Supercritical unit AGC coordinated control optimization method

Abstract

The invention discloses an AGC coordinated control optimization method for a supercritical unit. The method comprises the following steps: firstly, carrying out data collection and preprocessing; secondly, constructing a supercritical unit high-precision digital twinborn body; then designing and developing an AGC coordination control strategy for the supercritical unit; constructing and simulating a supercritical unit model based on a digital twinning technology; calculating and simulating field working conditions, and tracking the field working conditions in real time; and finally, carrying out optimization, testing and system verification on the AGC coordination control strategy. According to the AGC coordinated control optimization method for the supercritical unit, coordinated control optimization and accurate testing of the unit under multiple working conditions and multiple business states can be achieved, the AGC index adjusting performance and the unit operation reliability are effectively improved, and a powerful guarantee is provided for efficient and stable operation of the supercritical combined heat and power generation unit.

Description

Technical Field

[0001] This invention relates to the field of coordinated control technology for supercritical units, and in particular to an optimization method for coordinated control of AGC (Automatic Generative Control) for supercritical units. Background Technology

[0002] With the growth of energy demand and the development of the power industry, supercritical units play an important role in power production. However, supercritical units are typical complex thermodynamic systems characterized by nonlinearity, multiple parameters, and strong coupling, posing numerous challenges to their operation and control. Traditional control methods are insufficient to meet the requirements of modern power systems for unit flexibility, reliability, and efficiency.

[0003] Existing technologies for AGC coordinated control have certain limitations. Some generating units exhibit low load tracking accuracy and slow adjustment rates when responding to AGC load commands, affecting the frequency stability of the power system. For example, some traditional control strategies, when faced with rapid load changes, fail to adequately consider the complex coupling relationships between the unit's subsystems, resulting in a lack of timely and accurate adjustment of control parameters such as fuel and water supply, leading to significant load deviations.

[0004] In the field of simulation technology, although some simulation models of combined heat and power (CHP) units exist, most models lack accuracy and fail to accurately reflect the actual operating conditions of the units. Especially when simulating complex and fault conditions, the accuracy and reliability of existing simulation models are insufficient, failing to provide effective support for optimized control and fault diagnosis of the units. For example, some simulation models, when simulating the operating characteristics of units under deep peak-shaving conditions, are oversimplified and cannot accurately predict the stability and safety of the units under low loads.

[0005] The emergence of digital twin technology has provided new ideas and methods for the coordinated control optimization of supercritical units. By constructing a high-precision digital twin, accurate simulation and prediction of the unit's operating status can be achieved, providing a reliable basis for the formulation of optimized control strategies. However, research on coordinated control optimization test systems for 350MW supercritical units based on digital twin technology is still in its developmental stage, and a complete technical system has not yet been formed. Therefore, developing an efficient, accurate, and highly adaptable AGC coordinated control optimization method for supercritical units has significant practical implications. Summary of the Invention

[0006] The purpose of this invention is to provide an AGC coordinated control optimization method for supercritical units, which can realize unit coordinated control optimization and accurate testing under multiple operating conditions and business modes, effectively improve the adjustment performance of AGC indicators and the reliability of unit operation, and provide a strong guarantee for the efficient and stable operation of supercritical cogeneration units.

[0007] To achieve the above objectives, this invention provides an AGC coordinated control optimization method for supercritical units, comprising the following steps:

[0008] S1. Data collection and preprocessing;

[0009] S2. Construct a high-precision digital twin of the supercritical unit;

[0010] S3. Design and development of AGC coordinated control strategies for supercritical units;

[0011] S4. Construction and simulation of supercritical unit model based on digital twin technology;

[0012] S5. Calculate and simulate on-site working conditions, and track on-site working conditions in real time;

[0013] S6. Optimize, test, and verify the AGC coordination and control strategy.

[0014] Preferably, in S1, data collection includes collecting historical operating data of supercritical units under different operating conditions and business modes. Operating conditions include pure condensing, extraction condensing and cylinder cutting, and business modes include AGC and R mode, deep peak shaving and intraday start-up and shutdown.

[0015] The collected historical operating data of the supercritical unit includes the temperature parameter T. i Pressure parameter P j and flow parameter F k The temperature parameters include the main steam temperature and the reheat steam temperature; the pressure parameters include the main steam pressure and the steam drum pressure; and the flow parameters include the steam flow rate and the feedwater flow rate.

