A power plant power generation regulation method and device based on a PID algorithm and a medium

Abstract

The present application relates to a kind of power plant generation regulating method, equipment and medium based on PID algorithm, belong to power plant control technical field, including the following steps: obtaining unit operation data and power grid dispatching instruction;Estimate low calorific value of coal into furnace, determine boiler energy storage deviation;To dispatching instruction trifurcation decomposition, obtain high, medium, low frequency energy proportion;Based on dispatching instruction, obtain PID parameter by DDQN;Based on PID parameter, each frequency band energy proportion and boiler energy storage deviation, seek the bias increment of valve, secondary air volume, coal supply amount, and then generate valve opening degree, secondary air volume and coal supply amount instruction.The present application adjusts PID parameter by double deep Q network DDQN online, according to real-time load tracking error and its change rate dynamic optimization proportion, integral, differential gain, can effectively respond to dispatching instruction high frequency fluctuation, reduce output deviation, significantly improve power plant load tracking precision.

Description

Technical Field

[0001] This invention relates to a method, equipment, and medium for regulating power generation in thermal power plants based on a PID algorithm, belonging to the field of thermal power plant control technology. Background Technology

[0002] As a crucial link in grid frequency and load balance, the control quality of the power plant's power generation regulation system directly affects the safety and economy of unit operation. Currently, the mainstream coordinated control strategy is still based on PID controllers, combined with feedforward compensation to achieve coordinated regulation of turbine valves, air supply, and coal feed.

[0003] However, in actual operation, dispatch commands contain fluctuation components of different frequencies, and the response time constants of turbine valves, secondary air volume, and coal feed rate differ significantly (valves at the second level, secondary air volume at the ten-second level, and coal feed rate at the minute level). In existing technologies, load commands are often directly input to the PID controller as a single signal or distributed to each actuator after simple filtering. This fails to decompose the commands into reasonable frequency bands based on the dynamic characteristics of each stage, leading to overshoot caused by slow actuators (such as coal feed rate) responding to high-frequency disturbances, and increased wear due to frequent adjustments by fast actuators (such as valves) to low-frequency baseline components. While some existing solutions introduce load decomposition, the correspondence between the decomposed frequency bands and the three control quantities is fixed and lacks a dynamic weight allocation mechanism. This fails to adjust the compensation intensity of each loop based on real-time energy proportions, causing actuators to couple and interfere with each other in non-optimal frequency bands, potentially leading to system oscillations and reducing load tracking accuracy and unit stability. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention proposes a method, equipment, and medium for regulating power generation in thermal power plants based on a PID algorithm.

[0005] The technical solution of the present invention is as follows: On the one hand, this invention provides a method for regulating power generation in a thermal power plant based on a PID algorithm, comprising the following steps: To obtain operating data of power generation equipment in thermal power plants and dispatch instructions from the power grid; The estimated lower calorific value of coal fed into the furnace is obtained based on operational data; Determine boiler energy storage deviation based on the estimated lower calorific value of coal fed into the furnace; The scheduling command is decomposed into three frequency loads to obtain the proportion of low-frequency energy, medium-frequency energy, and high-frequency energy. The parameters of the PID algorithm are obtained based on the scheduling instructions using a dual-depth Q-network (DDQN). Based on the parameters of the PID algorithm, the high-frequency energy ratio, the medium-frequency energy ratio, the low-frequency energy ratio, and the boiler energy storage deviation, the valve command offset increment, the secondary air volume command offset increment, and the coal feed command offset increment are obtained. Valve opening command, secondary air volume command, and coal feed rate command are obtained based on valve command offset increment, secondary air volume command offset increment, and coal feed rate command offset increment.

[0006] Preferably, the operating data includes the active power of the steam turbine generator, the coal feed rate of the coal feeder, and the secondary air volume of the blower; The operating data and scheduling instructions are preprocessed to obtain preprocessed operating data and scheduling instructions, including standard active power, standard coal feed rate, standard secondary air volume, and filtered scheduling instructions.

