Boiler stable combustion regulation and control method based on multi-parameter fusion

Abstract

The invention relates to the technical field of programming auxiliary tools, and discloses a boiler stable combustion regulation and control method based on multi-parameter fusion, which comprises the following steps: monitoring coal quality data in real time, and generating a trigger signal when the fluctuation of the coal quality data meets a preset amplitude; when the trigger signal is monitored, first air door data are generated according to the coal quality data and the operation load; setting a sensor module to obtain flame shape data in real time; setting second air door data according to the flame shape data and the first air door data; setting a regulation and control strategy, and carrying out stable combustion regulation and control according to the second air door data. The flame shape data in the stable combustion regulation and control process are monitored in real time, the combustion stability of the boiler is improved, and flame extinguishing caused by stable combustion regulation and control is prevented.

Description

Technical Field

[0001] This application relates to the field of boiler combustion technology, and in particular to a boiler combustion stabilization control method based on multi-parameter fusion. Background Technology

[0002] In actual operation, coal-fired boilers often experience unpredictable fluctuations in the quality of the coal entering the furnace (such as changes in volatile matter, calorific value, and moisture content). These disturbances directly disrupt the stability of combustion within the furnace, leading to flame pulsation, uneven temperature distribution, and in severe cases, fire extinguishing accidents. Traditional boiler combustion control often relies on feedback regulation based on a few parameters such as outlet oxygen content and main steam pressure, which exhibits significant lag and struggles to respond promptly to sudden changes in coal quality.

[0003] In the existing technology, the stable combustion technology of coal-fired boilers has the following shortcomings: First, the relationship between coal quality monitoring and damper adjustment relies on a fixed empirical curve, which cannot adapt to complex dynamic processes; second, the judgment of flame state is relatively simple, and it is impossible to comprehensively obtain detailed flame data and evaluate flame quality based on flame shape data; third, during the adjustment process, excessive movements can easily introduce new disturbances that interfere with the flame shape, leading to unstable flame combustion. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a boiler combustion stabilization control method based on multi-parameter fusion, aiming to obtain a technical solution that can monitor flame quality during the combustion stabilization process through detailed analysis of flame shape, thereby improving combustion stability and boiler reliability.

[0005] In some embodiments of this application, a boiler combustion stabilization control method based on multi-parameter fusion is provided, characterized by comprising: Real-time monitoring of coal quality data; when the fluctuation of the coal quality data meets a preset range, a trigger signal is generated. When a trigger signal is detected, the first damper data is generated based on the coal quality data and operating load; Configure the sensor module to acquire flame shape data in real time; The second damper data is set based on the flame shape data and the first damper data; Set a control strategy and control the combustion stability based on the data from the second air damper.

[0006] In some embodiments of this application, when the coal quality data fluctuation meets a preset range, it includes: Extract time series data of various index parameters from coal quality data; Calculate the rate of change and the amount of change of the index parameter within the sliding time window; When the rate of change of an indicator parameter is continuously greater than the first threshold and the amount of change is greater than the second threshold, it is determined that the period of change has begun. Generate a trigger signal.

[0007] In some embodiments of this application, generating the first damper data includes: Pre-set coal load-air damper database; Obtain the boiler load data, and based on the coal load-damper database, query the damper opening corresponding to the load data and set it as the baseline damper opening; Acquire coal quality data and compare it with benchmark coal quality data, and set the offset data for coal quality parameters; The damper opening compensation value is set based on the offset data of the coal quality parameters; The first damper data is obtained by superimposing the reference damper opening degree with the damper opening compensation value.

[0008] In some embodiments of this application, the offset data for setting coal quality parameters includes: Extract benchmark index parameters from benchmark coal quality data; Calculate the difference between real-time index parameters and baseline index parameters in real-time coal quality data; Compare each difference with the amplitude data, and set the real-time indicator parameters whose differences are greater than the amplitude data as offset parameters; Set the differences between each offset parameter and its corresponding value as a structured offset data set; Associate each structured offset data set with the timestamp of the trigger signal.

