Intelligent disk monitoring method and device based on DCS (Distributed Control System) thermal power plant high-pressure heating system

Abstract

The invention provides an intelligent disk monitoring method and device based on a DCS (Distributed Control System) thermal power plant high-pressure heater system, and aims to solve the problems that the traditional constant value alarm has many false alarms, early-stage abnormity cannot be identified, the handling capacity of operators is insufficient and the like. According to the method, a multivariate regression and multivariate state estimation method is adopted, historical operation data modeling is conducted on the high-pressure heater under the stable working condition that the unit load is larger than or equal to 280 MW, and the key parameter deviation condition is monitored in real time. Through three types of functions of a fluctuation value, a reference value and a predicted value, three typical faults of high-pressure heater performance reduction, steam inlet pipeline valve jamming and leakage risk are accurately identified, grading early warning is triggered according to expert rules, and processing suggestions are synchronously pushed. The system is deployed and applied on a DCS platform, the high-pressure heater performance degradation trend is successfully found in advance, the cold source loss is effectively reduced, the manual monitoring intensity and the equipment maintenance cost are reduced, and the safety, the economical efficiency and the intelligent level of unit operation are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring and fault diagnosis technology for high-pressure heater systems in thermal power plants, and particularly to a smart monitoring method and device for high-pressure heater systems in thermal power plants based on DCS. Background Technology

[0002] As a core auxiliary equipment in the thermal cycle of a thermal power plant, the high-pressure heater system's operating status directly affects the unit's thermal efficiency (approximately 1% change in terminal temperature difference affects thermal efficiency by 0.2%), cold source loss, and equipment lifespan. With the development of smart power generation technology, traditional static setpoint alarm systems based on DCS are no longer sufficient to meet the refined operation and maintenance needs of modern power plants. Existing technologies typically employ a single-point threshold alarm strategy, constructing a basic monitoring framework through analog quantity acquisition modules and switch quantity monitoring units in a distributed control system (DCS). The indicator alarm system, as a core component, primarily relies on preset threshold parameters such as temperature, pressure, and water level for fault identification. Specifically, this technology system covers the entire process from data acquisition and signal processing to alarm triggering, including key aspects such as real-time database storage, historical data archiving, and setpoint alarm logic.

[0003] However, existing DCS alarm systems have systemic flaws. Specifically, traditional fixed-value alarm strategies do not consider the dynamic characteristics of load conditions, resulting in a large number of false alarms even during stable operation phases where the unit load is ≥280MW and the load change within 10 minutes is ≤35MW (a certain 600MW unit averages over 200 false alarms per day). Furthermore, the static alarm mechanism cannot capture the performance degradation process of the difference between the current and historical values ​​at the heater top (ΔT>2℃), triggering an alarm only when ΔT>5℃, by which time significant thermal efficiency loss has already occurred. Based on this, existing high-pressure heater leakage early warning systems have a 3-5 hour lag and cannot identify micro-leakage with water level changes <1mm, resulting in a fault diagnosis accuracy rate of less than 70%. Due to the lack of multi-parameter dynamic correlation analysis capabilities, operators need to manually analyze 10+ parameters, with decision-making time exceeding 15 minutes (the golden response time for accidents is <5 minutes), potentially leading to equipment damage and unplanned downtime risks. Industry data shows that high-pressure heating system failures account for more than 35% of auxiliary equipment failures in thermal power plants, with an average annual maintenance cost exceeding one million yuan. There is an urgent need to build an intelligent early warning system based on multi-technology collaboration. Summary of the Invention

[0004] The present invention aims to at least partially solve one of the technical problems in the related art.

[0005] Therefore, the first objective of this invention is to propose a smart monitoring method based on the DCS high-pressure heater system of a thermal power plant.

[0006] The second objective of this invention is to propose an intelligent monitoring device based on the DCS high-pressure heater system of a thermal power plant.

[0007] The third objective of this invention is to provide a computer device.

[0008] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above objectives, a first aspect of the present invention proposes a smart monitoring method based on a DCS high-pressure heater system in a thermal power plant, comprising: S1 collects multi-source operating data of the high-voltage heater system, including analog data, digital data and equipment status data; S2 cleans, aligns, and removes bad values ​​from the collected data to generate structured data, which is then stored in the real-time database and the historical database respectively. S3, based on structured data and thermodynamic principles, establishes a dynamic benchmark performance model to calculate the deviation between the real-time values ​​of key performance indicators and historical health benchmark values. S4 uses multi-parameter cross-validation logic to determine whether the deviation meets the preset fault warning conditions. If it does, it triggers a warning and generates a fault diagnosis conclusion.

[0010] In one embodiment of the present invention, S1 includes: S11 collects analog data including unit load, heater inlet and outlet feedwater temperature, condensate temperature, steam inlet temperature, extraction steam pressure, steam inlet pressure, and water level parameters. S12 collects switch data including the open / close status of the high-pressure heater inlet and outlet three-way valves, the steam inlet electric valve, and the steam inlet check valve.

[0011] In one embodiment of the present invention, S2 includes: S21 uses a bad value removal method based on the 3σ principle to filter out outlier data that deviates from the mean by more than three standard deviations. S22 uses an interpolation algorithm to align time-series data, ensuring that data with different sampling frequencies are synchronized at the same timestamp.

[0012] In one embodiment of the present invention, S3 includes: S31. A nonlinear relationship model between the upper and lower differences of the heater and the load and extraction steam pressure was established using a multiple regression algorithm. The model's coefficient of determination R² is greater than 0.92. S32, calculate the deviation between the real-time value of the entire high-pressure heater's temperature rise and the historical health baseline value. The deviation formula is:

[0013] in For real-time inlet water temperature, Based on historical baseline inlet water temperature, For real-time outlet water temperature, The historical baseline outlet water temperature.