[0016] Preferably, in S1, the preprocessing of the collected historical operating data specifically involves: using data cleaning techniques to remove outliers and using Kalman filtering to remove noise fluctuations in the data.

[0017] Preferably, in S2, a high-precision digital twin of the supercritical unit is constructed based on digital twin technology and using a big data iterative coupling modeling method, including the following steps:

[0018] S21. Use principal component analysis to extract features from the preprocessed historical running data;

[0019] S22. Constructing a high-precision digital twin of a supercritical unit using a method that combines big data iterative coupling modeling and integration, including the following steps:

[0020] S221. Based on the laws of conservation of energy and mass, establish preliminary models for each subsystem of the unit. The subsystems include the boiler system, turbine system, DCS control system, instrumentation and control system, electrical system, utility system, and heating system.

[0021] S222. An iterative optimization algorithm is adopted, and the preliminary model of the overall unit is iteratively optimized based on actual operating data. The iterative formula is as follows:

[0022] M n+1 =M n +α·αE(M n );

[0023] Among them, M n Let α be the learning rate, and ΔE(M) be the model for the nth iteration. n ) represents the model correction amount calculated based on energy and mass errors.

[0024] Preferably, S3 is as follows:

[0025] A thorough analysis of the operating characteristics and control requirements of supercritical units under different business modes and conditions was conducted. Based on the analysis results, an optimization algorithm for the AGC integrated coordinated control strategy was determined. The optimization algorithm is based on the predictive control principle, and a predictive model for the predictive unit was constructed.

[0026]

[0027] Where y represents the unit output variables, including load and pressure; u represents the control input variables, including fuel quantity and feedwater quantity; k represents the time step; p and q represent the model order; and a represents the time step. i and b j The model coefficients are identified through historical data, and the optimal control sequence is solved with the objective of minimizing the deviation between the predicted output and the target output.

[0028] Preferably, in S4, a supercritical unit model is constructed based on digital twin technology, and the various subsystems of the unit are simulated in full range and process at a 1:1 scale. It has the functions of an excitation-type virtual simulator system, including the simulation of monitoring, operation and control functions in the unit's central control room, as well as the local operation in start-up, shutdown and fault handling outside the central control room, including the simulation of manual opening and closing of valves and start-up and shutdown of pumps.

[0029] Preferably, S5 includes the following steps:

[0030] S51. Determine the core parameters for online tracking, including unit load L and main steam pressure P. s and main steam temperature T s ;

[0031] S52. Through analysis of the unit's operating mechanism and statistical analysis of a large amount of data, an online simulation operating condition calculation model is derived based on the first and second laws of thermodynamics:

[0032] S = f(L, P) s T s ,...);

[0033] Where f is a functional relationship constructed based on the unit's operating mechanism, determined through extensive data fitting and theoretical analysis;

[0034] S53. Real-time tracking of on-site working conditions enables online simulation.

[0035] Preferably, in S6, the AGC coordinated control strategy and logic configuration are optimized, tested, and the system verified by comparing actual operating data with simulation data. The verification index is the tracking error.

[0036]

[0037] Where m is the number of data points, L t For the actual load, L sim The load is the simulated load, and t is the AGC adjustment time.

[0038] Therefore, the present invention employs the above-mentioned AGC coordinated control optimization method for supercritical units, and the beneficial effects are as follows:

[0039] (1) Significantly improve AGC regulation performance: Through precise digital twins and optimized AGC coordinated control strategies, the unit can respond quickly when the grid load command changes, the load tracking error can be controlled within a small range, the AGC regulation time is shortened, the stability and reliability of the power system are effectively improved, and the unit's ability to participate in grid frequency regulation is enhanced.

[0040] (2) Improve the flexibility and adaptability of unit operation: The method of the present invention can adapt to various operating conditions and business formats. Whether it is normal operation or special operating conditions (such as deep peak shaving, intraday start-up and shutdown, etc.), it can achieve good coordinated control. When switching between different operating conditions, the unit operates smoothly, reducing equipment losses and operating risks caused by changes in operating conditions, and improving the overall operational flexibility of the unit and its adaptability to complex operating environments.

[0041] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0042] Figure 1 This is an overall flowchart of an embodiment of the AGC coordinated control optimization method for supercritical units according to the present invention;

[0043] Figure 2 This is a flowchart illustrating the construction of a high-precision digital twin of a supercritical unit, according to an embodiment of the AGC coordinated control optimization method for supercritical units of the present invention.

