Method and system for predicting combustion efficiency of water jacket heater

Abstract

The invention discloses a combustion efficiency prediction method and system for a water jacket heater, and particularly relates to the technical field of combustion efficiency prediction. The method comprises the following steps: acquiring a thermal image sequence of the furnace wall surface of a water jacket heater, extracting pixel temperature gradient and texture change characteristics, and identifying a furnace wall heat transfer abnormal area; performing wavelet decomposition on the temperature gradient characteristic of the furnace wall heat transfer abnormal area, and calculating the total energy of a high-frequency signal; on the basis of the correlation with the temperature fluctuation amplitude, the rule of scale formation and falling is extracted; evaluating the heat transfer efficiency fluctuation risk caused by the scale change by using a dynamic Bayesian network, and outputting a heat transfer fluctuation risk probability; constructing a coupling model between the hearth combustion process and the heat transfer fluctuation risk probability, and correcting a heat loss factor; and according to the corrected heat loss factor, the real-time prediction value of the combustion efficiency is output, and the operation energy efficiency management level of the water jacket heater can be effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of combustion efficiency prediction, and more particularly to a water jacket heating furnace combustion efficiency prediction method and system. BACKGROUND

[0002] The combustion efficiency evaluation of the existing water jacket heating furnace mainly depends on the steady-state heat transfer model or the empirical heat loss correction coefficient, and it is difficult to accurately reflect the non-steady-state heat transfer fluctuation problem caused by the dynamic change of the water scale on the furnace wall in the actual operation.

[0003] Under the working condition of frequent water quality change and load fluctuation, the formation and shedding speed of the water scale on the furnace wall surface are accelerated, which causes the instantaneous heat transfer performance of the furnace wall to fluctuate sharply, resulting in input distortion of the combustion efficiency calculation model, and causing the prediction result of the combustion efficiency to deviate from the actual working condition. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a water jacket heating furnace combustion efficiency prediction method and system to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: A water jacket heating furnace combustion efficiency prediction method, comprising the following steps: S1: collecting a thermal image sequence of the furnace wall surface of the water jacket heating furnace, extracting the temperature gradient and texture change characteristics of the pixels in the heat radiation region of the furnace wall, and identifying the abnormal heat transfer region of the furnace wall; S2: wavelet decomposing the temperature gradient characteristics of the abnormal heat transfer region of the furnace wall, obtaining the high-frequency signal component reflecting the water scale adhesion and shedding process, and calculating the high-frequency signal energy; S3: analyzing the correlation between the high-frequency signal energy and the temperature fluctuation amplitude of the furnace wall, and extracting the rule of water scale formation and shedding on the furnace wall; S4: based on the rule of water scale formation and shedding, using a dynamic Bayesian network to evaluate the heat transfer efficiency fluctuation risk of the furnace wall caused by the water scale change, and outputting the heat transfer fluctuation risk probability; S5: constructing a coupling model between the combustion process of the furnace and the heat transfer fluctuation risk probability, and correcting the heat loss factor in the combustion efficiency calculation; S6: updating the combustion efficiency prediction model parameters in real time according to the corrected heat loss factor, and outputting the real-time prediction value of the combustion efficiency of the water jacket heating furnace.

[0006] In a preferred embodiment, S1 specifically comprises: collecting a thermal image sequence of the furnace wall surface of the water jacket heating furnace, delineating the boundary position of the heat radiation region of the furnace wall, and determining the temperature value of each pixel point in the heat radiation region of the furnace wall; The temperature difference between any adjacent pixel points in the thermal radiation region of the furnace wall is calculated by using a spatial gradient calculation method to obtain the temperature gradient value of each pixel position; The energy, contrast and correlation indexes are extracted by using a gray level co-occurrence matrix method for each frame of thermal image in the thermal radiation region of the furnace wall to determine the texture change characteristics; The determination threshold is set according to the temperature gradient value and the texture change characteristics to identify the abnormal heat transfer region of the furnace wall.

[0007] In a preferred embodiment, S2, specifically: The temperature gradient value of each pixel position in the abnormal heat transfer region of the furnace wall is decomposed based on a multi-scale wavelet transform method to obtain the high-frequency wavelet coefficient and the low-frequency wavelet coefficient corresponding to the temperature gradient value; The high-frequency wavelet coefficient is corresponded to each pixel point in the abnormal heat transfer region of the furnace wall according to the spatial position to obtain the high-frequency signal component distribution of each pixel point; The high-frequency signal component distribution of each pixel point in the abnormal heat transfer region of the furnace wall is squared calculated, and the high-frequency signal component values of each pixel point after the square calculation are added point by point to determine the total energy value of the high-frequency signal component in the abnormal heat transfer region of the furnace wall.

[0008] In a preferred embodiment, S3, specifically: The total energy value of the high-frequency signal component in the abnormal heat transfer region of the furnace wall is taken as the abscissa, and the temperature fluctuation amplitude value of each pixel point in the thermal radiation region of the furnace wall is taken as the ordinate to construct a scatter plot between the total energy of the high-frequency signal component and the temperature fluctuation amplitude; The scatter plot is fitted and calculated by using a regression analysis method to obtain a function expression of the corresponding relationship between the total energy of the high-frequency signal component and the temperature fluctuation amplitude; According to the trend of the total energy of the high-frequency signal component changing with time and the function expression, the periodic law and change characteristics of the scale formation and shedding in the thermal radiation region of the furnace wall are determined.

[0009] In a preferred embodiment, S4, specifically: Based on the periodic law and change characteristics of the scale formation and shedding in the thermal radiation region of the furnace wall, a dynamic Bayesian network containing hidden state nodes and observation nodes is constructed; The scale adhesion period and the shedding period are taken as the hidden state nodes, and the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value are taken as the observation nodes, and the probability value of the furnace wall heat transfer efficiency fluctuation is determined by calculating the state transition probability and the observation probability in the dynamic Bayesian network.

[0010] In a preferred embodiment, S5, specifically: According to a calculation formula of heat loss in the furnace combustion process, a function relationship between the furnace combustion process and the probability value of the fluctuation of the furnace wall heat transfer efficiency is established; According to the probability value of the fluctuation of the furnace wall heat transfer efficiency, a heat loss factor in the calculation formula of the heat loss in the furnace combustion process is adjusted to determine a corrected value of the heat loss factor; Based on the corrected value of the heat loss factor, a corrected heat loss factor is determined.

[0011] In a preferred embodiment, S6, specifically: Based on the corrected heat loss factor, a correction amount of a to-be-corrected combustion efficiency prediction model parameter in the combustion efficiency prediction model is calculated; The correction amount of the combustion efficiency prediction model parameter is added to an initial value of the combustion efficiency prediction model parameter to obtain a real-time updated combustion efficiency prediction model parameter; According to the real-time updated combustion efficiency prediction model parameter, the temperature fluctuation amplitude value of each pixel point in the furnace wall heat radiation area of the water jacket heating furnace, the total energy value of the high-frequency signal component, and the probability value of the fluctuation of the furnace wall heat transfer efficiency are input into the combustion efficiency prediction model to output a real-time prediction value of the combustion efficiency of the water jacket heating furnace.

