Automatic power generation dispatching control method and device of thermal power generating unit

Abstract

The invention provides an automatic power generation dispatching control method and device of a thermal power generating unit, and relates to the technical field of intelligent power grids, and the method comprises the steps: collecting multi-source observation data of an electric side and a hearth combustion side, and extracting a hearth negative pressure fluctuation amplitude, a flame centroid jitter amplitude and a sound pressure fluctuation amplitude to construct a sound-combustion coupling stability index; establishing a coal consumption cost function and an emission cost function in combination with historical unit operation data; dividing an acoustic-fuel coupling stability domain according to an acoustic-fuel coupling stability index, monitoring a rising trend, and taking an acoustic-fuel coupling risk as a scheduling constraint triggering condition; a scheduling optimization model is reconstructed in a risk dominant scene, active power output, a coal mill combination and air distribution parameters are used as decision variables, an optimal track is solved with physical limits, power grid side constraints and a stability domain as limiting conditions, and finally the optimal track is issued to a coordination control layer. According to the method, the output curve can be actively deviated in the acoustic-fuel coupling instability risk rising stage to avoid acoustic resonance, and therefore the dynamic dispatching capacity of the unit is guaranteed.

Description

Technical Field

[0001] This invention relates to the technical field of smart grids, specifically to an automatic power generation dispatch and control method and device for thermal power units. Background Technology

[0002] The power generation dispatch and control of thermal power units usually relies on automatic generation control systems. The active power output of the units is dynamically adjusted according to the grid load demand and the coal consumption characteristics of the units, so that the units can operate as close to the output curve as possible while meeting the grid frequency stability constraints and reserve capacity constraints. Combustion efficiency is maintained by combining the start-up and shutdown strategy of the coal mill and the regulation of the air supply system. At the same time, continuous emission monitoring data is used as auxiliary feedback to achieve a comprehensive balance between economy and environmental protection. The dispatch side generally regards the units as quasi-steady-state power sources limited by the ramp rate, and achieves grid-side frequency regulation and peak shaving targets by optimizing the active power output setpoint.

[0003] When strong acoustic coupling disturbances occur in the combustion system, such as rapid changes in pulverized coal concentration leading to low-frequency oscillations in the furnace, automatic power generation dispatch control needs to temporarily deviate from the output curve. Current technologies lack explicit modeling and constraints on the combustion-acoustic coupling state within the unit's furnace at the dispatch layer. This causes automatic power generation dispatch to still forcibly follow commands based on economic priority logic when load fluctuates frequently or when the pulverizer and air distribution combinations fall into acoustically sensitive areas. This easily induces amplification of furnace negative pressure, combustion noise, and low-frequency oscillations in flame morphology, leading to combustion instability, drastic emission deterioration, and even triggering protective shutdowns. It fails to deviate from the output curve in time and avoid acoustic resonance conditions during periods of rising acoustic-combustion coupling instability risk, thus insufficient to guarantee the dynamic dispatch capability of thermal power units under complex disturbances. Summary of the Invention

[0004] This invention provides an automatic power generation dispatch control method and device for thermal power units, which can actively deviate from the output curve to avoid acoustic resonance during the stage of increased risk of acoustic-fuel coupling instability, thereby ensuring the dynamic dispatch capability of the unit.

[0005] In a first aspect of the present invention, an automatic power generation dispatch control method for thermal power units is provided, the method comprising: Collect multi-source observation data from both the electric side and the furnace combustion side; The multi-source observation data is preprocessed to extract the furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude, and sound pressure fluctuation amplitude, and an acoustic-fuel coupling stability index is constructed. Based on historical operating data, the unit's coal consumption cost function and emission cost function are constructed to predict the changes in fuel consumption and emissions of the unit under different active power outputs, different coal mill combinations, and different secondary air distribution methods. Based on the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations and different secondary air distribution methods, the acoustic-fuel coupling stability domain is divided. The operating point that induces acoustic-fuel coupling instability is set as the prohibited domain, and the operating point at the acoustic-fuel coupling instability boundary is set as the risk domain. When the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain, the current scheduling scenario is identified as an acoustic-fuel coupling constraint-dominated scenario. In the scenario dominated by acoustic-fuel coupling constraints, the scheduling optimization model is reconstructed, and the active power output setpoint of the unit, the start-stop status of the coal mill, and the allocation of primary air volume and secondary air volume are used as decision variables. The optimization objective is constructed based on the unit coal consumption cost function and the emission cost function, and the unit physical limit, grid-side constraints, and the acoustic-fuel coupling stability domain are used as optimization constraints. The optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectories obtained from solving the scheduling optimization model are sent down to the coordination control layer, so that the thermal power unit can avoid acoustically sensitive operating conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases.

[0006] In a second aspect of the invention, an automatic power generation dispatch control device for a thermal power unit is provided. The device is used to execute an automatic power generation dispatch control method for a thermal power unit as described in any of the above embodiments. The device includes an acquisition module, a processing module, and an output module, wherein: The acquisition module is used to collect multi-source observation data from the electric side and the furnace combustion side; The processing module is used to preprocess the multi-source observation data to extract the furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude and sound pressure fluctuation amplitude, and to construct an acoustic-flame coupling stability index. The processing module is used to construct the unit's coal consumption cost function and emission cost function based on historical operating data, so as to predict the changes in fuel consumption and emissions of the unit under different active power outputs, different coal mill combinations and different secondary air distribution methods. The processing module is used to divide the acoustic-fuel coupling stability domain according to the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations and different secondary air distribution methods, set the operating point that induces acoustic-fuel coupling instability as the prohibited domain, and set the operating point at the acoustic-fuel coupling instability boundary as the risk domain. The processing module is used to identify the current scheduling scenario as a scenario dominated by acoustic-fuel coupling constraints when the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain. The processing module is used to reconstruct the scheduling optimization model in the scenario dominated by acoustic-fuel coupling constraints. It takes the active power output setpoint of the unit, the start-stop status of the coal mill, and the allocation of primary air volume and secondary air volume as decision variables, and constructs the optimization objective based on the unit coal consumption cost function and the emission cost function. At the same time, it uses the unit physical limit, grid-side constraints and the acoustic-fuel coupling stability domain as optimization constraints. The output module is used to send the optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectory obtained by solving the scheduling optimization model to the coordination control layer, so that the thermal power unit can avoid acoustically sensitive operating conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases.

[0007] In a third aspect of the invention, an electronic device is provided, including a processor, a memory, a user interface, and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any of the preceding embodiments.

[0008] In a fourth aspect of the invention, a non-transitory computer-readable storage medium is provided, the computer-readable storage medium storing instructions that, when executed, perform the method as described in any of the preceding claims.

[0009] In summary, one or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages: 1. This invention introduces a furnace acoustic-fuel coupling stability index and establishes an acoustic-fuel coupling stability domain, explicitly incorporating the acoustic sensitivity of the unit's operating conditions into the scheduling constraints. When the acoustic-fuel coupling risk is observed to be increasing in real time and the operating conditions are approaching the instability boundary, the system automatically switches to an optimization mode dominated by acoustic-fuel coupling constraints. Under the premise of meeting the unit's physical limits and the grid's requirements, the system dynamically adjusts the active power output, coal mill combination, and air distribution mode, enabling automatic power generation scheduling to actively deviate from the traditional economic output curve and avoid the furnace acoustic resonance region in advance. This effectively maintains combustion stability and emission controllability and ensures dynamic scheduling capabilities under complex disturbance conditions.

[0010] 2. By performing unified time reference alignment, frequency band feature extraction, and normalized weighted combination on the furnace negative pressure time series, sound pressure time series, and flame video series, an acoustic-fuel coupling stability index is constructed. This enables the scheduling layer to quantify the coupling strength between furnace acoustic oscillation and flame oscillation in real time, thereby identifying combustion instability trends in advance and providing a dynamic risk characterization basis for subsequent scheduling optimization.

[0011] 3. By utilizing historical operating data to establish unit coal consumption cost functions and emission cost functions, the dispatching layer can accurately estimate changes in fuel consumption and emissions when predicting different future active power outputs, coal mill combinations, and secondary air distribution methods. This ensures that the dispatching optimization process achieves operational goals while meeting both economic and environmental requirements.

[0012] 4. By dividing the acoustic-combustion coupling stability domain based on the acoustic-combustion coupling stability index and setting stable operating points, risk operating points and prohibited operating points, the acoustically sensitive area is explicitly included in the scheduling constraints, avoiding the scheduling optimization from mistakenly selecting the prohibited operating conditions as the economically optimal solution, thereby providing a combustion-acoustic safety boundary guarantee.

[0013] 5. By identifying the continuous upward trend of the acoustic-fuel coupling stability index and combining it with the stability domain information of the operating point, the acoustic-fuel coupling constraint is used as the trigger condition for the dominant scheduling scenario, so that the scheduling control can switch from the economy-first strategy to the safety-first strategy in a timely manner, thereby actively avoiding dangerous areas when the acoustic-fuel coupling risk approaches the instability boundary.

[0014] 6. By reconstructing the scheduling optimization model in the scenario dominated by acoustic-fuel coupling constraints and introducing acoustic-fuel coupling stability domain constraints, the optimization process not only satisfies the physical limits of the unit and the constraints of the grid side, but also avoids entering the acoustic resonance region and limits the long-term residence in the risk region, thereby maintaining frequency regulation and peak shaving capabilities while ensuring combustion stability. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating an automatic power generation dispatch and control method for thermal power units disclosed in an embodiment of the present invention; Figure 2 This is a schematic diagram of a module of an automatic power generation dispatch and control device for a thermal power unit disclosed in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of the present invention.

[0016] Explanation of reference numerals in the attached drawings: 201, acquisition module; 202, processing module; 203, output module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] In the description of the embodiments of the present invention, words such as "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "for example" or "for instance" in the embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of words such as "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0019] In the description of the embodiments of the present invention, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0020] The existing thermal power unit power dispatch mainly relies on the automatic power generation control system to adjust the unit output based on economic efficiency and grid demand. However, since the combustion-acoustic coupling state of the furnace is not explicitly modeled and constrained, when the coal mill combination and air distribution combination enter the acoustically sensitive area or the rapid load fluctuation causes low-frequency oscillations in the furnace, they are still forced to follow the dispatch instructions. This can easily lead to combustion instability, emission deterioration, or even protective shutdown. It is impossible to actively deviate from the output curve to avoid acoustic resonance when the risk of acoustic-combustion coupling instability increases, thus making it difficult to ensure the dynamic safety margin and continuous dispatch capability of the unit.

[0021] This embodiment discloses an automatic power generation dispatch and control method for thermal power units, referring to... Figure 1 This includes the following steps S110-S170: S110 collects multi-source observation data from the power grid side and the furnace combustion side.

[0022] The automatic power generation dispatch and control method for thermal power units disclosed in this invention is applied to a server. The server includes, but is not limited to, electronic devices such as mobile phones, tablets, wearable devices, and PCs (Personal Computers), and can also be a backend server running the automatic power generation dispatch and control method for thermal power units. The server can be implemented using a standalone server or a server cluster composed of multiple servers.

