A method and system for operation control of thermal power plants based on multi-source data analysis

The multi-source data analysis-based operation control method for thermal power plants solves the problems of single data and lagging strategies in traditional thermal power plant control, enables precise monitoring and adjustment of boilers and turbines, improves combustion efficiency and equipment stability, and enhances the intelligent management of thermal power plants.

CN123113707APending Publication Date: 2026-11-17SICHUAN ENERGY INVESTMENT ZHONGJIANG GAS POWER GENERATION CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611101882.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-23
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Traditional thermal power plant operation and control methods rely on single data monitoring and lack multi-source data fusion analysis, resulting in inaccurate assessment of the impact of environmental changes and fuel quality fluctuations, lagging control strategies and slow response speed, making it difficult to achieve precise regulation and equipment stability.

Method used

By employing a multi-source data analysis method, various parameters of key equipment such as boilers and steam turbines are collected by deploying multiple sensors. Combustion efficiency, operating load stability, and vibration fatigue monitoring modules are constructed. Data preprocessing and threshold comparison analysis are performed to generate refined adjustment strategies.

Benefits of technology

It enables dynamic assessment and strategy optimization of boiler operating status, improves fuel utilization, reduces energy consumption and pollutant emissions, enhances the operational stability of power generation equipment and the lifespan of key components, and improves the level of intelligent management of thermal power plant operation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113707A_ABST
    Figure CN123113707A_ABST
Patent Text Reader

Abstract

This invention discloses a method and system for operation control of thermal power plants based on multi-source data analysis, belonging to the field of thermal power plant operation control technology. The system monitors boiler flue gas oxygen content, boiler temperature, ambient humidity, and fuel composition, while simultaneously collecting data on turbine speed, shaft vibration frequency, main steam temperature, and power generation load. The collected data undergoes outlier removal, signal smoothing, missing data completion, time-series reconstruction, and normalization. By calculating the combustion efficiency coefficient and comparing it with a preset threshold, the system determines the boiler combustion state and formulates control strategies. It quantifies the impact of load disturbances on turbine operation, assesses load stability, and adjusts operating strategies in case of anomalies. It identifies potential fatigue risks in turbine blades, conducts a comprehensive assessment combining vibration and corrosion factors, and generates maintenance plans in case of anomalies. This system achieves real-time dynamic monitoring and intelligent risk warning of the operating status of key equipment in thermal power plants, improving the safety and economy of unit operation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of thermal power plant operation control technology, specifically to a thermal power plant operation control method and system based on multi-source data analysis. Background Technology

[0002] In traditional thermal power plant operation and control, the system primarily relies on distributed sensor systems to monitor key parameters of core equipment such as boilers, turbines, and generators, including temperature, pressure, and speed, and then adjusts these parameters based on human experience or preset thresholds. This control method has the following limitations: The data dimensions are limited. Existing technologies typically only collect the operating parameters of the equipment itself, lacking comprehensive monitoring of environmental factors such as atmospheric humidity, fuel characteristics such as calorific value, sulfur content, and ash content, as well as the operating status of the electrical system such as voltage and power factor. This makes it impossible to accurately assess the impact of environmental changes and fuel quality fluctuations on unit efficiency.

[0003] The lack of multi-source data fusion and analysis capabilities is a significant issue. Thermal power plant operation involves the coupling of multiple disciplines such as thermodynamics and materials science, but current technologies have failed to achieve in-depth fusion and analysis of multi-source data related to environment, fuel, equipment, and electrical systems. For example, boiler combustion efficiency depends not only on the calorific value of the fuel but also on factors such as fuel and electrical load rates, and traditional control methods struggle to quantify the combined impact of these factors.

[0004] The control strategies suffer from lag and inefficiency. Traditional control strategies rely on manual experience for adjustment, resulting in slow response and low adjustment accuracy. For example, during sudden load changes, operators need to gradually adjust fuel supply, ventilation, and water supply, leading to large fluctuations in unit parameters and affecting power supply quality. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for the operation control of thermal power plants based on multi-source data analysis, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: a method and system for operation control of thermal power plants based on multi-source data analysis, comprising: a data acquisition module, a data processing module, a combustion efficiency evaluation module, a load response analysis module, and a vibration fatigue monitoring module; The data acquisition module is used to collect data on flue gas oxygen content, boiler temperature, ambient relative humidity, fuel volatile matter, fuel fixed carbon content, fuel ash content, and fuel sulfur content by monitoring key equipment in the operation control system of the thermal power plant, and to establish a first dataset; and to collect data on turbine speed, shaft vibration main frequency component, main steam temperature, and power generation load, and to establish a second dataset. The data processing module is used to preprocess the collected data, including outlier removal, signal smoothing, missing data completion, time series reconstruction and normalization. The combustion efficiency evaluation module is used to monitor the boiler operating efficiency, calculate the combustion efficiency coefficient RSx, compare it with the first threshold Q1, and determine whether the boiler's combustion efficiency is qualified. If it is not qualified, a strategy is given. The load response analysis module is used to quantify the impact of load disturbance on equipment operation by considering the unevenness of the turbine's operating response during load changes, calculate and obtain the operating load stability coefficient FHx, and compare it with the second threshold Q2 to determine whether the operating load fluctuation of the turbine generator set is normal. If it is abnormal, a strategy is given. The vibration fatigue monitoring module is used to identify the potential impact and fatigue risk of high-frequency disturbances on the turbine blade structure, calculate and obtain the impeller impact response coefficient YL, and compare and analyze it with the third threshold Q3 to determine whether the turbine blade is in normal condition. If it is abnormal, a strategy is given.