[0012] On the other hand, the present application provides a water jacket heating furnace combustion efficiency prediction system, comprising: A feature extraction module: a thermal image sequence of a water jacket heating furnace wall surface is collected, pixel temperature gradient and texture change features in a furnace wall heat radiation area are extracted, and a furnace wall heat transfer abnormal area is identified; A wavelet decomposition module: the temperature gradient features of the furnace wall heat transfer abnormal area are subjected to wavelet decomposition to obtain a high-frequency signal component reflecting the process of scale deposition and shedding, and the high-frequency signal energy is calculated; A rule identification module: by analyzing the correlation between the high-frequency signal energy and the temperature fluctuation amplitude of the furnace wall, the rules of scale formation and shedding on the furnace wall are extracted; A risk assessment module: based on the rules of scale formation and shedding, the risk of the fluctuation of the furnace wall heat transfer efficiency caused by the scale change is evaluated by using a dynamic Bayesian network, and a heat transfer fluctuation risk probability is output; A factor correction module: a coupling model between the furnace combustion process and the heat transfer fluctuation risk probability is constructed, and a heat loss factor in the calculation of the combustion efficiency is corrected; An efficiency prediction module: according to the corrected heat loss factor, the combustion efficiency prediction model parameter is real-time updated, and a real-time prediction value of the combustion efficiency of the water jacket heating furnace is output.

[0013] The technical effects and advantages of the water jacket heating furnace combustion efficiency prediction method and system of the present application are as follows: Through thermal image sequence acquisition and image feature extraction, the abnormal area of the furnace wall surface heat transfer is accurately identified, and the identification accuracy of the scale affected part is improved; the wavelet decomposition method is introduced to extract the high frequency signal component and calculate the high frequency signal energy, which effectively describes the dynamic characteristics in the process of scale adhesion and shedding; by analyzing the correlation between the high frequency signal energy and the temperature fluctuation amplitude, the periodic change rule of the scale is accurately extracted; the dynamic Bayesian network realizes the probabilistic evaluation of the furnace wall heat transfer efficiency fluctuation risk; the coupling model between the furnace combustion process and the heat transfer risk is constructed and the heat loss factor is corrected, which improves the dynamic response ability and stability of the combustion efficiency calculation; by updating the combustion efficiency prediction model parameters in real time, the real-time prediction value of the combustion efficiency is output, which enhances the accurate perception and dynamic regulation ability of the water jacket heating furnace operation energy efficiency, and improves the accuracy of the thermal efficiency evaluation under complex working conditions. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 FIG. 1 is a schematic diagram of a water jacket heating furnace combustion efficiency prediction method according to the present application; Figure 2 FIG. 2 is a structural schematic diagram of a water jacket heating furnace combustion efficiency prediction system according to the present application. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0016] Embodiment 1

[0017] Figure 1 A water jacket heating furnace combustion efficiency prediction method according to the present application is given, which includes the following steps: S1: Collecting thermal image sequence of the furnace wall surface of the water jacket heating furnace, extracting pixel temperature gradient and texture change feature in the furnace wall heat radiation area, and identifying the abnormal area of the furnace wall heat transfer; S2: Wavelet decomposition is performed on the temperature gradient feature of the abnormal area of the furnace wall heat transfer, the high frequency signal component reflecting the process of scale adhesion and shedding is obtained, and the high frequency signal energy is calculated; S3: By analyzing the correlation between the high frequency signal energy and the temperature fluctuation amplitude of the furnace wall, the rule of scale formation and shedding on the furnace wall is extracted; S4: Based on the rule of scale formation and shedding, the dynamic Bayesian network is used to evaluate the furnace wall heat transfer efficiency fluctuation risk caused by the change of scale, and the heat transfer fluctuation risk probability is output; S5: build a coupling model between the furnace combustion process and the heat transfer fluctuation risk probability, and correct the heat loss factor in the calculation of the combustion efficiency; S6: according to the corrected heat loss factor, real-time update the combustion efficiency prediction model parameters, output the real-time prediction value of the combustion efficiency of the water jacket heating furnace.

[0018] S1: collect the thermal image sequence of the water jacket heating furnace wall surface, extract the pixel temperature gradient and texture change characteristics in the furnace wall heat radiation area, identify the abnormal heat transfer area of the furnace wall, including: Collect the thermal image sequence of the water jacket heating furnace wall surface, demarcate the boundary position of the furnace wall heat radiation area, and determine the temperature value of each pixel point in the furnace wall heat radiation area; The infrared thermal imaging device is used to continuously shoot the surface of the water jacket heating furnace wall, and the time sequence thermal image is formed. The thermal image sequence needs to cover the whole furnace wall to ensure that the thermal radiation characteristics of the furnace wall under different working conditions are recorded comprehensively. For example, the thermal image of the furnace wall surface is continuously shot at a predetermined time interval, and the image data corresponding to multiple time points is obtained. The image data obtained by continuous shooting is the thermal image sequence of the furnace wall surface.

[0019] The boundary position of the furnace wall heat radiation area is demarcated for the thermal image sequence. The demarcation method of the boundary position of the furnace wall heat radiation area is as follows: first, select the area in the thermal image which is not affected by the heating of the furnace wall as the background area, calculate the average value of the temperature values of all pixel points in the background area, and obtain the average temperature value of the background area; compare the temperature value of each pixel point in the furnace wall area with the average temperature value of the background area, calculate the ratio of the temperature value of each pixel point to the average temperature value of the background area, and mark all the pixel points whose ratio is greater than a certain proportion of the average temperature value of the background area. The outermost line of all marked pixel points is taken as the boundary position of the furnace wall heat radiation area, and the certain proportion can be set to be between ten percent and twenty percent of the average temperature value of the background area.

[0020] After the boundary position of the furnace wall thermal radiation area is determined, the temperature value corresponding to each pixel point in the furnace wall thermal radiation area is determined one by one. Specifically, the infrared radiation intensity value corresponding to each pixel point in the furnace wall thermal radiation area is extracted from each frame of thermal image, and the extraction method is to read the intensity information of the infrared radiation data recorded by the pixel point by using the image data processing software; the infrared radiation intensity value extracted from each pixel point is converted into the corresponding actual temperature value through the calibration relationship of the infrared thermal imaging device itself, and the calibration relationship adopts the standard curve relationship calibrated by the infrared thermal imaging device at that time, that is, the result after the infrared radiation intensity value is modified by the specific calibration coefficient in the instrument, the environmental temperature compensation coefficient and the radiation correction factor is the actual temperature value of the pixel point. Repeat the above method to obtain the accurate actual temperature value of each pixel point in the furnace wall thermal radiation area.

