Unit load pre-control method based on battery energy storage state

Abstract

The present invention belongs to the technical field of automated control of energy storage in thermal power generation, and relates to a unit load pre-control method based on the battery energy storage state. The invention includes determining the state of a thermal power generation unit based on the AGC command and the current load of the thermal power generation unit, determining whether the energy storage output of the energy storage system satisfies the AGC command, and if it does not, establishing an initial thermal storage coefficient and a predicted thermal storage dataset of the thermal power generation unit based on the state of the thermal power generation unit, constructing a thermal storage coefficient prediction model for the thermal power generation unit, inputting operating data of the thermal power generation unit into a trained thermal storage coefficient prediction model for the thermal power generation unit to generate a predicted thermal storage coefficient for the thermal power generation unit, establishing a correspondence between the predicted thermal storage coefficient for the thermal power generation unit and the pre-load amount of the load command, generating a load command, and executing it by the thermal power generation unit. The present invention solves the problem in the prior art of assigning frequency adjustment tasks to energy storage systems and conventional thermal power generation units based on a fixed ratio coefficient, which results in excessively high unit loads.

Description

[Technical Field]

[0001] This invention belongs to the technical field of automated control of thermal power generation energy storage, and more specifically, relates to a unit load pre-control method based on the battery energy storage state. [Background technology]

[0002] With the expansion of wind and solar power grid connections, thermal power generation units are gradually changing their role in the power grid, providing more flexible and efficient auxiliary services. However, when thermal power generation units are involved in frequency regulation of the power grid, they can suffer serious wear due to the characteristics of the equipment, causing a short-term decrease in power throughput capacity, affecting the load regulation potential, further extending the system's response time, and reducing its frequency regulation capability, making it difficult to achieve the desired frequency regulation target. The development of battery energy storage technology provides important technical support to new energy power systems, and its rapid response characteristics can effectively reduce the frequency regulation burden on conventional thermal power generation units and improve the quality of frequency regulation services.

[0003] Currently, when a unit configured with battery energy storage receives an Automatic Generating Control (AGC) command from the power grid, it generally assigns frequency regulation tasks to the energy storage system and general thermal power generation units based on a fixed proportionality coefficient. However, this method has several limitations: for thermal power generation units, it is not possible to flexibly determine whether their output can meet load distribution commands based on the unit's actual operating conditions, nor can the unit's heat storage capacity be fully utilized, thereby preventing them from fully realizing their frequency regulation potential. For battery energy storage systems, if overcharging or over-discharging occurs in the energy storage system, its response capability is limited, and it cannot effectively execute the assigned commands. Furthermore, the number of charge-discharge cycles of the energy storage system changes with fluctuations in the frequency regulation signal, which can adversely affect the lifespan of the energy storage system. [Overview of the project] [Problems that the invention aims to solve]

[0004] The object of the present invention is to provide a method for pre-controlling unit loads based on the battery energy storage state, thereby solving the problem in the conventional art of assigning frequency adjustment tasks to energy storage systems and conventional thermal power generation units according to a fixed proportionality coefficient, which results in excessively high unit loads. [Means for solving the problem]

[0005] The technical solution employed in this invention is a unit load pre-control method based on the battery energy storage state, and is specifically implemented according to the following steps: Step 1: Determine the status of the thermal power generation unit based on the AGC command and the current load of the thermal power generation unit. Step 2: Determine whether the energy storage output of the energy storage system meets the AGC directive. If the energy storage output does not meet the AGC directive, perform steps 3 through 6. If the energy storage output meets the AGC directive, do not perform these steps. Step 3: Based on the state of the thermal power generation unit, obtain the initial heat storage coefficient of the thermal power generation unit, and further construct a predicted heat storage dataset for the thermal power generation unit. Step 4: Construct a thermal energy storage coefficient prediction model for thermal power generation units and train the thermal energy storage coefficient prediction model for thermal power generation units based on the thermal energy storage dataset of thermal power generation units predicted in Step 3. Step 5: Input the operating data of the thermal power generation unit into the thermal power generation unit heat storage coefficient prediction model trained in Step 4 to generate the predicted heat storage coefficient of the thermal power generation unit, and establish a correspondence between the predicted heat storage coefficient of the thermal power generation unit and the preceding load command. Step 6: Based on the state of the thermal power generation unit and the energy storage charge state of the energy storage system, a load command to be executed by the thermal power generation unit is generated, and based on the correspondence between the predicted thermal power generation unit's heat storage coefficient and the preceding load command, the corresponding preceding load command is determined based on the predicted thermal power generation unit's heat storage coefficient, and the load command to be executed by the thermal power generation unit is dynamically adjusted based on the preceding load command.

