Auxiliary frequency modulation control system for thermal power unit based on multi-source information fusion

Abstract

This invention relates to the field of automatic control technology for thermal power generating units, and discloses an auxiliary frequency regulation control system for thermal power generating units based on multi-source information fusion. The system includes a multi-source data acquisition module, a data preprocessing module, a multi-source information fusion module, an intelligent analysis module, an auxiliary frequency regulation control module, a full-cycle health management module, a predictive maintenance module, and a human-machine interaction module. By collecting multi-dimensional data on unit operation, and after preprocessing and fusion, the system calculates the frequency regulation contribution matching degree 'a' and the power deviation rate, which are used to dynamically optimize boiler combustion, turbine speed control valves, and coordinated control strategies to achieve second-level rapid response. Simultaneously, based on the remaining health cycle of the equipment, a full life cycle state evolution model of the equipment is constructed, a hierarchical early warning mechanism is established, and the predictive maintenance module is linked to generate maintenance strategies and optimize maintenance plans and spare parts inventory, thereby comprehensively improving the response speed, control accuracy, and operational reliability of the system's auxiliary frequency regulation.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for thermal power generating units, specifically to an auxiliary frequency regulation control system for thermal power generating units based on multi-source information fusion. Background Technology

[0002] With the large-scale grid connection of new energy sources (wind power and photovoltaics), the power grid's requirements for the rapid and precise frequency regulation capabilities of thermal power units are increasing. As an important support for power grid frequency regulation, the auxiliary frequency regulation performance of thermal power units is directly related to the safe and stable operation of the power grid.

[0003] However, existing auxiliary frequency regulation control systems for thermal power units suffer from the following technical deficiencies: First, data sources are limited, resulting in insufficient sensing capabilities. Traditional control systems primarily rely on a single grid frequency deviation signal or AGC command for adjustment, lacking comprehensive sensing of multi-source information such as the unit's internal combustion state, turbine parameters, and equipment health. This leads to incomplete control decision-making information and makes it difficult to achieve refined frequency regulation. Second, inherent response lag exists. Thermal power units rely on boiler combustion and mechanical inertia for energy conversion, resulting in a long energy conversion chain. Under traditional control strategies, it takes several seconds to tens of seconds after a frequency regulation command is issued to generate an effective power response, which is insufficient to meet the grid's requirements for second-level or even sub-second-level responses. Third, the evaluation system is one-sided and lacks closed-loop optimization. Existing technologies rely heavily on single power differences or instantaneous deviations to evaluate frequency regulation effects, making it difficult to quantify the unit's "actual contribution" to grid frequency fluctuations. They lack effective sensing of small-frequency fluctuations and lack a closed-loop mechanism of "evaluation-feedback-optimization," hindering continuous improvement in control performance. Fourth, equipment health and frequency regulation control are disconnected. Existing systems lack the ability to monitor and predict the lifespan of critical equipment (such as feedwater pumps, coal mills, and fans) throughout their entire lifecycle. Equipment maintenance relies on periodic inspections or post-failure repairs, failing to achieve predictive maintenance. Furthermore, frequency regulation control strategies do not consider equipment health constraints, making them susceptible to impacts on frequency regulation response capabilities due to sudden equipment failures. Fifth, safety and economy are difficult to balance. In pursuing rapid frequency regulation response, existing technologies often neglect equipment fatigue accumulation and energy consumption optimization, leading to shortened equipment lifespan, increased coal consumption, and a lack of multi-objective collaborative optimization techniques.

[0004] To address the shortcomings of the existing technology, this invention provides an auxiliary frequency regulation control system for thermal power units based on multi-source information fusion. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides an auxiliary frequency regulation control system for thermal power units based on multi-source information fusion. It has the advantages of holographic perception, rapid response, online quantitative evaluation, closed-loop self-optimization, full-cycle health management, and predictive maintenance. It solves the problems of existing technologies, such as single data source, delayed response, one-sided evaluation, separation of health and frequency regulation, and difficulty in coordinating safety and economy.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention provides the following technical solution: an auxiliary frequency regulation control system for thermal power units based on multi-source information fusion, comprising a multi-source data acquisition module, a data preprocessing module, a multi-source information fusion module, an intelligent analysis module, an auxiliary frequency regulation control module, a full-cycle health management module, a predictive maintenance module, and a human-machine interaction module;

[0009] The multi-source data acquisition module is used to collect multi-dimensional data during the operation of thermal power units, including power grid frequency regulation command data, unit operation data, and equipment health data.

