Depth scheduling method and system for deep peak regulation of coal-fired unit

Abstract

The invention relates to the technical field of power system resource optimization scheduling, in particular to a deep scheduling method and system for deep peak regulation of a coal-fired unit, and the method comprises the steps: collecting multi-dimensional operation parameters of the unit, a deep peak regulation instruction and a carbon price signal of a carbon emission permit trading market, and achieving timestamp alignment; constructing a five-dimensional security constraint model associated with the economic cost boundary; taking peak regulation income maximization as a target, combining fire coal purchase cost, carbon transaction cost and equipment maintenance cost to construct a target function, adopting a dynamic planning algorithm to solve an economic dispatching scheme, and executing a cross-system cooperation strategy; and on the basis of LSTM-based equipment life prediction and digital twin simulation, triggering a cross-system correction strategy. According to the method, the commercial cooperation problem of economy, environmental protection and equipment reliability in deep peak regulation is solved, and the peak regulation income maximization of the coal-fired unit in an electricity market environment is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power system resource optimization and scheduling, and in particular to a deep scheduling method and system for deep peak regulation of coal-fired units. Background Art

[0002] In the early days, coal-fired units mainly undertook base load power generation tasks in the power system, and their operating conditions were relatively stable. Their combustion systems, turbine steam distribution logic, and auxiliary equipment configurations were all designed and optimized around efficient and stable operation under rated load. The scheduling method focused on meeting the power generation needs of the rated load, and lacked adaptability considerations to the unit's deep load variation operation scenarios. A multi-dimensional safety constraint collaborative management mechanism for the furnace stable combustion boundary, SCR denitrification efficiency lower limit, turbine vibration threshold, carbon emission intensity upper limit, and secondary reheat system temperature difference safety zone during deep peak regulation was not established.

[0003] With the construction of new power systems, the installed capacity of renewable energy such as wind power and photovoltaics will exceed 45% by 2023. Their intermittent and volatile characteristics pose severe challenges to the flexibility of the power grid; at the same time, the launch of the carbon emission trading market has made carbon price signals an important variable for the economic dispatch of units. Although the application of secondary reheating technology in coal-fired units has improved energy efficiency, it has also increased the complexity of dispatching scenarios. Traditional peak-shaving technology, which only focuses on a single load regulation target, can no longer meet the multiple needs of stable operation of rated load in deep peak-shaving, low-carbonization carbon quota trading cost constraints, and energy efficiency improvement and optimization of secondary reheating systems.

[0004] Although some research and applications have focused on deep peak-shaving scheduling for coal-fired units, existing technologies still have significant deficiencies. First, there is a lack of comprehensive modeling of the multi-dimensional safety constraints of units during deep peak-shaving. A five-dimensional safety constraint model, encompassing the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the safety threshold of secondary reheat temperature difference, has not been constructed, effectively guaranteeing the stability and reliability of units under low-load operation. Second, for complex scenarios involving deep peak-shaving coupled with multiple factors such as the carbon trading market and secondary reheat systems, existing scheduling methods struggle to achieve coordinated control and optimization of multiple systems. Furthermore, they fail to integrate peak-shaving ancillary service revenue, coal procurement costs, carbon quota trading costs, and equipment maintenance costs to construct a function aimed at maximizing peak-shaving revenue. Furthermore, there is a lack of LSTM neural network-based prediction of the lifespan of equipment such as turbine blades and SCR catalysts, as well as cross-system correction strategies for safety boundary breach scenarios simulated using digital twins integrated with the APROS and EBSILON platforms. Consequently, the full potential of coal-fired units for flexible regulation in new power systems is not fully realized, limiting the large-scale integration of renewable energy and the low-carbon transformation of power systems.

[0005] The invention patent with publication number CN114444785B discloses a deep scheduling method and system for deep peak regulation of coal-fired units. By constructing a carbon emission intensity and transaction cost model, the objective function is to minimize the total peak regulation cost and the solution is solved in combination with multiple constraints. However, the efficiency loss of the secondary reheat system under deep peak regulation is not considered, and there is a lack of a linkage protection mechanism for turbine vibration and SCR temperature. Summary of the Invention

[0006] The purpose of the present invention is to solve the problems existing in the prior art in deep peak regulation of coal-fired units, such as blurred safety boundaries, loss of secondary reheating energy efficiency, insufficient coordination of multiple systems, delayed equipment maintenance and extensive economic scheduling, and to propose a deep scheduling method and system for deep peak regulation of coal-fired units.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions: A deep scheduling method for deep peak regulation of coal-fired units, comprising the following steps: Step S1: Collect multi-dimensional operating parameters of coal-fired units through a distributed control system, and simultaneously access the deep peak-shaving instructions sent by the grid dispatching end and the carbon price signal from the carbon emission trading market; Step S2: construct a five-dimensional safety constraint model associated with the economic cost boundary, including the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the secondary reheat temperature difference safety threshold; Step S3: With the goal of maximizing the revenue from peak-shaving auxiliary services, an objective function is constructed that includes coal procurement costs, carbon quota trading costs, and equipment maintenance costs. Constraints in each dimension are set based on a five-dimensional safety constraint model, and a dynamic programming algorithm is used to solve an economic dispatch plan that meets the requirements of the power market. Step S4: Based on the economic dispatch plan, implement cross-system economic coordination strategies, including fuel cost control, maintenance risk suppression, carbon trading linkage, energy consumption optimization, and thermal stress loss prevention and control; In step S5, LSTM-based equipment life prediction and digital twin simulation trigger cross-system correction strategies to maintain the safety boundary of commercial operations and reduce the risk of loss of peak-shaving auxiliary service revenue caused by equipment failure.

