Flue gas emission adjusting method and system for heating furnace
By constructing a machine learning model to dynamically predict pollutant concentration and adjust process parameters, the problem of heating furnace emission exceeding the standard is solved, and the pollutant emissions meet standards and the intelligent control of heating furnace operation is achieved.
Patent Information
- Application Number
- CN202310538095.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-12
- Publication Date
- 2024-11-12
AI Technical Summary
The sulfur dioxide and nitrogen oxide emitted by the heating furnace during operation often exceed the standard, mainly due to the unstable fuel properties and operation, and the existing technology is difficult to achieve automatic optimization and regulation, resulting in the pollutant emissions not meeting the standard.
By building a machine learning model based on historical process parameters, we dynamically predict pollutant concentrations and adjust process parameters when the standards are not met, we use the second relational model to generate a preset process parameter adjustment plan to adjust flue gas emissions in real time to ensure pollutant concentrations Meet the standards.
It realizes automatic adjustment of heating furnace fuel and operating parameters, ensures pollutant emissions meet standards, realizes real-time monitoring, intelligent perception and control, reduces the risk of exceeding the standard caused by delays in manual adjustment, has the characteristics of intelligence and informationization, and improves emissions Accuracy and timeliness.
Smart Images

Figure CN118936107A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of pollution control, and in particular relates to a fume emission regulating method and system for a heating furnace. Background Art
[0002] At present, the tail gas emissions of heating furnaces are subject to the Petroleum Refining Industry Pollutant Emission Standard (GB 31570-2015), and the sulfur dioxide standard limit in key areas is 50mg / m 3 、The standard limit of nitrogen oxides is 100mg / m 3 . However, according to the existing real-time data of pollutant emissions from heating furnaces in refining units, the actual emissions of sulfur dioxide and nitrogen oxides still exceed the standards. According to the analysis of the current pollutant emissions, the main factors causing the excessive emissions of pollutants from heating furnaces are the properties of the heating furnace fuel and the unstable operation of the heating furnace itself. Therefore, there is an urgent need for a method that can automatically optimize and adjust the heating furnace fuel and operating process parameters to ensure the stable operation of the heating furnace and achieve stable and standard pollutant emissions. Summary of the invention
[0003] In order to solve the above problems, an embodiment of the present invention proposes a flue gas emission regulation method for a heating furnace, comprising: obtaining historical process parameters of the heating furnace to be regulated, and constructing a first relationship model for predicting the concentrations of different pollutants in the flue gas emitted in the next time period based on the process parameters of the current time period, wherein the pollutants are one or more of sulfur dioxide and nitrogen oxides; obtaining the real-time process parameters of the heating furnace to be regulated, dynamically predicting the pollutant concentration using the first relationship model, and determining in real time whether the predicted pollutant concentration meets the standard; when it is determined that the standard is not met, adjusting the process parameters of the current heating furnace to be regulated according to a preset process parameter adjustment plan, thereby dynamically adjusting the flue gas emissions, wherein the preset process parameter adjustment plan is predicted based on the real-time process parameters and using a second relationship model, and the second relationship model is used to predict a plan for adjusting the process parameters of the current time period so that the pollutant concentration remains up to standard.
[0004] Preferably, the historical process parameters include but are not limited to gas pipeline network process parameters, heating furnace combustion process parameters and environmental protection facility process parameters.
[0005] Preferably, in the process of obtaining the historical process parameters of the heating furnace to be adjusted, it includes: screening and eliminating the data generated by the heating furnace in the shutdown state from the historical process parameters; using the preset process parameter threshold value that characterizes that the heating furnace to be adjusted is in the normal operation state, screening the normal operation data from the historical process parameters excluding the shutdown data to form new historical process parameters, and performing blank data supplementation processing for the current new historical process parameters; unifying the format of different types of parameters in the historical process parameters that have been supplemented, so as to construct the first relationship model using the historical process parameters with a unified format.
[0006] Preferably, the step of constructing the first relationship model includes: constructing a first preset model; using historical process parameters of different time periods as input data for model training, and using pollutant concentrations of different time periods as output data for model training, training the first preset model to obtain the first relationship model for predicting pollutant concentration, wherein the historical process parameters include but are not limited to furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate and amine liquid concentration.
[0007] Preferably, the first preset model is constructed based on a RBF neural network algorithm framework, and the RBF neural network algorithm uses a Gaussian function as a radial basis function.
[0008] Preferably, the step of constructing the first relationship model also includes: converting the historical process parameters into a first parameter characterizing the state of sulfur concentration to be discharged in the heating furnace to be adjusted, and using the first parameter to construct the first relationship model, wherein the first parameter is the ratio between the hydrogen sulfide concentration of the gas in the gas pipeline network process parameters and the sulfur dioxide emission concentration in the environmental protection facility process parameters.
[0009] Preferably, after constructing the first relationship model, the flue gas emission regulation method also includes: randomly selecting historical process parameters of multiple different time periods, and using the first relationship model currently constructed to generate predicted historical pollutant concentrations for each time period; extracting the actual pollutant concentrations of the multiple different time periods from the historical process parameters, and comparing and analyzing the predicted historical pollutant concentrations with the actual pollutant concentrations of the corresponding time periods, so as to determine whether the first relationship model currently constructed has the ability to predict pollutant concentrations, so as to dynamically predict pollutant concentrations using the first relationship model with the ability to predict pollutant concentrations, wherein when the mean square error between the predicted historical pollutant concentration and the actual pollutant concentration is less than a preset allowable error, it is determined that the first relationship model currently constructed has the ability to predict pollutant concentrations.
