Active regulation method and system for interaction between coal-fired generating unit and power grid load

Abstract

The embodiment of the present application relates to the technical field of coal-fired units, in particular to a kind of active regulation and control method and system of coal-fired unit and grid load interaction.The method comprises: if the fluctuation amplitude of grid load power is greater than the preset amplitude threshold by the load power prediction model of grid, then according to the fluctuation amplitude, the predicted time point of fluctuation amplitude, the maximum climbing rate of coal-fired unit, the upward reserve capacity and the downward reserve capacity to determine the target load power and load adjustment time of coal-fired unit, the target load power and load adjustment time are sent to coal-fired unit;Coal-fired unit determines the electric companion heating power of thick-walled component according to target load power and load adjustment time;According to the electric companion heating power and the preset electric companion heating algorithm, the outer wall of thick-walled component is heated to reduce the inner and outer wall temperature difference of thick-walled component to cope with fluctuation amplitude.The technical scheme of the present application can improve the regulation and control capability of coal-fired unit and grid when coping with load fluctuation.

Description

Technical Field

[0001] This invention relates to the field of coal-fired power unit technology, and in particular to an active control method and system for interaction between coal-fired power units and power grid load. Background Technology

[0002] In the power system, the proportion of new energy sources such as wind power and photovoltaics continues to increase. However, the output of these new energy sources is intermittent and highly volatile, which may affect the stable operation of the power grid. Currently, when the grid load fluctuates, the interaction between the power grid and coal-fired power units is mainly achieved through self-generation control and regulation. This regulation method passively responds to meet the grid's frequency and voltage requirements under power load fluctuations. However, when the grid load power fluctuates significantly and rapidly, the heat transfer rate of thick-walled components in coal-fired power units is much slower than the speed of the grid's response to the load demand signal. This results in a lag in the response of the coal-fired power units. Furthermore, the temperature of the inner wall of the thick-walled components rises rapidly, while the outer wall, due to lagging heat conduction, forms a large temperature difference between the inner and outer walls. Over time, this can shorten the service life of the equipment and create safety hazards.

[0003] Therefore, there is an urgent need for an active control method and system for the interaction between coal-fired power units and grid loads to solve the above-mentioned technical problems. Summary of the Invention

[0004] This invention provides an active control method and system for interaction between coal-fired power units and power grid load, which can improve the interactive control capability of coal-fired power units and power grid when facing load fluctuations.

[0005] In a first aspect, the present invention provides an active control method for the interaction between a coal-fired power unit and a power grid load, applied to an active control system for the interaction between a coal-fired power unit and a power grid load, the system comprising a power grid and a coal-fired power unit, including:

[0006] If the power grid predicts through a trained load power prediction model that the fluctuation range of the power grid load power will be greater than a preset amplitude threshold, then the target load power and load adjustment time of the coal-fired unit are determined based on the fluctuation range, the prediction time point that generates the fluctuation range, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity, and the target load power and load adjustment time are sent to the coal-fired unit.

[0007] The electric heating power of the thick-walled components of the coal-fired power unit is determined by a feedforward-feedback composite control algorithm based on the target load power and load adjustment time.

[0008] The coal-fired power unit heats the outer wall of the thick-walled component 1-2 hours in advance according to the electric tracing heating power and the preset electric tracing heating algorithm, so as to cope with the fluctuation range by reducing the temperature difference between the inner and outer walls of the thick-walled component.

[0009] Secondly, the present invention provides an active control system for interaction between a coal-fired power unit and a power grid load, the system comprising a power grid and a coal-fired power unit;

[0010] The power grid is used to predict, through a trained load power prediction model, that the fluctuation range of the power grid load power will be greater than a preset range threshold. Then, based on the fluctuation range, the prediction time point that generates the fluctuation range, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity, the target load power and load adjustment time of the coal-fired unit are determined, and the target load power and load adjustment time are sent to the coal-fired unit.

[0011] The coal-fired power unit is used to determine the electric heating power of the thick-walled components of the coal-fired power unit based on the target load power and load adjustment time, using a feedforward-feedback composite control algorithm; and

[0012] The outer wall of the thick-walled component is heated 1-2 hours in advance according to the electric heating power and the preset electric heating algorithm, so as to cope with the fluctuation range by reducing the temperature difference between the inner and outer walls of the thick-walled component.

[0013] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the method described in the first aspect of the present invention.

[0014] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the method described in the first aspect of the present invention.