[0021] The temperature difference between any adjacent pixel points in the furnace wall thermal radiation area is calculated by using the spatial gradient calculation method, and the temperature gradient value of each pixel position is obtained. Based on the temperature values of each pixel point in the furnace wall thermal radiation area, the temperature values of the pixel points adjacent to the pixel points in any position in the furnace wall thermal radiation area are determined; the absolute value of the temperature value difference between the pixel points and the adjacent pixel points is calculated respectively, and the absolute values of the calculated temperature differences are accumulated and then divided by the number of adjacent pixel points participating in the calculation to obtain the temperature gradient value corresponding to the pixel point. For example, the absolute values of the temperature differences between a certain pixel point and its adjacent pixel points in the four directions of up, down, left and right are summed, and then divided by the total number of adjacent pixel points participating in the difference calculation to obtain the spatial temperature gradient value of the pixel position. Through the above process, the temperature gradient values of all pixel positions in the furnace wall thermal radiation area are obtained.

[0022] The energy, contrast and correlation indicators are extracted by using the gray level co-occurrence matrix method for each frame of thermal image in the furnace wall thermal radiation area to determine the texture change characteristics. Each frame of thermal image is converted into a gray scale image form, specifically: the temperature value of each pixel point in the furnace wall thermal radiation area is taken as the gray scale intensity value. Based on the gray scale image, a gray level co-occurrence matrix is constructed. The construction method is: in the furnace wall thermal radiation area, taking each pixel point as a reference, the frequency of simultaneous occurrence of the gray scale intensity value of each pixel point and the gray scale intensity values of the right adjacent pixel position, the lower adjacent pixel position and the right lower diagonal adjacent pixel position is counted, and the counting result is recorded in the gray level co-occurrence matrix. The row and column of the gray level co-occurrence matrix respectively represent the gray scale intensity values of two adjacent pixel points, and each element represents the occurrence frequency of the gray scale value combination.

[0023] Based on the gray level co-occurrence matrix, the energy, contrast and correlation are calculated respectively. The calculation method of the energy index is that the square of all element values in the gray level co-occurrence matrix is calculated, and then the square values are added one by one to obtain the energy index value; the calculation method of the contrast index is that for each element in the gray level co-occurrence matrix, the difference value of the gray intensity value position corresponding to the element is calculated, that is, the difference between the corresponding row number and column number in the gray level co-occurrence matrix, and then the square of the difference value is multiplied by the element value, and the values calculated from all elements in the gray level co-occurrence matrix are added to obtain the contrast index value; the calculation method of the correlation index is that the average values of the gray intensity values in the row direction and the column direction of the gray level co-occurrence matrix are calculated respectively, for each element in the gray level co-occurrence matrix, the difference between the gray intensity value at the position of the element and the average value calculated above is calculated, and then the difference value is multiplied by the element value, and the sum of the values calculated from all elements in the gray level co-occurrence matrix is normalized to obtain the correlation index value.

[0024] The determination threshold is set according to the temperature gradient value and the texture change feature, and the abnormal heat transfer area of the furnace wall is identified. The average value of the temperature gradient value distribution of all pixel positions in the heat radiation area of the furnace wall is calculated, and the determination threshold of the temperature gradient value is set as one hundred and twenty percent of the average value; the average values of the energy index, the contrast index and the correlation index in the texture change feature are calculated, and the arithmetic average values of the energy index, the contrast index and the correlation index are multiplied by the proportion value of one hundred and twenty percent as the determination thresholds of the corresponding texture change features.

[0025] The determination is performed at each pixel position in the heat radiation area of the furnace wall, first, whether the temperature gradient value corresponding to the pixel position is greater than the determination threshold of the temperature gradient value is determined; then, whether at least one of the texture change feature indexes, that is, the energy index, the contrast index and the correlation index corresponding to the pixel position is greater than the determination threshold of the corresponding texture change feature is determined; if both of the above two conditions are met, that is, the temperature gradient value of the pixel position is greater than the determination threshold of the temperature gradient value, and any one of the three texture change feature indexes exceeds the corresponding determination threshold, the pixel position is marked as the abnormal heat transfer area of the furnace wall. The identification of the abnormal heat transfer area of the furnace wall is completed by the above method.

[0026] S2: Wavelet decomposition is performed on the temperature gradient feature of the abnormal heat transfer area of the furnace wall to obtain high-frequency signal components reflecting the process of scale deposition and shedding, and the energy of the high-frequency signal is calculated, including: Based on the multi-scale wavelet transform method, the temperature gradient value of each pixel position in the abnormal heat transfer area of the furnace wall is decomposed to obtain the high-frequency wavelet coefficient and the low-frequency wavelet coefficient corresponding to the temperature gradient value; The temperature gradient value corresponding to each pixel position in the abnormal heat transfer region of the furnace wall is selected to form a one-dimensional spatial distribution sequence of the temperature gradient value, i.e., a temperature gradient value sequence; the temperature gradient value sequence is decomposed in multiple levels step by step by using a multi-scale wavelet transform method; the wavelet transform method uses a predetermined wavelet basis function, such as an orthogonal wavelet basis function, to decompose the temperature gradient value sequence layer by layer according to the spatial scale from large to small; each layer of decomposition obtains high-frequency wavelet coefficients and low-frequency wavelet coefficients through convolution processing of the sequence; the high-frequency wavelet coefficients are used to represent the local characteristics of rapid changes in the temperature gradient value sequence, and the low-frequency wavelet coefficients represent the overall changes of the slow trend in the temperature gradient value sequence; the wavelet transform calculation method is that the temperature gradient value corresponding to the position of each pixel point is multiplied point by point with the wavelet basis function, and then the results are summed point by point to obtain the high-frequency wavelet coefficients; the temperature gradient value is multiplied point by point with the scale function, and then summed to obtain the low-frequency wavelet coefficients; the above calculation is repeated to complete the decomposition of all spatial scale levels, and finally all high-frequency wavelet coefficients and low-frequency wavelet coefficients of each pixel position in the abnormal heat transfer region of the furnace wall at multiple scales are obtained.