[0010] The data preprocessing module is responsible for cleaning, denoising, time-stamp alignment, outlier removal, and normalization of the collected multi-source data to establish a data stream with a unified time reference.

[0011] The multi-source information fusion module uses an adaptive weighted fusion algorithm to perform spatiotemporal fusion of preprocessed multi-source data to construct a unified information framework that reflects the overall operating status of the unit.

[0012] The intelligent analysis module receives the fused data output by the multi-source information fusion module and calculates the frequency modulation contribution consistency. a. Power deviation rate and remaining health cycle of equipment At the same time, the fused data is deeply analyzed to identify the characteristics of small frequency fluctuations in the power grid and generate frequency regulation demand forecasts;

[0013] The auxiliary frequency modulation control module determines the frequency modulation contribution matching degree based on the output of the intelligent analysis module. Power deviation rate The calculation results are used to dynamically optimize boiler combustion control parameters, turbine speed control valve opening, and coordinated control strategies.

[0014] The full-cycle health management module is based on the remaining health cycle of the device. Based on the calculation results, a full life cycle state evolution model of the equipment is constructed, the degradation trend of key components is monitored in real time, and a health status classification and early warning mechanism is established.

[0015] The predictive maintenance module, based on a health status grading and early warning mechanism, combined with full-cycle health management data and a fault mode library, predicts the remaining service life of the equipment and generates maintenance strategy recommendations.

[0016] The human-machine interaction module provides a visual monitoring interface for the system, displaying the frequency modulation control effect, equipment health status and maintenance early warning information, and supporting operator intervention and system parameter configuration.

[0017] Preferably, the power grid frequency regulation command data includes: power grid frequency deviation signal, AGC dispatch command, regional control error, planned output curve, frequency regulation mileage price signal, primary frequency regulation dead zone threshold and secondary frequency regulation response time requirement.

[0018] Preferably, the unit operating data includes: boiler combustion status data, turbine operating parameters, generator electrical parameters, valve opening feedback data, coal feeder speed data, forced draft and induced draft data, main steam pressure and temperature data, condenser vacuum data, ambient temperature data, feedwater pump outlet pressure and flow rate, high-pressure heater water level, deaerator pressure and temperature, coal mill current and outlet air-coal temperature, air preheater inlet and outlet flue gas differential pressure, desulfurization tower pH value and slurry density, reheat steam temperature and pressure, and shaft seal steam pressure and temperature.

[0019] Preferably, the equipment health data includes: equipment vibration spectrum data, bearing temperature field data, lubricating oil quality data, ultrasonic testing data of key welds, thermal deformation of water pump and fan casing, turbine blade tip clearance value, generator stator winding partial discharge signal, fatigue characteristic frequency amplitude of high-voltage motor bearing, wear index of coal mill liner, blower blade crack propagation monitoring data, and high-temperature pipeline creep strain data.

[0020] Preferably, the intelligent analysis module receives the fused data output by the multi-source information fusion module and calculates the frequency modulation contribution consistency. The calculation formula is as follows: In the formula, This represents the actual output power of the generator unit, that is, the active power actually generated by the generator unit at time t. Indicates the power output of the generator set via AGC commands. This indicates the ideal power corresponding to the grid frequency. Indicates the initial power.

[0021] Preferably, the intelligent analysis module receives the fused data output by the multi-source information fusion module and calculates the power deviation rate. The calculation formula is as follows: In the formula, This represents the rated power of the unit, used as a normalization reference. This represents the actual power of the unit at the i-th sampling time. This indicates that the ideal power corresponds to the grid frequency at the i-th sampling time. Indicates the number of sampling points. The standard deviation of the power grid frequency deviation, This represents the mean value of the power grid frequency deviation. This represents the frequency fluctuation sensitivity coefficient.

[0022] Preferably, the intelligent analysis module receives the fused data output by the multi-source information fusion module to calculate the remaining health cycle of the device. The calculation formula is as follows: In the formula, Indicates the remaining health cycle of the equipment. Indicates the standard health cycle of the equipment. Indicates the attenuation coefficient. Indicates the number of equipment health monitoring indicators. This represents the current monitoring value of the j-th health indicator; This represents the standard threshold for the j-th health indicator.