[0007] Preferably, the data acquisition module is used to monitor the combustion exhaust gas in the boiler flue in real time by deploying a flue gas analyzer above the boiler in a thermal power plant, and to collect the oxygen content of the flue gas. By deploying infrared temperature measuring devices and thermocouple arrays, the boiler furnace is monitored in real time, and the boiler temperature is collected. By deploying environmental humidity sensors, the relative humidity of the environment is collected. By deploying coal quality testing equipment, fuel samples in the coal conveying system are monitored in real time, and the volatile matter (VM), fixed carbon content (FC), ash content (Ash), and sulfur content (SC) of the fuel are collected to establish the first dataset. By deploying turbine condition monitoring sensors, the turbine shaft system is monitored in real time, and the turbine speed is collected. By deploying vibration acceleration sensors and FFT analysis technology, the turbine bearing housing is monitored in real time, and the main frequency component fv of shaft vibration is collected. By deploying main steam temperature sensors, the main steam pipeline is monitored in real time, and the main steam temperature Ts is collected. By deploying power metering and load analysis devices, the power generation bus is monitored in real time, and the power generation load Pg is collected. A second dataset is then established.

[0008] Preferably, the data processing module is used to preprocess the collected data, including: using outlier removal technology to remove abrupt changes and outliers during the monitoring process; using a sliding window and weighted average algorithm to smooth high-frequency disturbance signals; using time series interpolation and multi-source correlation completion methods to fill in intermittent missing data; performing signal synchronization and time series reconstruction for multi-channel sensor data; and normalizing all numerical variables.

[0009] Preferably, the combustion efficiency evaluation module includes a first calculation unit and a first analysis unit; The first computing unit is used to monitor the boiler operating efficiency and extract the oxygen content of the flue gas from the first dataset. Boiler temperature Ambient relative humidity value The volatile matter (VM), fixed carbon content (FC), ash content (Ash), and sulfur content (SC) of the fuel are dimensionlessly processed to calculate the combustion efficiency coefficient (RSx).

[0010] Preferably, the first analysis unit is used to preset a first threshold Q1 and compare the combustion efficiency coefficient RSx with the first threshold Q1 to obtain a first evaluation result, including: When the combustion efficiency coefficient RSx ≥ the first threshold Q1, it indicates that the boiler's combustion efficiency is qualified and should be continuously monitored. When the combustion efficiency coefficient RSx < the first threshold Q1, it indicates that the boiler's combustion efficiency is unqualified, triggering the first early warning command and generating the first strategy: by adjusting the ratio of primary air volume to secondary air volume to 1:1.8-1:2.2, the oxygen content of the flue gas is reduced. Control it between 3% and 4.5%; adjust the pulverized coal ratio into the furnace, prioritizing the addition of coal with high calorific value. Fuel with a concentration of ≥25 MJ / kg should account for no less than 20%; the air preheater regeneration channel should be activated to increase the temperature of the pre-combustion mixture by 30% and improve ignition stability.

[0011] Preferably, the load response analysis module includes a second calculation unit and a second analysis unit; The second calculation unit is used to extract the power generation load Pg and turbine speed from the second dataset by considering the unevenness of the turbine's operating response during load changes. and main steam temperature As an observed variable, the impact of load disturbance on equipment operation is quantified by calculating the change rate per unit time. After dimensionless processing, the operating load stability coefficient FHx is calculated and obtained.

[0012] Preferably, the second analysis unit is used to pre-set a second threshold Q2, and compare the operating load stability coefficient FHx with the second threshold Q2 to obtain a second evaluation result, including: When the operating load stability coefficient FHx ≤ the second threshold Q2, it indicates that the operating load fluctuation of the steam turbine generator unit is normal and should be continuously monitored. When the operating load stability coefficient FHx ≤ the second threshold Q2, it indicates abnormal fluctuations in the operating load of the steam turbine generator unit, triggering a second early warning command and generating a second strategy: limiting the rate of change of the turbine main steam valve opening to within 2% / s to mitigate the risk of high-frequency power surges eroding the steam passages, and using the SCADA monitoring system to monitor the power generation load. If the fluctuation is within ±8% for 5 consecutive minutes, the output shaft torque fluctuation should be controlled to be no higher than ±3%.

[0013] Preferably, the vibration fatigue monitoring module includes a third calculation unit and a third analysis unit; The third computing unit is used to identify the potential impact and fatigue risks of high-frequency disturbances on the turbine blade structure, and to extract the turbine speed from the second dataset. The acceleration change, the dominant frequency component of shaft system vibration Using the rate of change as a dynamic indicator, and combining the sulfur content SC of the fuel in the first dataset as a corrosion influencing factor, the impeller impact response coefficient YL was calculated after dimensionless processing.

[0014] Preferably, the third analysis unit is used to obtain a third evaluation result by pre-setting a third threshold Q3 and comparing the impeller impact response coefficient YL with the third threshold Q3. When the impeller impact response coefficient YL≤ the third threshold Q3, it indicates that the turbine blades are in normal condition and there is no risk of potential cumulative damage to the shaft structure caused by high-frequency vibration and speed impact in a high-sulfur environment. Continuous monitoring is required. When the third threshold Q3*130% ≥ the impeller impact response coefficient YL > the third threshold Q3, it indicates that the turbine blade condition is at level one abnormality, and there is a potential risk of cumulative damage to the shaft structure caused by high-frequency vibration and speed impact in a high-sulfur environment. This triggers the third early warning command and generates the third strategy: using laser multi-point interferometric vibration measurement technology to perform non-contact condition monitoring of the turbine blade, capturing the micro-perturbation characteristics of the vibration response curve in real time, and simultaneously increasing the sampling frequency of the shaft vibration signal to ≥1kHz, covering a high-frequency bandwidth of 200–500Hz, performing dynamic analysis of spectral anomalies, identifying potential resonance components and their evolution trends; adjusting the main steam valve response hysteresis curve in conjunction with the control, limiting the steam dynamic pressure slope to no more than 0.6MPa / s, indirectly suppressing the repeated stress loading on the blade structure by the surge in torque through the slow release of steam thrust impact process, thereby reducing the fatigue accumulation rate; and reducing the proportion of high-sulfur coal to ≤50% to alleviate the impact of increased corrosion on fatigue accumulation. When the impeller impact response coefficient YL > the third threshold Q3*130%, it indicates that the turbine blade is in a level two abnormal condition and the fatigue damage trend of the turbine blade has reached an irreversible level, triggering the fourth warning instruction and generating the fourth strategy: automatically triggering a maintenance work order to replace the target blade assembly.