[0027] The high-frequency wavelet coefficients are corresponded to each pixel point of the abnormal heat transfer region of the furnace wall according to the spatial position to obtain the high-frequency signal component distribution of each pixel point. The high-frequency wavelet coefficients are corresponded to each pixel point of the abnormal heat transfer region of the furnace wall according to the spatial position to obtain the high-frequency signal component distribution of each pixel point. Specifically, the high-frequency wavelet coefficients are corresponded to the spatial position information of the abnormal heat transfer region of the furnace wall one by one; the corresponding method is that the value of the corresponding high-frequency wavelet coefficient of each pixel position is determined one by one according to the spatial coordinate position of each pixel position in the image of the abnormal heat transfer region of the furnace wall; for example, each pixel position corresponds to a group of high-frequency wavelet coefficients in the multi-scale wavelet transform process, and the amplitude of the group of high-frequency wavelet coefficients is used as the intensity of the high-frequency signal component of the pixel point; the amplitude of the high-frequency wavelet coefficient of each pixel position is marked at the corresponding pixel position one by one to complete the spatial position mapping of the high-frequency signal component distribution of all pixel points in the abnormal heat transfer region of the furnace wall, and the high-frequency signal component distribution of each pixel point is obtained.

[0028] The high-frequency signal component distribution of each pixel point in the abnormal heat transfer region of the furnace wall is squared, and then the high-frequency signal component values of each pixel point after the square calculation are added point by point to determine the total energy value of the high-frequency signal component in the abnormal heat transfer region of the furnace wall; The high frequency signal component distribution of each pixel point in the abnormal heat transfer area of the furnace wall is obtained, the high frequency wavelet coefficient amplitude corresponding to each pixel point is squared, that is, the value of each high frequency signal component is multiplied by the value of each high frequency signal component, and the energy value of the high frequency signal component at each pixel position is obtained; the energy values of the high frequency signal components at all pixel positions in the abnormal heat transfer area of the furnace wall obtained after the square operation are added point by point; the final value obtained after the addition is the total energy value of the high frequency signal components in the abnormal heat transfer area of the furnace wall; for example, the high frequency wavelet coefficient amplitude of each pixel position is multiplied by itself to obtain the high frequency signal energy value of each pixel position, and then the addition operation is performed one by one, and the total energy value of the high frequency signal components on the whole abnormal heat transfer area of the furnace wall is finally obtained after the calculation of all pixel positions in the abnormal heat transfer area of the furnace wall is completed.

[0029] S3: By analyzing the correlation between the high frequency signal energy and the temperature fluctuation amplitude of the furnace wall, the rules of scale formation and scale shedding on the furnace wall are extracted, including: The total energy of the high frequency signal components in the abnormal heat transfer area of the furnace wall is taken as the abscissa, and the temperature fluctuation amplitude value of each pixel point in the heat radiation area of the furnace wall is taken as the ordinate, and a scatter plot between the total energy of the high frequency signal components and the temperature fluctuation amplitude is constructed. The total energy value of the high frequency signal components in the abnormal heat transfer area of the furnace wall is used to determine the abscissa value range in the scatter plot. For each pixel position in the heat radiation area of the furnace wall, the actual temperature values of each pixel position at continuous time points are determined, and the temperature fluctuation amplitude value of each pixel position is determined by calculating the difference between the maximum and minimum values of the temperature values of each pixel position at continuous time points. The calculation method of the temperature fluctuation amplitude is as follows: compare the temperature values of each pixel position at multiple continuous time points, select the maximum temperature value and the minimum temperature value, and then subtract the minimum temperature value from the maximum temperature value, and the result is the temperature fluctuation amplitude value of the pixel position. Repeat the above steps for all pixel positions in the heat radiation area of the furnace wall to obtain the complete temperature fluctuation amplitude value as the ordinate value range of the scatter plot. The total energy value of the high frequency signal components in the abnormal heat transfer area of the furnace wall and the temperature fluctuation amplitude value of each pixel position in the heat radiation area of the furnace wall are taken as the abscissa and ordinate of the scatter plot, and a scatter plot between the total energy of the high frequency signal components and the temperature fluctuation amplitude is constructed.

[0030] A regression analysis method is used to fit and calculate the scatter plot to obtain a function expression of the corresponding relationship between the total energy of the high frequency signal components and the temperature fluctuation amplitude; According to the data change trend reflected by the scatter diagram, a suitable function form is determined, for example, when the data points in the scatter diagram present an approximate linear relationship, a function with a linear form is selected as the basic function type; when the data points present a nonlinear trend, a polynomial function can be selected as the basic function type; if the data points present a significant rapid growth or rapid decay trend, an exponential function can be selected as the basic function type.

[0031] After the basic function type is determined, function fitting calculation is performed by using the least square method, specifically: first, the number of parameters contained in the function and the initial value of each parameter are determined; for each data point in the scatter diagram, the values of the data point in the horizontal coordinate and the vertical coordinate are determined respectively, and the horizontal coordinate value is substituted into the function expression to calculate the corresponding theoretical prediction value; the theoretical prediction value is subtracted from the actual value of the vertical coordinate of the data point to obtain the error value of the data point; then the error value is squared to obtain the error square value corresponding to the data point; the error square values of all data points in the scatter diagram are calculated one by one, and all the error square values are added one by one to obtain the total sum of the error square values; by gradually adjusting the values of the parameters in the function, the total sum of the error square values is repeatedly calculated until the optimal values of the parameters are obtained when the total sum of the error square values reaches the minimum.

[0032] After obtaining the optimal values of the parameters, the function expression is obtained by substituting the determined basic function type; the function expression describes the corresponding relationship between the total energy value of the high-frequency signal component in the abnormal heat transfer area of the furnace wall and the temperature fluctuation amplitude value of each pixel point in the thermal radiation area of the furnace wall.

[0033] For example, if a function with a linear form is selected as the basic function type, the parameters of the function include the proportional coefficient and the constant term of the linear function. By least square fitting, the proportional coefficient and the constant term values that make the scatter diagram data error square sum reach the minimum are determined.

[0034] According to the trend of the total energy of the high-frequency signal component changing with time and the function expression, the periodicity and variation characteristics of the scale formation and shedding in the thermal radiation area of the furnace wall are determined; The total energy values of the high-frequency signal component at multiple continuous time points in the abnormal heat transfer area of the furnace wall are continuously collected, and the interval between the multiple continuous time points can be set to several minutes to several tens of minutes, for example, the interval is set to five minutes, and the total energy values of the high-frequency signal component are continuously collected for several hours; The total energy values of the high-frequency signal component at the multiple continuous time points are plotted into a trend graph with time as the horizontal coordinate and the total energy value as the vertical coordinate; whether the total energy of the high-frequency signal component presents a significant periodic change trend is analyzed through the trend graph, and the judgment standard of the periodic trend is: If the total energy value of the high-frequency signal component shows a rising trend, followed by a falling trend, and the rising and falling trends continue to appear repeatedly, it is determined to have a periodic change characteristic; the duration of the periodic change characteristic is measured, i.e. the interval time from the beginning of the significant rise of the total energy value of the high-frequency signal component to the next significant rise is determined as the cycle length of the scale formation and shedding; The total energy value of the high-frequency signal component at each time is substituted into the functional expression between the total energy of the high-frequency signal component and the temperature fluctuation amplitude to calculate the temperature fluctuation amplitude value at the corresponding time in the heat radiation region of the furnace wall; The calculated temperature fluctuation amplitude value changes over time is plotted into a corresponding trend graph, which is compared and analyzed with the periodic change trend graph of the total energy of the high-frequency signal component; it is determined whether the temperature fluctuation amplitude at the corresponding time also significantly increases when the total energy value of the high-frequency signal component reaches a higher level; whether the temperature fluctuation amplitude at the corresponding time also significantly decreases when the total energy value of the high-frequency signal component decreases to a lower level; Through the comparison and analysis of the above two trend graphs for a plurality of continuous cycles, it is determined that the high-frequency signal energy increases accompanied by the increase of the temperature fluctuation amplitude during the scale adhesion process, and the high-frequency signal energy decreases accompanied by the decrease of the temperature fluctuation amplitude during the scale shedding process; The cycle length of the scale formation and shedding on the surface of the furnace wall is determined through repeated observation and analysis, for example, several hours or several days, and the maximum and minimum values of the temperature fluctuation amplitude in each cycle are recorded in detail, and then the regular characteristics and change characteristics of the scale formation and shedding are obtained.

