Economizer pipeline blockage risk monitoring method and system based on intelligent working condition identification

Abstract

The invention discloses an economizer pipeline blockage risk monitoring method and system based on intelligent working condition identification, and relates to the technical field of flue gas treatment, and the method comprises the steps: collecting working condition monitoring data in a preset monitoring time window; performing pipeline blockage risk prediction to obtain a current predicted pipeline blockage probability; calculating to obtain a blockage risk deviation degree; adjusting the working condition monitoring index and the monitoring time window to obtain an optimized working condition monitoring index and an optimized monitoring time window; and data monitoring and pipeline blockage risk prediction are conducted, and if the predicted pipeline blockage probability exceeds a preset threshold value, the jaw type disrotatory screening crusher is activated to conduct active blockage clearing. The technical problems that an existing economizer pipeline blockage monitoring method mostly depends on manual inspection or a traditional fixed threshold value alarm mode, dynamic changes of the economizer operation condition and differences of characteristic parameters in different blockage stages are ignored, and consequently early warning accuracy and timeliness are insufficient are solved.

Description

Technical Field

[0001] This application relates to the field of flue gas treatment technology, specifically to a method and system for monitoring the risk of economizer pipeline blockage based on intelligent identification of operating conditions. Background Technology

[0002] With increasingly stringent environmental protection requirements and continuous improvement in energy efficiency, the economizer, as a crucial heat exchange device in the boiler system of thermal power plants, significantly impacts the boiler's thermal efficiency and the safe and stable operation of the unit. However, during long-term operation of the economizer, impurities such as fly ash and unburned carbon particles carried in the flue gas can easily deposit and adhere to the inner walls of the pipes, leading to a reduction in the pipe's cross-sectional area and even blockage.

[0003] However, in existing technologies, monitoring of economizer pipeline blockage mostly relies on manual inspection or traditional fixed threshold alarm methods. These methods are highly subjective, have long inspection cycles, or set alarm thresholds based on only one or a few fixed operating parameters, ignoring the dynamic changes in economizer operating conditions and the differences in characteristic parameters at different blockage stages, resulting in insufficient accuracy and timeliness of early warnings. Summary of the Invention

[0004] This application provides a method and system for monitoring the risk of economizer pipeline blockage based on intelligent identification of operating conditions. This solves the technical problem that existing economizer pipeline blockage monitoring methods mostly rely on manual inspection or traditional fixed threshold alarm methods, ignoring the dynamic changes in economizer operating conditions and the differences in characteristic parameters at different blockage stages, resulting in insufficient accuracy and timeliness of early warning.

[0005] The technical solution to the above-mentioned technical problems in this application is as follows: Firstly, this application provides a method for monitoring the risk of economizer pipeline blockage based on intelligent identification of operating conditions, the method comprising: The starting point of the economizer's operating cycle is taken as the monitoring starting point, and operating condition monitoring data within the preset monitoring time window are collected according to the light chemical condition monitoring indicators. Based on the aforementioned operating condition monitoring data, predict the risk of pipeline blockage in the economizer and obtain the current predicted probability of pipeline blockage. Based on a preset blockage probability curve, the blockage risk deviation is calculated according to the current predicted pipeline blockage probability. Based on the aforementioned operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, the operating condition monitoring indicators and the monitoring time window are adjusted respectively to obtain optimized operating condition monitoring indicators and optimized monitoring time windows; Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized operating condition monitoring indicators. If the predicted pipeline blockage probability exceeds the preset threshold, the jaw counter-rotating screening crusher is activated to actively clear the blockage.

[0006] Secondly, this application provides an economizer pipeline blockage risk monitoring system based on intelligent operating condition identification, including: The data acquisition module is used to take the start time of the economizer's operating cycle as the monitoring starting point and collect operating condition monitoring data within a preset monitoring time window according to the light chemical condition monitoring indicators. The risk prediction module is used to predict the risk of pipeline blockage in the economizer based on the operating condition monitoring data and obtain the current predicted pipeline blockage probability. The deviation calculation module is used to calculate the deviation of the blockage risk based on the current predicted pipeline blockage probability, using a preset blockage probability curve as a benchmark. The parameter optimization module is used to adjust the operating condition monitoring indicators and the monitoring time window based on the operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, respectively, to obtain optimized operating condition monitoring indicators and optimized monitoring time window; The unblocking execution module is used to perform data monitoring and pipeline blockage risk prediction according to the optimized operating condition monitoring indicators within the optimized monitoring time window. If the predicted pipeline blockage probability exceeds the preset threshold, the jaw counter-rotating screen crusher is activated to perform active unblocking.

[0007] This application provides one or more technical solutions, which have at least the following technical effects or advantages: This application provides a method and system for monitoring economizer pipeline blockage risk based on intelligent operating condition identification. First, it uses light-duty operating condition monitoring indicators as a starting point, collecting light-duty operating condition data sequences within a preset monitoring time window as operating condition monitoring data, avoiding data redundancy and wasted computing resources caused by too many or too complex monitoring indicators. Second, it calls an adapted pipeline blockage risk prediction model, which is generated through training with a large amount of sample data, thus ensuring the accuracy of the prediction results. Third, it calculates the blockage risk deviation based on a preset blockage probability curve, which reflects the typical blockage probability change trend of the economizer during normal operation. By comparing the current predicted probability with the benchmark value, it determines whether the current blockage risk deviates from the normal level. Finally, based on the operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, the operating condition monitoring indicators and monitoring time window are dynamically adjusted to monitor the economizer operating condition more meticulously. Finally, within the optimized monitoring time window, continuous monitoring and risk prediction are carried out according to the optimized operating condition monitoring indicators. Once the predicted probability exceeds the preset threshold, the jaw counter-rotating screen crusher is activated to actively clear blockages. This crusher can effectively remove large particles and clumps of ash from the ash hopper outlet to the ash conveying pump, reducing the possibility of pipeline blockage from the source.

[0008] Through the above technical solution, this application realizes intelligent early warning and active prevention and control of the risk of blockage in the economizer pipeline, improves the safety and stability of economizer operation, and overcomes the defects of traditional monitoring methods, such as strong subjectivity and delayed early warning. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart illustrating the economizer pipeline blockage risk monitoring method based on intelligent operating condition identification provided in this application embodiment; Figure 2 This is a schematic diagram of the economizer pipeline blockage risk system based on intelligent operating condition identification provided in this application embodiment.

[0011] The components represented by each number in the attached diagram are explained below: Data acquisition module 11, risk prediction module 12, deviation calculation module 13, parameter optimization module 14, and blockage clearing execution module 15. Detailed Implementation

[0012] This application provides a method and system for monitoring the risk of economizer pipeline blockage based on intelligent identification of operating conditions. This addresses the technical problem that existing economizer pipeline blockage monitoring methods rely heavily on manual inspections or traditional fixed threshold alarms, neglecting the dynamic changes in economizer operating conditions and the differences in characteristic parameters at different blockage stages, resulting in insufficient accuracy and timeliness of early warnings.