[0015] This invention provides an active control method and system for the interaction between a coal-fired power unit and the power grid load. By predicting the future moment when the power grid load will fluctuate significantly through a load power prediction model, and by heating the thick-walled components of the coal-fired power unit 1 to 2 hours in advance, the temperature difference between the inner and outer walls of the thick-walled components of the boiler during the load change process is effectively reduced, thereby improving the response speed of the coal-fired power unit, reducing fatigue life loss during the load change process, and improving the safety of the unit during startup. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of an active control method for interaction between a coal-fired power unit and the power grid load, provided by an embodiment of the present invention;

[0018] Figure 2 This is a hardware architecture diagram of an electronic device provided in an embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram illustrating the principle of power system power generation and supply balance in this invention;

[0020] Figure 4 This is a schematic diagram showing the hybrid operation mode of the inter-storage type and direct-blown pulverizing system of the present invention and the timing relationship of electric heat tracing control;

[0021] Figure 5 This is a schematic diagram of the active control method for thick-walled components of a coal-fired power unit according to the present invention;

[0022] Figure 6 This is a schematic diagram of a new model for deep interaction between the power grid and coal-fired power units according to the present invention;

[0023] Figure 7 This is a schematic diagram comparing the old and new modes of deep interaction between the power grid and coal-fired power units in this invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0025] In current new power systems, the principle of balancing power generation and supply is as follows: Figure 3 As shown, the balance between consumer load (factories, etc.) and unadjustable power sources such as renewable energy (wind power, solar power, etc.) is represented by the net load of the grid. This net load is balanced with adjustable power sources in the grid. Therefore, the trends between adjustable power sources and net load are opposite. That is, when the net load is greater than zero, the amount of adjustable power generated decreases, and the adjustable power load becomes negative; conversely, when the net load is less than zero, the amount of adjustable power generated increases, and the adjustable power load becomes positive. The ultimate effect is to keep the frequency of grid nodes within the allowable range. This real-time control method requires adjustable power sources to respond immediately to the grid's demand for increased or decreased adjustable power load.

[0026] In existing regulation methods, adjustable power sources passively respond to meet the grid's demands for frequency and voltage. Power plants primarily using coal-fired units have limited rapid start-up and load-change capabilities due to the inherent flexibility of these units. This is mainly reflected in the fact that the pulverizing system of coal-fired units cannot quickly respond to the grid's increased load demands, as direct-fired pulverizing systems require a delay of approximately 10-15 minutes. Furthermore, due to the use of thick-walled components, the heat transfer rate of coal-fired units is much slower than the speed of the electrical signal indicating the grid's load demand. Rapid changes can cause significant thermal stress, leading to damage to these thick-walled components. To address the limitation of the response speed of direct-fired pulverizing systems, adopting an intermediate storage pulverizing system can significantly reduce the response speed requirements of coal-fired unit pulverizing systems for rapid load changes. Specifically, for example... Figure 4 As shown, by employing a hybrid approach of intermediate storage and direct-fired pulverizing in the combustion-side pulverizing system, pulverized coal from the intermediate storage is rapidly introduced to meet the corresponding rapid load changes. Once the direct-fired system has achieved its intended effect, the coal supply from the intermediate storage pulverizing system is gradually reduced, achieving rapid matching between fuel and load. Coupled with pre-emptive thermal stress control of thick-walled components, current coal-fired units can respond to grid demands more quickly, thereby supporting the safe and stable operation of the new power system.

[0027] In existing technologies, coal-fired power units employ a passive response approach to cope with large and rapid fluctuations in grid load, namely the AGC (Automatic Generation Control) mode. This mode, constrained by thick-walled components, suffers from response lag and impacts the lifespan of these components. Even after flexibility modifications, the response rate of coal-fired units currently does not exceed 1.5%~2% Pe / min. Given the lack of breakthroughs in the low-cost, large-scale application of new energy storage technologies, the grid's adjustable resources are scarce, severely limiting the power system's ability to accommodate a higher proportion of renewable energy. Currently, to meet the demands of higher grid-coal-fired unit interaction capabilities, energy storage technology is used to overcome the mismatch between the slow time-varying nature of coal-fired units and the rapid time-varying nature of grid demand. However, to achieve a load-changing capacity of 4~5% Pe / min for coal-fired units, the investment in energy storage technology is approximately 500-1000 yuan / kW, significantly increasing the cost of coal-fired units.

[0028] To address the aforementioned deficiencies in existing technologies, this invention discloses an active control method and system for interaction between coal-fired power units and grid load. This method aims to meet the deep interaction needs between the grid and coal-fired power units, enabling rapid load changes and rapid start-up and shutdown of the units. This allows for a new interaction mode between the coal-fired power units and the grid, 1-2 hours in advance. It provides a novel interactive response method between coal-fired power units and the grid, creating a new ancillary service model for the existing electricity market, and thus contributing to the development of new interactive technologies between renewable energy sources (wind and solar) and coal-fired power units in the power system. Figure 5 As shown, by preheating with electric heat tracing 1-2 hours in advance, the high inner wall temperature and low outer wall temperature of thick-walled components caused by passive rapid heating during the start-up phase can be transformed into a situation where the outer wall temperature is high and the inner wall temperature is low by using external electric heat tracing technology in advance. By preheating, the magnitude of thermal stress is changed from negative to positive, thereby reducing the stress amplitude during the start-up phase and achieving effective control of the start-up stress amplitude.