[0010] Preferably, the flue gas emission regulation method also includes: for different non-compliant pollutant concentrations, analyzing and comparing the first related parameters related to the compliance of pollutant concentration emissions in the historical process parameters, and the second related parameters related to the non-compliance of pollutant concentration emissions, so as to configure multiple process parameter adjustment schemes for each non-compliant pollutant concentration, respectively, to construct a second relationship model for generating the optimal process parameter adjustment scheme based on the predicted pollutant concentration, based on which, the optimal process parameter adjustment scheme is used as the preset process parameter adjustment scheme to dynamically adjust flue gas emissions.
[0011] Preferably, the process of constructing the second relationship model includes: constructing a second preset model; using the multiple process parameter adjustment schemes and substandard pollutant concentrations as input data for model training, and using the optimal process parameter adjustment scheme as output data for model training, training the second preset model to obtain the second relationship model.
[0012] Preferably, the second preset model is constructed based on a SVM algorithm framework.
[0013] Preferably, the flue gas emission regulation method also includes: acquiring the changing trend of the predicted pollutant concentration in real time, and analyzing the effective time of the process parameter adjustment according to the changing trend, so as to issue an early warning when the effective time is greater than the time between the current moment and the moment when the predicted pollutant concentration emission does not meet the standard.
[0014] On the other hand, the present invention also provides a flue gas emission regulation system for a heating furnace, the flue gas emission regulation system comprising the following modules: a model building module, which is used to obtain the historical process parameters of the heating furnace to be regulated, and construct a first relationship model for predicting the concentrations of different pollutants in the flue gas emitted in the next time period based on the process parameters of the current time period, wherein the pollutants are one or more of sulfur dioxide and nitrogen oxides; a pollutant concentration prediction module, which is used to obtain the real-time process parameters of the heating furnace to be regulated, and use the first relationship model to dynamically predict the pollutant concentration, and determine in real time whether the predicted pollutant concentration meets the standard; a process parameter adjustment module, which is used to adjust the process parameters of the current heating furnace to be regulated according to a preset process parameter adjustment plan when it is determined that the standard is not met, thereby dynamically adjusting the flue gas emissions, wherein the preset process parameter adjustment plan is predicted based on the real-time process parameters and using the second relationship model, and the second relationship model is used to predict a plan to adjust the process parameters of the current time period so that the pollutant concentration remains up to standard.
[0015] Compared with the prior art, one or more embodiments of the above scheme may have the following advantages or beneficial effects:
[0016] The present invention proposes a flue gas emission regulation method and system for a heating furnace. The method first utilizes the historical process parameters of the heating furnace to be regulated, dynamically predicts the pollutant concentration by constructing a machine learning model for predicting the pollutant concentration, and determines in real time whether the predicted pollutant concentration meets the standard. Then, when it is determined that it does not meet the standard, the process parameters of the current heating furnace to be regulated are adjusted according to the preset process parameter adjustment scheme, thereby dynamically adjusting the flue gas emission. At the same time, the historical process parameters of the heating furnace to be regulated are also utilized to predict the dynamic preset process parameter adjustment scheme corresponding to the predicted pollutant concentration that does not meet the standard by constructing a machine learning model for the preset process parameter adjustment scheme, thereby dynamically adjusting the process parameters of the heating furnace to be regulated so that the pollutant concentration remains up to standard. The present invention can automatically adjust the heating furnace fuel and related operating process parameters during the actual operation of the heating furnace, so that pollutants are discharged up to standard, and realize real-time monitoring, intelligent perception, intelligent decision-making and operation control and regulation of the operating status of the heating furnace. In addition, the present invention can analyze and adjust various situations of excessive flue gas from heating furnaces, and has the characteristics of high intelligence and informatization. While making accurate adjustments, it reduces the negative impact of continuous excessive emissions caused by manual inability to adjust excessive equipment in a timely manner, thereby achieving good results of accurate monitoring and timely emission reduction.
[0017] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0019] Figure 1 This is a step diagram of a flue gas emission regulation method for a heating furnace according to an embodiment of the present application.
[0020] Figure 2 This is a module block diagram of a flue gas emission control system for a heating furnace according to an embodiment of the present application. DETAILED DESCRIPTION
[0021] The following will describe the implementation methods of the present invention in detail with reference to the accompanying drawings and embodiments, so that the implementation process of how the present invention applies technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly. It should be noted that as long as there is no conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other, and the technical solutions formed are all within the protection scope of the present invention.
[0022] In addition, the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be performed in an order different from that shown here.
[0023] At present, the tail gas emissions of heating furnaces are subject to the Petroleum Refining Industry Pollutant Emission Standard (GB 31570-2015), and the sulfur dioxide standard limit in key areas is 50mg / m 3 、The standard limit of nitrogen oxides is 100mg / m 3 . However, according to the existing real-time data of pollutant emissions from heating furnaces in refining units, the actual emissions of sulfur dioxide and nitrogen oxides still exceed the standards. According to the analysis of the current pollutant emissions, the main factors causing the excessive emissions of pollutants from heating furnaces are the properties of the heating furnace fuel and the unstable operation of the heating furnace itself. Therefore, there is an urgent need for a method that can automatically optimize and adjust the heating furnace fuel and operating process parameters to ensure the stable operation of the heating furnace and achieve stable and standard pollutant emissions.