[0013] Example 1, as Figure 1 As shown in the embodiments of this application, a method for monitoring the risk of economizer pipeline blockage based on intelligent operating condition identification is provided, including: S10: Take the start time of the economizer's operating cycle as the monitoring starting point, and collect operating condition monitoring data within the preset monitoring time window according to the light chemical condition monitoring indicators; In this embodiment of the application, the lightweight chemical condition monitoring index refers to the set of core indicators selected from all monitorable operating parameters of the economizer that are highly correlated with pipeline blockage risk, have low data acquisition cost, and good real-time performance.

[0014] The preset monitoring time window is a fixed time length that is pre-set based on the historical operating data of the economizer and the typical blockage development cycle. For example, it can be set to 2 hours, which is used to continuously collect data on the basic operating conditions of the economizer in the early stage of monitoring.

[0015] Among them, the operating condition monitoring data collected within a preset monitoring time window according to the lightweight chemical operating condition monitoring indicators includes: Configure a set of operating condition monitoring indicators, which includes combustion operating condition monitoring indicators, heat transfer operating condition monitoring indicators, soot blowing operating condition monitoring indicators, and ash conveying operating condition monitoring indicators. The combustion operating condition monitoring indicators include furnace outlet flue gas temperature, total coal feed rate / load command, and key air volume ratio. The heat transfer operating condition monitoring indicators include economizer inlet flue gas temperature, economizer outlet flue gas temperature, and economizer flue gas temperature drop. The soot blowing operating condition monitoring indicators include soot blowing steam header pressure and cumulative operating time of a single soot blower. The ash conveying operating condition monitoring indicators include at least the peak pressure of the silo pump, pressure rise rate, duration of the conveying stage, ash hopper level, level drop rate, key valve action signals, and conveying air header pressure. Economizer flue gas temperature drop and peak pressure of silo pump delivery, which are selected from the set of operating condition monitoring indicators, are used as light chemical operating condition monitoring indicators. Lightweight chemical condition data sequences within a preset monitoring time window are collected according to the lightweight chemical condition monitoring indicators and used as operating condition monitoring data.

[0016] In this embodiment of the application, firstly, a set of operating condition monitoring indicators is constructed, which includes combustion operating condition monitoring indicators, heat transfer operating condition monitoring indicators, soot blowing operating condition monitoring indicators, and ash conveying operating condition monitoring indicators.

[0017] The combustion condition monitoring indicators include furnace outlet flue gas temperature, total coal feed / load command, and key air volume ratio, reflecting the boiler's combustion status. Unstable or incomplete combustion can lead to increased carbon content in fly ash, exacerbating the risk of ash accumulation and blockage in pipelines. Specifically, furnace outlet flue gas temperature reflects combustion intensity and flame center position; total coal feed / load command characterizes the overall boiler combustion load and is directly related to ash production; key air volume ratios, such as the primary air / secondary air ratio, affect pulverized coal burnout and ash characteristics.

[0018] The heat transfer condition monitoring indicators include the economizer inlet flue gas temperature, the economizer outlet flue gas temperature, and the economizer flue gas temperature drop. The change in flue gas temperature drop reflects the heat exchange efficiency of the economizer. When the pipeline is blocked, the heat exchange efficiency decreases, and the flue gas temperature drop will decrease accordingly.

[0019] The main monitoring indicators for soot blowing are the pressure of the soot blowing steam header and the cumulative running time of a single soot blower. Insufficient pressure or insufficient running time will make it difficult to effectively remove accumulated ash.

[0020] The specific monitoring indicators for ash conveying operation include peak pressure of the silo pump, rate of pressure rise, duration of conveying stage, ash hopper level, rate of ash level fall, key valve action signals, and pressure of the conveying air main pipe, which reflect the operating status of the ash conveying system.

[0021] Furthermore, within the aforementioned set of operating condition monitoring indicators, the economizer flue gas temperature drop and the peak pressure of the silo pump were selected as the monitoring indicators for light-duty chemical operations. The economizer flue gas temperature drop reflects the degree of ash accumulation inside the pipeline. When increased ash accumulation leads to impeded heat transfer, the flue gas temperature drop will decrease significantly. On the other hand, the peak pressure of the silo pump directly reflects the unobstructed flow of the pipeline during ash conveying. When blockage occurs, this pressure value will rise significantly.

[0022] After selecting the light-duty chemical condition monitoring indicators, data is continuously collected within the preset monitoring time window according to the set sampling frequency to form a light-duty chemical condition data sequence, which serves as the basic operating condition monitoring data for subsequent pipeline blockage risk prediction.

[0023] S20: Based on the operating condition monitoring data, predict the risk of pipeline blockage in the economizer and obtain the current predicted pipeline blockage probability; In this embodiment, the risk of pipeline blockage in the economizer is predicted to obtain the current predicted pipeline blockage probability. Specifically, a pre-trained pipeline blockage risk prediction model is invoked, and a light chemical condition data sequence is input into the pipeline blockage risk prediction model. After model prediction, the current predicted pipeline blockage probability is output.

[0024] Specifically, step S20 in the method includes: Based on the aforementioned operating condition monitoring data, the risk of pipeline blockage in the economizer is predicted, and the current predicted probability of pipeline blockage is obtained, including: Within the preset monitoring index-prediction model library, the appropriate pipeline blockage risk prediction model is called based on the light chemical condition monitoring index. The lightweight chemical condition data sequence is input into the pipeline blockage risk prediction model, and the current predicted pipeline blockage probability is output. The construction process of the adaptive pipeline blockage risk prediction model includes: Using the aforementioned lightweight chemical condition monitoring indicators as constraints, a sample lightweight chemical condition data sequence set is collected, and the proportion of historical pipeline blockage events in the economizer pipeline within a preset future time period under different sample lightweight chemical condition data sequences is obtained as the sample pipeline blockage probability, thus obtaining the sample pipeline blockage probability set. Using the sample lightweight chemical condition data sequence set as training input and the sample pipeline blockage probability set as supervision label, a deep learning model is trained until convergence to generate an adapted pipeline blockage risk prediction model.

[0025] In this embodiment, firstly, a preset monitoring index-prediction model library is constructed, which stores pipeline blockage risk prediction models corresponding to different combinations of monitoring indices. Then, based on the currently selected light-duty chemical condition monitoring indices, namely the economizer flue gas temperature drop and the peak pressure of the silo pump, the appropriate pipeline blockage risk prediction model is called from the model library.

[0026] Furthermore, the process of constructing the adaptation pipeline blockage risk prediction model is as follows: First, using lightweight chemical condition monitoring indicators as constraints, sample lightweight chemical condition data sequences were collected under different operating scenarios. These sequences included the dynamic changes in flue gas temperature drop and peak pressure of the silo pump under different loads, combustion states, and ash accumulation levels. Simultaneously, for each sample lightweight chemical condition data sequence, the proportion of actual blockage events occurring in the economizer pipeline within a preset future time period (e.g., the next 24 hours) was obtained through historical operating records, manual inspection, or other effective means. This proportion was used as the sample pipeline blockage probability, thus constructing a sample pipeline blockage probability set.

