A method and system for output optimization based on frequency modulation capability of a thermal power system

Abstract

The present application relates to the technical field of frequency modulation of thermal power system, and discloses a kind of output optimization method and system based on the frequency modulation capability of thermal power system, obtains the monitoring data of thermal power generating unit in preset monitoring period, calculates benchmark frequency modulation capability score;Collect the unit data of each thermal power generating unit coal mill outlet wind powder mixture, dynamically correct benchmark frequency modulation capability score, obtain current frequency modulation capability index;The fluctuation characteristic decomposition of instruction sequence is carried out, and the frequency modulation capacity demand prediction value and the frequency modulation rate demand prediction value are obtained;Establish the objective function of thermal power generating unit output distribution;Optimization solution space containing unit output upper and lower limit constraint, climbing rate constraint, rotating reserve capacity constraint and environmental protection emission concentration constraint is constructed, and the optimal output instruction sequence is obtained, the optimal output instruction sequence is issued to the distributed control system of thermal power generating unit, under the premise of guaranteeing frequency modulation response fast and environmental protection standard, significantly improve the frequency modulation resource utilization rate and economic performance of thermal power system.

Description

Technical Field

[0001] This invention relates to the field of frequency regulation technology for thermal power systems, and more specifically, to a method and system for optimizing output based on the frequency regulation capability of thermal power systems. Background Technology

[0002] Frequency regulation of thermal power systems refers to the key technology of rapidly adjusting the active power output of thermal power units to respond to grid frequency deviations and maintain power quality and system stability.

[0003] Existing technologies often employ a fixed-ratio allocation method based on the unit's rated capacity or human experience weighting, executing output adjustment commands according to preset frequency regulation priorities and response delays. However, frequency regulation decisions in existing technologies are frequently affected by environmental factors such as fluctuations in coal calorific value, changes in ambient temperature, and boiler coking conditions, resulting in low accuracy in assessing the unit's real-time frequency regulation capability. Furthermore, single-load parameter assessment methods contain limited equipment status information, affecting the rationality of matching frequency regulation demand and supply. While multi-parameter comprehensive assessment methods offer rich information, they suffer from inconsistent dimensions and high computational complexity, leading to lengthy frequency regulation optimization times and poor real-time performance, making it difficult to meet the rapid frequency regulation requirements under high-proportion renewable energy integration. Summary of the Invention

[0004] This invention provides a method and system for optimizing the output of thermal power systems based on their frequency regulation capabilities. It realizes the dynamic transformation of multi-source heterogeneous operating parameters into a dimensionless frequency regulation capability index, and uses this as a benchmark to construct a cost minimization objective function driven by the supply-demand ratio. It also incorporates the technical effects of equipment safety and environmental emission constraints, thereby significantly improving the utilization rate and economic performance of thermal power systems' frequency regulation resources while ensuring rapid frequency regulation response and environmental compliance.

[0005] To achieve the above objectives, the present invention provides a power output optimization method based on the frequency regulation capability of a thermal power system, comprising:

[0006] The monitoring data of each thermal power unit in the thermal power system within a preset monitoring period is obtained. The monitoring data includes actual load data, main steam temperature data, main steam pressure data, and coal feeder coal quantity data. Based on the monitoring data, the baseline frequency regulation capability score of each thermal power unit is calculated. Collect unit data on the air-coal mixture at the outlet of each thermal power unit's coal mill. The unit data includes real-time coal calorific value data, air preheater inlet ambient temperature data, and boiler heating surface coking thickness data. Use the unit data to dynamically correct the benchmark frequency regulation capability score to obtain the current frequency regulation capability index. Extract the sequence of automatic generation control commands issued by the power grid dispatch center to the thermal power system, and perform fluctuation characteristic decomposition on the command sequence to obtain the predicted value of frequency regulation capacity demand and the predicted value of frequency regulation rate demand. Based on the current frequency regulation capability index, the predicted value of frequency regulation capacity demand, and the predicted value of frequency regulation rate demand, an objective function for the power output allocation of thermal power units is established, with the optimization direction being the minimization of frequency regulation execution cost. An optimization solution space is constructed, which includes constraints on the upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints. The objective function is solved by nonlinear programming within the solution space to obtain the optimal output command sequence for each thermal power unit. The optimal output command sequence is then sent to the distributed control system of each thermal power unit.

[0007] Furthermore, based on the monitoring data, a baseline frequency regulation capability score is calculated for each thermal power unit, including: Determine the first difference between the maximum load value and the minimum load value of the thermal power unit within the preset monitoring period, and use the ratio of the first difference to the rated installed capacity of the thermal power unit as the load response amplitude benchmark value; Determine the maximum heating rate of the main steam temperature during the load increase process, and use the ratio of the maximum heating rate to the design heating rate as the benchmark value for temperature response speed. Determine the pressure stabilization deviation value of the main steam pressure data during the load reduction process, and use the ratio of the pressure stabilization deviation value to the design pressure value as the pressure stabilization accuracy benchmark value; The coal quantity adjustment time lag of the coal feeder during load change is inversely normalized to obtain the benchmark value of coal quantity response timeliness. The benchmark frequency regulation capability score is obtained by weighting and summing the benchmark values ​​of load response amplitude, temperature response speed, pressure stability accuracy, and coal quantity response timeliness.

[0008] Furthermore, the baseline frequency regulation capability score is dynamically corrected using real-time coal calorific value data, ambient temperature data, and coking thickness data, including: The deviation rate of calorific value between the real-time coal quality calorific value data of the coal mill outlet air-coal mixture and the design coal quality calorific value standard is determined, and the coal quality influence factor is obtained by multiplying the calorific value deviation rate using a preset coal quality correction coefficient. Determine the temperature deviation of the air preheater inlet ambient temperature data from the design ambient temperature range, and use a preset temperature sensitivity coefficient to multiply the temperature deviation to obtain the environmental impact factor. Determine the proportion of the boiler heating surface coking thickness data that exceeds the coking warning thickness, and use the preset coking penalty coefficient to multiply the proportion to obtain the equipment status influence factor. The current frequency regulation capability index is obtained by subtracting the sum of the coal quality impact factor, environmental impact factor, and equipment condition impact factor from the baseline frequency regulation capability score.