[0006] Preferably, the state of the thermal power generation unit in step 1 includes a load increase, a load decrease, and a stable state, where the load increase is when the AGC command is greater than the unit's current load, the load decrease is when the AGC command is less than the unit's current load, and the stable state is when the AGC command is equal to the unit's current load.

[0007] Preferably, in step 2, determining whether the energy storage output of the energy storage system satisfies the AGC directive is done as follows: If the energy storage charge state of the energy storage system is not within the adjustment range, the energy storage output of the energy storage system cannot satisfy the AGC directive.

[0008] Preferably, the formula for calculating the heat storage coefficient C of the initial thermal power generation unit in step 3 is as shown in formula (1):

number

[0009] Preferably, in step 3, the predicted heat storage dataset of the thermal power generation unit includes an input set {X i} and an output set {Y i}, and the input set {X i} is the unit parameters for 300 seconds before time t0 and the steam drum pressure and main steam flow rate of the unit at time t1. The unit parameters include load, main steam flow rate, feed water flow rate, main steam pressure set value, actual main steam pressure value, total air volume, fuel quantity, main steam temperature, furnace negative pressure, comprehensive valve position, load change rate, and steam drum pressure. Based on different times, i input samples are constructed to form the input set {X i}, The output set {Y i} is the heat storage coefficient of the thermal power generation unit at time t0. Based on different times, i output samples are constructed to form the output set {Y i}.

[0010] Preferably, the heat storage coefficient prediction model of the thermal power generation unit includes an input layer, a convolutional neural network layer, a bidirectional gated regression unit layer, a multi-head self-attention mechanism layer and an output layer. The convolutional neural network layer includes two convolutional layers and one pooling layer. The bidirectional gated regression unit layer includes two GRU networks. The multi-head self-attention mechanism layer and the output layer are connected via a fully connected layer. The unit load predictive control method based on the battery energy storage state according to claim 5, characterized in that.

[0011] Preferably, in step 4, based on the predicted heat storage dataset of the thermal power generation unit in step 3, the heat storage coefficient prediction model of the thermal power generation unit is trained, and specifically it is implemented according to the following steps: Step 4.1: Input the input set {X i} into the input layer. The vector of input features is 13×16, Step 4.2: Extract the feature vector of the data set by the convolutional layer, and reduce the extracted feature vector to one-dimensional data by the pooling layer, Step 4.3: The bidirectional gated recurrent unit layer performs bidirectional training on the one-dimensional data output in Step 4.2, learns deeper-level time series features between the unit operation data and the heat storage coefficient of the thermal power unit, and then uses all the one-dimensional data output after its training as the input data of the multi-head self-attention mechanism layer, Step 4.4: The multi-head self-attention mechanism layer assigns different weight values to the hidden state h t output from the bidirectional gated recurrent unit layer, [[ID=​​​​​​​​​​​​​​​​​​​​​​​​​​t is the hidden state of the GRU at time step t, and β t These are the weights of different time steps t, and a t This involves calculating the correlation score between the hidden state and the predicted vector at time step t, and a i This is the normalization factor for the time step t, Step 4.5 is specifically as shown in equation (4):