[0027] Then, the aforementioned sample light-duty chemical condition data sequence set was used as the training input for the neural network model, and the corresponding sample pipeline blockage probability set was used as the supervision label to select the neural network model for training. During the training process, the model parameters were continuously adjusted to minimize the loss function between the predicted probability and the actual sample probability until the model converged, ultimately generating an adapted pipeline blockage risk prediction model that can accurately predict the pipeline blockage probability based on the light-duty chemical condition data sequence.

[0028] For example, the steps for building and training an adaptive pipeline blockage risk prediction model based on a neural network are as follows: First, data preparation involves collecting a set of sample lightweight chemical condition data sequences and a set of sample pipeline blockage probabilities based on historical data.

[0029] Secondly, a Long Short-Term Memory (LSTM) network was chosen as the basic network architecture. The number of neurons in the model's input layer corresponds to the dimensions of the light-duty chemical operation monitoring indicators, namely two neurons, receiving sequential data on flue gas temperature drop and peak pressure delivered by the silo pump, respectively. The model has 2-3 hidden layers, each containing 64 to 128 neurons. A ReLU activation function is used to introduce nonlinear transformation capabilities, enhancing the model's ability to fit complex operating conditions. The output layer uses a single neuron, coupled with a Sigmoid activation function, mapping the output value to between 0 and 1 to represent the current predicted pipeline blockage probability.

[0030] Secondly, during the model training phase, the Adam optimizer was used for parameter optimization, with the learning rate initially set to 0.001 and dynamically adjusted based on the validation set loss during training. The binary cross-entropy loss function was chosen to measure the difference between the predicted probability and the sample pipeline blockage probability label.

[0031] Furthermore, during the training process, the sample lightweight chemical condition data sequence set is divided into a training set and a validation set in a 7:3 ratio. The training set is used to update the model parameters, while the validation set is used to monitor the model's generalization ability. When the validation set loss no longer decreases for 10 consecutive rounds, training is stopped, resulting in an adapted pipeline blockage risk prediction model.

[0032] Finally, after the light chemical condition data sequence collected within the preset monitoring time window is input into the adaptive pipeline blockage risk prediction model, the model outputs the current predicted pipeline blockage probability at the current moment based on the historical patterns and data characteristics it has learned.

[0033] S30: Based on the preset blockage probability curve, calculate the blockage risk deviation degree according to the current predicted pipeline blockage probability; In this embodiment, the preset blockage probability curve is obtained by statistical analysis and curve fitting of historical data of the economizer under long-term normal operation. It reflects the typical trend of pipeline blockage probability changing over time in a complete operating cycle of the economizer. The curve usually shows the characteristics of slow rise in the early stage, relatively stable in the middle stage, and gradual acceleration in the later stage due to equipment aging or ash accumulation.

[0034] Specifically, step S30 in the method includes: Based on the historical operating logs of the economizer, several historical blockage probability curves of the economizer's operating cycle were obtained through analysis. In a two-dimensional space, the mean of the several historical congestion probability curves is fitted simultaneously to obtain a preset congestion probability curve. Obtain the blockage probability baseline value within the preset blockage probability curve at the current moment, and use the ratio of the probability difference between the current predicted pipeline blockage probability and the blockage probability baseline value to the blockage probability baseline value as the blockage risk deviation.

[0035] In this embodiment of the application, firstly, blockage probability data for multiple complete operating cycles are extracted from the historical operating log of the economizer. Each cycle corresponds to a historical blockage probability curve, including the changes in blockage probability under various typical operating conditions such as different seasons, different load levels, and different coal types.

[0036] Secondly, in a two-dimensional space with time as the horizontal axis and blockage probability as the vertical axis, the arithmetic mean of the probability values ​​of all historical blockage probability curves at each corresponding moment is calculated to obtain the average blockage probability value at that moment. Then, using the average probability value as data points, curve fitting is performed using the least squares method to generate a smooth curve that represents the overall trend of blockage probability under normal economizer operation, i.e., the preset blockage probability curve. Based on the current monitoring moment, the corresponding blockage probability baseline value at that moment is found and obtained from the preset blockage probability curve.

[0037] Finally, calculate the difference between the current predicted pipeline blockage probability and the baseline value, and then divide this difference by the baseline value. The result is the blockage risk deviation, which can be positive or negative. The larger the risk deviation, the more severe the blockage trend. For example, if the current predicted probability is 0.3 and the baseline value is 0.2, then the blockage risk deviation is (0.3-0.2) / 0.2=0.5, or 50%, indicating that the current blockage risk is 50% higher than the normal level.

[0038] S40: Adjust the operating condition monitoring indicators and monitoring time window based on the operating condition monitoring data, the current predicted pipeline blockage probability and the blockage risk deviation, respectively, to obtain optimized operating condition monitoring indicators and optimized monitoring time window; In this embodiment of the application, the operating condition monitoring indicators and monitoring time window are adjusted based on the operating condition monitoring data obtained above, the current predicted pipeline blockage probability and the deviation of blockage risk, so as to achieve dynamic monitoring of the economizer pipeline blockage risk.

[0039] Specifically, step S40 in the method includes: The initial operating condition monitoring indicators and the initial monitoring time window are obtained by matching the current predicted pipeline blockage probability, wherein the duration of the initial monitoring time window is negatively correlated with the current predicted pipeline blockage probability; The initial operating condition monitoring indicators are adjusted based on the light-weight chemical operating condition fluctuation coefficient and the blockage risk deviation to obtain the optimized operating condition monitoring indicators. The initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation to obtain an optimized monitoring time window.

[0040] The number of monitoring indicators in the initial operating condition monitoring index is positively correlated with the current predicted pipeline blockage probability. The initial operating condition monitoring index is selected from top to bottom in the operating condition monitoring index sequence according to the corresponding number of monitoring indicators. The operating condition monitoring index sequence is obtained by sorting several operating condition monitoring indicators in the operating condition monitoring index set from largest to smallest according to their correlation with economizer pipeline blockage.

[0041] In this embodiment, firstly, initial operating condition monitoring indicators and initial monitoring time windows are obtained. The number of initial monitoring indicators and the length of the initial time window are pre-set for different probability intervals. For example, when the current predicted pipeline blockage probability is below 10%, the number of initial monitoring indicators is set to 2, i.e., maintaining the light-duty operating condition monitoring indicators, and the initial monitoring time window is set to 4 hours. When the current predicted pipeline blockage probability is between 10% and 30%, the number of initial monitoring indicators increases to 4, and the initial time window is shortened to 1.5 hours. When the current predicted pipeline blockage probability is above 30%, the number of initial monitoring indicators is set to all 8 indicators in the operating condition monitoring indicator set, and the initial time window is further shortened to 1 hour. The length of the initial monitoring time window is negatively correlated with the current predicted pipeline blockage probability; that is, the higher the predicted risk, the shorter the initial window.