[0009] Furthermore, fluctuation characteristic decomposition is performed on the automatic generation control command sequence to obtain the predicted values ​​of frequency regulation capacity demand and frequency regulation rate demand, including: The automatic generation control command sequence is divided into multiple command segment sequences by a time-domain sliding window. Calculate the cumulative change for each instruction segment sequence, and use the maximum value of all cumulative changes as the predicted value for frequency modulation capacity demand; Calculate the first-order difference sequence for each instruction segment sequence, extract the maximum rate of change and the average rate of change in the first-order difference sequence, and weight and fuse the two to obtain the predicted value of frequency regulation rate demand; wherein, the sliding window duration is determined based on the average fluctuation period in the historical data of power grid frequency fluctuation.

[0010] Furthermore, an objective function is established with the optimization direction of minimizing the frequency modulation execution cost, including: The sum of fuel cost and opportunity cost consumed by each thermal power unit in performing frequency regulation tasks per unit capacity is determined as the base price of frequency regulation cost; Determine the supply-demand ratio between the predicted frequency regulation capacity demand and the sum of the current frequency regulation capacity indices of all thermal power units, and multiply the frequency regulation cost base price with the supply-demand ratio to obtain the capacity cost item; Determine the rate matching deviation rate between the predicted frequency regulation rate demand and the average actual output change rate of each thermal power unit, and use the preset rate penalty coefficient to multiply the rate matching deviation rate to obtain the rate cost item. Adding the capacity cost term to the rate cost term yields the objective function expression.

[0011] Furthermore, an optimization solution space is constructed that includes constraints on unit output upper and lower limits, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints, including: Determine the minimum technical output value and maximum safe output value of each thermal power unit based on the equipment health status, and generate upper and lower limit constraints for output; Determine the uphill and downhill speed limits for each thermal power unit based on boiler thermal stress constraints, and generate ramp speed constraints. The total system spinning reserve requirement issued by the power grid dispatch center is determined, and the requirement is allocated according to the current frequency regulation capability index ratio of each thermal power unit to obtain the single unit spinning reserve capacity constraint; the upper limit of nitrogen oxide emission concentration and the lower limit of desulfurization tower efficiency are determined, and the two are combined to generate environmental emission concentration constraints.

[0012] Furthermore, the optimal output command sequence for each thermal power unit is obtained by performing nonlinear programming on the objective function within the optimization solution space, including: The objective function is processed using the Lagrange multiplier method, and the constraints are transformed into penalty terms and incorporated into the objective function to obtain the augmented objective function; An improved particle swarm optimization algorithm is used to iteratively optimize the augmented objective function. During the iteration process, the particle flight speed weighting coefficient is dynamically adjusted according to the current frequency regulation capability index of each thermal power unit. When the particle swarm aggregation degree exceeds the preset aggregation threshold and the number of consecutive times the global optimal solution fails to improve reaches the preset stagnation number, the iteration is terminated and the final particle position is decoded into the optimal output command sequence for each thermal power unit.

[0013] Furthermore, after issuing the optimal output command sequence to the distributed control systems of each thermal power unit, the process also includes: During the execution of the command, the power grid frequency deviation data is collected in real time. When the frequency deviation data exceeds the preset frequency regulation dead zone threshold, the next round of optimization calculation is performed.

[0014] Furthermore, determining whether the frequency deviation data exceeds the preset frequency modulation dead zone threshold includes: Determine the absolute value of the frequency deviation between the rated frequency value of the power grid and the actual frequency value collected; Determine the preset frequency modulation dead zone threshold range. The lower limit of the preset frequency modulation dead zone threshold range is used to trigger the upward frequency modulation demand, and the upper limit is used to trigger the downward frequency modulation demand. The absolute value of the frequency deviation is compared with the lower limit and the upper limit respectively. When the absolute value of the frequency deviation is greater than the lower limit or less than the upper limit, it is determined that the frequency deviation data exceeds the preset frequency tuning dead zone threshold, and a trigger signal is generated to start the next round of optimization calculation.

[0015] To achieve the above objectives, the present invention also provides an output optimization system based on the frequency regulation capability of a thermal power system, comprising: The data acquisition module is used to acquire monitoring data of each thermal power unit in the thermal power system within a preset monitoring period. The monitoring data includes actual load data, main steam temperature data, main steam pressure data, and coal feeder coal quantity data. Based on the monitoring data, the module calculates the benchmark frequency regulation capability score of each thermal power unit. The data acquisition module is used to collect unit data of the air-coal mixture at the outlet of the coal mill of each thermal power unit. The unit data includes real-time coal calorific value data, air preheater inlet ambient temperature data, and boiler heating surface coking thickness data. The unit data is used to dynamically correct the benchmark frequency regulation capability score to obtain the current frequency regulation capability index. The feature decomposition module is used to extract the automatic generation control command sequence issued by the power grid dispatch center to the thermal power system, perform fluctuation feature decomposition on the command sequence, and obtain the frequency regulation capacity demand prediction value and the frequency regulation rate demand prediction value. The function construction module is used to establish an objective function for the power allocation of thermal power units based on the current frequency regulation capability index, the predicted value of frequency regulation capacity demand, and the predicted value of frequency regulation rate demand. The objective function takes minimizing the frequency regulation execution cost as the optimization direction. The output optimization module is used to construct an optimization solution space that includes upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints. Within the solution space, the objective function is solved by nonlinear programming to obtain the optimal output command sequence for each thermal power unit. The optimal output command sequence is then sent to the distributed control system of each thermal power unit.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention discloses a method and system for optimizing the output of thermal power systems based on frequency regulation capability. The method involves acquiring monitoring data from thermal power units within a preset monitoring period and calculating a baseline frequency regulation capability score. It also involves collecting unit data on the air-coal mixture at the coal mill outlet of each thermal power unit, dynamically correcting the baseline frequency regulation capability score, and obtaining the current frequency regulation capability index. Furthermore, it involves performing fluctuation characteristic decomposition on the command sequence to obtain predicted values ​​for frequency regulation capacity demand and frequency regulation rate demand. An objective function for allocating output to thermal power units is established. Finally, an optimization solution space is constructed, including constraints on upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints, to obtain the optimal output command sequence. This optimal output command sequence is then distributed to the distributed control system of the thermal power units. This significantly improves the utilization rate and economic performance of frequency regulation resources in thermal power systems while ensuring rapid frequency regulation response and compliance with environmental standards. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating an output optimization method based on the frequency regulation capability of a thermal power system is shown in an embodiment of the present invention. Figure 2 A schematic diagram of the output optimization system based on the frequency regulation capability of a thermal power system is shown in an embodiment of the present invention. Detailed Implementation