[0014] To achieve the above objectives, a second aspect of the present invention provides a smart monitoring device based on a DCS high-pressure heater system in a thermal power plant, comprising: The multi-source data acquisition module is used to collect multi-source operating data of the high-voltage heater system, including analog data, digital data and equipment status data; The data cleaning and storage module is used to clean, time-series align, and remove bad values ​​from the collected data, generating structured data and storing it in the real-time database and historical database respectively. The dynamic benchmark performance model building module is used to build a dynamic benchmark performance model based on structured data and thermodynamic principles, and calculate the deviation between the real-time values ​​of key performance indicators and historical health benchmark values. The multi-parameter cross-validation and early warning module is used to determine whether the deviation meets the preset fault early warning conditions through multi-parameter cross-validation logic. If it does, an early warning is triggered and a fault diagnosis conclusion is generated.

[0015] The present invention discloses a smart monitoring method and device for high-pressure heater systems in DCS thermal power plants, which can realize accurate early warning and diagnosis of early faults in high-pressure heater systems, significantly improve the timeliness and accuracy of identifying leaks, valve jamming and performance degradation, and improve the safety and economy of thermal power plant operation.

[0016] To achieve the above objectives, a third aspect of this application provides a computer device, including a processor and a memory; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement a smart monitoring method based on a DCS thermal power plant high-pressure heater system as described in the first aspect embodiment.

[0017] To achieve the above objectives, the fourth aspect of this application proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a smart monitoring method based on a DCS thermal power plant high-pressure heater system as described in the first aspect embodiment.

[0018] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0019] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1This is a flowchart of a smart monitoring method based on a DCS high-pressure heater system in a thermal power plant according to an embodiment of the present invention; Figure 2 This is a system overall architecture diagram according to an embodiment of the present invention; Figure 3 This is a diagram of the intelligent monitoring and visualization interface of the high-voltage heater system according to an embodiment of the present invention; Figure 4 This is a structural diagram of a smart monitoring device based on a DCS high-pressure heater system in a thermal power plant according to an embodiment of the present invention; Figure 5 It is a computer device according to an embodiment of the present invention. Detailed Implementation

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0022] The following description, with reference to the accompanying drawings, describes a smart monitoring method and apparatus based on a DCS high-pressure heater system in a thermal power plant, according to an embodiment of the present invention.

[0023] Example 1 Figure 1 This is a flowchart of a smart monitoring method based on a DCS high-pressure heater system in a thermal power plant according to an embodiment of the present invention, such as... Figure 1 As shown, it includes: S1 collects multi-source operating data of the high-voltage heating system, including analog data, digital data, and equipment status data.

[0024] Specifically, in this invention, collecting multi-source operational data from the high-voltage heater system is a fundamental step in realizing the functions of the intelligent monitoring system. Its technical implementation principle is based on a real-time data interaction mechanism between the distributed control system (DCS) and the high-voltage heater system. This step establishes a bidirectional communication link with the high-voltage heater system's real-time database (RDB) and historical database (HDB) through software tools such as Ovation Analytics Studio and Ovation Intelligence Framework within the DCS system, enabling comprehensive perception of the high-voltage heater system's operational status.

[0025] Furthermore, the data acquisition and interface layer reads analog, digital, and equipment status data from the DCS system in real time via the standard OPC (OLE for Process Control) protocol or a dedicated communication interface. Analog data includes unit load, inlet / outlet feedwater temperature, condensate temperature, steam inlet temperature, extraction steam pressure, steam inlet pressure, and water level for each high-pressure heater (#6A / B, #7A / B, #8A / B). The sampling frequency is typically set to 1 to 10 seconds to meet real-time requirements. Digital data includes the open / closed status of the high-pressure heater inlet / outlet three-way valves, steam inlet electric valves, and check valves, used to determine whether the system is in normal operating mode. Equipment status data reflects whether the high-pressure heaters are in operation or disconnected, providing operational condition data for subsequent analysis.

[0026] Furthermore, analog data must meet the accuracy requirements for process variables in the IEC 61131-3 standard; temperature measurement error should be controlled within ±0.5℃, and pressure measurement error should be controlled within ±0.1MPa. Switching signals must conform to the definition standards of logic signals in the DCS system to ensure the accuracy and consistency of status acquisition. Timing alignment processing is also required during data acquisition to eliminate data deviations caused by asynchronous sampling times.

[0027] Furthermore, in practical applications, this step is deployed in the DCS system of thermal power plants, suitable for monitoring the operation of high-pressure heater systems under different load conditions. By collecting multi-source data, the system can provide high-quality, structured input for subsequent intelligent analysis engines, thereby achieving accurate identification and early warning of high-pressure heater performance degradation, valve jamming, and leakage risks.

[0028] Furthermore, this step ensures the system has a comprehensive understanding of the high-pressure heater's operating status, providing reliable data support for anomaly detection and fault diagnosis. Through real-time data acquisition and processing, the system can effectively reduce the misjudgment rate caused by missing or delayed data, improve the accuracy and timeliness of fault warnings, and thus enhance the safety and economy of unit operation.

[0029] Furthermore, S1 includes: S11 collects analog data including unit load, heater inlet and outlet feedwater temperature, condensate temperature, steam inlet temperature, extraction steam pressure, steam inlet pressure, and water level parameters.

[0030] Specifically, in the intelligent monitoring system of this invention, collecting analog data is a fundamental step in realizing high-pressure heater system performance monitoring and fault early warning. This step, through integration with the DCS system, acquires key analog parameters of the high-pressure heater system in real time, including unit load, heater inlet and outlet feedwater temperatures, condensate temperatures, steam inlet temperature, extraction steam pressure, steam inlet pressure, and water level. These parameters reflect the thermodynamic characteristics, fluid dynamics behavior, and equipment operating status of the high-pressure heater system, and are the core input data for subsequent intelligent analysis engines to perform performance evaluation, fault diagnosis, and optimization suggestions.

[0031] Furthermore, the data acquisition module establishes a bidirectional communication link with the DCS's historical stations and real-time database through software tools. The acquisition frequency is typically set to 1 to 10 times per second to ensure data real-time performance and continuity. During acquisition, the system uses the OPC protocol or the DCS's native interface to read data, ensuring compatibility with existing control systems. The acquired data format must conform to the IEEE 754 standard, with temperature data in degrees Celsius (…). The units are as follows: pressure data is in megapascals (MPa), and water level data is in millimeters (mm) to ensure data consistency and processability.