[0042] Meanwhile, the number of initial operating condition monitoring indicators is positively correlated with the current predicted pipeline blockage probability. The selection rule is based on an operating condition monitoring indicator sequence, which sorts all indicators in the monitoring indicator set from highest to lowest correlation with economizer pipeline blockage. The sorting criteria can be determined through historical data analysis or expert evaluation. For example, the sorting result might be: economizer flue gas temperature drop, peak pressure of the silo pump, economizer outlet flue gas temperature, ash hopper level, furnace outlet flue gas temperature, soot blowing steam header pressure, silo pump pressure rise rate, and total coal feed / load command. When four initial monitoring indicators need to be selected, the first four indicators are selected sequentially from the top of the sequence.

[0043] Secondly, the fluctuation coefficient of light-duty chemical conditions is calculated. This coefficient measures the degree of fluctuation of the light-duty chemical condition data series within a preset monitoring time window. It can be obtained by calculating the ratio of the standard deviation to the mean of the light-duty chemical condition data series, reflecting the stability of the current operating conditions. The larger the fluctuation coefficient, the more unstable the operating conditions, the higher the uncertainty of the blockage risk, and the more comprehensive the indicator monitoring and the shorter the monitoring window are required.

[0044] Then, the initial operating condition monitoring indicators are adjusted based on the light-duty chemical condition fluctuation coefficient and the blockage risk deviation to obtain optimized operating condition monitoring indicators. For example, the adjustment rule is as follows: when the absolute value of the blockage risk deviation is greater than a preset deviation threshold of 20% or the light-duty chemical condition fluctuation coefficient is greater than a preset fluctuation threshold of 15%, 1-2 monitoring indicators are added to the initial number of operating condition monitoring indicators. The added indicators are selected from the subsequent unselected indicators in the operating condition monitoring indicator sequence; if neither exceeds the threshold, the initial operating condition monitoring indicators remain unchanged.

[0045] For example, if the initial monitoring indicators are 4, when the deviation of the blockage risk reaches 30% and the fluctuation coefficient is 20%, the number of optimized monitoring indicators increases to 6, and the newly added indicators are the fifth and sixth indicators in the sequence.

[0046] Finally, the initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation to obtain the optimized monitoring time window. The adjustment formula can be set as: Optimized monitoring time window = Initial monitoring time window × (1 - α × Absolute value of blockage risk deviation - β × Light chemical condition fluctuation coefficient), where α and β are preset adjustment weights, such as α = 0.3, β = 0.2, and the optimized window length is not less than the preset shortest window and not more than the preset longest window, where the preset shortest window is 30 minutes and the preset longest window is 6 hours.

[0047] For example, if the initial monitoring time window is 1.5 hours, the blockage risk deviation is 50%, and the fluctuation coefficient is 10%, then the optimized window = 1.5 × (1 - 0.3 × 0.5 - 0.2 × 0.1) = 1.245 hours, about 75 minutes, which achieves a further shortening of the monitoring window when the risk is high and the operating conditions fluctuate greatly.

[0048] Specifically, adjusting the initial operating condition monitoring indicators based on the light-weight chemical condition fluctuation coefficient and the blockage risk deviation includes: If the light-duty operating condition fluctuation coefficient is less than or equal to the preset standard light-duty operating condition fluctuation coefficient, and the blockage risk deviation is less than or equal to 0, then the initial operating condition monitoring index will be used as the optimized operating condition monitoring index. If the light-duty chemical condition fluctuation coefficient is greater than the preset standard light-duty chemical condition fluctuation coefficient, the ratio of the light-duty chemical condition fluctuation coefficient to the preset standard light-duty chemical condition fluctuation coefficient is divided by K to obtain the first index compensation coefficient, wherein K is greater than or equal to 3 and less than or equal to 10. If the deviation of the blockage risk is less than or equal to 0, the first index compensation coefficient shall be used as the index compensation coefficient. If the deviation of the blockage risk is greater than 0, the sum of the deviation of the blockage risk and the first index compensation coefficient shall be used as the index compensation coefficient. The product of the index compensation coefficient and the number of monitoring indicators of the initial working condition monitoring index is rounded up to obtain the number of optimized monitoring indicators. The optimized working condition monitoring indicators are then reselected from the working condition monitoring index sequence according to the number of optimized monitoring indicators.

[0049] In this embodiment, firstly, a preset standard light-weight chemical condition fluctuation coefficient is set, for example, 0.15. This value is determined by statistical analysis of the light-weight chemical condition data sequence under long-term stable operation of the economizer, representing the fluctuation benchmark level when the operating condition is stable. When the light-weight chemical condition fluctuation coefficient is ≤0.15 and the blockage risk deviation is ≤0, it indicates that the current operating condition is stable and the blockage risk is not higher than the normal level. There is no need to add monitoring indicators, and the initial operating condition monitoring indicators are directly used as optimized operating condition monitoring indicators.

[0050] If the light-duty operating condition fluctuation coefficient is greater than 0.15, it indicates abnormal fluctuations in the current operating conditions, which may affect the accuracy of blockage risk prediction. In this case, the first indicator compensation coefficient is calculated as follows: First indicator compensation coefficient = (Light-duty operating condition fluctuation coefficient / Preset standard light-duty operating condition fluctuation coefficient) / K, where K is an adjustment coefficient that can weaken the impact of the light-duty operating condition fluctuation coefficient. Its value range is 3 ≤ K ≤ 10, for example, K = 5. Assuming the light-duty operating condition fluctuation coefficient is 0.25, then the first indicator compensation coefficient = (0.25 / 0.15) / 5 ≈ 0.333.

[0051] Furthermore, if the congestion risk deviation is ≤0, it indicates that the current congestion risk trend has not exceeded the normal level. In this case, only the impact of operating condition fluctuations is considered, and the first indicator compensation coefficient is used as the final indicator compensation coefficient. For example, the aforementioned first indicator compensation coefficient of 0.333 is the indicator compensation coefficient.

[0052] If the blockage risk deviation is greater than 0, it indicates that the current blockage risk is higher than the normal level. The combined impact of operating condition fluctuations and risk deviation needs to be considered. The blockage risk deviation should be added to the first indicator compensation coefficient to obtain the indicator compensation coefficient. For example, if the blockage risk deviation is 0.5 and the first indicator compensation coefficient is 0.333, then the indicator compensation coefficient = 0.5 + 0.333 = 0.833.

[0053] Next, after obtaining the index compensation coefficient, the number of optimized monitoring indicators is calculated by "the number of optimized monitoring indicators = the index compensation coefficient × the number of monitoring indicators of the initial working condition monitoring indicators", and then rounded up.

[0054] For example, if the initial number of monitoring indicators is 4 and the indicator compensation coefficient is 0.833, then the number of optimized monitoring indicators = 0.833 × 4 = 3.332, and the number of optimized monitoring indicators is 4 when rounded up.

[0055] Finally, based on the calculated number of optimized monitoring indicators, a corresponding number of indicators are reselected from the beginning to the end of the working condition monitoring indicator sequence as optimized working condition monitoring indicators.