[0018] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0019] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0020] 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 number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0021] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0022] The following is a description of preferred embodiments of the present invention in conjunction with the accompanying drawings.

[0023] like Figure 1 As shown, an embodiment of the present invention discloses a power output optimization method based on the frequency regulation capability of a thermal power system, comprising: S110: Obtain monitoring data of each thermal power unit in the thermal power system within a preset monitoring period. The monitoring data includes actual load data, main steam temperature data, main steam pressure data, and coal feeder coal quantity data. Calculate the baseline frequency regulation capability score of each thermal power unit based on the monitoring data. S120: Collect unit data of the air-coal mixture at the outlet of each thermal power unit's coal mill. The unit data includes real-time coal calorific value data, air preheater inlet ambient temperature data, and boiler heating surface coking thickness data. Use the unit data to dynamically correct the benchmark frequency regulation capability score to obtain the current frequency regulation capability index. S130: Extract the sequence of automatic generation control commands issued by the power grid dispatch center to the thermal power system, perform fluctuation characteristic decomposition on the command sequence, and obtain the predicted value of frequency regulation capacity demand and the predicted value of frequency regulation rate demand. S140: Based on the current frequency regulation capability index, the predicted value of frequency regulation capacity demand and the predicted value of frequency regulation rate demand, establish an objective function for the power output allocation of thermal power units, wherein the objective function takes the minimization of frequency regulation execution cost as the optimization direction; S150: Construct an optimization solution space that includes upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints. Perform nonlinear programming on the objective function within the solution space to obtain the optimal output command sequence for each thermal power unit. Send the optimal output command sequence to the distributed control system of each thermal power unit.

[0024] In some embodiments of this application, a baseline frequency regulation capability score for each thermal power unit is calculated based on the monitoring data, including: Determine the first difference between the maximum load value and the minimum load value of the thermal power unit within the preset monitoring period, and use the ratio of the first difference to the rated installed capacity of the thermal power unit as the load response amplitude benchmark value; Determine the maximum heating rate of the main steam temperature during the load increase process, and use the ratio of the maximum heating rate to the design heating rate as the benchmark value for temperature response speed. Determine the pressure stabilization deviation value of the main steam pressure data during the load reduction process, and use the ratio of the pressure stabilization deviation value to the design pressure value as the pressure stabilization accuracy benchmark value; The coal quantity adjustment time lag of the coal feeder during load change is inversely normalized to obtain the benchmark value of coal quantity response timeliness. The benchmark frequency regulation capability score is obtained by weighting and summing the benchmark values ​​of load response amplitude, temperature response speed, pressure stability accuracy, and coal quantity response timeliness.

[0025] In this embodiment, the preset monitoring period is set to 15 minutes according to the time resolution requirements for frequency regulation performance evaluation in the power grid dispatching operation procedure. This period can balance the integrity of data representation and the real-time performance of calculation.

[0026] In this embodiment, the first difference is obtained by directly subtracting the maximum and minimum load values ​​after traversing all load data points within the preset monitoring period. When calculating the load response amplitude benchmark, the first difference is divided by the unit's rated installed capacity of 500 MW, resulting in a dimensionless ratio of 0.32. This ratio directly reflects the unit's load adjustment range within the monitoring period. The maximum heating rate is used to screen all load increase periods within the monitoring period, calculating the maximum increase in main steam temperature per minute. For example, if the main steam temperature rises from 560℃ to 580℃ in 5 minutes during a load increase process, the heating rate is 4℃ per minute. The design heating rate is 3℃ per minute, and the ratio of the two is used as the temperature response speed benchmark. The pressure stabilization deviation during load reduction is calculated as the average deviation between the main steam pressure and its set value during all load reduction periods within the monitoring period. The design pressure is taken as the main steam pressure under rated operating conditions of 18 MPa, and the ratio of the two is 0.011, used as the pressure stability accuracy benchmark. The smaller this value, the more precise the pressure control. The coal quantity adjustment time lag is determined by the time difference between the moment the coal feeder command changes and the moment the actual coal quantity begins to change. The inverse normalization process uses an upper limit threshold method, setting a maximum allowable time lag of 120 seconds. The difference between the actual time lag and the upper limit is divided by the upper limit; the shorter the time lag, the higher the score. Load response amplitude accounts for 40% of the weight, temperature response speed for 30%, pressure stability accuracy for 20%, and coal quantity response timeliness for 10%. The final benchmark frequency regulation capability score ranges from 0 to 100 points.

[0027] The beneficial effects of the above technical solution are: by constructing and weighting the benchmark values ​​of four key operating parameters, a quantitative assessment of the frequency regulation capability of thermal power units can be achieved, providing objective data support for subsequent optimized allocation and avoiding the subjectivity of traditional experience-based scheduling.

[0028] In some embodiments of this application, the benchmark frequency regulation capability score is dynamically corrected using real-time coal calorific value data, ambient temperature data, and coking thickness data, including: The deviation rate of calorific value between the real-time coal quality calorific value data of the coal mill outlet air-coal mixture and the design coal quality calorific value standard is determined, and the coal quality influence factor is obtained by multiplying the calorific value deviation rate using a preset coal quality correction coefficient. Determine the temperature deviation of the air preheater inlet ambient temperature data from the design ambient temperature range, and use a preset temperature sensitivity coefficient to multiply the temperature deviation to obtain the environmental impact factor. Determine the proportion of the coking thickness on the boiler heating surface that exceeds the coking warning thickness, and use the preset coking penalty coefficient to multiply the proportion to obtain the equipment status influence factor. The current frequency regulation capability index is obtained by subtracting the sum of the coal quality impact factor, environmental impact factor, and equipment condition impact factor from the baseline frequency regulation capability score.