[0029] Please refer to Figure 1 This invention provides an active control method for the interaction between a coal-fired power unit and a power grid load. The method is applied to an active control system for the interaction between a coal-fired power unit and a power grid load. The system includes a power grid and the coal-fired power unit. The method includes:

[0030] Step 100: If the power grid predicts through the trained load power prediction model that the fluctuation range of the power grid load will be greater than the preset amplitude threshold, then the target load power and load adjustment time of the coal-fired unit are determined based on the fluctuation range, the prediction time point of the fluctuation range, the maximum ramp rate of the coal-fired unit, the upward reserve capacity and the downward reserve capacity, and the target load power and load adjustment time are sent to the coal-fired unit.

[0031] Step 102: The electric heating power of the thick-walled components of the coal-fired power unit is determined by the feedforward-feedback composite control algorithm based on the target load power and load adjustment time.

[0032] Step 104: The coal-fired unit preheats the outer wall of the thick-walled component according to the electric tracing heating power and the preset electric tracing heating algorithm, so as to cope with the fluctuation range by reducing the temperature difference between the inner and outer walls of the thick-walled component.

[0033] In this embodiment of the invention, a regulation strategy for coping with power grid load fluctuations is provided, namely a new mode of deep interaction between the power grid and coal-fired power units, as shown in the schematic diagram. Figure 6 As shown, a load power prediction model is used to predict grid load fluctuations 1-2 hours in advance. If a significant load fluctuation (the fluctuation amplitude exceeding a preset threshold) is predicted after a preset time period, the target load power required for the coal-fired power unit and the required load adjustment time are determined based on the future fluctuation amplitude and the predicted time. During the load adjustment time, the outer wall of the thick-walled components of the coal-fired power unit is heated to prevent large temperature differences during the ramp-up process. The electric tracing heating power during the heating process is calculated based on the target load power and the load adjustment time. This method can meet the deep interaction requirements of the power grid and the coal-fired power unit, enabling rapid load changes and start-up / shutdown of the coal-fired power unit, thereby facilitating early interaction and control between the coal-fired power unit and the power grid. The response rate of this new mode of deep interaction between the power grid and the coal-fired power unit, as demonstrated by this invention, is comparable to that of the old mode. Figure 7 As shown, through Figure 7 It can be seen that the new model can absorb grid fluctuations more quickly and greatly improve the response rate of coal-fired units to load fluctuations.

[0034] Understandably, in coal-fired power units, the boiler and turbine systems require heating of thick-walled components for active regulation. Thick-walled components refer to pressure-bearing or load-bearing members in coal-fired power units that require increased wall thickness to meet strength, rigidity, and fatigue resistance requirements due to high-temperature and high-pressure operating conditions. These components maintain the safe and stable operation of the unit and prevent phenomena such as tube rupture and vessel breakage. They include the boiler drum, superheater headers, reheater, high-temperature economizer, and the turbine high-pressure cylinder, regulating valve body, and main steam pipes, etc., with wall thicknesses generally greater than 20mm and exhibiting significant thermal inertia.

[0035] Temperature sensors on the outer side of each thick-walled component of the boiler measure the temperature in real time, and a temperature transmitter converts the temperature signal into an output current. Two identical temperature sensors should be present at each location, and the higher temperature signal is used. After the coal-fired unit receives the ramp-up task, it calculates the target heating temperature for the thick-walled components. Before boiler ignition or load change, the thick-walled components are heated a certain period in advance based on the difference between the measured temperature and the set value. By using electric heating elements installed on the outside of the thick-walled components to increase their wall temperature, sufficient wall temperature can be ensured, effectively reducing the stress amplitude during startup and ensuring startup safety. Additionally, after confirming the load change is complete, the pulverizing system manually switches from intermediate storage to direct-fired system. Temperature sensors on the outer wall of the turbine cylinder body measure the temperature in real time; two identical temperature sensors are installed at each location, and the higher temperature signal is used. Before the turbine receives steam or changes load, and after shutdown, to ensure the conditions for the next hot start-up, heating is performed for a period of time based on four parameters: the temperature difference between the inner walls of the upper and lower high-pressure cylinders, the temperature difference between the inner and outer walls of the upper high-pressure cylinder, the temperature difference between the inner and outer walls of the lower high-pressure cylinder, and the expansion difference of the high-pressure cylinder. Electric heating elements installed on the outer wall of the cylinder increase the wall temperature of thick-walled components. This method ensures sufficient wall temperature on the outer wall of the cylinder, effectively reducing the stress amplitude during startup and controlling the expansion difference, thus ensuring startup safety. When the temperature sensor reading of the thick-walled component reaches the set value for a certain period, or the reading significantly exceeds the set value, or after manual selection to shut down the system, the central control system generates a control signal to disconnect the power supply and stop the electric heating system.

