Power grid frequency control method and device of AGC system, equipment and storage medium

Abstract

The application provides a power grid frequency control method, device and equipment of an AGC system and a storage medium. In the application, the active power change values of generators and new energy stations in a control area of the AGC system and the corresponding power grid frequency change values under a severe power grid frequency fluctuation condition are selected as training samples, a neural network carrying a mapping relationship between the active power change values and the power grid frequency change values is obtained, under a current severe power grid frequency fluctuation condition, the neural network, the current active power change values of the generators and the new energy stations are used to predict the current power grid frequency change value, and the predicted current power grid frequency change value is superimposed into generation of a frequency control target value of the AGC system for the control area, so that the AGC system can respond to the power grid frequency change in advance and pre-regulate the power grid frequency change, improve the fast response capability of the AGC system under an emergency power grid frequency change condition, and strengthen the system stability.

Description

Technical Field

[0001] This application relates to the field of power dispatch automation technology, and in particular to a power grid frequency control method, device, AGC system, storage medium and computer program product for an AGC system. Background Technology

[0002] An AGC (Automatic Generation Control) system is a power system based on AGC, which falls under the category of secondary frequency regulation in power systems and is a crucial component of the Energy Management System (EMS). In response to random load disturbances in an AGC system, each AGC unit rapidly adjusts its output to stabilize the frequency of the supplied power. The overall structure of AGC systems is becoming increasingly complex, and the disturbances they face are becoming more diverse and complex. With the rapid development of new and renewable energy sources, wind power, solar power, and other renewable energy sources are being widely applied to power systems. As intermittent and highly fluctuating power sources, these sources cause more frequent and severe frequency fluctuations in the power grid. The power grid must rely on AGC to adjust the load power sources within its control area to maintain grid frequency stability, thereby ensuring the safe and stable operation of the power grid.

[0003] The AGC system uses traditional thermal or hydropower units as regulating units. However, when traditional hydropower and thermal power units operate under a wide range of varying conditions, the furnace coordination object exhibits strong nonlinear characteristics, resulting in a large frequency regulation response delay.

[0004] To address the issue of large frequency regulation response delay, current power grid frequency control strategies for AGC systems, such as PID control, fuzzy control, and sliding mode variable structure control, have not yet provided a good solution. They are unable to fundamentally reverse the problem of the relative lag in frequency regulation of AGC systems, resulting in low stability of AGC systems. Summary of the Invention

[0005] Therefore, it is necessary to provide a power grid frequency control method, device, AGC system, storage medium, and computer program product for the above-mentioned technical problems.

[0006] A method for controlling the power grid frequency of an AGC system, the method comprising:

[0007] The active power change values ​​of generator sets, active power change values ​​of new energy power plants, active power change values ​​of loads, and grid frequency change values ​​were selected when the grid frequency in the control area of ​​the AGC system fluctuated drastically.

[0008] The active power change values ​​of the generator set, the active power change values ​​of the new energy power station, the active power change values ​​of the load, and the grid frequency change values ​​are used as training samples to obtain a neural network carrying the mapping relationship between the active power change values ​​and the grid frequency change values.

[0009] If the power grid frequency is currently experiencing severe fluctuations, then the change in active power of the generator set under the current severe fluctuation situation shall be taken as the current change in active power of the generator set, the change in active power of the new energy power station under the current severe fluctuation situation shall be taken as the current change in active power of the new energy power station, and the change in active power of the load under the current severe fluctuation situation shall be taken as the current change in active power of the load.

[0010] The current active power change value of the generator set, the current active power change value of the new energy power station, and the current active power change value of the load are input into the neural network so that the neural network can predict the current grid frequency change value based on the mapping relationship between the active power change value it carries and the grid frequency change value.

[0011] The predicted current grid frequency change value is superimposed on the frequency control target of the AGC system for the control area to calculate the regional control demand for the control area. The regional control demand is then allocated to each AGC unit of the AGC system to complete the grid frequency control of the AGC system.

[0012] A power grid frequency control device for an AGC system, the device comprising:

[0013] The first data acquisition module is used to select the active power change value of the generator set, the active power change value of the new energy power station, the active power change value of the load, and the grid frequency change value in the control area of ​​the AGC system when the grid frequency fluctuates drastically.