[0044] Figure 3 This is a schematic diagram of the tracking field operating conditions in an embodiment of the supercritical unit AGC coordinated control optimization method of the present invention. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0046] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0047] As shown in the figure, an optimization method for AGC coordinated control of a supercritical unit includes the following steps:

[0048] S1. Data collection and preprocessing;

[0049] Data is collected from the thermal engineering training simulator system at the group-level power maintenance training base and the actual 350MW supercritical unit operation data acquisition system. This includes collecting historical operation data of various supercritical units under different operating conditions and business modes. The operating conditions include pure condensing, extraction condensing and cylinder cutting, while the business modes include AGC and R modes, deep peak shaving and intraday start-up and shutdown.

[0050] The collected historical operating data of the supercritical unit includes the temperature parameter T. i Pressure parameter P j and flow parameter F k The temperature parameters include the main steam temperature and the reheat steam temperature; the pressure parameters include the main steam pressure and the steam drum pressure; and the flow parameters include the steam flow rate and the feedwater flow rate.

[0051] Under pure condensation conditions, the temperature of the main steam collected is T. s The temperature range is 530℃-550℃, and the main steam pressure is P. s 24-26 MPa, steam flow rate F s Data such as extraction steam pressure P is collected under condensation conditions. c Steam extraction flow rate F c Data such as parameters; under deep peak shaving conditions, collect parameter data when operating at low load.

[0052] The preprocessing of collected historical operational data specifically involves, for example, for temperature data, if at a certain moment T... i If the temperature deviates from the average temperature under this operating condition by more than 3 times the standard deviation, data cleaning techniques are used to remove outliers, and Kalman filtering is used to remove noise fluctuations in the data.

[0053] S2. Based on digital twin technology, a high-precision digital twin of the supercritical unit is constructed using a big data iterative coupling modeling method, including the following steps:

[0054] S21. Use principal component analysis to extract features from the preprocessed historical operating data and discover the feature parameters that have a key impact on the unit's operating status.

[0055] S22. Constructing a high-precision digital twin of a supercritical unit using a method that combines big data iterative coupling modeling and integration, including the following steps:

[0056] S221. Based on the laws of conservation of energy and mass, establish preliminary models for each subsystem of the unit. The subsystems include the boiler system, turbine system, DCS control system, instrumentation and control system, electrical system, utility system, and heating system. For example, establish combustion and heat transfer models in the boiler system, and steam work model and speed regulation model in the turbine system.

[0057] Specifically, taking the boiler subsystem as an example, a combustion model is established based on the law of conservation of energy:

[0058] Q in =Q out +Q loss ;

[0059] Among them, Q in For the input of heat, Q out To output heat, Q loss For heat loss;

[0060] Input heat Q in With fuel quantity F f and fuel calorific value H f Related:

[0061] Q in =F f ·H f ;

[0062] Output heat and steam flow rate F s and vapor enthalpy H s Related:

[0063] Q out =F s ·H s ;

[0064] For the turbine subsystem, a model is established based on the principle of steam work, where steam performs work:

[0065] W = F s ·Δh;

[0066] Where Δh is the enthalpy difference between the steam inlet and outlet;

[0067] S222. An iterative optimization algorithm is adopted, and the preliminary model of the overall unit is iteratively optimized based on actual operating data. The iterative formula is as follows:

[0068] M n+1 =M n +α·ΔE(M n );

[0069] Among them, M n Let α be the learning rate, and ΔE(M) be the model for the nth iteration. n ) represents the model correction amount calculated based on energy and mass errors.

[0070] The model parameters are continuously adjusted to optimize and couple the models of each subsystem until the error between the model output and the actual operating data meets the preset accuracy requirements. For example, the boiler steam output is correlated with the steam consumption of the steam turbine. When the load prediction error is within ±2%, a high-precision digital twin is obtained.

[0071] S3. Design and develop AGC coordinated control strategies for supercritical units, specifically:

[0072] A thorough analysis of the operating characteristics and control requirements of supercritical units under different business modes and operating conditions reveals that, in AGC mode, the unit needs to respond quickly to grid load commands. d Based on the analysis results, an optimization algorithm for the AGC integrated coordinated control strategy was determined. This optimization algorithm is based on predictive control principles, and a predictive model for the predictive unit was constructed.

