Cleaning factor evaluation method, system, device and medium based on monitoring data

Abstract

The present invention relates to the technical field of boiler operation and maintenance, and in particular to a cleaning factor evaluation method, system, device and medium based on monitoring data; deploying monitoring equipment to obtain monitoring data of a coal-fired boiler, and performing data cleaning on the acquired data; establishing a parameter adjustment model based on historical boiler operation data and monitoring data, and optimizing boiler operation parameters; calculating the heat transfer coefficient of a heating surface, evaluating the operating state of the boiler, and using a neural network model to construct an ideal heat transfer coefficient prediction model; establishing a boiler heating surface cleaning factor calculation model, and dynamically adjusting boiler maintenance and cleaning strategies; the present invention solves the problem of untimely or excessive maintenance in traditional maintenance strategies, reduces maintenance costs and energy consumption, improves the heat transfer efficiency and operation stability of the boiler, and extends the service life of boiler equipment; the present invention significantly improves the overall operation efficiency and safety of the boiler system, and realizes intelligent boiler maintenance and management.

Description

Technical Field

[0001] The present invention relates to the technical field of boiler operation and maintenance, and in particular to a cleaning factor evaluation method, system, device and medium based on monitoring data. Background Art

[0002] Boilers play a vital role in industrial production and energy conversion, and their operating efficiency and safety directly affect production efficiency and energy utilization. With the continuous advancement of industrialization, traditional coal-fired boiler technology is also constantly evolving and improving. Early boilers mainly relied on manual operation and experience judgment, and had problems such as low efficiency, high energy consumption, and serious pollution. In order to improve the operating efficiency of boilers and reduce environmental pollution, modern boilers have gradually introduced automatic control technology and advanced monitoring equipment to realize the automation and intelligence of boiler operation. The advancement of computer technology and sensor technology has made it possible to monitor the operating status of boilers in real time and collect data, providing strong technical support for the optimized operation and maintenance of boilers.

[0003] Although modern boiler technology has made significant progress, there are still some problems in actual operation. For example, dust and coking problems on the heating surface will lead to a decrease in heat transfer efficiency, increase energy consumption, and affect the stable operation of the boiler. Traditional maintenance and cleaning strategies are often based on fixed time intervals or empirical judgments, which are difficult to respond to changes in actual operation in a timely manner, resulting in problems such as untimely maintenance or excessive maintenance. The existing boiler monitoring system has certain limitations in data processing and analysis, and it is difficult to make full use of historical operating data for optimization and adjustment. In response to these problems, the present invention proposes a cleaning factor evaluation method based on monitoring data to achieve dynamic evaluation of the boiler operating status and intelligent adjustment of the maintenance strategy, ensuring efficient and safe operation of the boiler under different load conditions. Summary of the invention

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

[0005] Therefore, the present invention provides a cleaning factor evaluation method based on monitoring data, which can solve the problem that the maintenance and cleaning strategies of traditional technical solutions are based on fixed time intervals or empirical judgments, and it is difficult to respond to changes in actual operation in a timely manner, resulting in untimely maintenance or excessive maintenance. The existing boiler monitoring system has certain limitations in data processing and analysis, and it is difficult to make full use of historical operation data for optimization and adjustment.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a cleaning factor evaluation method based on monitoring data, comprising: obtaining primary data; using the primary data to tune parameters to obtain secondary data; using the secondary data to perform calculation and evaluation; establishing a model, real-time monitoring, and dynamic adjustment.

[0007] As a preferred solution of the cleaning factor evaluation method based on monitoring data described in the present invention, the first-level data obtained includes boiler operating parameters, boiler historical data and heating surface monitoring data.

[0008] As a preferred solution of the cleaning factor evaluation method based on monitoring data described in the present invention, wherein: the parameter tuning includes temperature tuning, pressure tuning and emergency response;

[0009] The temperature optimization, if the deviation exceeds the threshold, determines the boiler heat transfer efficiency, if it decreases, adjusts the boiler operating parameters and the working fluid mass flow rate;

[0010] The pressure adjustment, if the deviation exceeds the threshold value, is caused by the decrease of fluid resistance and heat exchange efficiency, then the boiler operation compliance and main feed water flow are adjusted;

[0011] The emergency response, if both the temperature deviation and the pressure deviation exceed the threshold, and there is internal blockage and working fluid leakage, initiate the emergency response procedure, reduce combustion adjustments and check the boiler system.

[0012] As a preferred scheme of the cleaning factor evaluation method based on monitoring data described in the present invention, the secondary data obtained includes the heat transfer coefficient of the boiler heating surface, the cleaning factor, the boiler operating status, the maximum value of the interval grid heat transfer coefficient, the predicted value of the ideal heat transfer coefficient and the actual heat transfer coefficient.

[0013] As a preferred embodiment of the cleaning factor evaluation method based on monitoring data described in the present invention, the calculation evaluation includes comparing the actual heat transfer coefficient with the maximum value of the interval grid, judging the boiler operation status according to the difference, and dividing it into low-risk state A1, medium-risk state A2 and high-risk state A3, and then evaluating the boiler risk level according to the cleaning factor states B1, B2 and B3.