[0007] Preferably, the estimated lower calorific value of the coal fed into the furnace is obtained based on operational data. The specific steps are as follows: The preprocessed running data is input into the hidden layer of the Extreme Learning Machine (ELM) to obtain the hidden layer output vector; Obtain the supervision labels during training of Extreme Learning Machine (ELM); The monitoring tag is the equivalent calorific value of coal derived from the current active power of the steam turbine generator and the coal feed rate. The output weight vector is obtained based on the hidden layer output vector and the supervision label; The estimated lower calorific value of coal fed into the furnace is obtained based on the output weight vector and the hidden layer output vector.

[0008] Preferably, the boiler energy storage deviation is determined based on the estimated lower calorific value of the coal fed into the furnace. The specific steps are as follows: The input thermal power is obtained based on the lower heating value of the coal fed into the furnace and the standard coal feed rate; Equivalent output thermal power is obtained based on standard active power; Boiler energy storage deviation is obtained based on input thermal power and equivalent output thermal power.

[0009] Preferably, the scheduling command is decomposed into three frequency components to obtain the high-frequency energy ratio, the high-frequency energy ratio, and the high-frequency energy ratio. The specific steps are as follows: Based on the filtered scheduling instructions, the db4 wavelet Mallat algorithm is used to perform three-frequency load decomposition to obtain high-frequency detail coefficient sequences, mid-frequency detail coefficient sequences and low-frequency approximation coefficient sequences. The high-frequency energy percentage, high-frequency energy percentage, and high-frequency energy percentage are obtained based on the high-frequency detail coefficient sequence, the mid-frequency detail coefficient sequence, and the low-frequency approximation coefficient sequence.

[0010] Preferably, the db4 wavelet Mallat algorithm is used to perform three-frequency load decomposition based on the filtered scheduling instructions. The specific steps are as follows: The first-level decomposition based on the filtered scheduling instructions yields a high-frequency detail coefficient sequence and a mid-to-low-frequency approximation coefficient sequence. The second-level decomposition based on the mid-to-low frequency approximation coefficient sequence yields the mid-frequency detail coefficient sequence and the low-frequency approximation coefficient sequence.

[0011] Preferably, the parameters of the PID algorithm are obtained using a dual-depth Q-network (DDQN) based on the scheduling instructions. The specific steps are as follows: Load tracking error is obtained based on the grid-filtered dispatch instructions and the standard active power output of the steam turbine generator; The error change rate is obtained based on the load tracking error. Construct the DDQN state vector based on load tracking error, error rate of change, high-frequency detail coefficient sequence, and mid-frequency detail coefficient sequence; Based on the DDQN state vector, a normalized action vector is obtained through a Q network, including the proportional gain normalization increment, the integral gain normalization increment, and the differential gain normalization increment. The parameters of the PID algorithm, including proportional gain, integral gain, and derivative gain, are obtained based on the normalized action vector.

[0012] Preferably, the valve command offset increment of the steam turbine generator is obtained based on the high-frequency energy ratio and proportional gain; The secondary air volume command offset increment of the blower is obtained based on integral gain, intermediate frequency energy ratio, and boiler energy storage deviation. The feed rate command offset increment of the coal feeder is obtained based on differential gain and low-frequency energy ratio.

[0013] In another aspect, the present invention also provides an electronic device having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the power generation regulation method for thermal power plants based on the PID algorithm as described in any embodiment of the present invention.

[0014] In another aspect, the present invention also provides a computer-readable storage medium for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the PID algorithm-based power generation regulation method for thermal power plants as described in any embodiment of the present invention.

[0015] The present invention has the following beneficial effects: 1. This invention employs wavelet transform to decompose the dispatching command into three frequency components: high frequency, medium frequency, and low frequency, which correspond to the adjustment of turbine valves, secondary air volume of the forced draft fan, and coal feeder, respectively. This frequency division control strategy fully considers the time scale difference between boiler heat storage and turbine response, enabling each actuator to operate within its preferred frequency band, avoiding mutual interference and oscillation between different loops, and improving the overall coordinated control quality.

[0016] 2. This invention utilizes an extreme learning machine to estimate the lower heating value of the coal fed into the furnace in real time and calculates the boiler energy storage deviation accordingly. When the coal quality changes, the system can promptly detect the change in fuel calorific value and compensate the control quantity through feedback of the boiler energy storage deviation. This enables the unit to maintain a stable combustion state and coordinated control performance even when burning different types of coal or when the coal quality fluctuates significantly, thus broadening the unit's adaptability to fuel changes.