number

[0013] Preferably, the correspondence between the predicted thermal energy storage coefficient of the thermal power generation unit and the preceding load command is as follows: If the predicted thermal energy storage coefficient of the thermal power generation unit is -1, the preceding load command is -10; if the predicted thermal energy storage coefficient of the thermal power generation unit is -0.6, the preceding load command is -5; if the predicted thermal energy storage coefficient of the thermal power generation unit is -0.4, the preceding load command is -2, and if the predicted thermal energy storage coefficient of the thermal power generation unit is -0.1, The preceding load command is -0 and the heat storage coefficient is 0; the preceding load command is 0 and the predicted heat storage coefficient of the thermal power generation unit is 0.1; the preceding load command is 0 and the predicted heat storage coefficient of the thermal power generation unit is 0.4; the preceding load command is 2 and the predicted heat storage coefficient of the thermal power generation unit is 0.6; the preceding load command is 5 and the predicted heat storage coefficient of the thermal power generation unit is 1; the preceding load command is 10.

[0014] Preferably, in step 6, a load command to be executed by the thermal power generation unit is generated based on the state of the thermal power generation unit and the energy storage charge state of the energy storage system, and based on the correspondence between the predicted heat storage coefficient of the thermal power generation unit and the preceding load command, the preceding load command is determined based on the predicted heat storage coefficient of the thermal power generation unit, and the load command to be executed by the thermal power generation unit is dynamically adjusted based on the preceding load command. Specifically, If the thermal power generation unit is in a load-up state and the energy storage charge state of the energy storage system is less than the minimum output limit, a load command to be executed by the thermal power generation unit is generated, and the preceding amount of the load command corresponding to the thermal power generation unit's heat storage coefficient is increased. If the thermal power generation unit is in a load-down state and the energy storage charge state of the energy storage system is greater than the energy storage charge limit, the preceding amount of the load command corresponding to the thermal power generation unit's heat storage coefficient is decreased in the load command to be executed by the thermal power generation unit. [Effects of the Invention]

[0015] The present invention's unit load pre-control method based on the battery energy storage state reads relevant parameters of the energy storage battery and the thermal power generation unit in real time to determine the load increase, load decrease, and stable state of the thermal power generation unit based on the unit AGC command, and determines whether the load fluctuation requirements of the thermal power generation unit can be met based on the energy storage charge state (SOC) and output parameters of the energy storage battery. If the energy storage battery cannot meet the requirements, the heat storage coefficient of the thermal power generation unit is predicted in combination with the real-time operating status of the thermal power generation unit, and the amount of pre-control of the unit's load setting is automatically adjusted. This method makes full use of the heat stored in the thermal power generation unit while avoiding overcharging or over-discharging of the energy storage system, that is, it extends the service life of the energy storage system and improves the load adjustment performance index of the unit by comprehensively considering the energy storage state and unit status and utilizing the heat stored in the thermal power generation unit according to the operating status. [Brief explanation of the drawing]

[0016] [Figure 1] This is a flowchart of the unit load pre-control method based on the battery energy storage state of the present invention. [Figure 2] This is a schematic diagram of the structure of the thermal energy storage coefficient prediction model for a thermal power generation unit in the unit load pre-control method based on the battery energy storage state of the present invention. [Figure 3] This is a schematic diagram of the load change in response to the AGC command by the unit in Embodiment 3 of the present invention. [Modes for carrying out the invention]

[0017] To enable those skilled in the art to better understand the methods of the present invention, the technical methods in the embodiments of the present invention are described below clearly and completely with reference to the drawings of the embodiments. Clearly, the embodiments described are some, but not all, embodiments of the present invention. All other embodiments obtained based on the embodiments of the present invention without the creative effort of those skilled in the art should all fall within the scope of protection of the present invention.

[0018] The present invention will be described in detail below with reference to the drawings and specific embodiments.

[0019] (Example 1) The unit load pre-control method based on the battery energy storage state of the present invention is specifically carried out according to the following steps, as shown in Figure 1: Step 1: Determine the status of the thermal power generation unit based on the unit's Automatic Generating Control (AGC) command and the current load of the thermal power generation unit. In step 1, the state of the thermal power generation unit includes load increase, load decrease, and stable state, where load increase is when the AGC command is greater than the current load of the thermal power generation unit, load decrease is when the AGC command is less than the current load of the thermal power generation unit, and stable state is when the AGC command is equal to the current load of the thermal power generation unit.