[0014] As a preferred embodiment of the cleaning factor evaluation method based on monitoring data described in the present invention, the dynamic adjustment includes judging according to the operating status. If the status satisfies A1 and B1, it is judged to be a normal operating status; if the status satisfies A1, B2 and A2, B1, it is judged to be an optimized adjustment status; if the status satisfies A3 and the status satisfies B3, it is judged to be an emergency response status.

[0015] As a preferred solution of the cleaning factor evaluation method based on monitoring data described in the present invention, the operation status judgment includes adjustment under normal operation status, adjustment under optimization adjustment status and adjustment under emergency response status;

[0016] Adjust under the normal operating state to maintain the current operating state and ensure the high heat transfer efficiency of the boiler without adjustment; if abnormal parameter fluctuations are found, re-evaluate the system state;

[0017] When the conductivity decreases, the detection frequency is increased during the optimal adjustment state;

[0018] Adjustments are made under the emergency response state. If the system is still in the emergency response state and new abnormal conditions occur, the boiler is stopped and inspected for thorough repairs.

[0019] As a preferred solution of the cleaning factor evaluation system based on monitoring data described in the present invention, it includes: a data preprocessing module, a parameter adjustment module, a state evaluation module and a strategy adjustment module;

[0020] The data preprocessing module deploys monitoring equipment to obtain monitoring data of the coal-fired boiler, cleans the acquired data, removes incomplete, missing and unstable data, and ensures data quality;

[0021] The parameter adjustment module establishes a temperature prediction model and a pressure prediction model based on the cleaned data, predicts the outlet working medium temperature and pressure of the boiler, calculates the temperature and pressure deviations by comparing with the actual monitoring data, and adjusts the boiler operating parameters and the main feed water flow rate;

[0022] The state evaluation module combines the optimized boiler operating parameters and the monitoring data of the boiler heating surface to calculate the heat transfer coefficient of the heating surface, evaluate the operating state of the boiler, and use the neural network model to build an ideal heat transfer coefficient prediction model;

[0023] The strategy adjustment module establishes a cleaning factor calculation model, accesses monitoring data and operating parameters in real time, divides the operating status based on the output of the cleaning factor calculation model, and dynamically adjusts the maintenance and cleaning strategy of the boiler to ensure efficient and safe operation of the boiler system.

[0024] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and wherein when the processor executes the computer program, the steps of any one of the methods for evaluating a cleaning factor based on monitoring data are implemented.

[0025] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of any one of the methods for evaluating a cleaning factor based on monitoring data are implemented.

[0026] The beneficial effects of the present invention are as follows: through real-time monitoring and data cleaning technology, the integrity and reliability of the data are ensured, providing a basis for subsequent analysis and optimization. A temperature prediction model and a pressure prediction model are established to accurately predict the outlet working fluid temperature and pressure of the boiler, and compare them with the actual monitoring data to adjust the operating parameters. By calculating the actual heat transfer coefficient and the ideal heat transfer coefficient, the cleaning factor of the boiler is evaluated, the operating status of the boiler is dynamically judged, and corresponding maintenance and cleaning strategies are adopted. The ideal heat transfer coefficient is predicted using a neural network model, which improves the accuracy and reliability of the prediction and optimizes the operation and maintenance management of the boiler. The present invention effectively solves the problem of untimely or excessive maintenance in traditional maintenance strategies, reduces maintenance costs and energy consumption, improves the heat transfer efficiency and operating stability of the boiler, extends the service life of boiler equipment, improves the overall operating efficiency and safety of the boiler system, and realizes intelligent boiler maintenance and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0028] Figure 1 A schematic flow chart of a cleaning factor evaluation method based on monitoring data provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0030] Example 1

[0031] Reference Figure 1 , which is the first embodiment of the present invention, provides a cleaning factor evaluation method based on monitoring data, comprising:

[0032] S1: Get primary data.

[0033] Deploy monitoring equipment to obtain monitoring data of coal-fired boilers and clean the acquired data. The monitoring data includes boiler operating parameters, boiler historical operating data and boiler heating surface monitoring data.

[0034] It should be noted that boiler operating parameters include unit load, main feed water flow;

[0035] The monitoring data of the boiler heating surface include the inlet flue gas temperature, outlet flue gas temperature, inlet working fluid temperature, outlet working fluid temperature, inlet working fluid pressure, outlet working fluid pressure, working fluid mass flow rate and heat exchange area of ​​the heating surface;

[0036] The boiler heating surface includes economizer, low-temperature superheater, low-temperature reheater and final superheater. The specific combination is determined by the boiler structure.

[0037] Data cleaning includes missing value cleaning, threshold cleaning and change rate cleaning. Missing value cleaning includes eliminating data at incomplete and missing monitoring times, threshold cleaning includes eliminating data when the boiler is not started, and change rate cleaning includes eliminating data under unstable conditions during the boiler load adjustment process.

