current thermal power units are required to have wide load operation capabilities. The controlled characteristics of the reheat steam temperature object under wide load vary greatly, and the existing control strategy is difficult to take into account the high and low load operation characteristics. Summary of the invention

[0005] In view of the problem that thermal power units using existing reheat steam temperature control methods are all required to have wide load operation capabilities, and the control strategy is difficult to take into account both high and low load operation characteristics, the present invention provides a reheat steam temperature control method for thermal power units based on deep learning.

[0006] In order to achieve the above technical objectives, the technical solution adopted by the present invention is as follows:

[0007] A method for controlling reheat steam temperature of a thermal power unit based on deep learning, comprising the steps of:

[0008] S1. Collect historical operation data of the unit under wide load, and select steady-state operating data from the historical operation data of the unit under wide load;

[0009] S2, input the steady-state operating data of the unit into the LSTM neural network, and obtain the neural network model 1 through offline training;

[0010] S3, collect dynamic historical data of the unit under wide load operation, input it into the LSTM neural network, and obtain the neural network model 2 through offline training;

[0011] S4. Construct a reheat steam temperature control system under wide load of the unit based on the output data of neural network model 1 and neural network model 2 combined with the real-time operation data of the unit.

[0012] Furthermore, the judgment criteria for selecting the steady-state operating data in the historical operating data under the wide load of the unit are:

[0013]

[0014] in is the average load of this section of the unit, X i It represents the unit load at a certain moment in the historical operation data, σ represents the deviation threshold, and according to the demand, the threshold can be 5%, 1%, etc. The length of the data segment is set to 30 minutes. The above rules filter out 100 or more steady-state operating points.

[0015] Furthermore, the historical operation data under the unit wide load includes unit load, reheater inlet flue gas temperature, reheater inlet steam temperature, coal feed rate, coal lower calorific value, number of coal mills in operation and damper opening data values;

[0016] The unit load, reheater inlet flue gas temperature, reheater inlet steam temperature, coal feed rate, coal lower calorific value and number of coal mills in operation are used as inputs of neural network model 1, and the damper opening data value is used as output of neural network model 1.

[0017] Furthermore, the dynamic historical data of the unit under wide load operation include flue gas damper opening, reheater inlet flue gas temperature, reheater inlet steam temperature, unit load rate, unit load, number of coal mills in operation and reheater outlet steam temperature;

[0018] The flue gas damper opening, reheater inlet flue gas temperature, reheater inlet steam temperature, unit load change rate, unit load and number of coal mills in operation are the inputs of neural network model 2, and the reheater outlet steam temperature is the output of neural network model 2.

[0019] Furthermore, the reheat steam temperature control system under wide load of the unit includes a feedback loop module, a controller input module, a controller module and a feedforward signal module.

[0020] Further, the feedback loop module includes feedback loop module 1 and feedback loop module 2;

[0021] The feedback loop module 1 adopts the formula: let z2 = K*u, where K is a constant and u is the smoke damper opening control instruction;

[0022] The feedback loop module 2 adopts the formula: let z1=y1-y2, where y2 is the output of the neural network model 2, and y1 is the real-time measurement value of the reheater outlet steam temperature.

[0023] Furthermore, the controller input module adopts the formula: r2=r-z1-z2, where r is the set value of the reheater outlet steam temperature.

[0024] Furthermore, the feedforward control signal formula adopted by the feedforward signal module is: u=v1+v2, wherein v1 is the output of the neural network model 1, and v2 is the controller output.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The nonlinear characteristics of the reheat steam temperature controlled object under wide load are fitted through the neural network, and the nonlinear characteristics under wide load are compensated in the control strategy. The structure is suitable for the wide load operation of the reheat steam temperature control system. At the same time, the compensation effect of the large delay and large inertia link is improved through the compensation effect of the large delay and large inertia link through the neural network model 2, so as to realize the automatic and precise control of the reheat steam temperature under wide load. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is an overall flow chart of a method for controlling reheat steam temperature of a thermal power unit based on deep learning in an embodiment of the present invention;

[0028] Figure 2 Schematic diagram of a reheat steam temperature control method for a thermal power unit based on deep learning in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.

[0030] like Figure 1 As shown, this embodiment provides a method for controlling the reheat steam temperature of a thermal power unit based on deep learning, comprising the steps of:

[0031] S1. Collect historical operation data of the unit under wide load, and select steady-state operating data from the historical operation data of the unit under wide load;

[0032] S2, input the steady-state operating data of the unit into the LSTM neural network, and obtain the neural network model 1 through offline training;

[0033] S3, collect dynamic historical data of the unit under wide load operation, input it into the LSTM neural network, and obtain the neural network model 2 through offline training;

[0034] S4. Construct a reheat steam temperature control system under wide load of the unit based on the output data of neural network model 1 and neural network model 2 combined with the real-time operation data of the unit.

[0035] The judgment criteria for selecting steady-state operating data from the historical operating data under wide load of the unit are:

[0036]

[0037] in is the average load of this section of the unit, X i It represents the unit load at a certain moment in the historical operation data, σ represents the deviation threshold, and according to the demand, the threshold can be 5%, 1%, etc. The length of the data segment is set to 30 minutes. The above rules filter out 100 or more steady-state operating points.