[0056] Furthermore, the initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation, including: If the light chemical condition fluctuation coefficient is less than or equal to the preset standard light chemical condition fluctuation coefficient, and the blockage risk deviation is less than or equal to 0, then the initial monitoring time window will be used as the optimized monitoring time window. If the light chemical condition fluctuation coefficient is greater than the preset standard light chemical condition fluctuation coefficient, the ratio of the preset standard light chemical condition fluctuation coefficient to the light chemical condition fluctuation coefficient is divided by K to obtain the first window influence coefficient. If the deviation of the blockage risk is less than or equal to 0, the window compensation coefficient is obtained by subtracting the first window influence coefficient from 1. If the congestion risk deviation is greater than 0, the first window influence coefficient is added to the congestion risk deviation to obtain the window influence coefficient, and the window compensation coefficient is obtained by subtracting the window influence coefficient from 1. The product of the window compensation coefficient and the initial monitoring time window is used as the optimized monitoring time window.

[0057] In this embodiment of the application, firstly, when the light-duty chemical condition fluctuation coefficient is ≤ the preset standard light-duty chemical condition fluctuation coefficient and the blockage risk deviation is ≤ 0, it indicates that the current operating condition is stable and the blockage risk is within the normal trend range. There is no need to shorten the monitoring window, and the initial monitoring time window is directly used as the optimized monitoring time window.

[0058] If the fluctuation coefficient of light-duty operating conditions is greater than the preset standard fluctuation coefficient of light-duty operating conditions, it indicates that the current operating conditions are fluctuating significantly, which may increase the uncertainty of the blockage risk assessment. In this case, the first window influence coefficient is calculated as follows: First window influence coefficient = (Preset standard light-duty operating condition fluctuation coefficient / Light-duty operating condition fluctuation coefficient) / K, where K is the same adjustment coefficient as in the index compensation coefficient, 3 ≤ K ≤ 10, for example, K = 5. This coefficient reflects the degree of influence of operating condition fluctuations on the monitoring window. The greater the operating condition fluctuation, the greater the light-duty operating condition fluctuation coefficient, and the smaller the first window influence coefficient.

[0059] For example, if the preset standard light chemical condition fluctuation coefficient is 0.15 and the current light chemical condition fluctuation coefficient is 0.25, then the first window influence coefficient = (0.15 / 0.25) / 5 = 0.12.

[0060] Furthermore, if the congestion risk deviation is ≤0, it indicates that the current congestion risk has not exceeded the normal level. In this case, only the impact of operating condition fluctuations is considered, and the window compensation coefficient is obtained by subtracting the first window influence coefficient from 1. For example, if the first window influence coefficient is 0.12, then the window compensation coefficient = 1 - 0.12 = 0.88.

[0061] If the blockage risk deviation is greater than 0, it means that the current blockage risk is higher than the normal level. It is necessary to take into account the dual impact of operating condition fluctuations and risk deviations. The window impact coefficient is obtained by adding the first window impact coefficient to the blockage risk deviation, and then the window compensation coefficient is obtained by subtracting the window impact coefficient from 1.

[0062] For example, if the congestion risk deviation is 0.5 and the first window influence coefficient is 0.12, then the window influence coefficient = 0.5 + 0.12 = 0.62, and the window compensation coefficient = 1 - 0.62 = 0.38.

[0063] Finally, the window compensation coefficient is multiplied by the initial monitoring time window to obtain the optimized monitoring time window. Meanwhile, to ensure the effectiveness of monitoring and avoid windows that are too short or too long, the optimized monitoring time window must be limited to between the preset shortest and longest windows.

[0064] For example, if the initial monitoring time window is 2 hours and the window compensation coefficient is 0.38, then the optimized monitoring time window = 2 × 0.38 = 0.76 hours = 45.6 minutes. This value is higher than the preset shortest window of 30 minutes, so the final optimized monitoring time window is 45.6 minutes.

[0065] S50: Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized working condition monitoring indicators. If the predicted pipeline blockage probability exceeds the preset threshold, the jaw counter-rotating screening crusher is activated to actively clear the blockage.

[0066] In this embodiment, within the optimized monitoring time window, real-time data of the economizer operation is continuously collected according to the optimized operating condition monitoring indicators to dynamically predict the pipeline blockage risk, thereby obtaining the real-time predicted pipeline blockage probability within the optimized monitoring time window. If the real-time predicted pipeline blockage probability at any time within the window exceeds the preset blockage risk threshold, an active unblocking mechanism is triggered, activating the jaw counter-rotating screening crusher.

[0067] This crusher utilizes the combined action of jaw crushing and counter-rotating screening. Upstream material is first separated by a screening filter, allowing fine ash of the correct particle size to pass directly through. Larger coke lumps that are trapped fall into the counter-rotating crusher and are forcibly crushed until their particle size is reduced to below the safe conveying range of the ash conveying system. Finally, all materials are smoothly introduced into the downstream silo pump through the aforementioned square-to-round transition section.

[0068] Specifically, before the blockage enters the ash conveying pipeline, it is actively screened and pre-treated in size, thereby eliminating mechanical blockage caused by large pieces of material at the source. It can break up the blockage such as ash and slag that may form in the economizer pipeline, and through screening, the fine particles after crushing are discharged in time to avoid further accumulation of blockage.

[0069] During the operation of the jaw counter-rotating screen crusher, the operating condition monitoring indicators and pipeline blockage probability are continuously monitored and optimized until the predicted pipeline blockage probability drops below the preset threshold. Then, the crusher operation is stopped, thereby achieving timely intervention and effective control of the risk of economizer pipeline blockage and ensuring the safe and stable operation of the economizer.

[0070] Specifically, step S50 in the method includes: Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized operating condition monitoring indicators. If the predicted pipeline blockage probability does not exceed the preset threshold, the operating condition monitoring indicators and monitoring time window are optimized and adjusted based on the operating condition monitoring data, predicted pipeline blockage probability, and blockage risk deviation in the previous monitoring time window. If the predicted probability of pipeline blockage exceeds a preset threshold, the jaw counter-rotating screen crusher is activated to actively clear the blockage. The jaw counter-rotating screen crusher is connected in series between the economizer ash hopper outlet and the ash conveying silo pump. Its structure integrates a screening filter, a counter-rotating crushing mechanism, and a transition section from top to bottom.

[0071] In this embodiment, firstly, within the optimized monitoring time window, based on the determined optimized operating condition monitoring indicators, various real-time operating parameters of the economizer are collected and input into the economizer pipeline blockage risk prediction model. Based on the input optimized operating condition monitoring indicator data, this model dynamically calculates and outputs the real-time predicted pipeline blockage probability within the current optimized monitoring time window.

[0072] If, within the entire optimized monitoring time window, the predicted pipeline blockage probability calculated at any point does not exceed the preset blockage risk threshold, then the current economizer pipeline blockage risk is determined to be within a controllable range. At this point, the next round of optimization and adjustment processes for operating condition monitoring indicators and monitoring time windows begins. This adjustment uses all operating condition monitoring data collected within the previous complete monitoring time window, the final predicted pipeline blockage probability at the end of the window, and the blockage risk deviation calculated during this period as new inputs. The entire process from initial monitoring indicator and initial window setting to optimization and adjustment is then re-executed to adapt to potential changes in operating conditions.