[0036] In one embodiment of the present invention, the training process of the load power prediction model is as follows:

[0037] Based on historical operating data of power stations within the power consumption area of ​​coal-fired power units;

[0038] The power, load power, ambient temperature, ambient humidity, wind speed, light intensity and corresponding timestamp information from historical operating data are input into the data input layer to train the neural network model and obtain the trained load power prediction model. The input of the load power prediction model is the timestamp information and the output is the predicted load value. The timestamp information includes the weekday number and holiday information.

[0039] In this embodiment, the load forecasting process employs a deep learning prediction model based on Long Short-Term Memory (LSTM) networks. This involves collecting operational records from all power stations within the electricity consumption area and extracting historical operational data of the generating units, including power, load, temperature, humidity, wind speed, light intensity, and corresponding timestamp information (including weekday numbers and holiday markers). This data is then input into the data input layer. Specifically, when extracting historical operational data, sampling can be performed at preset time intervals (e.g., once per hour), with increased sampling frequency at specific time points (e.g., once per minute on weekends or holidays). When the grid load power fluctuates significantly, the sampling frequency is also increased accordingly to improve the granularity of historical operational data sampling during periods of load power fluctuation, thereby enhancing the accuracy of the load power forecasting model. It is understood that after the load power forecasting model is trained, it can also be updated at preset time intervals.

[0040] In one embodiment of the present invention, power, load power, ambient temperature, ambient humidity, wind speed, light intensity and corresponding timestamp information from historical operating data are input into the data input layer to train the neural network model and obtain a trained load power prediction model, including:

[0041] Based on the timestamp information, power, load power, ambient temperature, ambient humidity, wind speed, and light intensity are divided into time series to form time series data;

[0042] Timing data is input into the LSTM network layer, which includes at least one LSTM unit.

[0043] LSTM cells operate as follows:

[0044] The forget gate is represented as The input gate is represented as The output gate is represented as ,in, This is the hidden state from the previous moment. For the current input, and The weight values ​​obtained during training, and The bias obtained during training. It is the sigmoid activation function;

[0045] The hidden state of the last layer of the LSTM network is passed to the fully connected layer to obtain the predicted load value at a preset future time. ,in, for Forecast load value at time of day This represents the final hidden state of the LSTM network. and These are the parameters for the fully connected layer.

[0046] In this embodiment, the input data is standardized to eliminate dimensional differences, and missing data is imputed to construct a sample dataset with a unified timestamp. A time series analysis-based architecture is selected to build a load forecasting model. The preprocessed time series data is input into an LSTM network layer, which consists of at least one LSTM unit. Through the synergistic effect of forget gates, input gates, and output gates, long-term dependency features and periodic variation patterns in the load sequence are extracted to obtain load forecast results for the next 1-2 hours and generate a power load forecast curve. This power load forecast curve can predict whether there will be a significant change in the output of new energy sources or the power load, and inform coal-fired power plants of future load changes and estimated timeframes.

[0047] In one embodiment of the present invention, determining the target load power and load adjustment time of the coal-fired power unit based on the fluctuation amplitude, the predicted time point of the fluctuation amplitude, the maximum ramp rate of the coal-fired power unit, the upward reserve capacity, and the downward reserve capacity includes:

[0048] Based on the fluctuation range, the upward and downward reserve capacity of the coal-fired power units, the target load power is determined using the following formula:

[0049] ,

[0050] in, For the target load power, This represents the current load power. The fluctuation amplitude, For downside reserve capacity, Reserve capacity for upward movement;

[0051] The load adjustment time is determined based on the target load power and the maximum ramp rate of the coal-fired unit, using the following formula:

[0052] ,

[0053] in, For load adjustment time, The rate of change of power, , This represents the maximum ramp rate for coal-fired power units. The time required to reach the predicted time point, This is for the safety factor.

[0054] In this embodiment, the optimal target power adjustment strategy for dealing with fluctuations is determined by combining the upward and downward reserve capacity of the coal-fired unit with the target load power. The current power value is added to the predicted fluctuation amplitude, while taking into account the upward and downward reserve capacity of the coal-fired unit, so that the target load power is within the safe operating capacity range of the coal-fired unit.

[0055] The safety factor is introduced during load adjustment to provide a safety margin, ensuring that adjustments are completed before the power grid requires them, and allowing sufficient preheating time for the electric heat tracing system. Pre-cooling time. If... This indicates that there is ample time, allowing for a longer adjustment time than the minimum required time, which helps reduce thermal stress on the equipment and saves energy consumption for electric heat tracing. If This means that time is tight, and actions need to be started in advance to take on as many adjustment tasks as possible.

