An intelligent ammonia injection adjustment method for a boiler denitration control system

Abstract

The present invention provides an intelligent ammonia spraying adjustment method for a boiler denitration control system, which belongs to the field of gas separation technology, and includes: step 1: obtaining historical operation data and historical flue gas data, and obtaining historical denitration data; step 2: constructing an optimal denitration prediction model based on historical operation data, historical flue gas data, and historical denitration data; step 3: real-time collection of real-time operation data and real-time flue gas data, and real-time collection of real-time denitration data; step 4: determining the optimal ammonia spraying amount based on the real-time operation data; step 5: adjusting the ammonia spraying device based on the optimal ammonia spraying amount, monitoring the ammonia escape rate in real time, and correcting and adjusting the ammonia spraying device based on the ammonia escape rate and real-time flue gas data. Accurate ammonia spraying adjustment can be achieved, the denitration effect can be maximized, ammonia waste can be reduced, environmental pollution can be reduced, the intelligence and adaptability of the boiler denitration system can be improved, and the efficient and stable operation of the denitration control system can be ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of gas separation, and in particular to an intelligent ammonia spraying adjustment method for a boiler denitration control system. Background Art

[0002] In traditional boiler denitrification control systems, the regulation of ammonia injection usually relies on preset control rules or fixed operation strategies, lacking real-time monitoring and dynamic regulation of boiler operating status, flue gas composition, and denitrification effect. This traditional method may lead to ammonia waste or poor denitrification effect, especially when the boiler load changes greatly or the combustion process is complex, it is difficult to accurately adjust the ammonia injection amount.

[0003] With increasingly stringent environmental protection requirements, especially in the control of nitrogen oxide (NOx) emissions, boiler denitrification technology has gradually developed. Traditional denitrification technologies such as selective catalytic reduction (SCR) and selective non-catalytic reduction (SNCR) have been widely used, but the efficiency of these methods is often affected by boiler operating conditions. In order to improve the effect of the denitrification system and reduce the ammonia escape rate, intelligent control technology has emerged in recent years.

[0004] Therefore, the present invention provides an intelligent ammonia injection regulation method for a boiler denitrification control system. Summary of the invention

[0005] The present invention provides an intelligent ammonia spraying adjustment method for a boiler denitration control system, which constructs an optimal denitration prediction model by acquiring historical operation data, historical flue gas data, and historical denitration data, determines the optimal ammonia spraying amount based on the real-time operation data, the real-time flue gas data, the real-time denitration data, and the optimal denitration prediction model collected in real time, adjusts the ammonia spraying device according to the optimal ammonia spraying amount, monitors the ammonia escape rate in real time, and corrects and adjusts the ammonia spraying device according to the ammonia escape rate and the real-time flue gas data. It can achieve accurate ammonia spraying adjustment, maximize the denitration effect, reduce ammonia waste and reduce environmental pollution, improve the intelligence and adaptability of the boiler denitration system, ensure the efficient and stable operation of the denitration control system, and has a high degree of adaptability and energy-saving and environmental protection effects.

[0006] The present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, comprising:

[0007] Step 1: Obtain historical operation data during boiler combustion and historical flue gas data generated by boiler combustion, and obtain historical denitrification data of boiler denitrification;

[0008] Step 2: Based on the historical operation data, historical flue gas data and historical denitrification data, determine the corresponding first denitrification related vector, the historical operation vectors, the historical flue gas vectors and the historical denitrification vectors within all specified time periods, and construct an optimal denitrification prediction model, wherein the optimal denitrification prediction model takes the optimal ammonia injection amount as the output target;

[0009] Step 3: Real-time collection of real-time operation data during boiler combustion and real-time flue gas data generated by boiler combustion, and real-time collection of real-time denitrification data of boiler denitrification;

[0010] Step 4: Based on the real-time operation data, the real-time flue gas data, the real-time denitrification data and the optimal denitrification prediction model, determine the optimal ammonia injection amount based on the real-time operation data;

[0011] Step 5: Adjust the ammonia injection device based on the optimal ammonia injection amount, monitor the ammonia escape rate in real time, and make corrections and adjustments to the ammonia injection device based on the ammonia escape rate and real-time flue gas data.

[0012] According to the intelligent ammonia injection adjustment method of a boiler denitration control system provided by the present invention, historical operation data of the boiler combustion process is obtained, historical flue gas data generated by boiler combustion and historical denitration data of boiler denitration are obtained, including:

[0013] Obtain historical sub-operation data of the boiler combustion process and historical sub-flue gas data generated by the boiler combustion within multiple specified time periods, and obtain historical sub-denitrification data of boiler denitrification within multiple specified time periods;

[0014] Preprocess the historical sub-operation data, historical sub-flue gas data and historical sub-denitrification data within each specified time period;

[0015] The historical operation data is determined based on the historical sub-operation data within all specified time periods after preprocessing, the historical flue gas data is determined based on the historical sub-flue gas data within all specified time periods after preprocessing, and at the same time, the historical denitrification data is determined based on the historical sub-denitrification data within all specified time periods after preprocessing.

[0016] According to an intelligent ammonia injection adjustment method for a boiler denitration control system provided by the present invention, an optimal denitration prediction model is constructed, including:

[0017] Perform feature extraction on the historical sub-operation data within each specified time period after preprocessing to determine the historical operation vector within each specified time period. At the same time, perform feature extraction on the historical sub-smoke data within each specified time period after preprocessing to determine the historical smoke vector within each specified time period.