[0032] Furthermore, the system performs quality control on the collected data, including digital filtering (such as a low-pass filter with a cutoff frequency of 0.1Hz), out-of-value removal (such as the 3σ principle), and standardized conversion of engineering units. For example, temperature data needs to be filtered to remove values ​​exceeding ±5 Hz. For outliers, pressure data must be filtered to remove abnormal fluctuations exceeding ±0.2 MPa, and water level data must be filtered to remove abnormal deviations exceeding ±5 mm. Simultaneously, the system performs time-series alignment of the data to ensure consistency in timestamps across all parameters, supporting subsequent multi-parameter fusion analysis.

[0033] Furthermore, this step is widely applied to online monitoring and fault early warning of high-pressure heater systems in thermal power plants. When the unit load is stable (≥280MW and load change ≤35MW within 10 minutes) and the high-pressure heater is in normal operation, the system continuously collects the aforementioned analog data to determine whether there is a performance degradation in the high-pressure heater, or a risk of valve jamming or leakage in the steam inlet pipeline. For example, when there is a significant deviation between the upper and lower temperature differences of the heaters or the overall temperature rise of the high-pressure heaters and historical benchmark values, the system will trigger an early warning mechanism to provide timely diagnostic information to operators.

[0034] Furthermore, by acquiring high-precision, high-frequency data, a reliable data foundation is provided for the intelligent analysis engine, thereby enabling dynamic perception and anomaly identification of the high-pressure heater system's operating status. Compared to traditional static alarm methods, this invention significantly improves the accuracy and timeliness of fault early warning through multi-parameter fusion analysis, providing solid data support for subsequent fault diagnosis and optimized control, and effectively improving the safety and economy of thermal power unit operation.

[0035] S12 collects switch data including the open / close status of the high-pressure heater inlet and outlet three-way valves, the steam inlet electric valve, and the steam inlet check valve.

[0036] Specifically, in the intelligent monitoring system of this invention, the open / closed status of key valves in the high-pressure heater system is collected, including the high-pressure heater inlet and outlet three-way valves, the steam inlet electric valve, and the steam inlet check valve. The status of these valves directly reflects whether the high-pressure heater system is in normal operation, disconnected, or partially restricted, and is an important basis for judging whether the system's operating logic is complete and whether the equipment has experienced faults such as jamming or leakage.

[0037] Furthermore, this step establishes a communication link with the DCS real-time database through a software module in the DCS system, enabling real-time acquisition of the valve status. The acquisition frequency is typically set to once per second (1Hz) to ensure timely response to changes in system status. The acquired data is in Boolean format, representing "open" or "closed" states, corresponding to digital signal input (DI) points in the DCS system. The signal level is typically DC24V or DC48V, conforming to the definition of digital input signals in the IEC 61131-3 standard.

[0038] Furthermore, the acquired digital input data needs to be time-aligned with the analog input data to ensure that the timestamps of each signal are consistent during multi-parameter fusion analysis. Time alignment typically employs interpolation or synchronous sampling mechanisms, with time deviations controlled within ±50ms to meet the power system's requirements for real-time performance and data consistency. The acquired digital input data will be stored in a real-time database and used in the equipment status assessment module and multi-parameter fusion early warning module within the intelligent analysis engine layer.

[0039] Furthermore, it has significant engineering value, particularly during the operation of high-pressure heater systems in thermal power plants. For example, in the diagnostic logic for valve jamming in the high-pressure heater inlet pipeline, the system needs to confirm whether the three-way valve, electric valve, and check valve are all fully open to rule out misjudgments caused by valves not being fully open. Similarly, in the diagnosis of high-pressure heater leakage risks, collecting valve status data helps determine whether the system is in a stable operating state, thereby improving the accuracy of the diagnostic logic.

[0040] Furthermore, collecting switch data not only provides the system with crucial operational status information, but also provides a logical basis for subsequent intelligent diagnosis and early warning, making it an indispensable foundation for realizing intelligent monitoring of high-voltage heating systems.

[0041] S2 cleans, aligns, and removes bad values ​​from the collected data, generating structured data which is then stored in the real-time database and the historical database, respectively.

[0042] Specifically, in the data processing and storage layer, cleaning, time-series alignment, and bad value removal of the collected high-performance computing system data are key steps in constructing a high-quality structured dataset. This step is technically implemented based on data preprocessing algorithms and time-series consistency verification mechanisms, ensuring that the subsequent intelligent analysis engine layer can perform modeling and diagnosis based on accurate and consistent data.

[0043] Furthermore, the data cleaning process includes identifying and correcting invalid, outlier, and missing values ​​in the original data. Specifically, the system employs statistical threshold-based detection methods, such as the 3σ principle or the IQR (interquartile range) method, to identify outliers in the analog data. For example, if the high-pressure steam inlet pressure at a certain moment deviates from its average value within a 10-minute sliding window by more than a set threshold... If a value is not found, it is marked as a bad value and removed. Furthermore, the system supports rule-based cleaning strategies, such as performing state consistency checks on switch signals to ensure that the state signals of devices such as three-way valves, electric valves, and check valves are logically consistent.

[0044] Furthermore, time alignment is a key technology for ensuring the consistency of multi-source data across the time dimension. Due to factors such as different sensor sampling frequencies and communication delays, the original data may have timestamp offsets. The system employs timestamp interpolation and resampling techniques to uniformly align all data to a 1-second time resolution and eliminates instantaneous jitter through time window sliding averaging. This process follows the specifications for time synchronization and data alignment in the IEEE 1451 standard.

[0045] Furthermore, the processed data is structured into a unified data model, including fields such as timestamp, device identifier, parameter type, numerical value, and quality code, and stored in a real-time database and a historical database respectively according to the data's timeliness. The real-time database is used to support online analysis with sub-second response times, while the historical database is used for offline modeling and trend analysis.

