Boiler combustion self-adaptive adjusting method and system

Abstract

The invention relates to the technical field of boiler control, and discloses a boiler combustion self-adaptive adjusting method and system.The method comprises the steps that a plurality of data sets are constructed according to historical monitoring data of the boiler combustion process, and working condition characteristics are determined according to historical working condition data subsets in each data set; working condition similarity coefficients among different data sets are calculated, and a plurality of data sets are classified to obtain a plurality of data set-control groups; analyzing each data set-control group to obtain a first adjustment mapping table, a second adjustment mapping table and a third adjustment mapping table of each working condition feature, respectively constructing a first adjustment model, a second adjustment model and a third adjustment model, acquiring real-time monitoring data and judging whether to trigger a preset adjustment condition, and if yes, judging whether to trigger the preset adjustment condition; and a corresponding adjusting model is selected, an adjusting strategy is generated, and an adjusting instruction is issued, so that self-adaptive adjustment in the boiler combustion process is achieved, the energy utilization efficiency is improved, and pollutant emission is reduced.

Description

Technical Field

[0001] This application relates to the field of boiler control technology, and in particular to a boiler combustion adaptive adjustment method and system. Background Technology

[0002] Boilers, as core equipment for energy conversion, are widely used in power generation, industrial heating, and residential heating. However, traditional boiler combustion control methods rely heavily on fixed parameter settings and manual experience-based adjustments, making them ill-suited to complex and ever-changing operating conditions. During boiler operation, combustion efficiency and pollutant emission levels vary significantly under different conditions. Adopting a uniform control strategy can easily lead to energy waste and environmental pollution. For example, when boiler load changes abruptly, traditional control methods cannot adjust fuel supply and air ratio in a timely manner, resulting in incomplete combustion, which reduces thermal efficiency and increases the generation of pollutants such as nitrogen oxides and sulfur dioxide. Furthermore, traditional control methods lack in-depth analysis of historical operating data, making it difficult to summarize optimal control patterns under different operating conditions and providing a scientific basis for real-time regulation. Therefore, there is an urgent need for an adaptive boiler combustion control method and system that can automatically adjust control parameters according to real-time operating conditions, improving energy utilization efficiency and reducing pollutant emissions. Summary of the Invention

[0003] To address the aforementioned technical problems, this application provides a boiler combustion adaptive adjustment method and system. By constructing a first adjustment mapping table, a second adjustment mapping table, and a third adjustment mapping table with different operating conditions, and constructing a first adjustment model, a second adjustment model, and a third adjustment model, the system determines whether real-time monitoring data triggers preset adjustment conditions. If so, the system selects the corresponding adjustment model and issues an adjustment command to achieve adaptive adjustment of the boiler combustion process, thereby improving energy utilization efficiency and reducing pollutant emissions.

[0004] In some embodiments of this application, a boiler combustion adaptive adjustment method is provided, including:

[0005] Several datasets are constructed based on historical monitoring data of the boiler combustion process. Each dataset includes a subset of historical operating condition data, a subset of historical control data, a subset of historical efficiency data, and a subset of historical emission data.

[0006] The operating condition characteristics are determined based on a subset of historical operating condition data in each dataset. The operating condition similarity coefficient between different datasets is calculated. Several datasets are classified according to the operating condition similarity coefficient to obtain multiple dataset-control group.

[0007] For each dataset-control group, we analyzed the first, second, and third regulation mapping tables for each operating condition characteristic and constructed the first, second, and third regulation models respectively.

[0008] Acquire real-time monitoring data and determine whether preset adjustment conditions are triggered. If so, select the corresponding adjustment model, generate an adjustment strategy, and issue adjustment instructions.

[0009] In some embodiments of this application, several datasets are constructed based on historical monitoring data of the boiler combustion process, including:

[0010] Acquire several historical monitoring logs of the boiler combustion process, and set several collection time nodes for each historical monitoring log according to the preset monitoring time interval;

[0011] Historical monitoring data is collected from each historical monitoring log according to several collection time points, and the difference between the historical monitoring data at the previous adjacent collection time point is calculated.