[0008] A deep dispatching system for deep peak regulation of coal-fired units, comprising: The data acquisition module collects multi-dimensional operating parameters of coal-fired units through a distributed control system. It then uses the edge computing unit to simultaneously access the deep peak-shaving instructions sent by the grid dispatcher and the carbon price signals from the carbon emissions trading market, achieving timestamp alignment of all parameters. The safety modeling module builds a five-dimensional safety constraint model, including the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the secondary reheat temperature difference safety threshold; The optimization solution module constructs a function to maximize the benefits of peak-shaving auxiliary services. Based on a five-dimensional safety constraint model, it sets constraints on load, environmental protection, steam turbine, secondary reheat, and carbon emissions. It uses a dynamic programming algorithm to solve the economic dispatch plan, including peak-shaving output allocation, carbon quota decision-making, and marginal benefit optimization. The collaborative control module implements cross-system economic collaborative strategies based on the economic dispatch scheme to achieve fuel cost control, maintenance risk suppression, carbon trading linkage, energy consumption optimization, and thermal stress loss prevention and control; The safety correction module predicts equipment life based on an LSTM neural network and generates early warning signals that include an estimate of the loss of revenue from peak-shaving ancillary services. By integrating the digital twin of the APROS and EBSILON platforms, it simulates the safety boundary breach scenario of the secondary reheat system, triggering cross-system correction strategies to ensure that the loss of revenue from peak-shaving ancillary services is minimized.

[0009] The beneficial effects brought about by the technical solution provided by the present invention include at least: The present invention constructs a five-dimensional safety constraint model associated with the economic cost boundary, integrates the safety boundaries of steady combustion load, denitrification temperature, turbine vibration, carbon emission intensity and secondary reheating temperature difference, and can achieve coordinated management and control of multi-dimensional safety indicators under deep peak-shaving conditions, avoid operating risks under a single safety constraint, and improve the operating stability of the unit in a wide load range.

[0010] The present invention uses a dynamic programming algorithm to solve the objective function that integrates the peak-shaving auxiliary service revenue, coal procurement costs, carbon quota trading costs and equipment maintenance costs, responds to carbon price signals and deep peak-shaving instructions, and can optimize the comprehensive economic benefits of the unit, achieving the unity of economy and safety of deep peak-shaving.

[0011] By implementing a cross-system economic coordination strategy and outputting coordinated control instructions to the combustion, steam turbine, environmental protection, thermal auxiliary equipment and secondary reheat systems, the present invention can solve the problem of separate control of boilers, steam turbines and environmental protection islands in traditional scheduling and realize multi-system linkage regulation.

[0012] The present invention uses an LSTM neural network to predict the remaining life of equipment and generate early warning signals containing estimates of peak-shaving revenue losses. Combined with the digital twin simulation safety boundary breakthrough scenarios integrated with the APROS and EBSILON platforms, it can achieve early warning and cross-system correction of equipment failures, reduce catalyst replacement costs and the risk of water erosion on turbine blades, and minimize peak-shaving revenue losses caused by equipment failures.

[0013] The present invention improves the accuracy of carbon emission intensity calculation under high-sulfur coal conditions through the sulfur-carbon emission correction coefficient, and coordinates the linkage adjustment of the secondary reheat system to improve the carbon emission calculation accuracy, reduce the carbon intensity per unit power generation, help the unit adapt to the requirements of the carbon trading market, and achieve low-carbon operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0015] Figure 1 A method flow chart of a deep scheduling method for deep peak regulation of a coal-fired unit provided in an embodiment of the present invention; Figure 2 A system architecture diagram of a deep scheduling system for deep peak regulation of coal-fired units provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a method and system for deep peak shaving of coal-fired power plants according to the present invention, including its specific implementation, structure, features, and effectiveness. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0017] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0018] The following examples are for illustrative purposes only and are not intended to limit the scope of the present invention.

[0019] The following describes in detail a specific scheme of a deep scheduling method and system for deep peak regulation of a coal-fired unit provided by the present invention with reference to the accompanying drawings.

[0020] See also Figure 1 , which shows a method flow chart of a deep scheduling method for deep peak regulation of a coal-fired unit provided by one embodiment of the present invention, the method comprising the following steps: Step S1: Collect multi-dimensional operating parameters of coal-fired units through a distributed control system, and simultaneously access the deep peak-shaving instructions sent by the grid dispatching end and the carbon price signal from the carbon emission trading market; Step S1 also includes the following sub-steps: S1-1, collect boiler parameters, including main steam pressure, main steam temperature, furnace pressure and oxygen content; S1-2, collect environmental island parameters, including SCR inlet flue gas temperature, desulfurization tower outlet sulfur dioxide concentration and dust collector outlet dust concentration; S1-3, collecting turbine parameters, including turbine vibration value, turbine vacuum degree and high-pressure cylinder exhaust temperature; S1-4, collecting fuel and load parameters, including real-time coal consumption, instantaneous coal feed signal of coal feeder and generator output power; S1-5, collects secondary reheat system operating parameters, including secondary reheat steam temperature, secondary reheat steam pressure, secondary reheat desuperheating water volume, and secondary reheat system bypass valve opening; S1-6 uses an edge computing unit to synchronously collect deep peak-shaving instructions and carbon price signals from the carbon emission trading market, establishes a communication connection with the distributed control system through the OPC UA protocol, and based on the real-time clock signal of the communication interaction, performs time stamp calibration on the deep peak-shaving instructions and the carbon price signals from the carbon emission trading market with the multi-dimensional operating parameters collected by the distributed control system to ensure the consistency of the time dimension of all parameters.

[0021] It should be noted that the distributed control system adopts Siemens PCS 7 or ABB AC 800M system with redundant configuration.

[0022] The furnace pressure in the boiler parameters is collected through a Rosemount 3051 pressure transmitter with an accuracy of ±0.1% FS and a sampling frequency of 100 Hz to capture combustion stability fluctuations.

[0023] The secondary reheat steam temperature is measured using an S-type thermocouple, which is used in conjunction with a temperature transmitter to achieve a measurement accuracy of ±1°C, meeting the thermal stress calculation requirements of the secondary reheat system.

