Intelligent desulfurization control method and equipment based on dynamic data, medium and product

Abstract

The invention discloses an intelligent desulfurization control method and equipment based on dynamic data, a medium and a product. The method comprises the following steps: firstly, acquiring flue gas flow, SOconcentration and slurry pH value data at an inlet, carrying out de-noising and abnormal value processing to obtain effective data, and then adjusting the length of a sliding window according to the instantaneous change rate of the flue gas flow to realize time synchronization. And then combining the association between historical data and desulfurization efficiency to obtain the influence weight of each data on the desulfurization efficiency, calculating a credibility score and comparing the credibility score with a threshold value, and dynamically adjusting the flow of the slurry circulating pump and the limestone slurry supply rate according to the high-credibility data. According to the scheme, the problem of response lag of a traditional static control method is effectively solved, real-time synchronous data can be accurately acquired, the influence of each parameter on the desulfurization efficiency is quantified, the equipment operation parameters are rapidly and accurately adjusted according to the dynamic working condition, and the desulfurization control precision and stability are improved.

Description

Technical Field

[0001] This application relates to the field of industrial flue gas treatment, and in particular to intelligent desulfurization control methods, equipment, media and products based on dynamic data. Background Technology

[0002] With the acceleration of industrialization, sulfur dioxide emissions from industrial flue gas have become one of the main factors causing serious environmental pollution. To meet increasingly stringent environmental regulations, flue gas desulfurization (FGD) technology has been widely applied, among which wet FGD technology has become the mainstream desulfurization method due to its high desulfurization efficiency and mature operation. In the wet FGD process, the concentration of sulfur dioxide in the flue gas can be effectively reduced by adjusting the flow rate of the slurry circulation pump and the limestone slurry supply rate within the desulfurization tower.

[0003] In related technologies, a control method based on static parameter settings is commonly used to address the efficiency issues of wet desulfurization. This method controls the flow rate of the slurry circulation pump and the limestone slurry supply rate by setting preset fixed values ​​or proportional relationships to maintain the pH value and reactant concentration of the slurry. In this method, the control system mainly formulates parameter setting curves based on historical experience or experimental data, and meets the basic requirements of desulfurization by controlling the flow rate of the slurry circulation pump and the limestone slurry supply rate.

[0004] However, existing static parameter control methods exhibit significant response lag when facing dynamic operating conditions. When flue gas flow or sulfur dioxide concentration changes rapidly, the static parameter settings cannot adjust the slurry circulation pump flow or limestone slurry supply rate in a timely manner, resulting in a slow response of the control system to the operating conditions. This lag may lead to problems such as insufficient desulfurization reaction or excessive slurry supply, ultimately affecting desulfurization efficiency and stable equipment operation. Summary of the Invention

[0005] This application provides an intelligent desulfurization control method, equipment, medium, and product based on dynamic data, which solves the problem that the adjustment of slurry circulation pump and limestone slurry supply is lagging under dynamic operating conditions, making it difficult to adapt to rapid changes in flue gas flow and sulfur dioxide concentration, thereby improving the control accuracy and operational stability of desulfurization.

[0006] Firstly, this application provides an intelligent desulfurization control method based on dynamic data, comprising: acquiring flue gas flow rate, SO2 concentration, and slurry pH value data at the inlet of the desulfurization device; performing preprocessing operations on the flue gas flow rate, SO2 concentration, and slurry pH value data to obtain effective data, the preprocessing operations including removing high-frequency noise and handling outliers based on a sliding window; dynamically adjusting the length of the sliding window according to the instantaneous change rate of flue gas flow rate in the effective data to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value; and comparing historical flue gas flow rate, SO2 concentration, and slurry pH value data with actual desulfurization data. The correlation between sulfur efficiency and time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value is analyzed to obtain the influence weights of these parameters on desulfurization efficiency. Based on these influence weights, the reliability scores of the flue gas flow rate, SO2 concentration, and slurry pH value data are calculated. By comparing these reliability scores with a preset reliability threshold, high-reliability flue gas flow rate, SO2 concentration, and slurry pH value data are obtained. Based on these high-reliability flue gas flow rate, SO2 concentration, and slurry pH value data, the flow rate of the slurry circulation pump and the limestone slurry supply rate within the desulfurization unit are dynamically adjusted.

[0007] By employing the above technical solution, the acquired flue gas flow rate, SO2 concentration, and slurry pH data are preprocessed to remove high-frequency noise and outliers, eliminate interference factors, and restore the true state of the data. The sliding window length is dynamically adjusted based on the instantaneous change rate of flue gas flow rate. The window is shortened during drastic changes in operating conditions to quickly capture data features, and lengthened during stable operating conditions to enhance data stability. The data processing cycle is adaptively adjusted to ensure a suitable sampling range under different operating conditions. The correlation between historical data and desulfurization efficiency is analyzed, and combined with real-time data synchronized with time, the influence weight of each parameter on desulfurization efficiency is calculated. Based on this, data reliability is assessed, and high-reliability data is selected as input for the control strategy. Based on this high-quality data, the slurry circulation pump flow rate and limestone slurry supply rate are dynamically adjusted to achieve precise control of the desulfurization process, significantly improving the response speed to changes in operating conditions and the stability of desulfurization efficiency, effectively overcoming the response lag problem of traditional static control methods.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the length of the sliding window is dynamically adjusted based on the instantaneous rate of change of flue gas flow rate in the effective data to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value. Specifically, this includes: calculating the flue gas flow rate change rate of continuous data points in the effective data; comparing the flue gas flow rate change rate with a preset rate of change threshold to obtain the fluctuation of the flue gas flow rate change rate; dynamically adjusting the length of the sliding window based on the fluctuation of the flue gas flow rate change rate to obtain the adjusted sliding window length; and recalculating the timestamp of the effective data based on the adjusted sliding window length to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value.

[0009] By employing the above technical solution, the rate of change in flue gas flow can be calculated and compared with a threshold, enabling real-time capture of flow fluctuations. Dynamically adjusting the sliding window length based on these fluctuations allows for window reduction during sudden flow changes, enhancing sensitivity to such changes and making data processing more closely reflect actual operating conditions. Recalculating timestamps ensures time synchronization, guaranteeing that all parameters are analyzed on the same time base. This avoids control decision biases caused by time asynchrony and improves the accuracy of subsequent parameter correlation analysis and control adjustments.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, based on the correlation between historical flue gas flow rate, SO2 concentration, slurry pH value data and actual desulfurization efficiency, and combined with the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value, the influence weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value on desulfurization efficiency are obtained. Specifically, this includes: performing time-series segmentation on historical flue gas flow rate, SO2 concentration, and slurry pH value data to obtain multiple time window sequences. Each time window sequence includes the flue gas flow rate, SO2 concentration, and slurry pH value within the corresponding time period. The actual desulfurization efficiency was calculated. Parameter correlation was performed on the flue gas flow rate, SO2 concentration, and slurry pH value within the time window sequence to obtain a parameter correlation matrix. Based on this parameter correlation matrix and the actual desulfurization efficiency, a parameter contribution evaluation function was constructed to obtain the contribution weight coefficients of the flue gas flow rate, SO2 concentration, and slurry pH value to the desulfurization efficiency. Based on these contribution weight coefficients, and combined with the variation characteristics of the flue gas flow rate, SO2 concentration, and slurry pH value data synchronized at that time, the influence weights of the flue gas flow rate, SO2 concentration, and slurry pH value synchronized at that time on the desulfurization efficiency were obtained.

[0011] By employing the above technical solution, historical data is segmented into time window sequences, clearly defining different operating conditions. Calculating the parameter correlation matrix clarifies the interrelationships between parameters, and constructing a parameter contribution evaluation function, combined with actual desulfurization efficiency, quantifies the contribution of each parameter to the desulfurization effect. Based on the contribution weight coefficients and the changing characteristics of time-synchronized data, and considering the dynamic nature of real-time operating condition changes on parameters, the influence weights are made more consistent with the current operating conditions, providing a scientific basis for precise control.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, based on the contribution weight coefficient and combined with the variation characteristics of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value data, the influence weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value on desulfurization efficiency are obtained. Specifically, this includes: constructing a parameter coordination matrix based on the contribution weight coefficient and combined with the normalized values ​​of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value data; calculating the eigenvalues ​​of the parameter coordination matrix; correcting the contribution weight coefficient based on the magnitude of the eigenvalues ​​to obtain a corrected weight coefficient; and weighting the corrected weight coefficient with the normalized values ​​to obtain the influence weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value on desulfurization efficiency.

[0013] By employing the above technical solution, a parameter synergy matrix is ​​constructed based on contribution weight coefficients and normalized values, which comprehensively reflects the synergistic effect of each parameter at different levels. The contribution weight coefficients are corrected by calculating eigenvalues, highlighting parameter changes that significantly impact desulfurization efficiency and suppressing interference from secondary parameters. A weighted combination of normalized values ​​and corrected weight coefficients ensures that the influence weights more accurately reflect the actual impact of each parameter on desulfurization efficiency under current operating conditions, providing a reliable basis for subsequent reliability score calculations and enhancing the scientific nature of control decisions.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the reliability score of the flue gas flow rate, SO2 concentration, and slurry pH value data is calculated based on the influence weight. Specifically, this includes: calculating a weighted data value based on the influence weight, the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value; calculating the dynamic mean of the weighted data value within a preset time period using the moving exponential smoothing method; calculating the dynamic variance of the weighted data value using adaptive variance based on the dynamic mean; and calculating the reliability score of the flue gas flow rate, SO2 concentration, and slurry pH value data based on the degree of deviation between the weighted data value and the dynamic mean, combined with the dynamic variance.