[0023] First, the dispatch and status quantities that need to be monitored are determined on the grid side. Measurement loops are configured around the unit connection point. Bus voltage, active current, and reactive current are obtained through voltage transformers and current transformers installed on the generator output side and the high-voltage side of the plant service transformer. Simultaneously, the active power command of automatic generation control, the operating status of automatic generation control, and frequency regulation assessment related indicators are obtained on the communication link of the unit connected to the dispatch automatic generation control system. The above electrical quantities are sampled and digitized by the synchronous phasor measurement unit or high-precision measurement and control device. According to the set sampling period, the bus voltage time series, active power output time series, grid frequency time series, and automatic generation control command time series are generated. These are transmitted to the centralized data processing unit through the power plant's internal process bus, so that the active power output and grid operating status on the grid side are continuously recorded on the time axis.

[0024] On the combustion side of the furnace, to obtain multi-source observation data reflecting the combustion process and acoustic oscillation state, furnace negative pressure measuring points are arranged at typical locations in the upper, middle, and lower parts of the furnace. High-temperature differential pressure transmitters are used to collect furnace negative pressure signals, and broadband acoustic sensors are arranged near the observation holes in the burner area to collect furnace sound pressure signals. At the same time, protective industrial cameras are installed at appropriate locations to acquire flame video images. The furnace negative pressure signals and sound pressure signals are sent to the local acquisition unit after being subjected to anti-interference filtering and amplification conversion by the local signal conditioning device. They are synchronously sampled at a sampling period of milliseconds to tens of milliseconds and stamped with high-precision timestamps. Flame video images are continuously acquired at a preset frame rate with acquisition time information and transmitted to the image processing unit via industrial Ethernet. This enables the furnace negative pressure time series, sound pressure time series, and flame video time series to establish a correspondence with the grid-side data in the time dimension.

[0025] Regarding the fuel supply and air volume distribution status, the output of each coal mill is obtained by installing flow or power measuring devices at the outlet of each coal mill. The total primary air volume and branch air volume are obtained by installing flow measuring devices on the primary air fan outlet main pipe and each primary air branch. The total secondary air volume and air distribution ratio are obtained by configuring flow measuring devices on the secondary air main pipe, each burner layer and each side wall secondary air branch. At the same time, flue gas temperature measuring points, oxygen analyzers and continuous emission monitoring devices are arranged at the induced draft fan outlet and key sections of the flue to obtain information such as flue gas temperature, excess air coefficient, nitrogen oxide emission concentration and carbon monoxide emission concentration. The above-mentioned fuel side and air volume side signals are collected by the local acquisition unit at a sampling period coordinated with the furnace negative pressure and sound pressure. Together with the coal mill start-up and shutdown status signals and the fan operation status signals, they form time series data describing the coal powder supply path and air supply path, and are aggregated to the centralized data processing unit through the plant process network.

[0026] In the centralized data processing unit, the active power output time series, grid frequency time series, and automatic generation control command time series from the grid side, the furnace negative pressure time series, sound pressure time series, and flame video time series from the furnace combustion side, and the coal mill output time series, primary air volume time series, secondary air volume time series, flue gas temperature time series, oxygen content time series, and emission concentration time series from the fuel and air volume side are uniformly time-aligned. Based on a unified system clock or high-precision clock synchronization protocol, a common time axis is reconstructed for all observations, so that the grid side state, combustion side state, and emission side state at the same moment correspond one-to-one in the data structure. At the same time, quality checks and anomaly marking are performed on the raw observation data. Values ​​exceeding the range, values ​​that have not changed for a long time, and obvious noise data are identified and labeled, providing clean and time-consistent multi-source observation data input for subsequent acoustic-fuel coupling feature extraction, unit coal consumption cost modeling, and emission cost modeling.

[0027] Based on the aforementioned data collection and aggregation, various observation data are structured and stored according to categories such as active power output, active power commands, furnace negative pressure, furnace sound pressure, flame images, coal mill output, primary air volume, secondary air volume, flue gas temperature, oxygen content, and emission concentration. A comprehensive state vector containing both grid-side and furnace combustion-side observations is generated for each time step. This allows subsequent calculations of acoustic-fuel coupling stability indicators to simultaneously utilize synchronous information from both the grid and furnace combustion sides. When constructing the unit coal consumption cost function and emission cost function, the scheduling optimization model can reference active power output, coal mill combination, and air distribution mode on the same time axis. This establishes a sequential correlation between multi-source observation data from the grid and furnace combustion sides and acoustic-fuel coupling control decisions at the data level.

[0028] S120 preprocesses multi-source observation data to extract furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude, and sound pressure fluctuation amplitude, and constructs an acoustic-fuel coupling stability index.

[0029] In one possible implementation, multi-source observation data is preprocessed to extract the furnace negative pressure fluctuation amplitude, flame centroid jitter amplitude, and sound pressure fluctuation amplitude, and an acoustic-flame coupling stability index is constructed. Specifically, this includes: aligning the furnace negative pressure time series, sound pressure time series, and flame video sequence with a unified time reference and resampling them on the same time axis, ensuring a one-to-one correspondence between the furnace negative pressure discrete sequence, sound pressure discrete sequence, and flame image frame sequence at each discrete moment; performing detrending and bandpass filtering on the furnace negative pressure discrete sequence, and obtaining the furnace negative pressure fluctuation amplitude, which characterizes the intensity of pressure fluctuations related to the acoustic modal frequency band of the furnace, by calculating the effective value of the bandpass-filtered sequence within a sliding time window; performing detrending and bandpass filtering on the sound pressure discrete sequence, constructing an analytical signal from the bandpass-filtered sequence and extracting the envelope amplitude, and then calculating the average value within a sliding time window. The average envelope amplitude is used to obtain the sound pressure fluctuation amplitude, which characterizes the intensity of acoustic oscillation in the furnace. The flame video sequence is subjected to grayscale conversion, denoising, and brightness normalization, and the burner flame region is extracted. The pixel brightness of the flame region is weighted and summed to obtain the flame centroid coordinates. The offset of the flame centroid relative to the time average position is calculated within a sliding time window to obtain the flame centroid jitter amplitude, which characterizes the degree of overall flame oscillation and shape drift. Based on the reference amplitude and fluctuation range of the furnace negative pressure fluctuation amplitude, flame centroid jitter amplitude, and sound pressure fluctuation amplitude obtained statistically under stable combustion conditions, the furnace negative pressure fluctuation amplitude, sound pressure fluctuation amplitude, and flame centroid jitter amplitude are normalized. The normalized furnace negative pressure fluctuation amplitude, flame centroid jitter amplitude, and sound pressure fluctuation amplitude are combined in a weighted and nonlinear amplification manner to form an acoustic-combustion coupling stability index.

[0030] Specifically, a single global clock source is first used on both the power grid and the furnace combustion side to synchronize and calibrate the timestamps of all acquisition devices, ensuring that the timestamps of the furnace negative pressure time series, sound pressure time series, and flame video series all refer to the same time reference. Based on this, a unified sampling period is selected as the target time step. A common time axis is constructed under this target time step. For each discrete moment on the common time axis, the corresponding discrete values ​​of furnace negative pressure and sound pressure are calculated from the original furnace negative pressure and sound pressure time series through interpolation or extraction. The flame image frame with the timestamp closest to this discrete moment is selected from the flame video series. These three types of data are stored together in the same time index location. A unified time reference means that all observations use the same reference clock to avoid time offsets between different acquisition devices. Resampling refers to recalculating or selecting data points at new fixed time intervals, ensuring a one-to-one correspondence between the furnace negative pressure discrete sequence, sound pressure discrete sequence, and flame image frame sequence at each discrete moment. This provides a synchronous multi-source data foundation for subsequent joint analysis within a sliding time window.

[0031] First, the average value of the furnace negative pressure discrete sequence within each sliding time window is calculated, and this average value is subtracted from the current sampling point value to remove slowly changing static components and long-term drift components. Detrending processing refers to eliminating the overall translation or slowly changing trend from the time series, making the remaining signal more concentrated in the dynamically changing part. After completing the detrending processing, a pre-designed digital bandpass filter is used to perform a convolution operation on the detrended furnace negative pressure discrete sequence. The passband frequency range of the bandpass filter matches the acoustic mode frequency of the furnace, thereby suppressing extremely low-frequency changes and high-frequency noise unrelated to the acoustic mode, resulting in a bandpass filtered sequence that highlights the energy of the target frequency band. Within each sliding time window, the average value of the squared sampling points of the bandpass filtered sequence is calculated, and the square root of the average value is taken to obtain the furnace negative pressure fluctuation amplitude corresponding to that time window, which is used to characterize the pressure fluctuation intensity related to the furnace acoustic mode frequency band. If described symbolically, the furnace negative pressure fluctuation amplitude can be defined as follows within the nth sliding time window:

[0032] in, This represents the furnace negative pressure fluctuation amplitude corresponding to the nth sliding time window, where N represents the number of sampling points within that sliding time window. This represents the negative pressure value of the furnace after bandpass filtering at the k-th sampling point within the nth sliding time window. The form of summing the squares and taking the square root makes the larger pressure fluctuations contribute more to the result, thereby effectively characterizing the oscillation energy in the acoustic modal frequency band of the furnace.

[0033] Similarly, within each sliding time window, the discrete sound pressure sequence is first detrended to eliminate mean offset. Then, a digital bandpass filter matching the furnace acoustic modal frequency is used to filter the detrended discrete sound pressure sequence to obtain a bandpass-filtered sound pressure sequence that retains only the target frequency band components. Based on this, the instantaneous amplitude envelope of the sound pressure sequence is obtained by constructing an analytic signal. The analytic signal can be composed of the bandpass-filtered sound pressure sequence and its Hilbert transform. The Hilbert transform refers to constructing an imaginary part sequence orthogonal to the original sequence from the real-valued time sequence to form a complex analytic signal. Within the nth sliding time window, the magnitude of the analytic signal is averaged over time, and this average magnitude is used as the sound pressure fluctuation amplitude, reflecting the overall level of the furnace acoustic oscillation intensity. Symbolically, the sound pressure fluctuation amplitude within the nth sliding time window can be defined as:

[0034] in, This represents the sound pressure fluctuation amplitude corresponding to the nth sliding time window. This represents the analytic signal value of the k-th sampling point within the n-th sliding time window. This indicates the amplitude of the analytical signal. This indicates the number of sampling points within the sliding time window. By averaging the amplitude of the analytical signal, the positive and negative cancellation caused by sign flipping can be reduced while preserving the characteristics of acoustic modal energy change, making the sound pressure fluctuation amplitude more stably characterize the acoustic oscillation intensity inside the furnace.