[0020] Step 2: Determine whether the energy storage output of the energy storage system meets the AGC directive. When the AGC (Automatic Gain Control) command for a thermal power generation unit changes, the energy storage system adjusts its output in conjunction with the thermal power generation unit. Considering the service life of the energy storage system, the State of Charge (SOC) is generally maintained at 0.2 to 0.9, which is the adjustment range for the energy storage system. If the State of Control (SOC) of an energy storage system is not within the adjustment range, the energy storage system will not respond to the AGC directive. If the AGC directive is greater than the dead zone for energy storage adjustment and the SOC is within the acceptable range, the energy storage will respond to the AGC directive at its rated maximum power. The dead zone refers to a set range within which the energy storage system does not perform adjustment operations when changes in energy storage adjustment are within a certain range.

[0021] Step 3: If the State of Charge (SOC) is not within the adjustment range, obtain the initial thermal energy storage coefficient of the thermal energy storage unit based on the state of the thermal energy storage unit, and then build a predicted thermal energy storage dataset for the thermal energy storage unit. If the State of Charge (SOC) is outside the adjustment range, the time it takes for the thermal power generation unit to reach the AGC command will be longer. To fully utilize the output of the thermal power generation unit without reducing its response capability, the unit's lead load is calculated.

[0022] The leading load of a unit depends primarily on the heat storage status of the thermal power generation unit at the time the AGC command changes. When the thermal power generation unit has sufficient heat storage, it can increase the leading load, respond to the AGC command more quickly and accurately, and better meet the adjustment needs of the power grid. When the thermal power generation unit has insufficient heat storage, the short-term actual output of the thermal power generation unit is much smaller than that of a thermal power generation unit under normal conditions, and the output of the energy storage system is relatively small, making it impossible to quickly compensate for the shortfall in the actual output of the thermal power generation unit. This leads to a power shortage in the thermal power energy storage system and prevents the overall AGC indicator from being met.

[0023] The heat storage coefficient of an early thermal power generation unit is defined as the ratio of the change in steam flow rate stored in the water-cooled wall, steam drum, and superheater to the rate of pressure change in the steam drum, and the ratio C is given by equation (1):

number

[0024] Furthermore, the heat storage coefficient C0 of the thermal power generation unit when the unit AGC command t0 changes is determined by the combustion conditions of the thermal power generation unit, and the heat storage coefficient can be predicted based on the operating data of the thermal power generation unit due to the inertia of boiler combustion.

[0025] In step 3, the thermal storage dataset for the predicted thermal power generation unit is the input set {X i} and output set {Y i} includes the input set {X i{X} represents the thermal power generation unit parameters for 300 seconds prior to time t0 and the steam drum pressure and main steam flow rate of the thermal power generation unit at time t1, wherein the unit parameters include load, main steam flow rate, feedwater flow rate, main steam pressure setpoint, actual main steam pressure, total airflow, fuel amount, main steam temperature, furnace negative pressure, general valve position, load change rate, and steam drum pressure, and i input samples are configured based on different times to form the input set {X}. i} form, The aforementioned output set {Y i {Y} is the heat storage coefficient of the thermal power generation unit at time t0, and i output samples are constructed based on different times to form an output set {Y}. i It forms a}.

[0026] Step 4: Construct a thermal power generation unit heat storage coefficient prediction model and train the thermal power generation unit heat storage coefficient prediction model based on the thermal power generation unit heat storage dataset predicted in Step 3.

