Boiler combustion operation optimization method and system based on digital twin

Abstract

The present application relates to the technical field of boiler combustion optimization, and specifically to a method and system for optimizing boiler combustion operation based on digital twins, the method comprising: real-time collection of temperature, pressure and pollutant concentration at each moment of boiler combustion; selecting multiple moments before each moment, recording them as historical moments of each moment; determining the combustion influencing factor at each moment; determining the combustion efficiency factor at each moment; based on the combustion efficiency factor, using an integrated learning algorithm for prediction, determining the trend of changes in the predicted concentration of pollutants, and the digital twin model feeds back the boiler state to the control system to control the boiler combustion temperature. The present application can improve the accuracy of the prediction of pollutant emission concentration, thereby controlling the boiler combustion temperature.

Description

Technical Field

[0001] The present application relates to the technical field of boiler combustion optimization, and specifically to a boiler combustion operation optimization method and system based on digital twin. Background Art

[0002] As an advanced integrated model, digital twin technology can build a virtual model corresponding to the actual model based on the detailed data of the physical entity. The virtual model can receive and process a large amount of data from the boiler sensor in real time, so as to accurately reflect the current status and performance indicators of the actual model. For the operating status of the boiler, the in-depth analysis of the virtual model data through digital twin technology can not only monitor the real-time operation of the boiler, but also predict its future performance and potential problems. Furthermore, through the simulation function, the digital twin can also test different operating strategies and parameter adjustments in a virtual environment, optimize the combustion efficiency and thermal efficiency of the boiler, and reduce unnecessary energy consumption and environmental pollution. This simulation optimization process does not cause any risk to the actual boiler, but can provide operators with valuable operating experience and decision-making basis.

[0003] Without the use of digital twin technology, the predictive ability of boiler combustion operation is relatively weak due to the lack of comprehensive and in-depth real-time data monitoring and analysis of the boiler system. Without digital twins, it is difficult for operators to accurately predict the performance trend of the boiler, potential failure points, and fluctuations in carbon emissions. However, despite the use of digital twin technology, in order to meet environmental standards, boilers may need to implement complex emission control strategies such as selective catalytic reduction (SCR) or selective non-catalytic reduction (SNCR) to reduce pollutant emissions during boiler operation, and the performance of emission control equipment (such as denitrification devices, dust collectors) may fluctuate over time or perform inconsistently under different operating conditions, thereby affecting the combustion process and increasing the complexity of model prediction, resulting in poor prediction of pollutant emissions. Summary of the invention

[0004] In order to solve the above technical problems, a boiler combustion operation optimization method and system based on digital twin are provided to solve the existing problems.

[0005] The solution to the technical problem of this application is to provide a boiler combustion operation optimization method and system based on digital twin, including the following steps:

[0006] In a first aspect, an embodiment of the present application provides a boiler combustion operation optimization method based on digital twin, the method comprising the following steps:

[0007] Real-time collection of temperature, pressure and pollutant concentration at every moment of boiler combustion;

[0008] Select multiple moments before each moment and record them as the historical moments of each moment;

[0009] Determine the combustion influencing factor at each moment based on the changes and dispersion of local pressure and local temperature at different historical moments, as well as the correlation between pressure and temperature;

[0010] Determine the combustion efficiency factor at each moment based on the correlation between the temperature at each historical moment and the pollutant concentration, and the difference in pollutant concentration under different temperature conditions, combined with the combustion influencing factor;

[0011] Based on the combustion efficiency factor, an integrated learning algorithm is used for prediction to determine the changing trend of the predicted pollutant concentration. The digital twin model feeds back the boiler status to the control system to control the boiler combustion temperature.