[0024] Therefore, in order to solve the above problems, an embodiment of the present invention proposes a method and system for regulating flue gas emissions of a heating furnace. The method first uses the historical process parameters of the heating furnace to be regulated, dynamically predicts the pollutant concentration by constructing a machine learning model for predicting the pollutant concentration, and determines in real time whether the predicted pollutant concentration meets the standard. Then, when it is determined that it does not meet the standard, the process parameters of the current heating furnace to be regulated are adjusted according to the preset process parameter adjustment scheme, thereby dynamically adjusting the flue gas emissions. At the same time, the historical process parameters of the heating furnace to be regulated are also used to predict the dynamic preset process parameter adjustment scheme corresponding to the predicted pollutant concentration that does not meet the standard by constructing a machine learning model for the preset process parameter adjustment scheme, thereby dynamically adjusting the process parameters of the heating furnace to be regulated so that the pollutant concentration remains up to standard. The present invention can automatically adjust the heating furnace fuel and related operating process parameters during the actual operation of the heating furnace, so that pollutants are discharged in compliance with the standard, and realize real-time monitoring, intelligent perception, intelligent decision-making and operation control adjustment of the operating status of the heating furnace. In addition, the present invention can analyze and adjust various situations of excessive flue gas from heating furnaces, and has the characteristics of high intelligence and informatization. While making accurate adjustments, it reduces the negative impact of continuous excessive emissions caused by manual inability to adjust excessive equipment in a timely manner, thereby achieving good results of accurate monitoring and timely emission reduction.
[0025] Embodiment 1
[0026] Figure 1 This is a step diagram of the flue gas emission control method for a heating furnace according to an embodiment of the present application. Figure 1 To illustrate the various steps of this method.
[0027] like Figure 1 As shown, in step S110, the historical process parameters of the heating furnace to be adjusted are obtained, and a first relationship model for predicting the concentration of different pollutants in the flue gas discharged in the next period is constructed based on the process parameters of the current period. This implementation uses historical process parameters to obtain the concentration law of different pollutants in the flue gas discharged by the heating furnace to be adjusted, and forms a first relationship model based on the analyzed concentration law of different pollutants. Specifically, first, the historical process parameters stored in the historical operation process of the heating furnace to be adjusted, which characterize the historical operating status of the heating furnace combustion system and the desulfurization system supporting the heating furnace combustion system (for example: internal heating, environmental status of the desulfurization area, etc.). Afterwards, the first relationship model is constructed using the obtained historical process parameters, so that the process parameters of the current period are input into the first relationship model to predict and obtain the pollutant concentration corresponding to different types of pollutants in the flue gas discharged in the next period. In actual applications, there are often multiple types of pollutants in the flue gas discharged by the heating furnace. The present invention has a high degree of adaptability to actual working conditions and can obtain different pollutant concentrations corresponding to different pollutant types that match the actual working conditions. In a specific embodiment of the present application, the pollutant is one or more of sulfur dioxide and nitrogen oxides.
[0028] In order to obtain accurate pollutant concentration prediction data, after obtaining the historical process parameters, this embodiment does not directly apply the currently obtained historical process parameters to construct the first relationship model, but instead converts the currently obtained historical process parameters into a standard format, eliminates abnormal data, and supplements blank data to screen effective historical process parameters, thereby improving the accuracy of the first relationship model in predicting pollutant concentrations.
[0029] In the process of obtaining the historical process parameters of the heating furnace to be adjusted, first, the data generated when the heating furnace is in a shutdown state is screened and eliminated from the historical process parameters; then, the preset process parameter threshold that characterizes that the heating furnace to be adjusted is used to filter the normal operating data from the historical process parameters that have eliminated the shutdown data to form new historical process parameters, and blank data supplementation is performed for the current new historical process parameters; finally, the different types of parameters in the historical process parameters that have been supplemented are formatted in a unified manner to construct a first relationship model using the historical process parameters in a unified format.
[0030] In the process of processing the currently obtained historical process parameters, this embodiment first screens and removes the data generated by the heating furnace in the shutdown state from the historical process parameters, thereby realizing the data extraction of the historical process parameters of the operating condition section excluding the shutdown condition section. Next, the preset process parameter threshold is compared and analyzed with the historical process parameters of the operating condition section, and the historical process parameters within the preset process parameter threshold range that characterizes the heating furnace to be adjusted in a normal operating state are extracted, so as to screen the normal operating data from the historical process parameters (historical process parameters of the operating condition section) excluding the shutdown data to form new historical process parameters, and perform blank data supplementation processing for the current new historical process parameters through corresponding technical means (for example: using the historical process parameters collected by a data collection device different from the data collection device corresponding to the acquisition channel of the current new historical process parameters to fill the data gap of the corresponding period in the current new historical process parameters, etc.). In the present invention, the first relationship model is used to construct different types of process parameters from the heating furnace to be adjusted and its supporting system, and the data format of each type of process parameter is not exactly the same. Therefore, this embodiment unifies the data format of different types of parameters in the historical process parameters that have been supplemented, so as to construct a first relationship model using the historical process parameters with a unified format. In this way, after the first relationship model outputs the pollutant concentration prediction result, there is no need to perform format-related processing on the prediction result, and the obtained predicted decontamination concentration can be directly applied, which effectively improves the data processing efficiency.
[0031] Based on the design concept of using different types of process parameters from the heating furnace to be adjusted and its supporting system for constructing the first relationship model, in a specific embodiment of the present application, the historical process parameters include but are not limited to gas pipe network process parameters, heating furnace combustion process parameters and environmental protection facility process parameters.