[0073]

[0074] Where y represents the unit output variables, including load and pressure; u represents the control input variables, including fuel quantity and feedwater quantity; k represents the time step; p and q represent the model order; and a represents the time step. i and b j The model coefficients are identified through historical data, and the optimal control sequence is solved with the objective of minimizing the deviation between the predicted output and the target output.

[0075] Assuming p = 3 and q = 2, identify a using historical data. i and b j To minimize the predicted load L p With target load L d The objective function is defined as the deviation and the change in control quantity as the target.

[0076]

[0077] Where N is the prediction time domain, λ is the weighting coefficient, the optimal control sequence is solved, the adjustment strategy of control quantities such as fuel quantity and water supply is determined, and the logic configuration is performed.

[0078] S4. Construction and simulation of supercritical unit model based on digital twin technology;

[0079] A 350MW supercritical unit model was constructed based on digital twin technology. At a 1:1 scale, full-range, full-process simulation of each subsystem of the unit was performed. The model possesses the functionality of an excitation-based virtual simulator system, including simulation of monitoring, operation, and control functions within the unit's central control room, as well as local operations outside the control room for start-up, shutdown, and fault handling, including simulation of manual valve opening and closing and pump start-up and shutdown functions. For example, adjusting the valve opening degree O... v Change the flow rate according to the control logic F = K·O v K is the flow coefficient.

[0080] The simulation system provides comprehensive simulation of DCS, steam turbine, boiler, instrumentation and control system, electrical system, utility system (including desulfurization and denitrification), and heating system. It can also perform local 3D simulation of electrical equipment, allowing operators to intuitively understand the layout and operating status of electrical equipment.

[0081] In fault condition simulation, such as simulating a turbine sensor fault, the probability P of the fault occurrence is determined based on fault tree analysis. f And the degree of impact of the fault I f Modify the corresponding model parameters.

[0082] S5. Calculate and simulate the on-site working conditions, and track the on-site working conditions in real time, including the following steps:

[0083] S51. Determine the core parameters for online tracking, including unit load L and main steam pressure P. s and main steam temperature T s ;

[0084] S52. Through analysis of the unit's operating mechanism and statistical analysis of a large amount of data, an online simulation operating condition calculation model is derived based on the first and second laws of thermodynamics:

[0085] S = f(L, P) s T s ,...);

[0086] Where f is a functional relationship constructed based on the unit's operating mechanism, determined through extensive data fitting and theoretical analysis; the derivation of this formula is based on the law of conservation of energy and the unit's dynamic characteristic equations. For example, the relationship between unit load and main steam flow is derived from the principle of steam doing work in the turbine and the energy conversion relationship:

[0087] L = F s ·Δh·η

[0088] Where η is the unit efficiency, and the parameters in the function f are determined by data fitting.

[0089] S53. Real-time tracking of on-site working conditions enables online simulation.

[0090] Under high load conditions, the core parameter data of the on-site working conditions are collected in real time, substituted into the above relational model, and the online simulation working conditions are calculated and compared with the actual working conditions. Based on the comparison results, the digital twin and control strategy are adjusted to achieve real-time tracking of the on-site working conditions.

[0091] S6. Optimize, test, and verify the AGC coordination and control strategy;

[0092] The AGC coordinated control strategy and logic configuration were optimized, tested, and verified by comparing actual operating data with simulation data.

[0093] The verification metric is tracking error:

[0094]

[0095] Where m is the number of data points, L t For the actual load, L sim The load is the simulated load, and t is the AGC adjustment time.

[0096] If the unit is operated under different operating conditions, record the actual load L. t Pressure data P t Simultaneously, obtain the corresponding simulation data L from the simulation system. t，sim P t，sim .

[0097] Calculate and verify indicators, such as the online simulation model for load tracking error:

[0098]

[0099] If the online simulation working condition calculation model E L If the online simulation operating condition calculation model exceeds the preset value, adjust the weight coefficients in the coordinated control strategy of the online simulation operating condition calculation model AGC or the model parameters of the digital twin, and retest and verify until the requirements are met. For example, the tracking error E... L Less than ±2% of the online simulation working condition calculation model, and the adjustment time of the AGC online simulation working condition calculation model is less than 30 seconds, etc.