[0023] Preferably, the auxiliary frequency modulation control module is based on the frequency modulation contribution matching degree. Power deviation rate The adjusted control strategy is as follows:

[0024] (1) Power deviation rate based on real-time monitoring Optimize boiler combustion control parameters: when power deviation rate When the deviation is greater than 5%, the feeder speed and forced and induced draft are adjusted in real time to increase combustion intensity and quickly improve the unit's output power, thus reducing the deviation between the actual power and the commanded power; when the power deviation rate... When the concentration is ≤2%, maintain the current combustion parameters stable;

[0025] (2) Power deviation rate based on real-time monitoring Optimize turbine speed control valve opening: when power deviation rate When the steam rate is greater than 3% and the power grid frequency regulation command requires a rapid response, the opening of the turbine speed control valve is finely adjusted to change the steam intake.

[0026] (3) Frequency modulation contribution consistency based on real-time monitoring Optimize the coordination control strategy: when the frequency modulation contribution matches When the value is less than 0.7, adjust the boiler-turbine coordinated control logic to improve the linkage response speed of both and at the same time improve the frequency regulation contribution matching degree. Numerical value; when the frequency modulation contribution matches When the value is ≥0.9, optimize the coordinated control parameters.

[0027] Preferably, the full-cycle health management module is based on the remaining health cycle of the device. The calculation results are used to classify the equipment health level: when the equipment has a remaining health cycle When the remaining health cycle is ≥8000h, it is considered to be in a healthy state. Maintain the regular monitoring frequency, focusing on tracking basic health indicators such as equipment vibration and temperature to ensure stable equipment operation. When the remaining health cycle is <4000h... If the time is less than 8000 hours, it is considered to be in a sub-healthy state. The monitoring frequency should be increased, and the degradation trend of key components should be closely monitored.

[0028] Preferably, the full-cycle health management module is based on the remaining health cycle of the device. Calculation results monitor the degradation trend of key components: targeting the remaining health life of the equipment For equipment with low values, we focus on monitoring its vibration spectrum data, bearing temperature field data, and lubricating oil quality data. By combining the state evolution model, we analyze the degradation rate of key components, predict potential failure points, and output a cycle prediction report.

[0029] Compared with the prior art, the present invention provides an auxiliary frequency regulation control system for thermal power units based on multi-source information fusion, which has the following beneficial effects:

[0030] 1. This invention calculates the frequency modulation contribution matching degree. Power deviation rate It is used to identify the characteristics of low-frequency fluctuations in the power grid and generate frequency regulation demand forecasts, thereby achieving refined frequency regulation quantitative assessment and closed-loop self-optimization across the entire frequency band. On the one hand, the frequency regulation contribution consistency... By comparing the integral of the deviation from the actual power tracking command power with the integral of the ideal power change corresponding to grid frequency disturbances, the true contribution of the generating unit to grid frequency fluctuations is quantified, overcoming the one-sidedness of existing technologies that only consider instantaneous deviations; on the other hand, the power deviation rate... The formula introduces a small-frequency fluctuation sensing factor, which automatically increases the weight of the power deviation rate when there are small frequency fluctuations in the power grid. This enhances the system's ability to quantitatively perceive small frequency changes in the power grid and avoids the problem of traditional methods ignoring small-frequency fluctuations. The two calculated indicators together constitute a multi-dimensional quantitative evaluation system. Combined with the frequency regulation effect closed-loop evaluation module, a closed-loop process of evaluation-feedback-optimization is formed, which continuously improves control performance.

[0031] 2. This invention calculates the remaining health cycle of the device. It also integrates with the full-lifecycle health management module and predictive maintenance module to achieve proactive health management and predictive maintenance throughout the equipment's entire lifecycle. This includes monitoring the remaining health cycle of the equipment. The calculation formula comprehensively considers the deviation of the current monitored values ​​of multiple health indicators (vibration, temperature, oil, etc.) from the standard threshold, combined with the attenuation coefficient. This enables dynamic quantitative prediction of the remaining health life of equipment, based on real-time monitoring of the remaining health life of the equipment. The system classifies health into four levels: green, yellow, orange, and red, establishes a graded early warning mechanism, and limits the load change rate based on a dynamic frequency adjustment strategy according to the degradation rate. The predictive maintenance module, combined with a fault mode library, predicts the remaining service life and generates targeted maintenance strategies, optimizing maintenance plans and spare parts inventory configuration. Attached Figure Description

[0032] Figure 1 This is a system flowchart of the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Please see Figure 1 The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion includes a multi-source data acquisition module, a data preprocessing module, a multi-source information fusion module, an intelligent analysis module, an auxiliary frequency regulation control module, a full-cycle health management module, a predictive maintenance module, and a human-machine interaction module. The module operation process is as follows:

[0035] The multi-source data acquisition module is used to collect multi-dimensional data during the operation of thermal power units, including power grid frequency regulation command data, unit operation data, and equipment health data;

[0036] The power grid frequency regulation command data is received in real time from the power grid dispatch center via the dispatch data network through remote terminal units (RTUs), phasor measurement units (PMUs), and AGC workstations. At the same time, the power plant-side PMU devices collect the power grid frequency locally, specifically including: power grid frequency deviation signal, AGC dispatch command, regional control error (ACE), planned output curve, frequency regulation mileage price signal, primary frequency regulation dead zone threshold, and secondary frequency regulation response time requirements (ACE is added for more accurate frequency regulation demand calculation; the planned curve is used to predict load trends; the price signal supports economic optimization; and the dead zone and response time requirements are used for adaptive control strategies).

[0037] Unit operating data is acquired in real time from field sensors via distributed control systems (DCS), programmable logic controllers (PLC), fieldbuses (Profibus / FF, etc.), and various intelligent transmitters. Specifically, this includes boiler combustion status data (furnace pressure, flame intensity, oxygen content, CO / NO). x Concentration), turbine operating parameters (high-pressure cylinder exhaust temperature, intermediate-pressure cylinder inlet pressure, low-pressure cylinder exhaust enthalpy, rotor axial displacement, expansion differential), generator electrical parameters (terminal voltage, reactive power, power angle, excitation current), valve opening feedback data, coal feeder speed data, forced draft and induced draft volume data, main steam pressure and temperature data, condenser vacuum data, ambient temperature data, feedwater pump outlet pressure and flow rate, high-pressure heater water level, deaerator pressure and temperature, coal mill current and outlet air-coal temperature, air preheater inlet and outlet flue gas differential pressure, desulfurization tower pH value and slurry density, reheat steam temperature and pressure, shaft seal steam pressure and temperature (this category of data fully covers the entire process of boiler, turbine, generator and auxiliary equipment, supporting accurate modeling and frequency regulation control of nonlinear and large hysteresis systems).

[0038] Equipment health data is collected through a Condition Monitoring System (CMS), an online oil monitoring system, a wireless sensor network, portable offline testing instruments (regularly recorded), and embedded intelligent sensors. Specifically, this includes: equipment vibration spectrum data (frequency domain amplitude, dominant frequency offset, kurtosis index), bearing temperature field data (multi-point temperature, temperature difference change rate), lubricating oil quality data (viscosity, moisture, acid value, particulate contamination), ultrasonic testing data of key welds, thermal deformation of feedwater pump and fan casings, turbine blade tip clearance, generator stator winding partial discharge signal, fatigue characteristic frequency amplitude of high-voltage motor bearings, wear index of coal mill liner plates, crack propagation monitoring data of blower blades, and creep strain data of high-temperature pipelines (the data covers multiple dimensions of health indicators such as mechanical vibration, thermal stress, oil, non-destructive testing, and electrical insulation, providing more comprehensive feature inputs for full-cycle life prediction and predictive maintenance).

[0039] The data preprocessing module is responsible for cleaning, denoising, time-stamp alignment, outlier removal, and normalization of the collected multi-source data, establishing a data stream with a unified time reference, and eliminating data synchronization problems caused by differences in sensor sampling frequencies and transmission delays.

[0040] The multi-source information fusion module uses an adaptive weighted fusion algorithm to perform spatiotemporal fusion of preprocessed multi-source data, constructing a unified information framework that reflects the overall operating status of the unit, and realizing deep correlation and complementary enhancement of data from the boiler side, turbine side, and electrical side.

[0041] The intelligent analysis module receives the fused data output from the multi-source information fusion module and calculates the frequency modulation contribution consistency. a. Power deviation rate and remaining health cycle of equipment At the same time, the fused data is deeply analyzed to form quantitative evaluation results, identify the characteristics of small frequency fluctuations in the power grid, and generate frequency regulation demand forecasts.

[0042] (1) Calculate the frequency modulation contribution matching degree The calculation formula is as follows: In the formula, This represents the actual output power of the generator unit, that is, the active power actually generated by the generator unit at time t. This indicates the AGC (Automatic Generation Control) command power of the generating unit, i.e., the target power command issued by the power grid dispatching system. This represents the ideal power output corresponding to the grid frequency, specifically the ideal power output of the generating unit calculated based on the grid frequency deviation. Indicates the initial power.