[0018] Based on the historical sub-denitrification data in each specified time period after pretreatment, determine the historical ammonia injection amount, ammonia escape rate and nitrogen oxide removal rate in each specified time period;

[0019] Determine a historical denitration vector in each specified time period based on the historical ammonia injection amount, ammonia escape rate, and nitrogen oxide removal rate in each specified time period;

[0020] Determine an operation-denitrification correlation matrix, a flue gas-denitrification correlation matrix, an operation-flue gas correlation matrix, and a period correlation vector within each specified time period;

[0021] Determining a first denitrification correlation vector based on the period correlation vectors within all specified time periods and the operation-flue gas correlation matrix;

[0022] The optimal denitrification prediction model is constructed based on the first denitrification related vector, the historical operation vectors within all specified time periods, the historical flue gas vectors and the historical denitrification vectors.

[0023] According to an intelligent ammonia injection adjustment method for a boiler denitration control system provided by the present invention, an operation-denitration correlation matrix within each specified time period is determined, including:

[0024] Performing correlation analysis on the historical operation vectors and the historical denitrification vectors in each specified time period, determining the correlation value of each operation feature in the historical operation vectors and each feature in the historical denitrification vectors in each specified time period, and determining the operation-desnitrification correlation matrix in each specified time period;

[0025] ;

[0026] in, represents the operation-denitrification correlation matrix in the t-th specified time period, represents the i-th operation feature in the historical operation vector within the t-th specified time period, represents the jth feature of the historical denitrification vector in the tth specified time period, It represents the first correlation value of the i-th operation feature in the historical operation vector and the j-th feature in the historical denitrification vector within the t-th specified time period calculated based on the first correlation function, and N1 represents the number of operation features in the historical operation vector.

[0027] According to an intelligent ammonia injection adjustment method for a boiler denitration control system provided by the present invention, a flue gas-denitration correlation matrix within each specified time period is determined, including:

[0028] Performing correlation analysis on the historical flue gas vector and the historical denitrification vector in each specified time period, determining the correlation value of each flue gas feature in the historical flue gas vector and each feature in the historical denitrification vector in each specified time period, and determining the flue gas-denitrification correlation matrix in each specified time period;

[0029] ;

[0030] in, represents the flue gas-denitrification correlation matrix in the t-th specified time period, represents the kth smoke feature in the historical smoke vector within the tth specified time period, It represents the second correlation value of the kth flue gas feature in the historical flue gas vector and the jth feature in the historical denitrification vector within the tth specified time period calculated based on the second correlation function, and N2 represents the number of flue gas features in the historical flue gas vector.

[0031] According to an intelligent ammonia injection adjustment method for a boiler denitration control system provided by the present invention, an operation-flue gas correlation matrix and a period correlation vector within each specified time period are determined, including:

[0032] Performing correlation analysis on the historical operation vectors and the historical smoke vectors within each specified time period, determining the correlation value of each operation feature in the historical operation vectors and each smoke feature in the historical smoke vectors within each specified time period, and determining the operation-smoke correlation matrix;

[0033] ;

[0034] in, represents the operation-smoke correlation matrix in the t-th specified time period, represents a third correlation value of the i-th operation feature in the historical operation vector and the k-th smoke feature in the historical smoke vector within the t-th specified time period calculated based on the third correlation function;

[0035] Determining a period correlation vector based on the operation-denitrification correlation matrix and the flue gas-denitrification correlation matrix within each specified time period;

[0036] ;

[0037] ;

[0038] ;

[0039] in, represents the period-dependent vector of the t-th specified time period, represents the operation correlation vector of the tth specified time period, represents the smoke correlation vector of the tth specified time period, represents the first correlation threshold, represents the second correlation threshold, represents the average value of the first correlation value of the operation-denitrification correlation matrix for the t-th specified time period, represents the standard deviation of the first correlation value of the operation-denitrification correlation matrix for the t-th specified time period, , represents the average value of the second correlation value of the flue gas-denitrification correlation matrix in the t-th specified time period, represents the standard deviation of the second correlation value of the flue gas-denitrification correlation matrix for the t-th specified time period, They respectively represent the 1st related feature, the oth related feature, and the Nuth related feature in the periodic correlation vector of the tth specified time period, and tNu indicates that the number of related features of the tth specified time period is Nu.

[0040] According to an intelligent ammonia injection adjustment method for a boiler denitration control system provided by the present invention, a first denitration correlation vector is determined based on the period correlation vectors within all specified time periods and the operation-flue gas correlation matrix, including:

[0041] Determining a first weight of each correlation feature in each specified time period based on the period correlation vector in each specified time period and the operation-smoke correlation matrix;

[0042] ;

[0043] ;

[0044] ;

[0045] in, represents the first weight of the oth correlation feature in the periodic correlation vector of the tth specified time period, Represents the oth correlation feature in the periodic correlation vector of the tth specified time period The first indicator function of Represents the oth correlation feature in the periodic correlation vector of the tth specified time period The second indicator function of

[0046] Determine a second denitration correlation vector and a second weight of each correlation feature of the second denitration correlation vector based on the first weights of the period correlation vectors in all specified time periods and all correlation features in all specified time periods;