[0046] Furthermore, it provides a reliable data foundation for the intelligent analysis engine layer. Through data cleaning and alignment, the system effectively reduces the false alarm rate caused by data quality issues, improves the accuracy of model training and the robustness of diagnostic logic, thereby enabling early identification and accurate warning of high-precision system faults.

[0047] Furthermore, S2 includes: S21 uses a bad value removal method based on the 3σ principle to filter out outlier data that deviates from the mean by more than three standard deviations.

[0048] Furthermore, the specific technical implementation of this step is as follows: First, the system statistically analyzes the historical data of a certain operating parameter (such as the upper and lower temperature differences of the heater, temperature rise, and steam inlet pressure) under the same load conditions, and calculates its average value. and standard deviation Subsequently, the currently collected real-time data points were analyzed. Make a judgment if it satisfies If any data point falls outside the normal distribution, it is considered outlier and removed. This method is based on the property of a normal distribution, meaning that ideally, 99.73% of the data points should fall within the normal distribution. Data points outside the specified range are usually caused by noise or measurement error.

[0049] Furthermore, the system sets the rejection threshold to 3σ, which is three times the standard deviation. In practical applications, this method is suitable for analog parameters with continuous distribution characteristics in high-pressure heater systems, such as temperature, pressure, and water level. Under the conditions of a data acquisition frequency of 1 second / point and a historical data window length of 720 hours, the system can dynamically update the mean and standard deviation, ensuring that the rejection logic adapts to changes in operating conditions.

[0050] Furthermore, this step is widely used in the preprocessing stage of high-pressure heater system operating data, especially when the unit load is stable (≥280MW and 10-minute load change ≤35MW), to detect anomalies in high-pressure heater performance indicators. By eliminating abnormal data, the system can effectively avoid false alarms caused by transient interference or sensor failure, improving the accuracy of equipment condition assessment and fault diagnosis.

[0051] Furthermore, this step significantly improves the reliability of the data and the stability of the model. After removing outliers that deviate from the mean by more than three standard deviations, the performance model based on historical data in the intelligent analysis engine layer can more accurately reflect the true state of the equipment, thereby improving the identification accuracy of faults such as high-pressure heater performance degradation, valve jamming, and leakage risks. In its implementation at the Sutong Power Plant, this method effectively reduced the false judgment rate caused by data noise, improved the reliability of system early warning, and provided solid data support for realizing intelligent monitoring of the high-pressure heater system.

[0052] S22 uses an interpolation algorithm to align time-series data, ensuring that data with different sampling frequencies are synchronized at the same timestamp.

[0053] Specifically, in the data processing and storage layer, aligning time-series data using interpolation algorithms is a crucial step in realizing the multi-source data fusion analysis of the intelligent monitoring system for high-voltage systems. The technical principle behind this step is based on interpolation methods in time-series signal processing, aiming to solve the problem of inconsistent timestamps caused by different sensors or data acquisition channels having different sampling frequencies. This ensures that in subsequent intelligent analysis, all parameters can be compared and modeled on a unified time axis.

[0054] Furthermore, the system employs methods such as linear interpolation or spline interpolation to align the acquired analog and digital data in time. Specifically, the system first retrieves the raw time-series data from the DCS's real-time database for each measuring point, including key parameters such as feedwater temperature at the high-pressure heater inlet and outlet, condensate temperature, steam inlet pressure, extraction steam pressure, and water level. Since the sampling frequencies of these parameters may differ (e.g., some parameters are sampled once per second, others once per minute), the system maps all data to a unified time base, typically aligning it in seconds. During interpolation, the system estimates the parameter values ​​at unknown time points based on the known data values ​​using an interpolation algorithm, ensuring comparability of all data at the same timestamp.

[0055] Furthermore, the accuracy of the interpolation algorithm directly affects the reliability of subsequent models. In this system, the interpolation error is controlled within ±0.5%, and the time alignment step is 1 second to ensure that the temporal resolution of all parameters is consistent within a 10-minute sliding window. In addition, the system supports preprocessing of abnormal timestamps before interpolation, such as removing bad values ​​and filling in missing values, to improve the robustness of the interpolation results.

[0056] Furthermore, this step is widely applied to modules such as high-pressure heater performance evaluation, valve jamming diagnosis, and leakage early warning. For example, in diagnosing performance degradation in high-pressure heaters, the system needs to compare the terminal difference changes under the current and historical loads. If the timestamps are inconsistent, it will lead to misjudgment or missed judgment. Through interpolation alignment, the system can ensure that the parameters are comparable under the same load and operating conditions, thereby improving the accuracy of the diagnosis.

[0057] Furthermore, this step effectively addresses the inconsistency issue of multi-source heterogeneous data in the time dimension, providing a high-quality, structured data foundation for subsequent multi-parameter fusion early warning and fault diagnosis. Through precise time alignment, the system can identify performance anomalies, valve jamming, and leakage risks earlier, significantly improving the operational safety and economy of the high-pressure heater system.

[0058] S3, based on structured data and thermodynamic principles, establishes a dynamic benchmark performance model to calculate the deviation between the real-time values ​​of key performance indicators and historical health benchmark values.

[0059] Specifically, within the intelligent analysis engine layer, establishing a dynamic benchmark performance model based on structured data and thermodynamic principles is a core technical step for achieving health status assessment and fault early warning of high-voltage heater systems. This step involves collecting key operating parameters of the high-voltage heater system under different loads and conditions, combining them with a thermodynamic heat transfer model to construct a benchmark performance index system reflecting the health status of the equipment, and calculating in real time the deviation between the current performance index and historical health benchmark values, thereby achieving sensitive detection of changes in high-voltage heater performance.

[0060] Furthermore, this step first acquires cleaned, filtered, and time-aligned structured data through data processing and storage, including analog signals such as unit load, feedwater temperature at the high-pressure heater inlet / outlet, condensate temperature, steam inlet temperature, extraction steam pressure, steam inlet pressure, and water level, as well as digital signals such as valve open / close status. This data is aligned according to timestamps to ensure performance comparisons under the same load and similar operating conditions.