[0012] Historical monitoring data with a difference greater than a preset difference threshold and belonging to the working condition category are selected, and an initial monitoring dataset for the working condition category is constructed based on the selected historical monitoring data.

[0013] Calculate the coefficient of change of the initial monitoring dataset at the current data collection point;

[0014] If the coefficient of change is greater than the preset threshold for the coefficient of change, then the initial monitoring dataset at the current data collection time point is set as a subset of the historical operating condition data.

[0015] Construct a subset of historical control data based on historical monitoring data belonging to the control category at the current data collection time point;

[0016] Based on the historical monitoring data belonging to the efficiency category and the historical monitoring data belonging to the emission category at the historical feedback node corresponding to the current collection time node, a subset of historical efficiency data and a subset of historical emission data are constructed respectively. Combined with the subset of historical operating condition data and the subset of historical control data, a dataset is obtained.

[0017] Several datasets are generated sequentially.

[0018] In some embodiments of this application, the operating condition similarity coefficient between different datasets is calculated, and the datasets are classified according to the operating condition similarity coefficient to obtain multiple dataset-control groups, including:

[0019] Feature extraction is performed on a subset of historical working condition data in each dataset to obtain the corresponding working condition features, and each working condition feature is mapped to a corresponding weight coefficient.

[0020] The similarity analysis of the working condition features corresponding to different datasets is performed to obtain the number of similar working condition features in different datasets and the corresponding similarity coefficients. The working condition similarity coefficients between different datasets are then calculated by combining the weight coefficients of the corresponding working condition features.

[0021] Pre-set the threshold for the operating condition similarity coefficient;

[0022] Data sets with a similarity coefficient greater than the threshold for the working conditions are grouped into the same category, and a control group is constructed based on all data sets in that category.

[0023] Several datasets and control groups were constructed sequentially.

[0024] In some embodiments of this application, each dataset-control group includes a first dataset sequence, a second dataset sequence, and a third dataset sequence, specifically:

[0025] The historical monitoring data in the historical efficiency data subset of each dataset in the same dataset-control group are compared with the standard efficiency data range of the corresponding working condition characteristics, and the historical efficiency coefficient of each dataset is calculated based on the comparison results.

[0026] The standard efficiency data range includes several standard efficiency ranges, and each standard efficiency range is mapped to a corresponding preset combustion efficiency.

[0027] The historical monitoring data in the historical emission data subset of each dataset in the same dataset-control group are compared with the standard emission data range of the corresponding operating conditions, and the historical emission coefficient of each dataset is calculated based on the comparison results.

[0028] The historical comprehensive control coefficients for each dataset are calculated based on the historical efficiency coefficients and historical emission coefficients of each dataset.

[0029] Sort all datasets in the same dataset-control group according to their historical efficiency coefficients to obtain the first dataset sequence of the dataset-control group;

[0030] Sort all datasets in the same dataset-control group according to historical emission coefficients to obtain the second dataset sequence of the dataset-control group;

[0031] Sort all datasets in the same dataset-control group according to the historical comprehensive control coefficient to obtain the third dataset sequence of the dataset-control group.