[0014] The training module is used to take the active power change value of the generator set, the active power change value of the new energy power station, the active power change value of the load, and the grid frequency change value as training samples to obtain a neural network carrying the mapping relationship between the active power change value and the grid frequency change value.

[0015] The second data acquisition module uses the change in active power of the generator set under the current severe fluctuation of the power grid frequency as the current active power change value of the generator set, the change in active power of the new energy power station under the current severe fluctuation of the power grid frequency as the current active power change value of the new energy power station, and the change in active power of the load under the current severe fluctuation of the load as the current active power change value of the load.

[0016] The prediction module is used to input the current active power change value of the generator set, the current active power change value of the new energy power station, and the current active power change value of the load into the neural network, so that the neural network can predict the current grid frequency change value based on the mapping relationship between the active power change value it carries and the grid frequency change value.

[0017] The frequency superposition module is used to superimpose the predicted current power grid frequency change value onto the frequency control target of the AGC system for the control area, so as to calculate the area control demand for the control area, and allocate the area control demand to each AGC unit of the AGC system to complete the power grid frequency control of the AGC system.

[0018] An AGC system includes a memory and a processor, the memory storing a computer program and the processor executing the method described above.

[0019] A computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor using the methods described above.

[0020] A computer program product having a computer program stored thereon, the computer program being executed by a processor using the above-described method.

[0021] In the aforementioned AGC system's power grid frequency control method, device, AGC system, storage medium, and computer program product, the active power change values ​​of generators, renewable energy plants, and loads within the control area of ​​the AGC system under conditions of severe power grid frequency fluctuations, along with their corresponding power grid frequency change values, are selected as training samples. An artificial intelligence algorithm is used to fit a mathematical model between the power grid frequency change values ​​and the active power change values ​​of the generators, renewable energy plants, and loads, resulting in a neural network carrying the mapping relationship between the active power change values ​​and the power grid frequency change values. Then, when the power grid frequency experiences severe fluctuations, the mapping relationship carried by this neural network, the current active power change values ​​of the generator sets, renewable energy plants, and loads are used to predict the current power grid frequency change value. This predicted current power grid frequency change value is then superimposed onto the frequency control target of the AGC system for the control area. This allows the AGC system to respond to power grid frequency changes in advance and pre-adjust to these changes, improving the AGC system's rapid response capability to emergency power grid frequency changes and enhancing system stability. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the power grid frequency control method of an AGC system in one embodiment;

[0023] Figure 2 This is a schematic diagram of the training process of the backpropagation network in one embodiment;

[0024] Figure 3 This is a schematic diagram of the frequency superposition processing of an AGC system in one embodiment;

[0025] Figure 4 This is a structural block diagram of the power grid frequency control device of an AGC system in one embodiment;

[0026] Figure 5 This is an internal structure diagram of an AGC system in one embodiment. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0028] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0029] AGC (Automatic Generation Control) systems use traditional thermal or hydropower units as regulating units. However, when these units operate under varying conditions across a wide range of conditions, they exhibit strong nonlinear characteristics, resulting in significant frequency regulation response delays, low power generation rates, and an inability to quickly and accurately track the power output commands of dynamic AGC systems. Consequently, traditional AGC systems are prone to poor dynamic control performance of the power grid and failure to meet power / frequency targets. These problems are even more severe in control areas with high renewable energy penetration and insufficient regulation resources.

[0030] To address the issue of large frequency regulation response delay, current power grid frequency control strategies for AGC systems, such as PID control, fuzzy control, and sliding mode variable structure control, have not yet provided a good solution. They are unable to fundamentally reverse the problem of the relative lag in frequency regulation of AGC systems, resulting in low stability of AGC systems.

[0031] When performing grid frequency control, the AGC system first generates a total frequency control target value based on the disturbances in the control area, and then allocates the total frequency control target value to each AGC unit according to a certain allocation strategy. Current frequency control strategies mainly include PID control, fuzzy control, and sliding mode variable structure control, which mainly focus on improving the control within the AGC system without effectively incorporating the system's characteristics. This fails to fundamentally reverse the problem of relatively lagging regulation by AGC units.