[0012] Preferably, the determining of the combustion influencing factor at each moment includes:

[0013] Taking the pressure and temperature at each historical moment as the center, respectively, a local pressure window and a local temperature window of a preset size are set;

[0014] Analyze the pressure fluctuation changes in the local pressure window at each historical moment, and obtain the historical fluctuation moments at each moment;

[0015] Analyze the difference in pressure and the degree of dispersion in the local temperature window of each historical fluctuation moment at each moment, and determine the local temperature difference of each historical fluctuation moment at each moment;

[0016] Determine the local correlation of each historical fluctuation moment at each moment according to the correlation between the pressures at all historical moments in the local pressure window and the temperatures at all historical moments in the local temperature window at each moment;

[0017] Determining a local influence coefficient of each historical fluctuation moment at each moment based on the pressure fluctuation degree, the local temperature difference and the local correlation degree;

[0018] The average value of the local influence coefficients of all historical fluctuation moments at each moment is taken as the combustion influence factor at each moment.

[0019] Preferably, the obtaining of each historical fluctuation moment at each moment includes:

[0020] Determine the degree of dispersion of pressure at all historical moments within the local pressure window at each historical moment as the pressure fluctuation degree at each historical moment;

[0021] The historical moments when the pressure fluctuation degree is greater than the preset pressure fluctuation threshold are recorded as each historical fluctuation moment at each moment.

[0022] Preferably, determining the local temperature difference at each historical fluctuation moment at each moment includes:

[0023] Determine the degree of dispersion of the temperature at all historical moments within the local temperature window of each historical moment at each moment as the temperature dispersion at each historical moment;

[0024] The mean of the temperature dispersion at all historical moments at each moment is recorded as the first mean;

[0025] The difference between the temperature dispersion at each historical fluctuation moment at each moment and the first mean value is determined as the local temperature difference at each historical fluctuation moment at each moment.

[0026] Preferably, the local influence coefficient of each historical fluctuation moment at each moment is the product of the pressure fluctuation degree, the local temperature difference and the local correlation.

[0027] Preferably, the determining of the combustion efficiency factor at each moment includes:

[0028] The partial correlation analysis algorithm is used to calculate the partial correlation coefficient between the temperature and pollutant concentration at all historical moments at each moment;

[0029] Cluster the temperatures of all historical moments at each moment and obtain two clusters;

[0030] Analyze the differences between pollutant concentrations corresponding to all historical moments in each cluster and calculate the first difference;

[0031] Calculate the standard deviation of all temperatures in each cluster at the corresponding collection time; record the mean of the standard deviations of the two clusters as the second mean;

[0032] Determine the temperature influence coefficient at each moment based on the partial correlation coefficient, the first difference and the second mean;

[0033] The product of the temperature influence coefficient and the combustion influence factor is used as the combustion efficiency factor at each moment.

[0034] Preferably, the first difference is the difference between the mean values ​​of pollutant concentrations corresponding to all historical moments within the two clusters.

[0035] Preferably, the temperature influence coefficient at each moment is the product of the partial correlation coefficient, the first difference and the second mean.

[0036] Preferably, the determining of the change trend of the predicted pollutant concentration and the digital twin model feeding back the boiler status to the control system include:

[0037] The combustion efficiency factor at each moment is normalized, and the normalized result is used as the concentration weight at each moment;

[0038] According to the pollutant concentration at each moment and the concentration weight at the corresponding moment, an integrated learning algorithm is used for prediction;

[0039] Perform linear fitting based on all predicted pollutant concentrations to obtain the slope of the fitted line;

[0040] If the fitted slope is greater than the preset trend threshold, the pollutant concentration is on an upward trend, and the digital twin model provides feedback to the control system.

[0041] In the second aspect, an embodiment of the present application also provides a boiler combustion operation optimization system based on digital twins, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above-mentioned digital twin-based boiler combustion operation optimization methods.

[0042] This application has at least the following beneficial effects:

[0043] The present application determines the combustion influencing factor at each moment according to the changes and discreteness of local pressure and local temperature at different historical moments at each moment, as well as the correlation between pressure and temperature. The beneficial effect is that it takes into account the impact of the fluctuation of boiler temperature and pressure on combustion to evaluate the impact on combustion efficiency; the combustion efficiency factor at each moment is determined according to the correlation between the changes in temperature and pollutant concentration at each historical moment, as well as the difference in pollutant concentration under different temperature conditions. The beneficial effect is that it takes into account the significance of the impact of temperature on pollutant emissions after excluding the interference of pressure factors; based on the combustion efficiency factor, according to the changing trend of pollutant concentration predicted by the integrated learning algorithm, the digital twin model feeds back the boiler state to the control system to control the boiler combustion temperature. The beneficial effect is that it adjusts the weights in the prediction process of pollutant concentration at different moments, so that the model pays more attention to the changes in pollutant concentration that are more affected by temperature, which helps to improve the accuracy of pollutant emission prediction and achieve precise control of boiler combustion temperature, thereby minimizing environmental impact and ensuring an efficient and clean combustion process. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The boiler combustion operation optimization method based on digital twin of the present application is further described in detail below in conjunction with the accompanying drawings.