[0017] 3. This invention uses a dual-depth Q-network to adaptively adjust PID parameters online, enabling the controller to dynamically optimize proportional, integral, and derivative gains based on real-time load tracking errors and error change rates. Compared to traditional fixed-parameter PID control, this adaptive mechanism can effectively address the high-frequency fluctuations in power grid dispatch commands, reduce the deviation between actual output and dispatch commands, and thus significantly improve the load tracking accuracy of thermal power units. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the implementation of the method in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0021] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0023] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0024] Example 1: See Figure 1 This embodiment provides a method for regulating power generation in a thermal power plant based on a PID algorithm, including the following steps: S1. Obtain operating data of power generation equipment in thermal power plants and dispatch instructions from the power grid; In at least one embodiment, the power generation equipment of the thermal power plant includes a steam turbine generator, a coal feeder, and a blower, and the operating data includes the active power of the steam turbine generator, the coal feed rate of the coal feeder, and the secondary air volume of the blower. S2. Preprocess the running data and scheduling instructions to obtain preprocessed running data and scheduling instructions; The preprocessing of scheduling instructions specifically involves filtering, which can be expressed by the following formula: ; In the formula, Indicates that the power grid is The time-filtered dispatch command represents the target active power that the turbine generator should output at the moment, and serves as the benchmark for load tracking control. This represents the window length of the moving average filter. Indicates that the power grid is Time-based scheduling instructions; The preprocessing of the operational data specifically involves standardization, expressed by the following formula: ; ; ; In the formula, Indicates that the steam turbine generator is in The standard active power output at all times. Indicates that the steam turbine generator is in The active power output at all times. This indicates the rated active power of the steam turbine generator. Indicates that the coal feeder is in Standard coal feed rate at any given time Indicates that the coal feeder is in Coal feed rate at any given time This indicates the rated coal feed rate of the coal feeder. Indicates that the blower is in Standard secondary air volume at any time Indicates that the blower is in Secondary air volume at any time This indicates the rated air volume of the blower.

[0025] S3. Based on the preprocessed operating data, the Extreme Learning Machine (ELM) is used to obtain the estimated value of the lower heating value of the coal fed into the furnace. The specific steps are as follows: S301. Input the preprocessed running data into the hidden layer of the Extreme Learning Machine (ELM) to obtain the hidden layer output vector, expressed by the formula: ; In the formula, This indicates that the Extreme Learning Machine is in The hidden layer output vector at time 1. express Activation function This represents the hidden layer weight matrix. This represents the hidden layer bias vector; S302. Obtain the supervision labels during Extreme Learning Machine (ELM) training, expressed as a formula: ; In the formula, This indicates that the Extreme Learning Machine is in The monitoring labels, based on the current active power of the steam turbine generator and the coal feed rate, combined with the equivalent calorific value of the coal derived from the equipment efficiency, serve as the real-time targets for Extreme Learning Machine (ELM) training, enabling the network to track changes in coal quality. This represents the heat-to-work conversion efficiency coefficient of a steam turbine generator. Indicates in The boiler combustion efficiency coefficient at any given time; The boiler combustion efficiency coefficient is expressed by the formula: ; In the formula, Indicates the rated operating condition reference combustion efficiency. This represents the air volume influence coefficient. Indicates the air volume sensitivity coefficient; S303. Obtain the output weight vector based on the hidden layer output vector and the supervision label, expressed by the formula: ; In the formula, This indicates that the Extreme Learning Machine is in The output weight vector at each time step determines the contribution of each neuron in the hidden layer to the final heat output. This indicates that the Extreme Learning Machine is in The output weight vector at time 10:00. Represents the hidden layer output vector transpose; S304. Based on the output weight vector and the hidden layer output vector, the estimated value of the lower heating value of the coal fed into the furnace is obtained, expressed by the formula: ; In the formula, Indicates in The estimated lower heating value of coal fed into the furnace at a given time represents the heat released by the complete combustion of a unit mass of coal at that time.