[0032] In the step of constructing the first relationship model, the first preset model is first constructed; then, the historical process parameters of different time periods are used as input data for model training, and the pollutant concentrations of different time periods are used as output data for model training, and the first preset model is trained to obtain a first relationship model for predicting pollutant concentrations. Specifically, this embodiment is based on machine learning technology to construct a first preset model composed of a neural network algorithm framework. Afterwards, the historical process parameters of different time periods are used as input information of a preset machine learning model (first preset model), substituted into the first preset model, and the first preset model is trained. After the training is completed, the first relationship model is obtained. The first relationship model can calculate the pollutant concentrations of different time periods after the corresponding time period of the historical process parameters, generate the pollutant concentration prediction results of the corresponding time period, and obtain the pollutant concentration prediction value of the corresponding time period. In a specific embodiment of the present application, the historical process parameters used to train the first preset model include but are not limited to the furnace temperature, the furnace oxygen content, the burner outlet pressure, the fuel flow rate, the fuel flow rate and the amine liquid concentration.
[0033] In the embodiment of the present application, the first preset model is constructed based on the RBF neural network algorithm framework, and the RBF neural network algorithm uses a Gaussian function as a radial basis function. Specifically, in the RBF neural network structure, let X = [x1, x2, ... x m ] is the input vector of the neural network, let B = [b1, b2, ... b m ] is the basis width variable of the neural network, and is combined with a Gaussian function as a radial basis function to form a function characterizing the RBF neural network algorithm of this embodiment, wherein the function characterizing the RBF neural network algorithm is expressed by the following expression:
[0034]
[0035] Among them, R j represents the jth output variable, ||xc j || represents the Euclidean norm, x represents the input variable, c j represents the center of the Gaussian function, b represents the basis width parameter of the node, and m represents the number of input variables.
[0036] Next, in the process of training the first preset model, the output value of the RBF neural network is expressed using the following expression:
[0037] y m (k)=wh=w1h1+w2h2+.....+w m h m (2)
[0038] Among them, w represents the weight between the hidden layer and the output layer, y m (k) represents the actual output, k represents the time corresponding to the output value, and h represents the output of the hidden layer.
[0039] Further, in the step of constructing the first relationship model, the present invention also converts the historical process parameters into the first parameter characterizing the state of sulfur concentration to be discharged in the heating furnace to be adjusted, and constructs the first relationship model using the first parameter, wherein the first parameter is the ratio between the concentration of hydrogen sulfide in the gas gas in the process parameters of the gas pipeline network and the sulfur dioxide emission concentration in the process parameters of the environmental protection facilities. In a specific embodiment of the present invention, the first relationship model is constructed directly using the historical process parameters that have not been pre-processed by calculation. This method has the advantage of convenient calculation and can achieve the purpose of constructing the first relationship model. In order to improve the accuracy of the prediction results of sulfur dioxide concentration in pollutants, in another embodiment of the present invention, the first relationship model is constructed using the historical process parameters that have been pre-processed by calculation. Among them, in the process of calculation pre-processing, the process parameters with correlation in the historical process parameters used to construct the first relationship model are subjected to key parameter analysis, and the key parameters with strong correlation (sulfur concentration change index) that can accurately characterize the state of sulfur concentration to be discharged in the heating furnace to be adjusted are mined to construct the first relationship model, thereby effectively improving the accuracy of the pollutant concentration prediction results. It is known from the analysis that there is a correlation between the concentration of hydrogen sulfide in the gas pipeline network process parameters and the concentration of sulfur dioxide emission in the process parameters of the environmental protection treatment facilities. Therefore, in this embodiment, the ratio between the concentration of hydrogen sulfide in the gas pipeline network process parameters and the concentration of sulfur dioxide emission in the process parameters of the environmental protection treatment facilities is used as the first parameter (key parameter with strong correlation). The first parameter can effectively reflect the operation and desulfurization details of the heating furnace combustion system and the environmental protection treatment facilities supporting the heating furnace combustion system. Moreover, according to the Pearson parameter verification, the first parameter has a high correlation with the sulfur dioxide emission concentration at the outlet of the washing tower of the environmental protection treatment facility. Therefore, the first parameter can improve the prediction accuracy of the sulfur dioxide emission concentration.
[0040] After constructing the first relationship model, the present invention also determines the pollutant prediction capability of the constructed first relationship model to ensure that the current first relationship model can be used to predict pollutant concentrations. In the process of determining the pollutant prediction capability of the constructed first relationship model, historical process parameters of multiple different time periods are first randomly selected, and the predicted historical pollutant concentration of each time period is generated using the currently constructed first relationship model; then, the actual pollutant concentrations of multiple different time periods are extracted from the historical process parameters, and the predicted historical pollutant concentrations are compared and analyzed with the actual pollutant concentrations of the corresponding time periods, so as to determine whether the currently constructed first relationship model has the ability to predict pollutant concentrations, so as to dynamically predict pollutant concentrations using the first relationship model with the ability to predict pollutant concentrations, wherein when the mean square error between the predicted historical pollutant concentration and the actual pollutant concentration is less than the preset allowable error, it is determined that the currently constructed first relationship model has the ability to predict pollutant concentrations.