[0045] Figure 1 A flowchart of the steps of the boiler combustion operation optimization method based on digital twin provided in the embodiment of the present application;

[0046] Figure 2 A flowchart of the steps of the method for obtaining the combustion influence factor at each moment provided in an embodiment of the present application;

[0047] Figure 3 A schematic diagram of the temperature clustering results of all historical moments at each moment provided in an embodiment of the present application;

[0048] Figure 4 A flowchart of the steps of a method for obtaining the combustion efficiency factor at each moment provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of this application more clear, the following is a further detailed description of the boiler combustion operation optimization method and system based on digital twin proposed in this application in combination with the accompanying drawings and implementation examples. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application.

[0050] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0051] See also Figure 1 , which shows a flow chart of the steps of a boiler combustion operation optimization method based on digital twin provided by an embodiment of the present application, the method comprising the following steps:

[0052] Step 1: Collect the temperature, pressure and pollutant concentration of the boiler at every moment during combustion in real time.

[0053] Install the high temperature resistant thermocouple temperature sensor in the combustion chamber to collect the temperature data of the boiler in real time; install the strain gauge pressure sensor near the exhaust port of the boiler to collect the pressure data of the boiler in real time; install the electrochemical sensors for measuring nitrides, carbides and sulfides in a frame to integrate them into a sensor group to ensure that the concentrations of different pollutants in the flue gas can be measured at the same time, install it in the smoke outlet area, and accumulate the concentrations of different pollutants detected by the sensor group to obtain the total concentration of pollutants emitted by the boiler in real time, and set the collection interval of all sensors to t. According to the collection time sequence, the temperature, pressure and pollutant concentration at each moment are obtained.

[0054] Preferably, in this embodiment, the collection interval is 0.5 s, and the data collected at all times are normalized.

[0055] At this point, the temperature, pressure and pollutant concentration at each moment of boiler combustion are collected in real time.

[0056] Step 2, select multiple moments before each moment and record them as historical moments of each moment; determine the combustion influencing factor at each moment according to the changes and discreteness of the local pressure and local temperature at different historical moments of each moment, as well as the correlation between pressure and temperature.

[0057] Since the combustion efficiency of the boiler directly determines the degree of complete combustion of the fuel, by analyzing the combustion state, we can understand whether the fuel is fully burned. Temperature, pressure and pollutant concentration are key parameters that reflect the combustion state of the boiler. By monitoring the temperature changes in the combustion chamber, we can evaluate whether the combustion is uniform and whether there are hot spots or cold spots; the pressure data at the exhaust port helps monitor the back pressure and evaluate the performance of the flue gas emission system; and monitoring the concentration of pollutants such as nitrides, carbides and sulfides in the flue gas can evaluate the emissions during the combustion process. By analyzing the changing trends of these data, we can make accurate judgments on the combustion state of the boiler.

[0058] The temperature inside the boiler reflects the thermal efficiency of combustion. High temperature usually means more heat released by fuel combustion, while the pressure inside the boiler affects the stability of the combustion process. An appropriate pressure level helps maintain stable combustion, while too high or too low pressure will lead to a decrease in combustion efficiency. Therefore, by analyzing the changes in the temperature and pressure of the boiler to reflect the changes in the temperature and pressure of the combustion process, the moments with large local fluctuations are obtained, specifically:

[0059] Select multiple moments before each moment and record them as the historical moments of each moment;

[0060] Preferably, in this embodiment, the moment corresponding to the data within 30 minutes before each moment is taken as the historical moment. As for other implementation methods, the implementer can set it according to the actual situation.