[0026] In at least one embodiment, the initial output weight vector The elements are in The interval is randomly generated according to a uniform distribution.

[0027] S4. Determine the boiler energy storage deviation based on the estimated lower calorific value of the coal fed into the furnace. The specific steps are as follows: S401. The input thermal power is obtained based on the lower heating value of the coal fed into the furnace and the standard coal feed rate, expressed by the formula: ; In the formula, Indicates in The instantaneous input heat power released by fuel combustion and transferred to the boiler heating surface at any moment represents the effective heat energy released by fuel combustion per unit time. S402. Obtain the equivalent output heat power, expressed by the formula: ; In the formula, Indicates in The equivalent output heat power absorbed by the steam turbine generator from the boiler side at any given time is characterized as the heat that the boiler must release to meet the current power generation load. S403. The boiler energy storage deviation is obtained based on the input thermal power and the equivalent output thermal power, expressed by the formula: ; In the formula, Indicates in The boiler energy storage deviation at any given time is positive (energy storage is higher than the rated value) and negative (energy storage is lower than the rated value). Indicates in Boiler energy storage deviation at any given time It represents the time interval between two adjacent moments.

[0028] S5. Based on the preprocessed scheduling instructions, the db4 wavelet Mallat algorithm is used to perform a three-frequency load decomposition to obtain the high-frequency detail coefficient sequence, the mid-frequency detail coefficient sequence, and the low-frequency approximation coefficient sequence. The specific steps are as follows: S501. Based on the preprocessed scheduling instructions, the first-level decomposition yields the high-frequency detail coefficient sequence and the mid-to-low-frequency approximation coefficient sequence, expressed by the formula: ; ; In the formula, This represents the sequence of high-frequency detail coefficients obtained from the first-level wavelet decomposition. Each element corresponds to a frequency band. Hz represents the component with the fastest rate of change in the load command, corresponding to transient disturbances for which the steam turbine generator's inlet valves require a rapid response. This represents the first... One coefficient, This indicates the filtered sequence of scheduling instructions. The value at that location, This represents the sequence of low-to-mid-frequency approximation coefficients obtained from the first-level wavelet decomposition. Each element corresponds to a frequency band. Hz, This represents the first db4 wavelet low-pass decomposition filter. One coefficient, The eight non-zero coefficients used to iterate over the db4 filter are the shift values ​​in the convolution operation. S502. Based on the mid-to-low frequency approximation coefficient sequence, a second-level decomposition is performed to obtain the mid-frequency detail coefficient sequence and the low-frequency approximation coefficient sequence, expressed by the formula: ; ; In the formula, This represents the sequence of mid-frequency detail coefficients obtained from the second-level wavelet decomposition. Each element corresponds to a frequency band. Hz represents the medium-rate-of-change component in the load command, corresponding to the medium-frequency disturbances that need to be released or replenished by the boiler energy storage (secondary air volume of the blower). Represents the low-to-medium frequency approximation coefficient sequence exist The value at that location, This represents the sequence of low-frequency approximation coefficients obtained from the second-level wavelet decomposition. Each element corresponds to a frequency band. Hz represents the slowest changing baseline component in the load command, corresponding to the average load level that the coal feeder needs to track over a long period of time.

[0029] S6. Obtain the high-frequency energy ratio, high-frequency energy percentage, and high-frequency energy percentage based on the high-frequency detail coefficient sequence, mid-frequency detail coefficient sequence, and low-frequency approximation coefficient sequence. The specific steps are as follows: S601. Based on the high-frequency detail coefficient sequence, the mid-frequency detail coefficient sequence, and the low-frequency approximation coefficient sequence, the total energy of the high-frequency component, the total energy of the mid-frequency component, and the total energy of the low-frequency component are obtained, expressed by the formula: ; ; ; In the formula, This indicates the length of the time window for energy calculation. This represents the total energy of the high-frequency components within the window. This represents the total energy of the intermediate frequency components within the window. This represents the total energy of the low-frequency components within the window; S602. Normalize the total energy of the high-frequency components, the total energy of the mid-frequency components, and the total energy of the low-frequency components to obtain the high-frequency energy ratio, high-frequency energy percentage, and high-frequency energy percentage, which are expressed by the following formula: ; ; ; In the formula, This indicates the proportion of high-frequency energy and characterizes the control weight of the turbine generator valves. This indicates the proportion of medium-frequency energy and represents the control weight of the secondary air volume of the blower. This indicates the proportion of low-frequency energy and represents the control weight of the coal feeder's coal feed rate.