[0009] In some embodiments of this application, obtaining flame shape data includes: The predicted flame shape data for the changing time period is set based on the first damper data; The moments in the time period during which the preset flame shape data does not meet the preset conditions are extracted and set as the first time library; Obtain flame image data and temperature data at various times from the first time database; Calculate the similarity feature values ​​of the flame images at each time point, and divide the flame images at each time point into multiple groups according to the similarity feature values; The flame image data with the highest similarity feature value from each group is selected as the keyframe image; Extract pixel information of the flame region from each keyframe image; Based on the pixel information, the flame shape feature parameters are calculated; The flame shape feature parameters are combined to form the flame shape data.

[0010] In some embodiments of this application, the calculation of flame shape characteristic parameters includes: Calculate the centroid position of the flame and set the centroid position as the center position of the image data; Extract flame pixel region data from pixel information; Calculate the distance data between each flame pixel region and the center position, and set the weight coefficient of each flame pixel region based on the distance data; The ratio of the area data of the flame pixel region to the image area data is set as the flame coverage. The image is converted to grayscale to obtain pixel grayscale values, and the flame intensity is set based on the pixel grayscale values.

[0011] In some embodiments of this application, calculating the flame image similarity feature values ​​at each time point includes: Extract the grayscale data of the image and set the center position of the flame intensity based on the grayscale data of each pixel; Extract the fill level data of each image and set the fill level center position; A weighted center position is set based on the center position of the flame intensity, the center position of the fullness, and the center position of the image data; Set the vector angle between the weighted center positions of each keyframe image, and set the vector angle as a similarity feature value.

[0012] In some embodiments of this application, the setting of the control strategy includes: Extract the adjustment amount of each secondary damper from the second damper data; The historical adjustment count of each secondary damper is recorded in the historical operation data. Based on the historical adjustment count and adjustment amount, the first adjustment ratio of each secondary air damper is set to obtain the first adjustment amount; Extract the upper limit of the single adjustment threshold of the secondary damper, and calculate the number of adjustments for each secondary damper based on the upper limit of the threshold and the first adjustment amount; Calculate the second adjustment amount based on the first adjustment amount and the adjustment amount; The secondary adjustment scheme is set according to the second adjustment amount.

[0013] In some embodiments of this application, the setting of the secondary adjustment scheme includes: Obtain the second adjustment amount and set the initial adjustment rate for each secondary damper; Set the adjustment time interval; After each adjustment time interval, the initial adjustment rate is attenuated to obtain the attenuation rate.

[0014] In some embodiments of this application, the correction to obtain the attenuation rate includes: Based on the magnitude of the second adjustment amount, a decay coefficient variation is set for each secondary damper; Obtain boiler negative pressure data; At the beginning of each adjustment time interval, the real-time fluctuation amplitude of the boiler negative pressure data and the real-time offset distance of the center position are calculated. When the real-time fluctuation amplitude or real-time offset distance exceeds the corresponding preset safety limit, the attenuation coefficient is corrected according to the change of the attenuation coefficient; if the real-time fluctuation amplitude and real-time offset distance do not exceed the corresponding preset safety limit, the attenuation coefficient remains unchanged. The adjustment rate at the end of the previous time interval is calculated by using the corrected attenuation coefficient to obtain the adjustment rate that should be executed in the current time interval.