[0061] Taking the pressure and temperature at each historical moment as the center, respectively, a local pressure window and a local temperature window of a preset size are set;

[0062] Determine the degree of dispersion of pressure at all historical moments within the local pressure window at each historical moment as the pressure fluctuation degree at each historical moment;

[0063] Determine the degree of dispersion of the temperature at all historical moments within the local temperature window of each historical moment at each moment as the temperature dispersion at each historical moment;

[0064] The mean of the temperature dispersion at all historical moments at each moment is recorded as the first mean;

[0065] Record the historical moments when the pressure fluctuation is greater than a preset pressure fluctuation threshold as each historical fluctuation moment;

[0066] Preferably, in this embodiment, the size of the local pressure window and the local temperature window is set to 1×7. It should be noted that if the size of the local pressure window and the local temperature window is insufficient, the mean value in the window is used to fill it up as another implementation method. The implementer can set it according to the actual situation.

[0067] Preferably, in this embodiment, the standard deviation of the pressures of all historical moments within the local pressure window of each historical moment is calculated as the pressure fluctuation of each historical moment; the standard deviation of the temperatures of all historical moments within the local temperature window of each historical moment is calculated as the temperature dispersion of each historical moment; the preset pressure fluctuation threshold value is 0.3, and as other implementation methods, the implementer can set it according to actual conditions.

[0068] It can be understood that this embodiment reflects the degree of discreteness by calculating the standard deviation of the pressure at all moments within the local pressure window of each historical moment. As other implementation methods, implementers can adopt other methods of the prior art, such as variance, coefficient of variation, information entropy, etc. This embodiment does not impose any special restrictions on this.

[0069] Furthermore, the local temperature and pressure at all historical fluctuation moments are analyzed to reflect the influence of combustion stability at each moment of boiler combustion on combustion efficiency, specifically:

[0070] The difference between the temperature dispersion at each historical fluctuation moment at each moment and the first mean value is recorded as the local temperature difference at each historical fluctuation moment at each moment;

[0071] Determine the local correlation of each historical fluctuation moment at each moment according to the correlation between the pressures at all historical moments in the local pressure window and the temperatures at all historical moments in the local temperature window at each historical fluctuation moment at each moment;

[0072] Determine the local influence coefficient of each historical fluctuation moment at each moment by multiplying the product of the pressure fluctuation degree, the local temperature difference and the local correlation degree at each historical fluctuation moment at each moment;

[0073] The average of the local influence coefficients of all historical fluctuation moments at each moment is used as the combustion influence factor at each moment;

[0074] Preferably, in this embodiment, the absolute value of the difference between the temperature dispersion at each historical fluctuation moment and the first mean value at each moment is used as the local temperature difference at each historical fluctuation moment at each moment; the Pearson correlation coefficient between the local pressure window and the local temperature window at each historical fluctuation moment at each moment is calculated as the local correlation at each historical fluctuation moment at each moment; in this embodiment, the calculation method of the combustion influence factor at each moment is: Among them A t is the combustion influence factor at the tth moment, a t,k is the pressure fluctuation at the kth historical fluctuation moment at the tth moment, b t,k is the local temperature difference at the kth historical fluctuation moment at the tth moment, c t,k is the local correlation of the kth historical fluctuation moment at the tth moment, n k is the number of all historical fluctuation moments at time t, a t,k *b t,k *c t,k is the local influence coefficient of the kth historical fluctuation moment at the tth moment.

[0075] It can be understood that this embodiment measures the degree of correlation by calculating the Pearson correlation coefficient. As other implementations, implementers can adopt other methods in the prior art, such as cosine similarity, Spearman correlation coefficient, etc. This embodiment does not impose any special restrictions on this.

[0076] It should be noted that the pressure fluctuation reflects the degree of pressure fluctuation during the boiler combustion process. The greater the pressure fluctuation, the greater the local fluctuation of the boiler pressure, the greater the impact on the combustion efficiency, and the greater the combustion influence factor; the local temperature difference reflects the fluctuation of the local temperature during the period when the boiler pressure fluctuates greatly during the combustion process. The greater the local temperature difference, the greater the change in combustion efficiency at this time, the greater the impact of the boiler pressure change, and the greater the combustion influence factor; the local correlation reflects the degree of correlation between pressure changes and combustion efficiency changes; the larger the combustion influence factor, the more significant the impact of the volatility of the boiler pressure and temperature changes during the combustion process on the combustion efficiency.