[0032] In some embodiments of this application, the historical efficiency coefficient is calculated using the following formula:

[0033] ;

[0034] Where L1 is the historical efficiency coefficient, g1 is the first efficiency conversion coefficient, g2 is the second efficiency conversion coefficient, w1 is the number of historical monitoring data points within the corresponding standard efficiency data interval in the same historical efficiency data subset, and w2 is the total number of historical monitoring data points in the same historical efficiency data subset. This is the v1st historical monitoring data. This refers to the characteristic efficiency data within the standard efficiency data interval where the v1-th historical monitoring data falls. The weighting coefficient for the v1th historical monitoring data;

[0035] The formula for calculating the historical emission coefficient is as follows:

[0036] ;

[0037] Where L2 is the historical emission coefficient, u1 is the first emission conversion coefficient, u2 is the second emission conversion coefficient, w3 is the number of historical monitoring data points within the corresponding standard emission data range in the same historical emission data subset, and w4 is the total number of historical monitoring data points in the same historical efficiency data subset. This is the v2nd historical monitoring data. The characteristic emission data within the standard emission data range where the v2th historical monitoring data falls. The weighting coefficient for the v2th historical monitoring data.

[0038] In some embodiments of this application, each dataset-control group is analyzed to obtain a first adjustment mapping table, a second adjustment mapping table, and a third adjustment mapping table for each operating condition characteristic, including:

[0039] Each dataset in the same first dataset sequence is compared sequentially with the historical control data subset and historical efficiency data subset of other datasets to obtain the first control data difference of each dataset and the efficiency data difference of the historical control data subset of other datasets, and the first correlation matrix of each dataset is generated by combining the sequence order.

[0040] The initial adjustment mapping relationship between historical control data and historical efficiency data is determined based on the first association matrix of each dataset.

[0041] The initial adjustment mapping relationship between the historical control data and historical efficiency data determined by all datasets in the same first dataset sequence is summarized and integrated to obtain the first adjustment mapping relationship and generate the first adjustment mapping table under the working condition characteristics of the corresponding dataset-control group.

[0042] The first adjustment mapping table includes a first adjustment amount of several control data, each first adjustment amount is mapped to several efficiency data, the change in efficiency data and the increase in efficiency coefficient.

[0043] Each dataset in the same second dataset sequence is compared sequentially with the historical control data subset and historical emission data subset of other datasets to obtain the second control data difference of each dataset with the historical control data subset and the emission data difference of the historical emission data subset of other datasets. The second correlation matrix of each dataset is then generated by combining the sequence order.

[0044] The initial adjustment mapping relationship between historical control data and historical emission data is determined based on the second correlation matrix of each dataset;

[0045] The initial adjustment mapping relationship between the historical control data and historical emission data determined by all datasets in the same second dataset sequence is summarized and integrated to obtain the second adjustment mapping relationship and generate the second adjustment mapping table under the operating condition characteristics of the corresponding dataset-control group.

[0046] The second adjustment mapping table includes a second adjustment amount of several control data, each second adjustment amount is mapped to several emission data, the amount of change in emission data and the amount of increase in emission coefficient;

[0047] Each dataset in the same third dataset sequence is compared sequentially with the historical control data subset, historical efficiency data subset, and historical emission data subset of other datasets to obtain the third control data difference, efficiency data difference, and emission data difference of each dataset with the historical control data subset, historical efficiency data subset, and historical emission data subset of other datasets. The third correlation matrix of each dataset is then generated by combining the sequence order.

[0048] The initial adjustment mapping relationship between historical control data, historical efficiency data, and historical emission data is determined based on the third correlation matrix of each dataset;

[0049] The initial regulation mapping relationship of all datasets in the same third dataset sequence is summarized and integrated to obtain the third regulation mapping relationship and generate the third regulation mapping table under the operating condition characteristics of the corresponding dataset-control group.

[0050] The third adjustment mapping table includes a third adjustment amount of several control data, each third adjustment amount is mapped to several efficiency data and emission data, the change amount of efficiency data and emission data and the increase amount of comprehensive control coefficient.

[0051] In some embodiments of this application, a first regulation model, a second regulation model, and a third regulation model are constructed, including:

[0052] A first training set is constructed based on the first adjustment mapping table of all working conditions, and the first adjustment model is obtained by training the model based on the first training set.

[0053] The first training set uses operating condition characteristics, the increase in efficiency coefficient, and the change in several efficiency data as training input data, and the corresponding control data and the first adjustment amount of the control data as training output data.

