Online oxide skin concentration monitoring system, method, equipment and medium for supercritical boiler

Abstract

The invention discloses an online oxide skin concentration monitoring system, method, equipment and medium for a supercritical boiler, and the system comprises a data collection module which is used for obtaining various data of boiler smoke discharge and transmitting the data to a data processing module; the data processing module is used for receiving and preprocessing the data acquired by the data acquisition module; the intelligent prediction module is used for constructing a multi-parameter coupling model through an intelligent algorithm, inputting the preprocessed data into the multi-parameter coupling model and predicting the oxide skin concentration; and the early warning module is used for visually displaying the prediction result and automatically triggering an alarm mechanism when the oxide skin concentration exceeds a preset threshold value. Through multi-source sensor data fusion and intelligent algorithm modeling, real-time online monitoring of the oxide skin concentration is realized, the hysteresis problem of traditional offline detection is solved, the false alarm rate of the system is reduced, the early warning reliability is greatly improved, and an intelligent solution is provided for safe and economical operation of the supercritical boiler.

Description

Technical Field

[0001] The present invention relates to the field of boiler detection, and in particular to a system, method, equipment and medium for monitoring the online oxide scale concentration of a supercritical boiler. Background Art

[0002] With the adjustment of my country's energy structure and the continuous improvement of environmental protection requirements, the proportion of supercritical and ultra-supercritical units in thermal power generation continues to rise. These units have high operating parameters (pressure ≥22.1MPa, temperature ≥566℃) and thermal efficiency can reach over 45%, but they also face more severe material aging problems. Among them, the formation and shedding of high-temperature oxide scale on the heating surface of the boiler has become the primary hidden danger affecting the safe operation of the unit. Statistics show that domestic supercritical units have an average of more than 200 unplanned shutdown accidents caused by oxide scale problems each year, with direct economic losses reaching tens of billions of yuan. Even more seriously, the detached oxide scale particles will flush downstream pipelines with the flow of steam, which may cause secondary disasters such as damage to turbine blades.

[0003] Traditional offline detection methods rely on periodic boiler shutdowns or pipe sampling, which not only results in long inspection cycles but also limits the representativeness of the samples, making it impossible to capture the transient process of scale shedding. Secondly, existing online monitoring solutions often use single-parameter indirect inference algorithms, such as simple models based on temperature difference or pressure drop methods. Their measurement errors generally exceed 30% and are susceptible to load fluctuations. In terms of intelligent early warning, most systems still use fixed threshold alarm mechanisms, lacking deep learning of historical boiler operating data, making it difficult to predict scale accumulation trends. Of particular note, when the boiler is operating at peak load, the existing system's prediction accuracy for scale behavior under variable operating conditions is less than 60%, severely restricting the unit's flexible peak load regulation capabilities. These technical shortcomings often force power plant operators and maintenance personnel to adopt conservative operating strategies, which not only increases coal consumption but also affects grid dispatch requirements. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an online oxide scale concentration monitoring system for supercritical boilers to solve the problem that the oxide scale concentration in supercritical boilers is difficult to monitor accurately in real time and lacks intelligent early warning.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an online scale concentration monitoring system for a supercritical boiler, comprising:

[0008] Data acquisition module, used to obtain various data of boiler exhaust and transmit them to the data processing module;

[0009] A data processing module is used to receive and pre-process the data collected by the data acquisition module;

[0010] Intelligent prediction module, which is used to build a multi-parameter coupling model through intelligent algorithms, input pre-processed data into the multi-parameter coupling model, and predict the oxide scale concentration;

[0011] The early warning module is used to visualize the prediction results and automatically trigger the alarm mechanism when the oxide scale concentration exceeds the preset threshold.

[0012] As a preferred solution of the supercritical boiler online oxide scale concentration monitoring system of the present invention, various data of boiler exhaust gas are obtained, including:

[0013] Use data collection equipment to collect boiler exhaust gas temperature, pressure, flow rate, carbon monoxide and oxygen data;

[0014] The acquisition equipment includes a temperature sensor, a pressure sensor, a flow rate sensor and a gas composition analyzer.