[0035] First, each frame of the flame image is converted to grayscale, transforming the multi-channel color image into a single-channel grayscale image to reduce the impact of color components on subsequent analysis. Denoising involves using spatial or temporal filtering algorithms to reduce high-frequency noise introduced by the camera sensor, smoke disturbance, etc. Brightness normalization maps the brightness values ​​of the flame area to a uniform range to reduce the impact of differences in absolute brightness values ​​under different operating conditions or exposure conditions. Then, threshold segmentation or clustering segmentation methods are used to extract the burner flame area, separating the set of pixels belonging to the flame from the background. For each pixel within the flame area, a brightness-weighted sum is performed based on its horizontal and vertical coordinates and corresponding brightness value to calculate the flame centroid coordinates of a single frame. The flame centroid coordinates of consecutive frames are arranged chronologically. Within each sliding time window, the offset of the flame centroid relative to the average position of the flame centroid within that time window is calculated. The flame centroid jitter amplitude is obtained through the root mean square of the offset or an equivalent measure, used to characterize the overall oscillation and morphological drift of the flame. If described symbolically, the flame centroid jitter amplitude can be defined as follows within the nth sliding time window:

[0036] in, This represents the amplitude of the flame centroid jitter corresponding to the nth sliding time window. and This represents the horizontal and vertical coordinates of the flame centroid in the k-th frame of the flame image within the nth sliding time window. and This represents the average value of the flame centroid coordinates within the sliding time window. This indicates the number of image frames within the sliding time window. The square root of the sum of squared coordinate offsets measures the degree of spatial jitter of the flame's centroid around its average position, thus reflecting the overall oscillation intensity and morphological stability of the flame.

[0037] First, a period of time under stable combustion conditions is selected as the baseline operating condition. From the sliding time window corresponding to this baseline operating condition, the average level and fluctuation range of the three amplitudes are statistically analyzed. For example, the baseline average value and baseline fluctuation scale are obtained for the furnace negative pressure fluctuation amplitude, and the corresponding baseline average value and baseline fluctuation scale are also obtained for the flame centroid vibration amplitude and sound pressure fluctuation amplitude, respectively. Stable combustion condition refers to the operating state of the unit when combustion is stable, there are no abnormal oscillations in the furnace, emission indicators are within the normal range, and load changes are slow. During subsequent operation, the furnace negative pressure fluctuation amplitude and flame centroid vibration amplitude calculated within each sliding time window are... The amplitude of the sound pressure fluctuation is normalized using the reference average value and the reference fluctuation scale, respectively, so that these normalized amplitudes are close to the same reference level near the stable combustion conditions. When a certain observation increases significantly during operation, its corresponding normalized value will be significantly greater than the reference level. The normalization process can be carried out by dividing the current amplitude by "the sum of the reference average value and several times the reference fluctuation scale". This method can keep the normalized value close to one within the stable fluctuation range, and quickly increase the normalized value when there is an abnormal increase. This provides a dimensionless unified measure for the subsequent unified comparison and weighted combination of multi-source amplitudes.

[0038] By assigning weighting coefficients and nonlinear amplification exponents to the three normalized amplitudes respectively, they are combined into a single acoustic-combustion coupling stability index, which is used to comprehensively characterize the combined degree of furnace acoustic oscillation and combustion instability under the current operating conditions. The weighting coefficients are used to express the relative importance of different observations in the acoustic-combustion coupling risk assessment. For example, in some units, furnace negative pressure fluctuations are more sensitive to instability, so the normalized amplitude of furnace negative pressure can be given a greater weight. The nonlinear amplification exponent is used to maintain a relatively smooth response when the normalized amplitude slightly exceeds the reference level, and to rapidly increase the contribution to the overall index when the normalized amplitude significantly exceeds the reference level, thereby amplifying severe anomalies. If described symbolically, the acoustic-combustion coupling stability index can be represented in the nth sliding time window as:

[0039] in, This represents the acoustic-combustion coupling stability index corresponding to the nth sliding time window. , , These represent the normalized amplitudes of furnace negative pressure fluctuation, flame centroid vibration, and sound pressure fluctuation within the time window, respectively. , , These are non-negative weighting coefficients used to balance the influence of the three observations. , , The nonlinear amplification index is not less than one, which is used to adjust the sensitivity of each observation in the abnormal amplification range. Through this weighted nonlinear combination method, when all observations are in the stable range, the acoustic-fuel coupling stability index is maintained at a low level. When any one or more observations increase significantly at the same time, the acoustic-fuel coupling stability index will rise rapidly.

[0040] S130 constructs unit coal consumption cost function and emission cost function based on historical operating data, enabling the prediction of fuel consumption and emission changes of the unit under different active power outputs, different coal mill combinations and different secondary air distribution methods.

[0041] In one possible implementation, a unit coal consumption cost function and an emission cost function are constructed based on historical operating data to predict changes in fuel consumption and emissions under different active power outputs, different coal mill combinations, and different secondary air distribution methods. Specifically, this includes: establishing an operating condition database containing active power output time series, cumulative fuel consumption records, coal mill start-up and shutdown status sequences, and primary and secondary air volume distribution records; associating the average active power output, coal mill combination status, and secondary air distribution method within each stable operating time period with the corresponding fuel consumption increment and emission monitoring increment to form a historical operating condition sample set; within the historical operating condition sample set, establishing a functional relationship between the equivalent fuel consumption rate and operating condition characteristics based on the fuel consumption increment and average active power output, coal mill combination status, and secondary air distribution method within each stable operating time window; obtaining the unit coal consumption cost function through parameter optimization; and adjusting the unit coal consumption cost function according to active power output,... The system predicts fuel consumption changes based on the coal mill combination status and secondary air distribution method. In a historical operating condition sample set, a functional relationship is established between the comprehensive emission index and operating condition characteristics based on the average emission concentration within each stable operating time window, the average active power output within that time window, the coal mill combination status, and the secondary air distribution method. An emission cost function is obtained through parameter optimization, and this function predicts emission changes based on active power output, coal mill combination status, and secondary air distribution method. During real-time operation, the active power output setpoint, coal mill combination, and secondary air distribution method output from the scheduling optimization are used as input variables, sequentially input into the unit coal consumption cost function and emission cost function to obtain the fuel consumption cost and emission cost for each candidate operating condition in the future time domain. The changes in fuel consumption and emissions are quantitatively represented in the scheduling optimization objective function, enabling scheduling control to predict fuel consumption and emission changes under different operating conditions based on the unit coal consumption cost function and emission function.

[0042] Specifically, by analyzing the historical operating records of the generating units, key time series describing the electrical and combustion-side states of the units are unified into a single data structure. The active power output time series and cumulative fuel consumption records are extracted from the power plant's production information system. The active power output time series is considered a time function representing the active power output level provided by the unit to the grid at each moment, and the cumulative fuel consumption records are considered a time function representing the cumulative fuel input since a certain reference moment. The pulverizer start-up and shutdown state sequences and primary and secondary air volume allocation records are extracted from the pulverizer monitoring system and the fan control system, respectively. The pulverizer start-up and shutdown state sequences are considered discrete indicators of the change in the start-up and shutdown states of each pulverizer over time, and the primary and secondary air volume allocation records are considered as the total primary air volume and the total secondary air volume. The multidimensional time series of air volume and secondary air volume ratios between different burner layers and different sidewalls varies over time. Based on the above time series, according to the unit operation log and operating condition change records, the operation process is divided into several stable operating periods according to the time intervals of slow load changes, stable combustion, and parameter fluctuations within the allowable range. In each stable operating period, the average active power output is calculated and used as the average active power output of the stable operating period. The difference between the cumulative fuel consumption records at the beginning and end of the period is calculated as the fuel consumption increment of the period. The average values ​​of nitrogen oxide emission concentration, carbon monoxide emission concentration, and particulate matter emission concentration in the period are extracted from the continuous emission monitoring system and combined to form the emission monitoring increment or emission characteristics.

[0043] Simultaneously, during the stable operation period, the overall operating or shutdown status of each coal mill in the coal mill start-up and shutdown sequence is statistically analyzed to form the coal mill combination status. The average distribution ratio of secondary air volume among different burner layers and different sidewalls is also statistically analyzed to form the secondary air distribution mode. The average active power output, coal mill combination status, secondary air distribution mode, fuel consumption increment, and emission monitoring increment are linked and marked in the same data record to form a complete historical operating condition sample. By performing the above processing on all stable operation periods, a historical operating condition sample set containing a large number of samples is formed. The operating condition database is composed of these sample records with time index and operating condition feature tags.

[0044] For each stable operating time window in the historical operating condition sample set, the average fuel consumption rate of that time window is first calculated based on the fuel consumption increment and the time window length. Then, the equivalent fuel consumption rate is calculated based on the average active power output. The equivalent fuel consumption rate can be defined as the ratio of the fuel consumption rate to the average active power output within that time window, used to describe the fuel consumption level corresponding to a unit of active power output. Let the index of a certain operating condition sample be i, and its time window length be denoted as . The increase in fuel consumption is recorded as The average active power output is denoted as Then the fuel consumption rate can be expressed as The equivalent fuel consumption rate can be expressed as:

[0045] in, This represents the equivalent fuel consumption rate for the i-th operating condition sample. This indicates the increase in fuel consumption within the corresponding time window for this operating condition sample. Indicates the length of the time window. This represents the average active power output corresponding to the sample under this operating condition. A higher equivalent fuel consumption rate indicates greater fuel consumption under the same active power output conditions. Within the same sample under the same operating condition, the coal mill combination state is encoded as a discrete feature vector, and the secondary air distribution method is encoded as a continuous feature vector. For example, the coal mill combination state vector is composed of the start-stop state of each coal mill, and the air distribution feature vector is composed of the proportion of secondary air volume of each burner layer and each sidewall to the total secondary air volume. For all samples under the same operating condition, the operating condition features are... With equivalent fuel consumption rate In connection, the unit coal consumption cost function is described by defining a function with undetermined parameters. The unit coal consumption cost function is as follows:

[0046] in, This represents the coal consumption cost when the active power output is P, the coal mill combination state is M, and the secondary air distribution method is W. This represents the set of parameters to be estimated in the function. This represents a family of functions consisting of polynomial expansions, basis function combinations, or other nonlinear structures. By selecting all operating condition samples, treating the equivalent fuel consumption rate as the target output and the operating condition characteristics as the input, a custom loss function is defined, such as the sum of the squares of the differences between the predicted values ​​and the actual equivalent fuel consumption rates of all samples, and the parameter set is adjusted accordingly. Performing a minimization operation involves finding the set of parameters that minimizes the loss function using parameter optimization methods such as least squares or least squares with regularization. Thus, a definite unit coal consumption cost function is obtained. The unit's coal consumption cost function can predict changes in fuel consumption given active power output, coal mill combination status, and secondary air distribution method, and provide a quantitative coal consumption cost for subsequent scheduling optimization.