[0015] By adopting the above technical solution, weighted data values are calculated based on influence weights to highlight the role of key parameters. The sliding exponential smoothing method is used to calculate the dynamic mean, which can adapt to the dynamic changes of data and track the data trend in real time. The adaptive variance calculation takes into account the dynamic characteristics of data distribution and can more accurately reflect the data dispersion degree. By combining the dynamic mean and variance to calculate the credibility score, the reliability of data can be effectively identified, abnormal or unreliable data can be filtered, and high-credibility data can be retained for control adjustment, improving the accuracy and stability of control decisions.

[0016] Combined with some embodiments of the first aspect, in some embodiments, according to the high-credibility flue gas flow rate, SO2 concentration, and slurry pH value data, the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization device are dynamically adjusted. Specifically, it includes: calculating the real-time deviation value and the historical deviation cumulative value between the high-credibility SO2 concentration and the preset SO2 emission standard value; calculating the deviation weighted score according to the real-time deviation value and the historical deviation cumulative value; setting an adaptive adjustment coefficient based on the deviation weighted score, and the adaptive adjustment coefficient increases as the deviation weighted score increases; performing multi-parameter coupling calculation on the adaptive adjustment coefficient, the high-credibility flue gas flow rate data, and the slurry pH value data to obtain the flow rate adjustment increment and the slurry supply adjustment increment; calculating the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization device according to the flow rate adjustment increment and the slurry supply adjustment increment.

[0017] By adopting the above technical solution, the real-time deviation value and the historical deviation cumulative value are calculated to comprehensively measure the gap between the current and long-term desulfurization effects and the standards. The deviation weighted score comprehensively considers the immediate and historical situations, avoiding misjudgment caused by short-term fluctuations. The adaptive adjustment coefficient increases as the deviation increases, enabling the system to increase the adjustment intensity when the deviation is large and quickly respond to abnormal working conditions. The multi-parameter coupling calculation combines factors such as flow rate and pH value, comprehensively considering the complexity of the desulfurization process, making the adjustment increment more in line with actual requirements, and achieving precise dynamic adjustment.

[0018] Combined with some embodiments of the first aspect, in some embodiments, according to the flow rate adjustment increment and the slurry supply adjustment increment, the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization device are calculated. Specifically, it includes: performing trend analysis on the flow rate adjustment increment and the slurry supply adjustment increment to obtain the change trend characteristics and the dynamic adjustment step length of the adjustment parameters; generating the stage adjustment target value and the adjustment execution time sequence according to the dynamic adjustment step length, the flow rate adjustment increment, and the slurry supply adjustment increment; realizing the stage adjustment target value according to the adjustment execution time sequence, and adjusting the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization device.

[0019] By employing the above technical solutions, trend analysis is performed on the adjustment increments to identify parameter change trends and determine reasonable dynamic adjustment step sizes, avoiding over- or under-adjustment. Stage adjustment target values ​​and execution sequences are generated, decomposing the overall adjustment into ordered stages, enabling the system to smoothly transition to the target state. Achieving stage targets according to the execution sequence reduces system oscillations, ensures smooth adjustment of slurry circulation pump flow and limestone slurry supply rate, improves the adaptability and stability of the desulfurization system to dynamic operating conditions, and achieves efficient desulfurization.

[0020] In a second aspect, this application provides a server comprising: one or more processors and a memory; the memory being coupled to the one or more processors, the memory being used to store computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the server to perform the methods described in the first aspect and any possible implementation thereof.

[0021] Thirdly, this application provides a computer-readable storage medium including instructions that, when executed on a server, cause the server to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, this application provides a computer program product that, when run on a server, causes the server to perform the method described in the first aspect and any possible implementation thereof.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By employing techniques that acquire dynamic data from the desulfurization unit inlet, remove high-frequency noise, handle outliers using a sliding window, and dynamically adjust the sliding window length based on the instantaneous rate of change of flue gas flow to achieve data time synchronization, the system first removes high-frequency noise to prevent unnecessary data fluctuations. Outlier handling based on the sliding window eliminates random errors, ensuring data accuracy and reliability. Furthermore, the sliding window length is flexibly adjusted according to the instantaneous rate of change of flue gas flow. When flow changes drastically, the window is shortened to quickly capture changes; when flow is stable, the window is extended to improve data smoothness, achieving multi-parameter time synchronization. This effectively solves the problems of data processing lag and inability to adapt to rapid changes in flue gas flow and sulfur dioxide concentration in existing static parameter control methods under dynamic conditions. This results in acquiring more accurate, real-time, and synchronized effective data, laying the foundation for subsequent desulfurization efficiency improvements and precise control.

[0024] 2. By adopting the technical means of segmenting historical data in time series, calculating the parameter correlation matrix, constructing a parameter contribution evaluation function, and determining the influence weight in combination with the change characteristics of time-synchronized data, different working condition stage time window sequences are first formed by segmenting historical data in time series, the parameter correlation matrix is calculated to clarify the mutual influence relationship between parameters, and the parameter contribution evaluation function is constructed to quantify the contribution of each parameter in combination with the actual desulfurization efficiency. Then, in combination with the change characteristics of time-synchronized data, considering the dynamics of the influence of real-time working conditions on parameters, the weight is further optimized. It effectively solves the problem that it is difficult for the existing technology to accurately measure the actual influence of each parameter on desulfurization efficiency, and further realizes the technical effect of accurately quantifying the influence weight of each dynamic parameter on desulfurization efficiency, providing a scientific basis for data-based intelligent desulfurization control.

[0025] 3. By adopting the technical means of calculating the real-time and historical deviation values, setting an adaptive adjustment coefficient, and performing multi-parameter coupling calculation to determine the adjustment increment, the real-time deviation value and the historical deviation cumulative value are first calculated to comprehensively reflect the gap between the current and long-term desulfurization effects and the standard. The deviation weighted score combines the two to avoid misjudgment caused by short-term fluctuations. The set adaptive adjustment coefficient increases as the deviation increases, so that the system increases the adjustment intensity when the deviation is large. The multi-parameter coupling calculation comprehensively considers factors such as flow rate and pH value, and comprehensively considers the complexity of the desulfurization process. It effectively solves the problem that the existing static parameter control method cannot timely and reasonably adjust the flow rate of the slurry circulation pump and the limestone feeding rate under dynamic working conditions, and further realizes the technical effect of quickly and accurately adjusting the equipment operation parameters according to dynamic working conditions, improving desulfurization efficiency and system stability. Brief Description of the Drawings

[0026] Figure 1 is a schematic flow chart of using the intelligent desulfurization control method based on dynamic data in an embodiment of the present application; Figure 2 is another schematic flow chart of using the intelligent desulfurization control method based on dynamic data in an embodiment of the present application; Figure 3 is a schematic hardware structure diagram of an electronic device in an embodiment of the present application. Detailed Embodiments

[0027] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. As used in the specification and appended claims of the present application, the singular forms "a", "one kind", "the", "above-mentioned", "this" and "this one" are intended to include the plural forms as well, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations including one or more of the listed items.

[0028] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0029] This application aims to address the problem that, under complex operating conditions such as rapid fluctuations in flue gas flow and sulfur dioxide concentration and dynamic changes in slurry pH during industrial desulfurization, the control system with static parameters sets has a slow response and cannot adjust the slurry circulation pump flow and limestone slurry supply rate in a timely manner, leading to insufficient desulfurization reaction, excessive slurry supply, and ultimately low desulfurization efficiency.

[0030] In existing technologies, a control method based on static parameter settings is typically used to address desulfurization efficiency issues. This method controls the flow rate of the slurry circulation pump and the limestone slurry supply rate by setting preset fixed values ​​or proportional relationships to maintain the pH value and reactant concentration of the slurry. In this method, the control system mainly formulates parameter setting curves based on historical experience or experimental data, and meets the basic requirements of desulfurization by controlling the flow rate of the slurry circulation pump and the limestone slurry supply rate.

[0031] The intelligent desulfurization control method based on dynamic data in this application first acquires real-time data on flue gas flow rate, SO2 concentration, and slurry pH at the inlet of the desulfurization unit. This data undergoes high-frequency noise removal and outlier processing based on a dynamic sliding window to obtain valid data. The sliding window length is dynamically adjusted according to the instantaneous rate of change of flue gas flow rate to achieve data time synchronization. Then, based on the correlation analysis between historical data and actual desulfurization efficiency, the influence weight of each parameter on desulfurization efficiency is calculated. High-reliability monitoring data is obtained through reliability score evaluation and threshold screening. Finally, based on this high-reliability data, multi-parameter coupled calculation and adaptive adjustment are used to dynamically optimize the control parameters of slurry circulation pump flow rate and limestone slurry supply rate. This method, through data preprocessing, dynamic time window, parameter weight calculation, and adaptive control, significantly improves the response speed and control accuracy of the desulfurization equipment to changes in flue gas conditions, overcoming the lag problem of traditional static control methods.