[0035] S4: Based on the regularity of scale formation and shedding, the risk of fluctuation of the heat transfer efficiency of the furnace wall caused by the change of the scale is evaluated using a dynamic Bayesian network, and a heat transfer fluctuation risk probability is output, including: Based on the cycle regularity and change characteristics of the scale formation and shedding in the heat radiation region of the furnace wall, a dynamic Bayesian network including hidden state nodes and observation nodes is constructed; The scale adhesion period and the scale shedding period are taken as hidden state nodes, and the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value are taken as observation nodes, and the probability value of the fluctuation of the heat transfer efficiency of the furnace wall is determined through the calculation of the state transition probability and the observation probability in the dynamic Bayesian network.

[0036] The hidden state node is used to describe the state that cannot be directly observed but can be inferred through indirect data, and the observation node is used to describe the actual value that can be measured.

[0037] The construction process of the hidden state node is as follows: according to the periodic law of scale formation and falling in the heat radiation area of the furnace wall, the scale adhesion stage and the scale falling stage are set as two different hidden state nodes. Specifically, according to the cycle length of the scale adhesion stage and the scale falling stage, two different hidden state nodes are set in the dynamic Bayesian network to represent different states of the furnace wall heat transfer, for example, the stage in which the scale on the furnace wall surface gradually accumulates, resulting in gradually increasing temperature fluctuation amplitude, is set as the scale adhesion period; the stage in which the scale on the furnace wall surface gradually falls off, resulting in gradually decreasing temperature fluctuation amplitude, is set as the scale falling period. The scale adhesion stage and the scale falling stage are taken as hidden state nodes to represent different hidden states of the furnace wall heat transfer efficiency, so as to determine the definition and value range of the hidden state node.

[0038] The construction process of the observation node is as follows: the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value in the heat radiation area of the furnace wall are selected as the observation node. Specifically, the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value at each time are taken as two independent observation nodes in the dynamic Bayesian network; each observation node is represented by the actual measured value in the dynamic Bayesian network, for example, the total energy value of the high-frequency signal component obtained in the measurement process at each time and the value of the temperature fluctuation amplitude at the corresponding time are recorded in the observation node.

[0039] The state transition probability is calculated based on the dynamic Bayesian network. The state transition probability represents the change of the hidden state nodes over time, that is, the possibility value of the transition between the scale adhesion period and the scale falling period. The calculation method is as follows: according to the historical data of the scale adhesion stage and the scale falling stage on the furnace wall surface, the frequency of transition from the scale adhesion stage to the scale falling stage and the frequency of transition from the scale falling stage to the scale adhesion stage are counted respectively; the ratio of the frequency of transition in a specific stage to the total frequency of occurrence of the specific stage is the value of the state transition probability; for example, the ratio of the number of times that the scale changes from the adhesion stage to the falling stage in the statistical cycle to the number of times that the adhesion stage occurs is the probability value of the transition from the adhesion stage to the falling stage; the probability of the transition from the falling stage to the adhesion stage is also calculated in the same way to obtain the probability value of the transition from the falling stage to the adhesion stage. The historical data are the historical records of the mutual transition of the scale adhesion stage and the scale falling stage, including the starting time point, the ending time point and the number of transitions.

[0040] The observation probability is calculated based on the dynamic Bayesian network, and the observation probability is used to describe the possibility of the value corresponding to the observation node appearing when the hidden state node is in a specific state. The calculation method is that the historical record data of the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value are divided into multiple value intervals. For example, the total energy value of the high-frequency signal is divided into high, medium and low value intervals; the temperature fluctuation amplitude value is divided into large, medium and small value intervals; the division is based on uniform division according to the maximum and minimum values of the total energy value and the temperature fluctuation amplitude value in the historical record data.

[0041] The total energy value of the high-frequency signal component in the scale adhesion period is compared with each value interval of the determined total energy value of the high-frequency signal respectively, and the number of times of the total energy value of the high-frequency signal component appearing in each value interval is determined; the temperature fluctuation amplitude value is compared with each value interval of the temperature fluctuation amplitude value respectively, and the number of times of the temperature fluctuation amplitude value appearing in each value interval is determined.

[0042] The total number of historical records in the scale adhesion period is determined, which is the total number of historical records of all total energy values of the high-frequency signal or temperature fluctuation amplitude values in the scale adhesion period.

[0043] The occurrence frequency of each value interval is divided by the total number of historical records respectively to obtain the observation probability value of each value interval in the scale adhesion stage.

[0044] In the scale shedding period, the same method as in the scale adhesion stage is used to calculate the observation probability, that is, the value intervals are divided, the number of times of each value interval is counted, and then the number of times of each value interval is divided by the total number of historical records in the scale shedding period to obtain the observation probability value of each value interval in the scale shedding stage.

[0045] After obtaining the state transition probability and the observation probability, the probability value of the fluctuation of the furnace wall heat transfer efficiency is calculated by using the probability inference algorithm of the dynamic Bayesian network. The calculation method is that the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value measured at each time in the observation node are substituted into the dynamic Bayesian network, and then the state transition probability and the observation probability are calculated by using the Bayesian probability calculation formula, that is, the posterior probability is equal to the likelihood probability multiplied by the prior probability and then divided by the marginal probability of the observation data; specifically, the state transition probability is taken as the prior probability, and the observation probability is taken as the likelihood probability, and the posterior probability value of being in a specific hidden state node (scale adhesion or shedding state) is calculated.