[0027] Step 5: Input the operating data of the thermal power generation unit into the thermal power generation unit heat storage coefficient prediction model trained in Step 4 to generate the predicted heat storage coefficient of the thermal power generation unit, and establish a correspondence between the predicted heat storage coefficient of the thermal power generation unit and the preceding load command. The specific correspondence is as follows: [Table 1]

[0028] Step 6: Based on the state of the thermal power generation unit and the energy storage charge state of the energy storage system, generate a load command to be executed by the thermal power generation unit, and, in the correspondence between the predicted thermal power generation unit's heat storage coefficient and the preceding load command, determine the corresponding preceding load command based on the predicted thermal power generation unit's heat storage coefficient, and dynamically adjust the load command to be executed by the thermal power generation unit based on the preceding load command. Specifically, when a thermal power generation unit is in a load-up state and the energy storage charge state of the energy storage system is less than the minimum output limit, the load command executed by the thermal power generation unit increases the preceding amount of the load command corresponding to the thermal power generation unit's heat storage coefficient. When a thermal power generation unit is in a load-down state and the energy storage charge state of the energy storage system is greater than the energy storage charge limit, the load command executed by the thermal power generation unit decreases the preceding amount of the load command corresponding to the thermal power generation unit's heat storage coefficient.

[0029] (Example 2) In addition to Example 1, as shown in Figure 2, the thermal energy storage coefficient prediction model for a thermal power generation unit in the unit load pre-control method based on the battery energy storage state of the present invention includes a convolutional neural network (CNN) layer, a bidirectional gated recurrent unit (BiGRU) layer, a multihead self-attention (MSA) layer, and an output layer.

[0030] The convolutional neural network layer includes two convolutional layers and one pooling layer, and the multi-head self-awareness mechanism layer and the output layer are connected via a fully connected layer.

[0031] In Step 4, a thermal power generation unit heat storage coefficient prediction model is trained based on the thermal power generation unit heat storage dataset from Step 3, and this is specifically carried out according to the following steps: Step 4.1: Input Set {X i The input layer is}, and the input feature vector is 13 × 16. Step 4.2: Extract feature vectors from the dataset using a convolutional layer, and reduce the extracted feature vectors to one-dimensional data using a pooling layer. Step 4.3: The bidirectional gated regression unit layer consists of two GRU networks, one processing time-series data from front to back and the other processing time-series data from back to front, performing bidirectional training on the one-dimensional data output in Step 4.2 to learn deeper level time-series features between the unit operation data and the heat storage coefficient of the thermal power generation unit, and then using the one-dimensional data output after all the training as input data for the multi-head self-aware mechanism layer. Step 4.4: The multi-head self-aware mechanism layer outputs the hidden state h from the bidirectional gated regressive unit layer. t By assigning different weight values ​​to them, specifically, In step 1, for the hidden state ht at each time step, the function f(q,h t Using the correlation score a t Calculate, In step 2, a of all time steps t The normalized weights β at different time steps are obtained by applying softmax normalization to the data. t Obtained, In step 3, β t Using as a weight, all hidden states h t A weighted sum is performed on the hidden state h. t Assign different weight values ​​to them, Specifically, this is shown in equations (2) and (3):

number

number

[0032] Step 4.5: The output layer has hidden states h assigned different weight values ​​in Step 4.4. t After mapping Tanh as the activation function, the predicted heat storage coefficient of the thermal power generation unit is output, and the predicted heat storage coefficient of the thermal power generation unit is mapped within the (-1,1) interval to obtain the predicted heat storage coefficient of the thermal power generation unit, as shown specifically in equation (4):

number

[0033] Step 4.6: Output set {Y} of the predicted thermal power generation unit's heat storage coefficient i The method for pre-controlling unit load based on battery energy storage state according to claim 6, characterized in that steps 4.1 to 4.5 are repeated and multiple iterative training is performed until the thermal power generation unit's heat storage coefficient prediction model converges, by comparing it with}.

[0034] (Example 3) The unit load pre-control method based on the battery energy storage state in this embodiment is carried out according to the following steps: Step 1: Determine the status of the thermal power generation unit based on the AGC command and the current load of the thermal power generation unit. Based on the unit's Automatic Generating Control (AGC) command and the current load of the thermal power generation unit, the unit is determined to be in a state of increased load, decreased load, or stable state. If the AGC command is greater than the current load of the thermal power generation unit, it is an increased load; if the AGC command is less than the current load of the thermal power generation unit, it is a decreased load; and if the AGC command is equal to the current load of the thermal power generation unit, it is a stable state.