[0032] For ease of understanding, the intelligent desulfurization control method based on dynamic data provided in this application is described below: First, the flue gas flow rate, SO2 concentration, and slurry pH value data at the inlet are acquired, and valid data are obtained after noise reduction and outlier processing. Then, the sliding window length is adjusted according to the instantaneous change rate of the flue gas flow rate to achieve time synchronization. Next, the influence weight of each data on the desulfurization efficiency is obtained by combining historical data with desulfurization efficiency, a confidence score is calculated and compared with a threshold, and the slurry circulation pump flow rate and limestone slurry supply rate are dynamically adjusted based on high confidence data.

[0033] Figure 1 This is a flowchart illustrating the intelligent desulfurization control method based on dynamic data used in the embodiments of this application.

[0034] Please see Figure 1 The specific description of the intelligent desulfurization control method based on dynamic data is as follows: 101. Obtain flue gas flow rate, SO2 concentration and slurry pH data at the inlet of the desulfurization device.

[0035] Obtaining flue gas flow rate data is typically achieved using flow sensors installed at specific locations on the inlet pipe of the desulfurization unit. For example, a commonly used vortex flow meter utilizes the principle of fluid oscillation. When flue gas passes through, vortices are alternately generated on both sides of the flow meter's vortex generator. The frequency of the vortexes is linearly related to the fluid velocity (i.e., the flue gas flow rate). By detecting the vortex frequency, the flue gas flow rate can be accurately calculated.

[0036] For measuring SO2 concentration, non-dispersive infrared absorption sensors are commonly used. Their working principle is based on the strong absorption characteristic of SO2 gas by infrared light of a specific wavelength. When flue gas containing SO2 passes through the measuring chamber, the infrared light of that specific wavelength is absorbed by the flue gas. By detecting the change in infrared intensity before and after absorption, the SO2 concentration can be calculated. Taking a chemical plant as an example, the flue gas emitted during its production process contains different concentrations of SO2. Installing such a sensor at the inlet of the desulfurization unit allows for timely acquisition of dynamic changes in SO2 concentration.

[0037] The pH value of the slurry is obtained using a pH sensor. It consists of a reference electrode and a measuring electrode. When the sensor is immersed in the desulfurization slurry, it generates a corresponding potential difference based on the hydrogen ion concentration in the slurry. By measuring this potential difference and performing calculations, the pH value of the slurry can be determined. In the desulfurization process, the pH value of the slurry is crucial to the desulfurization reaction. For example, in the desulfurization system of a certain smelter, if the pH value is too high, it will lead to excessively rapid consumption of limestone and may also cause scaling on the equipment; if the pH value is too low, the desulfurization efficiency will decrease significantly.

[0038] 102. Perform preprocessing operations on the flue gas flow rate, SO2 concentration and slurry pH value data to obtain valid data. The preprocessing operations include removing high-frequency noise and handling outliers based on a sliding window.

[0039] High-frequency noise was removed from the flue gas flow rate, SO2 concentration, and slurry pH data to obtain denoised flue gas flow rate, SO2 concentration, and slurry pH data. High-frequency noise is usually generated by factors such as equipment vibration and electromagnetic interference, which can cause unstable fluctuations in the data and affect the accuracy of subsequent analysis and control. A common noise reduction method is to use digital filters, such as low-pass filters. A low-pass filter can be set with a cutoff frequency, allowing only signals below that frequency to pass through, while high-frequency noise signals above the cutoff frequency are filtered out. For example, in a desulfurization unit of a factory, electromagnetic interference caused by the operation of a large nearby motor caused high-frequency fluctuations in the collected flue gas flow rate data. By setting a low-pass filter with an appropriate cutoff frequency in the data processing stage, these high-frequency noises were successfully removed, making the flue gas flow rate data smooth and stable.

[0040] Based on the denoised flue gas flow rate, SO2 concentration, and slurry pH data, a sliding window is set. The median of the data points within the sliding window is compared with each data point within the window to identify outliers. The median of the data points within the sliding window is then used to replace the value of the outlier, yielding the valid data within the sliding window. A sliding window is a fixed-length subset of data that slides across a data sequence. Taking flue gas flow rate data as an example, assuming a sliding window length of 5 data points, as the window moves across the data sequence, statistical characteristics of the data within the window are calculated, such as the mean and standard deviation. If a data point deviates from the mean within the window by more than a certain multiple of the standard deviation (usually set to 3 times the standard deviation), the data point is considered an outlier. For example, in a set of continuously collected flue gas flow rate data, most data fluctuate between 500-600 cubic meters per hour, but suddenly a data point of 1000 cubic meters per hour appears. Sliding window calculations show that this point deviates from the mean within the window by far more than 3 times the standard deviation, thus it can be identified as an outlier. A common method for handling outliers is to replace them with the median value within the window.

[0041] 103. Based on the instantaneous change rate of flue gas flow rate in the effective data, dynamically adjust the length of the sliding window to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value.

[0042] The instantaneous rate of change of flue gas flow reflects the change in flue gas flow rate over a very short period of time. When calculating the instantaneous rate of change, the difference between two adjacent data points is usually divided by the time interval. For example, if the flue gas flow rate collected at a certain time t1 is F1, and the flue gas flow rate collected at the next adjacent time t2 is F2, with a time interval Δt of (t2-t1), the instantaneous rate of change of flue gas flow rate during this time interval is (F2-F1) / Δt.

[0043] Dynamically adjusting the sliding window length based on the calculated instantaneous rate of change allows for better capture of data trends. When the instantaneous rate of change in flue gas flow is large, it indicates a rapidly changing operating condition, and the sliding window length should be shortened. For example, in the desulfurization process of a steel plant, specific production stages such as blast furnace tapping can cause a sharp increase in flue gas flow within a short period, resulting in a larger instantaneous rate of change. If a longer sliding window is used, the data within the window will include data from many different operating conditions, failing to accurately reflect the current rapidly changing situation. Shortening the sliding window length allows for more timely and detailed tracking of rapid changes in flue gas flow, making subsequent analysis and control more targeted.

[0044] Conversely, when the instantaneous rate of change in flue gas flow is small, it indicates that the operating conditions are relatively stable. In this case, the sliding window length can be appropriately extended. Taking the stable power generation phase of a thermal power plant as an example, the flue gas flow changes relatively smoothly. Extending the sliding window length can include more data in the analysis, enhance the stability and representativeness of the data, and reduce the impact of random factors on data analysis.

[0045] After adjusting the sliding window length, data time synchronization is required. Time synchronization is crucial for accurately analyzing the relationship between flue gas flow rate, SO2 concentration, and slurry pH. Assuming the original sliding window length was a fixed 10 data points with a time span of 1 minute, and now the window length is adjusted to 5 data points based on the instantaneous rate of change, the time span becomes 0.5 minutes. At this point, the time stamp for each data point needs to be redefined to ensure that the time reference for all data points is consistent under the new window length.

[0046] 104. Based on the correlation between historical flue gas flow rate, SO2 concentration, slurry pH value and actual desulfurization efficiency, and combined with the time-synchronized flue gas flow rate, SO2 concentration and slurry pH value, the influence weights of the time-synchronized flue gas flow rate, SO2 concentration and slurry pH value on desulfurization efficiency are obtained.

[0047] First, collect data on flue gas flow rate, SO2 concentration, slurry pH value, and corresponding desulfurization efficiency over a period of time. This data should cover operating conditions under different circumstances to ensure the comprehensiveness and accuracy of the analysis results. For example, in the desulfurization system of a thermal power plant, collect the above data hourly for a month, including data under different operating conditions such as unit load changes and coal quality fluctuations.

[0048] Historical data reveals that when flue gas flow suddenly increases, desulfurization efficiency often decreases if the slurry circulation pump flow rate and limestone slurry supply rate are not adjusted in time. This indicates a close correlation between flue gas flow rate and desulfurization efficiency. Furthermore, when the slurry pH value is within a specific range, the desulfurization reaction is more complete, resulting in higher desulfurization efficiency, revealing the significant impact of slurry pH value on desulfurization efficiency.

[0049] When combining historical data with real-time data synchronized with time, it is necessary to comprehensively consider the changes in each parameter. Suppose that at a certain moment, the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value show that the current operating conditions are similar to a historically efficient desulfurization period, but with some subtle differences. In this case, the weights from the historical data cannot be simply applied; adjustments must be made based on the characteristics of the current real-time data. If the current SO2 concentration is slightly higher than the historical data, while the flue gas flow rate and slurry pH value are similar, then when calculating the influence weights, the weight of the SO2 concentration needs to be appropriately increased to reflect its greater impact on desulfurization efficiency under the current operating conditions.

[0050] The calculation of influencing weights needs to be adjusted based on the dynamic changes in real-time data. In actual operation, the desulfurization process is constantly changing and may be affected by various factors. For example, fuel replacement may cause changes in flue gas composition, thereby affecting the role of various parameters in desulfurization efficiency. By continuously adjusting the influencing weights in conjunction with real-time data, the system can better reflect the actual situation and improve the accuracy and adaptability of control.