[0077] Further, the step flow chart of the method for obtaining the combustion influence factor at each moment provided in the embodiment of the present application is as follows: Figure 2 shown.

[0078] At this point, the combustion influence factor at each moment is obtained.

[0079] Step 3, determining the combustion efficiency factor at each moment according to the correlation between the temperature and pollutant concentration at each historical moment, the difference in pollutant concentration under different temperature conditions, and the combustion influencing factor.

[0080] When the boiler is running, the efficiency and stability of combustion will have a direct impact on the pollutant emissions during combustion. Changes in temperature during the combustion process may change the generation mechanism of pollutants. For example, a high boiler temperature may promote NO x The formation of CO and HC may be increased when the boiler temperature is low. Therefore, when controlling the pressure effect, the influence of boiler combustion temperature change on pollutant emissions is analyzed, specifically:

[0081] A partial correlation analysis algorithm is used to calculate the partial correlation coefficient between the temperature and pollutant concentration at all historical moments at each moment while controlling for the pressure effect;

[0082] Cluster the temperatures of all historical moments at each moment and obtain two clusters;

[0083] Preferably, in this embodiment, K-means clustering algorithm is used for clustering, and the number of clusters is set to 2. As other implementation methods, the implementer can set it according to the actual situation.

[0084] Furthermore, the schematic diagram of the temperature clustering results of all historical moments at each moment provided in the embodiment of the present application is as follows: Figure 3 As shown, the horizontal axis is the acquisition order and the vertical axis is the temperature.

[0085] It is understandable that this embodiment only provides one clustering algorithm, namely, the K-means clustering algorithm. As other implementation methods, implementers can adopt other methods of the prior art, such as the hierarchical clustering algorithm, the DBSCAN clustering algorithm, etc. This embodiment does not impose any special restrictions on this.

[0086] Calculate the mean of the pollutant concentration corresponding to all historical moments in each cluster; record the difference of the mean between two clusters as the first difference;

[0087] Calculate the standard deviation of all temperatures in each cluster at the corresponding collection time, and record the mean of the standard deviations of the two clusters as the second mean;

[0088] Determine the product of the partial correlation coefficient, the first difference and the second mean as the temperature influence coefficient at each moment;

[0089] The product of the temperature influence coefficient and the combustion influence factor is used as the combustion efficiency factor at each moment.

[0090] Preferably, in this embodiment, the difference between the means of two clusters is taken as the first difference.

[0091] It should be noted that the partial correlation coefficient can exclude the impact of pressure on pollutant emissions, thereby more accurately evaluating the independent impact of temperature on pollutant emissions, reflecting the degree of correlation between changes in temperature and pollutant concentration. The first difference reflects the difference in pollutant emission levels in different temperature zones, and the second mean reflects the volatility of temperature on pollutant concentration under different temperature conditions. The larger the second mean, the more significant the impact of temperature changes on pollutant emissions; the larger the combustion influence factor, the greater the impact of fluctuations in air pressure and temperature during the combustion process on combustion efficiency, which means that the more unstable the combustion process is, the greater the comprehensive impact of temperature on pollutant emissions, the larger the combustion efficiency factor, and the more significant the impact of temperature changes on pollutant emissions.

[0092] Furthermore, the step flow chart of the method for obtaining the combustion efficiency factor at each moment provided in the embodiment of the present application is as follows: Figure 4 shown.

[0093] At this point, the combustion efficiency factor at each moment is obtained.

[0094] Step 4: Based on the combustion efficiency factor, an integrated learning algorithm is used to make predictions to determine the changing trend of the predicted pollutant concentration. The digital twin model feeds back the boiler status to the control system to control the boiler combustion temperature.