[0035] Step 2: Analyze whether the energy storage output of the thermal power generation unit and the energy storage system meets the AGC directive. As shown in Figure 3, when the unit AGC command changes at time t0, the AGC command forms a load command via the set speed limit, and the unit performs load fluctuations based on the load command. If the energy storage SOC is within the acceptable range, the energy storage P e The output is, and the load on the thermal power generation unit at this time is the energy storage load P1 of the superimposed energy storage system.

[0036] Step 3: Establish the heat storage coefficient of the thermal power generation unit and the predicted heat storage dataset for the thermal power generation unit.

[0037] As shown in Figure 3, when the unit responds to the AGC command, the heat storage is maintained for a short period; that is, the unit receives the change in the AGC command at time t0, the heat storage of the thermal power generation unit is consumed at time t1, and the unit depends on the energy response load of the fuel system. The heat storage coefficient C0 when the unit receives the AGC command change t0 is given by equation (5):

number

[0038] Step 4: Construct a thermal power generation unit heat storage coefficient prediction model and train the thermal power generation unit heat storage coefficient prediction model based on the thermal power generation unit heat storage dataset predicted in Step 3.

[0039] Step 5: Input the data from the unit operation into the thermal power generation unit heat storage coefficient prediction model trained in Step 4 to generate the predicted heat storage coefficient of the thermal power generation unit, and establish a correspondence between the predicted heat storage coefficient of the thermal power generation unit and the preceding load command.

[0040] Step 6: Based on the state of the thermal power generation unit and the energy storage charge state of the energy storage system, a load command to be executed by the thermal power generation unit is generated, and based on the correspondence between the predicted thermal power generation unit's heat storage coefficient and the preceding load command, the corresponding preceding load command is determined based on the predicted thermal power generation unit's heat storage coefficient, and the load command to be executed by the thermal power generation unit is dynamically adjusted based on the preceding load command.

[0041] Specifically, when a thermal power generation unit is in a load-up state and the energy storage system's State of Charge (SOC) is less than the minimum output limit, the load command for the thermal power generation unit increases the calculated lead amount of the load command, and as shown in Figure 3, the superimposed lead amount of the thermal power generation unit allows the thermal power generation unit to respond more quickly and accurately to AGC load requests. When a thermal power generation unit is in a load-down state and the energy storage system's SOC is greater than the energy storage charge limit, the thermal power generation unit responds more quickly to AGC commands by reducing the corresponding load based on the calculated lead amount of the load command.

[0042] It should be understood that, as used herein and in the appended claims, the terms “include” and “inclusive” indicate the presence of a described feature, whole, step, operation, element and / or assembly, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, assemblies and / or sets thereof.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art. The terms used herein are for the sole purpose of describing specific embodiments and are not intended to limit the invention. The terms “and / or” as used herein include any and all combinations of one or more related enumerated items.

[0044] The above description is merely a preferred embodiment of the present invention and does not impose any formal limitations on the present invention. Those skilled in the art can successfully implement the present invention as shown in the drawings of the specification and described above. However, those skilled in the art will also know that any equivalent changes, modifications, and advancements made using the above-mentioned technical content without departing from the scope of the proposed technical invention are all equivalent embodiments of the present invention. At the same time, any equivalent changes, modifications, and advancements made to the above embodiments based on the substantial art of the present invention still fall within the scope of protection of the proposed technical invention.