[0029] In this embodiment, the calorific value deviation rate is calculated as (real-time coal calorific value data - design coal calorific value standard) / design coal calorific value standard 100%. The preset coal quality correction coefficient is determined to be 0.8 based on the influence of coal quality on boiler response characteristics; this coefficient is obtained through comparative tests of frequency regulation performance under different coal qualities. The coal quality influence factor is the product of the absolute value of the calorific value deviation rate and the correction coefficient. The temperature offset is calculated by subtracting the midpoint value of the design range from the actual ambient temperature. For example, if the actual temperature is 28℃ and the design range is 20-25℃ with the midpoint at 22.5℃, then the temperature offset is 5.5℃. The preset temperature sensitivity coefficient is determined to be 0.05 per degree Celsius based on the slope of the curve showing boiler thermal efficiency changing with ambient temperature; this value comes from the multi-condition simulation results of the boiler thermal calculation software. The environmental influence factor is the temperature offset multiplied by the sensitivity coefficient; in this example, 5.5℃ × 0.05 = 0.275. This factor reflects the degree to which the ambient temperature deviation from the design value weakens the frequency regulation capability. The coking warning thickness is set at 50 mm according to the boiler operation procedures. The percentage exceeding the warning thickness is calculated by subtracting 50 mm from the actual coking thickness and dividing by 50 mm. For example, if the actual coking thickness is 65 mm, the percentage is (65-50) / 50 = 0.3. The preset coking penalty coefficient is determined to be 0.1 based on the impact of coking on the heat exchange efficiency of the heating surface. This coefficient is obtained by fitting the relationship curve between the cleanliness factor and the coking thickness. The current frequency regulation capability index is the baseline frequency regulation capability score minus the sum of three influencing factors. For example, if the baseline score is 85, the current index is 85 - (4.36 + 0.275 + 0.03) = 80.335. Dynamic correction makes the frequency regulation capability assessment closer to the real-time state.

[0030] The beneficial effects of the above technical solution are: by introducing quantitative corrections for three external disturbance factors—coal quality, ambient temperature, and equipment status—the interference of changes in operating conditions on frequency regulation capability assessment is eliminated, making the dynamic allocation of frequency regulation resources more accurate and reliable, and improving the stability of the power grid frequency.

[0031] In some embodiments of this application, fluctuation characteristic decomposition is performed on the automatic generation control command sequence to obtain frequency regulation capacity demand forecast and frequency regulation rate demand forecast, including: The automatic generation control command sequence is divided into multiple command segment sequences by a time-domain sliding window. Calculate the cumulative change for each instruction segment sequence, and use the maximum value of all cumulative changes as the predicted value for frequency modulation capacity demand; Calculate the first-order difference sequence for each instruction segment sequence, extract the maximum rate of change and the average rate of change in the first-order difference sequence, and weight and fuse the two to obtain the predicted value of frequency regulation rate demand; wherein, the sliding window duration is determined based on the average fluctuation period in the historical data of power grid frequency fluctuation.

[0032] In this embodiment, time-domain sliding window segmentation divides the continuous automatic generation control command sequence into non-overlapping segments with a fixed duration. The window duration is set to 5 minutes, which is obtained by performing spectral analysis on historical grid frequency fluctuation data over the past 30 days. The analysis shows that the power spectral density of the frequency fluctuation peaks around 0.0033 Hz, corresponding to a period of 5 minutes; therefore, this value is chosen as the window width. The command segment sequence is the set of AGC command points within each 5-minute window. For example, 15 minutes of data can be divided into 3 segments, each containing 75 command points (one point every 4 seconds). The cumulative change is calculated as the difference between the maximum and minimum values ​​of the commands within each segment. For example, if a segment's command increases from 350 MW to 420 MW and then decreases to 380 MW, the cumulative change is 420 - 350 = 70 MW. The predicted frequency regulation capacity demand is the maximum cumulative change across all segments. If the cumulative changes for the three segments are 70 MW, 85 MW, and 60 MW, respectively, the predicted value is 85 MW. This value represents the maximum frequency regulation capacity demand the system may encounter in the next cycle. The first-order difference sequence calculates the difference between adjacent command points within each segment, resulting in 74 difference values. For example, if the command within a segment changes from 350 MW to 355 MW, the difference is +5 MW. The maximum rate of change is the maximum absolute value in the difference sequence. For example, if the maximum value in the difference sequence is +8 MW every 4 seconds, this translates to a rate of 120 MW per minute. The average rate of change is the arithmetic mean of all values ​​in the difference sequence. In the weighted fusion process, the maximum rate of change value accounts for 70% of the weight, and the average rate of change value accounts for 30%. After fusion, the predicted frequency regulation rate demand is 120 × 0.7 + 30 × 0.3 = 93 MW / min. This weight allocation is based on the fact that the maximum rate represents the demand under extreme operating conditions, while the average rate represents the demand under normal operating conditions. Extreme operating conditions have a greater impact on the success or failure of frequency regulation, hence the higher weight. The 5-minute sliding window duration is determined based on the historical statistical characteristics of grid frequency fluctuations, ensuring that the window width matches the typical fluctuation cycle and that the decomposition results are representative of predictions.

[0033] The beneficial effects of the above technical solution are: by performing time-domain sliding window decomposition on the AGC instruction sequence, extracting two-dimensional demand characteristics of frequency regulation capacity and rate, quantifying the intensity of power grid frequency regulation tasks, providing accurate demand-side input for subsequent optimized allocation, and improving the matching degree of frequency regulation resources.