[0047] ;

[0048] ;

[0049] , ;

[0050] , ;

[0051] in, represents the second denitrification related vector under T specified time periods, They represent the 1st related feature, the pth related feature, and the Nmth related feature of the second denitrification related vector respectively, A second weight representing a p-th related feature of a second denitrification related vector; represents a function for fusing period-related vectors of T specified time periods; Represents the characteristic correlation analysis function of the t-th specified time period; represents the characteristic consistent analysis function of the t-th specified time period;

[0052] Based on the second weights of all relevant features of the second denitration related vector, all relevant features in the second denitration related vector are sequentially adjusted to determine the first denitration related vector. According to an intelligent ammonia injection adjustment method for a boiler denitration control system provided by the present invention, based on real-time operation data, real-time flue gas data, real-time denitration data and an optimal denitration prediction model, an optimal ammonia injection amount based on real-time operation data is determined, including:

[0053] Feature extraction is performed on real-time operation data, real-time flue gas data and real-time denitrification data respectively to determine a real-time operation vector, a real-time flue gas vector and a real-time denitrification vector;

[0054] The real-time operation vector, the real-time flue gas vector and the real-time denitrification vector are input into the optimal denitrification prediction model, and the optimal ammonia injection amount is output based on the optimal denitrification prediction model.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The best denitrification prediction model is constructed by acquiring historical operation data, historical flue gas data, and historical denitrification data. The best ammonia injection amount based on the real-time operation data, real-time flue gas data, real-time denitrification data, and the best denitrification prediction model are determined based on the real-time operation data. The ammonia injection device is adjusted according to the best ammonia injection amount, the ammonia escape rate is monitored in real time, and the ammonia injection device is corrected and adjusted according to the ammonia escape rate and real-time flue gas data. It can achieve precise ammonia injection adjustment, maximize the denitrification effect, reduce ammonia waste and reduce environmental pollution, improve the intelligence and adaptability of the boiler denitrification system, ensure the efficient and stable operation of the denitrification control system, and have a high degree of adaptability and energy-saving and environmental protection effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 It is a flow chart of an intelligent ammonia injection adjustment method for a boiler denitration control system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] Embodiment 1:

[0061] The embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, such as Figure 1 As shown, including:

[0062] Step 1: Obtain historical operation data during boiler combustion and historical flue gas data generated by boiler combustion, and obtain historical denitrification data of boiler denitrification;

[0063] Step 2: Based on the historical operation data, historical flue gas data and historical denitrification data, determine the corresponding first denitrification related vector, the historical operation vectors, the historical flue gas vectors and the historical denitrification vectors within all specified time periods, and construct an optimal denitrification prediction model, wherein the optimal denitrification prediction model takes the optimal ammonia injection amount as the output target;

[0064] Step 3: Real-time collection of real-time operation data during boiler combustion and real-time flue gas data generated by boiler combustion, and real-time collection of real-time denitrification data of boiler denitrification;

[0065] Step 4: Based on the real-time operation data, the real-time flue gas data, the real-time denitrification data and the optimal denitrification prediction model, determine the optimal ammonia injection amount based on the real-time operation data;

[0066] Step 5: Adjust the ammonia injection device based on the optimal ammonia injection amount, monitor the ammonia escape rate in real time, and make corrections and adjustments to the ammonia injection device based on the ammonia escape rate and real-time flue gas data.

[0067] In this embodiment, based on historical operation data, historical flue gas data and historical denitration data, an optimal denitration prediction model is constructed using statistical and machine learning methods. The model is used to predict the denitration effect under different operating conditions and determine the optimal ammonia injection amount.

[0068] In this embodiment, the operation data, flue gas data and denitration data of the boiler are collected in real time to monitor the operation status, flue gas composition and denitration effect of the boiler in real time. These real-time data are the basis for ensuring the accuracy of regulation.

[0069] In this embodiment, the optimal ammonia injection amount is determined based on real-time data and the optimal denitration prediction model. This amount is the ammonia injection amount that optimizes the denitration efficiency by comprehensively considering the current state of the boiler, the flue gas composition and the denitration effect.

[0070] In this embodiment, the ammonia injection device is adjusted according to the determined optimal ammonia injection amount, and the ammonia escape rate is monitored in real time. When the ammonia escape rate is too high, the ammonia injection device is corrected and adjusted according to the real-time flue gas data to ensure full utilization of ammonia and reduce pollutant emissions.

[0071] The beneficial effects of the above technical solution are as follows: the optimal denitration prediction model is constructed by acquiring historical operation data, historical flue gas data and historical denitration data, and the optimal ammonia injection amount based on the real-time operation data, real-time flue gas data, real-time denitration data and the optimal denitration prediction model are determined according to the real-time operation data, real-time flue gas data, real-time denitration data and the optimal denitration prediction model collected in real time, and the ammonia injection device is adjusted according to the optimal ammonia injection amount, and the ammonia escape rate is monitored in real time, and the ammonia injection device is corrected and adjusted according to the ammonia escape rate and real-time flue gas data. It can achieve precise ammonia injection adjustment, maximize the denitration effect, reduce ammonia waste and reduce environmental pollution, improve the intelligence and adaptability of the boiler denitration system, ensure the efficient and stable operation of the denitration control system, and have a high degree of adaptability and energy-saving and environmental protection effects.