[0030] S7. Based on the grid-filtered dispatch instructions, the standard active power output of the turbine generator, the high-frequency detail coefficient sequence, and the medium-frequency detail coefficient sequence, the parameters of the PID algorithm are obtained using a dual-depth Q-network (DDQN). The specific steps are as follows: S701. The load tracking error is obtained based on the grid-filtered dispatch command and the standard active power output of the turbine generator, expressed by the formula: ; In the formula, Indicates in The load tracking error at any given time characterizes the deviation between the actual power of the steam turbine generator and the set value. A positive value indicates insufficient output, and a negative value indicates excessive output. It is the core feedback quantity of the control. S702. Obtain the error change rate based on the load tracking error, expressed by the formula: ; In the formula, Indicates in The rate of change of error at any given time; S703. Based on load tracking error, error change rate, high-frequency detail coefficient sequence, and mid-frequency detail coefficient sequence, a DDQN state vector is constructed, expressed by the formula: ; In the formula, Indicates in The DDQN state vector at time t. This represents the largest element in the high-frequency detail coefficient sequence obtained from the first-level wavelet decomposition. This represents the largest element in the mid-frequency detail coefficient sequence obtained from the second-level wavelet decomposition. Representing vectors transpose; In at least one embodiment, each element of the DDQN state vector is normalized to eliminate the influence of dimensions. S704. Based on the DDQN state vector, a normalized action vector is obtained through a Q-network, including the proportional gain normalization increment, the integral gain normalization increment, and the differential gain normalization increment, expressed by the formula: ; In the formula, Indicates in Normalized action vector at time step, Indicates in The normalized increment of the proportional gain at time step, Indicates in The normalized increment of the integral gain at time step [time]. Indicates in The normalized increment of the differential gain at time step; S705. Obtain the parameters of the PID algorithm based on the normalized action vector, including proportional gain, integral gain, and derivative gain, expressed by the formula: ; ; ; In the formula, Indicates in The proportional gain at any given moment determines the instantaneous amplification factor of the current error on the turbine generator valve command. This represents the initial set value of the proportional gain. The scaling factor represents the proportional gain. Indicates in The integral gain at any given time determines the strength of the impact of the accumulated error on the secondary air volume command of the blower. This represents the initial set value of the integral gain. This represents the scaling factor for the integral gain. Indicates in The differential gain at any given time determines the strength of the impact of the error rate of change on the coal feeder's coal feed command. This represents the initial set value of the differential gain. This represents the scaling factor of the differential gain.

[0031] S8. Based on the PID algorithm, the parameters, high-frequency energy ratio, medium-frequency energy ratio, low-frequency energy ratio, and boiler energy storage deviation are used to obtain the valve command offset increment, secondary air volume command offset increment, and coal feed command offset increment. The specific steps are as follows: S801. The valve command offset increment of the steam turbine generator is obtained based on the high-frequency energy ratio and proportional gain, expressed by the formula: ; In the formula, Indicates that the steam turbine generator is in The valve command offset increment at any given moment represents the additional percentage opening (or reduction) required by the turbine generator inlet valve to drive its operation, focusing on rapid response to high-frequency components. This represents the feedforward coefficient for the valve error of the steam turbine generator. Indicates in The high-frequency detail coefficient sequence index corresponding to each time step; The high-frequency detail coefficient sequence index is expressed by the formula: ; In the formula, This represents the rounding function; S802. The secondary air volume command offset increment of the blower is obtained based on integral gain, intermediate frequency energy ratio, and boiler energy storage deviation, expressed by the formula: ; In the formula, Indicates that the blower is in The increment of the secondary air volume command offset at any given time represents the percentage increase (or decrease) in the secondary air volume of the blower, driving the adjustment of the blower's secondary air volume, focusing on medium-frequency energy storage compensation. This represents the feedforward coefficient for boiler energy storage deviation. Indicates in The mid-frequency detail coefficient sequence index corresponding to the time step; S803. The feed rate command offset increment of the coal feeder is obtained based on the differential gain and the proportion of low-frequency energy, expressed by the formula: ; In the formula, Indicates that the coal feeder is in The increment of the coal feed command offset at any given moment. Represents the fuel feedforward coefficient. Indicates the index of the low-frequency approximation coefficient sequence; The mid-frequency detail coefficient sequence index and the low-frequency approximation coefficient sequence index are expressed by the following formula: .