[0024] The edge computing unit uses Advantech's UNO-3083G industrial computer with a built-in OPC UA client server (compliant with the IEC 62541 standard). Timestamp alignment is achieved through the following mechanisms: At the hardware level, it is equipped with a GPS clock module with a synchronization accuracy of ±1μs; At the software level, the IEEE 1588 precision clock protocol is adopted to add nanosecond timestamps to deep peak-shaving instructions (Modbus TCP format) and carbon price signals (JSON format), and data synchronization is achieved through the Linux system real-time scheduling kernel.

[0025] Furnace pressure collection points are located at the four corners of the boiler's burner area. The pressure sampling tubes are Φ14×2 stainless steel tubes with built-in anti-clogging filters. Fluctuations in the furnace negative pressure are measured using a differential pressure transmitter. When the pressure fluctuation exceeds ±50 Pa, a combustion stability warning is triggered, providing data support for the 20% load stable combustion boundary.

[0026] The SCR inlet flue gas temperature is collected using an array of thermocouples (averaging the temperature at six points), located 1 meter in front of the SCR reactor inlet to ensure temperature uniformity. When the flue gas temperature falls below 300°C, the staged heat recovery system automatically activates.

[0027] The opening of the bypass valve in the secondary reheat system is collected by a high-precision displacement sensor, forming a closed-loop monitoring with the desuperheating water volume (electromagnetic flowmeter, accuracy ±0.5%), providing real-time data for coordinated control of the bypass valve, solving the lag problem of secondary reheat temperature difference control in existing technologies.

[0028] The communication between the edge computing unit and the DCS adopts the OPC UA security policy, including: Transport layer, using TCP / IP protocol, port 4840 (standard OPC UA port); The security layer uses AES-256 encryption and X.509 certificate authentication to prevent tampering of carbon price signals; The data model follows the OPC UA data model mapping IEC 61850-7-420, mapping peak-shaving instructions to OPC UA nodes to achieve seamless integration with distributed control systems.

[0029] Timestamp alignment uses a two-way timestamp correction algorithm. The edge computing unit sends a synchronization frame (including the local GPS clock, with an accuracy of ±1μs) to the distributed control system every 100ms. The distributed control system returns a response frame with its own clock. The edge computing unit calculates the delay compensation value through the Kalman filter algorithm to ensure that the end-to-end synchronization error is ≤2ms (including communication delay and protocol processing time).

[0030] Step S2: construct a five-dimensional safety constraint model associated with the economic cost boundary, including the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the secondary reheat temperature difference safety threshold; In step S2, the following sub-steps are also included: S2-1, based on the balance between furnace pressure fluctuation and peak-shaving auxiliary service income, the upper limit of stable combustion load is set at 20% of rated load to ensure that the low-load coal procurement cost does not exceed the peak-shaving auxiliary service income; S2-2, based on the correlation between SCR inlet flue gas temperature and catalyst maintenance cost, the lower limit of SCR denitrification temperature is set at 300°C to avoid a surge in replacement costs due to insufficient catalyst activity; S2-3, based on the mapping relationship between turbine vibration values ​​and blade maintenance costs, sets the turbine vibration upper limit to 76 μm to balance the benefits of peak-shaving auxiliary services with equipment maintenance costs; S2-4, calculates the carbon emission intensity cap based on the sulfur dioxide concentration at the desulfurization tower outlet, the instantaneous coal feed signal of the coal feeder, and the generator output power. The carbon emission intensity cap is linked to the carbon quota transaction cost to ensure that the carbon cost is controllable. The formula is expressed as follows:

[0031] in, Indicates the carbon emission intensity per unit of electricity generation; Indicates standard coal Emission factors; Indicates real-time coal consumption; Indicates the generating power of the unit; Indicates the sulfur-carbon emission correction factor; S2-5, based on the difference between the secondary reheat steam temperature and the high-pressure cylinder exhaust temperature, the secondary reheat temperature difference safety threshold is set. The secondary reheat temperature difference safety threshold is associated with the equipment maintenance cost, and the formula is expressed as:

[0032] in, Indicates the secondary reheat temperature difference safety threshold; Indicates the secondary reheat steam temperature; Indicates the exhaust temperature of the high-pressure cylinder; ≤80℃ means the safety upper limit of the secondary reheat temperature difference is 80℃. When ΔT>80℃, the secondary reheat system thermal stress over-limit warning is triggered.

[0033] It should be noted that combustion stability is determined based on furnace pressure fluctuations. When the unit load falls below 20%, a sharp drop in furnace temperature can easily lead to unstable combustion. Real-time monitoring of furnace pressure fluctuations (if fluctuations exceed ±30 Pa, unstable combustion is identified) has been verified through hot tests. The 660MW unit can maintain stable combustion at 20% load (132MW). Therefore, 20% is set as the upper limit for stable combustion to avoid the risk of flameout at low load.

[0034] The lower limit of the SCR denitration temperature refers to the minimum inlet flue gas temperature required to ensure the activity of the SCR denitration catalyst. This temperature is set at 300°C in this invention. The activity of the SCR catalyst is highly dependent on the inlet flue gas temperature. When the flue gas temperature falls below 300°C, the catalyst's NOx conversion rate decreases significantly, potentially leading to an increase in ammonia slip and, in turn, air preheater blockage. This not only affects denitration effectiveness but also increases catalyst replacement costs.

[0035] The turbine vibration limit refers to the maximum vibration value allowed to prevent faults such as erosion on turbine blades. This invention sets it at 76 μm. The final blades of the turbine's low-pressure cylinder are susceptible to erosion due to steam-carrying water under low-load conditions. Vibration is a key indicator of erosion risk. According to relevant standards, 76 μm is the safe threshold for turbine bearing vibration. When the turbine vibration value exceeds this limit, the blade erosion rate will significantly accelerate, increasing equipment maintenance costs.

[0036] The calculation formula is:

[0037] The logarithmic relationship reflects The effect of concentration on carbon emissions shows a marginal decreasing effect, and the logarithmic form is more consistent with the actual chemical reaction kinetics; It represents the sulfur dioxide concentration at the outlet of the desulfurization tower; the coefficient 0.175 represents the slope coefficient of sulfur content influence; 0.98 represents the benchmark carbon conversion rate.