[0043] The advantage is that it calculates the frequency modulation contribution matching degree. This indicator quantifies the true contribution of generating units to grid frequency fluctuations, thereby overcoming the one-sidedness of existing technologies that only consider instantaneous deviations. Indicates the start time of the frequency modulation command. This indicator represents the duration of the frequency regulation assessment cycle. It quantifies the unit's true contribution to grid frequency fluctuations by comparing the integral of the deviation from the actual power tracking command power with the integral of the ideal power change corresponding to grid frequency disturbances. The numerator reflects the power tracking accuracy, and the denominator reflects the intensity of grid disturbances. A smaller ratio indicates a more accurate unit response. The closer it is to 1, the higher the degree of agreement of the frequency modulation contribution, thus overcoming the one-sidedness of existing technologies that only consider instantaneous deviation.

[0044] (2) Calculate the power deviation rate The calculation formula is as follows: In the formula, This represents the rated power of the unit, used as a normalization reference. This represents the actual power of the unit at the i-th sampling time (actual power at time i), that is, the actual output power of the unit at the discrete sampling points. This represents the ideal power corresponding to the grid frequency at the i-th sampling time (ideal power at time i), which is the ideal power that the unit should generate calculated based on the grid frequency at time i. This represents the number of sampling points, i.e., the total number of samplings within the assessment period. The standard deviation of the power grid frequency deviation reflects the degree of dispersion of frequency fluctuations. It represents the mean of the power grid frequency deviation, reflecting the central trend of the frequency offset. Indicates the frequency fluctuation sensitivity coefficient (0 < ≤1), used to adjust the intensity of low-frequency fluctuation perception.

[0045] The advantage is that it allows for the calculation of power deviation rate. This formula introduces a small-frequency fluctuation sensing factor. When there are small frequency fluctuations in the power grid ( relatively (If the value is relatively large), this factor automatically increases the weight of the power deviation rate, enhances the system's ability to quantitatively perceive small frequency changes in the power grid, avoids the shortcomings of traditional methods that ignore small frequency fluctuations, and achieves refined frequency regulation assessment across the entire frequency band.

[0046] (3) Calculate the remaining health cycle of the equipment. The calculation formula is as follows: In the formula, Indicates the remaining health cycle of the equipment (unit: hours). This indicates the standard health cycle of the equipment (unit: h, calibrated according to the equipment model and operating parameters). This represents the attenuation coefficient (derived by fitting historical health data and fault data of the equipment). Indicates the number of equipment health monitoring indicators (such as vibration, temperature, etc.); This represents the current monitoring value of the j-th health indicator; This represents the standard threshold for the j-th health indicator.

[0047] The advantage is that it calculates the remaining health cycle of the device. This method aims to achieve comprehensive quantitative assessment and dynamic prediction of the health status of key equipment in thermal power units based on multiple indicators. The formula comprehensively considers the deviation of current monitoring values ​​from standard thresholds for multiple health monitoring indicators such as vibration, temperature, and oil quality. It also employs an exponential decay model to characterize the nonlinear accelerated degradation characteristics of cumulative equipment damage. Compared to traditional methods based on fixed service life or a single threshold, this method can reflect the actual deterioration trend of equipment in real time. When multiple indicators deteriorate simultaneously, the summation portion of the exponential term automatically increases, accelerating the remaining health cycle of the equipment. The attenuation of power can trigger a health warning in a timely manner, providing equipment constraint boundaries (such as limiting the rate of load change) for the auxiliary frequency control module. At the same time, it outputs power deviation warning information and equipment health status assessment results, thereby providing accurate data support for subsequent frequency control and periodic prediction, and realizing proactive predictive maintenance.

[0048] The auxiliary frequency modulation control module determines the frequency modulation contribution matching degree based on the output of the intelligent analysis module. Power deviation rate The calculation results dynamically optimize boiler combustion control parameters, turbine speed control valve opening, and coordinated control strategies to achieve rapid power response at the second or even sub-second level. The control strategy is as follows:

[0049] (1) Power deviation rate based on real-time monitoring Optimize boiler combustion control parameters: when power deviation rate When the deviation is >5% (significant small-frequency fluctuations or large frequency modulation deviations), the coal feeder speed and forced and induced draft volumes are adjusted in real time to increase combustion intensity and rapidly improve the unit's output power, thus reducing the deviation between actual and commanded power; when the power deviation rate... When the frequency regulation accuracy is ≤2% (meeting the standard), maintain the current combustion parameters to avoid unit fluctuations caused by excessive adjustment and ensure frequency regulation stability;