[0015] Compared with existing technologies, the boiler combustion stabilization control method based on multi-parameter fusion proposed in this application has the following advantages: A system capable of quantifying flame combustion status was constructed. By deploying multi-view sensor modules, the system can extract multi-dimensional shape feature parameters such as centroid position, fill degree, and pulsation intensity from flame images in real time, ensuring the stability of flame shape and the reliability of boiler combustion. Simultaneously, the shape feature parameters can be used to further verify and adjust load data caused by changes in coal quality. This mechanism ensures that control commands not only respond to external disturbances but also directly affect the flame morphology itself. Through intelligent collaborative adjustment of secondary air dampers at each level, the flame is actively shaped and stabilized in an ideal state, significantly improving the stability, economy, and overall reliability of the combustion process and boiler operation. This achieves refined and visualized monitoring of combustion stability. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a boiler combustion stabilization control method based on multi-parameter fusion in an embodiment of this application. Detailed Implementation

[0017] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate this application, but are not intended to limit the scope of this application.

[0018] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0019] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0020] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] like Figure 1 As shown in the figure, an embodiment of this application provides a boiler combustion stabilization control method based on multi-parameter fusion, comprising: Real-time monitoring of coal quality data; when fluctuations in coal quality data meet a preset range, a trigger signal is generated. When a trigger signal is detected, the first damper data is generated based on the coal quality data and operating load; Configure the sensor module to acquire flame shape data in real time; Set the second damper data based on the flame shape data and the first damper data; Set a control strategy and control the combustion stability based on the data from the second air damper.

[0022] Specifically, an online coal quality analyzer is installed at the outlet of the coal mill to detect and acquire coal quality data, including volatile matter, calorific value, and final calorific value. Once a violent fluctuation in coal quality exceeding the allowable range is detected, a trigger signal is immediately generated.

[0023] Specifically, when a trigger signal is received, the system begins to perform calculations. Based on the current coal quality data and the boiler's operating load, it quickly calculates a set of preliminary secondary damper adjustment target values. This set of data is called the first damper data.

[0024] Specifically, the sensor modules installed on the boiler furnace mainly consist of high-temperature industrial cameras and infrared thermometers located in the observation holes and key parts of the furnace. These devices operate at high frequency to acquire images and temperature information of the flame in real time. By analyzing these images, quantitative data such as the shape, brightness, and position of the flame can be extracted, i.e., flame shape data.

[0025] Specifically, the flame shape data is compared with the first damper data. Based on the actual shape of the flame, the first damper data is intelligently corrected. For example, if the "first damper data" suggests closing the upper damper, but the flame image shows that the flame center is already too low, the correction algorithm will reduce the degree of closure or even open it wider. The more accurate and safer instruction obtained after correction is the "second damper data".

[0026] Specifically, the system does not directly send the "secondary damper data" to the dampers for execution. Instead, it sets a sophisticated "control strategy," specifying how much the dampers can move each time, how fast they can move, and how many steps they need to complete the action. Based on this strategy and the "secondary damper data," the control system issues a series of smooth and orderly commands to coordinate the actions of each secondary damper, ultimately completing the entire stable combustion control process.

[0027] In some embodiments of this application, when the fluctuation of coal quality data meets a preset range, the following includes: Extract time series data of various index parameters from coal quality data; Calculate the rate of change and the amount of change of the index parameters within the sliding time window; When the rate of change of an indicator parameter is continuously greater than the first threshold and the amount of change is greater than the second threshold, it is determined that the period of change has begun. Generate a trigger signal.

[0028] Specifically, the control system retrieves data points from the database for each indicator parameter of the flame, such as volatile matter and calorific value, arranged in chronological order within 10 minutes; sets a sliding time window (1 minute in length) and lets this window slide on the time series. At each window position, it calculates the rate of change (i.e., the value at the end of the window minus the value at the beginning, then divided by time) and the amount of change (i.e., the absolute value of the end value minus the beginning value) of each indicator parameter (such as volatile matter) within the window.

[0029] Specifically, a dual judgment condition is set: a first threshold and a second threshold. The program will make a judgment when one or more indicator parameters have a rate of change greater than the first threshold for two consecutive calculation periods, and the total change during this period also exceeds the second threshold. The preferred first threshold is a rate of change of 0.8% per minute; the preferred second threshold is an absolute value of 2.5% for the change.