[0051] 105. Calculate the reliability scores of the flue gas flow rate, SO2 concentration, and slurry pH data based on the influence weights.

[0052] The influence weights reflect the importance of each parameter in the desulfurization process. Calculating the reliability score based on these weights can highlight the role of key parameters. For example, in the desulfurization system of a chemical plant, analysis revealed that SO2 concentration has a significant impact on desulfurization efficiency. This means that when calculating the reliability score, the accuracy and stability of SO2 concentration data have a more significant impact on the overall reliability.

[0053] Combining the dynamic mean and variance to calculate the reliability score allows for a comprehensive assessment of data reliability. For example, if the SO2 concentration data at a certain moment deviates little from the dynamic mean and also has a small variance, it indicates that the data is within the normal fluctuation range and has high reliability. Conversely, if the deviation and variance are both large, the data may be abnormal and has low reliability.

[0054] 106. By comparing the confidence score with a preset confidence threshold, high-confidence flue gas flow rate, SO2 concentration and slurry pH data are obtained.

[0055] The preset credibility threshold is determined based on actual needs and experience in the desulfurization process. Due to varying process requirements and expectations for data accuracy, the set credibility threshold will differ. For example, in the power industry, where desulfurization accuracy is extremely high, the credibility threshold may be set high to ensure emission compliance; only data with a high credibility score will be accepted. Conversely, in smaller enterprises with relatively lower desulfurization efficiency requirements but a greater focus on cost control, the credibility threshold may be relatively lower.

[0056] The process of comparing the confidence score with the threshold is a rigorous data screening process. Taking the desulfurization system of a steel plant as an example, if the calculated confidence score of the flue gas flow rate at a certain moment is 0.8, while the preset confidence threshold is 0.7, it indicates that the confidence of the flue gas flow rate data is high, meets the data quality requirements, and will be retained as the basis for subsequent control. Conversely, if the confidence score of the slurry pH value at a certain moment is 0.6, which is lower than the preset threshold, then the reliability of the data is questionable and further processing is required, such as re-collecting data or correcting it by combining it with other data.

[0057] There are several ways to handle data that does not reach the threshold. One common method is to interpolate and estimate based on data from previous and subsequent times. For example, if the SO2 concentration data at a certain time has low reliability, but the data from the previous and subsequent times have higher reliability, then a more reasonable SO2 concentration value for that time can be estimated using linear interpolation. Alternatively, data from other relevant parameters can be combined for comprehensive judgment and correction.

[0058] This screening process yields highly reliable and accurate data.

[0059] 107. Based on the highly reliable flue gas flow rate, SO2 concentration, and slurry pH data, dynamically adjust the slurry circulation pump flow rate and limestone slurry supply rate within the desulfurization unit.

[0060] High-reliability data on flue gas flow rate, SO2 concentration, and slurry pH provide a reliable basis for adjustments. For example, when high-reliability data shows an increase in current flue gas flow rate, SO2 concentration exceeding emission standards, and slurry pH being low, this indicates a need to increase the reaction intensity to reduce SO2 emissions. In this case, it is necessary to increase the slurry circulation pump flow rate to allow more slurry to participate in the reaction, while simultaneously increasing the limestone slurry supply rate to replenish the alkaline substances required for the desulfurization reaction, thereby enhancing the desulfurization effect.

[0061] If the dynamic adjustment process involves simply making a large, one-time adjustment based on current data, it may overreact and lead to slurry waste. Therefore, the adjustment process needs to be carried out gradually, and the effects of the adjustment must be monitored in real time. For example, after increasing the slurry circulation pump flow rate, it is necessary to closely monitor changes in desulfurization efficiency and fluctuations in slurry pH, and further fine-tune the flow rate and slurry supply rate based on actual feedback.

[0062] The intelligent desulfurization control method based on dynamic data in this application embodiment acquires and preprocesses data on flue gas flow rate, SO2 concentration, and slurry pH at the inlet of the desulfurization unit to remove high-frequency noise and outliers. It dynamically adjusts the sliding window length based on the instantaneous change rate of flue gas flow to achieve time synchronization. Historical data is used to determine the weight of each parameter's impact on desulfurization efficiency, and a reliability score is calculated to select high-reliability data. Finally, based on this data, the slurry circulation pump flow rate and limestone slurry supply rate within the desulfurization unit are dynamically adjusted to achieve precise control of the desulfurization process. This method not only effectively solves the problem of response lag in traditional static parameter control methods under dynamic conditions but also enhances the stability of desulfurization efficiency, ensuring stable and efficient operation of the desulfurization process under complex and variable conditions, reducing operating costs, minimizing environmental pollution, and meeting increasingly stringent environmental regulations.

[0063] The following describes another method provided by the embodiments of this application: First, the flue gas flow rate change rate is calculated, the sliding window length is dynamically adjusted, and the timestamp is recalculated to achieve data synchronization; then, the historical data is segmented into time series, and the influence weights are determined by correlation calculation and construction of an evaluation function, and the weighting is corrected; then, the confidence score is calculated based on the influence weights; finally, the deviation is calculated based on the high confidence data, an adaptive adjustment coefficient is set, and the slurry circulation pump flow rate and limestone slurry supply rate are precisely and dynamically adjusted through multi-parameter coupling and trend analysis.

[0064] Figure 2 This is another flowchart illustrating the intelligent desulfurization control method based on dynamic data used in the embodiments of this application.

[0065] Please see Figure 2 Another method for intelligent desulfurization control based on dynamic data is described in detail below: 201. Obtain flue gas flow rate, SO2 concentration and slurry pH value data at the inlet of the desulfurization device (this step has been explained in 101 and will not be repeated here).

[0066] 202. Perform preprocessing operations on the flue gas flow rate, SO2 concentration and slurry pH value data to obtain valid data. The preprocessing operations include removing high-frequency noise and outlier processing based on sliding window (this step has been described in 102 and will not be repeated here).

[0067] 203. Calculate the rate of change of flue gas flow rate for continuous data points in the valid data.

[0068] First, read consecutive data points from the database or cache that stores valid data. Assume the obtained flue gas flow data is a time series. For example, the flue gas flow rate collected at a certain time t1 is F1, and the flue gas flow rate collected at the next adjacent time t2 is F2. The time interval Δt is (t2-t1), and the instantaneous change rate of flue gas flow rate during this time interval is (F2-F1) / Δt.

[0069] During the calculation process, the validity of the data and anomalies must also be considered. If a data point deviates significantly from the normal range, such as due to a sudden data change caused by sensor malfunction, the server needs to handle it according to preset rules. For example, a data smoothing algorithm can be used to correct the abnormal data, or the anomaly point can be skipped when calculating the rate of change to ensure that the calculated rate of change accurately reflects the changes in flue gas flow. In addition, the server will store the calculated rate of change data for subsequent analysis and use.

[0070] 204. Compare the flue gas flow rate change rate with a preset change rate threshold to obtain the fluctuation of the flue gas flow rate change rate.

[0071] The preset rate of change threshold is determined based on a comprehensive consideration of factors such as the historical operating data of the desulfurization unit, equipment performance, and process requirements. Generally, two thresholds are set: an upper threshold Vmax and a lower threshold Vmin. For example, for a specific desulfurization unit, after analyzing a large amount of historical data and conducting experimental verification, the upper threshold Vmax is determined to be 2000 cubic meters / (hour·minute), and the lower threshold Vmin is determined to be -1500 cubic meters / (hour·minute).

[0072] Each calculated rate of change is compared to these two thresholds. When the calculated rate of change Vi is greater than Vmax, it indicates that the flue gas flow rate is in a state of rapid and drastic fluctuation. For example, in the desulfurization process of a steel plant, when the blast furnace starts tapping iron, a large amount of high-temperature flue gas rapidly enters the desulfurization unit, causing a sharp increase in flue gas flow rate. Assuming that the calculated rate of change at this time is 2500 cubic meters / (hour·minute), which is greater than the preset upper limit threshold of 2000 cubic meters / (hour·minute), the server determines that the flue gas flow rate is in a state of drastic upward fluctuation.

[0073] When the rate of change Vi is less than Vmin, it indicates that the flue gas flow rate is decreasing rapidly and is also in a state of violent fluctuation. For example, during the shutdown phase of production equipment, the flue gas flow rate will decrease rapidly. If the calculated rate of change is -1800 cubic meters / (hour·minute), which is less than the lower limit threshold of -1500 cubic meters / (hour·minute), the server will recognize this rapid and violent fluctuation.

[0074] If the rate of change Vi is between Vmin and Vmax, then the flue gas flow rate changes relatively smoothly. For example, during the stable power generation period of a thermal power plant, if the rate of change of flue gas flow rate remains between -500 cubic meters per hour per minute and 1000 cubic meters per hour per minute, the server will determine that the flue gas flow rate is in a stable fluctuation state.

[0075] 205. Based on the fluctuation of the flue gas flow rate, dynamically adjust the length of the sliding window to obtain the adjusted sliding window length.