[0056] In one embodiment of the present invention, the electric heating power of thick-walled components of a coal-fired power unit is determined by a feedforward-feedback composite control algorithm based on the target load power and load adjustment time, including:

[0057] The target temperature value for thick-walled components is determined using the following formula, based on the target load power and load adjustment time:

[0058]

[0059] in, for The target temperature value at any given time. This represents the initial temperature value for the thick-walled component. Let the target load power function be... For load adjustment time function;

[0060] The electric heating power for thick-walled components is determined based on the target temperature value, and is calculated using the following formula:

[0061] ,

[0062] in, This refers to the power of the electric heating tracing. For the equivalent heat capacity of thick-walled components, This is the equivalent heat loss coefficient. For ambient temperature, It is a dynamic gain function. .

[0063] In this embodiment, for The target temperature value of the thick-walled component at any given time is a dynamic temperature trajectory. Based on the target load power and load adjustment time, the target temperature change trajectory of the thick-walled component over a future period is calculated using the target temperature model. Based on the target temperature change trajectory, the electric heat tracing power is calculated using a feedforward-feedback composite control algorithm.

[0064] Among them, dynamic gain function Satisfy: When When it exceeds the preset threshold, Take the first preset gain value, when When less than or equal to a preset threshold, Take the second gain value. The first gain value is greater than the second gain value.

[0065] In one embodiment of the present invention, heating the outer wall of a thick-walled component according to the electric heating power and a preset electric heating algorithm includes:

[0066] The electric heating algorithm is as follows:

[0067] ,

[0068] in, , For the calculation period, Temperature parameters for thick-walled components. and For parameters in the current The parameter values ​​at each time point and the parameter values ​​from the previous sampling period. This is the proportionality coefficient. For electric heat tracing power in The increment function at time step, This refers to the power of the electric heating tracing. This represents the difference between the target load power and the current load power. The algorithm predicts the power increment for the next cycle based on the derivative of the parameters from the previous cycle, thus smoothly eliminating the parameter difference and preventing thermal shock.

[0069] In this embodiment, upon receiving the target load power and load adjustment time, the coal-fired power unit collects the following parameters: the outer wall temperature of the boiler's screen superheater, rear screen superheater, final stage superheater, high-temperature reheater, low-temperature reheater, low-temperature superheater, and economizer; the inner wall temperature of the upper and lower high-pressure cylinders of the turbine; the inner and outer wall temperature of the upper high-pressure cylinder; the inner and outer wall temperature of the lower high-pressure cylinder; and the expansion value of the high-pressure cylinder. The target power value is then... and the unit's target ramp rate The external wall temperature differences of the boiler's screen-type superheater, rear screen superheater, final stage superheater, high-temperature reheater, low-temperature reheater, low-temperature superheater, and economizer, as well as the internal wall temperature differences of the upper and lower high-pressure cylinders of the turbine, the internal and external wall temperature differences of the upper and lower high-pressure cylinders, and the expansion difference of the high-pressure cylinder are input into the electric heat tracing power algorithm, with the calculation step size period set to [value missing]. The algorithm input is Time and previous cycle Various parameter values ​​at different times are used to calculate... The derivatives of various parameters at time points. Predicting based on the derivatives of various parameters. The power controller outputs power increments at all times to smoothly eliminate differences in various parameters. It switches to thermal standby when the outer wall temperature reaches the target temperature.

[0070] Temperature parameters of thick-walled components include: the outer wall temperature difference of the screen-type superheater, the rear screen superheater, the final stage superheater, the high-temperature reheater, the low-temperature reheater, the low-temperature superheater, the economizer, as well as various parameters such as the inner wall temperature difference of the upper and lower cylinders of the turbine high-pressure cylinder, the inner and outer wall temperature difference of the upper cylinder of the high-pressure cylinder, the inner and outer wall temperature difference of the lower cylinder of the high-pressure cylinder, and the expansion difference of the high-pressure cylinder. The output power of the power controller (with a maximum value) (and the minimum value of 0).

[0071] In one embodiment of the present invention, it further includes:

[0072] During the preset time period Then, based on the temperature parameters of the thick-walled components, the actual ramp rate of the coal-fired unit is determined using the following formula:

[0073] ,

[0074] in, The actual climbing rate is given by [value], and the thermal stress is determined based on the temperature parameters of the thick-walled component. The target ramp rate is determined based on the target load power.

[0075] In this embodiment, an example is used for illustration. For instance, according to a cold air transit warning issued by a local meteorological station, the wind force is expected to continue to decrease after 1 hour, resulting in a sudden drop of 900MW in wind power. At this time, through the active control method of interaction between coal-fired units and grid load in this embodiment of the invention, electric heating is applied to thick-walled components during the load adjustment period (i.e., the preparation stage). After the preparation stage is completed, the wind force decreases, and the output of the coal-fired unit also changes accordingly, starting to climb at the actual climbing rate. At this time, the coal-fired unit can climb rapidly at a rate exceeding 4-6% Pe / min (in the traditional mode, a 900MW load gap requires the use of 3 1000MW units to climb at a rate of 1.5-2% Pe / min for 30 minutes, while in the new mode, only 2 1000MW units are needed to climb at a rate of 4% Pe / min for 12 minutes to meet the requirement, not only exceeding the task, but also greatly increasing the economy and safety margin by saving one climbing unit). Climbing begins. After a certain period of time, the real-time power of the coal-fired power unit undertaking the ramp-up task has reached the target power value, and the frequency deviation of the entire network has been detected. Actual climbing rate The electric heat tracing system was taken out of service, and the power grid returned to stable operation.