[0073] Conversely, if the predicted probability of pipeline blockage exceeds a preset threshold at any point within the optimized monitoring time window, the risk of blockage is deemed high, requiring proactive intervention. In this case, a start command is sent to the jaw counter-rotating screen crusher to activate the equipment for proactive unblocking operations.

[0074] Furthermore, the jaw counter-rotating screen crusher is installed in series between the economizer ash hopper outlet and the ash conveying silo pump in the economizer system, ensuring that fly ash and other materials discharged from the economizer ash hopper are processed by the crusher before entering the ash conveying silo pump. Its specific structural design integrates a screening filter, a counter-rotating crushing mechanism, and a transition section from top to bottom.

[0075] Specifically, the screening screen first intercepts and screens larger pieces of accumulated ash or slag; the counter-rotating crusher uses its high-speed rotating components to shear, squeeze and crush larger particles passing through the screen, refining them into smaller particles; the transition section between the top and bottom sections serves to connect and guide the flow, ensuring that the crushed fine particles can smoothly enter the downstream ash conveying silo pump, thereby effectively preventing large particles from clogging the pipeline.

[0076] In summary, compared with existing technologies, this application achieves precise and differentiated monitoring of economizer pipeline blockage risk by combining the lightweight chemical condition fluctuation coefficient and the blockage risk deviation, and dynamically optimizing the operating condition monitoring indicators and monitoring time window.

[0077] In summary, the embodiments of this application have at least the following technical effects: This application provides a method for monitoring economizer pipeline blockage risk based on intelligent operating condition identification. First, it uses light-weight chemical condition monitoring indicators as a starting point, collecting light-weight chemical condition data sequences within a preset monitoring time window as operating condition monitoring data, avoiding data redundancy and wasted computational resources caused by too many or too complex monitoring indicators. Second, it calls an adapted pipeline blockage risk prediction model, which is generated through training on a large amount of sample data, thus ensuring the accuracy of the prediction results. Third, it calculates the blockage risk deviation using a preset blockage probability curve as a benchmark. This preset curve reflects the typical blockage probability change trend of the economizer during normal operation. By comparing the current predicted probability with the benchmark value, it determines whether the current blockage risk deviates from the normal level. Finally, based on the operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, the operating condition monitoring indicators and monitoring time window are dynamically adjusted to monitor the economizer operating condition more meticulously. Finally, within the optimized monitoring time window, continuous monitoring and risk prediction are carried out according to the optimized operating condition monitoring indicators. Once the predicted probability exceeds the preset threshold, the jaw counter-rotating screen crusher is activated to actively clear blockages. This crusher can effectively remove large particles and clumps of ash from the ash hopper outlet to the ash conveying pump, reducing the possibility of pipeline blockage from the source.

[0078] Through the above technical solution, this application realizes intelligent early warning and active prevention and control of the risk of blockage in the economizer pipeline, improves the safety and stability of economizer operation, and overcomes the defects of traditional monitoring methods, such as strong subjectivity and delayed early warning.

[0079] Example 2, as Figure 2 As shown, based on the same inventive concept as the economizer pipeline blockage risk monitoring method based on intelligent operating condition identification provided in Embodiment 1, this application also provides an economizer pipeline blockage risk monitoring system based on intelligent operating condition identification, including: Data acquisition module 11 is used to take the start time of the economizer's operating cycle as the monitoring starting point and collect operating condition monitoring data within a preset monitoring time window according to the light chemical condition monitoring indicators. Risk prediction module 12 is used to predict the risk of pipeline blockage in the economizer based on the operating condition monitoring data and obtain the current predicted pipeline blockage probability. The deviation calculation module 13 is used to calculate the deviation of the blockage risk based on the current predicted pipeline blockage probability, using a preset blockage probability curve as a benchmark. The parameter optimization module 14 is used to adjust the operating condition monitoring indicators and the monitoring time window based on the operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, respectively, to obtain optimized operating condition monitoring indicators and optimized monitoring time window; The unblocking execution module 15 is used to perform data monitoring and pipeline blockage risk prediction according to the optimized working condition monitoring indicators within the optimized monitoring time window. If the predicted pipeline blockage probability exceeds the preset threshold, the jaw counter-rotating screen crusher is activated to perform active unblocking.

[0080] Furthermore, in one embodiment of the application, collecting operating condition monitoring data within a preset monitoring time window according to lightweight chemical operating condition monitoring indicators includes: Configure a set of operating condition monitoring indicators, which includes combustion operating condition monitoring indicators, heat transfer operating condition monitoring indicators, soot blowing operating condition monitoring indicators, and ash conveying operating condition monitoring indicators. The combustion operating condition monitoring indicators include furnace outlet flue gas temperature, total coal feed rate / load command, and key air volume ratio. The heat transfer operating condition monitoring indicators include economizer inlet flue gas temperature, economizer outlet flue gas temperature, and economizer flue gas temperature drop. The soot blowing operating condition monitoring indicators include soot blowing steam header pressure and cumulative operating time of a single soot blower. The ash conveying operating condition monitoring indicators include at least the peak pressure of the silo pump, pressure rise rate, duration of the conveying stage, ash hopper level, level drop rate, key valve action signals, and conveying air header pressure. Economizer flue gas temperature drop and peak pressure of silo pump delivery, which are selected from the set of operating condition monitoring indicators, are used as light chemical operating condition monitoring indicators. Lightweight chemical condition data sequences within a preset monitoring time window are collected according to the lightweight chemical condition monitoring indicators and used as operating condition monitoring data.

[0081] In one embodiment, the risk prediction module 12 is specifically used for: Based on the aforementioned operating condition monitoring data, the risk of pipeline blockage in the economizer is predicted, and the current predicted probability of pipeline blockage is obtained, including: Within the preset monitoring index-prediction model library, the appropriate pipeline blockage risk prediction model is called based on the light chemical condition monitoring index. The lightweight chemical condition data sequence is input into the pipeline blockage risk prediction model, and the current predicted pipeline blockage probability is output. The construction process of the adaptive pipeline blockage risk prediction model includes: Using the aforementioned lightweight chemical condition monitoring indicators as constraints, a sample lightweight chemical condition data sequence set is collected, and the proportion of historical pipeline blockage events in the economizer pipeline within a preset future time period under different sample lightweight chemical condition data sequences is obtained as the sample pipeline blockage probability, thus obtaining the sample pipeline blockage probability set. Using the sample lightweight chemical condition data sequence set as training input and the sample pipeline blockage probability set as supervision label, a deep learning model is trained until convergence to generate an adapted pipeline blockage risk prediction model.

[0082] In one embodiment, the deviation calculation module 13 is specifically used for: Based on the historical operating logs of the economizer, several historical blockage probability curves of the economizer's operating cycle were obtained through analysis. In a two-dimensional space, the mean of the several historical congestion probability curves is fitted simultaneously to obtain a preset congestion probability curve. Obtain the blockage probability baseline value within the preset blockage probability curve at the current moment, and use the ratio of the probability difference between the current predicted pipeline blockage probability and the blockage probability baseline value to the blockage probability baseline value as the blockage risk deviation.