[0072] Embodiment 2:

[0073] The embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, which obtains historical operation data during boiler combustion, historical flue gas data generated by boiler combustion, and historical denitration data of boiler denitration, including:

[0074] Obtain historical sub-operation data of the boiler combustion process and historical sub-flue gas data generated by the boiler combustion within multiple specified time periods, and obtain historical sub-denitrification data of boiler denitrification within multiple specified time periods;

[0075] Preprocess the historical sub-operation data, historical sub-flue gas data and historical sub-denitrification data within each specified time period;

[0076] The historical operation data is determined based on the historical sub-operation data within all specified time periods after preprocessing, the historical flue gas data is determined based on the historical sub-flue gas data within all specified time periods after preprocessing, and at the same time, the historical denitrification data is determined based on the historical sub-denitrification data within all specified time periods after preprocessing.

[0077] In this embodiment, historical data of the boiler is collected from multiple different time periods. The historical sub-operation data refers to the specific operation data generated by the boiler under different operating conditions, such as temperature, pressure, load, etc.; the historical sub-flue gas data is the concentration data of various components (such as nitrogen oxides, carbon dioxide, etc.) in the flue gas discharged by the boiler; the historical sub-denitrification data records the denitrification efficiency and emission effect data of the denitrification device in different time periods.

[0078] In this embodiment, after the data is collected, it needs to be preprocessed to ensure the data quality. Preprocessing includes operations such as removing noise, filling missing values, standardization and normalization. The preprocessed data can effectively eliminate external interference and improve the accuracy of data analysis.

[0079] In this embodiment, historical operation data, historical flue gas data and historical denitrification data are determined by integrating and summarizing preprocessing data within multiple time periods.

[0080] The beneficial effects of the above technical solution are: obtaining historical operation data during boiler combustion, obtaining historical flue gas data generated by boiler combustion, and historical denitrification data of boiler denitrification, which can enhance the representativeness of the data and provide accurate data basis for building the best denitrification prediction model.

[0081] Embodiment 3:

[0082] The embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, and constructs an optimal denitration prediction model, including:

[0083] Perform feature extraction on the historical sub-operation data within each specified time period after preprocessing to determine the historical operation vector within each specified time period. At the same time, perform feature extraction on the historical sub-smoke data within each specified time period after preprocessing to determine the historical smoke vector within each specified time period.

[0084] Based on the historical sub-denitrification data in each specified time period after pretreatment, determine the historical ammonia injection amount, ammonia escape rate and nitrogen oxide removal rate in each specified time period;

[0085] Determine a historical denitration vector in each specified time period based on the historical ammonia injection amount, ammonia escape rate, and nitrogen oxide removal rate in each specified time period;

[0086] Determine an operation-denitrification correlation matrix, a flue gas-denitrification correlation matrix, an operation-flue gas correlation matrix, and a period correlation vector within each specified time period;

[0087] Determining a first denitrification correlation vector based on the period correlation vectors within all specified time periods and the operation-flue gas correlation matrix;

[0088] The optimal denitrification prediction model is constructed based on the first denitrification related vector, the historical operation vectors within all specified time periods, the historical flue gas vectors and the historical denitrification vectors.

[0089] In this embodiment, feature extraction is performed on the historical sub-operation data and sub-flue gas data within each specified time period. By selecting key indicators (such as temperature, pressure, NOx concentration, etc.) and converting them into feature vectors, it is ensured that the core features of boiler operation and flue gas emissions can be represented.

[0090] In this embodiment, a historical operation vector is constructed based on the extracted operation data features, which will reflect the operation status of the boiler within a certain period; historical flue gas vector: based on the feature extraction of flue gas data, a historical flue gas vector is formed, which represents the main components and concentration changes of the flue gas discharged by the boiler.

[0091] In this embodiment, the ammonia injection amount (amount of injected ammonia), ammonia escape rate (emission ratio of unreacted ammonia) and nitrogen oxide removal rate (NOx removal rate) in each cycle are extracted from the denitrification data after pretreatment.

[0092] In this embodiment, a historical denitrification vector is constructed based on the extracted ammonia injection amount, ammonia escape rate and nitrogen oxide removal rate to measure the denitrification effect in different specified time periods.

[0093] In this embodiment, based on the cycle-related data, a cycle-related vector is derived, including all operation characteristics and flue gas characteristics having large correlation values ​​between the historical sub-operation data and the historical sub-flue gas data and the historical sub-denitrification data for each specified time period.

[0094] In this embodiment, the first denitrification related vector is combined with the historical operation vector, the historical flue gas vector and the historical denitrification vector to construct an optimal denitrification prediction model, which can predict the optimal denitrification effect and the optimal ammonia injection amount under different working conditions.

[0095] In this embodiment, the first denitrification related vector, historical operation vector, historical flue gas vector and historical denitrification vector in each historical period determined based on historical operation data, historical flue gas data and historical denitrification data are used as input samples, and the historical optimal ammonia injection amount in each historical period is used as output sample to train the neural network model to obtain the optimal denitrification prediction model, and the training samples exceed 1,000.