[0038] The secondary reheat temperature difference reflects the level of thermal stress. When ΔT exceeds 80°C, the reheater piping thermal stress exceeds the limit, potentially leading to creep rupture. 80°C is set as the threshold. When the temperature difference exceeds the limit, an alarm is triggered. Thermal stress is controlled by adjusting the desuperheating water flow and bypass valve opening to ensure equipment safety.

[0039] Step S3: With the goal of maximizing the revenue from peak-shaving auxiliary services, an objective function is constructed that includes coal procurement costs, carbon quota trading costs, and equipment maintenance costs. Constraints in each dimension are set based on a five-dimensional safety constraint model, and a dynamic programming algorithm is used to solve an economic dispatch plan that meets the requirements of the power market. In step S3, the following sub-steps are also included: S3-1, with the goal of maximizing the benefits of peak-shaving auxiliary services, construct an objective function that includes coal procurement costs, carbon quota trading costs, and equipment maintenance costs. The formula is expressed as:

[0040] in, It means cumulative calculation within the time range from time 1 to time N; max means finding the maximum value of the objective function; It represents the peak load compensation income obtained by the unit from participating in the power auxiliary service market; Indicates the coal purchase cost corresponding to the coal consumption; Indicates the carbon quota transaction cost corresponding to the carbon emissions of the unit; Represents the equipment maintenance cost, including the cost of secondary reheating efficiency loss; represents the objective function; The revenue from peak-shaving ancillary services is calculated in accordance with the market rules for power ancillary services and is implemented in different tiers. The formula for calculating the revenue from peak-shaving ancillary services is as follows:

[0041] in, Indicates the peak load output actually provided by the unit during the tth period; represents the peak load compensation unit price in the tth period; S3-2, based on the five-dimensional security constraint model, sets the corresponding conditions for each dimension constraint, including: Load constraint: Based on the upper limit of stable combustion load, the load variation range is set to 20%-100% of rated power to ensure that the load range of economic dispatch does not exceed the combustion stability limit; Environmental constraints, linked to the environmental performance requirements of the lower limit of the SCR denitrification temperature, set the NOx emission concentration to ≤25mg / Nm³ and the dust emission concentration to ≤2mg / Nm³, to ensure that environmental indicators are met during economic optimization; Steam turbine constraints, combined with the need to control equipment reliability based on the turbine vibration upper limit, set the load change rate to ≥1.5%Pe / min. This ensures peak load shaving service benefits while avoiding water erosion risks caused by excessive vibration. Secondary reheat constraints, based on the secondary reheat temperature difference safety threshold, set the secondary reheat steam temperature change rate to <1.5°C / min and the secondary reheat system bypass valve opening adjustment range to 10%-90%, ensuring that the thermal stress of the secondary reheat system during economic dispatch is within a safe range; Carbon emission constraints, linked to the low-carbon requirements of the carbon emission intensity cap, set carbon emission intensity ≤ carbon emission intensity cap to ensure that the economic scheduling plan meets the cost control objectives of the carbon trading market; S3-3 uses a dynamic programming algorithm to solve an economic dispatch plan that meets the requirements of the power market, including: Peak load distribution: Based on the deep peak load instruction and the upper limit of stable combustion load, the optimal output range for each period is predicted; Carbon quota decision-making, combining the carbon price signal and carbon emission intensity cap of the carbon emission rights trading market, analyzing the sensitivity of carbon quota transaction costs to the objective function, and determining the timing of buying and selling carbon quotas; Marginal benefit optimization, based on the economic relationship between turbine maintenance cost and secondary reheat maintenance cost, calculates the marginal balance point between equipment maintenance cost and peak-shaving auxiliary service income.

[0042] It should be noted that: in the scheduling process of deep peak shaving of coal-fired units, in order to achieve economic operation, the present invention constructs an objective function with the goal of maximizing the benefits of peak-shaving auxiliary services. This objective function comprehensively considers multiple economic factors such as coal procurement costs, carbon quota trading costs and equipment maintenance costs, and aims to improve the economic benefits of coal-fired units in the process of deep peak shaving by optimizing these cost factors. The objective function consists of four parts: peak-shaving auxiliary service benefits, coal procurement costs, carbon quota trading costs and equipment maintenance costs. Among them, peak-shaving auxiliary service benefits refer to the economic compensation obtained by the unit for participating in the power auxiliary service market, which is closely related to the peak-shaving capacity and market rules of the unit; coal procurement costs are directly related to coal consumption and coal prices; carbon quota trading costs depend on the carbon emissions of the unit and the carbon price in the carbon emission rights trading market; equipment maintenance costs cover the daily maintenance and overhaul of equipment and the additional equipment wear costs caused by deep peak shaving conditions.

[0043] Peak load ancillary service revenue is calculated based on the market rules for power ancillary services and is implemented in different tiers. Peak load compensation unit prices are set by load level, for example, 200 yuan / MWh for a 20%-30% load range, 100 yuan / MWh for a 30%-50% load range, and 30 yuan / MWh for a 50%-70% load range.

[0044] The coal purchase cost is calculated in real time based on the coal feeder's instantaneous coal feed signal and the main steam pressure / temperature. , combined with market coal prices Generate a cost term with the running time A , and adjusted through coal consumption correction coefficient under low load conditions.

[0045] Carbon quota trading costs are exported through scrubbers Concentration, real-time carbon price Dynamically calculate carbon emission intensity using power generation P , and introduce the sulfur-carbon emission correction factor for calibration, the final cost item is , digital twin correction is triggered when the carbon emission intensity exceeds 800g / kWh.

[0046] Equipment maintenance costs are primarily due to irreversible heat loss caused by excessive temperature differences. When ΔT exceeds the safety threshold (80°C), the thermal stress in the reheater pipe increases, and the steam enthalpy drop decreases, leading to an increase in the unit's heat rate. The calculation formula is:

[0047] Among them, ΔQ represents the efficiency loss of the reheat system, which is obtained by looking up ΔT in the table; B represents the low calorific value of standard coal; A represents the operating time; η represents the boiler efficiency, which is provided in real time by the distributed control system.