[0032] In one embodiment, such as Figure 1 As shown, a power grid frequency control method for an AGC system is provided. The method is illustrated using an AGC system as an example. In this embodiment, the power grid frequency and active power can be per-unit values. The method includes the following steps:

[0033] Step S101: Select the active power change values ​​of generator sets, active power change values ​​of new energy power plants, active power change values ​​of loads, and grid frequency change values ​​in the control area of ​​the AGC system when the grid frequency fluctuates drastically.

[0034] Step S102: The active power change value of the generator set, the active power change value of the new energy power station, the active power change value of the load, and the grid frequency change value are used as training samples to obtain a neural network carrying the mapping relationship between the active power change value and the grid frequency change value.

[0035] The drastic fluctuations in the power grid frequency mentioned in step S101 occur within a historical time period. Determining whether drastic fluctuations in the power grid frequency have occurred in the past can be achieved by obtaining the power grid frequency value at a specific historical moment and the value at the moment preceding that historical moment. If the difference between the power grid frequency value at a specific historical moment and the value at the moment preceding that historical moment is greater than a threshold, then it is determined that drastic fluctuations in the power grid frequency have occurred in the past.

[0036] Under the premise that the power control parameters of generator sets and new energy power stations remain unchanged, the active power change value and the grid frequency change value of each generator set, new energy power station and load under the condition of drastic fluctuation of grid frequency are selected as training samples of neural network. The calculation formulas for active power change value and grid frequency change value are shown in Equation (1).

[0037]

[0038] In equation (1), P G1 P R1 P L1 f1 represents the per-unit active power values ​​of the generator set, the new energy power station, the load, and the grid frequency at a given historical moment, based on the grid frequency value at a specific historical moment and the value at the previous historical moment, after detecting a fluctuation in the grid frequency. G2 P R2 P L2 f2 represents the per-unit active power of the generator set, the per-unit active power of the renewable energy power station, the per-unit active power of the load, and the per-unit frequency of the power grid at the previous moment; Δ s This is a threshold value set to detect fluctuations in the power grid frequency.

[0039] The neural network can be a back propagation network (BP). When the neural network can be a back propagation network, the active power change values ​​of the generator set, the active power change values ​​of the renewable energy power station, and the grid frequency change values ​​are used as training samples. Specifically, this includes using the active power change values ​​of the generator set, the renewable energy power station, and the load as training samples for the input layer of the back propagation network, and using the grid frequency change values ​​as training samples for the output layer of the back propagation network.

[0040] The number of hidden layer neurons in the backpropagation network is calculated according to formula (2);

[0041] N Y =2m+1 (2)

[0042] Where, N Y denoted as the number of hidden layer neurons, and m as the number of input signals to the backpropagation network;

[0043] The activation functions of the hidden layer and the output layer are shown in equation (3);

[0044]

[0045] Where y is the activation function parameter.

[0046] The input of the i-th node in the hidden layer is shown in equation (4);

[0047]

[0048] Among them, w ij x represents the weights from the i-th node in the hidden layer to the j-th node in the input layer; j θ is the j-th input of the input layer; i The threshold value for the i-th node in the hidden layer;

[0049] The output of the i-th node in the hidden layer is shown in equation (5);

[0050] uy i =σ(u) i (5)

[0051] The inputs to the output layer nodes are shown in equation (6);

[0052]

[0053] Among them, w 1j The weights from the output layer node to the j-th node in the input layer; uy j α is the j-th input to the output layer; α is the threshold of the output layer node;

[0054] The output of the output layer node is shown in equation (7);

[0055] y=σ(k) (7)

[0056] After inputting all training samples, the total training error is calculated as shown in equation (8);

[0057]

[0058] Among them, T p Let y be the expected output of the p-th sample. p This is the output of the p-th sample, where P is the number of samples.

[0059] The correction formulas for weights and thresholds are shown in equation (9);

[0060]

[0061] Where η is the learning rate.

[0062] The corrected weights and thresholds are shown in equation (10);

[0063]

[0064] Repeat the adjustment of weights and thresholds until the number of calculations t is reached or the error is less than a specified value T. At this point, the backpropagation network training is complete and it can be used to predict power grid frequency changes. For specific training procedures, refer to [reference needed]. Figure 2 .

[0065] Step S103: If the power grid frequency is currently experiencing severe fluctuations, then the change in active power of the generator set under the current severe fluctuation situation shall be taken as the current change in active power of the generator set, and the change in active power of the new energy power station under the current severe fluctuation situation shall be taken as the current change in active power of the new energy power station.