[0034] In some embodiments of this application, an objective function is established with the optimization direction of minimizing the frequency modulation execution cost, including: The sum of fuel cost and opportunity cost consumed by each thermal power unit in performing frequency regulation tasks per unit capacity is determined as the base price of frequency regulation cost; Determine the supply-demand ratio between the predicted frequency regulation capacity demand and the sum of the current frequency regulation capacity indices of all thermal power units, and multiply the frequency regulation cost base price with the supply-demand ratio to obtain the capacity cost item; Determine the rate matching deviation rate between the predicted frequency regulation rate demand and the average actual output change rate of each thermal power unit, and use the preset rate penalty coefficient to multiply the rate matching deviation rate to obtain the rate cost item. Adding the capacity cost term to the rate cost term yields the objective function expression.

[0035] In this embodiment, fuel cost is calculated by multiplying the additional coal consumed during frequency regulation by the standard coal price. For example, if a unit needs to consume an additional 0.5 tons of standard coal to perform a 10 MW frequency regulation task, and the standard coal price is 800 yuan per ton, then the fuel cost is 400 yuan. Opportunity cost is estimated using the benefit loss curve of the unit deviating from its economic load point. For example, if a unit's economic load point is 400 MW, but it needs to be reduced to 380 MW for frequency regulation, the coal consumption rate increases by 2 grams per kilowatt-hour according to the coal consumption-load characteristic curve. At an electricity price of 0.4 yuan per kilowatt-hour, the hourly loss is 304 yuan, which translates to an opportunity cost of 101 yuan for a 20-minute frequency regulation cycle. In this example, 400 + 101 = 501 yuan per MW. This base price reflects the comprehensive cost per unit of frequency regulation capacity. The capacity cost item is the frequency regulation cost base price multiplied by the supply-demand ratio. This cost item reflects the relative tightness of frequency regulation capacity demand and resource supply; the tighter the supply and demand, the higher the cost. The rate matching deviation rate is calculated by dividing the predicted frequency regulation rate demand of 93 MW / min by the average actual output change rate of each unit (60 MW / min) and then subtracting 1, resulting in 0.55. A value greater than 0 indicates that the system's frequency regulation rate demand exceeds the average supply capacity of the units. The preset rate penalty coefficient is determined to be 1.5 based on the severity of the impact of rate mismatch on grid frequency stability. This coefficient is calculated by simulating frequency deviation curves under different rate deviations using power system simulation software, and the penalty intensity corresponding to the critical point of frequency deviation exceeding the standard is taken. The rate cost item is the rate matching deviation rate multiplied by the penalty coefficient. This cost item is dimensionless and has the same dimensions as the capacity cost item.

[0036] The expression for the objective function is: OBJ is the objective function, C cap For capacity cost, C rate This is the rate cost item.

[0037] The beneficial effects of the above technical solution are: by constructing a two-dimensional cost objective function of capacity and rate, the frequency regulation demand of the power grid and the frequency regulation cost of the generating units are quantitatively correlated, and the optimized solution realizes the economic allocation of frequency regulation tasks, thereby reducing the total frequency regulation cost of the system.

[0038] In some embodiments of this application, an optimization solution space is constructed that includes constraints on upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints, including: Determine the minimum technical output value and maximum safe output value of each thermal power unit based on the equipment health status, and generate upper and lower limit constraints for output; Determine the uphill and downhill speed limits for each thermal power unit based on boiler thermal stress constraints, and generate ramp speed constraints. The total system spinning reserve requirement issued by the power grid dispatch center is determined, and the requirement is allocated according to the current frequency regulation capability index ratio of each thermal power unit to obtain the single unit spinning reserve capacity constraint; the upper limit of nitrogen oxide emission concentration and the lower limit of desulfurization tower efficiency are determined, and the two are combined to generate environmental emission concentration constraints.

[0039] In this embodiment, the minimum technical output value is determined based on the boiler's minimum stable combustion load thermal test report. The test is conducted after the unit's major overhaul, gradually reducing the load until combustion becomes unstable, resulting in flame flickering or excessive carbon monoxide content. At this point, the load value is increased by 5% as the minimum technical output, which is 40% of the rated capacity. The maximum safe output value is determined based on the turbine high-pressure cylinder regulating stage metal temperature and downstream pressure monitoring data. When generating the upper and lower output limits, the minimum technical output value is used as the lower limit and the maximum safe output value is used as the upper limit, forming a closed interval, such as [200, 525] MW. The optimization algorithm searches for the optimal unit output within this interval, and output solutions outside the interval are considered infeasible solutions and discarded directly. The uphill rate limit is set based on the boiler drum upper and lower wall temperature difference control requirements. The drum material is 20G carbon steel, and the allowable wall temperature difference is 50℃. By simulating the wall temperature difference under different load increase rates using thermal stress calculation software, it is determined that a rate of 3% per minute (15 MW per minute) can meet the wall temperature difference limit. The downhill rate limit is set based on boiler combustion stability and drum water level safety. Too rapid a load reduction will shorten the residence time of pulverized coal in the furnace, leading to incomplete combustion. Simultaneously, the drum water level will exhibit a false reading due to a sudden drop in steam flow. Therefore, 2% per minute (10 MW per minute) is determined as the safe lower limit. When generating the ramp rate constraint, the downhill rate limit is negative as the lower bound, and the ramp rate limit is the upper bound, forming a ramp range of [-10, +15] MW per minute, constraining the output variation of the unit at adjacent moments. Assuming a demand of 300 MW, this demand characterizes the reserve capacity required by the system to cope with sudden accidents or load changes. When allocating according to the current frequency regulation capability index ratio of each thermal power unit, if there are 5 units in the system with frequency regulation capability indices of 80, 90, 85, 75, and 70 points respectively, totaling 400 points, then the unit with an index of 80 points is allocated 300 × 80 / 400 = 60 MW of spinning reserve. The single-unit spinning reserve capacity constraint uses this allocation value as the lower limit of the unit's reserve. For example, if the unit's current output is 400 MW, its maximum allowable output shall not exceed 400 + 60 = 460 MW, ensuring that the unit has sufficient room for upward adjustment.