[0048] The constraints define the safe and economical operation boundaries of the unit, with a load range of 20%-100% to ensure combustion stability and denitrification efficiency, NOx ≤ 25mg / Nm³, dust ≤ 2mg / Nm³, in compliance with ultra-low emission requirements, variable load rate ≥ 1.5%Pe / min, to meet the rapid peak-shaving needs of the power grid, secondary reheat steam temperature change rate < 1.5℃ / min, secondary reheat system bypass valve opening adjustment range 10%-90%, to prevent thermal stress exceeding the limit and valve failure, and the carbon emission intensity upper limit is set to meet the low-carbon requirements of the carbon emission trading market. By setting carbon emission intensity ≤ carbon emission intensity upper limit, the carbon emissions of the unit can be reasonably controlled during the economic dispatch process to avoid problems such as increased carbon quota purchase costs due to excessive carbon emissions.

[0049] Dynamic programming is a classic optimization algorithm suitable for solving optimization problems involving multi-stage decision-making processes. In the economic scheduling problem of deep peak load regulation for coal-fired power plants, the scheduling process can be divided into multiple stages, each corresponding to a different time period or operating state. Dynamic programming can find the optimal solution to the overall problem by decomposing the original problem into a series of interrelated subproblems and solving each of these subproblems separately.

[0050] The dynamic programming algorithm predicts the optimal output range for each time period based on deep peak-shaving instructions and the upper limit of stable combustion load. It also optimizes output distribution by taking into account factors such as unit startup and shutdown costs to meet the grid's peak-shaving needs, ensure operational stability and economic efficiency, and maximize the benefits of peak-shaving ancillary services.

[0051] The dynamic programming algorithm combines carbon price signals from the carbon emissions trading market and the carbon emission intensity cap to analyze the sensitivity of carbon quota transaction costs to the objective function and determine the timing of buying and selling. Buying occurs when carbon prices are low and emissions are high, and selling occurs when carbon prices are high and emissions are low. This satisfies carbon emission constraints, optimizes transaction costs, and improves economic efficiency.

[0052] The dynamic programming algorithm, based on the economic relationship between turbine and secondary reheat maintenance costs, calculates the marginal break-even point between equipment maintenance costs and the revenue from peak-shaving ancillary services. Based on factors such as operating status, the maintenance plan and operating strategy are adjusted to achieve the optimal balance between costs and benefits, ensuring economical unit operation.

[0053] Step S4: Based on the economic dispatch plan, implement cross-system economic coordination strategies, including fuel cost control, maintenance risk suppression, carbon trading linkage, energy consumption optimization, and thermal stress loss prevention and control; In step S4, the following sub-steps are also included: S4-1: Based on the real-time ratio of coal purchase cost to peak load ancillary service income, switch the coal type and control the increase in coal purchase cost to not exceed the preset ratio of peak load ancillary service income; S4-2, when the turbine vibration value approaches the associated boundary of the blade maintenance cost, the throttling steam distribution mode is enabled to suppress the equipment maintenance cost; S4-3: Dynamically adjust the SCR denitrification efficiency according to the carbon price signal range of the carbon emission rights trading market, and control the denitrification cost and carbon quota trading cost to not exceed the preset proportion of peak-shaving auxiliary service income; S4-4, optimize the number of auxiliary units in operation of the thermal and auxiliary systems to ensure that the power saving benefits cover the equipment adjustment costs; S4-5, when the secondary reheat temperature difference approaches the safety threshold of the maintenance cost of the associated equipment, the secondary reheat desuperheating water and the secondary reheat system bypass valve are started to be linked to avoid economic losses from shutdown maintenance.

[0054] It should be noted that fuel cost control involves optimizing coal procurement and usage to reduce coal procurement costs while ensuring the unit's peak-shaving capacity and combustion stability. Real-time monitoring and adjustments ensure that increases in coal procurement costs do not exceed the growth in revenue from peak-shaving ancillary services, thereby maintaining the economic benefits of unit operation.

[0055] Maintenance risk mitigation involves monitoring equipment status and taking timely measures to reduce maintenance costs, thereby avoiding financial losses caused by equipment failures. By balancing the benefits of peak-shaving ancillary services with equipment maintenance costs, we can avoid financial losses from downtime and repairs due to equipment failures and ensure the long-term stable operation of the unit.

[0056] Carbon trading linkage dynamically adjusts denitrification efficiency to control carbon emission costs while meeting the low-carbon requirements of the carbon emissions trading market. By ensuring that denitrification costs and carbon quota trading costs do not exceed a predetermined proportion of peak-shaving ancillary service revenue, this ensures the controllability of carbon emission costs and improves the economic benefits of the units.

[0057] Energy consumption optimization refers to reducing energy consumption and improving the energy efficiency of the unit by optimizing the number of auxiliary units in operation. By optimizing the number of auxiliary units in operation, we ensure that the power savings cover the equipment adjustment costs and reduce the operating costs of the unit.

[0058] Thermal stress loss prevention and control involves adjusting the desuperheating water and bypass valves to prevent equipment damage caused by excessive thermal stress, ensuring safe equipment operation. This prevents economic losses from downtime and repairs caused by excessive thermal stress, ensuring continuous and stable operation of the unit and improving economic benefits.