[0083] Furthermore, the parameter optimization module 14 is specifically used for: The initial operating condition monitoring indicators and the initial monitoring time window are obtained by matching the current predicted pipeline blockage probability, wherein the duration of the initial monitoring time window is negatively correlated with the current predicted pipeline blockage probability; The initial operating condition monitoring indicators are adjusted based on the light-weight chemical operating condition fluctuation coefficient and the blockage risk deviation to obtain the optimized operating condition monitoring indicators. The initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation to obtain an optimized monitoring time window.

[0084] The number of monitoring indicators in the initial operating condition monitoring index is positively correlated with the current predicted pipeline blockage probability. The initial operating condition monitoring index is selected from top to bottom in the operating condition monitoring index sequence according to the corresponding number of monitoring indicators. The operating condition monitoring index sequence is obtained by sorting several operating condition monitoring indicators in the operating condition monitoring index set from largest to smallest according to their correlation with economizer pipeline blockage.

[0085] Furthermore, in one embodiment of the application, adjusting the initial operating condition monitoring indicators based on the lightweight chemical condition fluctuation coefficient and the blockage risk deviation includes: If the light-duty operating condition fluctuation coefficient is less than or equal to the preset standard light-duty operating condition fluctuation coefficient, and the blockage risk deviation is less than or equal to 0, then the initial operating condition monitoring index will be used as the optimized operating condition monitoring index. If the light-duty chemical condition fluctuation coefficient is greater than the preset standard light-duty chemical condition fluctuation coefficient, the ratio of the light-duty chemical condition fluctuation coefficient to the preset standard light-duty chemical condition fluctuation coefficient is divided by K to obtain the first index compensation coefficient, wherein K is greater than or equal to 3 and less than or equal to 10. If the deviation of the blockage risk is less than or equal to 0, the first index compensation coefficient shall be used as the index compensation coefficient. If the deviation of the blockage risk is greater than 0, the sum of the deviation of the blockage risk and the first index compensation coefficient shall be used as the index compensation coefficient. The product of the index compensation coefficient and the number of monitoring indicators of the initial working condition monitoring index is rounded up to obtain the number of optimized monitoring indicators. The optimized working condition monitoring indicators are then reselected from the working condition monitoring index sequence according to the number of optimized monitoring indicators.

[0086] Furthermore, the initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation, including: If the light chemical condition fluctuation coefficient is less than or equal to the preset standard light chemical condition fluctuation coefficient, and the blockage risk deviation is less than or equal to 0, then the initial monitoring time window will be used as the optimized monitoring time window. If the light chemical condition fluctuation coefficient is greater than the preset standard light chemical condition fluctuation coefficient, the ratio of the preset standard light chemical condition fluctuation coefficient to the light chemical condition fluctuation coefficient is divided by K to obtain the first window influence coefficient. If the deviation of the blockage risk is less than or equal to 0, the window compensation coefficient is obtained by subtracting the first window influence coefficient from 1. If the congestion risk deviation is greater than 0, the first window influence coefficient is added to the congestion risk deviation to obtain the window influence coefficient, and the window compensation coefficient is obtained by subtracting the window influence coefficient from 1. The product of the window compensation coefficient and the initial monitoring time window is used as the optimized monitoring time window.

[0087] Furthermore, the congestion clearing execution module 15 is specifically used for: Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized operating condition monitoring indicators. If the predicted pipeline blockage probability does not exceed the preset threshold, the operating condition monitoring indicators and monitoring time window are optimized and adjusted based on the operating condition monitoring data, predicted pipeline blockage probability, and blockage risk deviation in the previous monitoring time window. If the predicted probability of pipeline blockage exceeds a preset threshold, the jaw counter-rotating screen crusher is activated to actively clear the blockage. The jaw counter-rotating screen crusher is connected in series between the economizer ash hopper outlet and the ash conveying silo pump. Its structure integrates a screening filter, a counter-rotating crushing mechanism, and a transition section from top to bottom.

Claims

1. A method for monitoring the risk of economizer pipeline blockage based on intelligent operating condition identification, characterized in that, The methods include: The starting point of the economizer's operating cycle is taken as the monitoring starting point, and operating condition monitoring data within the preset monitoring time window are collected according to the light chemical condition monitoring indicators. Based on the aforementioned operating condition monitoring data, predict the risk of pipeline blockage in the economizer and obtain the current predicted probability of pipeline blockage. Based on a preset blockage probability curve, the blockage risk deviation is calculated according to the current predicted pipeline blockage probability. Based on the aforementioned operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, the operating condition monitoring indicators and the monitoring time window are adjusted respectively to obtain optimized operating condition monitoring indicators and optimized monitoring time windows; Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized operating condition monitoring indicators. If the predicted pipeline blockage probability exceeds the preset threshold, the jaw counter-rotating screening crusher is activated to actively clear the blockage.

2. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 1, characterized in that, According to the lightweight chemical condition monitoring indicators, operating condition monitoring data are collected within a preset monitoring time window, including: Configure a set of operating condition monitoring indicators, which includes combustion operating condition monitoring indicators, heat transfer operating condition monitoring indicators, soot blowing operating condition monitoring indicators, and ash conveying operating condition monitoring indicators. The combustion operating condition monitoring indicators include furnace outlet flue gas temperature, total coal feed rate / load command, and key air volume ratio. The heat transfer operating condition monitoring indicators include economizer inlet flue gas temperature, economizer outlet flue gas temperature, and economizer flue gas temperature drop. The soot blowing operating condition monitoring indicators include soot blowing steam header pressure and cumulative operating time of a single soot blower. The ash conveying operating condition monitoring indicators include at least the peak pressure of the silo pump, pressure rise rate, duration of the conveying stage, ash hopper level, level drop rate, key valve action signals, and conveying air header pressure. Economizer flue gas temperature drop and peak pressure of silo pump delivery, which are selected from the set of operating condition monitoring indicators, are used as light chemical operating condition monitoring indicators. Lightweight chemical condition data sequences within a preset monitoring time window are collected according to the lightweight chemical condition monitoring indicators and used as operating condition monitoring data.

3. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 2, characterized in that, Based on the aforementioned operating condition monitoring data, the risk of pipeline blockage in the economizer is predicted, and the current predicted probability of pipeline blockage is obtained, including: Within the preset monitoring index-prediction model library, the appropriate pipeline blockage risk prediction model is called based on the light chemical condition monitoring index. The lightweight chemical condition data sequence is input into the pipeline blockage risk prediction model, and the current predicted pipeline blockage probability is output. The construction process of the adaptive pipeline blockage risk prediction model includes: Using the aforementioned lightweight chemical condition monitoring indicators as constraints, a sample lightweight chemical condition data sequence set is collected, and the proportion of historical pipeline blockage events in the economizer pipeline within a preset future time period under different sample lightweight chemical condition data sequences is obtained as the sample pipeline blockage probability, thus obtaining the sample pipeline blockage probability set. Using the sample lightweight chemical condition data sequence set as training input and the sample pipeline blockage probability set as supervision label, a deep learning model is trained until convergence to generate an adapted pipeline blockage risk prediction model.

4. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 1, characterized in that, Based on a preset blockage probability curve, the blockage risk deviation is calculated according to the current predicted pipeline blockage probability, including: Based on the historical operating logs of the economizer, several historical blockage probability curves of the economizer's operating cycle were obtained through analysis. In a two-dimensional space, the mean of the several historical congestion probability curves is fitted simultaneously to obtain a preset congestion probability curve. Obtain the blockage probability baseline value within the preset blockage probability curve at the current moment, and use the ratio of the probability difference between the current predicted pipeline blockage probability and the blockage probability baseline value to the blockage probability baseline value as the blockage risk deviation.

5. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 2, characterized in that, Based on the aforementioned operating condition monitoring data, the current predicted pipeline blockage probability, and the deviation of blockage risk, the operating condition monitoring indicators and monitoring time window are adjusted respectively to obtain optimized operating condition monitoring indicators and optimized monitoring time window, including: Characteristic fluctuation analysis was performed on the economizer flue gas temperature drop sequence and the peak pressure sequence of the silo pump in the light chemical condition data sequence, respectively, and the temperature drop variation coefficient and pressure variation coefficient were calculated. The mean value was then used to calculate the light chemical condition fluctuation coefficient. The initial operating condition monitoring indicators and the initial monitoring time window are obtained by matching the current predicted pipeline blockage probability, wherein the duration of the initial monitoring time window is negatively correlated with the current predicted pipeline blockage probability; The initial operating condition monitoring indicators are adjusted based on the light-weight chemical operating condition fluctuation coefficient and the blockage risk deviation to obtain the optimized operating condition monitoring indicators. The initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation to obtain an optimized monitoring time window.

6. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 5, characterized in that, The number of monitoring indicators in the initial operating condition monitoring index is positively correlated with the current predicted pipeline blockage probability. The initial operating condition monitoring indicators are selected from top to bottom in the operating condition monitoring indicator sequence according to the corresponding number of monitoring indicators. The operating condition monitoring indicator sequence is obtained by sorting several operating condition monitoring indicators in the operating condition monitoring indicator set from largest to smallest correlation with economizer pipeline blockage.

7. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 6, characterized in that, The initial operating condition monitoring indicators are adjusted based on the light-duty operating condition fluctuation coefficient and the blockage risk deviation, including: If the light-duty operating condition fluctuation coefficient is less than or equal to the preset standard light-duty operating condition fluctuation coefficient, and the blockage risk deviation is less than or equal to 0, then the initial operating condition monitoring index will be used as the optimized operating condition monitoring index. If the light-duty chemical condition fluctuation coefficient is greater than the preset standard light-duty chemical condition fluctuation coefficient, the ratio of the light-duty chemical condition fluctuation coefficient to the preset standard light-duty chemical condition fluctuation coefficient is divided by K to obtain the first index compensation coefficient, wherein K is greater than or equal to 3 and less than or equal to 10. If the deviation of the blockage risk is less than or equal to 0, the first index compensation coefficient shall be used as the index compensation coefficient. If the deviation of the blockage risk is greater than 0, the sum of the deviation of the blockage risk and the first index compensation coefficient shall be used as the index compensation coefficient. The product of the index compensation coefficient and the number of monitoring indicators of the initial working condition monitoring index is rounded up to obtain the number of optimized monitoring indicators. The optimized working condition monitoring indicators are then reselected from the working condition monitoring index sequence according to the number of optimized monitoring indicators.

8. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 7, characterized in that, The initial monitoring time window is adjusted based on the light chemical condition fluctuation coefficient and the blockage risk deviation, including: If the light chemical condition fluctuation coefficient is less than or equal to the preset standard light chemical condition fluctuation coefficient, and the blockage risk deviation is less than or equal to 0, then the initial monitoring time window will be used as the optimized monitoring time window. If the light chemical condition fluctuation coefficient is greater than the preset standard light chemical condition fluctuation coefficient, the ratio of the preset standard light chemical condition fluctuation coefficient to the light chemical condition fluctuation coefficient is divided by K to obtain the first window influence coefficient. If the deviation of the blockage risk is less than or equal to 0, the window compensation coefficient is obtained by subtracting the first window influence coefficient from 1. If the congestion risk deviation is greater than 0, the first window influence coefficient is added to the congestion risk deviation to obtain the window influence coefficient, and the window compensation coefficient is obtained by subtracting the window influence coefficient from 1. The product of the window compensation coefficient and the initial monitoring time window is used as the optimized monitoring time window.

9. The economizer pipeline blockage risk monitoring method based on intelligent operating condition identification according to claim 1, characterized in that, Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized operating condition monitoring indicators. If the predicted pipeline blockage probability exceeds a preset threshold, the jaw counter-rotating screening crusher is activated for active unblocking, including: Within the optimized monitoring time window, data monitoring and pipeline blockage risk prediction are performed according to the optimized operating condition monitoring indicators. If the predicted pipeline blockage probability does not exceed the preset threshold, the operating condition monitoring indicators and monitoring time window are optimized and adjusted based on the operating condition monitoring data, predicted pipeline blockage probability, and blockage risk deviation in the previous monitoring time window. If the predicted probability of pipeline blockage exceeds a preset threshold, the jaw counter-rotating screen crusher is activated to actively clear the blockage. The jaw counter-rotating screen crusher is connected in series between the economizer ash hopper outlet and the ash conveying silo pump. Its structure integrates a screening filter, a counter-rotating crushing mechanism, and a transition section from top to bottom.

10. A risk monitoring system for economizer pipeline blockage based on intelligent operating condition identification, characterized in that: The method for monitoring economizer pipeline blockage risk based on intelligent operating condition identification as described in any one of claims 1-9 includes: The data acquisition module is used to take the start time of the economizer's operating cycle as the monitoring starting point and collect operating condition monitoring data within a preset monitoring time window according to the light chemical condition monitoring indicators. The risk prediction module is used to predict the risk of pipeline blockage in the economizer based on the operating condition monitoring data and obtain the current predicted pipeline blockage probability. The deviation calculation module is used to calculate the deviation of the blockage risk based on the current predicted pipeline blockage probability, using a preset blockage probability curve as a benchmark. The parameter optimization module is used to adjust the operating condition monitoring indicators and the monitoring time window based on the operating condition monitoring data, the current predicted pipeline blockage probability, and the blockage risk deviation, respectively, to obtain optimized operating condition monitoring indicators and optimized monitoring time window; The unblocking execution module is used to perform data monitoring and pipeline blockage risk prediction according to the optimized operating condition monitoring indicators within the optimized monitoring time window. If the predicted pipeline blockage probability exceeds the preset threshold, the jaw counter-rotating screen crusher is activated to perform active unblocking.

Images