[0061] Furthermore, the dynamic benchmark performance model is based on the principle of thermodynamic heat transfer and uses a multiple regression method to model historical health data. Key performance indicators output by the model include the heater top-end temperature difference, bottom-end temperature difference, and overall high-pressure heater temperature rise.

[0062] Furthermore, in practical applications, this step operates on top of the DCS system, with model deployment and real-time calculations achieved through the OIF platform. Its technical value lies in enabling dynamic assessment of the health status of the high-pressure heater system by quantifying performance deviations, providing data support for subsequent fault diagnosis and optimized control, thereby improving the safety, economy, and automation level of thermal power plant operation.

[0063] Furthermore, S3 includes: S31. A nonlinear relationship model between the upper and lower differences of the heater and the load and extraction steam pressure was established using a multiple regression algorithm. The model's coefficient of determination R² is greater than 0.92.

[0064] Specifically, in some implementations, this invention employs a multiple regression algorithm to establish a nonlinear relationship model between the heater's upper and lower pressure differences and the load and extraction steam pressure, thereby achieving accurate identification and early warning of performance changes in the high-pressure heater system. The technical principle behind this step is based on statistical modeling and thermodynamic characteristic analysis. By collecting historical data of the high-pressure heater system under different operating conditions, key variables (such as unit load, extraction steam pressure, heater upper and lower pressure differences, etc.) are extracted, and a mathematical mapping relationship between the variables is constructed using a multiple nonlinear regression method.

[0065] First, historical data of the high-pressure heater system under stable operating conditions is extracted from the DCS system. The data acquisition cycle is typically more than 10 minutes to ensure that the load variation does not exceed 35MW, meeting the operating condition consistency requirements for modeling. Subsequently, the collected data undergoes preprocessing, including digital filtering, outlier removal, and time-series alignment, to improve data quality. After data processing, a multivariate nonlinear regression algorithm (such as multinomial regression, support vector regression (SVR), or random forest regression) is used to analyze the differential pressure at the upper end of the heater. Difference between the lower end and the upper end With load extraction steam pressure Model the relationships between them.

[0066] Furthermore, this step is primarily used for health status assessment and early fault warning of the high-pressure heater system in thermal power plants. When the unit load stabilizes at 280MW or above, the system determines whether there is a risk of performance degradation, valve jamming, or leakage by comparing the model's predicted values ​​with the actual values. The technical value of this step lies in overcoming the insensitivity of traditional static alarm methods to slowly changing faults by establishing a high-precision nonlinear model, thereby improving the accuracy of fault identification and the timeliness of early warning, and providing reliable data support for subsequent fault diagnosis and operational optimization.

[0067] S32, calculate the deviation between the real-time value of the entire high-pressure heater's temperature rise and the historical health baseline value. The deviation formula is: .

[0068] Specifically, the technical implementation of this step is based on thermodynamic principles and historical data modeling, aiming to achieve dynamic assessment of the health status of the high-pressure heating system by quantifying the degree of deviation of performance indicators.

[0069] Furthermore, this step first acquires the real-time temperature rise data of the entire series of high-pressure heaters from the DCS system through the data acquisition and interface layer, including key parameters such as the inlet / outlet feedwater temperature and condensate temperature of each high-pressure heater (e.g., #6A / B, #7A / B, #8A / B). In the data processing and storage layer, this raw data undergoes digital filtering, outlier removal, and time-series alignment before being stored in a real-time database for use by the intelligent analysis engine.

[0070] Furthermore, this step, in practical applications, must meet specific operating conditions, such as stable unit load (load change ≤ 35MW within 10 minutes) and the high-pressure heater system being in normal operating condition (e.g., inlet / outlet three-way valves, inlet electric valve, and check valve are all fully open). Under these conditions, abnormal changes in deviation values ​​have higher diagnostic value.

[0071] Furthermore, by comparing and analyzing real-time and historical data, early deterioration trends in the performance of the high-pressure heater system can be accurately captured, providing a reliable basis for subsequent fault diagnosis and operational optimization. Its innovation lies in introducing a dynamic benchmark comparison mechanism, overcoming the limitations of traditional static threshold alarms, improving the system's early warning sensitivity and accuracy, and thus effectively ensuring the safe and economical operation of thermal power units.

[0072] S4 uses multi-parameter cross-validation logic to determine whether the deviation meets the preset fault warning conditions. If it does, it triggers a warning and generates a fault diagnosis conclusion.

[0073] Specifically, in some implementations, a key step in the intelligent monitoring system of this invention is to determine whether the deviation meets the preset fault warning conditions through multi-parameter cross-validation logic, and to trigger an warning and generate a fault diagnosis conclusion when the conditions are met. This step is based on the comparative analysis of thermodynamic performance models and historical data, combined with the dynamic change trends of multiple operating parameters, and employs a dual judgment mechanism of logic and threshold to achieve accurate identification of potential faults in the high-pressure heater system.

[0074] Furthermore, this step first extracts key parameters of the current operating status from the structured data provided by the data processing and storage layer, including the heater upper and lower temperature differences, the overall high-pressure heater temperature rise, inlet steam pressure, extraction steam pressure, and water level. These parameters are compared with historical health benchmark values ​​under the same load and similar operating conditions to calculate their deviation. For example, in the diagnosis of high-pressure heater performance degradation, if the difference between the current heater upper temperature difference and the historical value exceeds a preset threshold, or if the overall high-pressure heater temperature rise decreases beyond a certain range, an early warning is triggered. In the diagnosis of high-pressure heater inlet pipe valve jamming, the system further combines the deviation between the inlet steam pressure and the extraction steam pressure, as well as the mutual deviation between the inlet steam pressures of high-pressure heaters A and B, to perform multi-parameter cross-validation to improve the accuracy of the diagnosis.