[0030] In some embodiments of this application, generating first damper data includes: Pre-set coal load-air damper database; Obtain the boiler load data, and based on the coal load-damper database, query the damper opening corresponding to the load data and set it as the baseline damper opening; Acquire coal quality data and compare it with benchmark coal quality data, and set the offset data for coal quality parameters; Set the damper opening compensation value based on the offset data of coal quality parameters; The first damper data is obtained by superimposing the reference damper opening and the damper opening compensation value.

[0031] Specifically, a "coal load-damper database" is pre-established by collecting historical operating data. This database is the foundation of this invention. The database stores a large number of data records, each containing a specific boiler load and corresponding reference coal quality, along with recommended operating openings for each layer of secondary dampers (A, B, C, D, burnout air, etc.) to achieve optimal combustion. These openings are the reference damper openings. When a trigger signal is detected, the boiler load data is acquired, and the corresponding reference damper opening is obtained by indexing this load data in the database. Then, the current coal quality data is compared with the reference coal quality data in the database to quantify the difference, and the offset data is calculated. Based on this offset data, the reference and damper openings are corrected to obtain the first damper data.

[0032] In some embodiments of this application, offset data for coal quality parameters is set, including: Extract benchmark index parameters from benchmark coal quality data; Calculate the difference between real-time index parameters and baseline index parameters in real-time coal quality data; Compare each difference with the amplitude data, and set the real-time indicator parameters whose differences are greater than the amplitude data as offset parameters; Set the differences between each offset parameter and its corresponding value as a structured offset data set; Associate each structured offset data set with the timestamp of the trigger signal.

[0033] Specifically, the benchmark parameters include calorific value and volatile matter, where the calorific value is [calorific value] and the volatile matter is [volatile matter]. Specifically, the program acquires real-time data reported by the online coal quality analyzer, which consists of real-time indicator parameters. Then, it calculates the numerical difference between each real-time indicator parameter and its corresponding benchmark indicator parameter.

[0034] Specifically, amplitude data is obtained through a pre-defined amplitude data table within the system, which defines the permissible random fluctuation range for each coal quality indicator during normal operation. The program compares the differences calculated in the previous step with the corresponding amplitude data. If the difference of a parameter exceeds its set amplitude, the parameter is marked as an offset parameter, indicating an abnormal change. The offset parameters are then processed to obtain structured data, which clearly shows which coal quality characteristics have changed and by how much.

[0035] In some embodiments of this application, obtaining flame shape data includes: The predicted flame shape data for the changing time period is set based on the data from the first air damper; The moment when the preset flame shape data in the time period does not meet the preset conditions is set as the first time library; Obtain flame image data and temperature data at various times from the first time database; Calculate the similarity feature values ​​of the flame images at each time point, and divide the flame images at each time point into multiple groups according to the similarity feature values; The flame image data with the highest similarity feature value from each group was selected as the keyframe image; Extract pixel information of the flame region from each keyframe image; Based on pixel information, the flame shape feature parameters are calculated; Flame shape feature parameters are combined to form flame shape data.

[0036] Specifically, the theoretical changes in flame shape are predicted by generating the first air damper data. These flame shapes are then examined, and moments that do not meet the preset conditions are recorded to form a "first time database." These moments represent the time points when problems may occur during combustion.

[0037] Specifically, the preset conditions are set by statistically analyzing the flame position data and course temperature data under normal operating conditions. The model constructed from these data is set as the preset conditions, such as: the flame filling degree is greater than 70% and less than 90%, the combustion center should be between the height of the second secondary air damper and the height of the next secondary air damper in the furnace, and the temperature should reach above 1000 degrees Celsius.

[0038] Specifically, the first time database consists of various time points and corresponding flame image data and temperature distribution data.