[0076] When the flue gas flow rate fluctuates drastically—that is, when the rate of change exceeds the upper threshold or falls below the lower threshold—the server shortens the sliding window. This is because, under rapidly changing operating conditions, a shorter sliding window can capture instantaneous data changes more promptly and accurately. For example, in a smelter, when the furnace is charged, a large amount of flue gas is generated, causing a sudden and significant increase in the flue gas flow rate at the desulfurization unit inlet, with the rate of change far exceeding the upper threshold. In this case, the server shortens the original sliding window of 10 data points to 5 data points. The data within the window can more closely reflect the current rapidly changing operating conditions, avoiding the inclusion of too much data from different operating conditions due to an excessively long window. This makes subsequent data analysis and processing more targeted and allows for a rapid response to changes in flue gas flow.

[0077] Conversely, if the flue gas flow rate change is relatively stable, i.e., within the upper and lower thresholds, the server will appropriately lengthen the sliding window. For example, during the stable operation of a thermal power plant, the flue gas flow rate change is relatively stable, fluctuating within a preset threshold range. In this case, the server increases the sliding window length from 5 data points to 10 data points. A longer sliding window can include more data, enhancing data stability and representativeness, and reducing the impact of random factors on data analysis. Because under stable operating conditions, data fluctuations are small, increasing the amount of data can more accurately reflect the overall trend and characteristics of the data.

[0078] 206. Based on the adjusted sliding window length, recalculate the timestamp of the valid data to obtain the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value.

[0079] First, obtain the adjusted sliding window length information. Assume the original sliding window length was a fixed 10 data points, corresponding to a time span of 1 minute. After adjustment based on the flue gas flow rate change, the window length becomes 5 data points. At this point, the server needs to redetermine the time stamp for each data point based on the new window length.

[0080] Certain rules must be followed when recalculating timestamps. If the original data collection time intervals are uniform, then when shortening the window, data points can be selected proportionally and the time remarked. When lengthening the window, it may be necessary to supplement some data points using an interpolation algorithm and allocate timestamps appropriately. For example, when the window is lengthened, the server can use linear interpolation to determine the timestamps for the newly added data points. Assuming the original window length was 5 points with a time interval of 0.5 minutes, and it is now lengthened to 10 points, 5 new points need to be inserted between the original 1st and 2nd points. The server will calculate the timestamps and corresponding data values ​​of these 5 new points using linear interpolation based on the time and data characteristics of the original two points.

[0081] By recalculating the timestamp, it can be ensured that the three sets of data—flue gas flow rate, SO2 concentration, and slurry pH value—are completely synchronized in time.

[0082] 207. The historical flue gas flow rate, SO2 concentration, and slurry pH value data are segmented over time to obtain multiple time window sequences. The time window sequences include the flue gas flow rate, SO2 concentration, slurry pH value, and actual desulfurization efficiency within the corresponding time period.

[0083] First, a large amount of historical data was read from the database, covering information about the desulfurization unit at different operating stages. For example, flue gas flow rate, SO2 concentration, slurry pH value, and corresponding actual desulfurization efficiency data were obtained, collected every 15 minutes over the past year.

[0084] The historical data is segmented into time-series segments according to preset rules. The length of the time window for each segment can be set according to actual conditions, commonly 5 minutes, 10 minutes, 30 minutes, etc. Assuming the server sets the time window length to 10 minutes, the data is divided into 10-minute intervals starting from the start time of the historical data. For example, the data from minutes 1 to 10 is divided into the first time window sequence, which includes the flue gas flow rate, SO2 concentration, slurry pH value data at each collection moment within these 10 minutes, as well as the actual desulfurization efficiency of the desulfurization unit within these 10 minutes. Then, the data from minutes 11 to 20 is divided into the second time window sequence, and so on.

[0085] During the segmentation process, it is essential to ensure the integrity and continuity of data within each time window sequence. If data is missing at a certain point in time, the server will employ appropriate methods to supplement or process it, based on the characteristics of the data and its relationship with preceding and following data. For example, for missing flue gas flow data, if the changes in preceding and following data are relatively stable, the server may use linear interpolation to estimate the missing value; if the data fluctuates significantly, it may combine data from other relevant parameters for a comprehensive estimation.

[0086] Through this time-series segmentation operation, the server obtains multiple time window sequences. These sequences are like data "slices," each representing the operating status of the desulfurization unit within a specific time period.

[0087] 208. Perform parameter correlation calculations on the flue gas flow rate, SO2 concentration, and slurry pH value in the time window sequence to obtain the parameter correlation matrix.

[0088] The server performs calculations for each time window sequence. Taking one time window sequence as an example, suppose the sequence contains data from n collection times, i.e., n sets of flue gas flow rate data (F1, F2, ..., Fn), SO2 concentration data (C1, C2, ..., Cn), and slurry pH value data (P1, P2, ..., Pn). The server will use an appropriate algorithm to calculate the correlation between the parameters, a commonly used method being the Pearson correlation coefficient algorithm.

[0089] Taking the correlation between flue gas flow rate and SO2 concentration as an example, the formula for calculating the Pearson correlation coefficient r is: The r FC Pearson correlation coefficient, representing the correlation between flue gas flow rate and SO2 concentration; It is the average value of the flue gas flow rate within that time window; It is the average SO2 concentration.

[0090] The correlations between flue gas flow rate and SO2 concentration, flue gas flow rate and slurry pH, and SO2 concentration and slurry pH were calculated sequentially. For example, the Pearson correlation coefficient between flue gas flow rate and SO2 concentration was 0.8 within a certain time window series, indicating a strong positive correlation between the two, meaning that SO2 concentration tends to increase as flue gas flow rate increases. However, the correlation coefficient between flue gas flow rate and slurry pH was -0.5, indicating a certain negative correlation, meaning that slurry pH may decrease as flue gas flow rate increases.

[0091] After calculating the correlation between parameters within each time window sequence, the server organizes these correlation values ​​into a parameter correlation matrix.

[0092] 1. Based on the parameter correlation matrix and the actual desulfurization efficiency, construct a parameter contribution evaluation function to obtain the contribution weight coefficients of the flue gas flow rate, SO2 concentration and slurry pH value to the desulfurization efficiency.

[0093] When constructing the evaluation function, the server comprehensively considers the relationship between the correlation of each parameter and the desulfurization efficiency. Let the flue gas flow rate be x1, the SO2 concentration be x2, the slurry pH value be x3, and the actual desulfurization efficiency be y. The initial function constructed is y = β0 + β1x1 + β2x2 + β3x3 + ∈, where β0 is a constant term, β1, β2, and β3 are weighting coefficients to be determined, and ∈ is the error term.

[0094] The least squares optimization algorithm was used to continuously adjust the β value based on historical data, minimizing the sum of squared errors between the predicted function value and the actual desulfurization efficiency. For example, in the data from this thermal power plant, after multiple iterations, β1 = 0.3, β2 = 0.5, and β3 = 0.2 were obtained. These β values ​​represent the contribution weighting coefficients of each parameter to the desulfurization efficiency, indicating that during this period, SO2 concentration had the greatest impact on desulfurization efficiency, followed by flue gas flow rate, while slurry pH value was relatively small, but all of them made significant contributions to desulfurization efficiency.

[0095] 210. Based on the contribution weight coefficient, and combined with the normalized values ​​of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH data, construct a parameter coordination matrix.

[0096] First, the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH data are normalized. Normalization aims to unify data of different magnitudes and ranges onto a standard scale for fair comparison and analysis. For example, in the desulfurization system of a thermal power plant, the flue gas flow rate may range from several hundred to several thousand cubic meters per hour, while the SO2 concentration may range from tens to hundreds of ppm, and the slurry pH value falls within a certain acidity / alkalinity range. Through normalization, the server converts these data into a standard interval of [0,1], resulting in normalized flue gas flow rate data.

[0097] The server constructs a parameter coordination matrix based on contribution weight coefficients and normalized data. Assume the contribution weight coefficients are a (flue gas flow rate), b (SO2 concentration), and c (slurry pH), and the normalized flue gas flow rate, SO2 concentration, and slurry pH data are F... norm C norm P normThe parameter coordination matrix constructed by the server may be a three-dimensional matrix, whose elements can represent the interaction relationship of each parameter under different weights. For example, a certain element in the matrix may represent the degree of influence of the synergistic effect of flue gas flow rate and SO2 concentration on desulfurization efficiency under the current flue gas flow rate weight a, SO2 concentration weight b, and slurry pH value weight c.

[0098] 211. Calculate the eigenvalues ​​of the parameter cooperative matrix.

[0099] The eigenvalues ​​of the constructed parametric cooperative matrix are calculated. During the calculation, the server employs specific algorithms, such as the power method or the QR algorithm. These algorithms can accurately solve for the eigenvalues ​​of the matrix. Assume the calculated eigenvalues ​​are λ1, λ2, and λ3 (for a third-order parametric cooperative matrix).

[0100] The magnitude and properties of eigenvalues ​​contain a wealth of information. Larger eigenvalues ​​usually correspond to more important eigenvectors in the matrix, and these eigenvectors are closely related to the synergistic relationships between the parameters. For example, if λ1 is the largest eigenvalue, then the combination of parameters represented by its corresponding eigenvector may play a dominant role in influencing desulfurization efficiency. This means that under this combination, the synergistic effect of the parameters has the most significant impact on desulfurization efficiency.