[0076] Therefore, based on the method proposed in this study, by predicting future moments when the grid load will fluctuate significantly using a load power prediction model, and by heating the thick-walled components of the coal-fired unit 1-2 hours in advance, the temperature difference between the inner and outer walls of the boiler's thick-walled components during load changes can be effectively reduced. This meets the needs of deep interaction between the grid and the coal-fired unit, enabling rapid load changes and start-up / shutdown of the coal-fired unit, thus facilitating advance interaction and regulation between the coal-fired unit and the grid. Furthermore, it reduces fatigue life loss during load changes and improves the safety of unit startup.

[0077] According to another embodiment, the present invention provides an active control system for interaction between a coal-fired power unit and a power grid load, the system comprising a power grid and a coal-fired power unit;

[0078] The power grid is used to predict, through a trained load power prediction model, that the fluctuation range of the power grid load power will be greater than a preset range threshold. Then, based on the fluctuation range, the prediction time point that generates the fluctuation range, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity, the target load power and load adjustment time of the coal-fired unit are determined, and the target load power and load adjustment time are sent to the coal-fired unit.

[0079] The coal-fired power unit is used to determine the electric heating power of the thick-walled components of the coal-fired power unit based on the target load power and load adjustment time, using a feedforward-feedback composite control algorithm; and

[0080] The outer wall of the thick-walled component is preheated according to the electric heating power and the preset electric heating algorithm, so as to cope with the fluctuation amplitude by reducing the temperature difference between the inner and outer walls of the thick-walled component.

[0081] According to another embodiment, the present invention provides an active control device for interaction between a coal-fired power unit and the power grid load. Figure 2 A schematic block diagram of an active control device for interaction between a coal-fired power unit and the grid load, according to one embodiment, is shown. It will be understood that this device can be implemented using any device, equipment, platform, or cluster of devices with computing and processing capabilities. Figure 2 As shown, the device includes: a load power prediction module 200, a heating power prediction module 202, and a thick-walled component heating module 204. The main functions of each component are as follows:

[0082] The load power prediction module is used to determine the target load power and load adjustment time of the coal-fired unit based on the fluctuation range, the prediction time point at which the fluctuation range is generated, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity if the fluctuation range of the power grid is predicted to be greater than the preset amplitude threshold by the trained load power prediction model.

[0083] The heating power prediction module is used to determine the electric heating power of the thick-walled components of the coal-fired unit based on the target load power and load adjustment time through a feedforward-feedback composite control algorithm.

[0084] A thick-walled component heating module is used to heat the outer wall of the thick-walled component according to the electric heating power and a preset electric heating algorithm, so as to cope with the fluctuation amplitude by reducing the temperature difference between the inner and outer walls of the thick-walled component.

[0085] As a preferred embodiment, the training process of the load power prediction model is as follows:

[0086] Based on historical operating data of power stations within the power consumption area of ​​the coal-fired power unit;

[0087] The power, load power, ambient temperature, ambient humidity, wind speed, light intensity and corresponding timestamp information from the historical operating data are input into the data input layer to train the neural network model and obtain the trained load power prediction model. The input of the load power prediction model is the timestamp information and the output is the predicted load value. The timestamp information includes weekday identifier and holiday identifier.

[0088] In one preferred embodiment, the step of inputting the power, load power, ambient temperature, ambient humidity, wind speed, light intensity and corresponding timestamp information from the historical operating data into the data input layer to train the neural network model and obtain the trained load power prediction model includes:

[0089] Based on the timestamp information, the power, load power, ambient temperature, ambient humidity, wind speed, and light intensity are divided into time series data.

[0090] The timing data is input into an LSTM network layer, the LSTM network layer including at least one LSTM unit;

[0091] The LSTM unit operates as follows:

[0092] The forget gate is represented as The input gate is represented as The output gate is represented as ,in, This is the hidden state from the previous moment. For the current input, and The weight values ​​obtained during training, and The bias obtained during training. It is the sigmoid activation function;

[0093] The hidden state of the last layer of the LSTM network is passed to the fully connected layer to obtain the predicted load value at a preset future time. ,in, for Forecast load value at time of day This represents the final hidden state of the LSTM network. and These are the parameters for the fully connected layer.