[0015] Preferably, a method for operating control of a thermal power plant based on multi-source data analysis includes the following steps: Step 1: By monitoring key equipment in the operation and control system of the thermal power plant, collect data on flue gas oxygen content, boiler temperature, ambient relative humidity, fuel volatile matter, fuel fixed carbon content, fuel ash content, and fuel sulfur content to establish the first dataset; collect data on turbine speed, shaft vibration main frequency component, main steam temperature, and power generation load to establish the second dataset. Step 2: Preprocess the collected data, including outlier removal, signal smoothing, missing data completion, time series reconstruction, and normalization. Step 3: By monitoring the boiler operating efficiency, calculate and obtain the combustion efficiency coefficient RSx, and compare it with the first threshold Q1 to determine whether the boiler's combustion efficiency is qualified. If it is not qualified, a strategy will be given. Step 4: By addressing the unevenness of the turbine's operating response during load changes, quantify the impact of load disturbances on equipment operation, calculate and obtain the operating load stability coefficient FHx, and compare it with the second threshold Q2 to determine whether the operating load fluctuation of the turbine generator set is normal. If abnormal, provide a strategy. Step 5: Identify the potential impact and fatigue risks of high-frequency disturbances on the turbine blade structure, calculate and obtain the impeller impact response coefficient YL, and compare and analyze it with the third threshold Q3 to determine whether the turbine blade condition is normal. If it is abnormal, a strategy is given.

[0016] This invention provides a method and system for operation control of thermal power plants based on multi-source data analysis. It has the following beneficial effects: (1) The operation control method and system of thermal power plant based on multi-source data analysis collects key data such as boiler flue gas, fuel composition and furnace temperature by deploying multi-source sensors, and constructs combustion efficiency coefficient RSx to realize dynamic evaluation and strategy optimization of boiler operation status. It can effectively improve fuel utilization rate and reduce excess oxygen content in flue gas, thereby reducing energy consumption and pollutant emissions.

[0017] (2) The operation control method and system of thermal power plant based on multi-source data analysis can quantitatively evaluate the operation response law of steam turbine under load disturbance by constructing the operating load stability coefficient FHx. It can accurately identify the unbalanced operation state, adjust the main steam valve opening and load response rhythm in a timely manner, effectively alleviate the risk of steam shock and power fluctuation, and improve the operation stability of power generation equipment.

[0018] (3) The operation control method and system of thermal power plant based on multi-source data analysis integrates multiple factors such as turbine speed impact, shaft vibration frequency and sulfur content to construct impeller impact response coefficient YL, identify the risk of combined effects of high frequency vibration and corrosion on blade structure, realize early warning of fatigue damage trend, support graded response and maintenance decision, and significantly extend the service life of key components such as turbine blades.

[0019] (4) The operation control method and system of thermal power plant based on multi-source data analysis establishes first and second datasets, integrates data from multiple subsystems such as boiler, steam turbine and power output, realizes system-level monitoring and modular analysis, and each calculation module has threshold comparison and strategy triggering functions. It can generate refined adjustment measures according to the operating status, and improve the overall intelligent and refined management level of thermal power plant operation. Attached Figure Description

[0020] Figure 1 This is a block diagram and flowchart of a thermal power plant operation control system based on multi-source data analysis according to the present invention. Figure 2 This is a schematic diagram illustrating the steps of a thermal power plant operation control method based on multi-source data analysis according to the present invention. Detailed Implementation

[0021] 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.

[0022] Example 1 Please see Figure 1 The present invention provides a thermal power plant operation control system based on multi-source data analysis, including: a data acquisition module, a data processing module, a combustion efficiency evaluation module, a load response analysis module, and a vibration fatigue monitoring module; The data acquisition module is used to collect data on flue gas oxygen content, boiler temperature, ambient relative humidity, fuel volatile matter, fuel fixed carbon content, fuel ash content, and fuel sulfur content by monitoring key equipment in the operation control system of the thermal power plant, and to establish a first dataset; and to collect data on turbine speed, shaft vibration main frequency component, main steam temperature, and power generation load, and to establish a second dataset. The data processing module is used to preprocess the collected data, including outlier removal, signal smoothing, missing data completion, time series reconstruction and normalization. The combustion efficiency evaluation module is used to monitor the boiler operating efficiency, calculate the combustion efficiency coefficient RSx, compare it with the first threshold Q1, and determine whether the boiler's combustion efficiency is qualified. If it is not qualified, a strategy is given. The load response analysis module is used to quantify the impact of load disturbance on equipment operation by considering the unevenness of the turbine's operating response during load changes, calculate and obtain the operating load stability coefficient FHx, and compare it with the second threshold Q2 to determine whether the operating load fluctuation of the turbine generator set is normal. If it is abnormal, a strategy is given. The vibration fatigue monitoring module is used to identify the potential impact and fatigue risk of high-frequency disturbances on the turbine blade structure, calculate and obtain the impeller impact response coefficient YL, and compare and analyze it with the third threshold Q3 to determine whether the turbine blade is in normal condition. If it is abnormal, a strategy is given.

[0023] In this embodiment, by constructing a multi-functional operation analysis system including a combustion efficiency assessment module, a load response analysis module, and a vibration fatigue monitoring module, it is possible to achieve a coordinated assessment of boiler combustion status, turbine operation balance, and blade structural safety. This enables the generation of graded control strategies for different abnormal operation scenarios, which not only improves system operation efficiency and safety but also significantly enhances the ability of thermal power plants to cope with dynamic load fluctuations and structural fatigue risks.