[0038] Furthermore, data cleaning is an important step to ensure data integrity and reliability. During boiler operation, monitoring equipment may cause missing or incomplete data due to various reasons. For example, sensor failure, data transmission interruption or storage device problems can lead to incomplete data. Monitoring data may be affected by various noises or abnormal conditions, resulting in abnormal values. For example, equipment failure, sudden changes in operating conditions or human errors can lead to data anomalies. Boiler operation data involves multiple parameters and data from different time periods, which need to be consistent. Data cleaning can eliminate non-stable operating condition data during boiler load adjustment changes to ensure data consistency in time and space. After data cleaning, the remaining data is more accurate and reliable, which helps to improve the accuracy of the prediction model and parameter adjustment model. When establishing the temperature prediction model and pressure prediction model below, the cleaned data can more accurately reflect the actual operating status of the boiler, thereby improving the prediction accuracy and robustness of the model.

[0039] S2: Use the primary data to tune parameters and obtain the secondary data.

[0040] Based on the historical operation data and monitoring data of the boiler, a parameter adjustment model is established to optimize the boiler operation parameters;

[0041] It should be noted that the establishment of the parameter adjustment model includes a temperature prediction model and a pressure prediction model;

[0042] The temperature prediction model includes the energy balance principle based on the heat transfer process, considering that the outlet working fluid temperature prediction and heat flow calculation are coupled and cannot be solved directly, selecting key features and performing feature engineering, selecting time series analysis and LSTM deep learning algorithms, using training data for model training, using test data to evaluate model performance, performing cross-validation to ensure model stability, inputting validation data into the trained model, predicting the outlet working fluid temperature in real time, and updating the model regularly to maintain accuracy. The specific process of establishing the temperature prediction model is as follows:

[0043] The features most relevant to the prediction target are selected from the original data. When predicting the boiler outlet working fluid temperature, the Pearson correlation coefficient statistical method is used to screen the features. The key features include the inlet working fluid temperature, working fluid mass flow rate and heat flow.

[0044] The data set is divided into training set, validation set and test set with a ratio of 6:2:2. The data set includes all key features and target variables. The target variables include the inlet flue gas temperature, outlet flue gas temperature, outlet working fluid temperature, inlet working fluid pressure, outlet working fluid pressure and heat exchange area of ​​the heating surface.

[0045] The temperature prediction model formula is expressed as:

[0046] ,

[0047] in, is the forget gate activation vector, is the input gate activation vector, is the output gate activation vector, is the cell state, is the hidden state, The input vector for the temperature prediction model contains the data set for the current time step. is the cell candidate vector, For different weight matrices, For different bias vectors, For time The predicted value of boiler outlet working medium temperature, is the hyperbolic tangent function.

[0048] The pressure prediction model includes the consideration of pressure loss and energy balance in boiler operation based on the principles of fluid mechanics and thermodynamics, converting continuous time series data into sequence data for gated recurrent unit GRU and time series model input. The data is divided into training set and test set. The loss function is used to optimize the model during training. The input layer receives time series data, the GRU layer processes time series data, and the fully connected layer converts the output of the GRU layer into the predicted value of the outlet working fluid pressure. The specific process of establishing the pressure prediction model is as follows:

[0049] Collect relevant time series data, ensure that the data has a timestamp and that each time point is recorded, divide the data into a training set and a test set, and when performing pressure prediction, the input data includes the inlet working fluid pressure, outlet working fluid pressure, unit load, heat exchange area, inlet flue gas temperature, outlet flue gas temperature, and main feed water flow. If the data granularity is too fine or too coarse, resampling can be performed. For example, resample minute-level data to hour-level data, or resample hour-level data to day-level data.

[0050] When performing time series forecasting, continuous time series data needs to be converted into multiple time windows, each of which contains multiple continuous time step data, and an initial time series length is determined based on previous domain knowledge or data analysis.

[0051] During real-time operation, the performance of the model is monitored regularly, especially on the validation set, and the length of the time series is dynamically modified so that it can make accurate judgments for the next step. If the error rate on the validation set causes the model false alarm rate to exceed 7%, the length of the time series is reduced to 70%, and the current length is set as the new time series length. The validation set is retested using this length. If the model false alarm rate is less than 4%, it indicates that the current length meets the requirements and the current length is used. If the model false alarm rate exceeds 4% but is less than 7%, it indicates that the current length is close to meeting the requirements, and the time series length is reduced to 80%, and the current length is set as the new time series length. Use this length to test the validation set until the false alarm rate is less than 4%.

[0052] Model optimization and tuning: Use test set data to evaluate the performance of the model. Common evaluation indicators include mean square error (MSE), mean absolute error (MAE), etc. According to the evaluation results, adjust the hyperparameters and structure of the model, and continuously optimize the model through repeated experiments to obtain the best prediction performance.

[0053] Boiler operating parameter tuning includes predicting the outlet working fluid temperature through the boiler temperature prediction model, comparing it with the actual monitoring data and calculating the temperature deviation. If the temperature deviation exceeds the threshold, it is determined whether it is caused by the decrease in boiler heat transfer efficiency. If the boiler heat transfer efficiency decreases, the boiler operating parameters and working fluid mass flow rate are adjusted to adjust the outlet working fluid temperature. If the boiler heat transfer efficiency does not decrease, there is no need to adjust the boiler operating parameters and working fluid mass flow rate.