[0094] As a preferred embodiment, determining the target load power and load adjustment time of the coal-fired power unit based on the fluctuation amplitude, the predicted time point of the fluctuation amplitude, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity includes:

[0095] Based on the fluctuation range, the upward and downward reserve capacity of the coal-fired unit, the target load power is determined using the following formula:

[0096]

[0097] in, For the target load power, This represents the current load power. The fluctuation amplitude, For downside reserve capacity, Reserve capacity for upward movement;

[0098] The load adjustment time is determined based on the target load power and the maximum ramp rate of the coal-fired unit, using the following formula:

[0099]

[0100] in, For load adjustment time, The rate of change of power, , This represents the maximum ramp rate for coal-fired power units. The time required to reach the predicted time point, This is for the safety factor.

[0101] As a preferred embodiment, determining the electric heating power of the thick-walled components of the coal-fired unit using a feedforward-feedback composite control algorithm based on the target load power and load adjustment time includes:

[0102] The target temperature value of the thick-walled component is determined using the following formula based on the target load power and load adjustment time:

[0103]

[0104] in, for The target temperature value at any given time. This is the initial temperature value of the thick-walled component. Let the target load power function be... For load adjustment time function;

[0105] The electric heating power of the thick-walled component is determined based on the target temperature value, and is calculated using the following formula:

[0106]

[0107] in, This refers to the power of the electric heating tracing. For the equivalent heat capacity of thick-walled components, This is the equivalent heat loss coefficient. For ambient temperature, It is a dynamic gain function. .

[0108] In a preferred embodiment, heating the outer wall of the thick-walled component according to the electric heating power and a preset electric heating algorithm includes:

[0109] The electric heating algorithm is as follows:

[0110]

[0111] in, , For the calculation period, The temperature parameters of the thick-walled component. and For parameters in the current The parameter values ​​at each time point and the parameter values ​​from the previous sampling period. This is the proportionality coefficient. For electric heat tracing power in The increment function at time step, This refers to the power of the electric heating tracing. The difference between the target load power and the current load power.

[0112] As a preferred embodiment, it also includes:

[0113] During the preset time period Tr Then, based on the temperature parameters of the thick-walled component, the actual ramp rate of the coal-fired unit is determined using the following formula:

[0114] ,

[0115] in, The actual climbing rate is denoted by , and the thermal stress is determined based on the temperature parameters of the thick-walled component. The target ramp rate is determined based on the target load power.

[0116] According to another embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed in a computer, causes the computer to perform a combination Figure 1 The method described.

[0117] According to another embodiment, an electronic device is also provided, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, it implements a combination... Figure 1 The method described.

[0118] In summary, this invention provides an active control method and system for the interaction between coal-fired power units and grid load. By predicting future moments of significant grid load fluctuations through a load power prediction model, and pre-heating the thick-walled components of the coal-fired power unit, the temperature difference between the inner and outer walls of these components during load changes is effectively reduced. This improves the response speed of the coal-fired power unit, reduces fatigue life losses during load changes, and enhances the safety of unit startup. The response time can be shortened by more than 25%. According to the load power prediction model, when a sudden change in renewable energy output caused by environmental factors such as wind and solar power is predicted, leading to grid frequency fluctuations, the boiler and turbine have already completed their preparations. The coal-fired power unit can then rapidly change load at a rate of 4%-6% Pe / min, more quickly absorbing grid fluctuations caused by sudden changes in renewable energy output. This reduces fatigue losses of thick-walled components by more than 20%, lowers investment costs by more than 90% compared to energy storage, and allows for parallel operation with existing DCS without complex modifications. This effectively enhances the renewable energy absorption capacity and provides efficient peak-shaving support for grids with a high proportion of renewable energy.

[0119] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0120] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in this invention can be implemented using hardware, software, firmware, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium.

[0121] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. An active control method for interaction between a coal-fired power unit and a power grid load, applied to an active control system for interaction between a coal-fired power unit and a power grid load, the system comprising a power grid and a coal-fired power unit, characterized in that, include: If the power grid predicts through a trained load power prediction model that the fluctuation range of the power grid load power will be greater than a preset amplitude threshold, then the target load power and load adjustment time of the coal-fired unit are determined based on the fluctuation range, the prediction time point that generates the fluctuation range, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity, and the target load power and load adjustment time are sent to the coal-fired unit. The electric heating power of the thick-walled components of the coal-fired power unit is determined by a feedforward-feedback composite control algorithm based on the target load power and load adjustment time. The coal-fired power unit preheats the outer wall of the thick-walled component according to the electric tracing heating power and the preset electric tracing heating algorithm, so as to cope with the fluctuation amplitude by reducing the temperature difference between the inner and outer walls of the thick-walled component; The step of determining the electric heating power of the thick-walled components of the coal-fired power unit using a feedforward-feedback composite control algorithm based on the target load power and load adjustment time includes: The target temperature value of the thick-walled component is determined using the following formula based on the target load power and load adjustment time: in, for The target temperature value at any given time. This is the initial temperature value of the thick-walled component. Let the target load power function be... For load adjustment time function; The electric heating power of the thick-walled component is determined based on the target temperature value, and is calculated using the following formula: in, This refers to the power of the electric heating tracing. For the equivalent heat capacity of thick-walled components, This is the equivalent heat loss coefficient. For ambient temperature, It is a dynamic gain function. .