[0024] Example 2 This embodiment is an explanation based on Embodiment 1. Specifically, the data acquisition module is used to monitor the combustion exhaust gas in the boiler flue in real time by deploying a flue gas analyzer above the boiler in a thermal power plant, and to collect the oxygen content of the flue gas. By deploying infrared temperature measuring devices and thermocouple arrays, the boiler furnace is monitored in real time, and the boiler temperature is collected. By deploying environmental humidity sensors, the relative humidity of the environment is collected. By deploying coal quality testing equipment, fuel samples in the coal conveying system are monitored in real time, and the volatile matter (VM), fixed carbon content (FC), ash content (Ash), and sulfur content (SC) of the fuel are collected to establish the first dataset. By deploying turbine condition monitoring sensors, the turbine shaft system is monitored in real time, and the turbine speed is collected. By deploying vibration acceleration sensors and FFT analysis technology, the turbine bearing housing is monitored in real time, and the main frequency component fv of shaft vibration is collected. By deploying main steam temperature sensors, the main steam pipeline is monitored in real time, and the main steam temperature Ts is collected. By deploying power metering and load analysis devices, the power generation bus is monitored in real time, and the power generation load Pg is collected. A second dataset is then established.

[0025] In this embodiment, by deploying flue gas analyzers, infrared temperature measuring devices, thermocouple arrays, humidity sensors, coal quality testing equipment, and turbine condition monitoring sensors at key locations in the thermal power plant, a first dataset containing information such as boiler combustion exhaust gas, furnace temperature, ambient humidity, and fuel composition, and a second dataset containing information such as turbine speed, shaft vibration frequency, main steam temperature, and power generation load are constructed. This enables high-precision, real-time monitoring of the thermal power plant's operating status throughout the entire process, significantly improving the system's ability to perceive energy efficiency fluctuations, operational anomalies, and equipment fatigue, and providing comprehensive and reliable data support for subsequent operation assessments and strategy generation.

[0026] Example 3 This embodiment is an explanation of Embodiment 2. Specifically, the data processing module is used to preprocess the collected data, including: using outlier removal technology to remove abrupt changes and outliers during the monitoring process; using a sliding window and weighted average algorithm to smooth high-frequency disturbance signals; using time series interpolation and multi-source correlation completion methods to fill in intermittent missing data; performing signal synchronization and time-series reconstruction for multi-channel sensor data; and normalizing all numerical variables.

[0027] In this embodiment, the data processing module performs systematic preprocessing on multi-source monitoring data, employs outlier removal techniques to eliminate mutations and outliers, utilizes sliding window and weighted average algorithms to smooth high-frequency disturbances, combines time series interpolation and multi-source correlation methods to complete missing data, and performs signal synchronization and time-series reconstruction on multi-channel sensor data. Simultaneously, variable normalization is implemented, thereby significantly improving data quality and time-series consistency. This ensures that subsequent combustion assessment, load response analysis, and vibration monitoring modules operate stably with high-precision, high-robust data support, effectively enhancing the overall system's analytical accuracy and response efficiency.

[0028] Example 4 This embodiment is an explanation of Embodiment 1. Specifically, the combustion efficiency evaluation module includes a first calculation unit and a first analysis unit. The first computing unit is used to monitor the boiler operating efficiency and extract the oxygen content of the flue gas from the first dataset. Boiler temperature Ambient relative humidity value The volatile matter (VM), fixed carbon content (FC), ash content (Ash), and sulfur content (SC) of the fuel are dimensionlessly processed to calculate the combustion efficiency coefficient (RSx), as shown in the following formula:

[0029] In the formula, Indicates the calorific value of the fuel. This indicates the oxygen content of the flue gas. Indicates the boiler temperature. This indicates the relative humidity value of the environment;

[0030] In the formula, VM represents the volatile matter content of the fuel, FC represents the fixed carbon content of the fuel, Ash represents the ash content of the fuel, and SC represents the sulfur content of the fuel. , w3 and Indicates the weighting coefficient. , , , and ; , w3 and The weighted coefficients, representing characteristic parameters of fuel coal, are derived from statistical modeling and multiple regression analysis of extensive experimental calorific value data from typical coal samples, including bituminous coal, lignite, and anthracite. This analysis extracts the dominance of components such as volatile matter (VM), fixed carbon (FC), ash (Ash), and sulfur content (Sc) on fuel calorific value. Combining coal quality inspection reports, relevant national standards such as "Industrial Analysis Methods for Coal" and "Standards for Determination of Fuel Calorific Value," and referencing the coal quality-calorific value correspondence in the historical operating database of coal-fired power plants, along with industry expert experience, the fitting weights for each component are comprehensively determined. The coefficient values ​​aim to improve the accuracy and generalization ability of calorific value estimation, making it applicable to complex and variable operating scenarios with varying coal compositions, and providing a reliable basis for combustion efficiency assessment.

[0031] In this embodiment, the first calculation unit in the combustion efficiency evaluation module calculates the combustion efficiency coefficient RSx based on multiple source parameters collected centrally from the first dataset, including flue gas oxygen content, boiler temperature, ambient relative humidity, and fuel components such as volatile matter VM, fixed carbon content FC, ash content Ash, and sulfur content SC. This coefficient effectively reflects the current thermal conversion efficiency of the boiler and the degree of fuel compatibility. Combined with the comparison analysis of RSx with a preset threshold by the first analysis unit, abnormal fluctuations in the boiler combustion state can be accurately identified, and the air-fuel ratio and combustion control strategy can be adjusted in a timely manner. This significantly improves the boiler combustion efficiency, reduces energy loss and pollution emissions, and enhances the economic and environmental benefits of thermal power plant operation.

[0032] Example 5 This embodiment is an explanation of Embodiment 4. Specifically, the first analysis unit is used to preset a first threshold Q1 and compare the combustion efficiency coefficient RSx with the first threshold Q1 to obtain a first evaluation result, including: When the combustion efficiency coefficient RSx ≥ the first threshold Q1, it indicates that the boiler's combustion efficiency is qualified and should be continuously monitored. When the combustion efficiency coefficient RSx < the first threshold Q1, it indicates that the boiler's combustion efficiency is unqualified, triggering the first early warning command and generating the first strategy: by adjusting the ratio of primary air volume to secondary air volume to 1:1.8-1:2.2, the oxygen content of the flue gas is reduced. Control it between 3% and 4.5%; adjust the pulverized coal ratio into the furnace, prioritizing the addition of coal with high calorific value. Fuel with a concentration of ≥25 MJ / kg should account for no less than 20%; the air preheater regeneration channel should be activated to increase the temperature of the pre-combustion mixture by 30% and improve ignition stability.