[0041] Specifically, this embodiment first randomly selects historical process parameters of multiple different time periods from the historical process parameters used to construct the first relationship model, and uses the currently constructed first relationship model to calculate the predicted historical pollutant concentration corresponding to each time period, that is, the current pollutant concentration is directly generated by the first relationship model. Then, the actual pollutant concentrations of multiple different time periods corresponding to the predicted historical pollutant concentrations are extracted from the historical process parameters, that is, the current pollutant concentration is essentially obtained by collecting data collected by the corresponding data collection device at the construction site. After obtaining the pollutant concentration data from different acquisition methods for the corresponding time period, this embodiment compares and analyzes the predicted historical pollutant concentration with the actual pollutant concentration for the corresponding time period to verify the output value of the first relationship model, thereby determining whether the currently constructed first relationship model has the ability to predict pollutant concentrations, so as to use the first relationship model with the ability to predict pollutant concentrations to dynamically predict pollutant concentrations and ensure the reliability of the prediction results. Among them, when the mean square error between the predicted historical pollutant concentration and the actual pollutant concentration is less than the preset allowable error, it is determined that the currently constructed first relationship model has the ability to predict pollutant concentrations. In the embodiment of the present application, when the output value (predicted pollutant concentration) y at time k m When the mean square error between y(k) and the actual value (actual pollutant concentration) y(k) at time k is less than the allowable error, it is proved that the output value (predicted pollutant concentration) is available, and further proves that the first relationship model is available, wherein the mean square error is calculated using the following expression:
[0042] N(k)=1 / 2[y(k)-y m (k)] 2 (3)
[0043] Where N(k) represents the mean square error.
[0044] In step S120, the real-time process parameters of the heating furnace to be adjusted are obtained, and the pollutant concentration is dynamically predicted using the first relationship model, and it is determined in real time whether the predicted pollutant concentration meets the standard. First, the real-time process parameters generated by the heating furnace to be adjusted during the current operation process, which represent the real-time operating status of the heating furnace combustion system and the desulfurization system and other systems supporting the heating furnace combustion system, are obtained, and the real-time process parameters are converted to a standard format, abnormal data is eliminated, and blank data is supplemented in a similar manner to the historical process parameters to screen effective real-time process parameters, so as to further improve the accuracy of the obtained pollutant concentration prediction results. Then, the processed real-time process parameters are used as the input information of the first relationship model to obtain the pollutant concentration prediction result, thereby realizing the dynamic prediction of the pollutant concentration. Finally, this embodiment presets a compliance threshold that represents that the corresponding type of pollutant concentration is in a compliance state, and by comparing the real-time predicted pollutant concentration with the compliance threshold, the purpose of real-time determination of whether the predicted pollutant concentration meets the standard is achieved.
[0045] In step S130, when it is determined that the standard is not met, the process parameters of the current heating furnace to be adjusted are adjusted according to the preset process parameter adjustment scheme, thereby dynamically adjusting the flue gas emission. Specifically, the present embodiment presets a preset process parameter adjustment scheme for adjusting the concentration of pollutants that do not meet the standard to meet the standard, and when it is determined that the standard is not met, the process parameters of the current heating furnace to be adjusted are adjusted using the preset process parameter adjustment scheme, thereby dynamically adjusting the flue gas emission so that the concentration of the pollutants emitted is always the standard concentration. In the embodiment of the present application, the preset process parameter adjustment scheme is predicted based on the real-time process parameters and using the second relationship model, and the second relationship model is used to predict the scheme for adjusting the process parameters of the current time period so that the pollutant concentration remains up to standard. Accordingly, the present invention, while obtaining accurate pollutant concentration prediction results, also instantly generates and executes a preset process parameter adjustment scheme suitable for adjusting the corresponding non-standard pollutant concentration, thereby achieving timely control of the pollutant concentration exceeding the standard emission condition to be occurred, and effectively avoiding the occurrence of pollutant concentration exceeding the standard emission situation.
[0046] Further, the present invention analyzes and compares the first relevant parameters related to the pollutant concentration emission reaching the standard and the second relevant parameters related to the pollutant concentration emission not reaching the standard in the historical process parameters for different non-standard pollutant concentrations, so as to configure a plurality of process parameter adjustment schemes for each non-standard pollutant concentration, so as to construct a second relationship model for generating the best process parameter adjustment scheme according to the predicted pollutant concentration, based on which, the best process parameter adjustment scheme is used as the preset process parameter adjustment scheme to dynamically adjust the flue gas emission. In this embodiment, the historical process parameters are divided into the first relevant parameters related to the pollutant concentration emission reaching the standard and the second relevant parameters related to the pollutant concentration emission not reaching the standard, and then, for different representative pollutant concentrations, the first relevant parameters and the second relevant parameters are analyzed and compared to obtain the parameters such as the data difference in the two types of parameters, and then, according to the aforementioned data difference and other parameters, for each non-standard pollutant concentration, a plurality of process parameter adjustment schemes that can adjust the non-standard pollutant concentration to the standard pollutant for emission are matched. Finally, a plurality of process parameter adjustment schemes for each non-standard pollutant concentration are used to construct a second relationship model for generating the best process parameter adjustment scheme according to the predicted pollutant concentration, so as to predict the best process parameter adjustment scheme that can obtain the best adjustment effect. In this way, this embodiment uses the optimal process parameter adjustment scheme as the preset process parameter adjustment scheme to dynamically adjust the flue gas emissions, thereby obtaining the best adjustment effect, thereby achieving all-round accurate monitoring and timely emission reduction of pollutants.
[0047] In the process of constructing the second relationship model, the second preset model is first constructed; then, a variety of process parameter adjustment schemes and non-standard pollutant concentrations are used as input data for model training, and the optimal process parameter adjustment scheme is used as output data for model training to train the second preset model and obtain the second relationship model. Specifically, this embodiment is based on machine learning technology to construct a second preset model composed of a classification algorithm framework. Afterwards, a variety of process parameter adjustment schemes and non-standard pollutant concentrations are used as input information of a preset machine learning model (second preset model), substituted into the second preset model, and training of the second preset model is started. After the training is completed, the second relationship model is obtained. The second relationship model can optimize the process parameter adjustment schemes used to adjust the corresponding non-standard pollutant concentrations and obtain the corresponding optimal process parameter adjustment schemes.