[0015] As a preferred solution of the supercritical boiler online oxide scale concentration monitoring system of the present invention, the data processing module is used to receive the data collected by the data acquisition module and perform preprocessing, including:

[0016] Perform data cleaning on various boiler exhaust data, remove outliers, and fill in missing data through linear interpolation;

[0017] After data cleaning, the Min-Max normalization method is used to normalize the various data of boiler exhaust to eliminate dimensional differences;

[0018] The dynamic change rate of exhaust temperature and pressure is calculated through a sliding time window to generate a time series feature vector.

[0019] The beneficial effects of this preferred technical solution are: through data cleaning and normalization processing, sensor noise and dimensional differences are effectively eliminated, and the availability of original data is improved; the generation of time series feature vectors breaks through the limitations of traditional static analysis, can accurately capture the dynamic operating status of the boiler, and provide high-precision input for subsequent predictions.

[0020] As a preferred solution of the supercritical boiler online oxide scale concentration monitoring system of the present invention, a multi-parameter coupling model is constructed by an intelligent algorithm, including:

[0021] A multi-parameter coupling model was constructed using the random forest algorithm and the long short-term memory network algorithm. The random forest algorithm was used to process static parameters, and the long short-term memory network algorithm was used to analyze time series data.

[0022] The contribution of different parameters is dynamically weighted through the attention mechanism to establish a nonlinear mapping relationship between exhaust temperature, CO concentration and oxide scale thickness.

[0023] Supervised learning is performed based on historical data of boiler exhaust gas, and a loss function is set to optimize the multi-parameter coupling model.

[0024] The beneficial effects of this preferred technical solution are: the multi-parameter coupling model adopts the random forest algorithm and the long short-term memory network algorithm, taking into account the analysis of static parameters and dynamic timing characteristics, and the prediction accuracy is improved compared with the single model; the introduction of the attention mechanism makes the weight distribution of key parameters more reasonable, enhancing the robustness of the model under variable working conditions.

[0025] As a preferred solution of the online scale concentration monitoring system for supercritical boilers of the present invention, the pre-processed data is input into a multi-parameter coupling model to predict the scale concentration, including:

[0026] The pre-processed data is input into the multi-parameter coupling model in the form of a time series matrix, and continuous prediction is performed in each cycle to output the instantaneous predicted value of the oxide scale concentration. The inference engine ensures that the prediction process is completed quickly.

[0027] A sliding window is used to smooth the instantaneous prediction value to eliminate short-term fluctuations and obtain a stable concentration change trend;

[0028] Calculate the hourly scale deposition rate, dynamically assess the scale accumulation rate, and automatically shorten the sliding window when changes in boiler load are detected;

[0029] The confidence interval of the prediction result is calculated through multiple random inferences. When the width of the confidence interval exceeds the confidence threshold range, the automatic review mechanism is triggered.

[0030] The beneficial effects of this preferred technical solution are: the combination of real-time reasoning and sliding window smoothing enables the prediction results to maintain a response speed of seconds while avoiding misjudgments caused by data jitter; the confidence interval trigger mechanism can reduce the system's false alarm rate and greatly improve the reliability of early warnings.

[0031] As a preferred solution of the supercritical boiler online oxide scale concentration monitoring system of the present invention, the prediction results are visualized, including:

[0032] The scale concentration curve and the linkage change trend of key parameters are displayed in real time through the visual interface;

[0033] Visualize the spatial distribution of scale deposition based on boiler pipe GIS coordinates;

[0034] Automatically export reports containing the number of concentration violations, peak statistics, and maintenance recommendations.

[0035] As a preferred solution of the online oxide scale concentration monitoring system for supercritical boilers of the present invention, when the oxide scale concentration exceeds a preset threshold, an alarm mechanism is automatically triggered, including:

[0036] When the oxide scale concentration exceeds the first threshold range, a text message is sent to notify relevant staff;

[0037] When the oxide scale concentration exceeds the second threshold range, the DCS system is linked to reduce the boiler load;

[0038] Automatically associate historical alarm records, mark high-frequency exceeding-standard areas as priority inspection points, and call pre-stored cleaning solutions to push them to the operation terminal.

[0039] In a second aspect, the present invention provides a method for monitoring the online scale concentration of a supercritical boiler, comprising: obtaining various data of boiler exhaust gas;

[0040] Preprocess the collected data;

[0041] A multi-parameter coupling model is constructed through an intelligent algorithm, and the pre-processed data is input into the multi-parameter coupling model to predict the oxide scale concentration;

[0042] The prediction results are displayed visually, and when the oxide scale concentration exceeds the preset threshold, the alarm mechanism is automatically triggered.