[0033] The first threshold Q1 is derived from the following: By statistically analyzing long-term operating data of multiple boilers under different operating loads, fuel types, and environmental conditions, the typical range of combustion efficiency coefficient RSx under high-efficiency and low-efficiency operating conditions is extracted. Combined with thermal test results and combustion optimization experimental data, the critical efficiency inflection point is identified. At the same time, referring to the qualified benchmark values ​​of combustion efficiency in national and industry standards such as the "Technical Specification for Energy Saving of Industrial Boilers" and "Operating Procedures for Thermal Power Boilers", combined with the experience judgment of industry experts and feedback from on-site operators, and comprehensively considering fuel calorific value utilization, safety emission limits, and economic operation requirements, a reasonable combustion efficiency discrimination threshold Q1 is determined to accurately identify boilers operating under low-efficiency conditions, and to achieve timely early warning and strategy optimization for efficiency anomalies.

[0034] In this embodiment, by setting a first threshold Q1 and comparing and analyzing the combustion efficiency coefficient RSx with it, the intelligent judgment of the boiler combustion efficiency status can be realized. If the efficiency is low, a first strategy including air volume ratio adjustment, coal calorific value ratio optimization and air preheater activation is automatically generated, thereby effectively improving the uniformity of combustion temperature and ignition stability, reducing the excess oxygen content in flue gas, improving boiler thermal efficiency, and enhancing the system's adaptability to fuel fluctuations.

[0035] Example 6 This embodiment is an explanation of Embodiment 1. Specifically, the load response analysis module includes a second calculation unit and a second analysis unit. The second calculation unit is used to extract the power generation load Pg and turbine speed from the second dataset by considering the unevenness of the turbine's operating response during load changes. and main steam temperature As an observed variable, the impact of load disturbance on equipment operation is quantified by calculating the magnitude of its change per unit time. After dimensionless processing, the operating load stability coefficient FHx is calculated using the following formula:

[0036] In the formula, Indicates power generation load. Indicates the turbine speed. This indicates the main steam temperature.

[0037] In this embodiment, by constructing an operating load stability coefficient FHx, the dynamic changes of power generation load Pg, turbine speed, and main steam temperature within a unit time are comprehensively quantified, which can sensitively capture the system response characteristics under load disturbances. Compared with the traditional judgment method that only relies on the single variable change of load, it has a stronger ability to identify operational imbalances, which can effectively assist the control system in realizing the adjustment of main steam valve opening and the optimization of load response, thereby reducing the intensity of steam impact, suppressing mechanical fatigue and energy efficiency fluctuations caused by fluctuations, and improving the stability and reliability of the turbine generator set under variable load operating conditions.

[0038] Example 7 This embodiment is an explanation of embodiment 6. Specifically, the second analysis unit is used to preset a second threshold Q2 and compare the operating load stability coefficient FHx with the second threshold Q2 to obtain a second evaluation result, including: When the operating load stability coefficient FHx ≤ the second threshold Q2, it indicates that the operating load fluctuation of the steam turbine generator unit is normal and should be continuously monitored. When the operating load stability coefficient FHx ≤ the second threshold Q2, it indicates abnormal fluctuations in the operating load of the steam turbine generator unit, triggering a second early warning command and generating a second strategy: limiting the rate of change of the turbine main steam valve opening to within 2% / s to mitigate the risk of high-frequency power surges eroding the steam passages, and using the SCADA monitoring system to monitor the power generation load. If the fluctuation is within ±8% for 5 consecutive minutes, the output shaft torque fluctuation should be controlled to be no higher than ±3%.

[0039] The second threshold Q2 is derived from the statistical analysis of turbine load disturbance response data from a large number of thermal power plant operating conditions. This analysis extracts the amplitude and frequency characteristics of the interconnected fluctuations in power generation load, turbine speed, and main steam temperature during load changes. Combined with the dynamic stability margin and main steam valve response inertia under actual operating conditions, a reasonable load fluctuation threshold is determined. Referring to relevant national or industry safety standards for thermal equipment operation, load regulation performance parameters provided by turbine manufacturers, and the experience of industry experts in dynamic disturbance control, this second threshold Q2 is formulated to effectively identify abnormal response behavior of turbine generator sets during load disturbances. It assists the system in providing accurate early warnings and generating response strategies, thereby ensuring unit operational stability and the safety of load response.

[0040] In this embodiment, by comparing the operating load stability coefficient FHx with the second threshold Q2, the operating fluctuation state of the steam turbine generator set under load disturbance is accurately identified. When an abnormal fluctuation is detected, the system automatically triggers a second early warning command and generates a multi-dimensional strategy that includes limiting the rate of change of the turbine main steam valve opening, continuously monitoring the power generation load, and controlling shaft torque fluctuations. This can effectively alleviate the risk of high-frequency power surges and steam channel scouring caused by sudden load changes, reduce equipment fatigue and the probability of control system overshoot, thereby improving the response flexibility and safety of the steam turbine generator set under dynamic operating conditions.

[0041] Example 8 This embodiment is an explanation of Embodiment 1. Specifically, the vibration fatigue monitoring module includes a third calculation unit and a third analysis unit. The third computing unit is used to identify the potential impact and fatigue risks of high-frequency disturbances on the turbine blade structure, and to extract the turbine speed from the second dataset. The acceleration change, the dominant frequency component of shaft system vibration Using the rate of change as a dynamic indicator, and combining the sulfur content (SC) of the fuel in the first dataset as a corrosion influencing factor, the impeller impact response coefficient YL was calculated after dimensionless processing, as shown in the following formula:

[0042] In the formula, Indicates the turbine speed. The second derivative of the turbine rotational speed represents the instantaneous impact of the turbine rotational speed. The frequency component represents the dominant frequency of the vibration, and t represents time t. The term represents the rate of change of the dominant vibration frequency over time, and SC represents the sulfur content of the fuel. and This represents the empirical adjustment factor.