[0039] Specifically, the image data with the highest similarity feature values ​​from each group is selected as the keyframe image. This involves dividing the image data according to different characteristics and selecting the image with the strongest features within each group as the representative image of that group, which is then set as the keyframe image. This keyframe image completely and comprehensively records all the features of that group. From these keyframe images, pixel information such as the contour and brightness of the flame region is extracted. Based on this pixel information, feature parameters that can quantitatively describe the flame morphology (such as centroid position and degree of fullness) are calculated. Finally, these feature parameters are combined and packaged to form the flame shape data.

[0040] In some embodiments of this application, flame shape characteristic parameters are calculated, including: Calculate the centroid position of the flame and set it as the center position of the image data; Extract flame pixel region data from pixel information; Calculate the distance data between each flame pixel region and the center position, and set the weight coefficient of each flame pixel region based on the distance data; The ratio of the flame pixel area data to the image area data is set as the flame coverage. The image is converted to grayscale to obtain pixel grayscale values, and the flame intensity is set based on the pixel grayscale values.

[0041] Specifically, the system reads the grayscale image of each keyframe image. The image size is 1024x768 pixels; higher temperatures result in higher brightness. It scans every pixel in the image, searching for all pixels with a grayscale value greater than 200. The system calculates the average coordinates of these flame pixels in the image plane; for example, if the average X-coordinate of all flame pixels is 512 pixels and the average Y-coordinate is 300 pixels (Y=0 at the top of the image), then the system sets this coordinate point (512, 300) as the centroid of the current flame. Simultaneously, the system records the center position of the image itself.

[0042] Specifically, the system calculates the straight-line distance from the center point of each flame region to the centroid of the entire flame (512, 300). For example, if there are three flame regions, region A is 50 pixels from the centroid, region B is 150 pixels, and region C is 10 pixels. The system assigns a weight coefficient to each region based on this distance data: the closer the distance, the larger the weight coefficient. Region C has a weight of 1.0, region A has a weight of 0.8, and region B has a weight of 0.5.

[0043] Specifically, flame coverage is calculated by counting all pixels identified as flames, resulting in the total number of pixels in the flame pixel region. Simultaneously, the total number of pixels in the effective observation area of ​​the image (i.e., the image area data) is obtained. Dividing the total number of pixels in the flame pixel region by the total number of pixels in the effective observation area yields the flame coverage. This value directly reflects the proportion of the flame within the observation field of view; a low coverage level may indicate a risk of flame dispersion or extinguishing.

[0044] Specifically, calculating flame intensity begins by converting the image to grayscale, transforming the RGB color value of each pixel into a single grayscale value. Then, statistical analysis is performed on the grayscale values ​​of all flame pixels. For example, the average and standard deviation of the grayscale values ​​for all flame pixels are calculated. The average grayscale value characterizes the overall brightness (intensity) of the flame; for instance, an average grayscale value of 230 indicates a very bright and intense flame. The standard deviation of the grayscale values ​​reflects the pulsation or stability of the flame; a larger standard deviation indicates greater differences in brightness across points, suggesting the flame may be flickering or unstable. The system combines these two values ​​and sets them together as a characteristic parameter characterizing flame intensity.

[0045] In some embodiments of this application, the calculation of flame image similarity feature values ​​at various times includes: Extract the grayscale data of the image and set the center position of the flame intensity based on the grayscale data of each pixel; Extract the fill level data of each image and set the fill level center position; A weighted center position is set based on the center position of flame intensity, the center position of fullness, and the center position of image data; Set the vector angle between the weighted center positions of each keyframe image, and set the vector angle as a similar feature value.

[0046] Specifically, determine the center position of the flame intensity and calculate the coordinates of the system's equilibrium point. For example, the flame is brighter on the right (higher grayscale value) and slightly darker on the left. After weighted calculation, the resulting equilibrium point may be located at (600, 350), and this point is called the center position of the flame intensity.

[0047] Specifically, determine the center position of the fire coverage, assign a weight of 1 to each flame pixel, and a weight of 0 to non-flame pixels. Calculate the geometric center of all flame pixels.