[0059] Step S5: Based on LSTM-based equipment life prediction and digital twin simulation, cross-system correction strategies are triggered to maintain the safety boundary of commercial operations and reduce the risk of loss of peak-shaving auxiliary service revenue due to equipment failure. In step S5, the following sub-steps are also included: S5-1 uses an LSTM neural network to perform time series analysis on turbine vibration values, SCR catalyst activity decay data, boiler tube wall temperature, and secondary reheat system operating parameters to predict the remaining life of the turbine last-stage blades, reheater tube wall, and SCR catalyst. When the predicted life falls below a preset threshold, an early warning signal is generated, including an estimate of potential loss of peak-shaving auxiliary service revenue. S5-2 integrates the thermal system dynamic model with real-time operating data to build a digital twin that includes the secondary reheat system. The digital twin integrates the thermal system models of the APROS and EBSILON platforms to simulate the following scenarios: Dynamic distribution of secondary reheat temperature difference under deep peak regulation conditions; Quantification of economic losses when the operating parameters of the secondary reheat system fluctuate beyond the safety threshold; The decline in peak-shaving capacity caused by equipment degradation and the forecast of peak-shaving ancillary service revenue; S5-3: If the digital twin simulation detects a safety boundary breach, the following corrective measures are executed, including: When the secondary reheat temperature difference exceeds the safety threshold, the load change rate, secondary reheat desuperheating water volume, secondary reheat system bypass valve opening, and turbine steam distribution parameters are adjusted to balance thermal stress and output with the goal of minimizing the loss of peak-shaving auxiliary service revenue; When the turbine vibration value and the secondary reheat temperature difference exceed the limit, the blade anti-erosion protection is activated, the auxiliary engine drive mode is switched, and the heat source distribution is optimized to ensure that the increase in equipment maintenance costs does not exceed the preset proportion of the peak-shaving auxiliary service income; When the carbon emission intensity exceeds the standard, the heat exchange efficiency of the secondary reheat system and the operating parameters of the pulverizer are adjusted in a coordinated manner to make the carbon quota trading cost return to the controllable range of the objective function.

[0060] It should be noted that the activity of the SCR catalyst is inversely calculated based on the denitrification efficiency and ammonia escape rate. The denitrification efficiency is calculated based on the SCR inlet flue gas temperature and the dust concentration at the dust collector outlet, and the ammonia escape rate is calculated based on the environmental protection island parameters and the feedback data of the ammonia injection flow control instruction.

[0061] The boiler tube wall temperature is calculated through the model, and based on the operating parameters of the secondary reheat system, the boiler heating surface temperature field is inverted through the APROS digital twin.

[0062] LSTM neural networks are a special type of recurrent neural network that can effectively process and predict long-term dependencies in time series data. In this paper, the LSTM neural network is used to perform time series analysis on turbine vibration values, SCR catalyst activity decay data, boiler tube wall temperature, and secondary reheat system operating parameters to predict the remaining life of the turbine last-stage blades, reheater tube walls, and SCR catalyst.

[0063] The input data for the LSTM neural network includes turbine vibration values, SCR catalyst activity decay data, boiler tube wall temperature, and secondary reheat system operating parameters. This data is collected through the distributed control system and preprocessed before being input into the LSTM neural network.

[0064] LSTM neural networks predict the remaining life of equipment by analyzing the time series of input data and learning long-term dependencies within the data. The network is trained based on historical data, and the error between the predicted and actual remaining life is minimized by adjusting the network's weights and biases.

[0065] When the predicted lifespan falls below a preset threshold, the LSTM neural network generates a warning signal containing an estimate of potential loss of revenue from peak-shaving services. This warning signal is transmitted to the control center via the system's internal communication module. The warning signal includes the device name, predicted remaining lifespan, and an estimate of potential loss of revenue from peak-shaving services.

[0066] A digital twin is a virtual model that can reflect and predict the state and behavior of a physical system in real time. In this paper, the digital twin is used to simulate scenarios such as the dynamic distribution of secondary reheat temperature differences under deep peak-shaving conditions, quantify economic losses when secondary reheat system operating parameter fluctuations exceed safety thresholds, and predict the decline in peak-shaving capacity caused by equipment degradation and the revenue generated by peak-shaving ancillary services.

[0067] The digital twin can simulate the temporal evolution of the secondary reheat temperature difference under deep peak-shaving conditions, helping operators understand the thermal stress trends in the secondary reheat system. It can also predict the economic losses incurred when secondary reheat system operating parameters fluctuate beyond safety thresholds, including equipment repair costs, downtime losses, and lost revenue from peak-shaving ancillary services. The digital twin can assess the impact of equipment degradation on peak-shaving capacity and predict changes in peak-shaving ancillary service revenue, providing a basis for adjusting equipment maintenance and operating strategies.

[0068] Trigger cross-system correction strategies based on digital twin simulation results.

[0069] See also Figure 2 , which shows a system architecture diagram of a deep scheduling system for deep peak regulation of coal-fired units provided by one embodiment of the present invention, including: The data acquisition module collects multi-dimensional operating parameters of coal-fired units through a distributed control system. It then uses the edge computing unit to simultaneously access the deep peak-shaving instructions sent by the grid dispatcher and the carbon price signals from the carbon emissions trading market, achieving timestamp alignment of all parameters. The safety modeling module builds a five-dimensional safety constraint model, including the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the secondary reheat temperature difference safety threshold; The optimization solution module constructs a function to maximize the benefits of peak-shaving auxiliary services. Based on a five-dimensional safety constraint model, it sets constraints on load, environmental protection, steam turbine, secondary reheat, and carbon emissions. It uses a dynamic programming algorithm to solve the economic dispatch plan, including peak-shaving output allocation, carbon quota decision-making, and marginal benefit optimization. The collaborative control module implements cross-system economic collaborative strategies based on the economic dispatch scheme to achieve fuel cost control, maintenance risk suppression, carbon trading linkage, energy consumption optimization, and thermal stress loss prevention and control; The safety correction module predicts equipment life based on an LSTM neural network and generates early warning signals that include an estimate of the loss of revenue from peak-shaving ancillary services. By integrating the digital twin of the APROS and EBSILON platforms, it simulates the safety boundary breach scenario of the secondary reheat system, triggering cross-system correction strategies to ensure that the loss of revenue from peak-shaving ancillary services is minimized.

[0070] In this way, a deep scheduling method and system for deep peak regulation of coal-fired units can be realized.