[0043] Empirical moderating factor and The data is derived from historical monitoring data of multiple steam turbine generator sets under different operating loads, fuel conditions, and vibration environments. Regression fitting and sensitivity analysis were performed to extract the coupling relationship between speed impact, vibration frequency fluctuations, and corrosion factors. Combined with system structural characteristics and fatigue failure mechanisms, the weight ratios of each factor in the impeller impact response coefficient YL were adjusted. Referring to typical blade vibration tolerance indicators provided by turbine manufacturers, domestic and international steam turbine operating condition monitoring standards, and industry expert experience, the value ranges of α and β were comprehensively determined to ensure that the response coefficient has reasonable sensitivity and stability to the coupling risk of impact and corrosion, thereby achieving accurate characterization and early warning of structural fatigue trends.

[0044] In this embodiment, the vibration fatigue monitoring module constructs an impeller impact response coefficient YL, integrating multiple dynamic and corrosion factors such as the second derivative of the turbine speed, the dominant frequency component of shaft vibration and its rate of change, and the sulfur content SC of the fuel, to achieve a multi-dimensional assessment of the load-bearing state of the turbine blade structure. This method can not only identify the instantaneous impact risk caused by high-frequency disturbances, but also combine the corrosion effect brought by sulfur-containing fuel to comprehensively reflect the fatigue accumulation trend of the blades, effectively supporting early warning and condition classification response, thereby extending the service life of the impeller structure and improving the reliability of unit operation and the accuracy of maintenance decisions.

[0045] Example 9 This embodiment is an explanation of Embodiment 8. Specifically, the third analysis unit is used to obtain a third evaluation result by pre-setting a third threshold Q3 and comparing the impeller impact response coefficient YL with the third threshold Q3. When the impeller impact response coefficient YL≤ the third threshold Q3, it indicates that the turbine blades are in normal condition and there is no risk of potential cumulative damage to the shaft structure caused by high-frequency vibration and speed impact in a high-sulfur environment. Continuous monitoring is required. When the third threshold Q3*130% ≥ the impeller impact response coefficient YL > the third threshold Q3, it indicates that the turbine blade condition is at level one abnormality, and there is a potential risk of cumulative damage to the shaft structure caused by high-frequency vibration and speed impact in a high-sulfur environment. This triggers the third early warning command and generates the third strategy: using laser multi-point interferometric vibration measurement technology to perform non-contact condition monitoring of the turbine blade, capturing the micro-perturbation characteristics of the vibration response curve in real time, and simultaneously increasing the sampling frequency of the shaft vibration signal to ≥1kHz, covering a high-frequency bandwidth of 200–500Hz, performing dynamic analysis of spectral anomalies, identifying potential resonance components and their evolution trends; adjusting the main steam valve response hysteresis curve in conjunction with the control, limiting the steam dynamic pressure slope to no more than 0.6MPa / s, indirectly suppressing the repeated stress loading on the blade structure by the surge in torque through the slow release of steam thrust impact process, thereby reducing the fatigue accumulation rate; and reducing the proportion of high-sulfur coal to ≤50% to alleviate the impact of increased corrosion on fatigue accumulation. When the impeller impact response coefficient YL > the third threshold Q3*130%, it indicates that the turbine blade is in a level two abnormal condition and the fatigue damage trend of the turbine blade has reached an irreversible level, triggering the fourth warning instruction and generating the fourth strategy: automatically triggering a maintenance work order to replace the target blade assembly.

[0046] The third threshold Q3 is derived from the statistical analysis of a large amount of turbine operation monitoring data and blade vibration fatigue events. This involves a comprehensive assessment of the blade structural state and fatigue damage risk corresponding to the impeller impact response coefficient YL under different operating conditions. Combined with industry vibration monitoring standards and material fatigue safety specifications, along with expert experience, reasonable thresholds for multiple safety risk levels are determined. Referring to relevant mechanical vibration safety standards, corrosion fatigue assessment specifications, and blade durability limit parameters provided by equipment manufacturers, the third threshold Q3 is set to scientifically distinguish between normal, first-level abnormal, and second-level abnormal states, guiding the implementation of early warning response and maintenance strategies, and ensuring the long-term safe and stable operation of turbine blades and critical shaft structures.

[0047] In this embodiment, the vibration fatigue monitoring module constructs the impeller impact response coefficient YL and performs multi-level comparative analysis in conjunction with a third threshold Q3 to achieve graded diagnosis of turbine blade fatigue state. When YL is within the normal range, the system operates stably. When YL exceeds Q3 but does not exceed 130% of Q3, laser multi-point interferometric vibration measurement and high-spectral anomaly analysis are automatically triggered to accurately capture the micro-perturbation resonance trend. This, combined with limiting the rate of change of steam pressure and reducing the proportion of high-sulfur coal, fundamentally alleviates the risk of impact and corrosion coupling. When YL exceeds 130% of Q3, a maintenance work order is triggered to automatically execute the blade assembly replacement operation, effectively preventing fatigue damage from expanding to an uncontrollable range. This solution significantly improves the accuracy and response efficiency of fatigue state identification of key turbine components, and enhances the operational safety and maintenance intelligence level of the system in complex vibration and corrosion environments.