[0096] In this embodiment, the method for obtaining the historical optimal ammonia injection amount is:

[0097] The historical operation data, historical flue gas data and historical denitrification data in each historical period are taken as a group of data;

[0098] All group data are classified, and the minimum ammonia injection amount is extracted from the historical ammonia injection amounts corresponding to all group data under the same classification as the optimal ammonia injection amount under the corresponding classification.

[0099] For example, the first group: historical operation data is: data 01, historical flue gas data is: data 02, historical denitrification data is: data 03;

[0100] The second group: historical operation data is: data 11, historical flue gas data is: data 12, historical denitrification data is: data 13;

[0101] The third group: historical operation data is: data 21, historical flue gas data is: data 22, historical denitrification data is: data 23;

[0102] At this time, the results after classification are: the first group and the second group are in one category, and the third group is in another category. At this time, the corresponding minimum historical ammonia injection amount is extracted from the first group and the second group as the optimal ammonia injection amount for the corresponding classification.

[0103] The corresponding historical ammonia injection amount is extracted from the third group as the optimal ammonia injection amount.

[0104] In this embodiment, the classification method is implemented based on the principle of data similarity, which may be cluster analysis, and belongs to the prior art.

[0105] The beneficial effects of the above technical solution are as follows: by building an optimal denitrification prediction model based on historical operation data, historical flue gas data and historical denitrification data, the relationship between boiler operation, flue gas emissions and denitrification effects can be accurately determined, thereby improving the intelligence and adaptability of the denitrification system.

[0106] Embodiment 4:

[0107] An embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, which determines an operation-denitration correlation matrix within each specified time period, including:

[0108] Performing correlation analysis on the historical operation vectors and the historical denitrification vectors in each specified time period, determining the correlation value of each operation feature in the historical operation vectors and each feature in the historical denitrification vectors in each specified time period, and determining the operation-desnitrification correlation matrix in each specified time period;

[0109] ;

[0110] in, represents the operation-denitrification correlation matrix in the t-th specified time period, represents the i-th operation feature in the historical operation vector within the t-th specified time period, represents the jth feature of the historical denitrification vector in the tth specified time period, It represents the first correlation value of the i-th operation feature in the historical operation vector and the j-th feature in the historical denitrification vector within the t-th specified time period calculated based on the first correlation function, and N1 represents the number of operation features in the historical operation vector.

[0111] In this embodiment, the historical operation vector and the historical denitration vector in each specified time period are subjected to correlation analysis. By calculating the correlation value between each operation feature (such as boiler temperature, pressure, etc.) and the denitration effect feature (such as historical ammonia injection amount, ammonia escape rate, and nitrogen oxide removal rate), an operation-denitration correlation matrix is ​​obtained. This matrix reveals the relationship between the boiler operation state and the denitration effect, and helps to understand how different operating parameters affect the denitration effect.

[0112] In this embodiment, the first correlation function may be determined by a Pearson correlation coefficient, a Spearman rank correlation coefficient, or a cosine similarity.

[0113] The beneficial effects of the above technical solution are as follows: determining the operation-denitrification correlation matrix within each specified time period can provide high-quality data basis for determining the period-related vector, improve the intelligence level of boiler operation, and achieve more accurate ammonia injection adjustment and denitrification efficiency optimization.

[0114] Embodiment 5:

[0115] The embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, which determines a flue gas-denitration correlation matrix within each specified time period, including:

[0116] Performing correlation analysis on the historical flue gas vector and the historical denitrification vector in each specified time period, determining the correlation value of each flue gas feature in the historical flue gas vector and each feature in the historical denitrification vector in each specified time period, and determining the flue gas-denitrification correlation matrix in each specified time period;

[0117] ;

[0118] in, represents the flue gas-denitrification correlation matrix in the t-th specified time period, represents the kth smoke feature in the historical smoke vector within the tth specified time period, It represents the second correlation value of the kth flue gas feature in the historical flue gas vector and the jth feature in the historical denitrification vector within the tth specified time period calculated based on the second correlation function, and N2 represents the number of flue gas features in the historical flue gas vector.

[0119] In this embodiment, the historical flue gas vector and the historical denitrification vector within each specified time period are subjected to correlation analysis. The correlation between each flue gas feature (such as nitrogen oxide concentration, oxygen concentration, etc.) and the denitrification effect feature (such as historical ammonia injection amount, ammonia escape rate, and nitrogen oxide removal rate) is calculated to obtain a flue gas-denitrification correlation matrix. This can help analyze how different flue gas components affect the denitrification effect.

[0120] In this embodiment, the second correlation function may be determined by a Pearson correlation coefficient, a Spearman rank correlation coefficient, or a cosine similarity.

[0121] In this embodiment, They respectively represent the historical ammonia injection amount, ammonia escape rate, and nitrogen oxide removal rate in the tth specified time period.

[0122] The beneficial effects of the above technical solution are as follows: determining the flue gas-denitrification correlation matrix within each specified time period can provide high-quality data basis for determining the periodic correlation vector, improve the intelligence level of boiler operation, and achieve more accurate ammonia injection adjustment and denitrification efficiency optimization.