[0050] (2) Power deviation rate based on real-time monitoring Optimize turbine speed control valve opening: when power deviation rate When the power deviation is greater than 3% and the grid frequency regulation command requires a rapid response, fine-tune the turbine speed control valve opening (adjustment range and power deviation rate). (Positive correlation in numerical values) directly changes the steam intake, enabling rapid fine-tuning of power, compensating for the lag in boiler combustion regulation, and facilitating sub-second response;

[0051] (3) Frequency modulation contribution consistency based on real-time monitoring Optimize the coordination control strategy: when the frequency modulation contribution matches When the deviation between the unit's frequency regulation contribution and the command is large (<0.7), adjust the boiler-turbine coordinated control logic to improve the linkage response speed of both, prioritize the execution accuracy of the frequency regulation command, and improve the consistency of the frequency regulation contribution. Numerical value; when the frequency modulation contribution matches When the contribution rate is ≥0.9 (frequency regulation contribution meets the standard), optimize and coordinate control parameters to balance the frequency regulation effect and the economic efficiency of unit operation;

[0052] The advantage is that the frequency modulation contribution matches the data through real-time monitoring. Power deviation rate The system collaboratively achieves dynamic calibration. By combining the values ​​of both, the unit's frequency regulation response coefficient is calibrated in real time, and the deviation of control parameters is corrected. This ensures that rapid response and high-precision control can be achieved in different frequency regulation scenarios (such as small frequency fluctuations and large frequency regulation commands), thereby completely solving the defect of lag in traditional control response. The module can also link the real-time data stream channel of the multi-source data acquisition module to realize the real-time issuance and execution feedback of control commands.

[0053] The full-cycle health management module is based on the remaining health cycle of the device. Based on the calculation results, a full lifecycle state evolution model of the equipment is constructed to monitor the degradation trend of key components in real time, establish a health status classification and early warning mechanism, and combine the remaining health cycle of the equipment. Targeted adjustments and monitoring will be carried out, as follows:

[0054] (1) Based on the remaining health cycle of the equipment The calculation results are used to classify the equipment health level: when the equipment has a remaining health cycle When the remaining health cycle is ≥8000h, it is considered to be in a healthy state. Maintain the regular monitoring frequency, focusing on tracking basic health indicators such as equipment vibration and temperature to ensure stable equipment operation. When the remaining health cycle is <4000h... If the time is less than 8000 hours, it is considered to be in a sub-healthy state. The monitoring frequency should be increased, and the degradation trend of key components (such as bearings and valves) should be closely monitored.

[0055] (2) Based on the remaining health cycle of the equipment Calculation results optimize state evolution model parameters: based on the remaining health cycle of the equipment. Real-time changing data is used to dynamically correct the decay coefficient in the equipment health degradation model. Update model parameters to improve the prediction accuracy of the remaining health cycle of the equipment and ensure that the model can accurately reflect the actual operating status of the equipment.

[0056] (3) Based on the remaining health cycle of the equipment The calculation results are used to establish tiered early warning thresholds, combined with the remaining health cycle of the equipment. The numerical settings include a three-level early warning threshold, with the first level being an early warning (device remaining health cycle). ≥6000h): Indicates the equipment is in good condition and requires no special intervention; Level 2 warning (2000h ≤ remaining health cycle of the equipment) <6000h): Indicates slight degradation of the equipment, issuing a warning signal and suggesting increased inspections; Level 3 warning (remaining health cycle of the equipment). <2000h): This indicates severe equipment degradation, triggering an emergency warning and prompting the predictive maintenance module to prepare for pre-processing.

[0057] (4) Based on the remaining health cycle of the equipment Calculation results monitor the degradation trend of key components: targeting the remaining health life of the equipment For equipment with low values, we focus on monitoring its vibration spectrum data, bearing temperature field data, and lubricating oil quality data. By combining these with the state evolution model, we analyze the degradation rate of key components, predict potential failure points, and provide accurate basis for subsequent predictive maintenance, filling the gap in the lack of full-cycle health management of equipment in existing technologies. At the same time, we output cycle prediction reports to support the predictive maintenance module.

[0058] The predictive maintenance module is based on a health status classification and early warning mechanism. It combines full-cycle health management data and a failure mode library to predict the remaining service life of the equipment, generate maintenance strategy suggestions, and optimize maintenance plans and spare parts inventory configuration.