[0046] In the scale adhesion state, the temperature fluctuation amplitude value and the high frequency signal component total energy value of each pixel position in the historical record data in the heat radiation area of the furnace wall are taken as sample data, and the determination standard of the fluctuation of the furnace wall heat transfer efficiency is set, for example, the temperature fluctuation amplitude exceeding 120% of the average value of the historical temperature fluctuation is taken as the standard; all the historical record data in the scale adhesion state are analyzed to determine the number of data satisfying the fluctuation determination standard, and then the number of data satisfying the fluctuation determination standard is divided by the total number of the historical record data in the scale adhesion state to obtain the conditional probability of the obvious fluctuation of the furnace wall heat transfer efficiency in the scale adhesion state. In the scale adhesion state, the temperature fluctuation amplitude value and the high frequency signal component total energy value of each pixel position in the historical record data in the heat radiation area of the furnace wall are taken as sample data, and the determination standard of the fluctuation of the furnace wall heat transfer efficiency is set, for example, the temperature fluctuation amplitude exceeding 120% of the average value of the historical temperature fluctuation is taken as the standard; all the historical record data in the scale adhesion state are analyzed to determine the number of data satisfying the fluctuation determination standard, and then the number of data satisfying the fluctuation determination standard is divided by the total number of the historical record data in the scale adhesion state to obtain the conditional probability of the obvious fluctuation of the furnace wall heat transfer efficiency in the scale adhesion state.

[0047] The posterior probabilities of the scale adhesion state and the scale shedding state are multiplied by the conditional probability of the fluctuation of the furnace wall heat transfer efficiency in the corresponding state respectively, and then added to obtain the probability value of the fluctuation of the furnace wall heat transfer efficiency. The probability value of the fluctuation of the furnace wall heat transfer efficiency represents the possibility of the change of the furnace wall heat transfer efficiency in the two implicit states of the scale adhesion and the scale shedding; if the probability value of the fluctuation of the furnace wall heat transfer efficiency is high, it indicates that the fluctuation of the furnace wall heat transfer efficiency at the current moment is obvious, and vice versa.

[0048] The calculation method of the marginal probability is that the product of the prior probability of the scale adhesion state and the observation probability of the scale adhesion state is added to the product of the prior probability of the scale shedding state and the observation probability of the scale shedding state.

[0049] S5: A coupling model between the furnace combustion process and the heat transfer fluctuation risk probability is constructed, and the heat loss factor in the calculation of the combustion efficiency is corrected, including: According to the calculation formula of the heat loss in the furnace combustion process, a functional relationship between the furnace combustion process and the probability value of the fluctuation of the furnace wall heat transfer efficiency is established; The calculation formula of the heat loss in the furnace combustion process contains a heat loss factor. The calculation formula of the heat loss in the furnace combustion process is that the total heat loss in the furnace combustion process is equal to the total heat generated by combustion multiplied by the heat loss factor. The total heat generated by the furnace combustion is determined by the product of the actual combustion amount of the fuel and the heat value of the fuel, and the heat loss factor is a proportional coefficient representing the proportion of the heat loss to the external environment due to the ineffective utilization of the heat in the furnace; the larger the heat loss factor, the higher the proportion of the heat loss in the furnace combustion process, and the lower the combustion efficiency of the water jacket heating furnace.

[0050] The method for establishing the function relationship is: taking the probability value of the fluctuation of the furnace wall heat transfer efficiency as the input variable of the function relationship, taking the value of the heat loss in the furnace combustion process as the output variable of the function relationship, and determining the form of the function relationship. The method for determining the function relationship is: setting the basic function form of the function relationship, for example, selecting a function form with linear characteristics. Specifically, the probability value of the fluctuation of the furnace wall heat transfer efficiency and the corresponding heat loss in the furnace combustion process are obtained through historical record data, and the probability value of the fluctuation of the furnace wall heat transfer efficiency and the corresponding heat loss in the furnace combustion process are substituted into the selected basic function form to calculate the error between the probability value of the fluctuation of the furnace wall heat transfer efficiency and the heat loss in the furnace combustion process. The error calculation method is: squaring the difference between the heat loss in the furnace combustion process and the heat loss in the furnace combustion process predicted by the function relationship, and then adding the error squares of all historical data points one by one to obtain the error sum of squares. By adjusting the parameters of the basic function form, the error sum of squares value is minimized, that is, the optimal parameters of the basic function form are obtained. For example, if the basic function form is selected as a linear function, the slope value and the intercept value of the linear function corresponding to the minimum error sum of squares value calculated from the historical record data are obtained.

[0051] Adjust the heat loss factor in the calculation formula of the heat loss in the furnace combustion process according to the probability value of the fluctuation of the furnace wall heat transfer efficiency to determine the correction value of the heat loss factor. The adjustment method of the heat loss factor is: when the probability value of the fluctuation of the furnace wall heat transfer efficiency increases, it means that the instability of the furnace wall heat transfer efficiency increases, and the heat loss in the furnace combustion process also increases accordingly, so the heat loss factor needs to be increased; when the probability value of the fluctuation of the furnace wall heat transfer efficiency decreases, it means that the furnace wall heat transfer efficiency tends to be stable, and the heat loss in the furnace combustion process decreases, so the heat loss factor needs to be decreased; the adjustment method of the heat loss factor is: taking the absolute value of the difference between the probability value of the fluctuation of the furnace wall heat transfer efficiency and the historical average probability value of the fluctuation of the furnace wall heat transfer efficiency as the adjustment coefficient; multiplying the adjustment coefficient by the historical average value of the heat loss factor to obtain the correction value of the heat loss factor; the historical average value of the heat loss factor is the arithmetic mean of the heat loss factor in the period of the historical record data pairs. For example, if the probability value of the fluctuation of the furnace wall heat transfer efficiency is greater than the historical average probability value by a certain percentage, for example, more than ten percent of the historical average probability value, at this time the historical average value of the heat loss factor is increased by a certain percentage, for example, five percent; on the contrary, it is decreased by a certain percentage; through the above calculation method, the correction value of the heat loss factor is obtained.

[0052] Based on the correction value of the heat loss factor, a corrected heat loss factor is determined. The correction value of the heat loss factor is added to the historical average value of the heat loss factor to obtain the corrected heat loss factor.

[0053] S6: Real-time update the combustion efficiency prediction model parameters according to the corrected heat loss factor, and output the real-time prediction value of the combustion efficiency of the water jacket heating furnace, including: Calculate the correction amount of the combustion efficiency prediction model parameter to be corrected in the combustion efficiency prediction model based on the corrected heat loss factor; Calculate the correction amount of the combustion efficiency prediction model parameter to be corrected in the combustion efficiency prediction model based on the corrected heat loss factor: the combustion efficiency prediction model contains multiple parameters, such as fuel combustion completeness parameter, furnace heat transfer efficiency parameter, flue gas heat loss parameter and incomplete combustion heat loss parameter, etc., each parameter affects the accuracy of the output of the combustion efficiency prediction model; the combustion efficiency prediction model is obtained by predicting the combustion efficiency of the water jacket heating furnace through the functional relationship between the parameters and the input values.