[0095] Furthermore, based on the real-time collected temperature, pressure and pollutant emission data during boiler combustion, the combustion efficiency factor at each moment is calculated, and the pollutant emission concentration is predicted to reflect the temperature change trend during boiler combustion, specifically:

[0096] According to the design drawings and specifications of the boiler, a geometric model of the boiler is established, and tools such as computational fluid dynamics (CFD) are used to simulate the flow and heat transfer inside the boiler, and the real-time data collected by sensors are integrated into the established digital twin model.

[0097] The pollutant concentration at all times is used as input, and an ensemble learning algorithm is used for training. The pollutant concentration at each time is assigned the same initial weight to ensure that the algorithm gives equal consideration to each sample and there is no uneven weight due to the number of times certain samples appear in the data set.

[0098] Preferably, in this embodiment, an adaptive enhancement (Adaboost ensemble learning) algorithm is used, and three weak classifiers are set, namely, random forest, decision tree and naive Bayes classification. As other implementation methods, implementers can adopt other ensemble learning algorithms of the prior art, such as the extreme gradient boosting algorithm (XGboost), LightGBM algorithm, etc. This embodiment does not impose any special restrictions on this.

[0099] The combustion efficiency factor at each moment is normalized, and the normalized result is used as the concentration weight at each moment. Combined with the pollutant concentration at the corresponding moment, an ensemble learning algorithm is used for prediction;

[0100] Perform linear fitting based on all predicted pollutant concentrations to obtain the slope of the fitted line;

[0101] If the fitting slope is greater than the preset trend threshold, it means that the emission concentration of pollutants is showing a significant upward trend due to the influence of temperature. At this time, the digital twin model will provide feedback to the control system to adjust the combustion parameters or activate the pollutant control equipment;

[0102] Preferably, in this embodiment, the Softmax function can be used to normalize the combustion efficiency factors at all times, converting all combustion efficiency factors into probability distributions in the interval (0,1), and the sum of all probability distributions is 1, which meets the requirements of the sample initial value of the Adaboost ensemble learning algorithm; secondly, when predicting the pollutant concentration, the pollutant concentration in the next half hour is predicted; the preset trend threshold value is 0.1, and as other implementation methods, the implementer can set it according to the actual situation.

[0103] By adjusting the initial weights of pollutant concentrations at all times, we can ensure that the model pays more attention to concentrations that are greatly affected by temperature during the prediction process, thereby improving the accuracy of the prediction.

[0104] Furthermore, the specific method of adjusting the combustion parameters or activating the pollutant control device is: feedback to the control system of the boiler combustion temperature, so that the control system outputs a control signal to adjust the combustion parameters, such as increasing the air supply or adjusting the fuel supply to ensure more complete combustion and reduce pollutants generated by incomplete combustion, or outputting a control signal to activate or adjust the selective catalytic reduction (SCR) system or the selective non-catalytic reduction (SNCR) system to ensure accurate control of pollutant concentrations. Ensure that the boiler combustion process meets production needs while minimizing the impact on the environment and achieving efficient and clean combustion.

[0105] Based on the same inventive concept as the above-mentioned method, an embodiment of the present application also provides a boiler combustion operation optimization system based on digital twin, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above-mentioned digital twin-based boiler combustion operation optimization methods.

[0106] It should be understood that although Figure 1The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0107] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0108] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the present application. It should be pointed out that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application without departing from the content of the technical solution of the present application, shall fall within the protection scope of the technical solution of the present application.

Claims

1. A boiler combustion operation optimization method based on digital twins, characterized in that: The method comprises the following steps: Real-time collection of temperature, pressure and pollutant concentration at every moment of boiler combustion; Select multiple moments before each moment and record them as the historical moments of each moment; Determine the combustion influencing factor at each moment based on the changes and dispersion of local pressure and local temperature at different historical moments, as well as the correlation between pressure and temperature; Determine the combustion efficiency factor at each moment based on the correlation between the temperature and pollutant concentration at each historical moment, and the difference in pollutant concentration under different temperature conditions, and in combination with the combustion influencing factor; Based on the combustion efficiency factor, an integrated learning algorithm is used for prediction to determine the changing trend of the predicted pollutant concentration. The digital twin model feeds back the boiler status to the control system to control the boiler combustion temperature.