[0054] A second training set is constructed based on the second adjustment mapping table of all working conditions, and the second adjustment model is obtained by training the model based on the second training set.

[0055] The second training set uses operating condition characteristics, the increase in emission coefficients, and the change in several emission data as training input data, and the corresponding control data and the second adjustment of the control data as training output data.

[0056] A third training set is constructed based on the third adjustment mapping table of all working conditions, and the third adjustment model is obtained by training the model based on the third training set.

[0057] The third training set uses operating condition characteristics, the increase in the comprehensive control coefficient, and changes in several emission and efficiency data as training input data, and the corresponding control data and the third adjustment of the control data as training output data.

[0058] In some embodiments of this application, the preset adjustment conditions include:

[0059] The preset adjustment conditions include a first adjustment condition, a second adjustment condition, and a third adjustment condition;

[0060] Pre-set efficiency coefficient thresholds and emission coefficient thresholds;

[0061] When the real-time efficiency coefficient is less than the efficiency coefficient threshold and the real-time emission coefficient is not less than the emission coefficient threshold, the first adjustment condition is triggered and the first adjustment model is selected for adaptive adjustment.

[0062] When the real-time efficiency coefficient is not less than the efficiency coefficient threshold and the real-time emission coefficient is less than the emission coefficient threshold, the second adjustment condition is triggered and the second adjustment model is selected for adaptive adjustment.

[0063] When the real-time efficiency coefficient is less than the efficiency coefficient threshold and the real-time emission coefficient is less than the emission coefficient threshold, the third adjustment condition is triggered and the third adjustment model is selected for adaptive adjustment.

[0064] In some embodiments of this application, if so, acquiring real-time monitoring data and determining whether a preset adjustment condition is triggered, selecting a corresponding adjustment model, generating an adjustment strategy, and issuing an adjustment command includes:

[0065] Acquire real-time monitoring data and determine real-time operating condition characteristics;

[0066] The real-time efficiency coefficient is calculated based on the real-time monitoring data belonging to the efficiency category, and the real-time emission coefficient is calculated based on the real-time monitoring data belonging to the emission category.

[0067] The real-time efficiency coefficient and real-time emission coefficient are compared with the efficiency coefficient threshold and emission coefficient threshold corresponding to the real-time operating condition characteristics, respectively, and the preset adjustment conditions are determined based on the comparison results.

[0068] If the first adjustment condition is triggered, the first adjustment model is selected, and the real-time operating condition characteristics, real-time efficiency coefficient difference, efficiency data to be adjusted and the corresponding expected change are input into the first adjustment model to obtain the data to be controlled and the corresponding first adjustment amount and generate the adjustment strategy.

[0069] If the second adjustment condition is triggered, the second adjustment model is selected, and the real-time operating condition characteristics, real-time emission coefficient difference, emission data to be adjusted and the corresponding expected change are input into the second adjustment model to obtain the data to be controlled and the corresponding second adjustment amount and generate the adjustment strategy.

[0070] If the third adjustment condition is triggered, the third adjustment model is selected, and the real-time operating condition characteristics, real-time comprehensive control coefficient difference, efficiency data to be adjusted, emission data to be adjusted, and the corresponding expected change are input into the third adjustment model to obtain the data to be controlled and the corresponding third adjustment quantity and generate the adjustment strategy.

[0071] The adjustment strategy is converted into adjustment commands and sent to the boiler combustion control system.

[0072] In some embodiments of this application, a boiler combustion adaptive control system is also included:

[0073] The construction module is used to construct several datasets based on historical monitoring data of the boiler combustion process. Each dataset includes a subset of historical operating condition data, a subset of historical control data, a subset of historical efficiency data, and a subset of historical emission data.

[0074] The classification model is used to determine the working condition characteristics based on a subset of historical working condition data in each dataset, calculate the working condition similarity coefficient between different datasets, classify several datasets according to the working condition similarity coefficient, and obtain multiple datasets-control groups.