[0047] For each stable operating time window within the historical operating condition sample set, the emission concentrations of nitrogen oxides, carbon monoxide, and particulate matter recorded by the emission monitoring system are used. The average values ​​of each emission concentration are calculated within that time window. Correspondences are established between these average emission concentrations and their corresponding average active power output, coal mill combination status, and secondary air distribution method. To describe the overall emission level, the emission concentrations can be linearly or non-linearly combined into a comprehensive emission index for each operating condition sample based on the regulatory weights or environmental costs of different emission factors. For example, let the average nitrogen oxide emission concentration of the i-th operating condition sample be... The average concentration of carbon monoxide emissions was The average concentration of smoke and dust emissions was The comprehensive emission index can be defined as:

[0048] in, This represents the comprehensive emission index for the i-th operating condition sample. , , This represents the weighting coefficients of nitrogen oxides, carbon monoxide, and particulate matter in environmental assessments or economic penalties; when constructing the emission cost function, it incorporates comprehensive emission indicators. As the target output, the average active power output, coal mill combination status, and secondary air distribution method are considered as input features. The parameterized form of the emission cost function is set in a manner similar to that of the unit coal consumption cost function:

[0049] in, This represents the emission cost when the active power output is P, the coal mill combination state is M, and the secondary air distribution method is W. This represents the set of undetermined parameters in the emission cost function. This represents a family of functions that can be composed of polynomials, piecewise functions, or other nonlinear models; it is the set of parameters that minimizes the error between the predicted value of the emission cost function and the comprehensive emission index by solving over all historical operating condition samples. Thus, a definite emission cost function is obtained. This enables the emission cost function to predict emission changes based on active power output, coal mill combination status, and secondary air distribution method, and to continue tracking the predicted values ​​of each individual emission factor when needed, in order to refine the constraints on different emission types in scheduling optimization.

[0050] During real-time operation, the active power output setpoint, pulverizer combination, and secondary air distribution method generated by the scheduling optimization module in each rolling optimization cycle are used as input variables. These are then sequentially input into the unit's coal consumption cost function and emission cost function to predict the fuel consumption cost and emission cost of each candidate operating condition in the future time domain. Specifically, after the rolling optimization time domain is discretized into several future scheduling periods, the scheduling optimization module generates multiple candidate schemes for each future scheduling period. Each candidate scheme sets a set of active power output setpoints, pulverizer combinations, and secondary air distribution methods. These active power output setpoints, pulverizer combinations, and secondary air distribution methods of the candidate schemes are then substituted into the unit's coal consumption cost function. With emission cost function The fuel consumption cost and emission cost of the candidate operating condition during the scheduling period are obtained. By accumulating or weighting the fuel consumption cost and emission cost of the candidate operating conditions for each period in the entire future time domain, these values ​​are written into the scheduling optimization objective function. This allows the scheduling optimization objective function to consider not only the current moment but also the changes in fuel consumption and emissions corresponding to different decision paths over a future period when searching for the optimal solution. In this way, scheduling control quantifies the changes in fuel consumption and emissions in the objective function, enabling automatic generation scheduling to make forward-looking predictions and trade-offs on changes in fuel consumption and emissions under different operating conditions based on the unit coal consumption cost function and emission cost function, while satisfying grid security constraints and acoustic-fuel coupling constraints. This achieves synergistic optimization of economic efficiency and environmental protection.

[0051] S140 divides the acoustic-fuel coupling stability domain based on the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations, and different secondary air distribution methods. The operating point that induces acoustic-fuel coupling instability is set as the prohibited domain, and the operating point at the boundary of acoustic-fuel coupling instability is set as the risk domain.

[0052] In one possible implementation, the acoustic-fuel coupling stability domain is divided based on the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations, and different secondary air distribution methods. Operating points that induce acoustic-fuel coupling instability are set as prohibited domains, and operating points at the boundary of acoustic-fuel coupling instability are set as risk domains. Specifically, this includes: constructing an acoustic-fuel coupling sample set covering different active power outputs, different coal mill combinations, and different secondary air distribution methods based on real-time calculation results of the acoustic-fuel coupling stability index; recording the active power output, coal mill combination, secondary air distribution method, and corresponding acoustic-fuel coupling stability index time series at each operating point; collecting all acoustic-fuel coupling stability index samples within a sliding time window for each combined operating condition determined by the active power output range, coal mill combination code, and secondary air distribution mode; and based on the acoustic-fuel coupling stability index... The stability index samples are calculated to obtain the sample mean and sample standard deviation. An effective acoustic-fuel coupling stability assessment value is constructed based on the sample mean and sample standard deviation. The effective acoustic-fuel coupling stability assessment value is compared with a preset stability upper threshold and a risk upper threshold. When the effective acoustic-fuel coupling stability assessment value corresponds to a value below the stability upper threshold, the combined operating condition is marked as a stable operating condition point. When the effective acoustic-fuel coupling stability assessment value is between the stability upper threshold and the risk upper threshold, the combined operating condition is marked as a risky operating condition point. When the effective acoustic-fuel coupling stability assessment value exceeds the risk upper threshold, the combined operating condition is marked as a prohibited operating condition point. The stable operating condition point, risky operating condition point, and prohibited operating condition point are mapped to a multi-dimensional operating condition space to form an acoustic-fuel coupling stability domain. The multi-dimensional operating condition space uses active power output, coal mill combination, and secondary air distribution mode as independent variables.

[0053] Specifically, based on the real-time calculation results of the acoustic-combustion coupling stability index, when organizing the records of each operation and test, the active power output, coal mill combination, and secondary air distribution method at each moment are first correlated with the corresponding acoustic-combustion coupling stability index value at that moment according to the timestamp. The time axis is divided into several discrete moments, and a working point record is formed at each discrete moment. For each working point, the active power output value of the unit at that moment is recorded, the start-up and shutdown status of each coal mill is recorded and encoded as the current coal mill combination, and the primary air volume and secondary air volume at each burner layer are recorded. The distribution ratio between each side wall is summarized as the current secondary air distribution mode. At the same time, the acoustic-fuel coupling stability index values ​​calculated by sliding time window around the operating point over a period of time are arranged into an acoustic-fuel coupling stability index time series in chronological order. Finally, by aggregating all operating point records, an acoustic-fuel coupling sample set covering different active power outputs, different coal mill combinations, and different secondary air distribution modes is formed. Each element in the acoustic-fuel coupling sample set contains operating condition characteristics and an acoustic-fuel coupling stability index time series, thus providing basic data for subsequent statistical analysis of acoustic-fuel coupling stability by combined operating conditions.

[0054] When statistically analyzing each combined operating condition determined by the active power output range, the coal mill combination code, and the secondary air distribution mode, the continuous active power output values ​​are first segmented according to the pre-divided active power output range. The coal mill combination is encoded using a fixed-length state vector. A large number of instantaneous secondary air distribution methods are summarized into a limited number of secondary air distribution modes through pattern clustering or range segmentation. A combined operating condition is identified by a triple of "active power output range + coal mill combination code + secondary air distribution mode". After a combined operating condition is identified, the acoustic-fuel coupling sample set is traversed to select all operating points within the effective range of the combined operating condition. That is, operating points where the active power output falls within the corresponding active power output range, the coal mill combination code is completely consistent with the combined operating condition, and the secondary air distribution mode is classified as the secondary air distribution mode of the combined operating condition. The acoustic-fuel coupling stability index samples within all sliding time windows corresponding to these operating points are collected. These samples are regarded as the acoustic-fuel coupling stability performance of the combined operating condition in different time windows for subsequent statistical analysis.

[0055] When statistically analyzing the sample set of acoustic-combustion coupling stability indices corresponding to a certain combination of operating conditions, we first assume that the combination of operating conditions has a total of [number missing]. A number of acoustic-combustion coupling stability index samples are denoted as follows: The sample average value of the acoustic-fuel coupling stability index for this combined operating condition is calculated to characterize the long-term average level. Its expression is:

[0056] in, This represents the sample average value of the acoustic-fuel coupling stability index for combined operating condition c. This indicates the number of samples for the acoustic-fuel coupling stability index corresponding to this combined operating condition. This represents the acoustic-fuel coupling stability index calculated under the i-th sliding time window in this combined operating condition. After obtaining the sample mean, to measure the degree of fluctuation of the acoustic-fuel coupling stability index around the mean, the sample standard deviation is calculated. The expression for the sample standard deviation is:

[0057] in, This represents the sample standard deviation of the acoustic-fuel coupling stability index for combined operating condition c. and Same as before, denominator This is used to obtain a more reasonable fluctuation estimate when the sample size is limited. The larger the sample mean, the higher the long-term acoustic-fuel coupling risk of the combined operating conditions. The larger the sample standard deviation, the more severe the acoustic-fuel coupling fluctuation of the combined operating conditions.

[0058] When constructing an effective acoustic-fuel coupling stability assessment value based on the sample mean and sample standard deviation, in order to integrate the long-term average level and fluctuation intensity into a single index, an effective acoustic-fuel coupling stability assessment value is defined for each combined operating condition. The sample mean and sample standard deviation are added in a weighted manner, expressed as follows:

[0059] in, This represents the effective acoustic-fuel coupling stability assessment value corresponding to combined operating condition c. This represents the sample average value of the acoustic-fuel coupling stability index for combined operating condition c. This represents the sample standard deviation of the acoustic-fuel coupling stability index for combined operating condition c. This represents a non-negative weighting coefficient used to adjust the influence of the volatility term in the evaluation value. When the value is large, even if the average value of the combined working condition with large fluctuations is not very high, its effective acoustic-fuel coupling stability assessment value will be significantly increased. Through this linear combination method, the effective acoustic-fuel coupling stability assessment value reflects both the overall level of the acoustic-fuel coupling stability index of the combined working condition over a long period of time and the risk of strong oscillations that may occur in the combined working condition in a short period of time.

[0060] When comparing the effective acoustic-fuel coupling stability assessment value based on the preset stability upper threshold and risk upper threshold, two ordered threshold parameters are first set according to the unit's operating experience, safety procedures, and sensitivity to acoustic-fuel coupling instability. These are the stability upper threshold and the risk upper threshold. and risk upper limit threshold and satisfy For each combined operating condition, the effective acoustic-fuel coupling stability assessment value is calculated. ,if Less than or equal to the stability upper limit threshold If the combined operating condition is considered to be within the acceptable range in terms of both long-term average level and fluctuation intensity, and the acoustic-combustion coupling state of the combined operating condition is consistent with or close to the normal combustion state, then the combined operating condition is marked as a stable operating condition point. The marking result of the stable operating condition point will be used in subsequent scheduling optimization to indicate that the operating condition can be selected as a priority area without additional acoustic-combustion coupling penalty.