[0048] Specifically, the system takes three points from the image data—the center position (a fixed reference point, such as (512, 384)), the center position of flame intensity, and the center position of flame fill—and weights them according to a preset importance ratio. The image center accounts for 0.2 (emphasizing spatial reference), the intensity center accounts for 0.5 (emphasizing the energy core), and the fill center accounts for 0.3 (emphasizing the shape). Therefore, the weighted average coordinates, for example (550, 360), become a comprehensive weighted center position for the image. This point integrates spatial, energy, and shape information.

[0049] Specifically, the similarity feature value is calculated by determining the center position of each keyframe image and drawing two vectors from the origin of the corresponding image. The angle between these two vectors is then calculated. This angle is defined as the "similarity feature value" that measures the similarity of the flame shapes in the two images. The smaller the angle, the more similar the images are considered. Based on this value, the system can then divide all keyframe images into several groups with similar shapes.

[0050] In some embodiments of this application, a control strategy is set, including: Extract the adjustment amount of each secondary damper from the second damper data; The historical adjustment count of each secondary damper is recorded in the historical operation data. The first adjustment ratio of each secondary damper is set according to the historical adjustment number and adjustment amount to obtain the first adjustment amount; Extract the upper limit of the single adjustment threshold of the secondary damper, and calculate the number of adjustments for each secondary damper based on the upper limit of the threshold and the first adjustment amount; The second adjustment is calculated based on the first adjustment amount and the adjustment amount; The secondary adjustment scheme is set according to the second adjustment amount.

[0051] Specifically, the historical adjustment count is the total number of times each secondary damper performs adjustment actions within the past cycle (30 days).

[0052] Specifically, the difference between the current actual opening degree and the target opening degree is set as the total adjustment amount.

[0053] Specifically, the higher the total number of first adjustment ratio adjustment actions, the higher the importance, and thus the higher the first adjustment ratio.

[0054] Specifically, the upper limit of the single adjustment threshold is preferably 3 percent opening, which is the maximum value of the range that can be adjusted for this action.

[0055] In some embodiments of this application, a secondary adjustment scheme is set, including: Obtain the second adjustment amount and set the initial adjustment rate for each secondary damper; Set the adjustment time interval; After each adjustment time interval, the initial adjustment rate is adjusted by attenuation to obtain the attenuation rate.

[0056] Specifically, the adjustment time interval is the necessary time interval between each adjustment action to ensure that the result of this operation is observed and to prevent excessively large single adjustment amplitudes. A value of 10 seconds is preferred.

[0057] Specifically, the initial adjustment rate is set based on the second adjustment amount of each damper and the maximum safe mechanical speed allowed by the equipment, preferably 4% per minute.

[0058] In some embodiments of this application, the attenuation rate is modified, including: Based on the magnitude of the second adjustment amount, a decay coefficient variation is set for each secondary damper; Obtain boiler negative pressure data; At the beginning of each adjustment time interval, calculate the real-time fluctuation amplitude of the boiler negative pressure data and the real-time offset distance of the center position; When the real-time fluctuation amplitude or real-time offset distance exceeds the corresponding preset safety limit, the attenuation coefficient is corrected according to the change of the attenuation coefficient. If the real-time fluctuation amplitude and real-time offset distance do not exceed the corresponding preset safety limit, the attenuation coefficient remains unchanged. The adjustment rate at the end of the previous time interval is calculated by using the corrected attenuation coefficient to obtain the adjustment rate that should be executed in the current time interval.

[0059] Specifically, after each time interval, the system will not continue to adjust at the initial rate. Instead, it will "attenuate" the current adjustment rate based on the actual response of the boiler, reducing the rate to achieve a smooth adjustment.