[0054] Furthermore, the outlet working medium pressure is predicted by the boiler pressure prediction model, compared with the actual monitoring data and the pressure deviation is calculated. If the pressure deviation exceeds the threshold value, if it is caused by the increase of fluid resistance in the boiler or the decrease of heat exchange efficiency, the operating load and main feed water flow of the boiler are adjusted to adjust the outlet working medium pressure. If the fluid resistance in the boiler has not increased and the heat exchange efficiency has not decreased, there is no need to adjust the operating load and main feed water flow of the boiler.

[0055] Evaluation of boiler heat transfer efficiency: Check the heat transfer coefficients in historical and current data to determine whether the current heat transfer coefficient is significantly lower than the historical optimal heat transfer coefficient or the ideal heat transfer coefficient. If the current heat transfer coefficient is low, it may indicate a decrease in heat transfer efficiency. Evaluate the cleanliness of the heating surface by calculating the cleaning factor. A low cleaning factor usually indicates that the heating surface is dusted or coked, resulting in a decrease in heat transfer efficiency. If the cleaning factor is lower than the set threshold, the possibility of a decrease in heat transfer efficiency is further confirmed. Physically inspect the boiler heating surface to see if there is dust accumulation, coking, or other physical obstacles, which usually lead to a decrease in heat transfer efficiency. If physical obstacles are found, the cause of the decrease in heat transfer efficiency is further confirmed.

[0056] Combining the temperature deviation, cleaning factor, heat transfer efficiency evaluation and physical inspection results, it is comprehensively determined whether the temperature or pressure deviation exceeds the threshold due to the decrease in heat transfer efficiency.

[0057] Furthermore, if both the temperature deviation and the pressure deviation exceed the threshold, if there is internal blockage in the boiler or leakage of the working fluid, the emergency response procedure is initiated, and the combustion adjustment, inspection and maintenance of the boiler system are reduced or stopped. If there is no internal blockage in the boiler or leakage of the working fluid, the boiler operating parameters are adjusted;

[0058] Evaluation of fluid resistance in boilers. Use acoustic wave detection equipment to evaluate changes in fluid resistance in boilers. Under ideal conditions without resistance (such as newly installed boilers or boilers that have been thoroughly cleaned), record the baseline propagation time and attenuation of the sound waves. Start the acoustic wave transmitter and transmit the sound wave signal to the inside of the boiler pipe. When the sound wave propagates in the boiler pipe, it will be reflected, scattered and attenuated when it encounters obstacles such as dust accumulation, coking or blockage. The receiver receives the sound wave signal propagated to the other end and records the sound wave propagation time and attenuation intensity. Compare the actual measured sound wave propagation time with the baseline propagation time. If the actual propagation time increases significantly, it means that the sound wave encounters greater resistance during propagation, which may be due to dust accumulation, coking or blockage in the pipe. Compare the actual measured sound wave attenuation intensity with the baseline attenuation intensity. If the attenuation intensity increases significantly, it means that the sound wave loses a lot of energy during propagation, which may be due to the increase in sound wave scattering and absorption caused by obstacles in the pipe.

[0059] According to the change of sound wave propagation time, the propagation speed of sound waves in each section of the pipeline is calculated to locate the specific location of increased resistance. Combined with the change of attenuation intensity, the severity of resistance is evaluated to determine whether it is slight dust accumulation or severe coking or blockage. Based on the test results, a detailed report is generated to describe the location and degree of increased resistance.

[0060] After optimizing and adjusting the boiler operating parameters, re-input the temperature prediction model and pressure prediction model, and iterate the cycle until the temperature and pressure deviations are within the preset threshold range. The optimized boiler operating parameters and actual monitoring data are saved in the database as historical data.

[0061] S3: Computational evaluation using secondary data.

[0062] Combine the optimized boiler operating parameters and the monitoring data of the boiler heating surface to calculate the heat transfer coefficient of the heating surface, evaluate the operating status of the boiler, and use the neural network model to build an ideal heat transfer coefficient prediction model;

[0063] It should be noted that the evaluation of the operating status of the boiler includes calculating the heat transfer coefficient of the boiler heating surface, counting the maximum and minimum values ​​of the unit load and main feed water flow, obtaining the operating range, dividing the unit load range and the main feed water flow range into several intervals, traversing the historical data, and obtaining the maximum value of the heat transfer coefficient of each interval grid:

[0064] ,

[0065] ,

[0066] ,

[0067] in, For the The maximum value of the heat transfer coefficient corresponding to the grid, are the upper and lower limits of the load interval of the i-th unit, , are the upper and lower limits of the jth main water supply flow interval, are the index values ​​of unit load and main feedwater flow, For unit load and main feedwater flow The heat transfer coefficient under the condition, is the constraint condition, max is the maximum value function;