[0101] For example, the sign of the eigenvalue can also reflect the direction of the synergistic effect of the parameters. A positive eigenvalue may indicate that the synergistic effect between the parameters has a positive promoting effect on desulfurization efficiency, while a negative eigenvalue may mean that the synergistic effect under certain parameter combinations will inhibit desulfurization efficiency. By analyzing these eigenvalues, the server can more clearly understand the changing trend of desulfurization efficiency under different parameter synergies.

[0102] 212. Based on the magnitude of the eigenvalue, the contribution weight coefficient is corrected to obtain the corrected weight coefficient.

[0103] The server first analyzes the relationship between the eigenvalues. Taking the desulfurization system of a steel plant as an example, assume the calculated three eigenvalues ​​are λ1>λ2>λ3. The eigenvector corresponding to the larger eigenvalue λ1 reflects a strong synergistic effect among the parameters, and this effect has a more significant impact on desulfurization efficiency. The server will then adjust the contribution weight coefficients of each parameter accordingly based on the magnitude of the eigenvalues.

[0104] For parameters associated with the largest eigenvalue λ1, the server will appropriately increase their contribution weight coefficients. For example, if the synergistic effect of flue gas flow rate and SO2 concentration is significant in this eigenvector, the server may increase the weight coefficients of these two parameters. Conversely, for parameters associated with smaller eigenvalues, the server will appropriately decrease their weight coefficients. For instance, if the slurry pH value accounts for a large proportion in the eigenvector corresponding to the smaller eigenvalue λ3, but the synergistic effect represented by this eigenvector has a relatively weak impact on desulfurization efficiency, the server will decrease the contribution weight coefficient of the slurry pH value.

[0105] By adjusting the contribution weight coefficient based on the size of the eigenvalue, the server obtains the corrected weight coefficient.

[0106] 213. The corrected weighting coefficient is weighted and combined with the normalized value to obtain the influence weights of the time-synchronized flue gas flow rate, SO2 concentration and slurry pH value on the desulfurization efficiency.

[0107] Obtain the previously calculated correction weighting coefficients and the normalized values ​​of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value. Taking the desulfurization system of a thermal power plant as an example, after the previous calculations, the correction weighting coefficient for flue gas flow rate is 0.45, the correction weighting coefficient for SO2 concentration is 0.38, and the correction weighting coefficient for slurry pH value is 0.17. Meanwhile, the normalized values ​​of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value at a certain moment are 0.6, 0.7, and 0.5, respectively.

[0108] When performing weighted combination, corresponding calculations are performed based on this data. The corrected weighting coefficient for each parameter is multiplied by its corresponding normalized value, and then these products are considered together to determine the influence weight. For flue gas flow rate, its influence weight on desulfurization efficiency is a corrected weighting coefficient of 0.45 multiplied by a normalized value of 0.6; for SO2 concentration, it is a corrected weighting coefficient of 0.38 multiplied by a normalized value of 0.7; and for slurry pH value, it is a corrected weighting coefficient of 0.17 multiplied by a normalized value of 0.5.

[0109] This weighted combination method fully considers the weight of each parameter and the current real-time data status. If the correction weight coefficient of a certain parameter is high, it indicates that it is usually more important in the desulfurization process; while its normalized value reflects the actual level of the parameter at the current moment. For example, if the normalized value of SO2 concentration is high at a certain moment, and its correction weight coefficient is also large, then the role of SO2 concentration will be highlighted when calculating the influence weight, meaning that under the current operating conditions, SO2 concentration has a more critical impact on desulfurization efficiency.

[0110] 214. Based on the influence weights, the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value, the weighted data values ​​are calculated.

[0111] After determining the influence weights of time-synchronized flue gas flow rate, SO2 concentration, and slurry pH on desulfurization efficiency, weighted data values ​​were calculated based on these influence weights and the corresponding time-synchronized data. Taking the desulfurization system of a chemical plant as an example, the server obtained time-synchronized flue gas flow rate of 800 cubic meters per hour, SO2 concentration of 300 ppm, and slurry pH of 5.5 at a certain moment. The influence weights of these three parameters on desulfurization efficiency were calculated to be 0.4, 0.35, and 0.25, respectively.

[0112] When calculating the weighted data values, each parameter value is multiplied by its corresponding influence weight. For flue gas flow rate, the weighted value is 800 cubic meters per hour multiplied by 0.4; for SO2 concentration, it is 300 ppm multiplied by 0.35; and for slurry pH value, it is 5.5 multiplied by 0.25. This calculation highlights the importance of parameters with larger influence weights in the data. For example, if analysis shows that SO2 concentration has a larger influence weight on desulfurization efficiency under a certain operating condition, then the value of SO2 concentration will contribute more to the final result when calculating the weighted data values, better reflecting the key role of this parameter in the current desulfurization process.

[0113] This calculation method considers the varying degrees of importance of different parameters in the desulfurization process, making the results more representative of the overall impact of the data on desulfurization efficiency. Without weighted calculations, directly using the raw data would fail to reflect the different impacts of each parameter in the desulfurization process, potentially leading to inaccurate data assessments.

[0114] 215. Calculate the dynamic mean of the weighted data values ​​over a preset time period using the sliding exponential smoothing method.

[0115] After obtaining the weighted data values, the server uses a moving exponential smoothing method to calculate its dynamic mean over a preset time period. Taking the desulfurization system of a steel plant as an example, the server obtains a series of weighted data values. Assuming a preset time period of 10 minutes, data is collected every minute, resulting in 10 weighted data values. The core idea of ​​the moving exponential smoothing method is to assign different weights to data from different times, with newer data having higher weights, thereby highlighting real-time changes in the data.

[0116] When applying the sliding exponential smoothing method, a smoothing coefficient is first set. This coefficient is typically between 0 and 1, for example, set to 0.3. For the first data point, the dynamic mean is equal to the weighted value of that data point. Starting from the second data point, the server calculates the dynamic mean based on the smoothing coefficient. Specifically, the new dynamic mean equals the smoothing coefficient multiplied by the current weighted data value, plus (1 - smoothing coefficient) the previously calculated dynamic mean.

[0117] Assuming the first weighted data point has a value of 100, then the dynamic mean at the first time point is also 100. When the second data point has a weighted value of 110, the dynamic mean = 0.3 × 110 + (1 - 0.3) × 100 = 103. As data is continuously collected, the server continuously updates the dynamic mean according to this rule. The dynamic mean adjusts constantly with the addition of new data, responding quickly to data changes. If the weighted data value suddenly increases, the dynamic mean will rise rapidly due to the larger weight of the new data; conversely, if the data value decreases, the dynamic mean will also decrease accordingly, but the magnitude of the decrease will be influenced by previous data and will not be too drastic.

[0118] 216. Based on the dynamic mean, the dynamic variance of the weighted data values ​​is calculated using adaptive variance.

[0119] After obtaining the dynamic mean of the weighted data values ​​using the moving exponential smoothing method, adaptive variance is used to calculate the dynamic variance of these data values. Taking the desulfurization system of a cement plant as an example, the server has already calculated the dynamic mean of the weighted data values ​​over a certain period. The calculation of adaptive variance takes into account the dynamic characteristics of the data, flexibly adjusting the calculation method as the data changes. During the calculation process, the server compares the difference between each weighted data value and the dynamic mean.

[0120] If the data fluctuations are relatively stable, the adaptive variance calculation method will focus on reflecting the degree of dispersion in this stability. Assuming that over a period of time, the weighted data values ​​fluctuate slightly around the dynamic mean, the server will calculate the variance according to specific adaptive rules based on the deviations of these data points from the dynamic mean. For example, when data fluctuations are small, the adaptive variance calculation will appropriately reduce the weight of individual data points with larger deviations to avoid these outliers excessively affecting the overall variance, thus more accurately reflecting the stable fluctuation state of the data.

[0121] When data experiences significant fluctuations, adaptive variance calculations become more sensitive to these changes. For example, in cement plant production, sudden changes in raw materials can cause substantial fluctuations in desulfurization system data, increasing the deviation between the weighted data values ​​and the dynamic mean. In this case, adaptive variance calculations will take these deviations into account, allowing the variance value to rise rapidly and accurately reflect the intensified data fluctuations. This enables the server to promptly detect abnormal data fluctuations.

[0122] 217. Based on the degree of deviation between the weighted data value and the dynamic mean, and in conjunction with the dynamic variance, calculate the reliability score of the flue gas flow rate, SO2 concentration, and slurry pH value data.

[0123] First, calculate the degree of deviation between the weighted data values ​​and the dynamic mean. This degree of deviation reflects how far the current data deviates from the average level. If the deviation is small, it indicates that the data is relatively stable and consistent with the overall trend; conversely, a large deviation indicates that the data has deviated from the normal range and may indicate anomalies.

[0124] Next, a comprehensive analysis is conducted using dynamic variance. Dynamic variance reflects the dispersion of the data; the larger the variance, the more dispersed the data distribution and the worse the data stability. For example, if the dynamic variance of a certain parameter is large, it means that the data for that parameter fluctuates greatly, and its reliability may be relatively low.