[0075] Furthermore, the load threshold involved in this step is ≥280MW, and the load change within 10 minutes does not exceed 35MW to ensure the system is in a stable operating state. For high-pressure heater leakage risk diagnosis, the system requires that under the same load, the difference between the upper temperature difference and the historical value is ≥3°C, while the difference between the lower temperature difference and the historical value is ≤-1°C, and the absolute value of the high-pressure heater water level deviation is ≤1mm, in order to detect subtle signs of leakage. These parameter settings are based on the typical operating characteristics of high-pressure heater systems in thermal power plants and comply with thermal equipment performance testing standards such as ASME PTC 12.1.

[0076] Furthermore, this step is widely used in the online monitoring and fault early warning of high-pressure heater systems in thermal power plants. The system runs on a DCS platform and achieves deep integration with the DCS through OIF. It can identify the risk of high-pressure heater performance degradation, valve jamming, or leakage before operators notice any abnormalities, and generate structured diagnostic conclusions, which are pushed to the operator station or mobile terminal to assist operators in making rapid responses and decisions.

[0077] Furthermore, through multi-parameter fusion analysis, the system effectively overcomes the problems of high false alarm rate and delayed response of traditional single-point threshold alarms, achieving accurate identification and location of early faults in the high-pressure heater system. For example, the system successfully captured early signals of high-pressure heater performance degradation and avoided increased cold source losses through optimization suggestions, significantly improving the safety and economy of unit operation.

[0078] An intelligent monitoring method for high-pressure heater systems in thermal power plants based on DCS (Distributed Control System) is provided in this invention. This method can accurately identify faults such as performance degradation of high-pressure heaters, valve jamming, and early leakage, enabling early warning and diagnosis of faults and improving the operational safety and economy of high-pressure heater systems in thermal power plants.

[0079] Example 2 The overall architecture of a smart monitoring system based on a DCS thermal power plant high-pressure heater system according to an embodiment of the present invention is as follows from bottom to top: Figure 2 As shown, it includes: a data acquisition and interface layer, a data processing and storage layer, an intelligent analysis engine layer, and a human-computer interaction and application layer. Each layer communicates and exchanges data with the others via an internal data bus.

[0080] Specifically, the data acquisition and interface layer DCS, as the foundation of the system, is responsible for integrating with the power plant's existing DCS system. Its function is to collect multi-source data related to the high-voltage heater system in real time, including but not limited to: Analog data includes: unit load, #6A / B high-pressure heater inlet / outlet feedwater temperature, #6A / B high-pressure heater condensate temperature, #6A / B high-pressure heater steam inlet temperature, #6A / B high-pressure heater extraction steam pressure, #6A / B high-pressure heater steam inlet pressure, #6A / B high-pressure heater water level, #7A / B high-pressure heater inlet / outlet feedwater temperature, #7A / B high-pressure heater condensate temperature, #7A / B high-pressure heater steam inlet temperature, #7A / B high-pressure heater extraction steam pressure, #7 / B high-pressure heater steam inlet pressure, #7A / B high-pressure heater water level, #8A / B high-pressure heater inlet / outlet feedwater temperature, #8A / B high-pressure heater condensate temperature, #8A / B high-pressure heater steam inlet temperature, #8A / B high-pressure heater extraction steam pressure, #8A / B high-pressure heater steam inlet pressure, and #8A / B high-pressure heater water level. Switching data: such as the open / closed status of the high-pressure heater inlet / outlet three-way valve, #6A / B inlet electric valve, #6A / B inlet check valve, #7A / B inlet electric valve, #7A / B inlet check valve, #8A / B inlet electric valve, and #8A / B inlet check valve. Equipment status data: such as the high-pressure heater's disconnection / commissioning status. The data processing and storage layer (DCS) acts as the system's data hub, receiving raw data from the data acquisition layer. Its function is to perform data cleaning, digital filtering, engineering unit conversion, and outlier removal on the raw data, and to perform time-series alignment on data from different sources to ensure the quality of the acquired data. The processed data is categorized and stored in a real-time database (for storing short-term high-frequency data, supporting real-time analysis) and a historical database (for storing long-term operational data, supporting model training and trend analysis). This layer provides a high-quality, structured data foundation for upper-layer intelligent analysis. The Intelligent Analysis Engine (DCS) layer is the core of this invention. Its function is to analyze and calculate the preprocessed data, realizing the transformation from "data" to "criteria".

[0081] Furthermore, the intelligent analysis engine comprises several functionally coordinated modules: Equipment Status Assessment Module: Its function is to establish a baseline performance model for the high-pressure heater system. Based on thermodynamic principles, this module dynamically calculates key performance indicators (KPIs) such as heater top-end temperature difference, heater bottom-end temperature difference, and overall high-pressure heater temperature rise, and compares the real-time calculated values ​​with historical health baseline values ​​under the same load and similar operating conditions. By quantifying the deviation, this module can score the overall health status of the high-pressure heaters on a percentage scale, intuitively reflecting the degree of performance degradation of each heater. Multi-Parameter Fusion Early Warning Module: This module is the main inventive point of this invention. Its function is to overcome the limitations of traditional single-point threshold alarms. Through built-in comprehensive fault diagnosis logic, it fuses and analyzes multiple related parameters, thereby achieving accurate early warning of early, slowly changing faults. This module specifically executes the following three core diagnostic logics: Diagnosis of performance degradation in high-performance equipment (with the following conditions): 1) The DCS unit load is ≥280MW; and the load change within 10 minutes is ≤35MW.

[0082] 2) Both DCSA and column B are in operation (AND): a) The inlet and outlet three-way valves of DCSA and B-series high-pressure heaters are both in the fully open position; b) The DCS corresponding high-pressure steam inlet electric valve is in the fully open position; c) The DCS corresponding high-pressure steam inlet check valve is in the fully open position.