[0068] The heat transfer coefficient formula of the heated surface is expressed as:

[0069] ,

[0070] ,

[0071] ,

[0072] in, is the heat transfer coefficient of the heated surface, is the logarithmic mean temperature difference, is a logarithmic function, is the correction of the heat transfer coefficient due to unit load and main feed water flow rate, is the heat flow, is the heat exchange area, P is the unit load, The main water supply flow rate, is the inlet flue gas temperature, is the outlet flue gas temperature, is the outlet working fluid temperature, is the inlet working fluid temperature, is the working fluid mass flow rate, is the outlet working fluid specific enthalpy, is the specific enthalpy of the inlet working fluid;

[0073] According to the actual heat transfer coefficient The maximum heat transfer coefficient is obtained by statistics , the boiler operation status is divided into low-risk status A1, medium-risk status A2 and high-risk status A3;

[0074] like , then the boiler operation state is judged to be A1 state;

[0075] like , then the boiler operation state is judged to be A2 state;

[0076] like , then the boiler operation state is judged to be A3 state.

[0077] The cleanliness factor reflects the cleanliness and heat transfer efficiency of the boiler heating surface. 0.8 is selected as the upper threshold because when the heat transfer efficiency is high, the boiler system usually runs stably, and a cleanliness factor close to or greater than 0.8 indicates that the boiler heating surface is relatively clean. 0.5 is selected as the lower threshold because when the heat transfer efficiency drops significantly, there may be serious problems in the system, and a cleanliness factor less than 0.5 usually indicates that the heating surface is seriously contaminated or coked.

[0078] Furthermore, the ideal heat transfer coefficient prediction model is constructed based on a neural network. The neural network contains two hidden layers and the input vector , the hidden layer formula is expressed as:

[0079] ,

[0080] ,

[0081] in, is the ReLU activation function, , are the weight matrices for the first and second hidden layers, are the bias vectors of the first and second hidden layers, , They are the neurons of the first and second hidden layers respectively;

[0082] The output layer formula is expressed as:

[0083] ,

[0084] in, is the sigmoid activation function, is the weight matrix of the output layer, is the bias vector of the output layer, is the ideal heat transfer coefficient.

[0085] S4: Establish models, monitor in real time, and make dynamic adjustments.

[0086] Establish a boiler heating surface cleaning factor calculation model, access the boiler heating surface monitoring data and boiler operating parameters in real time, and dynamically adjust the boiler maintenance and cleaning strategy;

[0087] It should be noted that the calculation model for the cleanliness factor of the boiler heating surface includes the cleanliness factor being the ratio of the actual heat transfer coefficient of the boiler heating surface to the ideal heat transfer coefficient. The calculation formula is expressed as:

[0088] ,

[0089] in, is the cleaning factor;

[0090] like , then the cleaning factor is judged to be in the B1 state;

[0091] like , then the cleaning factor is judged to be in the B2 state;

[0092] like , then the cleaning factor is judged to be in B3 state;

[0093] in, They are different state thresholds, and their specific values ​​are determined based on historical data and actual operating conditions, usually through statistical analysis and experimental calibration.

[0094] Dynamic adjustment of boiler maintenance and cleaning strategies includes: if the status satisfies A1 and B1, the system is judged to be in normal operation; if the status satisfies A1 and B2 or A2 and B1, the system is judged to be in optimization adjustment; if the status satisfies A3 or B3, the system is judged to be in emergency response;

[0095] When in normal operation (A1 and B1), the heat transfer efficiency is high and the current operation status is maintained without maintenance or adjustment. Regular data collection and monitoring are maintained to ensure that the cleaning factor and heat transfer coefficient are within the normal range. The current operating parameters and monitoring data are recorded in the database as the basic data for subsequent analysis and optimization. The monitoring data is reviewed regularly. If abnormal fluctuations in parameters are found, analysis is performed and the system status is re-evaluated.

[0096] When in the optimization and adjustment state (A1 and B2 or A2 and B1), there is dust or coking on the heating surface, and the heat transfer efficiency decreases. The monitoring frequency of the heating surface is increased, and the monitoring data is obtained and analyzed in time. According to the changes in the cleaning factor and the monitoring data, the operating parameters of the boiler are appropriately adjusted, the combustion parameters are optimized, the feed water flow is adjusted, the heat transfer efficiency is improved, and regular cleaning of the heating surface is planned and arranged to prevent dust or coking. After the adjustment, the system status is evaluated again. If the adjusted parameters still do not achieve the expected effect, optimization adjustments are made or upgraded to the emergency response state;

[0097] When in emergency response state (A3 or B3), the heating surface is seriously dusted or coked, start the emergency cleaning procedure, carry out comprehensive cleaning and maintenance of the heating surface to ensure that the heat transfer efficiency of the boiler returns to normal levels, and re-evaluate the system status based on the monitoring data after cleaning. If the system is still in emergency response state or new abnormalities occur, stop the boiler operation, carry out detailed system inspection, troubleshoot and solve potential blockages or leakage problems, and implement thorough repair measures.