[0032] S9. The valve opening command, secondary air volume command, and coal feed rate command are obtained by rate limiting the valve command offset increment, secondary air volume command offset increment, and coal feed rate command offset increment. The specific steps are as follows: S901. Rate limiting is applied to the valve command offset increment, secondary air volume command offset increment, and coal feed command offset increment to obtain the actual valve command offset increment, actual secondary air volume command offset increment, and actual coal feed command offset increment, expressed by the formula: ; ; ; In the formula, Indicates that the steam turbine generator is in The actual valve command offset increment at any given moment. Indicates that the blower is in The actual secondary air volume command offset increment at any given time. Indicates that the coal feeder is in The actual coal feed instruction offset increment at any given moment. This represents a saturation function that restricts the input to a preset range, for example, by... Limited to the range Inside, This indicates the maximum permissible increment of the turbine generator control valve. This indicates the maximum permissible increment of the blower. This indicates the maximum permissible increment of the coal feeder; S902. Based on the actual valve command offset increment, the actual secondary air volume command offset increment, and the actual coal feed rate command offset increment, the valve opening command, secondary air volume command, and coal feed rate command are obtained, expressed by the following formula: ; ; ; In the formula, Indicates that the steam turbine generator is in The valve opening command at any given time. Indicates that the steam turbine generator is in The valve opening command at any given time. Indicates that the blower is in Secondary air volume command at any time. Indicates that the blower is in Secondary air volume command at any time. Indicates that the coal feeder is in The coal feed rate instruction at any given time. Indicates that the coal feeder is in The coal feed rate instruction at any given time.

[0033] S10. The instant reward value is obtained based on the actual valve command offset increment, the actual secondary air volume command offset increment, and the actual coal feed command offset increment, expressed by the formula: ; ; In the formula, Indicates in Instant reward value at any moment , , This represents the reward weighting coefficient. Indicates in The dead zone penalty function value at time 1. Represents the maximum value function. This represents the dead zone threshold for load tracking; The parameters of the dual-depth Q network (DDQN) are updated based on the instant reward value, including the Q network and the DDQN target network. This is existing technology and will not be described in detail here.

[0034] Example 2: This embodiment provides an electronic device that stores a computer program. When the computer program is executed by a processor, it implements the power generation regulation method for thermal power plants based on the PID algorithm as described in any embodiment of the present invention.

[0035] Example 3: This embodiment provides a computer-readable storage medium for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the PID algorithm-based power generation regulation method for thermal power plants as described in any embodiment of the present invention.

[0036] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0037] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0038] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0039] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0040] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for regulating power generation in a thermal power plant based on a PID algorithm, characterized in that, Includes the following steps: To obtain operating data of power generation equipment in thermal power plants and dispatch instructions from the power grid; The estimated lower calorific value of coal fed into the furnace is obtained based on operational data; Determine boiler energy storage deviation based on the estimated lower calorific value of coal fed into the furnace; The scheduling command is decomposed into three frequency loads to obtain the proportion of low-frequency energy, medium-frequency energy, and high-frequency energy. The parameters of the PID algorithm are obtained based on the scheduling instructions using a dual-depth Q-network (DDQN). Based on the parameters of the PID algorithm, the high-frequency energy ratio, the medium-frequency energy ratio, the low-frequency energy ratio, and the boiler energy storage deviation, the valve command offset increment, the secondary air volume command offset increment, and the coal feed command offset increment are obtained. Valve opening command, secondary air volume command, and coal feed rate command are obtained based on valve command offset increment, secondary air volume command offset increment, and coal feed rate command offset increment.