2. The method according to claim 1, characterized in that, The training process for the load power prediction model is as follows: Based on historical operating data of power stations within the power consumption area of ​​the coal-fired power unit; The power, load power, ambient temperature, ambient humidity, wind speed, light intensity and corresponding timestamp information from the historical operating data are input into the data input layer to train the neural network model and obtain the trained load power prediction model. The input of the load power prediction model is the timestamp information and the output is the predicted load value. The timestamp information includes weekday identifier and holiday identifier.

3. The method according to claim 2, characterized in that, The process involves inputting the power, load power, ambient temperature, ambient humidity, wind speed, light intensity, and corresponding timestamp information from the historical operating data into the data input layer to train the neural network model and obtain a trained load power prediction model, including: Based on the timestamp information, the power, load power, ambient temperature, ambient humidity, wind speed, and light intensity are divided into time series data. The timing data is input into an LSTM network layer, the LSTM network layer including at least one LSTM unit; The LSTM unit operates as follows: The forget gate is represented as The input gate is represented as The output gate is represented as ,in, This is the hidden state from the previous moment. For the current input, and The weights obtained during training, and The bias obtained during training. It is the sigmoid activation function; The hidden state of the last layer of the LSTM network is passed to the fully connected layer to obtain the predicted load value at a preset future time. ,in, for Forecast load value at time of day This represents the final hidden state of the LSTM network. and These are the parameters for the fully connected layer.

4. The method according to claim 1, characterized in that, The determination of the target load power and load adjustment time of the coal-fired power unit based on the fluctuation amplitude, the predicted time point of the fluctuation amplitude, the maximum ramp rate of the coal-fired power unit, the upward reserve capacity, and the downward reserve capacity includes: Based on the fluctuation range, the upward and downward reserve capacity of the coal-fired unit, the target load power is determined using the following formula: in, For the target load power, This represents the current load power. The fluctuation amplitude, For downside reserve capacity, Reserve capacity for upward movement; The load adjustment time is determined based on the target load power and the maximum ramp rate of the coal-fired unit, using the following formula: in, For load adjustment time, The rate of change of power, , This represents the maximum ramp rate for coal-fired power units. The time required to reach the predicted time point. This is for the safety factor.

5. The method according to claim 1, wherein heating the outer wall of the thick-walled component according to the electric heating power and a preset electric heating algorithm comprises: The electric heating algorithm is as follows: in, , For the calculation period, The temperature parameters of the thick-walled component. For parameters in the current The parameter values ​​at each time point and the parameter values ​​from the previous sampling period. This is the proportionality coefficient. For electric heat tracing power in The increment function at time step, This refers to the power of the electric heating tracing. The difference between the target load power and the current load power.

6. The method according to claim 5, characterized in that, Also includes: During the preset time period Then, based on the temperature parameters of the thick-walled component, the actual ramp rate of the coal-fired unit is determined using the following formula: in, The actual climbing rate is denoted by , and the thermal stress is determined based on the temperature parameters of the thick-walled component. The target ramp rate is determined based on the target load power.

7. An active control system for interaction between a coal-fired power unit and the power grid load, characterized in that, The system includes a power grid and coal-fired power units; The power grid is used to predict, through a trained load power prediction model, that the fluctuation range of the power grid load power will be greater than a preset range threshold. Then, based on the fluctuation range, the prediction time point that generates the fluctuation range, the maximum ramp rate of the coal-fired unit, the upward reserve capacity, and the downward reserve capacity, the target load power and load adjustment time of the coal-fired unit are determined, and the target load power and load adjustment time are sent to the coal-fired unit. The coal-fired power unit is used to determine the electric heating power of the thick-walled components of the coal-fired power unit based on the target load power and load adjustment time, using a feedforward-feedback composite control algorithm; and The outer wall of the thick-walled component is preheated according to the electric heating power and the preset electric heating algorithm, so as to cope with the fluctuation amplitude by reducing the temperature difference between the inner and outer walls of the thick-walled component; The electric heating power is calculated as follows: The target temperature value of the thick-walled component is determined using the following formula based on the target load power and load adjustment time: in, for The target temperature value at any given time. This is the initial temperature value of the thick-walled component. Let the target load power function be... For load adjustment time function; The electric heating power of the thick-walled component is determined based on the target temperature value, and is calculated using the following formula: in, This refers to the power of the electric heating tracing. For the equivalent heat capacity of thick-walled components, This is the equivalent heat loss coefficient. For ambient temperature, It is a dynamic gain function. .

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program, implements the method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the method of any one of claims 1-6.

Images