[0098] The above is a schematic scheme of the cleaning factor evaluation method based on monitoring data of this embodiment. It should be noted that the technical scheme of the cleaning factor evaluation system based on monitoring data and the technical scheme of the cleaning factor evaluation method based on monitoring data belong to the same concept. The details of the technical scheme of the cleaning factor evaluation system based on monitoring data in this embodiment that are not described in detail can be found in the description of the technical scheme of the cleaning factor evaluation method based on monitoring data.

[0099] Example 2

[0100] The second embodiment of the present invention is different from the previous embodiment in that:

[0101] The cleaning factor evaluation method based on monitoring data of the present invention can be implemented in the form of a software functional unit and sold or used as an independent product. The software product can be stored in a computer-readable storage medium and executed by different devices (such as a personal computer, a server or a network device, etc.). Computer-readable storage media include various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk, and an optical disk. In this way, users can easily install and execute the software to realize functions such as real-time monitoring of the boiler operating status, calculation of the heat transfer coefficient, evaluation of the cleaning factor, and dynamic adjustment of the boiler maintenance strategy.

[0102] In the present invention, the logic and steps of the functional modules are composed of instructions executed by a computer and stored in a computer-readable medium. The software is implemented by an instruction execution system, device or equipment (such as a computer system, a system including a processor, etc.), including processes such as data collection, data cleaning, construction of temperature and pressure prediction models, evaluation and dynamic adjustment of boiler operating status, etc. Specifically, the operation model determines whether the operating status and cleaning factor of the boiler are in the preset green, yellow or red range based on the collected data, and triggers different maintenance or emergency response measures based on the data change trend to ensure the efficient and safe operation of the boiler system.

[0103] In addition, the technical solution of the present invention can also be implemented through hardware, software, firmware or a combination thereof. For example, in terms of hardware, electronic devices, magnetic devices, memory, etc. can be used to realize real-time processing of boiler monitoring data through programming; in terms of software, the program stored in the memory will perform data collection, processing and judgment according to the set steps, and trigger corresponding adjustment and maintenance measures; if it is implemented in the form of firmware, the instructions will be embedded in the hardware device to directly control the operation of the boiler monitoring equipment. This combination of software and hardware can efficiently complete the monitoring and optimization of the boiler system and improve the safety and intelligence level of equipment operation.

[0104] Through the integration of software products and hardware equipment, the invented functional modules can be executed on different computer or device platforms, flexibly adapting to the needs of different boiler systems, and realizing comprehensive equipment monitoring and intelligent management. This implementation method can not only improve the efficiency of the system, but also continuously improve the intelligent level of boiler operation management through software updates and functional expansion, ensure the stability, heat transfer efficiency and safety of the boiler, avoid excessive maintenance or untimely maintenance, optimize the operating performance of the boiler, and extend the service life of the equipment.

[0105] Example 3

[0106] The third embodiment of the present invention provides a cleaning factor evaluation system based on monitoring data, including: a data preprocessing module 100, a parameter adjustment module 200, a state evaluation module 300 and a strategy adjustment module 400;

[0107] The data preprocessing module is responsible for deploying monitoring equipment and obtaining monitoring data of coal-fired boilers. These data include boiler operating parameters (such as unit load, main feed water flow, etc.), boiler historical data, and heating surface monitoring data (such as inlet flue gas temperature, outlet working fluid temperature, working fluid pressure, etc.); the acquired data needs to be cleaned to ensure the integrity and accuracy of the data. The data cleaning process includes: removing missing values, excluding incomplete or missing data, removing data when the boiler is not started, and removing unstable operating data that occurs during boiler load adjustment; this step ensures the quality of the monitoring data used in subsequent analysis and modeling, and greatly improves the accuracy of model training and parameter adjustment.

[0108] The parameter adjustment module optimizes the operation parameters of the boiler by establishing a temperature prediction model and a pressure prediction model based on the cleaning data provided;

[0109] Temperature prediction model: Based on the energy balance principle of heat transfer, combined with the LSTM deep learning algorithm, the boiler outlet working fluid temperature is predicted; by comparing with the actual monitoring data, the temperature deviation is calculated. If the deviation exceeds the set threshold, the model will determine whether the boiler heat transfer efficiency has decreased, and adjust the boiler's operating parameters (such as working fluid mass flow rate, etc.) to adjust the temperature;

[0110] Pressure prediction model: Combines the principles of fluid mechanics and thermodynamics to predict the boiler outlet working fluid pressure. By comparing with the actual monitoring data, calculate the pressure deviation. If the deviation exceeds the threshold, further analyze whether it is caused by increased fluid resistance or decreased heat exchange efficiency, and then adjust the boiler load and main feed water flow;

[0111] These adjustments ensure that the temperature and pressure fluctuations of the boiler are within the allowable range under different loads and operating conditions, thereby optimizing the operating efficiency and stability of the boiler.