[0066] The step of determining whether the power grid frequency is currently experiencing drastic fluctuations may include: obtaining the current value of the power grid frequency and the value at the previous time; if the difference between the current value and the previous value of the power grid frequency is greater than a threshold, then it is determined that the power grid frequency is currently experiencing drastic fluctuations. This threshold may be Δ. s .

[0067] After determining that the power grid frequency is currently experiencing severe fluctuations based on the difference between the current value and the previous value, the method for obtaining the change value of the active power of the generator set under the current severe fluctuations may include: obtaining the value of the active power of the generator set at the current time and the value at the previous time, so as to obtain the change value of the active power of the generator set under the current severe fluctuations.

[0068] After determining that the power grid frequency is currently experiencing severe fluctuations based on the difference between the current value and the previous value, the method for obtaining the active power change value of the new energy power station under the current severe fluctuation situation may include: obtaining the active power value of the new energy power station at the current time and the previous value to obtain the current active power change value of the new energy power station.

[0069] After determining that the grid frequency is currently experiencing severe fluctuations based on the difference between the grid frequency value at the current moment and the value at the previous moment, the method for obtaining the active power change value of the load under the current severe fluctuation situation may include: obtaining the active power value of the load at the current moment and the value at the previous moment, so as to obtain the active power change value of the load under the current severe fluctuation situation.

[0070] Step S104: Input the current active power change value of the generator set, the current active power change value of the new energy power station, and the current active power change value of the load into the neural network, so that the neural network can predict the current grid frequency change value based on the mapping relationship between the active power change value it carries and the grid frequency change value.

[0071] Step S105: The predicted current grid frequency change value is superimposed on the frequency control target of the AGC system for the control area to calculate the area control demand for the control area, and the area control demand is allocated to each AGC unit of the AGC system to complete the grid frequency control of the AGC system.

[0072] Figure 3In this diagram, Δf represents the predicted current grid frequency change, Δf′ represents the superimposed grid frequency prediction, ACE represents the frequency control target for the control area formed by superimposing the predicted current grid frequency change, PI represents the automatic controller for real-time load tracking, and ARR represents the regional control demand for the total output command within the control area, obtained based on the frequency control target after low-pass filtering. This regional control demand is used to determine the adjustment amount of active power for generator sets, renewable energy plants, and loads within the control area. After the regional control demand is allocated to each AGC unit, each AGC unit adjusts the active power of the generator sets, renewable energy plants, and loads within its corresponding control area based on the regional control demand allocated to it. After each AGC unit adjusts the active power of its corresponding generator sets, renewable energy plants, and loads, the grid frequency of the AGC system changes, thus completing the grid frequency control of the AGC system.

[0073] Furthermore, the area control requirements are allocated to each AGC unit of the AGC system to complete the grid frequency control of the AGC system. Specifically, this includes: allocating the area control requirements to each AGC unit of the AGC system according to a preset AGC allocation algorithm to complete the grid frequency control of the AGC system.

[0074] In the aforementioned AGC system's grid frequency control method, the active power change values ​​of generators, renewable energy plants, and loads within the AGC system's control area under conditions of severe grid frequency fluctuations, along with their corresponding grid frequency change values, are selected as training samples. An artificial intelligence algorithm is used to fit a mathematical model between the grid frequency change values ​​and the active power change values ​​of the generators, renewable energy plants, and loads, resulting in a neural network carrying the mapping relationship between the active power change values ​​and the grid frequency change values. Then, when the grid frequency experiences severe fluctuations, the mapping relationship carried by this neural network, the current active power change values ​​of the generator sets, renewable energy plants, and loads are used to predict the current grid frequency change value. This predicted current grid frequency change value is then superimposed onto the frequency control target of the AGC system for the control area. This allows the AGC system to respond to grid frequency changes in advance and pre-adjust to them, improving the AGC system's rapid response capability to emergency grid frequency changes and enhancing system stability.

[0075] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0076] In one embodiment, such as Figure 4 As shown, a power grid frequency control device for an AGC system is provided, comprising:

[0077] The first data acquisition module 401 is used to select the active power change value of the generator set, the active power change value of the new energy power station, and the grid frequency change value in the control area of ​​the AGC system when the grid frequency fluctuates drastically.