[0059] The predictive maintenance module contains a fault mode library with typical fault information for various key equipment in thermal power units, including but not limited to: bearing wear faults (corresponding to abnormal vibration spectrum and excessive bearing temperature), lubricating oil deterioration faults (corresponding to abnormal lubricating oil quality data), valve jamming faults (corresponding to abnormal valve opening feedback data), coal feeder faults (corresponding to coal feeder speed fluctuations and abnormal forced and induced draft), main steam system faults (corresponding to abnormal main steam pressure and temperature), and condenser faults (corresponding to abnormal condenser vacuum). This covers various fault types, characteristics, and handling solutions that may occur throughout the equipment's entire lifecycle. Based on this information, the module predicts the remaining service life of the equipment (based on the equipment's remaining health cycle). The calculation results are combined with the degradation patterns of similar faults in the fault mode library for correction, generating targeted maintenance strategy recommendations: For Level 1 early warning equipment, it is recommended to maintain routine daily inspections and conduct a comprehensive inspection once a quarter; for Level 2 early warning equipment, it is recommended to increase the frequency of inspections (once a week), conduct targeted testing on degraded components, and prepare spare parts in advance; for Level 3 early warning equipment, it is recommended to immediately shut down for maintenance, replace severely degraded key components, and avoid the fault from expanding and affecting the frequency regulation response.

[0060] Meanwhile, maintenance plans are optimized based on maintenance strategy recommendations to avoid downtime losses caused by blind maintenance. Maintenance cycles are adjusted reasonably based on the probability of failure in the failure mode database. Spare parts inventory configuration is optimized, with increased inventory reserves for high-frequency failure components (such as bearings, lubricating oil, and valves) and reasonable control of inventory for low-frequency failure components to reduce inventory costs. The entire maintenance process is recorded, and the feedback and evaluation of maintenance effects are carried out to ensure that the equipment is in good operating condition and to avoid the impact of equipment failure on frequency modulation response speed and frequency modulation effect, thus forming a closed loop of equipment management of "prediction-maintenance-evaluation".

[0061] The human-machine interface module provides a visual monitoring interface for the system, displaying the frequency modulation control effect, equipment health status, and maintenance early warning information. It supports operator intervention and system parameter configuration; the interface can display the frequency modulation contribution matching degree in real time. Power deviation rate Remaining health cycle of equipment The calculation results, along with graded early warning signals and maintenance strategy suggestions, enable operators to monitor the system's operating status in real time and manually intervene in control parameters and maintenance plans when necessary, thereby improving the system's operational flexibility and reliability.

[0062] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An auxiliary frequency regulation control system for thermal power units based on multi-source information fusion, characterized in that, It includes a multi-source data acquisition module, a data preprocessing module, a multi-source information fusion module, an intelligent analysis module, an auxiliary frequency modulation control module, a full-cycle health management module, a predictive maintenance module, and a human-computer interaction module; The multi-source data acquisition module is used to collect multi-dimensional data during the operation of thermal power units, including power grid frequency regulation command data, unit operation data, and equipment health data. The data preprocessing module is responsible for cleaning, denoising, time-stamp alignment, outlier removal, and normalization of the collected multi-source data to establish a data stream with a unified time reference. The multi-source information fusion module uses an adaptive weighted fusion algorithm to perform spatiotemporal fusion of preprocessed multi-source data to construct a unified information framework that reflects the overall operating status of the unit. The intelligent analysis module receives the fused data output by the multi-source information fusion module and calculates the frequency modulation contribution consistency. a. Power deviation rate and remaining health cycle of equipment At the same time, the fused data is deeply analyzed to identify the characteristics of small frequency fluctuations in the power grid and generate frequency regulation demand forecasts; The auxiliary frequency modulation control module determines the frequency modulation contribution matching degree based on the output of the intelligent analysis module. Power deviation rate The calculation results are used to dynamically optimize boiler combustion control parameters, turbine speed control valve opening, and coordinated control strategies. The full-cycle health management module is based on the remaining health cycle of the device. Based on the calculation results, a full life cycle state evolution model of the equipment is constructed, the degradation trend of key components is monitored in real time, and a health status classification and early warning mechanism is established. The predictive maintenance module, based on a health status grading and early warning mechanism, combined with full-cycle health management data and a fault mode library, predicts the remaining service life of the equipment and generates maintenance strategy recommendations. The human-machine interaction module provides a visual monitoring interface for the system, displaying the frequency modulation control effect, equipment health status and maintenance early warning information, and supporting operator intervention and system parameter configuration.

2. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The power grid frequency regulation command data includes: power grid frequency deviation signal, AGC dispatch command, regional control error, planned output curve, frequency regulation mileage price signal, primary frequency regulation dead zone threshold and secondary frequency regulation response time requirements.

3. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The unit operating data includes: boiler combustion status data, turbine operating parameters, generator electrical parameters, valve opening feedback data, coal feeder speed data, forced draft and induced draft data, main steam pressure and temperature data, condenser vacuum data, ambient temperature data, feedwater pump outlet pressure and flow rate, high-pressure heater water level, deaerator pressure and temperature, coal mill current and outlet air-coal temperature, air preheater inlet and outlet flue gas differential pressure, desulfurization tower pH value and slurry density, reheat steam temperature and pressure, and shaft seal steam pressure and temperature.

4. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The equipment health data includes: equipment vibration spectrum data, bearing temperature field data, lubricating oil quality data, ultrasonic testing data of key welds, thermal deformation of water pump and fan casings, turbine blade tip clearance value, generator stator winding partial discharge signal, fatigue characteristic frequency amplitude of high-voltage motor bearings, wear index of coal mill liner, blower blade crack propagation monitoring data, and high-temperature pipeline creep strain data.

5. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The intelligent analysis module receives the fused data output by the multi-source information fusion module and calculates the frequency modulation contribution consistency. The calculation formula is as follows: In the formula, This represents the actual output power of the generator unit, that is, the active power actually generated by the generator unit at time t. Indicates the power output of the generator set via AGC commands. This indicates the ideal power corresponding to the grid frequency. Indicates the initial power.

6. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The intelligent analysis module receives the fused data output by the multi-source information fusion module and calculates the power deviation rate. The calculation formula is as follows: In the formula, This represents the rated power of the unit, used as a normalization reference. This represents the actual power of the unit at the i-th sampling time. This indicates that the ideal power corresponds to the grid frequency at the i-th sampling time. Indicates the number of sampling points. The standard deviation of the power grid frequency deviation, This represents the mean value of the power grid frequency deviation. This represents the frequency fluctuation sensitivity coefficient.

7. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The intelligent analysis module receives the fused data output from the multi-source information fusion module to calculate the remaining health cycle of the device. The calculation formula is as follows: In the formula, Indicates the remaining health cycle of the equipment. Indicates the standard health cycle of the equipment. Indicates the attenuation coefficient. Indicates the number of equipment health monitoring indicators. This represents the current monitoring value of the j-th health indicator; This represents the standard threshold for the j-th health indicator.

8. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The auxiliary frequency modulation control module is based on the frequency modulation contribution matching degree. Power deviation rate The adjusted control strategy is as follows: (1) Power deviation rate based on real-time monitoring Optimize boiler combustion control parameters: when power deviation rate When the deviation is greater than 5%, the feeder speed and forced and induced draft are adjusted in real time to increase combustion intensity and quickly improve the unit's output power, thus reducing the deviation between the actual power and the commanded power; when the power deviation rate... When the concentration is ≤2%, maintain the current combustion parameters stable; (2) Power deviation rate based on real-time monitoring Optimize turbine speed control valve opening: when power deviation rate When the steam rate is greater than 3% and the power grid frequency regulation command requires a rapid response, the opening of the turbine speed control valve is finely adjusted to change the steam intake. (3) Frequency modulation contribution consistency based on real-time monitoring Optimize the coordination control strategy: when the frequency modulation contribution matches When the value is less than 0.7, adjust the boiler-turbine coordinated control logic to improve the linkage response speed of both and at the same time improve the frequency regulation contribution matching degree. Numerical value; when the frequency modulation contribution matches When the value is ≥0.9, optimize the coordinated control parameters.

9. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The full-cycle health management module is based on the remaining health cycle of the device. The calculation results are used to classify the equipment health level: when the equipment has a remaining health cycle When the remaining health cycle is ≥8000h, it is considered to be in a healthy state. Maintain the regular monitoring frequency, focusing on tracking basic health indicators such as equipment vibration and temperature to ensure stable equipment operation. When the remaining health cycle is <4000h... If the time is less than 8000 hours, it is considered to be in a sub-healthy state. The monitoring frequency should be increased, and the degradation trend of key components should be closely monitored.

10. The auxiliary frequency regulation control system for thermal power units based on multi-source information fusion according to claim 1, characterized in that, The full-cycle health management module is based on the remaining health cycle of the device. Calculation results monitor the degradation trend of key components: targeting the remaining health life of the equipment For equipment with low values, we focus on monitoring its vibration spectrum data, bearing temperature field data, and lubricating oil quality data. By combining the state evolution model, we analyze the degradation rate of key components, predict potential failure points, and output a cycle prediction report.

Images