[0061] When the effective acoustic-fuel coupling stability assessment value is between the stability upper limit threshold and the risk upper limit threshold, it indicates that the combined operating condition has exceeded the safety margin of the stable operating point in terms of the long-term average level or fluctuation intensity of the acoustic-fuel coupling stability index, but has not yet reached the severity level that significantly induces acoustic-fuel coupling instability. At this time, the combined operating condition is marked as a risk operating point. A risk operating point indicates that although the unit may still be able to withstand the combined operating condition in the short term, there is a significant risk of acoustic-fuel coupling instability if it remains in this condition for a long time or if there is further load disturbance. When the effective acoustic-fuel coupling stability assessment value exceeds the risk upper limit threshold, the risk is considered to be at risk. If the combined operating condition is marked as a prohibited operating condition, it indicates that the combined operating condition has shown significant high-risk characteristics in the historical samples. Both the average value and the degree of fluctuation have exceeded the acceptable range. Therefore, the combined operating condition is marked as a prohibited operating condition point. The prohibited operating condition point means that the unit should be explicitly prohibited from operating under this combined operating condition in the scheduling optimization model, because it is very easy to trigger furnace acoustic resonance or serious combustion instability under this combined operating condition.

[0062] When mapping stable operating points, risky operating points, and prohibited operating points to a multi-dimensional operating space and forming an acoustic-fuel coupling stability domain, the multi-dimensional operating space is constructed using active power output, coal mill combination, and secondary air distribution mode as independent variables. The active power output axis is discretized according to the active power output range, the coal mill combination axis is arranged according to the codes of all possible coal mill combinations, and the secondary air distribution mode axis is divided according to the secondary air distribution mode, so that each grid cell or each discrete point in the multi-dimensional operating space corresponds to a specific combination operating condition. After completing the classification of combination operating conditions, stable operating points are marked as stable regions in the multi-dimensional operating space, risky operating points are marked as risk regions in the multi-dimensional operating space, and prohibited operating conditions are marked as... Points are marked as prohibited regions in the multidimensional operating space. Through interpolation, neighborhood expansion, or clustering, the region classification is smoothly expanded in areas not directly covered by samples, so that the acoustic-fuel coupling stability domain presents a distribution pattern of stable, risk, and prohibited regions with continuous boundaries in the multidimensional operating space. The acoustic-fuel coupling stability domain transforms discrete statistical results into queryable operating space partitions through this mapping process. This enables the scheduling optimization model to quickly find the stable, risk, or prohibited region to which the corresponding point belongs in the multidimensional operating space when given any set of active power output, coal mill combination, and secondary air distribution mode. Thus, explicit control of the acoustic-fuel coupling instability risk is achieved in the scheduling optimization constraints.

[0063] S150, when the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain, the current scheduling scenario is identified as an acoustic-fuel coupling constraint-dominated scenario.

[0064] In one possible implementation, when the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain, the current scheduling scenario is identified as an acoustic-fuel coupling constraint-dominated scenario. Specifically, this includes: using a sliding time window to perform monotonicity and growth rate analysis on multiple consecutive sampling points of the time series; when the acoustic-fuel coupling stability index of multiple consecutive sampling points is greater than or equal to the acoustic-fuel coupling stability index of the previous sampling point, and the increment of the acoustic-fuel coupling stability index of the current sampling point relative to the acoustic-fuel coupling stability index of the starting sampling point of the local time window exceeds a preset increase threshold, it is determined that the time series has a continuous upward trend; based on the current active power output setpoint, the current coal mill combination, and the current secondary air distribution method, the current operating point is mapped to the acoustic-fuel coupling stability domain, and it is determined that the current operating point belongs to the stable domain, the risk domain, or the boundary of the prohibited domain; when the time series has a continuous upward trend and the current operating point belongs to the risk domain or the boundary of the prohibited domain, the current scheduling scenario is marked as an acoustic-fuel coupling constraint-dominated scenario.

[0065] Specifically, when performing monotonicity and growth rate analysis on the time series of acoustic-fuel coupling stability index using a sliding time window, a local time window of fixed length is first set for the time series of acoustic-fuel coupling stability index. This local time window slides point by point on the time axis, with each slide including the most recent consecutive sampling points. Within any local time window, starting from the earliest sampling point, the acoustic-fuel coupling stability index values ​​of adjacent sampling points are compared point by point. When the acoustic-fuel coupling stability index of each subsequent sampling point is greater than or equal to the acoustic-fuel coupling stability index of its predecessor sampling point, the change within this local time window is determined to be monotonically non-decreasing. At the same time, the acoustic-fuel coupling stability index of the current sampling point and the acoustic-fuel coupling stability index of the starting sampling point of this local time window are calculated. The difference between indicators is compared with a pre-set rise threshold. When the difference exceeds the pre-set rise threshold, the local time window is considered to not only satisfy monotonicity without decline, but also to have a significant overall rise. A sliding time window refers to a local interval with a fixed length on the time axis and whose start and end times move with the current sampling point. Monotonicity analysis determines whether the time series continues to rise or at least does not decline within the local interval. Growth rate analysis determines whether the numerical difference between the first and last sampling points is large enough. By simultaneously satisfying the two conditions of "multiple consecutive sampling points monotonically not declining" and "the increment of the first and last points exceeds the rise threshold", the time segment corresponding to the local time window is determined to have a continuous upward trend, thereby avoiding misjudgment caused by short-term noise or a single outlier.

[0066] When mapping the current operating point to the acoustic-combustion coupling stability domain based on the current active power output setpoint, the current coal mill combination, and the current secondary air distribution method, the following steps are first taken: First, the active power output setpoint, the current coal mill start / stop status, and the current secondary air distribution method parameters output by the scheduling optimization module within the current scheduling cycle are read. The active power output setpoint is mapped to active power output coordinates in the multi-dimensional operating space. The current coal mill start / stop status is encoded as a coal mill combination code consistent with the acoustic-combustion coupling stability domain. The current secondary air distribution method is converted into an airflow distribution mode on different burner layers and different sidewalls. Then, the position of the current operating point in the multi-dimensional operating space is determined based on these three dimensions. The acoustic-combustion coupling stability domain is based on the active power output setpoint. The multi-dimensional operating condition partitioning structure is constructed with output, coal mill combination, and secondary air distribution mode as independent variables. Each position is pre-marked as part of a stable domain, risk domain, or prohibited domain. By looking up tables or interpolating, it can be determined whether the current operating point is in the stable domain, risk domain, or inside or near the boundary of the prohibited domain. Operating points at the boundary of the prohibited domain refer to those that are geometrically close to the inside of the prohibited domain but have not yet completely fallen into the prohibited domain. They are usually identified by the distance threshold between the prohibited domain representative point or the prohibited domain envelope. In historical statistics, such operating points often show a highly sensitive state of near acoustic-fuel coupling instability. Therefore, they need to be included in the key attention scope along with the risk domain when identifying scheduling scenarios.

[0067] When the time series of the acoustic-fuel coupling stability index is determined to have a continuous upward trend within the current local time window and the current operating point is located at the boundary of the risk domain or the prohibited domain, the current scheduling scenario is marked as an acoustic-fuel coupling constraint-dominated scenario. First, within the current scheduling cycle, the two judgment results are combined. On the one hand, the monotonicity and growth rate analysis of the sliding time window provides a Boolean judgment on "whether there is a continuous upward trend." On the other hand, the acoustic-fuel coupling stability domain mapping provides a classification result for "the region to which the current operating point belongs." When the continuous upward trend is determined to be true and the operating condition classification result is either at the boundary of the risk domain or the prohibited domain, it indicates that the acoustic-fuel coupling stability index has not only been continuously rising in the recent period, but also that the corresponding operating condition... The system is already in a high-risk area near the boundary of acoustic-fuel coupling instability. Under this condition, the current scheduling operation status is marked as an acoustic-fuel coupling constraint-dominated scenario. An acoustic-fuel coupling constraint-dominated scenario means that in subsequent scheduling optimization and control execution, the priority of acoustic-fuel coupling-related constraints and acoustic-fuel coupling stability risk suppression requirements is raised to the same or even higher level as economic and grid-side requirements. The acoustic-fuel coupling penalty weight in the scheduling optimization objective function will be increased, and the region in the feasible region that is close to the prohibited region will be significantly compressed or avoided, so that the scheduling control prioritizes driving the unit operating conditions away from the acoustically sensitive region. Thus, in the stage of rapid increase in acoustic-fuel coupling instability risk, the scheduling mode is switched from economic-dominated to acoustic-fuel coupling constraint-dominated.

[0068] S160 reconstructs the scheduling optimization model in a scenario dominated by acoustic-fuel coupling constraints. It takes the unit's active power output setpoint, the coal mill's start-up and shutdown status, and the allocation of primary and secondary air volumes as decision variables. It constructs optimization objectives based on the unit's coal consumption cost function and emission cost function, and uses the unit's physical limits, grid-side constraints, and acoustic-fuel coupling stability domain as optimization constraints.

[0069] In one possible implementation, the scheduling optimization model is reconstructed in a scenario dominated by acoustic-fuel coupling constraints. The unit's active power output setpoint, the pulverizer's on / off status, and the allocation of primary and secondary air volumes are used as decision variables. An optimization objective is constructed based on the unit's coal consumption cost function and emission cost function. Simultaneously, the unit's physical limits, grid-side constraints, and the acoustic-fuel coupling stability domain are used as optimization constraints. Specifically, this includes: discretizing the rolling optimization time domain according to a uniform time length, and using the unit's active power output setpoint, pulverizer's on / off status, and the allocation of primary and secondary air volumes in each discrete time period as decision variables to be optimized; applying the unit's coal consumption cost function and emission cost function to the decision variable setpoint for each scheduling period to calculate the fuel consumption cost and the comprehensive emission cost, and then... The deviation of the active power output setpoint from the active power output command of the automatic generation control constitutes the tracking cost of the automatic generation control. The cost of the acoustic-fuel coupling stability penalty constructed based on the acoustic-fuel coupling stability index is used as the cost term in the optimization objective. The minimum stable output of the unit, the maximum rated output of the unit, the unit ramp rate, the minimum continuous running time and minimum downtime of the coal mill, the upper and lower limits of the total primary air volume and the total secondary air volume, and the upper and lower limits of the secondary air volume allocation ratio are used as physical limit constraints of the unit. The grid active power balance, reserve capacity requirements and frequency regulation assessment constraints are used as grid-side constraints. The operating condition combination corresponding to the prohibited operating condition point in the acoustic-fuel coupling stability domain is used as the prohibited value, and the time continuous residence of the operating condition combination corresponding to the risk operating condition point is restricted as the acoustic-fuel coupling stability domain constraint.

[0070] Specifically, when the rolling optimization time domain is discretized according to a uniform time length, firstly, after the dominant acoustic-fuel coupling constraint scenario is identified, a prediction duration looking forward from the current scheduling time is selected. This prediction duration is then divided into several scheduling periods of equal length, forming a series of discrete time period indices arranged in chronological order. Each discrete time period corresponds to the unit's operational decision within a future time interval. In each discrete time period, the unit's active power output setpoint is used as a continuous decision variable to describe the unit's planned active power output level to the grid within that time period, while the coal mill's on / off status is used as a discrete decision variable to represent the time period. The start-up and shutdown combination of each coal mill uses the allocation of primary and secondary air volumes as continuous decision variables to represent the total primary air volume, total secondary air volume, and the distribution ratio of secondary air among each burner layer and each sidewall within a given time period. By simultaneously configuring three categories of decision variables—unit active power output setpoint, coal mill start-up and shutdown status, and primary and secondary air volume allocation—at each discrete time period, a joint decision vector is formed that describes the unit's electrical side output trajectory and combustion side coal and air distribution trajectory in the future time domain. This achieves unified planning of the unit's active power output setpoint path, coal mill start-up and shutdown path, and air distribution path in the time dimension.