[0048] In the embodiment of the present application, the second preset model is constructed based on the SVM (support vector machine) algorithm framework.
[0049] Furthermore, this embodiment also analyzes the prediction accuracy of the first relationship model and the second relationship model based on historical process parameters, and adjusts the prediction accuracy of the corresponding relationship model based on the analysis results, thereby effectively improving the applicability of the constructed first relationship model and the second relationship model in the actual construction site.
[0050] Further, the present invention also obtains the change trend of the predicted pollutant concentration in real time, and analyzes the effective time of the process parameter adjustment according to the change trend, so as to give an early warning when the effective time satisfies the time between the current moment and the predicted pollutant concentration emission not reaching the standard. In practical applications, after the process parameters are adjusted, the change of the pollutant concentration requires a certain reaction time. Therefore, the present embodiment collects each real-time predicted pollutant concentration and obtains the change trend of the predicted pollutant concentration, and then obtains the inferred reaction time to reach each pollutant concentration according to the historical process parameters. Then, based on the inferred reaction time, the effective time of the process parameter adjustment is analyzed in combination with the change trend to determine whether it can be adjusted in time before the flue gas emission with the predicted pollutant concentration not reaching the standard is discharged, and when the effective time satisfies the time between the current moment and the predicted pollutant concentration emission not reaching the standard (when the adjustment cannot be made in time), an early warning is given and the currently generated optimal process parameter adjustment plan is output, and the corresponding personnel are instructed to take certain measures to achieve the expected adjustment effect through other adjustment methods in the form of instant messaging such as text messages or emails.
[0051] Embodiment 2
[0052] Based on the smoke emission regulating method for a heating furnace described in the above-mentioned embodiment 1, an embodiment of the present invention further provides a smoke emission regulating system for a heating furnace (hereinafter referred to as “smoke emission regulating system”). Figure 2 This is a module block diagram of a flue gas emission control system for a heating furnace according to an embodiment of the present application.
[0053] like Figure 2As shown, the flue gas emission control system in the embodiment of the present invention includes: a model building module 21, a pollutant concentration prediction module 22 and a process parameter adjustment module 23. Specifically, the model building module 21 is implemented according to the method described in step S110 above, and is configured to obtain the historical process parameters of the heating furnace to be adjusted, and to construct a first relationship model for predicting the concentrations of different pollutants in the flue gas discharged in the next period according to the process parameters of the current period, wherein the pollutants are one or more of sulfur dioxide and nitrogen oxides; the pollutant concentration prediction module 22 is implemented according to the method described in step S120 above, and is configured to obtain the real-time process parameters of the heating furnace to be adjusted, and dynamically predict the pollutant concentration using the first relationship model, and determine in real time whether the predicted pollutant concentration meets the standard; the process parameter adjustment module 23 is implemented according to the method described in step S130 above, and is configured to adjust the process parameters of the current heating furnace to be adjusted according to the preset process parameter adjustment plan when it is determined that the standard is not met, so as to dynamically adjust the flue gas emission, wherein the preset process parameter adjustment plan is predicted according to the real-time process parameters and using the second relationship model, and the second relationship model is used to predict the plan for adjusting the process parameters of the current period so that the pollutant concentration remains up to standard.
[0054] Embodiment 3
[0055] In the embodiment of the present application, the flue gas emission regulation method for a heating furnace described in the first embodiment is applied to a heating furnace in an actual refining environment. The heating furnace uses a gas pipe network to supply gas, the high-pressure gas pipe network pressure is 0.5Mpa, the low-pressure gas pipe network pressure is 0.1Mpa, and the gas output is 40,000t / a. Among them, the main production unit for producing gas is a catalytic cracking unit, and a small part comes from a polypropylene unit.
[0056] First, the furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate, amine liquid concentration and the corresponding sulfur concentration change index are used as the input data of the first relationship model, and the predicted sulfur dioxide concentration is used as the output data of the first relationship model. At the same time, the furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate, amine liquid concentration and a variety of process parameter adjustment schemes, and non-standard pollutant concentrations are used as input data of the second relationship model, and the predicted preset process parameter adjustment scheme together with the standard pollutant concentration and its corresponding furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate and amine liquid concentration are used as output data of the second relationship model. The adjustment results shown in Table 1 are obtained:
[0057] Table 1
[0058]
[0059]
[0060] Secondly, the furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate and amine liquid concentration are used as input data of the first relationship model, and the predicted sulfur dioxide concentration is used as the output data of the first relationship model. At the same time, the furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate, amine liquid concentration and a variety of process parameter adjustment schemes, and non-standard pollutant concentrations are used as input data of the second relationship model, and the predicted preset process parameter adjustment scheme together with the standard pollutant concentration and its corresponding furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate and amine liquid concentration are used as output data of the second relationship model. The adjustment results shown in Table 2 are obtained:
[0061] Table 2
[0062] Serial number Average error of predicted sulfur dioxide emission concentration Number of times sulfur dioxide emission concentration exceeded the standard date 1 6.26% 1 February 4, 2023 2 6.33% 2 February 5, 2023 3 7.56% 2 February 6, 2023
[0063] Finally, a heating furnace that did not implement the flue gas emission control method for the heating furnace was used as a control group. The number of times exceeding the standard occurred was obtained as shown in Table 3:
[0064] Table 3
[0065] Serial number Number of times sulfur dioxide emission concentration exceeded the standard date 1 3 February 7, 2023 2 4 February 8, 2023 3 2 February 9, 2023
[0066] It can be seen that the flue gas emission regulation method for a heating furnace according to the present invention directly uses the historical process parameters that have not been pre-processed by calculation to construct the first relationship model, and the sulfur dioxide emission concentration prediction accuracy is the highest. When the historical process parameters that have been pre-processed by calculation are used to construct the first relationship model for parameter regulation, there is no exceeding data within three days. Therefore, the present invention effectively improves the prediction accuracy of sulfur dioxide emission concentration and reduces the number of exceeding the sulfur dioxide emission concentration.