[0054] The method for calculating the correction amount of the combustion efficiency prediction model parameter is: based on the change of the heat loss factor in the furnace combustion process, the correction amount of the flue gas heat loss parameter in the combustion efficiency prediction model is determined; the calculation method of the correction amount is: calculating the difference between the corrected heat loss factor and the historical average heat loss factor; dividing the difference by the historical average heat loss factor to obtain a correction coefficient, multiplying the correction coefficient by the initial value of the flue gas heat loss parameter in the combustion efficiency prediction model parameter to calculate the correction amount of the flue gas heat loss parameter. For example, if the corrected heat loss factor is increased by 5% compared with the historical average heat loss factor, the flue gas heat loss parameter in the combustion efficiency prediction model is also increased by 5%; if the corrected heat loss factor is lower than the historical average heat loss factor, the flue gas heat loss parameter value is correspondingly reduced. Through the above calculation method, the correction amount of each parameter in the combustion efficiency prediction model is obtained, which ensures that the combustion efficiency prediction model can reflect the actual change of the furnace heat loss.

[0055] Add the correction amount of the combustion efficiency prediction model parameter to the initial value of the combustion efficiency prediction model parameter to obtain the real-time updated combustion efficiency prediction model parameter; The initial value of the combustion efficiency prediction model parameter is the parameter reference value determined according to the actual furnace combustion in the historical measurement period, such as the historical average value of the fuel combustion completeness parameter, the furnace heat transfer efficiency parameter, the flue gas heat loss parameter and the incomplete combustion heat loss parameter; add the correction amount of the combustion efficiency prediction model parameter to the initial value of the combustion efficiency prediction model parameter to obtain the real-time updated combustion efficiency prediction model parameter; ensure that the combustion efficiency prediction model parameter can dynamically follow the actual furnace combustion condition for timely updating and adjusting.

[0056] According to the real-time updated combustion efficiency prediction model parameters, the temperature fluctuation amplitude value, the total energy value of the high-frequency signal component and the probability value of the furnace wall heat transfer efficiency fluctuation of each pixel point in the furnace wall thermal radiation region of the water jacket heating furnace are input into the combustion efficiency prediction model, and the real-time prediction value of the combustion efficiency of the water jacket heating furnace is output. The temperature fluctuation amplitude value, the total energy value of the high-frequency signal component and the probability value of the furnace wall heat transfer efficiency fluctuation are taken as the input variables of the combustion efficiency prediction model, and are substituted into the combustion efficiency prediction model.

[0057] Specifically, if a linear function is taken as the function relationship of the combustion efficiency prediction model, the product of the fuel combustion completeness parameter and the furnace wall temperature fluctuation amplitude value is added to the product of the furnace heat transfer efficiency parameter and the probability value of the furnace wall heat transfer efficiency fluctuation, and then the product of the flue gas heat loss parameter and the total energy value of the high-frequency signal component is subtracted, and then the product of the incomplete combustion heat loss parameter and the probability value of the furnace wall heat transfer efficiency fluctuation is subtracted, to obtain the real-time prediction value of the combustion efficiency of the water jacket heating furnace.

[0058] Embodiment 2 The difference between the embodiment 2 and the embodiment 1 of the present application is that the embodiment 2 introduces a water jacket heating furnace combustion efficiency prediction system.

[0059] Figure 2 The structure diagram of the water jacket heating furnace combustion efficiency prediction system is given, and the water jacket heating furnace combustion efficiency prediction system comprises: The feature extraction module: a thermal image sequence of the furnace wall surface of the water jacket heating furnace is collected, the pixel temperature gradient and texture change features in the furnace wall thermal radiation region are extracted, and the furnace wall heat transfer abnormal region is identified; The wavelet decomposition module: the temperature gradient features of the furnace wall heat transfer abnormal region are subjected to wavelet decomposition, the high-frequency signal component reflecting the scale deposition and shedding process is obtained, and the high-frequency signal energy is calculated; The rule identification module: the correlation between the high-frequency signal energy and the temperature fluctuation amplitude of the furnace wall is analyzed, and the scale formation and shedding rule on the furnace wall is extracted; The risk assessment module: based on the scale formation and shedding rule, the dynamic Bayesian network is used to evaluate the furnace wall heat transfer efficiency fluctuation risk caused by the scale change, and the heat transfer fluctuation risk probability is output; The factor correction module: a coupling model between the furnace combustion process and the heat transfer fluctuation risk probability is constructed, and the heat loss factor in the combustion efficiency calculation is corrected; The efficiency prediction module: according to the corrected heat loss factor, the combustion efficiency prediction model parameters are updated in real time, and the real-time prediction value of the combustion efficiency of the water jacket heating furnace is output.

[0060] The above formulas are all dimensionless numerical calculations, the formulas are obtained by collecting a large amount of data to simulate a formula of the most recent real situation, and the preset parameters and threshold values in the formula are set by a person skilled in the art according to actual conditions.

[0061] The above embodiments can be realized wholly or partially by software, hardware, firmware or any other combination. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transferred from one computer readable storage medium to another, for example, the computer instructions can be transferred from one website, computer, server or data center to another website, computer, server or data center by wired (for example, infrared, wireless, microwave, etc.) mode. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center and the like containing one or more available medium sets. The available medium can be a magnetic medium (for example, floppy disk, hard disk, magnetic tape), an optical medium (for example, DVD) or a semiconductor medium. The semiconductor medium can be a solid state disk.

[0062] Those skilled in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0063] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and module can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0064] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the division of the above-described device embodiments is merely a logical function division, and there can be another division manner for the actual implementation, for example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different modules can be indirect couplings or communication connections through some interfaces, devices or modules, and can be electrical, mechanical or other forms.

[0065] The modules illustrated as separated components can or can not be physically separated, and the components illustrated as modules can or can not be physical modules, and can be located in one place, or can be distributed on multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments.

[0066] In addition, each functional module in each embodiment of the present application can be integrated into a processing module, or each module can be physically present alone, or two or more modules can be integrated into one module.

[0067] If the functions are realized in the form of software function modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, and includes several instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program codes that can be stored in the medium.

[0068] The above description is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, and all should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0069] Finally: the above only for the preferred embodiments of the present application, and not for limiting the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application, should be included in the scope of protection of the present application.