[0040] The beneficial effects of the above technical solution are: by constructing a four-dimensional constraint space of equipment, thermal stress, backup, and environmental protection, it ensures that the optimization solution is physically achievable, equipment safety is guaranteed, and environmental protection indicators are compliant, thus avoiding the problem that the theoretical optimal solution cannot be implemented.

[0041] In some embodiments of this application, the optimal output command sequence for each thermal power unit is obtained by performing nonlinear programming on the objective function within the optimization solution space, including: The objective function is processed using the Lagrange multiplier method, and the constraints are transformed into penalty terms and incorporated into the objective function to obtain the augmented objective function; An improved particle swarm optimization algorithm is used to iteratively optimize the augmented objective function. During the iteration process, the particle flight speed weighting coefficient is dynamically adjusted according to the current frequency regulation capability index of each thermal power unit. When the particle swarm aggregation degree exceeds the preset aggregation threshold and the number of consecutive times the global optimal solution fails to improve reaches the preset stagnation number, the iteration is terminated and the final particle position is decoded into the optimal output command sequence for each thermal power unit.

[0042] In this embodiment, the Lagrange multiplier method is used to construct penalty functions for the four constraints (output upper and lower limits, ramp rate, spindle reserve, and environmental emissions). The penalty term for the output upper and lower limit constraints is (amount exceeding the limit)² × penalty coefficient. The penalty term for the ramp rate constraints is (amount exceeding the limit)² × penalty coefficient. The penalty term for the spindle reserve constraints is (amount of reserve shortage)² × penalty coefficient. The penalty term for the environmental emissions constraints is (amount of concentration exceeding the standard)² × penalty coefficient. The augmented objective function is the original objective function value of 85.995 plus the sum of all penalty terms, resulting in 37535.995. The penalty terms significantly increase the objective function value, forcing the optimization algorithm to search the feasible region.

[0043] The augmented objective function is:

[0044] OB Jaug To augment the objective function, λ j P is the corresponding penalty coefficient. j Let be the penalty function for the j-th constraint.

[0045] The improved particle swarm optimization algorithm uses a particle swarm size of 50, where each particle position is an N-dimensional vector (N being the number of units), with each dimension representing the output value of a unit. During iterative optimization, particles update their flight speed based on individual and global optimum. In the standard particle swarm speed update formula, the inertia weight, individual learning factor, and social learning factor are set to 0.729, 1.494, and 1.494, respectively. When dynamically adjusting the particle flight speed weight coefficients, the current frequency regulation capability index of each unit is obtained, and the system average index is calculated. For units with an index higher than the average, the corresponding dimension speed weight is multiplied by 1.1; for units with an index lower than the average, it is multiplied by 0.9. For example, the weight of a unit with an index of 70 is adjusted to 0.9, allowing for a slower and more refined search. The particle swarm aggregation degree is calculated as the standard deviation of all particle positions. The number of consecutive iterations without improvement of the global optimum is counted. After each iteration, the global optimum fitness value is recorded. If the optimal value remains unchanged for 20 consecutive iterations, the optimization is considered stagnant. The preset number of stagnations (20) is determined based on algorithm convergence speed testing, balancing computational accuracy and time overhead. After the iteration is terminated, the components of each dimension of the final particle position are decoded into the corresponding power output command of the unit. For example, the particle's component of 450 MW is decoded into the optimal power output of Unit 3 of 450 MW, forming a complete optimal power output command sequence.

[0046] The beneficial effects of the above technical solution are: by improving the particle swarm optimization algorithm to solve the augmented objective function, dynamically adjusting the search strategy to adapt to the differences in frequency regulation capabilities of the units, intelligently judging the convergence conditions, and obtaining a globally optimal power allocation scheme that satisfies multiple constraints within 30 seconds, thereby improving the economy of frequency regulation.

[0047] In some embodiments of this application, after the optimal output command sequence is sent to the distributed control system of each thermal power unit, the method further includes: During the execution of the command, the power grid frequency deviation data is collected in real time. When the frequency deviation data exceeds the preset frequency regulation dead zone threshold, the next round of optimization calculation is performed.

[0048] In some embodiments of this application, determining whether the frequency deviation data exceeds a preset frequency modulation dead zone threshold includes: Determine the absolute value of the frequency deviation between the rated frequency value of the power grid and the actual frequency value collected; Determine the preset frequency modulation dead zone threshold range. The lower limit of the preset frequency modulation dead zone threshold range is used to trigger the upward frequency modulation demand, and the upper limit is used to trigger the downward frequency modulation demand. The absolute value of the frequency deviation is compared with the lower limit and the upper limit respectively. When the absolute value of the frequency deviation is greater than the lower limit or less than the upper limit, it is determined that the frequency deviation data exceeds the preset frequency tuning dead zone threshold, and a trigger signal is generated to start the next round of optimization calculation.

[0049] In this embodiment, the rated frequency of the power grid is fixed at 50 Hz according to the power quality standard. This value is the standard frequency of industrial frequency AC power, with an allowable deviation range of ±0.2 Hz. The actual frequency value is collected from the digital frequency transmitter on the 500 kV bus side of the substation. This transmitter uses hardware phase-locked loop technology and has a measurement accuracy of 0.001 Hz. The absolute value of the frequency deviation is calculated as the absolute value of the actual frequency value minus the rated frequency value. The lower limit of the preset frequency regulation dead zone threshold range is set to -0.033 Hz, corresponding to a turbine speed of -2 revolutions per minute (3000 rpm baseline). This value is used to trigger upward frequency regulation demand, i.e., when the frequency is too low, the generator needs to increase its output. The upper limit is set to +0.033 Hz, corresponding to a turbine speed of +2 revolutions per minute. This value is used to trigger downward frequency regulation demand, i.e., when the frequency is too high, the generator needs to decrease its output. 0.033 Hz corresponds to the initial stage of power grid frequency control deviation. After entering this range, secondary frequency regulation needs to be initiated to prevent further frequency deterioration. When comparing the absolute value of the frequency deviation with the lower limit, if the absolute value of the deviation (0.033 Hz) is greater than the absolute value of the lower limit (0.033 Hz in actual logic, the frequency value of 49.967 Hz is less than 50 - 0.033 Hz), it is determined that the upward frequency modulation condition is met. When comparing with the upper limit, if the absolute value of the frequency deviation is greater than the upper limit (actual frequency above 50.033 Hz), it is determined that the downward frequency modulation condition is met. The comparison logic is implemented using an OR gate circuit; if either condition is met, the frequency deviation data is determined to exceed the preset frequency modulation dead zone threshold. The trigger signal is a digital signal, connected to the AGC command receiving module of the optimization master station via hard-wiring. The signal is active high and lasts for more than 100 milliseconds to ensure reliable triggering. When starting the next round of optimization calculation, the remaining process of the current cycle is interrupted, and the AGC command sequence is immediately re-extracted for fluctuation feature decomposition to achieve closed-loop rolling optimization.