2. The boiler combustion operation optimization method based on digital twin according to claim 1 is characterized in that: The determining of the combustion influencing factor at each moment includes: Taking the pressure and temperature at each historical moment as the center, respectively, a local pressure window and a local temperature window of a preset size are set; Analyze the pressure fluctuation changes in the local pressure window of each historical moment at each moment, determine the discrete degree of pressure at all historical moments in the local pressure window of each historical moment at each moment as the pressure fluctuation degree of each historical moment, and obtain each historical fluctuation moment at each moment; Analyze the difference in pressure and the degree of dispersion in the local temperature window of each historical fluctuation moment at each moment, and determine the local temperature difference of each historical fluctuation moment at each moment; Determine the local correlation of each historical fluctuation moment at each moment according to the correlation between the pressures at all historical moments in the local pressure window and the temperatures at all historical moments in the local temperature window at each moment; Determining a local influence coefficient of each historical fluctuation moment at each moment based on the pressure fluctuation degree, the local temperature difference and the local correlation degree; The average value of the local influence coefficients of all historical fluctuation moments at each moment is taken as the combustion influence factor at each moment.

3. The boiler combustion operation optimization method based on digital twin according to claim 2 is characterized in that: The obtaining of each historical fluctuation moment at each moment includes: The historical moments when the pressure fluctuation degree is greater than the preset pressure fluctuation threshold are recorded as each historical fluctuation moment at each moment.

4. The boiler combustion operation optimization method based on digital twin according to claim 2 is characterized in that: Determining the local temperature difference at each historical fluctuation moment at each moment includes: Determine the degree of dispersion of the temperature at all historical moments within the local temperature window of each historical moment at each moment as the temperature dispersion at each historical moment; The mean of the temperature dispersion at all historical moments at each moment is recorded as the first mean; The difference between the temperature dispersion at each historical fluctuation moment at each moment and the first mean value is determined as the local temperature difference at each historical fluctuation moment at each moment.

5. The boiler combustion operation optimization method based on digital twin according to claim 2, characterized in that: The local influence coefficient of each historical fluctuation moment at each moment is the product of the pressure fluctuation degree, the local temperature difference and the local correlation degree.

6. The boiler combustion operation optimization method based on digital twin according to claim 1, characterized in that: Determining the combustion efficiency factor at each moment includes: The partial correlation analysis algorithm is used to calculate the partial correlation coefficient between the temperature and pollutant concentration at all historical moments at each moment; Cluster the temperatures of all historical moments at each moment and obtain two clusters; Analyze the differences between pollutant concentrations corresponding to all historical moments in each cluster and calculate the first difference; Calculate the standard deviation of all temperatures in each cluster at the corresponding collection time; record the mean of the standard deviations of the two clusters as the second mean; Determine the temperature influence coefficient at each moment based on the partial correlation coefficient, the first difference and the second mean; The product of the temperature influence coefficient and the combustion influence factor is used as the combustion efficiency factor at each moment.

7. The boiler combustion operation optimization method based on digital twin according to claim 6 is characterized in that: The first difference is the difference between the mean values ​​of pollutant concentrations corresponding to all historical moments within the two clusters.

8. The boiler combustion operation optimization method based on digital twin according to claim 6, characterized in that: The temperature influence coefficient at each moment is the product of the partial correlation coefficient, the first difference and the second mean.

9. The boiler combustion operation optimization method based on digital twin according to claim 1, characterized in that: The determination of the change trend of the predicted pollutant concentration and the digital twin model feeding back the boiler status to the control system include: The combustion efficiency factor at each moment is normalized, and the normalized result is used as the concentration weight at each moment; According to the pollutant concentration at each moment and the concentration weight at the corresponding moment, an integrated learning algorithm is used for prediction; Perform linear fitting based on all predicted pollutant concentrations to obtain the slope of the fitted line; If the fitted slope is greater than the preset trend threshold, the pollutant concentration is on an upward trend, and the digital twin model provides feedback to the control system.

10. A boiler combustion operation optimization system based on digital twins, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the boiler combustion operation optimization method based on digital twin as described in any one of claims 1 to 9.

Images