[0083] 3) Compare the DCS data with historical values ​​under the same load. If any of the following conditions are met, an alarm will be triggered: a) Under the same load conditions of DCS, the current value of the heater top temperature difference minus the historical value of the heater top temperature difference is ≥3℃; b) Under the same load conditions of DCS, the current value of the heater lower end temperature difference minus the historical value of the heater lower end temperature difference is ≥3.5℃; c) Under the same load conditions of DCS, the historical value of the high-temperature heater rise of the entire train minus the current value of the high-temperature heater rise of the entire train is ≥15℃.

[0084] An alarm is triggered when this logical condition is met. The scientific principle behind this is that the deterioration of these performance indicators directly reflects fouling or internal damage to the heat transfer surface.

[0085] Diagnosis of valve jamming in high-pressure steam inlet pipeline (criteria for judgment): 1) The unit load is ≥280MW; and the load change within 10 minutes is ≤35MW.

[0086] 2) Both columns A and B are in operation (AND): a) The inlet and outlet three-way valves of the high-pressure heaters in columns A and B are both in the fully open position; b) The corresponding high-pressure steam inlet electric valve is in the fully open position; c) The corresponding high-pressure steam inlet check valve is in the fully open position.

[0087] 3) Under the same load, the historical value - the high-temperature rise ≥ 2℃; 4) The deviation between the high-pressure heater inlet pressure and the corresponding extraction pressure is ≥0.2MPa, or the deviation between the high-pressure heater inlet pressure of columns A and B is ≥0.1MPa.

[0088] 5) Under the same load, the absolute value of the difference between the historical value and the inlet water temperature of the high-pressure heater is <15℃.

[0089] If this logical condition is met, the overall judgment is that the steam inlet pipeline valve is stuck. This logic uses multi-parameter cross-validation to accurately locate the source of the fault.

[0090] High-pressure heater leakage risk diagnosis (criteria, and): 1) The unit load is ≥280MW; and the load change within 10 minutes is ≤35MW.

[0091] 2) Both columns A and B are in operation (AND): a) The inlet and outlet three-way valves of the high-pressure heaters in columns A and B are both in the fully open position; b) The corresponding high-pressure steam inlet electric valve is in the fully open position; c) The corresponding high-pressure steam inlet check valve is in the fully open position.

[0092] 3) Under the same load conditions, if the deviation of the #6 high-pressure heater inlet steam pressure from the historical value is <0.15MPa, and the deviation of the #7 and #8 high-pressure heater inlet steam pressure from the historical value is <0.2MPa, and the absolute value of the high-pressure heater water level deviation is ≤1mm, the opening of the high-pressure heater normal drain regulating valve increases by more than 10% compared with the historical value.

[0093] 4) Under the same load conditions, the temperature difference at the upper end of the heater minus the historical value is ≥3℃ and the temperature difference at the lower end of the heater minus the historical value is ≤-1℃.

[0094] When this logical condition is met, a comprehensive judgment is made that there is an early risk of leakage in the high-pressure heater. This logic can detect subtle signs of leakage before traditional water level alarms are triggered.

[0095] The fault diagnosis and reasoning module: Its function is to further characterize the fault after the early warning module triggers an alarm. This module has built-in diagnostic rules based on an expert knowledge base and fault tree, and can automatically infer the most likely cause of the fault based on the alarm type and related parameter combinations, and provide confidence levels to provide decision support for operators. The operation optimization suggestion module: Its function is to proactively improve economic efficiency. Based on the current unit load and high-pressure heater health status, this module provides operators with optimization suggestions for the high-pressure heater water level setpoint (including risk assessment after water level setpoint optimization) based on a preset optimization algorithm, aiming to reduce the high-pressure heater terminal difference and improve the system's thermal economy while ensuring safe operation. The human-machine interaction and application layer (DCS) serves as the interface between the system and users. Its function is to present the results of intelligent analysis to operators in an intuitive and effective way and support human-machine collaborative control. It includes, for example... Figure 3 As shown: Intelligent monitoring visualization interface: Displays key status, health scores, early warning information, diagnostic conclusions, and optimization suggestions of the high-voltage heating system in an integrated, graphical manner (e.g., flowcharts, trend charts, dashboards). Early warning and diagnostic report push: Proactively pushes tiered alarm information and detailed diagnostic reports via DCS operator station and mobile terminal. Human-machine collaborative control interface: After obtaining authorization from the operator, optimized setpoints can be automatically and securely transmitted back to the corresponding adjustment loop of the DCS through data acquisition and interface layer, achieving closed-loop optimization control.

[0096] This invention relates to a smart monitoring system for high-pressure heaters in DCS-based thermal power plants. During cost reduction and efficiency improvement periods, the system issues alerts indicating a decline in high-pressure heater performance and provides suggested solutions. Through timely intervention, it prevents further performance degradation and increased cooling loss. The implementation of this smart monitoring system reduces labor costs and equipment maintenance costs.

[0097] Example 3 To achieve the above embodiments, such as Figure 4 As shown, this embodiment also provides a smart monitoring device 10 based on the DCS thermal power plant high-pressure heater system, including: The multi-source data acquisition module 100 is used to acquire multi-source operating data of the high-voltage heater system, including analog data, switch data and equipment status data; The data cleaning and storage module 200 is used to clean, time-series align, and remove bad values ​​from the collected data, generate structured data, and store it in the real-time database and historical database respectively. The dynamic benchmark performance model building module 300 is used to build a dynamic benchmark performance model based on structured data and thermodynamic principles, and to calculate the deviation between the real-time values ​​of key performance indicators and historical health benchmark values. The multi-parameter cross-validation and early warning module 400 is used to determine whether the deviation meets the preset fault early warning conditions through multi-parameter cross-validation logic. If it does, an early warning is triggered and a fault diagnosis conclusion is generated.

[0098] Furthermore, the multi-source data acquisition module 100 is also used for: The collected analog data includes unit load, heater inlet and outlet feedwater temperature, condensate temperature, steam inlet temperature, extraction steam pressure, steam inlet pressure, and water level parameters. The collected switch data includes the open / closed status of the high-pressure heater inlet and outlet three-way valves, the steam inlet electric valve, and the steam inlet check valve.