[0048] Example 10 A method for operation control of thermal power plants based on multi-source data analysis; please refer to [link / reference]. Figure 2 This includes the following steps: Step 1: By monitoring key equipment in the operation and control system of the thermal power plant, collect data on flue gas oxygen content, boiler temperature, ambient relative humidity, fuel volatile matter, fuel fixed carbon content, fuel ash content, and fuel sulfur content to establish the first dataset; collect data on turbine speed, shaft vibration main frequency component, main steam temperature, and power generation load to establish the second dataset. Step 2: Preprocess the collected data, including outlier removal, signal smoothing, missing data completion, time series reconstruction, and normalization. Step 3: By monitoring the boiler operating efficiency, calculate and obtain the combustion efficiency coefficient RSx, and compare it with the first threshold Q1 to determine whether the boiler's combustion efficiency is qualified. If it is not qualified, a strategy will be given. Step 4: By addressing the unevenness of the turbine's operating response during load changes, quantify the impact of load disturbances on equipment operation, calculate and obtain the operating load stability coefficient FHx, and compare it with the second threshold Q2 to determine whether the operating load fluctuation of the turbine generator set is normal. If abnormal, provide a strategy. Step 5: Identify the potential impact and fatigue risks of high-frequency disturbances on the turbine blade structure, calculate and obtain the impeller impact response coefficient YL, and compare and analyze it with the third threshold Q3 to determine whether the turbine blade condition is normal. If it is abnormal, a strategy is given.

[0049] In this embodiment, a complete thermal power plant operation control method is constructed through a multi-step collaborative control process, encompassing data acquisition, preprocessing, multi-dimensional coefficient evaluation, and strategy generation. The system first establishes a multi-source dataset covering boiler, turbine, and fuel characteristics, and performs high-quality preprocessing on the acquired data, including anomaly removal, signal smoothing, time-series completion, and normalization, to improve data availability and model response stability. Based on this, the combustion efficiency coefficient RSx, operating load stability coefficient FHx, and impeller impact response coefficient YL are calculated sequentially and compared with corresponding thresholds Q1, Q2, and Q3 to identify combustion efficiency anomalies, load disturbance risks, and vibration fatigue trends, respectively. Each module is equipped with a differentiated control strategy to achieve early intervention and graded response to problems, significantly improving the intelligent perception capability, system stability, and equipment safety of thermal power plant operation.

[0050] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0051] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. A thermal power plant operation control system based on multi-source data analysis, characterized in that, include: Data acquisition module, data processing module, combustion efficiency evaluation module, load response analysis module, and vibration fatigue monitoring module; The data acquisition module is used to collect data on flue gas oxygen content, boiler temperature, ambient relative humidity, fuel volatile matter, fuel fixed carbon content, fuel ash content, and fuel sulfur content by monitoring key equipment in the operation control system of the thermal power plant, and to establish a first dataset; and to collect data on turbine speed, shaft vibration main frequency component, main steam temperature, and power generation load, and to establish a second dataset. The data processing module is used to preprocess the collected data, including outlier removal, signal smoothing, missing data completion, time series reconstruction and normalization. The combustion efficiency evaluation module is used to monitor the boiler operating efficiency, calculate the combustion efficiency coefficient RSx, compare it with the first threshold Q1, and determine whether the boiler's combustion efficiency is qualified. If it is not qualified, a strategy is given. The load response analysis module is used to quantify the impact of load disturbance on equipment operation by considering the unevenness of the turbine's operating response during load changes, calculate and obtain the operating load stability coefficient FHx, and compare it with the second threshold Q2 to determine whether the operating load fluctuation of the turbine generator set is normal. If it is abnormal, a strategy is given. The vibration fatigue monitoring module is used to identify the potential impact and fatigue risk of high-frequency disturbances on the turbine blade structure, calculate and obtain the impeller impact response coefficient YL, and compare and analyze it with the third threshold Q3 to determine whether the turbine blade is in normal condition. If it is abnormal, a strategy is given.

2. The thermal power plant operation control system based on multi-source data analysis according to claim 1, characterized in that, The data acquisition module is used to monitor the combustion exhaust gas in the boiler flue in real time by deploying a flue gas analyzer above the boiler in a thermal power plant, and to collect the oxygen content of the flue gas. By deploying infrared temperature measuring devices and thermocouple arrays, the boiler furnace is monitored in real time, and the boiler temperature is collected. By deploying environmental humidity sensors, the relative humidity of the environment is collected. By deploying coal quality testing equipment, fuel samples in the coal conveying system are monitored in real time, and the volatile matter (VM), fixed carbon content (FC), ash content (Ash), and sulfur content (SC) of the fuel are collected to establish the first dataset. By deploying turbine condition monitoring sensors, the turbine shaft system is monitored in real time, and the turbine speed is collected. By deploying vibration acceleration sensors and FFT analysis technology, the turbine bearing housing is monitored in real time, and the main frequency component fv of shaft vibration is collected. By deploying main steam temperature sensors, the main steam pipeline is monitored in real time, and the main steam temperature Ts is collected. By deploying power metering and load analysis devices, the power generation bus is monitored in real time, and the power generation load Pg is collected. A second dataset is then established.

3. The thermal power plant operation control system based on multi-source data analysis according to claim 2, characterized in that, The data processing module is used to preprocess the collected data, including: using outlier removal technology to remove abrupt changes and outliers during the monitoring process; using a sliding window and weighted average algorithm to smooth high-frequency disturbance signals; using time series interpolation and multi-source correlation completion methods to fill in intermittent missing data; performing signal synchronization and time series reconstruction for multi-channel sensor data; and normalizing all numerical variables.

4. The thermal power plant operation control system based on multi-source data analysis according to claim 1, characterized in that, The combustion efficiency evaluation module includes a first calculation unit and a first analysis unit; The first computing unit is used to monitor the boiler operating efficiency and extract the oxygen content of the flue gas from the first dataset. Boiler temperature Ambient relative humidity value The volatile matter (VM), fixed carbon content (FC), ash content (Ash), and sulfur content (SC) of the fuel are dimensionlessly processed to calculate the combustion efficiency coefficient (RSx).