[0050] The beneficial effects of the above technical solution are: by setting a clear frequency dead zone threshold and intelligent triggering mechanism, the unit avoids excessive response to small frequency fluctuations, reduces equipment wear and coal waste, and ensures rapid start-up optimization when the frequency deviation exceeds the safe range, thereby improving the frequency recovery capability under large disturbances.

[0051] To further illustrate the technical concept of this invention, the technical solution of this invention will now be described in conjunction with specific application scenarios.

[0052] Correspondingly, such as Figure 2 As shown, this application also provides an output optimization system based on the frequency regulation capability of a thermal power system, comprising: The data acquisition module is used to acquire monitoring data of each thermal power unit in the thermal power system within a preset monitoring period. The monitoring data includes actual load data, main steam temperature data, main steam pressure data, and coal feeder coal quantity data. Based on the monitoring data, the module calculates the benchmark frequency regulation capability score of each thermal power unit. The data acquisition module is used to collect unit data of the air-coal mixture at the outlet of the coal mill of each thermal power unit. The unit data includes real-time coal calorific value data, air preheater inlet ambient temperature data, and boiler heating surface coking thickness data. The unit data is used to dynamically correct the benchmark frequency regulation capability score to obtain the current frequency regulation capability index. The feature decomposition module is used to extract the automatic generation control command sequence issued by the power grid dispatch center to the thermal power system, perform fluctuation feature decomposition on the command sequence, and obtain the frequency regulation capacity demand prediction value and the frequency regulation rate demand prediction value. The function construction module is used to establish an objective function for the power allocation of thermal power units based on the current frequency regulation capability index, the predicted value of frequency regulation capacity demand, and the predicted value of frequency regulation rate demand. The objective function takes minimizing the frequency regulation execution cost as the optimization direction. The output optimization module is used to construct an optimization solution space that includes upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints. Within the solution space, the objective function is solved by nonlinear programming to obtain the optimal output command sequence for each thermal power unit. The optimal output command sequence is then sent to the distributed control system of each thermal power unit.

[0053] In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0054] Although the invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the embodiments disclosed in this invention can be combined with each other in any way. The fact that not all of these combinations are described in this specification is merely for the sake of brevity and resource conservation.

[0055] It will be understood by those skilled in the art that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing power output based on the frequency regulation capability of a thermal power system, characterized in that, include: The monitoring data of each thermal power unit in the thermal power system within a preset monitoring period is obtained. The monitoring data includes actual load data, main steam temperature data, main steam pressure data, and coal feeder coal quantity data. Based on the monitoring data, the baseline frequency regulation capability score of each thermal power unit is calculated. Collect unit data on the air-coal mixture at the outlet of each thermal power unit's coal mill. The unit data includes real-time coal calorific value data, air preheater inlet ambient temperature data, and boiler heating surface coking thickness data. Use the unit data to dynamically correct the benchmark frequency regulation capability score to obtain the current frequency regulation capability index. Extract the sequence of automatic generation control commands issued by the power grid dispatch center to the thermal power system, and perform fluctuation characteristic decomposition on the command sequence to obtain the predicted value of frequency regulation capacity demand and the predicted value of frequency regulation rate demand. Based on the current frequency regulation capability index, the predicted value of frequency regulation capacity demand, and the predicted value of frequency regulation rate demand, an objective function for the power output allocation of thermal power units is established, with the optimization direction being the minimization of frequency regulation execution cost. An optimization solution space is constructed, which includes constraints on the upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints. The objective function is solved by nonlinear programming within the solution space to obtain the optimal output command sequence for each thermal power unit. The optimal output command sequence is then sent to the distributed control system of each thermal power unit.

2. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 1, characterized in that, Based on the monitoring data, a baseline frequency regulation capability score for each thermal power unit is calculated, including: Determine the first difference between the maximum load value and the minimum load value of the thermal power unit within the preset monitoring period, and use the ratio of the first difference to the rated installed capacity of the thermal power unit as the load response amplitude benchmark value; Determine the maximum heating rate of the main steam temperature during the load increase process, and use the ratio of the maximum heating rate to the design heating rate as the benchmark value for temperature response speed. Determine the pressure stabilization deviation value of the main steam pressure data during the load reduction process, and use the ratio of the pressure stabilization deviation value to the design pressure value as the pressure stabilization accuracy benchmark value; The coal quantity adjustment time lag of the coal feeder during load change is inversely normalized to obtain the benchmark value of coal quantity response timeliness. The benchmark frequency regulation capability score is obtained by weighting and summing the benchmark values ​​of load response amplitude, temperature response speed, pressure stability accuracy, and coal quantity response timeliness.

3. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 1, characterized in that, The baseline frequency regulation capability score is dynamically corrected using real-time coal calorific value data, ambient temperature data, and coking thickness data, including: The deviation rate of calorific value between the real-time coal quality calorific value data of the coal mill outlet air-coal mixture and the design coal quality calorific value standard is determined, and the coal quality influence factor is obtained by multiplying the calorific value deviation rate using a preset coal quality correction coefficient. Determine the temperature deviation of the air preheater inlet ambient temperature data from the design ambient temperature range, and use a preset temperature sensitivity coefficient to multiply the temperature deviation to obtain the environmental impact factor. Determine the proportion of the coking thickness on the boiler heating surface that exceeds the coking warning thickness, and use the preset coking penalty coefficient to multiply the proportion to obtain the equipment status influence factor. The current frequency regulation capability index is obtained by subtracting the sum of the coal quality impact factor, environmental impact factor, and equipment condition impact factor from the baseline frequency regulation capability score.

4. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 1, characterized in that, Fluctuation characteristic decomposition of the automatic generation control command sequence yields predicted values ​​for frequency regulation capacity demand and frequency regulation rate demand, including: The automatic generation control command sequence is divided into multiple command segment sequences by a time-domain sliding window. Calculate the cumulative change for each instruction segment sequence, and use the maximum value of all cumulative changes as the predicted value for frequency modulation capacity demand; Calculate the first-order difference sequence for each instruction segment sequence, extract the maximum rate of change and the average rate of change in the first-order difference sequence, and weight and fuse the two to obtain the predicted value of frequency regulation rate demand; wherein, the sliding window duration is determined based on the average fluctuation period in the historical data of power grid frequency fluctuation.

5. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 1, characterized in that, Establish an objective function with the optimization direction of minimizing the frequency modulation execution cost, including: The sum of fuel cost and opportunity cost consumed by each thermal power unit in performing frequency regulation tasks per unit capacity is determined as the base price of frequency regulation cost; Determine the supply-demand ratio between the predicted frequency regulation capacity demand and the sum of the current frequency regulation capacity indices of all thermal power units, and multiply the frequency regulation cost base price with the supply-demand ratio to obtain the capacity cost item; Determine the rate matching deviation rate between the predicted frequency regulation rate demand and the average actual output change rate of each thermal power unit, and use the preset rate penalty coefficient to multiply the rate matching deviation rate to obtain the rate cost item. Adding the capacity cost term to the rate cost term yields the objective function expression.

6. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 1, characterized in that, An optimization solution space is constructed that includes constraints on unit output upper and lower limits, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints, including: Determine the minimum technical output value and maximum safe output value of each thermal power unit based on the equipment health status, and generate upper and lower limit constraints for output; Determine the uphill and downhill speed limits for each thermal power unit based on boiler thermal stress constraints, and generate ramp speed constraints. The total system spinning reserve requirement issued by the power grid dispatch center is determined, and the requirement is allocated according to the current frequency regulation capability index ratio of each thermal power unit to obtain the single unit spinning reserve capacity constraint; the upper limit of nitrogen oxide emission concentration and the lower limit of desulfurization tower efficiency are determined, and the two are combined to generate environmental emission concentration constraints.

7. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 6, characterized in that, The optimal output command sequence for each thermal power unit is obtained by nonlinear programming of the objective function within the optimization solution space, including: The objective function is processed using the Lagrange multiplier method, and the constraints are transformed into penalty terms and incorporated into the objective function to obtain the augmented objective function; An improved particle swarm optimization algorithm is used to iteratively optimize the augmented objective function. During the iteration process, the particle flight speed weighting coefficient is dynamically adjusted according to the current frequency regulation capability index of each thermal power unit. When the particle swarm aggregation degree exceeds the preset aggregation threshold and the number of consecutive times the global optimal solution fails to improve reaches the preset stagnation number, the iteration is terminated and the final particle position is decoded into the optimal output command sequence for each thermal power unit.

8. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 1, characterized in that, After the optimal output command sequence is sent to the distributed control systems of each thermal power unit, the following steps are also included: During the execution of the command, the power grid frequency deviation data is collected in real time. When the frequency deviation data exceeds the preset frequency regulation dead zone threshold, the next round of optimization calculation is performed.

9. The power output optimization method based on the frequency regulation capability of a thermal power system according to claim 8, characterized in that, When determining whether the frequency deviation data exceeds the preset frequency tuning dead zone threshold, the following steps are taken: Determine the absolute value of the frequency deviation between the rated frequency value of the power grid and the actual frequency value collected; Determine the preset frequency modulation dead zone threshold range. The lower limit of the preset frequency modulation dead zone threshold range is used to trigger the upward frequency modulation demand, and the upper limit is used to trigger the downward frequency modulation demand. The absolute value of the frequency deviation is compared with the lower limit and the upper limit respectively. When the absolute value of the frequency deviation is greater than the lower limit or less than the upper limit, it is determined that the frequency deviation data exceeds the preset frequency tuning dead zone threshold, and a trigger signal is generated to start the next round of optimization calculation.

10. A power output optimization system based on the frequency regulation capability of a thermal power system, applied to the power output optimization method based on the frequency regulation capability of a thermal power system as described in any one of claims 1-9, characterized in that, include: The data acquisition module is used to acquire monitoring data of each thermal power unit in the thermal power system within a preset monitoring period. The monitoring data includes actual load data, main steam temperature data, main steam pressure data, and coal feeder coal quantity data. Based on the monitoring data, the module calculates the benchmark frequency regulation capability score of each thermal power unit. The data acquisition module is used to collect unit data of the air-coal mixture at the outlet of the coal mill of each thermal power unit. The unit data includes real-time coal calorific value data, air preheater inlet ambient temperature data, and boiler heating surface coking thickness data. The unit data is used to dynamically correct the benchmark frequency regulation capability score to obtain the current frequency regulation capability index. The feature decomposition module is used to extract the automatic generation control command sequence issued by the power grid dispatch center to the thermal power system, perform fluctuation feature decomposition on the command sequence, and obtain the frequency regulation capacity demand prediction value and the frequency regulation rate demand prediction value. The function construction module is used to establish an objective function for the power allocation of thermal power units based on the current frequency regulation capability index, the predicted value of frequency regulation capacity demand, and the predicted value of frequency regulation rate demand. The objective function takes minimizing the frequency regulation execution cost as the optimization direction. The output optimization module is used to construct an optimization solution space that includes upper and lower limits of unit output, ramp rate constraints, spinning reserve capacity constraints, and environmental emission concentration constraints. Within the solution space, the objective function is solved by nonlinear programming to obtain the optimal output command sequence for each thermal power unit. The optimal output command sequence is then sent to the distributed control system of each thermal power unit.

Images