[0075] The analysis module is used to analyze each dataset-control group to obtain the first regulation mapping table, the second regulation mapping table and the third regulation mapping table for each working condition feature, and to construct the first regulation model, the second regulation model and the third regulation model respectively.

[0076] The adjustment module is used to acquire real-time monitoring data and determine whether preset adjustment conditions are triggered. If so, it selects the corresponding adjustment model, generates an adjustment strategy, and issues adjustment instructions.

[0077] The boiler combustion adaptive adjustment method and system of this application have the following advantages compared with the prior art:

[0078] By constructing a first adjustment mapping table, a second adjustment mapping table, and a third adjustment mapping table with different operating conditions, and constructing a first adjustment model, a second adjustment model, and a third adjustment model, it is determined whether the real-time monitoring data triggers the preset adjustment conditions. If so, the corresponding adjustment model is selected and an adjustment command is issued to achieve adaptive adjustment of the boiler combustion process, improve energy utilization efficiency, and reduce pollutant emissions. Attached Figure Description

[0079] Figure 1 This is a schematic flowchart of a boiler combustion adaptive adjustment method according to an embodiment of this application. Detailed Implementation

[0080] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate this application, but are not intended to limit the scope of this application.

[0081] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0082] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0083] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0084] like Figure 1 As shown in the figure, an adaptive boiler combustion control method according to an embodiment of this application includes:

[0085] Step S101: Construct several datasets based on historical monitoring data of the boiler combustion process. Each dataset includes a subset of historical operating condition data, a subset of historical control data, a subset of historical efficiency data, and a subset of historical emission data.

[0086] Step S102: Determine the working condition characteristics based on the subset of historical working condition data in each dataset, calculate the working condition similarity coefficient between different datasets, classify several datasets according to the working condition similarity coefficient, and obtain multiple datasets-control groups;

[0087] Step S103: Analyze each dataset-control group to obtain the first regulation mapping table, the second regulation mapping table and the third regulation mapping table for each working condition feature, and construct the first regulation model, the second regulation model and the third regulation model respectively;

[0088] Step S104: Obtain real-time monitoring data and determine whether the preset adjustment conditions are triggered. If so, select the corresponding adjustment model, generate the adjustment strategy, and issue the adjustment command.

[0089] In some embodiments of this application, several datasets are constructed based on historical monitoring data of the boiler combustion process, including:

[0090] Acquire several historical monitoring logs of the boiler combustion process, and set several collection time nodes for each historical monitoring log according to the preset monitoring time interval;

[0091] Historical monitoring data is collected from each historical monitoring log according to several collection time points, and the difference between the historical monitoring data at the previous adjacent collection time point is calculated.

[0092] Historical monitoring data with a difference greater than a preset difference threshold and belonging to the working condition category are selected, and an initial monitoring dataset for the working condition category is constructed based on the selected historical monitoring data.

[0093] Calculate the coefficient of change of the initial monitoring dataset at the current data collection point;

[0094] If the coefficient of change is greater than the preset threshold for the coefficient of change, then the initial monitoring dataset at the current data collection time point is set as a subset of the historical operating condition data.

[0095] Construct a subset of historical control data based on historical monitoring data belonging to the control category at the current data collection time point;

[0096] Based on the historical monitoring data belonging to the efficiency category and the historical monitoring data belonging to the emission category at the historical feedback node corresponding to the current collection time node, a subset of historical efficiency data and a subset of historical emission data are constructed respectively. Combined with the subset of historical operating condition data and the subset of historical control data, a dataset is obtained.

[0097] Several datasets are generated sequentially.