[0071] When applying the unit coal consumption cost function and emission cost function to the decision variable set for each scheduling period, at each discrete time period index k, the unit active power output setpoint, pulverizer start / stop status, and primary and secondary air volume allocation for that period are substituted into the unit coal consumption cost function to obtain the fuel consumption cost for that period. The same set of decision variables is then substituted into the emission cost function to obtain the comprehensive emission cost for that period. Simultaneously, the automatic generation control tracking cost is constructed based on the deviation between the unit active power output setpoint and the automatic generation control active power output command for that period. For example... The degree of deviation from the dispatch command can be amplified by the square or absolute value of the deviation. Then, a penalty cost for acoustic-fuel coupling stability can be constructed based on the magnitude of the acoustic-fuel coupling stability index for that period. By assigning different amplification coefficients to the acoustic-fuel coupling stability index in different intervals, conditions approaching the acoustic-fuel coupling instability boundary or entering the risk domain can obtain higher costs in the objective function. Over the entire rolling optimization time domain, the weighted sum of the fuel consumption cost, comprehensive emission cost, automatic generation control tracking cost, and acoustic-fuel coupling stability penalty cost for all dispatch periods is used as the optimization objective. The objective function can be expressed as:

[0072] Where J represents the total optimization cost within the rolling optimization time domain, T represents the number of discrete scheduling periods within the rolling optimization time domain, and k represents the index of the currently considered scheduling period. , , , These represent the weight coefficients of the unit's coal consumption cost, emission cost, acoustic-fuel coupling stability penalty cost, and automatic generation control tracking cost in the objective function, respectively. This represents the active power output setpoint of the generating unit during the k-th scheduling period. This indicates the start / stop status of the coal mill during the k-th scheduling period. This represents the set of primary and secondary air volume allocation parameters for the k-th scheduling period. This represents the active power output command for automatic generation control during the k-th scheduling period. This represents the acoustic-fuel coupling stability index corresponding to the k-th scheduling period. This represents a function that predicts changes in fuel consumption under current operating conditions based on the unit's coal consumption cost function. This represents a function that predicts changes in emissions based on the emission cost function under current operating conditions. This represents a function that maps the acoustic-fuel coupling stability index to the acoustic-fuel coupling stability penalty cost value. This represents a function that maps the deviation of the unit's active power output setpoint from the active power output command of the automatic generation control to the tracking cost value of the automatic generation control. By minimizing the value of the objective function J, the scheduling optimization can provide the optimal decision sequence for the unit's active power output setpoint, the coal mill's start-up and shutdown status, and the allocation of primary and secondary air volumes, taking into account fuel consumption, emission levels, the tracking performance of the automatic generation control, and the stability risks of acoustic-fuel coupling.

[0073] When using unit physical limits, grid-side constraints, and acoustic-fuel coupling stability domain as optimization constraints, regarding unit physical limits, the minimum stable output and maximum rated output of the unit are applied to the active power output setpoint for each scheduling period, forming upper and lower limit constraints. Simultaneously, the allowable forward and reverse ramp rates limit the variation in active power output setpoint between adjacent scheduling periods, thereby ensuring that the variation in unit active power output does not exceed the allowable range of equipment mechanics and thermal inertia. Furthermore, the maximum stable output of the coal mill is applied to the coal mill during both start-up and shutdown states. The constraints of short continuous operating time and minimum downtime ensure that any coal mill, once put into operation in a certain period, will run continuously for at least a certain number of periods, and once shut down in a certain period, will remain shut down for at least a certain number of periods, thus avoiding frequent start-ups and shutdowns of the coal mill. Upper and lower limits are imposed on the total primary air volume and the total secondary air volume to ensure that the fans operate within a safe and effective range. At the same time, upper and lower limits are set on the secondary air volume distribution ratio, and the sum of these limits is also constrained to ensure that the air volume distribution of each burner layer and each side wall is neither too biased to one side nor violates the conservation of total air volume.

[0074] Regarding grid-side constraints, the active power balance conditions of the grid in the region where the generating unit is located during each dispatch period are used as constraints. The active power output setpoint of the generating unit, the planned output of other generating units in the same region, and the load demand are all incorporated into the active power balance equation, so that the power output matches the sum of the grid load and grid losses in each dispatch period. The grid reserve capacity requirement is used as a constraint, and the upper limit of the active power output setpoint of the generating unit is linked to the reserve capacity, so that the generating unit can meet the output requirements of the dispatch command while retaining the necessary upward or downward adjustment space to meet the needs of frequency control and emergency backup. Frequency regulation assessment constraints are used as an additional condition, linking the unit's ability to participate in frequency regulation with the tracking error of automatic generation control. In the objective function, a cost is imposed on the behavior that deviates from the active power output command of automatic generation control, and the cumulative degree of long-term deviation is limited in the constraints, so that the generating unit can still meet the grid-side assessment requirements for frequency regulation performance while the acoustic-fuel coupling constraint is dominant.

[0075] Regarding the stability domain constraint of acoustic-fuel coupling, the stable operating points, risky operating points, and prohibited operating points obtained through previous statistics on acoustic-fuel coupling stability indices are mapped to a multi-dimensional operating condition space with active power output, coal mill combination status, and secondary air distribution mode as independent variables. The decision variable combination for each scheduling period is then considered. All of these can be found in the multi-dimensional operating condition space. Hard constraints are imposed on decision combinations in the prohibited operating condition set, that is, these decision combinations are prohibited from being selected during any scheduling period, so that the unit will not run into the operating condition that induces acoustic-fuel coupling instability in the optimization result. For decision combinations in the risky operating condition set, time constraints are constructed by limiting the continuous residence time. For example, the number of scheduling periods in which the unit appears consecutively in the risky operating condition point or its neighborhood cannot exceed a preset upper limit. This allows the scheduling optimization to briefly cross the risk domain when necessary, but does not allow the unit to stay near the boundary of the risk domain for a long time. In the objective function, the cost corresponding to the risky operating condition point is further amplified by the acoustic-fuel coupling stability penalty value. Under the premise of satisfying the physical limits of the unit and the constraints of the grid side, the scheduling optimization tends to push the active power output setpoint of the unit, the start and stop status of the coal mill, and the allocation of primary air volume and secondary air volume to the stable region in the acoustic-fuel coupling stability domain. This achieves active avoidance of acoustically sensitive operating conditions in the scenario dominated by acoustic-fuel coupling constraints.

[0076] S170 sends the optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectory obtained from the scheduling optimization model to the coordination control layer, so that the thermal power unit can avoid acoustically sensitive conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases.

[0077] In one possible implementation, the optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectories obtained from the scheduling optimization model are sent to the coordination control layer to adjust the operation of the boiler combustion system and turbine system. This allows the thermal power unit to avoid acoustically sensitive conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases. Specifically, this includes: smoothing and checking the consistency of the optimal active power output trajectory, optimal coal mill combination trajectory, and optimal primary and secondary air distribution trajectories to ensure that the changes in active power output between adjacent time steps meet the unit ramp-up rate limit and that the changes in primary and secondary air distribution meet the correspondence between the coal mill start-up and shutdown states; sending the processed optimal active power output trajectory, optimal coal mill combination trajectory, and optimal primary and secondary air distribution trajectories to the coordination control layer, and in the coordination... In the control layer, the optimal active power output trajectory is decomposed into the boiler main setpoint, the turbine control valve opening setpoint, and the generator excitation voltage setpoint; the optimal coal mill combination trajectory is converted into start-stop commands and load increase / decrease rate settings for each coal mill; the optimal primary and secondary air distribution trajectories are converted into primary and secondary air fan outputs and burner layer and side wall damper settings; during the execution of the trajectory distribution by the coordination control layer, the optimal primary and secondary air distribution trajectories are corrected by feedforward and feedback based on real-time furnace negative pressure, oxygen content, flue gas temperature, and acoustic-fuel coupling stability indicators; and when the acoustic-fuel coupling stability indicators do not decrease as expected, the optimal active power output setting and the optimal coal mill combination are adjusted so that the boiler combustion system and turbine system avoid acoustically sensitive conditions and maintain dispatch control under the condition of increased acoustic-fuel coupling instability risk.

[0078] Specifically, when performing smoothing and consistency checks on the optimal active power output trajectory, the optimal coal mill combination trajectory, and the optimal primary and secondary air distribution trajectory, the optimal active power output setpoint corresponding to each time step is first read sequentially on the time axis according to the discrete time steps of the scheduling optimization. The active power output increment between adjacent time steps is calculated, and this increment is compared with the maximum output change corresponding to the unit's allowed forward ramp rate and reverse ramp rate. When it is found that the active power output increment exceeds the allowable range, the originally abrupt output change is decomposed into multiple gradual step sizes that meet the ramp rate limit by inserting transition time steps on the time axis or interpolating and stretching the output values ​​of multiple adjacent time steps, thereby obtaining a smooth active power output trajectory that can be executed within the unit's physical constraints; for the optimal coal mill When checking the consistency between the combined trajectory and the optimal primary and secondary air distribution trajectory, the start-up and shutdown status of the coal mill at each time step is used as the benchmark. It is verified whether there are coal or air distribution settings for the stopped coal mill within the same time step. It is also verified whether, during the critical start-up and shutdown period of the coal mill, settings for the primary and secondary air volumes that are inconsistent with the high-load operation of the coal mill are made before the coal mill has built up pressure or fully unloaded. When inconsistencies occur, the corresponding air volume adjustment time is delayed or advanced to ensure that the changes in primary and secondary air distribution meet the correspondence of the start-up and shutdown status of the coal mill. If necessary, the start-up and shutdown plans of individual coal mills at individual time steps are fine-tuned to form the optimal coal mill combined trajectory and the optimal primary and secondary air distribution trajectory that are physically feasible and consistent with the fuel supply path.

[0079] When the optimal active power output trajectory, optimal coal mill combination trajectory, and optimal primary and secondary air distribution trajectory, after smoothing and consistency checks, are sent to the coordination control layer, the trajectory data for the entire time period are packaged into a set of reference operating plans according to time sequence through the predefined communication interface of the unit control system. Within the coordination control layer, the optimal active power output trajectory is decomposed into three types of internal setpoints according to the functional decomposition principle: boiler main setpoint, turbine valve opening setpoint, and generator excitation voltage setpoint. The boiler main setpoint is used to determine the target evaporation rate or main steam flow rate of the boiler at each time step, ensuring that the energy input on the combustion side is consistent with the target evaporation rate or main steam flow rate. The power output demand is matched; the turbine control valve opening is given to determine the target value of the turbine inlet steam flow at each time step to ensure that the turbine mechanical output is consistent with the active power output on the electric side; the generator excitation voltage is given to adjust the excitation system output according to the changes in active power output and the grid voltage requirements, so that the generator terminal voltage and reactive power output are maintained within the allowable range; through this decomposition process, the coordination control layer maps the "active power output" abstract quantity of the scheduling layer into control objectives that can be directly executed by the three subsystems of boiler, turbine and generator, so that each subsystem can clearly understand its specific adjustment task under the optimal active power output trajectory at the execution layer.