[0112] The state evaluation module calculates the heat transfer coefficient of the boiler heating surface and evaluates the operating state of the boiler by combining the boiler operating parameters optimized by the parameter adjustment module and the boiler heating surface monitoring data;

[0113] By calculating the actual heat transfer coefficient of the boiler and comparing it with the ideal heat transfer coefficient, the status assessment module can determine whether the heat transfer efficiency of the boiler meets expectations;

[0114] Use the neural network model to build an ideal heat transfer coefficient prediction model, so as to dynamically predict the ideal heat transfer coefficient of the boiler under different operating conditions, and provide data support for the subsequent cleaning factor evaluation;

[0115] Based on the comparison between the actual heat transfer coefficient and the ideal heat transfer coefficient, it is evaluated whether the boiler is in a low risk (A1), medium risk (A2) or high risk (A3) state, providing a basis for decision-making.

[0116] The strategy adjustment module establishes a cleaning factor calculation model by accessing the boiler's monitoring data and operating parameters in real time, and dynamically adjusts the boiler's maintenance and cleaning strategies;

[0117] The cleanliness factor calculation model is based on the ratio of the actual heat transfer coefficient of the boiler to the ideal heat transfer coefficient, and is used to evaluate the cleanliness of the boiler heating surface;

[0118] According to the change of cleaning factor, the strategy adjustment module can determine whether the boiler needs maintenance or cleaning, and adjust the maintenance frequency and cleaning strategy according to the operating status of the boiler;

[0119] If the system is in normal operation (A1 and B1), no adjustment is required and the current state is maintained; if it is in optimization adjustment (A1 and B2 or A2 and B1), the monitoring of the heating surface is increased, regular cleaning is carried out and the operating parameters are optimized; if it is in emergency response state (A3 or B3), the emergency cleaning procedure is started, and comprehensive cleaning and maintenance are carried out to ensure that the heat transfer efficiency of the boiler is restored to normal levels;

[0120] Through the collaborative work of these modules, the boiler maintenance and cleaning strategies can be dynamically adjusted to ensure that the boiler system always maintains an efficient and safe operating state under different load conditions.

[0121] Example 4

[0122] As the fourth embodiment of the present invention, a cleaning factor evaluation method based on monitoring data is provided. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0123] This embodiment aims to verify the beneficial effects of the cleaning factor evaluation method and system based on monitoring data over the traditional boiler maintenance method. The test objects are two coal-fired boilers of the same model, one of which uses the method of the present invention (experimental boiler) and the other uses the traditional maintenance method (control boiler). The detection equipment deployed in the experiment includes temperature sensors, pressure sensors, flow meters and data acquisition systems, which are used to monitor the operating parameters of the boiler in real time.

[0124] The experimental boiler dynamically adjusts the boiler maintenance and cleaning strategy according to the method of the present invention. The control boiler is maintained and cleaned according to the traditional fixed cycle without real-time data monitoring and dynamic parameter adjustment.

[0125] The experimental conditions were that the experimental boiler and the control boiler were under the same operating load, and all variables were exactly the same except for the maintenance and cleaning strategies. The experimental results are shown in Tables 1 and 2.

[0126] Table 1. Data comparison table a

[0127]

[0128] Table 2. Data comparison table b

[0129]

[0130] From the above table data, it can be seen that the experimental boiler and the control boiler show obvious differences under the same operating load. The specific analysis is as follows:

[0131] Temperature deviation and heat transfer efficiency: The outlet flue gas temperature of the experimental boiler in each test was lower than that of the control boiler, indicating that the heat transfer efficiency was higher. The experimental boiler was able to adjust operating parameters in real time to optimize the heat transfer process. The outlet working fluid temperature in the experimental boiler was significantly higher than that of the control boiler, indicating that the heat transfer of the experimental boiler was more effective and improved the overall efficiency of the boiler.

[0132] Pressure deviation and fluid resistance: The outlet working fluid pressure of the experimental boiler is higher than that of the control boiler, indicating that the fluid resistance is smaller. The method of the present invention can dynamically adjust the operating parameters and reduce the influence of fluid resistance. The control boiler adopts the traditional fixed period maintenance method, which cannot adjust and optimize the operating parameters in time, resulting in relatively large fluid resistance.

[0133] Cleaning factor and maintenance strategy: The heat transfer coefficient of the experimental boiler is significantly higher than that of the control boiler, which indicates that the heating surface of the experimental boiler is cleaner and has higher heat transfer efficiency. By calculating the cleaning factor, the experimental boiler can promptly detect dust accumulation and coking problems on the heating surface, dynamically adjust maintenance and cleaning strategies, and maintain efficient operation. The control boiler uses a fixed cycle cleaning and cannot be adjusted in time according to actual conditions, resulting in poor heat transfer efficiency and cleaning conditions.

[0134] Based on the above experiments, the experimental boiler was pressure tested under different operating loads, and the test data output is shown in Table 3.

[0135] Table 3. Pressure test data table

[0136]

[0137] Pressure deviation and fluid resistance: By comparing the actual measured outlet working fluid pressure with the predicted pressure, it is found that the pressure prediction model has a high accuracy. The pressure deviation under each operating condition is small, indicating that the model can effectively predict the outlet pressure of the boiler. Under the high load condition of 150 MW, the actual pressure deviation is slightly larger, but still within the acceptable range, indicating that the fluid resistance increases slightly under high load conditions.