[0043] In a third aspect, the present invention provides an electronic device, comprising:

[0044] memory and processor;

[0045] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the online oxide scale concentration monitoring system for supercritical boilers are implemented.

[0046] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the supercritical boiler online oxide scale concentration monitoring system.

[0047] Compared with the existing technology, the present invention has the following beneficial effects: the present invention realizes real-time online monitoring of oxide scale concentration through multi-source sensor data fusion and intelligent algorithm modeling, and solves the problem of lag in traditional offline detection; adopts a multi-parameter coupling model that mixes random forest and long short-term memory network algorithms, and combines it with the attention mechanism to improve the prediction accuracy compared with the single parameter method, effectively overcoming the defects of high false alarm rate and poor adaptability of the existing technology; through confidence interval evaluation and dynamic review mechanism, the system false alarm rate is reduced, the early warning reliability is greatly improved, and an intelligent solution is provided for the safe and economical operation of supercritical boilers. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 This is a schematic diagram of the framework of each module of an online oxide scale concentration monitoring system for a supercritical boiler according to an embodiment of the present invention.

[0050] Figure 2 This is a schematic diagram of the change trend of oxide scale in condensate during the unit startup purge process of an online oxide scale concentration monitoring system for a supercritical boiler according to an embodiment of the present invention.

[0051] Explanation of reference numerals: 100, data acquisition module; 200, data processing module; 300, intelligent prediction module; 400, early warning module. DETAILED DESCRIPTION

[0052] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0053] Example 1, reference Figure 1 , which is an embodiment of the present invention, provides an online scale concentration monitoring system for a supercritical boiler, comprising: a data acquisition module 100 for acquiring various data of boiler exhaust and transmitting the data to a data processing module 200;

[0054] The data processing module 200 is used to receive and pre-process the data collected by the data collection module 100;

[0055] An intelligent prediction module 300 is used to construct a multi-parameter coupling model using an intelligent algorithm, input pre-processed data into the multi-parameter coupling model, and predict the oxide scale concentration;

[0056] The early warning module 400 is used to visualize the prediction results and automatically trigger an alarm mechanism when the oxide scale concentration exceeds a preset threshold.

[0057] It should be noted that the present invention realizes intelligent monitoring of the oxide scale concentration of supercritical boilers by constructing a full-process monitoring system for data acquisition, processing, prediction and early warning. The data acquisition module uses multiple sensors to work together to ensure the acquisition of comprehensive parameters such as boiler exhaust temperature, pressure, flow rate and gas composition, providing a reliable data basis for subsequent analysis. The data processing module effectively eliminates noise interference and dimensional differences through data cleaning and normalization processing, thereby improving data quality; the intelligent prediction module innovatively adopts a hybrid model of random forest and long short-term memory network algorithms, combined with the attention mechanism to dynamically weight key parameters, which can improve prediction accuracy and adaptability to working conditions; the early warning module realizes real-time monitoring and timely early warning of oxide scale risks through visual display and hierarchical alarm mechanism. This systematic design not only solves the problems of monitoring lag and insufficient accuracy of traditional methods, but also realizes accurate prediction of oxide scale deposition status through intelligent algorithms, providing reliable protection for the safe operation of supercritical boilers.

[0058] It should also be noted that the method of the present invention can not only realize the intelligent monitoring of the oxide scale concentration of supercritical boilers, but also realize the intelligent monitoring of the oxide scale concentration of ultra(super)critical boilers.

[0059] Example 2, reference Figure 1-Figure 2 , which is an embodiment of the present invention, provides an online oxide scale concentration monitoring system for a supercritical boiler based on the above embodiment.

[0060] In the embodiment of the present invention, the data acquisition module 100 is used to obtain various data of boiler exhaust and transmit them to the data processing module 200, which specifically includes:

[0061] Use data collection equipment to collect boiler exhaust gas temperature, pressure, flow rate, carbon monoxide and oxygen data;

[0062] The acquisition equipment includes temperature sensors, pressure sensors, flow rate sensors and gas composition analyzers.

[0063] It should be noted that by real-time monitoring of multiple key parameters during boiler operation, accurate and real-time monitoring of scale concentration can be achieved, thereby improving boiler operation efficiency.