[0060] Specifically, the attenuation coefficient variation is set as follows: Let the second adjustment amount of damper n be ΔDn, which is the difference between the target opening and the current opening. A baseline variation is set: Define a baseline variation Δλ0 (preferably 0.05) corresponding to a maximum adjustment amount ΔDmax. That is, when the adjustment amount reaches its maximum, the attenuation coefficient can be corrected by a maximum of 0.05. The linear variation is calculated: Calculate the basic attenuation coefficient variation Δλbn of the damper. Formula: Δλbn = Δλ0 * (|ΔDn| / ΔDmax).

[0061] Specifically, the system first acquires the data sequence. At the beginning of each adjustment time interval (Tc), the system reads the boiler negative pressure value sequence {Pt, P{t-1}, ..., P{t-N+1}} and the flame centroid height sequence {Yt, Y{t-1}, ..., Y{t-N+1}} from the database for the most recent N sampling periods (e.g., N=6, corresponding to the most recent 1 minute). Then, the system calculates the real-time fluctuation amplitude: the standard deviation of the negative pressure sequence is calculated as a measure of the fluctuation amplitude. Formula: ΔPi=([Σ(Pi-Pg)] / (N-1)), where Pg is the average value of the sequence, and i is the value from t to t-(N+1).

[0062] Specifically, the dynamic correction attenuation coefficient is applied when either the real-time fluctuation amplitude or the real-time offset distance exceeds its corresponding preset safety limit. This indicates a significant disturbance in the current adjustment speed. The attenuation coefficient is immediately increased as a correction. When neither the real-time fluctuation amplitude nor the real-time offset distance exceeds its preset safety limit, it indicates a smooth adjustment process without significant disturbance. The system will maintain the attenuation coefficient unchanged and continue according to the original attenuation rhythm.

[0063] Specifically, the fluctuation range is determined by first collecting data from 5 minutes ago to the current time, then collecting 10 data points every 6 seconds for a total of 50 sampling points, extracting the load data from these sampling points, calculating the standard deviation of the load data, and setting this standard deviation as the fluctuation range.

[0064] The above are merely preferred embodiments of this application. It should be noted that, for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of this application, and these improvements and substitutions should also be considered within the scope of protection of this application.

Claims

1. A boiler combustion stabilization control method based on multi-parameter fusion, characterized in that, include: Real-time monitoring of coal quality data; when the fluctuation of the coal quality data meets a preset range, a trigger signal is generated. When a trigger signal is detected, the first damper data is generated based on the coal quality data and operating load; Configure the sensor module to acquire flame shape data in real time; The second damper data is set based on the flame shape data and the first damper data; Set a control strategy and control the combustion stability based on the data from the second air damper.

2. The boiler combustion stability control method based on multi-parameter fusion as described in claim 1, characterized in that, When the fluctuation of the coal quality data meets the preset range, it includes: Extract time series data of various index parameters from coal quality data; Calculate the rate of change and the amount of change of the index parameter within the sliding time window; When the rate of change of an indicator parameter is continuously greater than the first threshold and the amount of change is greater than the second threshold, it is determined that the period of change has begun. Generate a trigger signal.

3. The boiler combustion stability control method based on multi-parameter fusion as described in claim 2, characterized in that, The generation of the first damper data includes: Pre-set coal load-air damper database; Obtain the boiler load data, and based on the coal load-damper database, query the damper opening corresponding to the load data and set it as the baseline damper opening; Acquire coal quality data and compare it with benchmark coal quality data, and set the offset data for coal quality parameters; The damper opening compensation value is set based on the offset data of the coal quality parameters; The first damper data is obtained by superimposing the reference damper opening degree with the damper opening compensation value.

4. The boiler combustion stability control method based on multi-parameter fusion as described in claim 3, characterized in that, The offset data for setting coal quality parameters includes: Extract benchmark index parameters from benchmark coal quality data; Calculate the difference between real-time index parameters and baseline index parameters in real-time coal quality data; Compare each difference with the amplitude data, and set the real-time indicator parameters whose differences are greater than the amplitude data as offset parameters; Set the differences between each offset parameter and its corresponding value as a structured offset data set; Associate each structured offset data set with the timestamp of the trigger signal.