[0080] When the optimal coal mill combination trajectory is converted into start / stop commands and load ramp-up / queue settings for each coal mill, the coordination control layer reads the coal mill combination status at each time step, identifies the changes in the start / stop status of each coal mill between adjacent time steps, generates a "start command" for coal mills that change from stopped to running, and sets the start preparation time and acceleration time based on the coal mill's mechanical characteristics, lubricating oil temperature, and sealing air establishment; generates a "stop command" for coal mills that change from running to stopped, and sets the unloading time and deceleration time based on the residual coal powder and grinding roller inertia in the coal mill; during continuous operation, based on the load distribution intention implicit in the optimal coal mill combination trajectory, load ramp-up / queue settings are set for each running coal mill to avoid excessively rapid load changes in a single coal mill, which could lead to excessively rich or lean fuel in the corresponding burner area; the above start / stop commands and load ramp-up / queue settings are executed through the coal mill body controller, realizing the orderly input, smooth unloading, and coordinated load sharing of coal mills on the time axis, so that the operating status of the coal mill group is consistent with the upper-level optimal coal mill combination trajectory.

[0081] When converting the optimal primary and secondary air distribution trajectories into primary and secondary air fan outputs and burner layer and sidewall damper opening settings, the coordination control layer first extracts the target total primary air volume and target total secondary air volume for each time step from the trajectory. Then, based on the unit's designed excess air coefficient control logic, it verifies whether the ratio of these target air volumes to the corresponding fuel supply for each time step is reasonable. Once confirmed to be reasonable, the target total primary air volume is allocated to specific primary air fans, achieving the target output by setting the primary air fan speed or guide vane opening setpoint. For the total secondary air volume, it is determined based on the different burners given in the trajectory. The secondary air distribution ratio of the burner layer and different side walls decomposes the total air volume into the air volume target of each branch, and then converts these air volume targets into the opening set value of the corresponding damper and the operating point setting of the secondary air fan, so that each burner layer and each side wall receives air supply consistent with the optimal air distribution trajectory at each time step. During the execution, the coordination control layer monitors the fan operation status and the actual damper opening in real time, compares the actual air volume with the target air volume, and uses closed-loop regulation to reduce the deviation, ensuring that the output of the primary and secondary air fans and the opening of the dampers of the burner layer and side walls can stably follow the optimal primary and secondary air distribution trajectory.

[0082] Throughout the execution of the issued trajectory, the coordination and control layer continuously collects the latest measurements of furnace negative pressure, oxygen content, flue gas temperature, and acoustic-combustion coupling stability indicators. These measurements are compared with the expected trends implied in the issued trajectory to construct a feedforward and feedback correction logic: the feedforward part predicts the possible directions of change in furnace negative pressure, oxygen content, and flue gas temperature based on the air distribution trends at future time steps in the trajectory, making slight pre-adjustments to the primary and secondary air distribution trajectories to offset unfavorable trends in advance; the feedback part fine-tunes the air distribution settings at the current time step based on the magnitude of deviations from the expected values ​​in the real-time measured furnace negative pressure, oxygen content, flue gas temperature, and acoustic-combustion coupling stability indicators. When acoustic-combustion coupling stability is detected... If the combustion coupling stability index fails to decrease as expected or even continues to rise after adjusting the air distribution, the coordination control layer triggers a higher-level adjustment strategy. This strategy moderately slows down the rate of increase in active power output or switches some high-risk coal mill combinations to low-risk coal mill combinations. Under the premise of maintaining grid security and meeting dispatch instructions, priority is given to restoring the acoustic stability of the combustion system. Through the above-mentioned feedforward and feedback correction mechanism, the boiler combustion system and turbine system can dynamically deviate from acoustically sensitive conditions based on real-time feedback while executing the optimal trajectory under the condition of increased acoustic-fuel coupling instability risk. This effectively avoids acoustically sensitive conditions and maintains the overall dispatch control objective without failure.

[0083] This embodiment also discloses an automatic power generation dispatch and control device for thermal power units, referring to... Figure 2 The device includes an acquisition module 201, a processing module 202, and an output module 203. It is used to execute any of the above-described automatic power generation dispatch control methods for thermal power units, wherein: Acquisition module 201 is used to collect multi-source observation data from the electric side and the furnace combustion side; The processing module 202 is used to preprocess the multi-source observation data to extract the furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude and sound pressure fluctuation amplitude, and to construct an acoustic-flame coupling stability index. The processing module 202 is used to construct the unit's coal consumption cost function and emission cost function based on historical operating data, so as to predict the changes in fuel consumption and emissions of the unit under different active power outputs, different coal mill combinations and different secondary air distribution methods. Processing module 202 is used to divide the acoustic-fuel coupling stability domain according to the acoustic-fuel coupling stability index under different active power output, different coal mill combinations and different secondary air distribution methods, set the operating point that induces acoustic-fuel coupling instability as the prohibited domain, and set the operating point at the acoustic-fuel coupling instability boundary as the risk domain. Processing module 202 is used to identify the current scheduling scenario as a scenario dominated by acoustic-fuel coupling constraints when the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain. Processing module 202 is used to reconstruct the scheduling optimization model in the scenario dominated by acoustic-fuel coupling constraints. It takes the active power output setpoint of the unit, the start-stop status of the coal mill, and the allocation of primary air volume and secondary air volume as decision variables, and constructs the optimization objective based on the unit coal consumption cost function and the emission cost function. At the same time, it uses the unit physical limit, grid-side constraints and the acoustic-fuel coupling stability domain as optimization constraints. Output module 203 is used to send the optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectory obtained by solving the scheduling optimization model to the coordination control layer, so that the thermal power unit can avoid acoustically sensitive conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases.

[0084] It should be noted that the above embodiments of the apparatus are only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0085] This embodiment also discloses an electronic device, as shown in the reference. Figure 3 The electronic device may include: at least one processor 301, at least one communication bus 302, user interface 303, network interface 304, and at least one memory 305.

[0086] The communication bus 302 is used to enable communication between these components.

[0087] The user interface 303 may include a display screen and a camera. Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0088] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0089] The processor 301 may include one or more processing cores. The processor 301 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory 305, and by calling data stored in memory 305. Optionally, the processor 301 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 301 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 301 and may be implemented as a separate chip.

[0090] The memory 305 may include random access memory (RAM) or read-only memory. Optionally, the memory may include a non-transitory computer-readable storage medium. The memory 305 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 305 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), instructions for implementing the various method embodiments described above, etc.; the data storage area may store data involved in the various method embodiments described above, etc. Optionally, the memory 305 may also be at least one storage device located remotely from the aforementioned processor 301. As a computer storage medium, the memory 305 may include an operating system, a network communication module, a user interface 303 module, and an application program for an automatic power generation dispatch control method for thermal power units.

[0091] exist Figure 3In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 301 can be used to call the application program of an automatic power generation scheduling and control method for a thermal power unit stored in the memory 305. When executed by one or more processors 301, the electronic device executes one or more methods as described in the above embodiments.

[0092] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, as some steps can be performed in other orders or simultaneously according to the present invention. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0093] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0094] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between apparatuses or units may be electrical or other forms.

[0095] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0096] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0097] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory 305 and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned memory 305 includes various media capable of storing program code, such as a USB flash drive, external hard drive, magnetic disk, or optical disk.

[0098] The present invention also discloses a non-transitory computer-readable storage medium storing instructions. When executed by one or more processors 301, these instructions cause an electronic device to perform one or more methods as described in the above embodiments.

[0099] The above are merely exemplary embodiments of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truths. This invention is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. An automatic power generation dispatch and control method for thermal power units, characterized in that, The method includes: Collect multi-source observation data from both the electric side and the furnace combustion side; The multi-source observation data is preprocessed to extract the furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude, and sound pressure fluctuation amplitude, and an acoustic-fuel coupling stability index is constructed. Based on historical operating data, the unit's coal consumption cost function and emission cost function are constructed to predict the changes in fuel consumption and emissions of the unit under different active power outputs, different coal mill combinations, and different secondary air distribution methods. Based on the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations and different secondary air distribution methods, the acoustic-fuel coupling stability domain is divided. The operating point that induces acoustic-fuel coupling instability is set as the prohibited domain, and the operating point at the acoustic-fuel coupling instability boundary is set as the risk domain. When the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain, the current scheduling scenario is identified as an acoustic-fuel coupling constraint-dominated scenario. In the scenario dominated by acoustic-fuel coupling constraints, the scheduling optimization model is reconstructed, and the active power output setpoint of the unit, the start-stop status of the coal mill, and the allocation of primary air volume and secondary air volume are used as decision variables. The optimization objective is constructed based on the unit coal consumption cost function and the emission cost function, and the unit physical limit, grid-side constraints, and the acoustic-fuel coupling stability domain are used as optimization constraints. The optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectories obtained from solving the scheduling optimization model are sent down to the coordination control layer, so that the thermal power unit can avoid acoustically sensitive operating conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases.

2. The automatic power generation dispatch and control method for thermal power units according to claim 1, characterized in that, The preprocessing of the multi-source observation data to extract the furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude, and sound pressure fluctuation amplitude, and to construct an acoustic-combustion coupling stability index, specifically includes: The furnace negative pressure time series, sound pressure time series and flame video series are aligned with a unified time reference and resampled on the same time axis, so that the furnace negative pressure discrete sequence, sound pressure discrete sequence and flame image frame sequence correspond one-to-one at each discrete moment. The furnace negative pressure discrete sequence is subjected to detrending and bandpass filtering. The effective value of the bandpass filtered sequence within the sliding time window is used to obtain the furnace negative pressure fluctuation amplitude, which is used to characterize the intensity of the pressure fluctuation related to the acoustic mode frequency band of the furnace. The discrete sound pressure sequence is subjected to detrending and bandpass filtering. The bandpass filtered sequence is used to construct an analytical signal and extract the envelope amplitude. The sound pressure fluctuation amplitude, which characterizes the intensity of acoustic oscillation in the furnace, is obtained by calculating the average envelope amplitude within a sliding time window. The flame video sequence is subjected to grayscale conversion, noise reduction and brightness normalization processing, and the burner flame area is extracted. The pixel brightness of the flame area is weighted and summed to obtain the flame centroid coordinates. The offset of the flame centroid relative to the time average position is calculated within the sliding time window to obtain the flame centroid jitter amplitude, which is used to characterize the degree of flame overall swaying and shape drift. Based on the baseline amplitude and fluctuation range of the furnace negative pressure fluctuation amplitude, the flame centroid vibration amplitude and the sound pressure fluctuation amplitude obtained statistically under stable combustion conditions, the furnace negative pressure fluctuation amplitude, the sound pressure fluctuation amplitude and the flame centroid vibration amplitude are normalized. The normalized furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude, and sound pressure fluctuation amplitude are combined in a weighted and nonlinear amplification manner to form the acoustic-flame coupling stability index.