[0064] In an optional embodiment, the various boiler data obtained may also include parameters such as flue gas humidity, smoke dust concentration, and NOx content;

[0065] In an optional embodiment, the acquisition device may also be an infrared thermal imager, a laser spectrum analyzer, an ultrasonic flow meter, etc.

[0066] In the embodiment of the present invention, the data processing module 200 is used to receive the data collected by the data collection module 100 and perform pre-processing, specifically including:

[0067] Perform data cleaning on various boiler exhaust data, remove outliers, and fill in missing data through linear interpolation;

[0068] After data cleaning, the Min-Max normalization method is used to normalize the various data of boiler exhaust to eliminate dimensional differences;

[0069] The dynamic change rate of exhaust temperature and pressure is calculated through a sliding time window to generate a time series feature vector.

[0070] It should be noted that through data cleaning and normalization, sensor noise and dimensional differences can be effectively eliminated, and the availability of original data can be improved; the generation of time series feature vectors breaks through the limitations of traditional static analysis, can accurately capture the dynamic operating status of the boiler, and provide high-precision input for subsequent predictions.

[0071] In an optional embodiment, various data of boiler exhaust gas may be preprocessed using the 3σ criterion or box plot method, spline interpolation or KNN interpolation method, Z-score standardization or decimal scaling standardization, and wavelet transform or Fourier transform;

[0072] For example, the 3σ criterion is used to calculate the range of the mean ± 3 times the standard deviation to identify outliers; the box plot method determines the outlier threshold based on the quartiles and 1.5 times the interquartile range. Missing values ​​for continuously varying parameters such as temperature and pressure are imputed using spline interpolation. Z-score normalization is used to transform the data by (x - μ) / σ to address outliers. Wavelet transforms can be used to extract local time-frequency features of the data and perform non-stationary time series data analysis.

[0073] In an optional embodiment, the preprocessing methods can be flexibly combined and applied according to the actual data characteristics;

[0074] For example, temperature data is processed using the 3σ criterion for outlier removal, spline interpolation for gap filling, and wavelet transform for feature extraction. Gas concentration data is processed using boxplots for outlier removal, KNN interpolation for gap filling, and Z-score normalization. This targeted preprocessing strategy can improve data quality by over 30%, providing more reliable input for subsequent modeling.

[0075] In an embodiment of the present invention, a multi-parameter coupling model is constructed by an intelligent algorithm, specifically including:

[0076] A multi-parameter coupling model was constructed using the random forest algorithm and the long short-term memory network algorithm. The random forest algorithm was used to process static parameters, and the long short-term memory network algorithm was used to analyze time series data.

[0077] The contribution of different parameters is dynamically weighted through the attention mechanism to establish a nonlinear mapping relationship between exhaust temperature, CO concentration and oxide scale thickness.

[0078] Supervised learning is performed based on historical data of boiler exhaust, and the loss function is set to optimize the multi-parameter coupling model.

[0079] Specifically, the random forest algorithm sets 100 decision trees, the maximum depth of each tree is 10, and the Gini coefficient is used as the splitting criterion; the long short-term memory network adopts a two-layer structure with 128 hidden units in each layer, and the dropout rate is set to 0.2 to prevent overfitting.

[0080] The attention mechanism uses scaled dot-product attention, dynamically adjusting the weights of various parameters by calculating the similarity score of query-key-value pairs. The supervised learning phase uses the Adam optimizer, with an initial learning rate set to 0.001 and a cosine annealing strategy for learning rate adjustment. The Huber loss function combines the advantages of mean squared error and absolute error, making it more robust to outliers. Early stopping is used during model training, and the model parameters with the best performance on the validation set are retained.

[0081] In an optional implementation, a multi-parameter coupled model can be constructed using a gradient boosted decision tree and a gated recurrent unit. Specifically, the gradient boosted decision tree is configured with 150 decision trees, a learning rate of 0.05, and a maximum depth of 8. The gated recurrent unit adopts a single-layer structure with 256 hidden units and a dropout rate of 0.3.

[0082] In another optional implementation, support vector regression and temporal convolutional networks can be used to construct a multi-parameter coupled model. Specifically, support vector regression uses an RBF kernel function, with the regularization parameter C set to 1.0 and the kernel coefficient gamma set to 0.1; the temporal convolutional network is constructed with four residual blocks, each containing 64 filters, a convolution kernel size of 3, and a dilation factor of 2.