Claims

1. A method for pre-controlling unit load based on the battery energy storage state, which is specifically carried out according to the following steps: Step 1: Determine the status of the thermal power generation unit based on the AGC directive and the current load of the thermal power generation unit. Step 2: Determine whether the energy storage output of the energy storage system meets the AGC directive. If the energy storage output does not meet the AGC directive, perform steps 3 through 6. If the energy storage output meets the AGC directive, do not perform these steps. Step 3: Based on the state of the thermal power generation unit, obtain the initial heat storage coefficient of the thermal power generation unit, and further construct a predicted heat storage dataset for the thermal power generation unit. Step 4: Construct a thermal energy storage coefficient prediction model for thermal power generation units and train the thermal energy storage coefficient prediction model for thermal power generation units based on the thermal energy storage dataset of thermal power generation units predicted in Step 3. Step 5: Input the operating data of the thermal power generation unit into the thermal power generation unit heat storage coefficient prediction model trained in Step 4 to generate the predicted heat storage coefficient of the thermal power generation unit, and establish a correspondence between the predicted heat storage coefficient of the thermal power generation unit and the preceding load command. Step 6: A unit load lead control method based on battery energy storage state, which generates a load command to be executed by a thermal power generation unit based on the state of the thermal power generation unit and the energy storage charge state of the energy storage system, determines the corresponding lead amount of the load command based on the predicted heat storage coefficient of the thermal power generation unit in the correspondence between the predicted heat storage coefficient of the thermal power generation unit and the lead amount of the load command, and dynamically adjusts the load command to be executed by the thermal power generation unit based on the lead amount of the load command.

2. The unit load pre-control method based on battery energy storage state according to claim 1, characterized in that the state of the thermal power generation unit in step 1 includes load increase, load decrease, and stable state, wherein the load increase is when the AGC command is greater than the current load of the thermal power generation unit, the load decrease is when the AGC command is less than the current load of the thermal power generation unit, and the stable state is when the AGC command is equal to the current load of the thermal power generation unit.

3. The unit load pre-control method based on battery energy storage state according to claim 1, characterized in that in step 2, determining whether the energy storage output of the energy storage system satisfies the AGC command means that, specifically, if the energy storage charge state of the energy storage system is not within the adjustment range, the energy storage output of the energy storage system cannot satisfy the AGC command.

4. The formula for calculating the heat storage coefficient C of the initial thermal power generation unit in step 3 is as follows: [Math 1] （１）、 The unit load pre-control method based on the battery energy storage state according to claim 1, characterized in that, in the formula, Δm represents the main steam flow rate, p is the steam drum pressure, and t is time.

5. In step 3, the heat storage dataset of the predicted thermal power generation unit is the input set {X ｉ } and output set {Y ｉ } includes, The aforementioned input set {X ｉ } is time t ０ The unit parameters and time t for the previous 300 seconds １ The unit parameters are the steam drum pressure and main steam flow rate at time, and the unit parameters include load, main steam flow rate, feedwater flow rate, main steam pressure setpoint, actual main steam pressure, total airflow, fuel amount, main steam temperature, furnace negative pressure, general valve position, load change rate, and steam drum pressure, and i input samples are constructed based on different times to form an input set {X ｉ } form, The output set {Y ｉ} is the heat storage coefficient of the thermal power generation unit at time t ０ , and constructs i output samples based on different times to form the output set {Y ｉ}, and the unit load leading control method based on the battery energy storage state according to claim 1 is characterized by this.

6. The thermal power generation unit's heat storage coefficient prediction model includes an input layer, a convolutional neural network layer, a bidirectional gated regression unit layer, a multi-head self-aware mechanism layer, and an output layer. The unit load pre-control method based on battery energy storage state according to claim 5, characterized in that the convolutional neural network layer includes two convolutional layers and one pooling layer, the bidirectional gated recurrent unit layer includes two GRU networks, and the multi-head self-awareness mechanism layer and the output layer are connected via a fully connected layer.