[0125] When calculating the confidence score, the degree of bias and dynamic variance are integrated. Appropriate weights are assigned to each based on their magnitude. Data with both high bias and high dynamic variance receives a lower confidence score because it deviates from the average level and exhibits significant fluctuations, indicating lower reliability. Conversely, data with low bias and low dynamic variance is assigned a higher confidence score, suggesting that the data is more stable and reliable.

[0126] 218. By comparing the confidence score with the preset confidence threshold, high-confidence flue gas flow rate, SO2 concentration and slurry pH data are obtained (this step has been explained in 106 and will not be repeated here).

[0127] 219. Calculate the real-time deviation and historical cumulative deviation between the high-confidence SO2 concentration and the preset SO2 emission standard value.

[0128] It reads highly reliable SO2 concentration data and preset SO2 emission standards. These preset SO2 emission standards are determined based on environmental regulations and enterprise production requirements, and serve as an important benchmark for measuring desulfurization effectiveness. For example, environmental regulations in a certain region stipulate that the SO2 emission limit for a certain type of enterprise is 100 mg / m³. 3 This is the SO2 emission standard value preset by the company's desulfurization system.

[0129] The real-time deviation value reflects the difference between the current high-confidence SO2 concentration and the emission standard value. The real-time deviation value is obtained by subtracting the preset SO2 emission standard value from the current high-confidence SO2 concentration value.

[0130] Historical cumulative deviation is a cumulative record of the deviation between SO2 concentration and emission standards over a period of time. The server reads high-reliability SO2 concentration data from storage devices for each previous moment, calculates the deviation from the emission standard value at each moment, and then sums them up. Assuming the deviation for the past 5 moments is 5 mg / m³... 3 8mg / m 3 -3mg / m 3 (Indicates below standard), 10mg / m³ 3 12mg / m 3 Therefore, the cumulative historical deviation is 5 + 8 - 3 + 10 + 12 = 32 mg / m³ 3 In this way, the server can gain a comprehensive understanding of how SO2 concentration deviates from the standard over a period of time.

[0131] Real-time deviation values ​​and historical cumulative deviation values ​​provide the server with information from different dimensions. Real-time deviation values ​​allow the server to promptly detect whether the current operating status of the desulfurization system meets the standards, while historical cumulative deviation values ​​help the server analyze the long-term stability and trends of the desulfurization system. If the real-time deviation value is small but the historical cumulative deviation value is large, it indicates that although the current desulfurization effect is close to the standard, there are some unstable factors in the long term, requiring further optimization of the desulfurization process. Conversely, if the real-time deviation value is large but the historical cumulative deviation value is small, it may be due to an unforeseen situation causing the current poor desulfurization effect, requiring timely emergency adjustment measures.

[0132] 220. Calculate the deviation weighted score based on the real-time deviation value and the historical cumulative deviation value.

[0133] When calculating the weighted score for deviations, different weights are assigned to real-time deviation values ​​and historical cumulative deviation values ​​based on the actual situation. This is because real-time deviation values ​​reflect the immediate state of the desulfurization equipment, while historical cumulative deviation values ​​reflect stability, and the two have different importance in evaluating the desulfurization effect. Real-time deviation values ​​better reflect the current operating conditions and may be given a relatively higher weight; historical cumulative deviation values ​​help to grasp long-term trends and also play an indispensable role. For example, the server can set the weight of real-time deviation values ​​to 0.6 and the weight of historical cumulative deviation values ​​to 0.4.

[0134] Calculating the deviation-weighted score is a weighted summation process. The server multiplies the real-time deviation value by its corresponding weight, and then adds the accumulated historical deviation value multiplied by its weight to obtain the final deviation-weighted score. Assume the real-time deviation value is 20 mg / m³. 3 The historical cumulative deviation value is 30 mg / m³. 3 According to the above weighting settings, the deviation weighted score is 20×0.6+30×0.4=12+12=24.

[0135] 221. An adaptive adjustment coefficient is set based on the deviation weighted score, and the adaptive adjustment coefficient increases as the deviation weighted score increases.

[0136] First, establish the mapping relationship between the deviation-weighted score and the adaptive adjustment coefficient. Since the adaptive adjustment coefficient is required to increase with the deviation-weighted score, a linear function is used to achieve this relationship. The adaptive adjustment coefficient is set as deviation-weighted score × 0.1 (here, 0.1 is a proportional coefficient set based on actual conditions and experience; it may vary for different desulfurization equipment). If the deviation-weighted score is 20, then the adaptive adjustment coefficient is 20 × 0.1 = 2.

[0137] In addition to linear relationships, the server can also employ nonlinear functions based on the characteristics of the desulfurization system. For example, using an exponential function, the adaptive adjustment coefficient increases slowly when the deviation weighted score is small, but increases rapidly when the deviation weighted score is large. Such a nonlinear relationship can better adapt to the adjustment needs of the desulfurization equipment under different deviation levels.

[0138] 222. Perform multi-parameter coupling calculations on the adaptive adjustment coefficient, the high-reliability flue gas flow data, and the slurry pH data to obtain the flow adjustment increment and the slurry supply adjustment increment.

[0139] In multi-parameter coupled calculations, the calculation is constructed based on the inherent principles of the desulfurization process and historical operating data. Adaptive adjustment coefficients, flue gas flow data, and slurry pH data are organically combined. For example, a multivariate function incorporating these parameters is used for the calculation; the form of the function and the weights of the parameters are determined through analysis of a large amount of historical data and simulation experiments.

[0140] Assuming the adaptive adjustment coefficient is K, the high-confidence flue gas flow rate is F, and the slurry pH is P, the server-built calculation might resemble the following form: Flow rate adjustment increment ΔQ = K × a × F + b × P + c, Slurry supply adjustment increment ΔM = K × d × F + e × P + f. Here, a, b, c, d, e, and f are coefficients fitted based on the characteristics of the desulfurization equipment and historical data. These coefficients reflect the degree of influence of each parameter on the flow rate adjustment increment and the slurry supply adjustment increment.

[0141] In actual calculations, the server will substitute the real-time adaptive adjustment coefficient, flue gas flow rate data, and slurry pH value data. For example, if the current adaptive adjustment coefficient K = 1.5, the high-confidence flue gas flow rate F = 800 cubic meters / hour, and the slurry pH value P = 5.8, and the calculated values ​​are a = 0.001, b = 10, c = -5, d = 0.002, e = 8, and f = -3, then the flow rate adjustment increment ΔQ = 1.5 × 0.001 × 800 + 10 × 5.8 - 5 = 59 cubic meters / hour, and the slurry supply adjustment increment ΔM = 1.5 × 0.002 × 800 + 8 × 5.8 - 3 = 2.4 + 46.4 - 3 = 45.8.

[0142] 223. Based on the flow rate adjustment increment and slurry supply adjustment increment, the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization unit are calculated.

[0143] Trend analysis is performed on the flow rate adjustment increment and slurry supply adjustment increment to obtain the changing trend characteristics of the adjustment parameters and the dynamic adjustment step size. By analyzing the changes of these increments over a period of time, the changing trend characteristics of the adjustment parameters are obtained. For example, if the flow rate adjustment increment shows a gradually increasing trend at several consecutive time points, it indicates that the flow rate of the slurry circulation pump needs to be continuously increased to cope with changes in SO2 concentration in the flue gas or other operating conditions. At the same time, by determining the dynamic adjustment step size, this step size determines the magnitude of each adjustment. The determination of the dynamic adjustment step size should comprehensively consider various factors, such as the equipment's response speed, stability requirements, and the equipment's adjustable range. If the equipment has high requirements for response to changes and allows for rapid adjustment, then the dynamic adjustment step size can be appropriately increased; conversely, if the equipment prioritizes stability and avoids fluctuations caused by excessive adjustment, the dynamic adjustment step size will be relatively small.

[0144] Based on the dynamic adjustment step size, the flow rate adjustment increment, and the slurry supply adjustment increment, stage adjustment target values ​​and adjustment execution sequences are generated. The stage adjustment target values ​​decompose the overall adjustment task into multiple smaller, staged targets, allowing the equipment to smoothly transition to the final target state. For example, assuming the current slurry circulation pump flow rate is 500 cubic meters per hour, the flow rate adjustment increment is 50 cubic meters per hour, and the dynamic adjustment step size is 20 cubic meters per hour, then the server might set the stage adjustment target values ​​as follows: first, adjust the flow rate to 520 cubic meters per hour; second, adjust it to 540 cubic meters per hour; and third, adjust it to 550 cubic meters per hour (close to the final target of 550 cubic meters per hour). Simultaneously, the server determines the execution sequence of each stage adjustment target, specifying when the corresponding adjustment operation will be performed.