[0083] In an embodiment of the present invention, the preprocessed data is input into a multi-parameter coupling model to predict the oxide scale concentration, specifically including:

[0084] The pre-processed data is input into the multi-parameter coupling model in the form of a time series matrix, and continuous prediction is performed in each cycle to output the instantaneous predicted value of the oxide scale concentration. The inference engine ensures that the prediction process is completed quickly.

[0085] A sliding window is used to smooth the instantaneous prediction value to eliminate short-term fluctuations and obtain a stable concentration change trend;

[0086] Calculate the hourly scale deposition rate, dynamically assess the scale accumulation rate, and automatically shorten the sliding window when changes in boiler load are detected;

[0087] The confidence interval of the prediction result is calculated through multiple random inferences. When the width of the confidence interval exceeds the confidence threshold range, the automatic review mechanism is triggered.

[0088] For example, the pre-processed data is input into the model in real time in the form of a time series matrix. The system automatically performs a prediction every 30 seconds and uses a high-performance inference engine to ensure that the prediction response time is controlled within 50 milliseconds. To eliminate the interference of data fluctuations, the system uses sliding window technology to perform exponential weighted smoothing on the prediction results. The standard window length is 1 hour and the smoothing coefficient is set to 0.3. When a change of more than 10% in the boiler load is detected, the system intelligently shortens the window to 30 minutes to improve the response sensitivity. At the same time, the system performs 100 random inferences using the Monte Carlo Dropout method to establish a 95% confidence interval. When the interval width exceeds 10% of the predicted value, it automatically starts a three-level review process including sensor calibration and backup model verification to ensure the reliability of the early warning.

[0089] It should be noted that the present invention uses a time series matrix as input and continuously predicts each cycle. In combination with an inference engine, it can quickly obtain the instantaneous value of the oxide scale concentration, making monitoring almost real-time; the sliding window smoothing process effectively filters short-term fluctuations, clearly presents the stable change trend of the concentration, and facilitates the grasp of the overall trend; calculates the hourly deposition rate, automatically adjusts the window in combination with the boiler load change, accurately and dynamically evaluates the accumulation speed, and can respond in time when the load changes; multiple random inferences are used to calculate the confidence interval, and the over-threshold triggers automatic review, which greatly improves the prediction reliability. The coordination of multiple links makes the oxide scale concentration prediction accurate, fast and stable, providing strong data support for the safe operation of the boiler.

[0090] In an alternative approach, using a gradient boosting decision tree and gated recurrent units to build a multi-parameter coupled model, the scale concentration prediction process is as follows: the prediction frequency is increased to 15 seconds, and a dynamic sliding window technique is used with a 30-minute window base length, which automatically adjusts based on the load change rate. Confidence assessment is performed using bootstrap resampling, with confidence intervals calculated using 50 sampling cycles with replacement.

[0091] In another optional fact method, if support vector regression and time convolutional network are used to construct a multi-parameter coupling model, the specific process of predicting scale concentration is as follows: a sliding time window of 2 hours is used, the results are smoothed by Kalman filtering, and a Bayesian optimization algorithm is introduced to automatically adjust the window length to 10-60 minutes, with a confidence threshold of 5%-15%.

[0092] In the embodiment of the present invention, the early warning module 400 is used to visualize the prediction results, specifically including:

[0093] The scale concentration curve and the linkage change trend of key parameters are displayed in real time through the visual interface;

[0094] Visualize the spatial distribution of scale deposition based on boiler pipe GIS coordinates;

[0095] Automatically export reports containing the number of concentration violations, peak statistics, and maintenance recommendations.

[0096] For example, the early warning module utilizes an advanced multi-level interactive visualization system, enabling comprehensive monitoring and early warning capabilities through a three-screen intelligent display solution. The left side of the system's main interface displays a real-time dynamic curve of scale concentration changes, using intuitive red, yellow, and green safety zone indicators. Trend curves for key parameters such as exhaust temperature and CO concentration are intelligently overlaid, allowing operators to conduct in-depth analysis using timeline zooming and parameter selection.

[0097] The middle area is based on a high-precision GIS coordinate system, and the positioning error is controlled within the range of ±5cm to construct a three-dimensional model of the boiler pipe. The gradient color thermal map from blue to red intuitively presents the distribution of oxide scale deposition thickness in the range of 0-2mm in different parts. It supports 360-degree rotation and arbitrary cross-section viewing, helping operation and maintenance personnel quickly locate problem areas.