[0078] The training module 402 is used to take the active power change value of the generator set, the active power change value of the new energy power station and the grid frequency change value as training samples to obtain a neural network carrying the mapping relationship between the active power change value and the grid frequency change value.

[0079] The second data acquisition module 403 is used to, if the power grid frequency is currently fluctuating drastically, take the change value of the active power of the generator set under the current drastic fluctuation as the current active power change value of the generator set, and take the change value of the active power of the new energy power station under the current drastic fluctuation as the current active power change value of the new energy power station.

[0080] The prediction module 404 is used to input the current active power change value of the generator set and the current active power change value of the new energy power station into the neural network, so that the neural network can predict the current grid frequency change value based on the mapping relationship between the active power change value and the grid frequency change value it carries.

[0081] The frequency superposition module 405 is used to superimpose the predicted current power grid frequency change value onto the frequency control target value generated by the AGC system for the control area, and to distribute the superimposed frequency control target value to each AGC unit of the AGC system to complete the power grid frequency control of the AGC system.

[0082] In one embodiment, the device further includes a severe fluctuation judgment module, used to obtain the value of the power grid frequency at the current moment and the value at the previous moment; if the difference between the value of the power grid frequency at the current moment and the value at the previous moment is greater than a threshold, then it is determined that the power grid frequency is currently experiencing severe fluctuations.

[0083] In one embodiment, the device further includes a generator set data acquisition module, which is used to determine, based on the difference between the current value of the grid frequency and the previous value, that the grid frequency is currently experiencing severe fluctuations, and then acquire the current value of the generator set's active power and the previous value to obtain the change value of the generator set's active power under the current severe fluctuation situation.

[0084] In one embodiment, the device further includes a new energy power station data acquisition module, which is used to determine, based on the difference between the current value of the power grid frequency and the previous value, that the power grid frequency is currently experiencing a drastic fluctuation, and then acquire the current value of the active power of the new energy power station and the previous value to obtain the current active power change value of the new energy power station.

[0085] In one embodiment, the neural network is a backpropagation network; the training module 402 is further configured to use the active power change value of the generator set and the active power change value of the new energy power station as training samples for the input layer of the backpropagation network, and the grid frequency change value as training samples for the output layer of the backpropagation network.

[0086] In one embodiment, the frequency superposition module 405 is further configured to allocate the regional control requirements to each AGC unit of the AGC system according to a preset AGC allocation algorithm, so as to complete the grid frequency control of the AGC system.

[0087] Specific limitations regarding the power grid frequency control device of the AGC system can be found in the limitations of the power grid frequency control method for the AGC system mentioned above, and will not be repeated here. Each module in the aforementioned power grid frequency control device of the AGC system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of the AGC system in hardware form or independent of it, or they can be stored in the memory of the AGC system in software form, so that the processor can call and execute the corresponding operations of each module.

[0088] In one embodiment, an AGC system is provided, the internal structure of which can be shown as follows: Figure 5As shown, the AGC system includes a processor, memory, and network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores the power grid frequency control data for the AGC system. The network interface is used for communication with external terminals via a network connection. The AGC system also includes input / output interfaces, which are connection circuits between the processor and external devices for exchanging information; they are connected to the processor via a bus and are referred to as I / O interfaces. When the computer program is executed by the processor, it implements a power grid frequency control method for the AGC system.

[0089] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the AGC system to which the present application is applied. A specific AGC system may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0090] In one embodiment, an AGC system is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the various method embodiments described above.

[0091] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the various method embodiments described above.

[0092] In one embodiment, a computer program product is provided having a computer program stored thereon, the computer program being executed by a processor of the steps described in the various method embodiments above.