3. The automatic power generation dispatch and control method for thermal power units according to claim 1, characterized in that, The method of constructing unit coal consumption cost functions and emission cost functions based on historical operating data enables the prediction of fuel consumption and emission changes of the unit under different active power outputs, different coal mill combinations, and different secondary air distribution methods. Specifically, this includes: Establish a working condition database that includes active power output time series, cumulative fuel consumption records, coal mill start-up and shutdown status sequences, and primary air volume and secondary air volume distribution records. Associate and mark the average active power output, coal mill combination status, and secondary air distribution method in each stable operating period with the fuel consumption increment and emission monitoring increment in the corresponding period to form a historical working condition sample set. In the historical operating condition sample set, based on the fuel consumption increment and average active power output, coal mill combination status and secondary air distribution method within each stable operating time window, a functional relationship between the equivalent fuel consumption rate and operating condition characteristics is established. The unit coal consumption cost function is obtained through parameter optimization, and the unit coal consumption cost function predicts fuel consumption changes based on active power output, coal mill combination status and secondary air distribution method. In the historical operating condition sample set, a functional relationship between comprehensive emission indicators and operating condition characteristics is established based on the average emission concentration within each stable operating time window, the average active power output within that time window, the coal mill combination status, and the secondary air distribution method. The emission cost function is obtained through parameter optimization, and the emission cost function is used to predict emission changes based on the active power output, the coal mill combination status, and the secondary air distribution method. During real-time operation, the active power output setpoint, coal mill combination, and secondary air distribution mode output by the scheduling optimization are used as input variables and sequentially input into the unit coal consumption cost function and the emission cost function to obtain the fuel consumption cost and emission cost of each candidate operating condition in the future time domain. The changes in fuel consumption and emissions are quantitatively represented in the scheduling optimization objective function, so that the scheduling control can predict the changes in fuel consumption and emissions under different operating conditions based on the unit coal consumption cost function and the emission cost function.

4. The automatic power generation dispatch and control method for thermal power units according to claim 1, characterized in that, The acoustic-fuel coupling stability domain is divided based on the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations, and different secondary air distribution methods. The operating point that induces acoustic-fuel coupling instability is set as the prohibited domain, and the operating point at the boundary of acoustic-fuel coupling instability is set as the risk domain. Specifically, this includes: Based on the real-time calculation results of the acoustic-fuel coupling stability index, an acoustic-fuel coupling sample set covering different active power outputs, different coal mill combinations, and different secondary air distribution methods is constructed, and the active power output, coal mill combination, secondary air distribution method, and corresponding acoustic-fuel coupling stability index time series are recorded at each operating point. For each combined operating condition determined by the active power output range, coal mill combination coding and secondary air distribution mode, all acoustic-combustion coupling stability index samples were collected within the sliding time window. Calculate the sample mean and sample standard deviation based on the acoustic-combustion coupling stability index sample; An effective acoustic-fuel coupling stability assessment value is constructed based on the sample mean and the sample standard deviation. The effective acoustic-fuel coupling stability assessment value is compared with the preset stability upper limit threshold and risk upper limit threshold. When the effective acoustic-fuel coupling stability assessment value corresponds to a value below the stability upper limit threshold, the combined operating condition is marked as a stable operating condition point. When the effective acoustic-fuel coupling stability assessment value is between the stability upper limit threshold and the risk upper limit threshold, the combined operating condition is marked as a risk operating condition point; when the effective acoustic-fuel coupling stability assessment value exceeds the risk upper limit threshold, the combined operating condition is marked as a prohibited operating condition point. The stable operating point, the risky operating point, and the prohibited operating point are mapped to a multi-dimensional operating space to form an acoustic-fuel coupling stability domain. The multi-dimensional operating space uses active power output, coal mill combination, and secondary air distribution mode as independent variables.

5. The automatic power generation dispatch and control method for thermal power units according to claim 1, characterized in that, When the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain, the current scheduling scenario is identified as an acoustic-fuel coupling constraint-dominated scenario, specifically including: Using a sliding time window, the monotonicity and growth rate of the time series are analyzed among multiple consecutive sampling points. When the acoustic-fuel coupling stability index of the multiple consecutive sampling points is greater than or equal to the acoustic-fuel coupling stability index of the previous sampling point, and the increment of the acoustic-fuel coupling stability index of the current sampling point relative to the acoustic-fuel coupling stability index of the starting sampling point of the local time window exceeds a preset rise threshold, it is determined that the time series has a continuous upward trend. Based on the current active power output setpoint, the current coal mill combination, and the current secondary air distribution method, the current operating point is mapped to the acoustic-combustion coupling stability domain, and it is determined whether the current operating point belongs to the stable domain, the risk domain, or the boundary of the prohibited domain. When the time series shows a continuous upward trend and the current operating point belongs to the risk domain or the boundary of the prohibited domain, the current scheduling scenario is marked as the acoustic-fuel coupling constraint-dominated scenario.

6. The automatic power generation dispatch and control method for thermal power units according to claim 1, characterized in that, The reconstructed scheduling optimization model in the scenario dominated by acoustic-fuel coupling constraints uses the unit's active power output setpoint, the pulverizer's on / off status, and the allocation of primary and secondary air volumes as decision variables. An optimization objective is constructed based on the unit's coal consumption cost function and emission cost function. Simultaneously, the unit's physical limits, grid-side constraints, and the acoustic-fuel coupling stability domain are used as optimization constraints. Specifically, this includes: The rolling optimization time domain is discretized according to a uniform time length, and the unit's active power output setpoint, the coal mill's start-up and shutdown status, and the allocation of primary and secondary air volumes in each discrete time period are all used as decision variables to be optimized. The unit coal consumption cost function and the emission cost function are respectively applied to the decision variable set of each scheduling period to calculate the fuel consumption cost and the comprehensive emission cost. The deviation of the unit active power output set value from the active power output command of the automatic generation control constitutes the automatic generation control tracking cost. The acoustic-fuel coupling stability penalty cost constructed based on the acoustic-fuel coupling stability index is used as the cost term in the optimization objective. The physical limit constraints of the unit are the minimum stable output, maximum rated output, ramp rate, minimum continuous operation time and minimum downtime of the coal mill, upper and lower limits of the total primary air volume and total secondary air volume, and upper and lower limits of the secondary air volume distribution ratio. The grid side constraints are the active power balance of the power grid, the reserve capacity requirement and the frequency regulation assessment constraints. The operating condition combination corresponding to the prohibited operating condition point in the acoustic-fuel coupling stability domain is used as the prohibited value, and the time continuous residence of the operating condition combination corresponding to the risk operating condition point is restricted as the acoustic-fuel coupling stability domain constraint.

7. The automatic power generation dispatch and control method for thermal power units according to claim 1, characterized in that, The optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectories obtained from solving the scheduling optimization model are sent to the coordination control layer to adjust the operation of the boiler combustion system and turbine system. This enables the thermal power unit to avoid acoustically sensitive operating conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases. Specifically, this includes: The optimal active power output trajectory, the optimal coal mill combination trajectory, and the optimal primary air and secondary air distribution trajectories are smoothed and checked for consistency, so that the changes in active power output in adjacent time steps meet the unit ramp-up rate limit, and the changes in primary air and secondary air distribution meet the correspondence between the coal mill start-up and shutdown states. The processed optimal active power output trajectory, optimal coal mill combination trajectory, and optimal primary and secondary air distribution trajectory are sent to the coordination control layer. In the coordination control layer, the optimal active power output trajectory is decomposed into boiler main setpoint, turbine valve opening setpoint, and generator excitation voltage setpoint. The optimal coal mill combination trajectory is converted into start / stop commands and load increase / decrease rate settings for each coal mill. The optimal primary air and secondary air distribution trajectories are converted into the output of the primary air fan and the secondary air fan, as well as the opening settings of the burner layer and the side wall damper. During the execution of the trajectory by the coordination and control layer, the optimal primary and secondary air distribution trajectories are corrected by feedforward and feedback based on the real-time furnace negative pressure, oxygen content, flue gas temperature and acoustic-combustion coupling stability index. If the acoustic-combustion coupling stability index does not decrease as expected, the optimal active power output setting and the optimal coal mill combination are adjusted so that the boiler combustion system and turbine system avoid acoustically sensitive conditions and maintain dispatch control under the condition of increased acoustic-combustion coupling instability risk.

8. An automatic power generation dispatch and control device for thermal power units, characterized in that, The device is used to execute an automatic power generation dispatch control method for a thermal power unit as described in any one of claims 1-7. The device includes an acquisition module, a processing module, and an output module, wherein: The acquisition module is used to collect multi-source observation data from the electric side and the furnace combustion side; The processing module is used to preprocess the multi-source observation data to extract the furnace negative pressure fluctuation amplitude, flame centroid vibration amplitude and sound pressure fluctuation amplitude, and to construct an acoustic-flame coupling stability index. The processing module is used to construct the unit's coal consumption cost function and emission cost function based on historical operating data, so as to predict the changes in fuel consumption and emissions of the unit under different active power outputs, different coal mill combinations and different secondary air distribution methods. The processing module is used to divide the acoustic-fuel coupling stability domain according to the acoustic-fuel coupling stability index under different active power outputs, different coal mill combinations and different secondary air distribution methods, set the operating point that induces acoustic-fuel coupling instability as the prohibited domain, and set the operating point at the acoustic-fuel coupling instability boundary as the risk domain. The processing module is used to identify the current scheduling scenario as a scenario dominated by acoustic-fuel coupling constraints when the time series of the acoustic-fuel coupling stability index shows a continuous upward trend and the corresponding operating point is located at the boundary of the risk domain or the prohibited domain. The processing module is used to reconstruct the scheduling optimization model in the scenario dominated by acoustic-fuel coupling constraints. It takes the active power output setpoint of the unit, the start-stop status of the coal mill, and the allocation of primary air volume and secondary air volume as decision variables, and constructs the optimization objective based on the unit coal consumption cost function and the emission cost function. At the same time, it uses the unit physical limit, grid-side constraints and the acoustic-fuel coupling stability domain as optimization constraints. The output module is used to send the optimal active power output trajectory, optimal coal mill combination, and optimal primary and secondary air distribution trajectory obtained by solving the scheduling optimization model to the coordination control layer, so that the thermal power unit can avoid acoustically sensitive operating conditions and maintain scheduling control when the risk of acoustic-fuel coupling instability increases.

9. An electronic device, characterized in that, The device includes a processor, a communication bus, a user interface, a network interface, and a memory. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The communication bus is used to enable communication between the components within the electronic device. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1-7.

10. A non-transitory computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed, perform the method as described in any one of claims 1-7.

Images