[0138] Heat transfer efficiency and cleaning factor: As the operating load increases, the heat transfer coefficient gradually increases, indicating that the heat transfer efficiency of the boiler increases. The cleaning factor is close to 1 under all operating conditions, indicating that the heating surface remains in a good clean state and the heat transfer efficiency is high. This is due to real-time monitoring and dynamic adjustment of maintenance strategies, which can clean and maintain the heating surface in a timely manner.

[0139] The experimental boiler can effectively evaluate the pressure changes of the boiler through real-time monitoring and pressure prediction models, and adjust the operating parameters in time to ensure that the pressure is within a reasonable range. By calculating the cleaning factor, the boiler system can dynamically adjust the cleaning and maintenance strategy to keep the heating surface running efficiently and avoid a decrease in heat transfer efficiency.

[0140] 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 the 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 method for evaluating the cleanliness factor of a variable load boiler based on monitoring data, characterized in that: include, Obtaining primary data, including deploying monitoring equipment to obtain monitoring data of coal-fired boilers, and performing data cleaning on the acquired data. The monitoring data includes boiler operating parameters, boiler historical operating data, and boiler heating surface monitoring data; Use primary data to tune parameters and obtain secondary data, including establishing parameter adjustment models based on historical boiler operation data and monitoring data, tuning boiler operation parameters, and establishing parameter adjustment models including temperature prediction models and pressure prediction models; Utilize secondary data for calculation and evaluation, including combining optimized boiler operating parameters and boiler heating surface monitoring data to calculate the heating surface heat transfer coefficient, evaluate the boiler operating status, and use a neural network model to build an ideal heat transfer coefficient prediction model; Establish models, monitor in real time, and make dynamic adjustments, including establishing a boiler heating surface cleaning factor calculation model, real-time access to boiler heating surface monitoring data and boiler operating parameters, and dynamic adjustment of boiler maintenance and cleaning strategies; The first-level data obtained include boiler operating parameters, boiler historical data and heating surface monitoring data; the parameter tuning includes temperature tuning, pressure tuning and emergency response; The temperature optimization, if the deviation exceeds the threshold, determines the boiler heat transfer efficiency, if it decreases, adjusts the boiler operating parameters and the working fluid mass flow rate; The pressure adjustment, if the deviation exceeds the threshold value, is caused by the decrease of fluid resistance and heat exchange efficiency, then the boiler operation compliance and main feed water flow are adjusted; The temperature prediction model selects time series analysis and LSTM deep learning algorithm, uses training data for model training, uses test data to evaluate model performance, performs cross-validation to ensure model stability, inputs validation data into the trained model, predicts outlet working fluid temperature in real time, and regularly updates the model to maintain accuracy; when predicting boiler outlet working fluid temperature, uses Pearson correlation coefficient statistical method to screen features, and key features include inlet working fluid temperature, working fluid mass flow rate and heat flow; The pressure prediction model converts continuous time series data into sequence data of gated recurrent unit GRU and time series model input. The data is divided into training set and test set. The loss function is used to optimize the model during the training process. The input layer receives the time series data, the GRU layer processes the time series data, and the fully connected layer converts the output of the GRU layer into the outlet working fluid pressure prediction value; The calculation and evaluation includes comparing the actual heat transfer coefficient with the maximum value of the interval grid, judging the boiler operation status according to the difference, and dividing it into low risk state A1, medium risk state A2 and high risk state A3, and then evaluating the boiler risk level according to the cleaning factor states B1, B2 and B3; The dynamic adjustment includes judging according to the operating status. If the status satisfies A1 and B1, it is judged as a normal operating status; if the status satisfies A1, B2 and A2, B1, it is judged as an optimized adjustment status; if the status satisfies A3 and the status satisfies B3, it is judged as an emergency response status.

2. The method for evaluating the cleaning factor of a variable load boiler based on monitoring data according to claim 1, characterized in that: The operation status judgment includes adjustment under normal operation status, adjustment under optimization adjustment status and adjustment under emergency response status; The above adjustment under the normal operation state maintains the current operation state and ensures the high heat transfer efficiency of the boiler without adjustment. If abnormal parameter fluctuations are found, re-evaluate the system status; When the conductivity decreases, the detection frequency is increased during the optimal adjustment state; Adjustments are made under the emergency response state. If the system is still in the emergency response state and new abnormal conditions occur, the boiler is stopped and inspected for thorough repairs.

3. A system based on the variable load boiler cleaning factor evaluation method based on monitoring data according to any one of claims 1-2, characterized in that: include: A data preprocessing module (100), a parameter adjustment module (200), a state evaluation module (300) and a strategy adjustment module (400).

4. 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 variable load boiler cleaning factor assessment method based on monitoring data described in any one of claims 1 to 2 are implemented.

5. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for evaluating the cleaning factor of a variable load boiler based on monitoring data as described in any one of claims 1 to 2.

Images