[0098] In this embodiment, several historical monitoring logs are obtained by dividing continuous monitoring logs according to a preset monitoring period. The preset monitoring period can be flexibly adjusted according to factors such as the running time of the boiler combustion process and the monitoring accuracy requirements. By dividing the logs into multiple historical monitoring logs, the accuracy of capturing changes in the operating conditions of the combustion process is improved, providing a more accurate and comprehensive data foundation for the subsequent construction of datasets and control groups.

[0099] In this embodiment, the coefficient of variation is calculated based on the number of selected historical monitoring data points, the differences between historical monitoring data points, and the preset weighting coefficients for the corresponding historical monitoring data points. Specifically, the differences between the selected historical monitoring data points and the corresponding preset weighting coefficients are weighted and transformed to obtain the coefficient of variation. The larger the differences between historical monitoring data points, the more data points there are, and the larger the preset weighting coefficients are, the larger the coefficient of variation will be, and vice versa. The coefficient of variation = Where n is the number of historical monitoring data points selected. Let be the difference of the i-th historical monitoring data, qi be the weighting coefficient of the difference of the i-th historical monitoring data, and z be the change conversion coefficient, which refers to converting the difference of historical monitoring data into a value with the same dimension as the change coefficient.

[0100] In this embodiment, the historical monitoring data for the operating condition category includes, but is not limited to, boiler load data, fuel supply data, fuel characteristic parameters, ambient temperature, atmospheric pressure, and air humidity. The historical monitoring data for the control category covers burner adjustment commands and damper opening commands. The historical monitoring data for the efficiency category mainly includes boiler thermal efficiency and fuel utilization rate. The historical monitoring data for the emission category involves the concentration of various pollutants in the flue gas, such as sulfur dioxide, nitrogen oxides, and particulate matter.

[0101] In this embodiment, the feedback time node corresponding to the acquisition time node is determined comprehensively based on factors such as the response characteristics of the boiler combustion system, the execution time of control commands, and the reaction time of efficiency data and emission data to control commands. By clearly defining the feedback time node, historical monitoring data of efficiency category and emission category after the historical control data at the corresponding acquisition time node can be accurately obtained, so as to ensure the integrity and accuracy of the constructed dataset, and thus provide reliable data support for the subsequent construction of the regulation model.

[0102] In this embodiment, by constructing a dataset, a set of historical monitoring data under different operating conditions is obtained. In-depth analysis of these datasets can uncover the inherent relationships and changing patterns of control data, efficiency data, and emission data under different operating conditions, providing solid data support for the subsequent construction of regulation models.

[0103] In some embodiments of this application, the operating condition similarity coefficient between different datasets is calculated, and the datasets are classified according to the operating condition similarity coefficient to obtain multiple dataset-control groups, including:

[0104] Feature extraction is performed on a subset of historical working condition data in each dataset to obtain the corresponding working condition features, and each working condition feature is mapped to a corresponding weight coefficient.

[0105] The similarity analysis of the working condition features corresponding to different datasets is performed to obtain the number of similar working condition features in different datasets and the corresponding similarity coefficients. The working condition similarity coefficients between different datasets are then calculated by combining the weight coefficients of the corresponding working condition features.

[0106] Pre-set the threshold for the operating condition similarity coefficient;

[0107] Data sets with a similarity coefficient greater than the threshold for the working conditions are grouped into the same category, and a control group is constructed based on all data sets in that category.

[0108] Several datasets and control groups were constructed sequentially.

[0109] In this embodiment, the operating conditions include, but are not limited to, boiler load level, fuel supply rate, fuel calorific value characteristics, ambient temperature conditions, atmospheric pressure conditions, and air humidity.

[0110] In this embodiment, the dataset-control group includes at least two datasets, and the datasets within the same dataset-control group have high operating condition similarity, such as similar boiler load, similar fuel supply, and similar ambient temperature and atmospheric pressure. By grouping datasets with similar operating conditions together, the efficiency performance and emission characteristics of different control strategies for boiler combustion under such operating conditions can be analyzed more effectively, providing a more targeted data foundation for the construction of subsequent regulation models. At the same time, different datasets-control groups represent different operating condition categories, which helps to comprehensively cover various operating conditions that may be encountered during boiler combustion.