[0100] Therefore, the above-mentioned supercritical unit AGC coordinated control optimization method can effectively improve the load tracking accuracy and AGC adjustment rate of the unit, enhance the adaptability and flexibility of the unit in complex operating environments, reduce the operating risk of the unit, and extend the service life of the unit.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for optimizing AGC coordinated control in a supercritical unit, characterized in that, Includes the following steps: S1. Data collection and preprocessing; S2. Construct a high-precision digital twin of the supercritical unit; S3. Design and development of AGC coordinated control strategies for supercritical units; S4. Construction and simulation of supercritical unit model based on digital twin technology; S5. Calculate and simulate on-site working conditions, and track on-site working conditions in real time; S6. Optimize, test, and verify the AGC coordination and control strategy.

2. The AGC coordinated control optimization method for supercritical units according to claim 1, characterized in that, In S1, data collection includes collecting historical operating data of supercritical units under different operating conditions and business modes. Operating conditions include pure condensing, extraction condensing and cylinder cutting, and business modes include AGC and R modes, deep peak shaving and intraday start-up and shutdown. The collected historical operating data of the supercritical unit includes the temperature parameter T. i Pressure parameter P j and flow parameter F k The temperature parameters include the main steam temperature and the reheat steam temperature; the pressure parameters include the main steam pressure and the steam drum pressure; and the flow parameters include the steam flow rate and the feedwater flow rate.

3. The AGC coordinated control optimization method for supercritical units according to claim 2, characterized in that, In S1, the preprocessing of the collected historical operating data specifically involves: using data cleaning techniques to remove outliers and using Kalman filtering to remove noise fluctuations in the data.

4. The AGC coordinated control optimization method for supercritical units according to claim 3, characterized in that, In S2, a high-precision digital twin of the supercritical unit is constructed based on digital twin technology and using a big data iterative coupling modeling method, including the following steps: S21. Use principal component analysis to extract features from the preprocessed historical running data; S22. Constructing a high-precision digital twin of a supercritical unit using a method that combines big data iterative coupling modeling and integration, including the following steps: S221. Based on the laws of conservation of energy and mass, establish preliminary models for each subsystem of the unit. The subsystems include the boiler system, turbine system, DCS control system, instrumentation and control system, electrical system, utility system, and heating system. S222. An iterative optimization algorithm is adopted, and the preliminary model of the overall unit is iteratively optimized based on actual operating data. The iterative formula is as follows: M n+1 =M n +α·ΔE(M n ); Among them, M n Let α be the learning rate, and ΔE(M) be the model for the nth iteration. n ) represents the model correction amount calculated based on energy and mass errors.

5. The AGC coordinated control optimization method for supercritical units according to claim 4, characterized in that, S3 specifically refers to: A thorough analysis of the operating characteristics and control requirements of supercritical units under different business modes and conditions was conducted. Based on the analysis results, an optimization algorithm for the AGC integrated coordinated control strategy was determined. The optimization algorithm is based on the predictive control principle, and a predictive model for the predictive unit was constructed. Where y represents the unit output variables, including load and pressure; u represents the control input variables, including fuel quantity and feedwater quantity; k represents the time step; p and q represent the model order; and a represents the time step. i and b j The model coefficients are identified through historical data, and the optimal control sequence is solved with the objective of minimizing the deviation between the predicted output and the target output.

6. The AGC coordinated control optimization method for supercritical units according to claim 5, characterized in that, In S4, a supercritical unit model is constructed based on digital twin technology. The various subsystems of the unit are simulated in full range and process at a 1:1 scale. It has the functions of an excitation-type virtual simulator system, including the simulation of monitoring, operation and control functions in the unit's central control room, as well as local operations in start-up, shutdown and fault handling outside the central control room, including the simulation of manual opening and closing of valves and start-up and shutdown of pumps.

7. The AGC coordinated control optimization method for supercritical units according to claim 6, characterized in that, S5 includes the following steps: S51. Determine the core parameters for online tracking, including unit load L and main steam pressure P. s and main steam temperature T s ; S52. Through analysis of the unit's operating mechanism and statistical analysis of a large amount of data, an online simulation operating condition calculation model is derived based on the first and second laws of thermodynamics: S=f(L,P s ,T s ,...); Where f is a functional relationship constructed based on the unit's operating mechanism, determined through extensive data fitting and theoretical analysis; S53. Real-time tracking of on-site working conditions enables online simulation.

8. The AGC coordinated control optimization method for supercritical units according to claim 7, characterized in that, In S6, the AGC coordinated control strategy and logic configuration are optimized, tested, and validated by comparing actual operating data with simulation data. The validation metric is tracking error. Where m is the number of data points, L t For the actual load, L sim The load is the simulated load, and t is the AGC adjustment time.

Images