2. The power generation regulation method for thermal power plants based on PID algorithm according to claim 1, characterized in that, The operating data includes the active power of the steam turbine generator, the coal feed rate of the coal feeder, and the secondary air volume of the blower. The operating data and scheduling instructions are preprocessed to obtain preprocessed operating data and scheduling instructions, including standard active power, standard coal feed rate, standard secondary air volume, and filtered scheduling instructions.

3. The power generation regulation method for thermal power plants based on PID algorithm according to claim 1, characterized in that, The specific steps for obtaining the estimated lower calorific value of coal fed into the furnace based on operational data are as follows: The preprocessed running data is input into the hidden layer of the Extreme Learning Machine (ELM) to obtain the hidden layer output vector; Obtain the supervision labels during training of Extreme Learning Machine (ELM); The monitoring tag is the equivalent calorific value of coal derived from the current active power of the steam turbine generator and the coal feed rate. The output weight vector is obtained based on the hidden layer output vector and the supervision label; The estimated lower calorific value of coal fed into the furnace is obtained based on the output weight vector and the hidden layer output vector.

4. The power generation regulation method for thermal power plants based on PID algorithm according to claim 1, characterized in that, The boiler energy storage deviation is determined based on the estimated lower calorific value of the coal fed into the furnace. The specific steps are as follows: The input thermal power is obtained based on the lower heating value of the coal fed into the furnace and the standard coal feed rate; Equivalent output thermal power is obtained based on standard active power; Boiler energy storage deviation is obtained based on input thermal power and equivalent output thermal power.

5. The power generation regulation method for thermal power plants based on PID algorithm according to claim 1, characterized in that, The scheduling command is decomposed into three frequency components to obtain the high-frequency energy ratio, high-frequency energy ratio, and high-frequency energy ratio. The specific steps are as follows: Based on the filtered scheduling instructions, the db4 wavelet Mallat algorithm is used to perform three-frequency load decomposition to obtain high-frequency detail coefficient sequences, mid-frequency detail coefficient sequences and low-frequency approximation coefficient sequences. The high-frequency energy percentage, high-frequency energy percentage, and high-frequency energy percentage are obtained based on the high-frequency detail coefficient sequence, the mid-frequency detail coefficient sequence, and the low-frequency approximation coefficient sequence.

6. The power generation regulation method for thermal power plants based on PID algorithm according to claim 5, characterized in that, Based on the filtered scheduling instructions, the db4 wavelet Mallat algorithm is used to perform three-frequency load decomposition. The specific steps are as follows: The first-level decomposition based on the filtered scheduling instructions yields a high-frequency detail coefficient sequence and a mid-to-low-frequency approximation coefficient sequence. The second-level decomposition based on the mid-to-low frequency approximation coefficient sequence yields the mid-frequency detail coefficient sequence and the low-frequency approximation coefficient sequence.

7. The power generation regulation method for thermal power plants based on PID algorithm according to claim 1, characterized in that, The parameters of the PID algorithm are obtained based on the scheduling instructions using a dual-depth Q-network (DDQN). The specific steps are as follows: Load tracking error is obtained based on the grid-filtered dispatch instructions and the standard active power output of the steam turbine generator; The error change rate is obtained based on the load tracking error. Construct the DDQN state vector based on load tracking error, error rate of change, high-frequency detail coefficient sequence, and mid-frequency detail coefficient sequence; Based on the DDQN state vector, a normalized action vector is obtained through a Q network, including the proportional gain normalization increment, the integral gain normalization increment, and the differential gain normalization increment. The parameters of the PID algorithm, including proportional gain, integral gain, and derivative gain, are obtained based on the normalized action vector.

8. The power generation regulation method for thermal power plants based on PID algorithm according to claim 7, characterized in that, Valve command offset increment of steam turbine generator is obtained based on high-frequency energy ratio and proportional gain; The secondary air volume command offset increment of the blower is obtained based on integral gain, intermediate frequency energy ratio, and boiler energy storage deviation. The feed rate command offset increment of the coal feeder is obtained based on differential gain and low-frequency energy ratio.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 8.

Images