[0067] Embodiment 4
[0068] In a specific embodiment of the present application, the concentration of nitrogen oxides at the flue gas outlet of the heating furnace is monitored and obtained in real time. When the concentration of nitrogen oxides exceeds the standard, is about to exceed the standard or has exceeded the standard, the real-time process parameters related to the internal environment of the furnace (such as: the temperature in the furnace, the oxygen content and the flow rate and velocity of the fuel in the furnace) are used as the input data of the first relationship model, and the corresponding real-time process parameters are used as the input data of the second relationship model to obtain the pressure and viscosity of the fuel at the burner outlet, the load of the heating furnace (affecting the temperature in the furnace), the air volume of the mixed fuel (affecting the oxygen content in the furnace), and the optimal preset process parameter adjustment scheme for adjusting the concentration of nitrogen oxides, so as to control the heating furnace to operate according to the current scheme in real time. When the concentration of nitrogen oxides is controlled within the limit, it is prompted that the exceeding of the standard is over; when the concentration of nitrogen oxides does not reach the standard within the effective time, it is prompted that the exceeding of the standard is pre-warning until the nitrogen oxides reach the standard emission.
[0069] Embodiment 5
[0070] In a specific embodiment of the present application, the concentration of hydrogen sulfide at the outlet of the desulfurization system is monitored and obtained in real time. When the hydrogen sulfide has exceeded the standard or is about to exceed the standard, the real-time process parameters related to the internal environment of the furnace (such as: the mixing ratio of natural gas and gas and the flow rate of fuel) are used as the input data of the first relationship model, and the corresponding real-time process parameters are used as the input data of the second relationship model to obtain the mixing ratio of the fuel (the concentration and flow ratio of natural gas and gas respectively), and the optimal preset process parameter adjustment scheme for adjusting the concentration of hydrogen sulfide, so as to control the mixing ratio and flow rate of the fuel in real time according to the current scheme to appropriately reduce the concentration of gas, thereby achieving the purpose of reducing sulfur dioxide emissions at the source. When the hydrogen sulfide concentration at the outlet of the desulfurization system is controlled within the limit, it is prompted that the exceeding standard is over; when the concentration of hydrogen sulfide does not reach the standard within the effective time, it is prompted that the exceeding standard is warned until the hydrogen sulfide reaches the standard emission.
[0071] The present invention discloses a method and system for regulating flue gas emissions of a heating furnace. The method first utilizes the historical process parameters of the heating furnace to be regulated, dynamically predicts the pollutant concentration by constructing a machine learning model for predicting the pollutant concentration, and determines in real time whether the predicted pollutant concentration meets the standard. Then, when it is determined that it does not meet the standard, the process parameters of the current heating furnace to be regulated are adjusted according to the preset process parameter adjustment scheme, thereby dynamically adjusting the flue gas emissions. At the same time, the historical process parameters of the heating furnace to be regulated are also utilized, and a machine learning model for the preset process parameter adjustment scheme is constructed to predict the dynamic preset process parameter adjustment scheme corresponding to the predicted pollutant concentration that does not meet the standard, thereby dynamically adjusting the process parameters of the heating furnace to be regulated so that the pollutant concentration remains up to standard. The present invention can automatically adjust the heating furnace fuel and related operating process parameters during the actual operation of the heating furnace, so that pollutants are discharged up to standard, and realize real-time monitoring, intelligent perception, intelligent decision-making and operation control adjustment of the operating status of the heating furnace. In addition, the present invention can analyze the excessive flue gas conditions of various heating furnaces, and promptly give corresponding optimal parameter adjustments to adjust the excessive status. It has the characteristics of intelligence, rapid response, and high degree of informatization. While making accurate adjustments, it reduces the negative impact of continuous excessive emissions due to manual inability to adjust the excessive equipment in a timely manner, and achieves the good results of accurate monitoring and timely emission reduction. Without adding or replacing pollution control facilities, it achieves the purpose of reducing the generation and emission of pollutants such as nitrogen oxides and sulfur dioxide from the source and in the production process.
[0072] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any person familiar with the technology within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
[0073] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims of the present invention.
[0074] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
[0075] Although the embodiments disclosed in the present invention are as above, the above contents are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.
Claims
1. A method for regulating flue gas emissions from a heating furnace, characterized in that: include: Obtaining historical process parameters of the heating furnace to be adjusted, and constructing a first relationship model for predicting the concentrations of different pollutants in the flue gas emitted in the next period according to the process parameters of the current period, wherein the pollutants are one or more of sulfur dioxide and nitrogen oxides; Acquire the real-time process parameters of the heating furnace to be adjusted, dynamically predict the pollutant concentration using the first relationship model, and determine in real time whether the predicted pollutant concentration meets the standard; When it is determined that the standards are not met, the process parameters of the current heating furnace to be adjusted are adjusted according to a preset process parameter adjustment plan, thereby dynamically adjusting the flue gas emissions, wherein the preset process parameter adjustment plan is predicted based on the real-time process parameters and using a second relationship model, and the second relationship model is used to predict a plan to adjust the process parameters of the current time period so that the pollutant concentration remains up to standard.