[0071] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A deep scheduling method for deep peak regulation of coal-fired units, characterized in that: The method includes: Step S1: Collect multi-dimensional operating parameters of coal-fired units through a distributed control system, and simultaneously access the deep peak-shaving instructions sent by the grid dispatching end and the carbon price signal from the carbon emission trading market; Step S2: construct a five-dimensional safety constraint model associated with the economic cost boundary, including the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the secondary reheat temperature difference safety threshold; Step S3: With the goal of maximizing the revenue from peak-shaving auxiliary services, an objective function is constructed that includes coal procurement costs, carbon quota trading costs, and equipment maintenance costs. Constraints in each dimension are set based on a five-dimensional safety constraint model, and a dynamic programming algorithm is used to solve an economic dispatch plan that meets the requirements of the power market. Step S4: Based on the economic dispatch plan, implement cross-system economic coordination strategies, including fuel cost control, maintenance risk suppression, carbon trading linkage, energy consumption optimization, and thermal stress loss prevention and control; In step S5, LSTM-based equipment life prediction and digital twin simulation trigger cross-system correction strategies to maintain the safety boundary of commercial operations and reduce the risk of loss of peak-shaving auxiliary service revenue caused by equipment failure.

2. The deep scheduling method for deep peak regulation of coal-fired units according to claim 1, characterized in that: In step S1, the following sub-steps are also included: S1-1, collect boiler parameters, including main steam pressure, main steam temperature, furnace pressure and oxygen content; S1-2, collect environmental island parameters, including SCR inlet flue gas temperature, desulfurization tower outlet sulfur dioxide concentration and dust collector outlet dust concentration; S1-3, collecting turbine parameters, including turbine vibration value, turbine vacuum degree and high-pressure cylinder exhaust temperature; S1-4, collecting fuel and load parameters, including real-time coal consumption, instantaneous coal feed signal of coal feeder and generator output power; S1-5, collects secondary reheat system operating parameters, including secondary reheat steam temperature, secondary reheat steam pressure, secondary reheat desuperheating water volume, and secondary reheat system bypass valve opening; S1-6 uses an edge computing unit to synchronously collect deep peak-shaving instructions and carbon price signals from the carbon emission trading market, establishes a communication connection with the distributed control system through the OPC UA protocol, and based on the real-time clock signal of the communication interaction, performs time stamp calibration on the deep peak-shaving instructions and the carbon price signals from the carbon emission trading market and the multi-dimensional operating parameters collected by the distributed control system to ensure the consistency of the time dimension of all parameters.

3. The deep scheduling method for deep peak regulation of coal-fired units according to claim 1, characterized in that: In step S2, the following sub-steps are also included: S2-1, based on the balance between furnace pressure fluctuation and peak-shaving auxiliary service income, the upper limit of stable combustion load is set at 20% of rated load to ensure that the low-load coal procurement cost does not exceed the peak-shaving auxiliary service income; S2-2, based on the correlation between SCR inlet flue gas temperature and catalyst maintenance cost, the lower limit of SCR denitrification temperature is set at 300°C to avoid a surge in replacement costs due to insufficient catalyst activity; S2-3, based on the mapping relationship between turbine vibration values ​​and blade maintenance costs, sets the turbine vibration upper limit to 76 μm to balance the benefits of peak-shaving auxiliary services with equipment maintenance costs; S2-4, calculate the carbon emission intensity cap based on the sulfur dioxide concentration at the desulfurization tower outlet, the instantaneous coal feed signal of the coal feeder, and the generator output power. The carbon emission intensity cap is linked to the carbon quota transaction cost to ensure that the carbon cost is controllable. The formula is expressed as follows: in, Indicates the carbon emission intensity per unit of electricity generation; Indicates standard coal Emission factors; Indicates real-time coal consumption; Indicates the generating power of the unit; Indicates the sulfur-carbon emission correction factor; S2-5, based on the difference between the secondary reheat steam temperature and the high-pressure cylinder exhaust steam temperature, a secondary reheat temperature difference safety threshold is set. The secondary reheat temperature difference safety threshold is associated with the equipment maintenance cost, and the formula is expressed as: in, Indicates the secondary reheat temperature difference safety threshold; Indicates the secondary reheat steam temperature; Indicates the exhaust temperature of the high-pressure cylinder; ≤80℃ means the safety upper limit of the secondary reheat temperature difference is 80℃. When ΔT>80℃, the secondary reheat system thermal stress over-limit warning is triggered.

4. The deep scheduling method for deep peak regulation of coal-fired units according to claim 1, characterized in that: In step S3, the following sub-steps are also included: S3-1, with the goal of maximizing the benefits of peak-shaving auxiliary services, construct an objective function that includes coal procurement costs, carbon quota trading costs, and equipment maintenance costs. The formula is expressed as: in, It means cumulative calculation within the time range from time 1 to time N; max means finding the maximum value of the objective function; It represents the peak load compensation income obtained by the unit from participating in the power auxiliary service market; Indicates the coal purchase cost corresponding to the coal consumption; Indicates the carbon quota transaction cost corresponding to the carbon emissions of the unit; represents the equipment maintenance cost; represents the objective function; The peak-shaving auxiliary service revenue is calculated in conjunction with the power auxiliary service market rules and is implemented in different tiers. The peak-shaving auxiliary service revenue calculation formula is expressed as follows: in, Indicates the peak load output actually provided by the unit during the tth period; represents the peak load compensation unit price in the tth period; S3-2, based on the five-dimensional security constraint model, sets the corresponding conditions for each dimension constraint, including: Load constraint: Based on the upper limit of stable combustion load, the load variation range is set to 20%-100% of rated power to ensure that the load range of economic dispatch does not exceed the combustion stability limit; Environmental constraints, linked to the environmental performance requirements of the lower limit of the SCR denitrification temperature, set the NOx emission concentration to ≤25mg / Nm³ and the dust emission concentration to ≤2mg / Nm³, to ensure that environmental indicators are met during economic optimization; Steam turbine constraints, combined with the need to control equipment reliability based on the turbine vibration upper limit, set the load change rate to ≥1.5%Pe / min. This ensures peak load shaving service benefits while avoiding water erosion risks caused by excessive vibration. Secondary reheat constraints, based on the secondary reheat temperature difference safety threshold, set the secondary reheat steam temperature change rate to <1.5°C / min and the secondary reheat system bypass valve opening adjustment range to 10%-90%, ensuring that the thermal stress of the secondary reheat system during economic dispatch is within a safe range; Carbon emission constraints, linked to the low-carbon requirements of the carbon emission intensity cap, set carbon emission intensity ≤ carbon emission intensity cap to ensure that the economic scheduling plan meets the cost control objectives of the carbon trading market; S3-3 uses a dynamic programming algorithm to solve an economic dispatch plan that meets the requirements of the power market, including: Peak load distribution: Based on the deep peak load instruction and the upper limit of stable combustion load, the optimal output range for each period is predicted; Carbon quota decision-making, combining the carbon price signal and carbon emission intensity cap of the carbon emission rights trading market, analyzing the sensitivity of carbon quota transaction costs to the objective function, and determining the timing of buying and selling carbon quotas; Marginal benefit optimization, based on the economic relationship between turbine maintenance cost and secondary reheat maintenance cost, calculates the marginal balance point between equipment maintenance cost and peak-shaving auxiliary service income.