[0098] The right-hand area automatically generates a comprehensive monitoring report containing statistics on concentration violations, peak data records, and intelligent maintenance recommendations. For example, based on historical data analysis, the system recommends purging the No. 3 high-temperature reheater within 72 hours, among other specific maintenance recommendations. This integrated visualization design not only enables a multi-dimensional display of monitoring data but also significantly improves the efficiency and accuracy of operational and maintenance decisions through intelligent analysis.

[0099] In an embodiment of the present invention, the early warning module 400 automatically triggers an alarm mechanism when the oxide scale concentration exceeds a preset threshold, specifically including:

[0100] When the scale concentration exceeds the preset threshold, the alarm mechanism is automatically triggered, including:

[0101] When the oxide scale concentration exceeds the first threshold range, a text message is sent to notify relevant staff;

[0102] When the oxide scale concentration exceeds the second threshold range, the DCS system is linked to reduce the boiler load;

[0103] Automatically associate historical alarm records, mark high-frequency exceeding-standard areas as priority inspection points, and call pre-stored cleaning solutions to push them to the operation terminal.

[0104] In an optional embodiment, the first threshold range is 80%-100% of the design value. When the concentration exceeds this range, the system will immediately send an early warning text message to five preset contacts, including the on-duty engineer and the equipment supervisor. The text message contains the exceeded value, specific location, and trend chart.

[0105] In an optional embodiment, the second threshold range is 100%-120% exceeding the design value. At this time, the system will automatically link with the power plant DCS control system to reduce the boiler load by 10%-30% according to the preset program, and at the same time trigger an audible and visual alarm and display a red warning on the large screen in the main control room.

[0106] In another optional embodiment, the first threshold range can be set to 70%-90% of the design value, and the second threshold range can be set to 90%-110%. The specific values ​​can be flexibly adjusted through the management interface according to parameters such as boiler model and operating years.

[0107] like Figure 2 As shown, from Figure 2 As can be seen in the figure, the monitoring system of the present invention can accurately capture the dynamic changes in scale concentration. The curve shows the fluctuations of the detection value at different time periods, such as the obvious peak of the detection value around 7:12:00, which shows that the system can keenly identify sudden concentration changes and promptly report abnormalities.

[0108] The relatively stable values ​​during other periods also demonstrate that the system is capable of stable monitoring under normal conditions. Through such continuous and detailed monitoring, operations and maintenance personnel can clearly understand the changing trends in scale concentration, facilitate early risk prediction and response measures, effectively ensure the safe operation of supercritical (ultracritical) boilers, improve the accuracy and timeliness of scale concentration monitoring, and help boilers operate more reliably and efficiently.

[0109] Example 3. The above is a schematic diagram of a system for monitoring the online oxide scale concentration of a supercritical boiler. It should be noted that the technical solution of this method for monitoring the online oxide scale concentration of a supercritical boiler is based on the same concept as the technical solution of the system for monitoring the online oxide scale concentration of a supercritical boiler described above. For details not described in detail in the technical solution of the method for monitoring the online oxide scale concentration of a supercritical boiler in this example, please refer to the description of the technical solution of the system for monitoring the online oxide scale concentration of a supercritical boiler described above.

[0110] This embodiment also provides a method for monitoring the online scale concentration of a supercritical boiler, comprising:

[0111] Obtain various data of boiler exhaust;

[0112] Preprocess the collected data;

[0113] A multi-parameter coupling model is constructed through an intelligent algorithm, and the pre-processed data is input into the multi-parameter coupling model to predict the oxide scale concentration;

[0114] The prediction results are displayed visually, and when the oxide scale concentration exceeds the preset threshold, the alarm mechanism is automatically triggered.

[0115] This embodiment also provides an electronic device suitable for online scale concentration monitoring of supercritical boilers, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement an online scale concentration monitoring system for supercritical boilers as proposed in the above embodiment.

[0116] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the system for online scale concentration monitoring of a supercritical boiler as proposed in the above embodiment is implemented.