5. A thermal power plant operation control system based on multi-source data analysis according to claim 4, characterized in that, The first analysis unit is used to preset a first threshold Q1, and compare the combustion efficiency coefficient RSx with the first threshold Q1 to obtain a first evaluation result, including: When the combustion efficiency coefficient RSx ≥ the first threshold Q1, it indicates that the boiler's combustion efficiency is qualified and should be continuously monitored. When the combustion efficiency coefficient RSx < the first threshold Q1, it indicates that the boiler's combustion efficiency is unqualified, triggering the first early warning command and generating the first strategy: by adjusting the ratio of primary air volume to secondary air volume to 1:1.8-1:2.2, the oxygen content of the flue gas is reduced. Control it between 3% and 4.5%; adjust the pulverized coal ratio into the furnace, prioritizing the addition of coal with high calorific value. Fuel with a concentration of ≥25 MJ / kg should account for no less than 20%; the air preheater regeneration channel should be activated to increase the temperature of the pre-combustion mixture by 30% and improve ignition stability.

6. The thermal power plant operation control system based on multi-source data analysis according to claim 1, characterized in that, The load response analysis module includes a second calculation unit and a second analysis unit; The second calculation unit is used to extract the power generation load Pg and turbine speed from the second dataset by considering the unevenness of the turbine's operating response during load changes. and main steam temperature As an observed variable, the impact of load disturbance on equipment operation is quantified by calculating the change rate per unit time. After dimensionless processing, the operating load stability coefficient FHx is calculated and obtained.

7. A thermal power plant operation control system based on multi-source data analysis according to claim 6, characterized in that, The second analysis unit is used to pre-set a second threshold Q2 and compare the operating load stability coefficient FHx with the second threshold Q2 to obtain a second evaluation result, including: When the operating load stability coefficient FHx ≤ the second threshold Q2, it indicates that the operating load fluctuation of the steam turbine generator unit is normal and should be continuously monitored. When the operating load stability coefficient FHx ≤ the second threshold Q2, it indicates abnormal fluctuations in the operating load of the steam turbine generator unit, triggering a second early warning command and generating a second strategy: limiting the rate of change of the turbine main steam valve opening to within 2% / s to mitigate the risk of high-frequency power surges eroding the steam passages, and using the SCADA monitoring system to monitor the power generation load. If the fluctuation is within ±8% for 5 consecutive minutes, the output shaft torque fluctuation should be controlled to be no higher than ±3%.

8. A thermal power plant operation control system based on multi-source data analysis according to claim 1, characterized in that, The vibration fatigue monitoring module includes a third calculation unit and a third analysis unit; The third computing unit is used to identify the potential impact and fatigue risks of high-frequency disturbances on the turbine blade structure, and to extract the turbine speed from the second dataset. The acceleration change, the dominant frequency component of shaft system vibration Using the rate of change as a dynamic indicator, and combining the sulfur content SC of the fuel in the first dataset as a corrosion influencing factor, the impeller impact response coefficient YL was calculated after dimensionless processing.

9. A thermal power plant operation control system based on multi-source data analysis according to claim 8, characterized in that, The third analysis unit is used to obtain a third evaluation result by pre-setting a third threshold Q3 and comparing the impeller impact response coefficient YL with the third threshold Q3. When the impeller impact response coefficient YL≤ the third threshold Q3, it indicates that the turbine blades are in normal condition and there is no risk of potential cumulative damage to the shaft structure caused by high-frequency vibration and speed impact in a high-sulfur environment. Continuous monitoring is required. When the third threshold Q3*130% ≥ the impeller impact response coefficient YL > the third threshold Q3, it indicates that the turbine blade condition is at level one abnormality, and there is a potential risk of cumulative damage to the shaft structure caused by high-frequency vibration and speed impact in a high-sulfur environment. This triggers the third early warning command and generates the third strategy: using laser multi-point interferometric vibration measurement technology to perform non-contact condition monitoring of the turbine blade, capturing the micro-perturbation characteristics of the vibration response curve in real time, and simultaneously increasing the sampling frequency of the shaft vibration signal to ≥1kHz, covering a high-frequency bandwidth of 200–500Hz, performing dynamic analysis of spectral anomalies, identifying potential resonance components and their evolution trends; adjusting the main steam valve response hysteresis curve in conjunction with the control, limiting the steam dynamic pressure slope to no more than 0.6MPa / s, indirectly suppressing the repeated stress loading on the blade structure by the surge in torque through the slow release of steam thrust impact process, thereby reducing the fatigue accumulation rate; and reducing the proportion of high-sulfur coal to ≤50% to alleviate the impact of increased corrosion on fatigue accumulation. When the impeller impact response coefficient YL > the third threshold Q3*130%, it indicates that the turbine blade is in a level two abnormal condition and the fatigue damage trend of the turbine blade has reached an irreversible level, triggering the fourth warning instruction and generating the fourth strategy: automatically triggering a maintenance work order to replace the target blade assembly.

10. A method for operating control of a thermal power plant based on multi-source data analysis, comprising a thermal power plant operation control system based on multi-source data analysis according to claims 1-9, characterized in that, Includes the following steps: Step 1: By monitoring key equipment in the operation and control system of the thermal power plant, collect data on flue gas oxygen content, boiler temperature, ambient relative humidity, fuel volatile matter, fuel fixed carbon content, fuel ash content, and fuel sulfur content to establish the first dataset; collect data on turbine speed, shaft vibration main frequency component, main steam temperature, and power generation load to establish the second dataset. Step 2: Preprocess the collected data, including outlier removal, signal smoothing, missing data completion, time series reconstruction, and normalization. Step 3: By monitoring the boiler operating efficiency, calculate and obtain the combustion efficiency coefficient RSx, and compare it with the first threshold Q1 to determine whether the boiler's combustion efficiency is qualified. If it is not qualified, a strategy will be given. Step 4: By addressing the unevenness of the turbine's operating response during load changes, quantify the impact of load disturbances on equipment operation, calculate and obtain the operating load stability coefficient FHx, and compare it with the second threshold Q2 to determine whether the operating load fluctuation of the turbine generator set is normal. If abnormal, provide a strategy. Step 5: Identify the potential impact and fatigue risks of high-frequency disturbances on the turbine blade structure, calculate and obtain the impeller impact response coefficient YL, and compare and analyze it with the third threshold Q3 to determine whether the turbine blade condition is normal. If it is abnormal, a strategy is given.