2. The flue gas emission control method according to claim 1, characterized in that: The historical process parameters include, but are not limited to, gas pipeline network process parameters, heating furnace combustion process parameters, and environmental protection facility process parameters.
3. The flue gas emission control method according to claim 2, characterized in that: The process of obtaining the historical process parameters of the heating furnace to be adjusted includes: Filter and eliminate data generated when the heating furnace is in a shutdown state from the historical process parameters; Using a preset process parameter threshold value that indicates that the heating furnace to be adjusted is in a normal operating state, normal operating data is selected from historical process parameters excluding shutdown data to form new historical process parameters, and blank data supplement processing is performed for the current new historical process parameters; Different types of parameters in the historical process parameters that have been supplemented are processed in a unified format, so as to construct the first relationship model using the historical process parameters in a unified format.
4. The flue gas emission control method according to claim 2 or 3, characterized in that: The step of constructing the first relationship model includes: constructing a first preset model; The first preset model is trained by using the historical process parameters of different time periods as input data for model training, and the pollutant concentrations of different time periods as output data for model training, to obtain the first relationship model for predicting pollutant concentrations, wherein the historical process parameters include but are not limited to furnace temperature, furnace oxygen content, burner outlet pressure, fuel flow rate, fuel flow rate and amine liquid concentration.
5. The flue gas emission control method according to claim 4, characterized in that: The first preset model is constructed based on the RBF neural network algorithm framework, and the RBF neural network algorithm uses a Gaussian function as a radial basis function.
6. The flue gas emission control method according to any one of claims 2 to 5, characterized in that: The step of constructing the first relationship model further includes: The historical process parameters are converted into a first parameter characterizing the state of sulfur concentration to be discharged in the heating furnace to be adjusted, and the first relationship model is constructed using the first parameter, wherein the first parameter is the ratio between the hydrogen sulfide concentration of the gas in the gas pipeline network process parameters and the sulfur dioxide emission concentration in the environmental protection facility process parameters.
7. The flue gas emission control method according to any one of claims 1 to 6, characterized in that: After constructing the first relationship model, the flue gas emission regulation method further includes: Randomly select historical process parameters of multiple different time periods, and use the first relationship model currently constructed to generate predicted historical pollutant concentrations for each time period; The actual pollutant concentrations of the multiple different time periods are extracted from the historical process parameters, and the predicted historical pollutant concentrations are compared and analyzed with the actual pollutant concentrations of the corresponding time periods, so as to determine whether the first relationship model currently constructed has the ability to predict pollutant concentrations, so as to dynamically predict pollutant concentrations using the first relationship model with the ability to predict pollutant concentrations, wherein: When the mean square error between the predicted historical pollutant concentration and the actual pollutant concentration is less than a preset allowable error, it is determined that the first relationship model currently constructed has the ability to predict pollutant concentration.
8. The flue gas emission control method according to any one of claims 1 to 7, characterized in that: The flue gas emission regulation method further comprises: For different non-compliant pollutant concentrations, the first related parameters related to the compliance of pollutant concentration emissions and the second related parameters related to the non-compliance of pollutant concentration emissions in the historical process parameters are analyzed and compared, so as to configure multiple process parameter adjustment schemes for each non-compliant pollutant concentration, respectively, to construct a second relationship model for generating the best process parameter adjustment scheme based on the predicted pollutant concentration, based on which, the best process parameter adjustment scheme is used as the preset process parameter adjustment scheme to dynamically adjust flue gas emissions.
9. The flue gas emission control method according to claim 8, characterized in that: The process of constructing the second relationship model includes: constructing a second preset model; The plurality of process parameter adjustment schemes and substandard pollutant concentrations are used as input data for model training, and the optimal process parameter adjustment scheme is used as output data for model training to train the second preset model and obtain the second relationship model.
10. The flue gas emission control method according to claim 9, characterized in that: The second preset model is constructed based on the SVM algorithm framework.
11. The flue gas emission control method according to any one of claims 1 to 10, characterized in that: The flue gas emission regulation method further comprises: The changing trend of the predicted pollutant concentration is obtained in real time, and the effective time of the process parameter adjustment is analyzed according to the changing trend, so as to issue an early warning when the effective time is greater than the time between the current moment and the moment when the predicted pollutant concentration emission does not meet the standard.
12. A flue gas emission control system for a heating furnace, characterized in that: The flue gas emission control system includes the following modules: A model building module, which is used to obtain historical process parameters of the heating furnace to be adjusted, and to build a first relationship model for predicting the concentrations of different pollutants in the flue gas discharged in the next period according to the process parameters of the current period, wherein the pollutants are one or more of sulfur dioxide and nitrogen oxides; A pollutant concentration prediction module, which is used to obtain the real-time process parameters of the heating furnace to be adjusted, dynamically predict the pollutant concentration using the first relationship model, and determine in real time whether the predicted pollutant concentration meets the standard; A process parameter adjustment module is used to adjust the process parameters of the current heating furnace to be adjusted according to a preset process parameter adjustment plan when it is determined that the standards are not met, so as to dynamically adjust the flue gas emissions, wherein the preset process parameter adjustment plan is predicted based on the real-time process parameters and using a second relationship model, and the second relationship model is used to predict a plan to adjust the process parameters of the current time period so that the pollutant concentration remains up to standard.