Claims

1. A method for predicting the combustion efficiency of a water-jacketed heating furnace, characterized in that: The steps include: S1: Collect thermal image sequences of the water-jacket heating furnace wall surface, extract pixel temperature gradients and texture change features within the heat radiation area of ​​the furnace wall, and identify abnormal heat transfer areas on the furnace wall; S2: Perform wavelet decomposition on the temperature gradient characteristics of the abnormal heat transfer area on the furnace wall to obtain the high-frequency signal components reflecting the scale adhesion and shedding process, and calculate the high-frequency signal energy; S3: By analyzing the correlation between the high-frequency signal energy and the temperature fluctuation amplitude of the furnace wall, the law of scale formation and shedding on the furnace wall is extracted; S4: Based on the laws of scale formation and shedding, a dynamic Bayesian network is used to evaluate the risk of heat transfer efficiency fluctuations in the furnace wall caused by scale changes, and the probability of heat transfer fluctuation risk is output; S5: Construct a coupling model between the furnace combustion process and the heat transfer fluctuation risk probability, and correct the heat loss factor in the combustion efficiency calculation; S6: Based on the corrected heat loss factor, the combustion efficiency prediction model parameters are updated in real time, and the real-time prediction value of the combustion efficiency of the water jacket heating furnace is output.

2. The method for predicting combustion efficiency of a water jacket heating furnace according to claim 1, characterized in that: S1, specifically: Collect thermal image sequences of the water-jacket heating furnace wall surface, delineate the boundary position of the furnace wall heat radiation area, and determine the temperature value of each pixel point in the furnace wall heat radiation area; The temperature difference between any adjacent pixel points in the heat radiation area of ​​the furnace wall is calculated using the spatial gradient calculation method to obtain the temperature gradient value of each pixel position; The gray-level co-occurrence matrix method is used to extract the energy, contrast and correlation indexes of each frame of thermal image in the heat radiation area of ​​the furnace wall to determine the texture change characteristics. The judgment threshold is set according to the temperature gradient value and texture change characteristics to identify the abnormal heat transfer area of ​​the furnace wall.

3. The method for predicting combustion efficiency of a water jacket heating furnace according to claim 2, characterized in that: S2, specifically: Based on the multi-scale wavelet transform method, the temperature gradient value of each pixel position in the abnormal heat transfer area of ​​the furnace wall is decomposed to obtain the high-frequency wavelet coefficients and low-frequency wavelet coefficients corresponding to the temperature gradient value. The high-frequency wavelet coefficients are mapped to each pixel point in the abnormal heat transfer area of ​​the furnace wall according to the spatial position to obtain the high-frequency signal component distribution of each pixel point; The high-frequency signal component distribution of each pixel point in the abnormal heat transfer area of ​​the furnace wall is squared, and then the high-frequency signal component values ​​of each pixel point after square calculation are accumulated point by point to determine the total energy value of the high-frequency signal component in the abnormal heat transfer area of ​​the furnace wall.

4. The method for predicting combustion efficiency of a water jacket heating furnace according to claim 3, characterized in that: S3, specifically: The total energy value of the high-frequency signal component in the abnormal heat transfer area of ​​the furnace wall is used as the horizontal axis, and the temperature fluctuation amplitude value of each pixel point in the heat radiation area of ​​the furnace wall is used as the vertical axis to construct a scatter plot between the total energy of the high-frequency signal component and the temperature fluctuation amplitude. The regression analysis method is used to fit the scatter plot and obtain the functional expression of the corresponding relationship between the total energy of the high-frequency signal component and the temperature fluctuation amplitude; According to the temporal variation trend and function expression of the total energy of the high-frequency signal components, the periodic law and variation characteristics of scale formation and shedding in the heat radiation area of ​​the furnace wall are determined.

5. The method for predicting combustion efficiency of a water jacket heating furnace according to claim 4, characterized in that: S4, specifically: Based on the periodic laws and changing characteristics of scale formation and shedding in the heat radiation area of ​​the furnace wall, a dynamic Bayesian network consisting of hidden state nodes and observation nodes is constructed. The scale attachment period and the scale shedding period are used as hidden state nodes, and the total energy value of the high-frequency signal component and the temperature fluctuation amplitude value are used as observation nodes. The probability value of the furnace wall heat transfer efficiency fluctuation is determined by calculating the state transition probability and observation probability in the dynamic Bayesian network.

6. The method for predicting combustion efficiency of a water jacket heating furnace according to claim 5, characterized in that: S5, specifically: Based on the calculation formula of heat loss during furnace combustion, a functional relationship between the furnace combustion process and the probability value of the fluctuation of furnace wall heat transfer efficiency is established; According to the probability value of the fluctuation of the heat transfer efficiency of the furnace wall, the heat loss factor in the calculation formula of the heat loss during the furnace combustion process is adjusted to determine the corrected value of the heat loss factor; Based on the corrected value of the heat loss factor, a corrected heat loss factor is determined.

7. The method for predicting combustion efficiency of a water jacket heating furnace according to claim 6, characterized in that: S6, specifically: Calculating a correction amount of a combustion efficiency prediction model parameter to be corrected in the combustion efficiency prediction model based on the corrected heat loss factor; Adding the correction amount of the combustion efficiency prediction model parameter to the initial value of the combustion efficiency prediction model parameter to obtain the combustion efficiency prediction model parameter after real-time update; According to the real-time updated combustion efficiency prediction model parameters, the temperature fluctuation amplitude value of each pixel point in the heat radiation area of ​​the water jacket heating furnace wall, the total energy value of the high-frequency signal component and the probability value of the furnace wall heat transfer efficiency fluctuation are input into the combustion efficiency prediction model, and the real-time prediction value of the water jacket heating furnace combustion efficiency is output.

8. A water jacket heating furnace combustion efficiency prediction system, used to implement a water jacket heating furnace combustion efficiency prediction method according to any one of claims 1 to 7, characterized in that: include: Feature extraction module: collects thermal image sequences of the water-jacket heating furnace wall surface, extracts pixel temperature gradients and texture change features within the heat radiation area of ​​the furnace wall, and identifies abnormal heat transfer areas on the furnace wall; Wavelet decomposition module: Performs wavelet decomposition on the temperature gradient characteristics of the abnormal heat transfer area on the furnace wall to obtain the high-frequency signal components reflecting the scale adhesion and shedding process, and calculates the high-frequency signal energy; Pattern recognition module: By analyzing the correlation between high-frequency signal energy and the temperature fluctuation amplitude of the furnace wall, the pattern of scale formation and shedding on the furnace wall is extracted; Risk Assessment Module: Based on the laws of scale formation and shedding, a dynamic Bayesian network is used to assess the risk of fluctuations in furnace wall heat transfer efficiency caused by scale changes, and output the probability of heat transfer fluctuation risk; Factor correction module: Constructs a coupling model between the furnace combustion process and the heat transfer fluctuation risk probability, and corrects the heat loss factor in the combustion efficiency calculation; Efficiency prediction module: Based on the corrected heat loss factor, the combustion efficiency prediction model parameters are updated in real time, and the real-time prediction value of the combustion efficiency of the water jacket heating furnace is output.

Images