[0123] Embodiment 6:

[0124] The embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, which determines an operation-flue gas correlation matrix and a period correlation vector within each specified time period, including:

[0125] Performing correlation analysis on the historical operation vectors and the historical smoke vectors within each specified time period, determining the correlation value of each operation feature in the historical operation vectors and each smoke feature in the historical smoke vectors within each specified time period, and determining the operation-smoke correlation matrix;

[0126] ;

[0127] in, represents the operation-smoke correlation matrix in the t-th specified time period, represents a third correlation value of the i-th operation feature in the historical operation vector and the k-th smoke feature in the historical smoke vector within the t-th specified time period calculated based on the third correlation function;

[0128] Determining a period correlation vector based on the operation-denitrification correlation matrix and the flue gas-denitrification correlation matrix within each specified time period;

[0129] ;

[0130] ;

[0131] ;

[0132] in, represents the period-dependent vector of the t-th specified time period, represents the operation correlation vector of the tth specified time period, represents the smoke correlation vector of the tth specified time period, represents the first correlation threshold, represents the second correlation threshold, represents the average value of the first correlation value of the operation-denitrification correlation matrix for the t-th specified time period, represents the standard deviation of the first correlation value of the operation-denitrification correlation matrix for the t-th specified time period, , represents the average value of the second correlation value of the flue gas-denitrification correlation matrix in the t-th specified time period, represents the standard deviation of the second correlation value of the flue gas-denitrification correlation matrix for the t-th specified time period, They respectively represent the 1st related feature, the oth related feature, and the Nuth related feature in the periodic correlation vector of the tth specified time period, and tNu indicates that the number of related features of the tth specified time period is Nu.

[0133] In this embodiment, the historical operation vectors and historical flue gas vectors in each specified time period are subjected to correlation analysis. By calculating the correlation between the operation characteristics of the boiler and the flue gas components, an operation-flue gas correlation matrix is ​​obtained to reveal the relationship between the boiler operation status and the flue gas emissions.

[0134] In this embodiment, the third correlation function can be determined by Pearson correlation coefficient, Spearman rank correlation coefficient or cosine similarity.

[0135] In this embodiment, the first correlation threshold Determined based on all first correlation values ​​in the operation-denitrification correlation matrix for all specified time periods.

[0136] In this embodiment, the second correlation threshold Determined based on all second correlation values ​​in the flue gas-denitrification correlation matrix within all specified time periods.

[0137] In this embodiment, Express The results are sorted from largest to smallest.

[0138] The beneficial effects of the above technical solution are as follows: determining the operation-flue gas correlation matrix and period correlation vector within each specified time period can provide high-quality data basis for determining the first denitrification correlation vector, improve the intelligence level of boiler operation, and achieve more accurate ammonia injection adjustment and denitrification efficiency optimization.

[0139] Embodiment 7:

[0140] An embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, which determines a first denitration correlation vector based on period correlation vectors within all specified time periods and an operation-flue gas correlation matrix, including:

[0141] Determining a first weight of each correlation feature in each specified time period based on the period correlation vector in each specified time period and the operation-smoke correlation matrix;

[0142] ;

[0143] ;

[0144] ;

[0145] in, represents the first weight of the oth correlation feature in the periodic correlation vector of the tth specified time period, Represents the oth correlation feature in the periodic correlation vector of the tth specified time period The first indicator function of Represents the oth correlation feature in the periodic correlation vector of the tth specified time period The second indicator function of

[0146] Determine a second denitration correlation vector and a second weight of each correlation feature of the second denitration correlation vector based on the first weights of the period correlation vectors in all specified time periods and all correlation features in all specified time periods;

[0147] ;

[0148] ;

[0149] , ;

[0150] , ;

[0151] in, represents the second denitrification related vector under T specified time periods, They represent the 1st related feature, the pth related feature, and the Nmth related feature of the second denitrification related vector respectively, A second weight representing a p-th related feature of a second denitrification related vector; represents a function for fusing period-related vectors of T specified time periods; Represents the characteristic correlation analysis function of the t-th specified time period; represents the characteristic consistent analysis function of the t-th specified time period;

[0152] All relevant features in the second denitrification related vector are sequentially adjusted based on the second weights of all relevant features in the second denitrification related vector to determine the first denitrification related vector. In this embodiment, based on the periodic related vector and the operation-flue gas correlation matrix in each specified time period, the relevant features in each period are analyzed to determine the first weight of each relevant feature in the periodic related vector in each specified time period. The first weight reflects the relative importance of a specific feature to the denitrification effect in a certain time period.

[0153] In this embodiment, in the second denitration related vector, the order is adjusted based on the second weight of each related feature. This adjustment is to optimize the denitration effect by rearranging the related features in the order of weights, and finally determine the first denitration related vector.

[0154] The beneficial effects of the above technical solution are as follows: the first denitrification related vector is determined based on the period-related vectors within all specified time periods and the operation-flue gas correlation matrix, which enables more refined feature analysis and optimization, and realizes dynamic and precise optimization of the denitrification effect, making the adjustment of the ammonia injection amount more flexible and the denitrification effect more precise, thereby improving the intelligence level of the denitrification system, reducing ammonia waste and reducing environmental pollution.