[0117] The storage medium proposed in this embodiment and the online scale concentration monitoring system for supercritical boilers proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0118] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general hardware, and of course can also be implemented by hardware. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0119] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A supercritical boiler online scale concentration monitoring system, characterized in that: include: A data acquisition module (100) is used to obtain various data of boiler exhaust and transmit them to a data processing module (200); A data processing module (200) is used to receive the data collected by the data collection module (100) and perform pre-processing; An intelligent prediction module (300) is used to construct a multi-parameter coupling model through an intelligent algorithm, input the pre-processed data into the multi-parameter coupling model, and predict the oxide scale concentration; The early warning module (400) is used to visualize the prediction results and automatically trigger an alarm mechanism when the oxide scale concentration exceeds a preset threshold.

2. The supercritical boiler online scale concentration monitoring system according to claim 1, characterized in that: Obtain various data of boiler exhaust, including: Use data collection equipment to collect boiler exhaust gas temperature, pressure, flow rate, carbon monoxide and oxygen data; The acquisition equipment includes a temperature sensor, a pressure sensor, a flow rate sensor and a gas composition analyzer.

3. The supercritical boiler online scale concentration monitoring system according to claim 2, characterized in that: The data processing module (200) is used to receive the data collected by the data collection module (100) and perform pre-processing, including: Perform data cleaning on various boiler exhaust data, remove outliers, and fill in missing data through linear interpolation; After data cleaning, the Min-Max normalization method is used to normalize the various data of boiler exhaust to eliminate dimensional differences; The dynamic change rate of exhaust temperature and pressure is calculated through a sliding time window to generate a time series feature vector.

4. The supercritical boiler online scale concentration monitoring system according to claim 3, characterized in that: Build multi-parameter coupling models through intelligent algorithms, including: A multi-parameter coupling model was constructed using the random forest algorithm and the long short-term memory network algorithm. The random forest algorithm was used to process static parameters, and the long short-term memory network algorithm was used to analyze time series data. The contribution of different parameters is dynamically weighted through the attention mechanism to establish a nonlinear mapping relationship between exhaust temperature, CO concentration and oxide scale thickness. Supervised learning is performed based on historical data of boiler exhaust gas, and a loss function is set to optimize the multi-parameter coupling model.

5. The supercritical boiler online scale concentration monitoring system according to claim 4, characterized in that: The pre-processed data is fed into a multi-parameter coupled model to predict the scale concentration, including: The pre-processed data is input into the multi-parameter coupling model in the form of a time series matrix, and continuous prediction is performed in each cycle to output the instantaneous predicted value of the oxide scale concentration. The inference engine ensures that the prediction process is completed quickly. A sliding window is used to smooth the instantaneous prediction value to eliminate short-term fluctuations and obtain a stable concentration change trend; Calculate the hourly scale deposition rate, dynamically assess the scale accumulation rate, and automatically shorten the sliding window when changes in boiler load are detected; The confidence interval of the prediction result is calculated through multiple random inferences. When the width of the confidence interval exceeds the confidence threshold range, the automatic review mechanism is triggered.

6. The supercritical boiler online scale concentration monitoring system according to claim 5, characterized in that: The prediction results are visualized, including: The scale concentration curve and the linkage change trend of key parameters are displayed in real time through the visual interface; Visualize the spatial distribution of scale deposition based on boiler pipe GIS coordinates; Automatically export reports containing the number of concentration violations, peak statistics, and maintenance recommendations.

7. The supercritical boiler online scale concentration monitoring system according to claim 1, characterized in that: When the scale concentration exceeds the preset threshold, the alarm mechanism is automatically triggered, including: When the oxide scale concentration exceeds the first threshold range, a text message is sent to notify relevant staff; When the oxide scale concentration exceeds the second threshold range, the DCS system is linked to reduce the boiler load; Automatically associate historical alarm records, mark high-frequency exceeding-standard areas as priority inspection points, and call pre-stored cleaning solutions to push them to the operation terminal.

8. A method for monitoring the online oxide scale concentration of a supercritical boiler, using the online oxide scale concentration monitoring system for a supercritical boiler according to any one of claims 1 to 7, characterized in that: include: Obtain various data of boiler exhaust; Preprocess the collected data; A multi-parameter coupling model is constructed through an intelligent algorithm, and the pre-processed data is input into the multi-parameter coupling model to predict the oxide scale concentration; The prediction results are displayed visually, and when the oxide scale concentration exceeds the preset threshold, the alarm mechanism is automatically triggered.

9. An electronic device comprising: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the supercritical boiler online oxide scale concentration monitoring system according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the steps of the supercritical boiler online oxide scale concentration monitoring system according to any one of claims 1 to 7.

Images