5. The boiler combustion stability control method based on multi-parameter fusion as described in claim 1, characterized in that, The acquisition of flame shape data includes: The predicted flame shape data for the changing time period is set based on the first damper data; The moments in the time period during which the preset flame shape data does not meet the preset conditions are extracted and set as the first time library; Obtain flame image data and temperature data at various times from the first time database; Calculate the similarity feature values ​​of the flame images at each time point, and divide the flame images at each time point into multiple groups according to the similarity feature values; The flame image data with the highest similarity feature value from each group is selected as the keyframe image; Extract pixel information of the flame region from each keyframe image; Based on the pixel information, the flame shape feature parameters are calculated; The flame shape feature parameters are combined to form the flame shape data.

6. The boiler combustion stabilization control method based on multi-parameter fusion as described in claim 5, characterized in that, The calculated flame shape characteristic parameters include: Calculate the centroid position of the flame and set the centroid position as the center position of the image data; Extract flame pixel region data from pixel information; Calculate the distance data between each flame pixel region and the center position, and set the weight coefficient of each flame pixel region based on the distance data; The ratio of the area data of the flame pixel region to the image area data is set as the flame coverage. The image is converted to grayscale to obtain pixel grayscale values, and the flame intensity is set based on the pixel grayscale values.

7. The boiler combustion stabilization control method based on multi-parameter fusion as described in claim 6, characterized in that, The calculation of similarity feature values ​​of flame images at each time point includes: Extract the grayscale data of the image and set the center position of the flame intensity based on the grayscale data of each pixel; Extract the fill level data of each image and set the fill level center position; A weighted center position is set based on the center position of the flame intensity, the center position of the fullness, and the center position of the image data; Set the vector angle between the weighted center positions of each keyframe image, and set the vector angle as a similarity feature value.

8. The boiler combustion stabilization control method based on multi-parameter fusion as described in claim 1, characterized in that, The set control strategy includes: Extract the adjustment amount of each secondary damper from the second damper data; The historical adjustment count of each secondary damper is recorded in the historical operation data. Based on the historical adjustment count and adjustment amount, the first adjustment ratio of each secondary air damper is set to obtain the first adjustment amount; Extract the upper limit of the single adjustment threshold of the secondary damper, and calculate the number of adjustments for each secondary damper based on the upper limit of the threshold and the first adjustment amount; Calculate the second adjustment amount based on the first adjustment amount and the adjustment amount; The secondary adjustment scheme is set according to the second adjustment amount.

9. The boiler combustion stabilization control method based on multi-parameter fusion as described in claim 8, characterized in that, The set secondary adjustment scheme includes: Obtain the second adjustment amount and set the initial adjustment rate for each secondary damper; Set the adjustment time interval; After each adjustment time interval, the initial adjustment rate is attenuated to obtain the attenuation rate.

10. The boiler combustion stabilization control method based on multi-parameter fusion as described in claim 9, characterized in that, The correction to obtain the attenuation rate includes: Based on the magnitude of the second adjustment amount, a decay coefficient variation is set for each secondary damper; Obtain boiler negative pressure data; At the beginning of each adjustment time interval, the real-time fluctuation amplitude of the boiler negative pressure data and the real-time offset distance of the center position are calculated. When the real-time fluctuation amplitude or real-time offset distance exceeds the corresponding preset safety limit, the attenuation coefficient is corrected according to the change of the attenuation coefficient; if the real-time fluctuation amplitude and real-time offset distance do not exceed the corresponding preset safety limit, the attenuation coefficient remains unchanged. The adjustment rate at the end of the previous time interval is calculated by using the corrected attenuation coefficient to obtain the adjustment rate that should be executed in the current time interval.

Images