[0155] Embodiment 8:

[0156] The embodiment of the present invention provides an intelligent ammonia injection adjustment method for a boiler denitration control system, which determines the optimal ammonia injection amount based on the real-time operation data based on the real-time operation data, the real-time flue gas data, the real-time denitration data and the optimal denitration prediction model, including:

[0157] Feature extraction is performed on real-time operation data, real-time flue gas data and real-time denitrification data respectively to determine a real-time operation vector, a real-time flue gas vector and a real-time denitrification vector;

[0158] The real-time operation vector, the real-time flue gas vector and the real-time denitrification vector are input into the optimal denitrification prediction model, and the optimal ammonia injection amount is output based on the optimal denitrification prediction model.

[0159] In this embodiment, the real-time operation data includes the real-time status information of the boiler, such as temperature, pressure, load, etc. By analyzing and processing these data, key features that can effectively describe the boiler operation status are extracted to form a real-time operation vector. The vector represents the main features of the current boiler operation status.

[0160] In this embodiment, the real-time flue gas data includes components in the flue gas discharged by the boiler, such as nitrogen oxides (NOx), carbon dioxide (CO2), etc. By extracting the features of the flue gas data, a real-time flue gas vector is generated to reflect the concentration and changes of the flue gas components.

[0161] In this embodiment, the real-time denitration data includes real-time monitoring values ​​of the denitration effect, such as denitration efficiency, ammonia escape rate, etc. These data are processed to generate a real-time denitration vector, which represents the denitration effect of the current denitration system.

[0162] In this embodiment, the extracted real-time operation vector, real-time flue gas vector and real-time denitration vector are used as input data and input into the previously constructed optimal denitration prediction model.

[0163] In this embodiment, the output result of the optimal denitration prediction model will predict the optimal ammonia injection amount based on the current state of the boiler (operation, flue gas and denitration effect data). The optimal ammonia injection amount can ensure the maximum removal of nitrogen oxides while avoiding ammonia escape or waste caused by excessive ammonia injection.

[0164] The beneficial effects of the above technical solution are as follows: based on real-time operation data, real-time flue gas data, real-time denitrification data and the optimal denitrification prediction model, the optimal ammonia injection amount based on the real-time operation data is determined, and intelligent and dynamic ammonia injection amount adjustment is realized. It can optimize the denitrification efficiency according to real-time changes, reduce ammonia waste and environmental pollution, and at the same time improve the economy and operation stability of the boiler.

[0165] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0166] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent ammonia injection adjustment method for a boiler denitration control system, characterized in that: include: Step 1: Obtain historical operation data during boiler combustion and historical flue gas data generated by boiler combustion, and obtain historical denitrification data of boiler denitrification; Step 2: Based on the historical operation data, historical flue gas data and historical denitrification data, determine the corresponding first denitrification related vector, the historical operation vectors, the historical flue gas vectors and the historical denitrification vectors within all specified time periods, and construct an optimal denitrification prediction model, wherein the optimal denitrification prediction model takes the optimal ammonia injection amount as the output target; Step 3: Real-time collection of real-time operation data during boiler combustion and real-time flue gas data generated by boiler combustion, and real-time collection of real-time denitrification data of boiler denitrification; Step 4: Based on the real-time operation data, the real-time flue gas data, the real-time denitrification data and the optimal denitrification prediction model, determine the optimal ammonia injection amount based on the real-time operation data; Step 5: Adjust the ammonia injection device based on the optimal ammonia injection amount, monitor the ammonia escape rate in real time, and make corrections and adjustments to the ammonia injection device based on the ammonia escape rate and real-time flue gas data; Among them, the optimal denitrification prediction model is constructed, including: Perform feature extraction on the historical sub-operation data within each specified time period after preprocessing to determine the historical operation vector within each specified time period. At the same time, perform feature extraction on the historical sub-smoke data within each specified time period after preprocessing to determine the historical smoke vector within each specified time period. Based on the historical sub-denitrification data in each specified time period after pretreatment, determine the historical ammonia injection amount, ammonia escape rate and nitrogen oxide removal rate in each specified time period; Determine a historical denitration vector in each specified time period based on the historical ammonia injection amount, ammonia escape rate, and nitrogen oxide removal rate in each specified time period; Determine an operation-denitrification correlation matrix, a flue gas-denitrification correlation matrix, an operation-flue gas correlation matrix, and a period correlation vector within each specified time period; Determining a first denitrification correlation vector based on the period correlation vectors within all specified time periods and the operation-flue gas correlation matrix; An optimal denitrification prediction model is constructed based on the first denitrification related vector, the historical operation vectors within all specified time periods, the historical flue gas vectors, and the historical denitrification vectors.

2. The intelligent ammonia injection adjustment method for a boiler denitration control system according to claim 1 is characterized in that: Obtain historical operation data during boiler combustion, historical flue gas data generated by boiler combustion, and historical denitrification data of boiler denitrification, including: Obtain historical sub-operation data of the boiler combustion process and historical sub-flue gas data generated by the boiler combustion within multiple specified time periods, and obtain historical sub-denitrification data of boiler denitrification within multiple specified time periods; Preprocess the historical sub-operation data, historical sub-flue gas data and historical sub-denitrification data within each specified time period; The historical operation data is determined based on the historical sub-operation data within all specified time periods after preprocessing, the historical flue gas data is determined based on the historical sub-flue gas data within all specified time periods after preprocessing, and at the same time, the historical denitrification data is determined based on the historical sub-denitrification data within all specified time periods after preprocessing.