[0099] Furthermore, the data cleaning and storage module 200 is also used for: A 3σ-based outlier removal method is used to filter out outlier data that deviates from the mean by more than three standard deviations. Interpolation algorithms are used to align time-series data, ensuring that data with different sampling frequencies are synchronized at the same timestamp.

[0100] Furthermore, the dynamic benchmark performance model building module 300 is also used for: A nonlinear relationship model between the upper and lower temperature differences of the heater and the load and extraction steam pressure was established using a multiple regression algorithm. The model's coefficient of determination R² is greater than 0.92. The deviation between the real-time value of the overall high-temperature heater rise and the historical health baseline value is calculated using the following formula:

[0101] in For real-time inlet water temperature, Based on historical baseline inlet water temperature, For real-time outlet water temperature, The historical baseline outlet water temperature.

[0102] An intelligent monitoring device based on the DCS high-pressure heater system of a thermal power plant, according to an embodiment of the present invention, can realize accurate early warning and diagnosis of early faults in the high-pressure heater system, significantly improve the timeliness and accuracy of identifying leakage, valve jamming and performance degradation, and improve the safety and economy of thermal power plant operation.

[0103] Example 4 The present invention also provides an electronic device such as Figure 5 As shown, it includes a processor and a memory. The memory stores executable instructions. When the processor executes the instructions, it implements the above-mentioned intelligent monitoring method based on the DCS thermal power plant high-pressure heater system.

[0104] Example 5 The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned intelligent monitoring method based on a DCS thermal power plant high-pressure heater system.

[0105] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," DCS "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0106] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A smart monitoring method based on a DCS superheater system of a thermal power plant, characterized in that, The method comprises the following steps: S1, collecting multi-source operation data of the high-pressure heater system, including analog quantity data, switch quantity data and equipment state data; S2, cleaning, time sequence alignment and bad value elimination are performed on the collected data, structured data is generated and stored into a real-time database and a historical database respectively; S3, a dynamic benchmark performance model is established based on the structured data and thermodynamic principles, and a deviation between a real-time value of a key performance index and a historical health benchmark value is calculated; S4, whether the deviation meets a preset fault early warning condition is judged through multi-parameter cross verification logic, and if yes, early warning is triggered and a fault diagnosis conclusion is generated.

2. The method of claim 1, wherein, The S1 comprises: S11, collecting analog quantity data including unit load, heater inlet and outlet feed water temperature, drain temperature, inlet steam temperature, extraction pressure, inlet steam pressure and water level parameters; S12, collecting switch quantity data including open / close state of high-pressure heater inlet and outlet three-way valve, inlet steam electric valve and inlet steam check valve.

3. The method of claim 1, wherein, The S2 comprises: S21, a bad value elimination method based on 3σ principle is adopted to filter abnormal data deviating from three times of standard deviation of mean value; S22, interpolation algorithm is used to align time sequence data, so that data with different sampling frequencies are synchronized at the same time stamp.

4. The method of claim 1, wherein, The S3 comprises: S31, a multivariate regression algorithm is used to establish a nonlinear relationship model of heater upper end difference, lower end difference, load and extraction pressure, and a model determination coefficient R² is greater than 0.92; S32, a deviation between a real-time value of whole high-pressure heater temperature rise and a historical health benchmark value is calculated, and the deviation formula is: wherein is a real-time inlet water temperature, is a historical baseline inlet water temperature, is a real-time outlet water temperature, is a historical baseline outlet water temperature.

5. A smart monitoring device based on a DCS superheating system of a thermal power plant, characterized in that, The method comprises the following steps: A multi-source data collection module is configured to collect multi-source operation data of the high-pressure heater system, including analog quantity data, switch quantity data and equipment state data; A data cleaning and storage module is configured to clean, time sequence align and eliminate bad values of the collected data, generate structured data and store the structured data into a real-time database and a historical database respectively; A dynamic benchmark performance model establishment module is configured to establish a dynamic benchmark performance model based on the structured data and thermodynamic principles, and calculate a deviation between a real-time value of a key performance index and a historical health benchmark value; A multi-parameter cross verification and early warning module is configured to judge whether the deviation meets a preset fault early warning condition through multi-parameter cross verification logic, and if yes, early warning is triggered and a fault diagnosis conclusion is generated.

6. The apparatus of claim 5, wherein, The multi-source data collection module is further configured to: Collect analog quantity data including unit load, heater inlet and outlet feed water temperature, drain temperature, inlet steam temperature, extraction pressure, inlet steam pressure and water level parameters; Collect switch quantity data including open / close state of high-pressure heater inlet and outlet three-way valve, inlet steam electric valve and inlet steam check valve.

7. The apparatus of claim 5, wherein, The data cleaning and storage module is further configured to: Adopt a bad value elimination method based on 3σ principle to filter abnormal data deviating from three times of standard deviation of mean value; Align time sequence data through interpolation algorithm to ensure that data with different sampling frequencies are synchronized at the same time stamp.

8. The apparatus of claim 5, wherein, The dynamic benchmark performance model establishment module is further configured to: Adopt a multivariate regression algorithm to establish a nonlinear relationship model of heater upper end difference, lower end difference, load and extraction pressure, and a model determination coefficient R² is greater than 0.92; Calculate a deviation between a real-time value of whole high-pressure heater temperature rise and a historical health benchmark value, and the deviation formula is: wherein is a real-time inlet water temperature, is a historical baseline inlet water temperature, is a real-time outlet water temperature, is a historical baseline outlet water temperature.

9. A computer device, comprising: comprising a processor and a memory; wherein the processor executes a program corresponding to executable program code stored in the memory by reading the executable program code, for realizing the method for wisdom monitoring and supervision of a high-pressure heater system of a DCS-based thermal power plant according to any one of claims 1-4.

10. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to realize the method for wisdom monitoring and supervision of a high-pressure heater system of a DCS-based thermal power plant according to any one of claims 1-4. The program is executed by the processor to realize the method for wisdom monitoring and supervision of a high-pressure heater system of a DCS-based thermal power plant according to any one of claims 1-4.

Images