[0038] The historical operation data under the unit wide load includes unit load, reheater inlet flue gas temperature, reheater inlet steam temperature, coal feed rate, coal lower calorific value, number of coal mills in operation and damper opening data values;

[0039] The unit load, reheater inlet flue gas temperature, reheater inlet steam temperature, coal feed rate, coal lower calorific value and number of coal mills in operation are used as inputs of neural network model 1, and the damper opening data value is used as output of neural network model 1.

[0040] The dynamic historical data of the unit under wide load operation include flue gas damper opening, reheater inlet flue gas temperature, reheater inlet steam temperature, unit load rate, unit load, number of coal mills in operation and reheater outlet steam temperature;

[0041] The flue gas damper opening, reheater inlet flue gas temperature, reheater inlet steam temperature, unit load change rate, unit load and number of coal mills in operation are the inputs of neural network model 2, and the reheater outlet steam temperature is the output of neural network model 2.

[0042] The reheat steam temperature control system under wide load of the unit includes a feedback loop module, a controller input module, a controller module and a feedforward signal module.

[0043] The feedback loop module includes a feedback loop module 1 and a feedback loop module 2;

[0044] The feedback loop module 1 adopts the formula: let z2 = K*u, where K is a constant and u is the smoke damper opening control instruction;

[0045] The feedback loop module 2 adopts the formula: let z1=y1-y2, where y2 is the output of the neural network model 2, and y1 is the real-time measurement value of the reheater outlet steam temperature.

[0046] The controller input module adopts the formula: r2=r-z1-z2, where r is the set value of the reheater outlet steam temperature.

[0047] The feedforward control signal formula adopted by the feedforward signal module is: u=v1+v2, where v1 is the output of the neural network model 1, and v2 is the controller output.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The nonlinear characteristics of the reheat steam temperature controlled object under wide load are fitted through the neural network, and the nonlinear characteristics under wide load are compensated in the control strategy. The structure is suitable for the wide load operation of the reheat steam temperature control system. At the same time, the compensation effect of the large delay and large inertia link is improved through the compensation effect of the large delay and large inertia link through the neural network model 2, so as to realize the automatic and precise control of the reheat steam temperature under wide load.

[0050] The above is a detailed introduction to a deep learning-based reheat steam temperature control method for a thermal power unit provided by the present application. The description of the specific embodiment is only used to help understand the method and its core idea of ​​the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A method for controlling reheat steam temperature of a thermal power unit based on deep learning, characterized in that: Includes steps: S1. Collect historical operation data of the unit under wide load, and select steady-state operating data from the historical operation data of the unit under wide load; S2, input the steady-state operating data of the unit into the LSTM neural network, and obtain the neural network model 1 through offline training; S3, collect dynamic historical data of the unit under wide load operation, input it into the LSTM neural network, and obtain the neural network model 2 through offline training; S4. Construct a reheat steam temperature control system under wide load of the unit based on the output data of neural network model 1 and neural network model 2 combined with the real-time operation data of the unit.

2. A method according to claim 1, characterized in that: The judgment criteria for selecting steady-state operating data from the historical operating data under wide load of the unit are: in is the average load of this section of the unit, X i It represents the unit load at a certain moment in the historical operation data, σ represents the deviation threshold, and the length of the data segment is set to 30 minutes.

3. A method according to claim 1, characterized in that: The historical operation data under the unit wide load includes unit load, reheater inlet flue gas temperature, reheater inlet steam temperature, coal feed rate, coal lower calorific value, number of coal mills in operation and damper opening data values; The unit load, reheater inlet flue gas temperature, reheater inlet steam temperature, coal feed rate, coal lower calorific value and number of coal mills in operation are used as inputs of neural network model 1, and the damper opening data value is used as output of neural network model 1.

4. A method according to claim 1, characterized in that: The dynamic historical data of the unit under wide load operation include flue gas damper opening, reheater inlet flue gas temperature, reheater inlet steam temperature, unit load rate, unit load, number of coal mills in operation and reheater outlet steam temperature; The flue gas damper opening, reheater inlet flue gas temperature, reheater inlet steam temperature, unit load change rate, unit load and number of coal mills in operation are the inputs of neural network model 2, and the reheater outlet steam temperature is the output of neural network model 2.

5. A method according to claim 1, characterized in that: The reheat steam temperature control system under wide load of the unit includes a feedback loop module, a controller input module, a controller module and a feedforward signal module.

6. A method according to claim 1, characterized in that: The feedback loop module includes a feedback loop module 1 and a feedback loop module 2; The feedback loop module 1 adopts the formula: let z2 = K*u, where K is a constant and u is the smoke damper opening control instruction; The feedback loop module 2 adopts the formula: let z1=y1-y2, where y2 is the output of the neural network model 2, and y1 is the real-time measurement value of the reheater outlet steam temperature.

7. A method according to claim 1, characterized in that: The controller input module adopts the formula: r2=r-z1-z2, where r is the set value of the reheater outlet steam temperature.

8. A method according to claim 1, characterized in that: The feedforward control signal formula adopted by the feedforward signal module is: u=v1+v2, where v1 is the output of the neural network model 1, and v2 is the controller output.

Images