[0093] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0094] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0095] The above embodiments are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method of grid frequency control for an AGC system, characterized by, The method comprises: selecting, in the case of a sharp fluctuation of the power grid frequency in a control area of an AGC system, an active power change value of a generator set in the control area, an active power change value of a new energy field station, an active power change value of a load, and a power grid frequency change value; using the active power change value of the generator set, the active power change value of the new energy field station, the active power change value of the load, and the power grid frequency change value as training samples to obtain a neural network carrying a mapping relationship between the active power change value and the power grid frequency change value; if the power grid frequency is currently fluctuating sharply, using the active power change value of the generator set in the case of the current sharp fluctuation as the current active power change value of the generator set, using the active power change value of the new energy field station in the case of the current sharp fluctuation as the current active power change value of the new energy field station, and using the active power change value of the load in the case of the current sharp fluctuation as the current active power change value of the load; inputting the current active power change value of the generator set, the current active power change value of the new energy field station, and the current active power change value of the load into the neural network to enable the neural network to predict a current power grid frequency change value based on the mapping relationship between the active power change value and the power grid frequency change value carried by the neural network itself; superimposing the predicted current power grid frequency change value on a frequency control target of the AGC system for the control area to calculate a regional control demand for the control area, and distributing the regional control demand to each AGC unit of the AGC system to complete power grid frequency control of the AGC system.

2. The method of claim 1, wherein, The step of determining that the power grid frequency is currently fluctuating sharply comprises: obtaining a value of the power grid frequency at a current time and a value of the power grid frequency at a previous time; if a difference between the value of the power grid frequency at the current time and the value of the power grid frequency at the previous time is greater than a threshold value, determining that the power grid frequency is currently fluctuating sharply.

3. The method of claim 2, wherein, The step of obtaining the active power change value of the generator set in the case of the current sharp fluctuation comprises: after determining that the power grid frequency is currently fluctuating sharply based on a difference between the value of the power grid frequency at the current time and the value of the power grid frequency at the previous time, obtaining a value of the active power of the generator set at the current time and a value of the active power of the generator set at the previous time to obtain the active power change value of the generator set in the case of the current sharp fluctuation.

4. The method of claim 2, wherein, The step of obtaining the active power change value of the new energy field station in the case of the current sharp fluctuation comprises: after determining that the power grid frequency is currently fluctuating sharply based on a difference between the value of the power grid frequency at the current time and the value of the power grid frequency at the previous time, obtaining a value of the active power of the new energy field station at the current time and a value of the active power of the new energy field station at the previous time to obtain the current active power change value of the new energy field station.

5. The method of claim 1, wherein, The neural network is a back propagation network. The active power change value of the generator set, the active power change value of the new energy station, the active power change value of the load, and the grid frequency change value are taken as training samples. The active power change value of the generator set, the active power change value of the new energy station, and the active power change value of the load are taken as training samples of an input layer of a back propagation network, and the grid frequency change value is taken as a training sample of an output layer of the back propagation network.

6. The method according to any one of claims 1 to 5, characterized in that, The regional control demand is distributed to each AGC unit of the AGC system to complete grid frequency control of the AGC system. The regional control demand is distributed to each AGC unit of the AGC system to complete grid frequency control of the AGC system according to a preset AGC distribution algorithm.

7. A grid frequency control device for an AGC system, characterized by The device comprises: The first data acquisition module is configured to select, in a case where the grid frequency in a control region of the AGC system fluctuates sharply, an active power change value of a generator set in the control region, an active power change value of a new energy station, an active power change value of a load, and a grid frequency change value. The training module is configured to take the active power change value of the generator set, the active power change value of the new energy station, the active power change value of the load, and the grid frequency change value as training samples to obtain a neural network carrying a mapping relationship between the active power change value and the grid frequency change value. The second data acquisition module is configured to, if the grid frequency currently fluctuates sharply, take an active power change value of the generator set in a case where the grid frequency currently fluctuates sharply as a current active power change value of the generator set, take an active power change value of the new energy station in the case where the grid frequency currently fluctuates sharply as a current active power change value of the new energy station, and take an active power change value of the load in the case where the grid frequency currently fluctuates sharply as a current active power change value of the load. The prediction module is configured to input the current active power change value of the generator set, the current active power change value of the new energy station, and the current active power change value of the load into the neural network to enable the neural network to predict a current grid frequency change value based on the mapping relationship between the active power change value and the grid frequency change value carried by the neural network. The frequency superposition module is configured to superimpose the predicted current grid frequency change value on a frequency control target of the AGC system for the control region to calculate a regional control demand for the control region, and distribute the regional control demand to each AGC unit of the AGC system to complete grid frequency control of the AGC system.

8. An AGC system comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to implement the method in any one of claims 1 to 6.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method in any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the method in any one of claims 1 to 6.

Images