3. The intelligent ammonia injection adjustment method for a boiler denitration control system according to claim 1 is characterized in that: Determine the operation-denitrification correlation matrix for each specified time period, including: Performing correlation analysis on the historical operation vectors and the historical denitrification vectors in each specified time period, determining the correlation value of each operation feature in the historical operation vectors and each feature in the historical denitrification vectors in each specified time period, and determining the operation-desnitrification correlation matrix in each specified time period; ; in, represents the operation-denitrification correlation matrix in the t-th specified time period, represents the i-th operation feature in the historical operation vector within the t-th specified time period, represents the jth feature of the historical denitrification vector in the tth specified time period, It represents the first correlation value of the i-th operation feature in the historical operation vector and the j-th feature in the historical denitrification vector within the t-th specified time period calculated based on the first correlation function, and N1 represents the number of operation features in the historical operation vector.

4. The intelligent ammonia injection adjustment method for a boiler denitration control system according to claim 1 is characterized in that: Determine the flue gas-denitrification correlation matrix for each specified time period, including: Performing correlation analysis on the historical flue gas vector and the historical denitrification vector in each specified time period, determining the correlation value of each flue gas feature in the historical flue gas vector and each feature in the historical denitrification vector in each specified time period, and determining the flue gas-denitrification correlation matrix in each specified time period; ; in, represents the flue gas-denitrification correlation matrix in the t-th specified time period, represents the kth smoke feature in the historical smoke vector within the tth specified time period, It represents the second correlation value of the kth flue gas feature in the historical flue gas vector and the jth feature in the historical denitrification vector within the tth specified time period calculated based on the second correlation function, and N2 represents the number of flue gas features in the historical flue gas vector.

5. The intelligent ammonia injection adjustment method for a boiler denitration control system according to claim 1 is characterized in that: Determine the operation-smoke correlation matrix and period correlation vector for each specified time period, including: Performing correlation analysis on the historical operation vectors and the historical smoke vectors within each specified time period, determining the correlation value of each operation feature in the historical operation vectors and each smoke feature in the historical smoke vectors within each specified time period, and determining the operation-smoke correlation matrix; ； in, represents the operation-smoke correlation matrix in the t-th specified time period, represents a third correlation value of the i-th operation feature in the historical operation vector and the k-th smoke feature in the historical smoke vector within the t-th specified time period calculated based on the third correlation function; Determining a period correlation vector based on the operation-denitrification correlation matrix and the flue gas-denitrification correlation matrix within each specified time period; ； ； ； in, represents the period-dependent vector of the t-th specified time period, represents the operation correlation vector of the tth specified time period, represents the smoke correlation vector of the tth specified time period, represents the first correlation threshold, represents the second correlation threshold, represents the average value of the first correlation value of the operation-denitrification correlation matrix for the t-th specified time period, represents the standard deviation of the first correlation value of the operation-denitrification correlation matrix for the t-th specified time period, , represents the average value of the second correlation value of the flue gas-denitrification correlation matrix in the t-th specified time period, represents the standard deviation of the second correlation value of the flue gas-denitrification correlation matrix for the t-th specified time period, They respectively represent the 1st related feature, the oth related feature, and the Nuth related feature in the periodic correlation vector of the tth specified time period, and tNu indicates that the number of related features of the tth specified time period is Nu.

6. The intelligent ammonia injection adjustment method for a boiler denitration control system according to claim 5 is characterized in that: The first denitrification correlation vector is determined based on the period correlation vectors in all specified time periods and the operation-flue gas correlation matrix, including: Determining a first weight of each correlation feature in each specified time period based on the period correlation vector in each specified time period and the operation-smoke correlation matrix; ； ； ； in, represents the first weight of the oth correlation feature in the periodic correlation vector of the tth specified time period, Represents the oth correlation feature in the periodic correlation vector of the tth specified time period The first indicator function of Represents the oth correlation feature in the periodic correlation vector of the tth specified time period The second indicator function of Determine a second denitration correlation vector and a second weight of each correlation feature of the second denitration correlation vector based on the first weights of the period correlation vectors in all specified time periods and all correlation features in all specified time periods; ； ； ， ； ， ； in, represents the second denitrification related vector under T specified time periods, They represent the 1st related feature, the pth related feature, and the Nmth related feature of the second denitrification related vector respectively, A second weight representing a p-th related feature of a second denitrification related vector; represents a function for fusing period-related vectors of T specified time periods; Represents the characteristic correlation analysis function of the t-th specified time period; represents the characteristic consistent analysis function of the t-th specified time period; All relevant features in the second denitration related vector are sequentially adjusted based on the second weights of all relevant features of the second denitration related vector to determine the first denitration related vector.

7. The intelligent ammonia injection adjustment method for a boiler denitration control system according to claim 1 is characterized in that: Based on real-time operation data, real-time flue gas data, real-time denitrification data and the best denitrification prediction model, the best ammonia injection amount based on real-time operation data is determined, including: Feature extraction is performed on real-time operation data, real-time flue gas data and real-time denitrification data respectively to determine a real-time operation vector, a real-time flue gas vector and a real-time denitrification vector; The real-time operation vector, the real-time flue gas vector and the real-time denitrification vector are input into the optimal denitrification prediction model, and the optimal ammonia injection amount is output based on the optimal denitrification prediction model.

Images