5. The deep scheduling method for deep peak regulation of coal-fired units according to claim 1, characterized in that: In step S4, the following sub-steps are also included: S4-1: Based on the real-time ratio of coal purchase cost to peak load ancillary service income, switch the coal type and control the increase in coal purchase cost to not exceed the preset ratio of peak load ancillary service income; S4-2, when the turbine vibration value approaches the associated boundary of the blade maintenance cost, the throttling steam distribution mode is enabled to suppress the equipment maintenance cost; S4-3: Dynamically adjust the SCR denitrification efficiency according to the carbon price signal range of the carbon emission rights trading market, and control the denitrification cost and carbon quota trading cost to not exceed the preset proportion of peak-shaving auxiliary service income; S4-4, optimize the number of auxiliary units in operation of the thermal and auxiliary systems to ensure that the power saving benefits cover the equipment adjustment costs; S4-5, when the secondary reheat temperature difference approaches the safety threshold of the maintenance cost of the associated equipment, the secondary reheat desuperheating water and the secondary reheat system bypass valve are started to be linked to avoid economic losses from shutdown maintenance.

6. The deep scheduling method for deep peak regulation of coal-fired units according to claim 1, characterized in that: In step S5, the following sub-steps are also included: S5-1 uses an LSTM neural network to perform time series analysis on turbine vibration values, SCR catalyst activity decay data, boiler tube wall temperature, and secondary reheat system operating parameters to predict the remaining life of the turbine last-stage blades, reheater tube wall, and SCR catalyst. When the predicted life falls below a preset threshold, an early warning signal is generated, including an estimate of potential loss of peak-shaving auxiliary service revenue. S5-2 integrates the thermal system dynamic model with real-time operating data to build a digital twin of the secondary reheat system. The digital twin integrates the thermal system models of the APROS and EBSILON platforms to simulate the following scenarios: Dynamic distribution of secondary reheat temperature difference under deep peak regulation conditions; Quantification of economic losses when the operating parameters of the secondary reheat system fluctuate beyond the safety threshold; The decline in peak-shaving capacity caused by equipment degradation and the forecast of peak-shaving ancillary service revenue; S5-3: If the digital twin simulation detects a safety boundary breach, the following corrective measures are executed, including: When the secondary reheat temperature difference exceeds the safety threshold, the load change rate, secondary reheat desuperheating water volume, secondary reheat system bypass valve opening, and turbine steam distribution parameters are adjusted to balance thermal stress and output with the goal of minimizing the loss of peak-shaving auxiliary service revenue; When the turbine vibration value and the secondary reheat temperature difference exceed the limit, the blade anti-erosion protection is activated, the auxiliary engine drive mode is switched, and the heat source distribution is optimized to ensure that the increase in equipment maintenance costs does not exceed the preset proportion of the peak-shaving auxiliary service income; When the carbon emission intensity exceeds the standard, the heat exchange efficiency of the secondary reheat system and the operating parameters of the pulverizer are adjusted in a coordinated manner to make the carbon quota trading cost return to the controllable range of the objective function.

7. A deep scheduling system for deep peak shaving of coal-fired units, used to implement the deep scheduling method for deep peak shaving of coal-fired units according to claim 1, characterized in that: include: The data acquisition module collects multi-dimensional operating parameters of coal-fired units through a distributed control system. It then uses the edge computing unit to simultaneously access the deep peak-shaving instructions sent by the grid dispatcher and the carbon price signals from the carbon emissions trading market, achieving timestamp alignment of all parameters. The safety modeling module builds a five-dimensional safety constraint model, including the upper limit of stable combustion load, the lower limit of SCR denitrification temperature, the upper limit of turbine vibration, the upper limit of carbon emission intensity, and the secondary reheat temperature difference safety threshold; The optimization solution module constructs a function to maximize the benefits of peak-shaving auxiliary services. Based on a five-dimensional safety constraint model, it sets constraints on load, environmental protection, steam turbine, secondary reheat, and carbon emissions. It uses a dynamic programming algorithm to solve the economic dispatch plan, including peak-shaving output allocation, carbon quota decision-making, and marginal benefit optimization. The collaborative control module implements cross-system economic collaborative strategies based on the economic dispatch scheme to achieve fuel cost control, maintenance risk suppression, carbon trading linkage, energy consumption optimization, and thermal stress loss prevention and control; The safety correction module predicts equipment life based on an LSTM neural network and generates early warning signals that include an estimate of the loss of revenue from peak-shaving ancillary services. By integrating the digital twin of the APROS and EBSILON platforms, it simulates the safety boundary breach scenario of the secondary reheat system, triggering cross-system correction strategies to ensure that the loss of revenue from peak-shaving ancillary services is minimized.

Images