[0145] The adjustment is executed according to the specified timing sequence to achieve the target values ​​for each stage, adjusting the flow rate of the slurry circulation pump and the limestone slurry supply rate within the desulfurization unit. During execution, the server sends control commands to relevant equipment to precisely control the slurry circulation pump to adjust the flow rate and to control the delivery speed of the limestone slurry supply equipment to adjust the slurry supply rate. During adjustment, the server also monitors the operating status of the desulfurization system in real time, including key indicators such as desulfurization efficiency, SO2 concentration, and slurry pH. If the actual adjustment effect does not match expectations, the server will promptly adjust subsequent adjustment strategies, such as recalculating the adjustment increment, adjusting the dynamic adjustment step size, or modifying the stage adjustment target value, to ensure that the adjustment of the slurry circulation pump flow rate and limestone slurry supply rate is accurate and stable, enabling the desulfurization system to operate efficiently and stably, meeting the requirements of the desulfurization process, reducing sulfur dioxide emissions, and achieving environmental standards.

[0146] The intelligent desulfurization control method based on dynamic data in this application embodiment acquires and preprocesses data on flue gas flow rate, SO2 concentration, and slurry pH value at the inlet of the desulfurization device to remove high-frequency noise and outliers. It dynamically adjusts the sliding window length based on the instantaneous change rate of flue gas flow to achieve data synchronization. Historical data is combined to determine the weight of each parameter's impact on desulfurization efficiency. A reliability score is calculated to select high-reliability data. The deviation is then calculated based on the high-reliability data, and an adaptive adjustment coefficient is set. Through multi-parameter coupling and trend analysis, the adjustment increment is determined, achieving precise dynamic adjustment and accurate control of the desulfurization process. This not only effectively solves the problem of lag in response under dynamic conditions in traditional static parameter control methods but also enhances the stability of desulfurization efficiency, ensuring stable and efficient operation of the desulfurization process under complex and variable conditions, reducing environmental pollution, and meeting increasingly stringent environmental regulations.

[0147] The intelligent desulfurization control method based on dynamic data provided in the above embodiments can be executed by a server, which is composed of electronic equipment. The electronic equipment in this embodiment of the invention is described below from a hardware processing perspective; please refer to [link to relevant documentation]. Figure 3 This is a schematic diagram of a hardware structure of an electronic device in an embodiment of this application.

[0148] It should be noted that, Figure 3 The structure of the electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0149] like Figure 3 As shown, the electronic device includes a Central Processing Unit (CPU) 301, which can perform various appropriate actions and processes according to a program stored in Read-Only Memory (ROM) 302 or a program loaded from storage portion 308 into Random Access Memory (RAM) 303, such as performing the methods described in the above embodiments. The Random Access Memory (RAM) 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An Input / Output (I / O) interface 305 is also connected to the bus 304.

[0150] The following components are connected to the input / output (I / O) interface 305: an input section 306 including audio input devices, push-button switches, etc.; an output section 307 including displays, audio output devices, indicator lights, etc.; a storage section 308 including hard disks, etc.; and a communication section 309 including network interface cards such as LAN (Local Area Network) cards, modems, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the input / output (I / O) interface 305 as needed. Removable media 311, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on the drive 310 as needed so that computer programs read from them can be installed into the storage section 308 as needed.

[0151] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by the Central Processing Unit (CPU) 301, it performs the various functions defined in the present invention.

[0152] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0153] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0154] Specifically, the electronic device of this embodiment includes a processor and a memory. The memory is coupled to one or more processors and is used to store computer program code. The computer program code includes computer instructions. One or more processors call the computer instructions to cause the electronic device to perform the method provided in the above embodiment.

[0155] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The storage medium carries one or more computer programs that, when executed by a processor of the electronic device, cause the electronic device to implement the methods provided in the above embodiments.

[0156] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0157] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0158] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0159] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A smart desulfurization control method based on dynamic data, characterized in that, include: Acquire data on flue gas flow rate, SO2 concentration, and slurry pH at the inlet of the desulfurization unit; The flue gas flow rate, SO2 concentration, and slurry pH data are preprocessed to obtain valid data. The preprocessing operation includes removing high-frequency noise and handling outliers based on a sliding window. Based on the instantaneous change rate of flue gas flow rate in the effective data, the length of the sliding window is dynamically adjusted to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value. Based on the correlation between historical flue gas flow rate, SO2 concentration, slurry pH value data and actual desulfurization efficiency, and combined with the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value, the influence weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value on desulfurization efficiency are obtained. Based on the influence weights, the reliability scores of the flue gas flow rate, SO2 concentration, and slurry pH data are calculated. By comparing the confidence score with a preset confidence threshold, high-confidence flue gas flow rate, SO2 concentration, and slurry pH value data are obtained. Based on the highly reliable flue gas flow rate, SO2 concentration, and slurry pH data, the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization unit are dynamically adjusted.

2. The method according to claim 1, characterized in that, Based on the instantaneous change rate of flue gas flow rate in the effective data, the length of the sliding window is dynamically adjusted to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value, specifically including: Calculate the rate of change of flue gas flow rate for consecutive data points in the valid data; The flue gas flow rate change rate is compared with a preset change rate threshold to obtain the fluctuation of the flue gas flow rate change rate; The length of the sliding window is dynamically adjusted based on the fluctuation of the flue gas flow rate to obtain the adjusted sliding window length. Based on the adjusted sliding window length, the timestamps of the valid data are recalculated to obtain time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value.

3. The method according to claim 1, characterized in that, The correlation between historical flue gas flow rate, SO2 concentration, slurry pH value data and actual desulfurization efficiency is established. Combined with time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value, the influence weights of these factors on desulfurization efficiency are obtained, specifically including: Historical flue gas flow rate, SO2 concentration, and slurry pH value data are segmented over time to obtain multiple time window sequences. The time window sequences include flue gas flow rate, SO2 concentration, slurry pH value, and actual desulfurization efficiency within the corresponding time period. The parameter correlation matrix is ​​obtained by performing parameter correlation calculations on the flue gas flow rate, SO2 concentration, and slurry pH value in the time window sequence. Based on the parameter correlation matrix and the actual desulfurization efficiency, a parameter contribution evaluation function is constructed to obtain the contribution weight coefficients of the flue gas flow rate, SO2 concentration and slurry pH value to the desulfurization efficiency. Based on the contribution weight coefficients and combined with the variation characteristics of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value data, the influence weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value on desulfurization efficiency are obtained.

4. The method according to claim 3, characterized in that, Based on the aforementioned contribution weighting coefficients, and combined with the variation characteristics of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH data, the influence weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH on desulfurization efficiency are obtained, specifically including: Based on the contribution weight coefficient, and combined with the normalized values ​​of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH data, a parameter coordination matrix is ​​constructed. Calculate the eigenvalues ​​of the parameter cooperative matrix; The contribution weight coefficient is corrected based on the magnitude of the feature value to obtain the corrected weight coefficient. The weighted combination of the corrected weighting coefficient and the normalized value yields the weights of the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value on the desulfurization efficiency.

5. The method according to claim 1, characterized in that, Based on the aforementioned influence weights, the reliability scores for the flue gas flow rate, SO2 concentration, and slurry pH data are calculated, specifically including: Based on the influence weights, the time-synchronized flue gas flow rate, SO2 concentration, and slurry pH value, the weighted data values ​​are calculated. The dynamic mean of the weighted data values ​​over a preset time period is calculated using the sliding exponential smoothing method. Based on the dynamic mean, the dynamic variance of the weighted data values ​​is calculated using adaptive variance. Based on the degree of deviation between the weighted data values ​​and the dynamic mean, and in conjunction with the dynamic variance, the reliability scores of the flue gas flow rate, SO2 concentration, and slurry pH value data are calculated.

6. The method according to claim 1, characterized in that, Based on the highly reliable flue gas flow rate, SO2 concentration, and slurry pH data, the flow rate of the slurry circulation pump and the limestone slurry supply rate within the desulfurization unit are dynamically adjusted, specifically including: Calculate the real-time deviation and historical cumulative deviation between the highly reliable SO2 concentration and the preset SO2 emission standard value; Calculate the deviation weighted score based on the real-time deviation value and the historical cumulative deviation value; An adaptive adjustment coefficient is set based on the deviation-weighted score, and the adaptive adjustment coefficient increases as the deviation-weighted score increases; The adaptive adjustment coefficient, the high-reliability flue gas flow rate data, and the slurry pH value data are coupled and calculated using multiple parameters to obtain the flow rate adjustment increment and the slurry supply adjustment increment. Based on the flow rate adjustment increment and the slurry supply adjustment increment, the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization unit are calculated.

7. The method according to claim 6, characterized in that, Based on the aforementioned flow rate adjustment increment and slurry supply adjustment increment, the flow rate of the slurry circulation pump and the limestone slurry supply rate within the desulfurization unit are calculated, specifically including: Trend analysis was performed on the flow rate adjustment increment and slurry supply adjustment increment to obtain the changing trend characteristics of the adjustment parameters and the dynamic adjustment step size; Based on the dynamic adjustment step size, the flow rate adjustment increment, and the slurry supply adjustment increment, the stage adjustment target value and the adjustment execution sequence are generated. The target values ​​for the stage adjustment are achieved according to the adjustment execution sequence, thereby adjusting the flow rate of the slurry circulation pump and the limestone slurry supply rate in the desulfurization unit.

8. A server, characterized in that, The server includes: One or more processors and memory; The memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the server to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the server, the server causes the server to perform the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the server, the server performs the method as described in any one of claims 1-7.

Images