7. In step 4, a thermal power generation unit heat storage coefficient prediction model is trained based on the predicted thermal power generation unit heat storage dataset from step 3, and is specifically carried out according to the following steps: Step 4.1: Input Set { X ｉ The input layer is then populated with}, and the input feature vector is 13 × 16. Step 4.2: Extract the feature vectors of the dataset using a convolutional layer, and reduce the extracted feature vectors to one-dimensional data using a pooling layer. Step 4.3: The bidirectional gated regression unit layer performs bidirectional training on the one-dimensional data output in Step 4.2 to learn deeper level time-series features between the operating data of the thermal power generation unit and the heat storage coefficient of the thermal power generation unit, and then uses the one-dimensional data output after all training as input data for the multi-head self-aware mechanism layer. Step 4.4: The multi-head self-aware mechanism layer outputs the hidden state h from the bidirectional gated regression unit layer. ｔ Assign different weight values ​​to them, Step 4.5: The output layer has hidden states h assigned different weight values ​​in Step 4.

4. ｔ After mapping Tanh as the activation function, the predicted heat storage coefficient of the thermal power generation unit is output, the predicted heat storage coefficient of the thermal power generation unit is mapped within the interval (-1, 1), and the predicted heat storage coefficient of the thermal power generation unit is obtained. Step 4.6: Set the predicted thermal energy storage coefficient of the thermal power generation unit to output {Y ｉ The method for controlling a unit load in advance based on the battery energy storage state according to claim 6, characterized in that steps 4.1 to 4.5 are repeated and multiple iterative training is performed until the thermal power generation unit's heat storage coefficient prediction model converges, compared with the above.

8. In step 4.4, the hidden state h ｔ Different weight values ​​are assigned to them, as shown specifically in equations (2) and (3): [Math 2] （２）、 [Math 3] （３）、 In the equation, f is the GRU network, q is the prediction vector of the GRU network, and h ｔ is the GRU hidden state at time step t, and β ｔ These are the weights of different time steps t, and a ｔ This involves calculating the correlation score between the hidden state and the predicted vector at time step t, and a ｉ This is the normalization factor for the time step t, Step 4.5 is specifically as shown in equation (4): [Math 4] （４）、 The unit load pre-control method based on the battery energy storage state according to claim 7, characterized in that, in the formula, x is the input value and e is the base of the natural logarithm.

9. The correspondence between the predicted thermal power generation unit's heat storage coefficient and the preceding load command is as follows: If the predicted thermal power generation unit's heat storage coefficient is -1, and the preceding load command is -10, If the predicted thermal power generation unit's heat storage coefficient is -0.6, and the preceding load command is -5, If the predicted thermal power generation unit's heat storage coefficient is -0.4, and the preceding load command is -2, If the predicted thermal power generation unit's heat storage coefficient is -0.1, and the preceding load command is -0, If the predicted thermal power generation unit's heat storage coefficient is 0, the preceding load command is 0, If the predicted thermal power generation unit's heat storage coefficient is 0.1, and the preceding load command is 0, If the predicted heat storage coefficient of the thermal power generation unit is 0.4, and the preceding load command is 2, If the predicted heat storage coefficient of the thermal power generation unit is 0.6, and the preceding load command is 5, The unit load advance control method based on battery energy storage state according to claim 1, characterized in that when the predicted heat storage coefficient of the thermal power generation unit is 1, the advance amount of the load command is 10.

10. In step 6, generating a load command to be executed by the thermal power generation unit based on the state of the thermal power generation unit and the energy storage charge state of the energy storage system, and determining the corresponding load command's preceding amount based on the predicted thermal power generation unit's thermal storage coefficient in the correspondence between the predicted thermal power generation unit's thermal storage coefficient and the preceding amount of the load command, and dynamically adjusting the load command to be executed by the thermal power generation unit based on the preceding amount of the load command, specifically involves: The unit load advance control method based on the battery energy storage state according to claim 1, characterized in that, when the thermal power generation unit is in a load increase state and the energy storage charge state of the energy storage system is less than the minimum output limit, a load command to be executed by the thermal power generation unit is generated and the advance amount of the load command corresponding to the heat storage coefficient of the thermal power generation unit is increased, and when the thermal power generation unit is in a load decrease state and the energy storage charge state of the energy storage system is greater than the energy storage charge limit, a load command to be executed by the thermal power generation unit is generated and the advance amount of the load command corresponding to the heat storage coefficient of the thermal power generation unit is decreased.

Images