[0111] In this embodiment, similar operating condition features refer to operating condition parameters with similar values ​​and the same trend of change. The similarity coefficient is obtained by quantitatively evaluating the numerical differences and the degree of agreement of the changing trends between similar operating condition features. The smaller the numerical differences and the higher the degree of agreement, the larger the quantified similarity coefficient, and vice versa. The operating condition similarity coefficient is obtained by summing the number of similar operating condition features, the similarity coefficient, and the corresponding weight coefficient. That is, the more operating condition features, the larger the similarity coefficient, and the larger the weight coefficient, the larger the operating condition similarity coefficient, and vice versa.

[0112] In this embodiment, the operating condition similarity coefficient = Where m is the number of similar working condition features, ks is the similarity coefficient of the s-th similar working condition feature, and qs is the weight coefficient of the s-th working condition feature.

[0113] In this embodiment, the constructed dataset-control group can clearly demonstrate the combustion efficiency and pollutant emissions under different control strategies in different operating conditions, providing strong data support for the subsequent construction of regulation models.

[0114] In some embodiments of this application, each dataset-control group includes a first dataset sequence, a second dataset sequence, and a third dataset sequence, specifically:

[0115] The historical monitoring data in the historical efficiency data subset of each dataset in the same dataset-control group are compared with the standard efficiency data range of the corresponding working condition characteristics, and the historical efficiency coefficient of each dataset is calculated based on the comparison results.

[0116] The standard efficiency data range includes several standard efficiency ranges, and each standard efficiency range is mapped to a corresponding preset combustion efficiency.

[0117] The historical monitoring data in the historical emission data subset of each dataset in the same dataset-control group are compared with the standard emission data range of the corresponding operating conditions, and the historical emission coefficient of each dataset is calculated based on the comparison results.

[0118] The historical comprehensive control coefficients for each dataset are calculated based on the historical efficiency coefficients and historical emission coefficients of each dataset.

[0119] Sort all datasets in the same dataset-control group according to their historical efficiency coefficients to obtain the first dataset sequence of the dataset-control group;

[0120] Sort all datasets in the same dataset-control group according to historical emission coefficients to obtain the second dataset sequence of the dataset-control group;

[0121] Sort all datasets in the same dataset-control group according to the historical comprehensive control coefficient to obtain the third dataset sequence of the dataset-control group.

[0122] In this embodiment, the standard efficiency data range refers to several efficiency data ranges that the boiler combustion efficiency should reach under normal operating conditions for each operating condition. The correspondence between each efficiency data in the standard efficiency data range and the preset combustion efficiency is set based on historical data. The standard emission data range refers to the emission data range that the concentration of pollutants emitted by the boiler combustion should meet within the environmental protection standard range for each operating condition.

[0123] In this embodiment, the combustion efficiency is sorted from smallest to largest under the corresponding operating conditions and historical control data subset based on the first dataset sequence. The emission situation is sorted from worst to best under the corresponding operating conditions and historical control data subset based on the second dataset sequence. The third dataset sequence comprehensively considers combustion efficiency and pollutant emissions to obtain the comprehensive sorting result under the corresponding operating conditions and historical control data subset (i.e., from low combustion efficiency and poor emission situation to high combustion efficiency and good emission situation).

[0124] In some embodiments of this application, the historical efficiency coefficient is calculated using the following formula:

[0125] ;

[0126] Where L1 is the historical efficiency coefficient, g1 is the first efficiency conversion coefficient, g2 is the second efficiency conversion coefficient, w1 is the number of historical monitoring data points within the corresponding standard efficiency data interval in the same historical efficiency data subset, and w2 is the total number of historical monitoring data points in the same historical efficiency data subset. This is the v1st historical monitoring data. This refers to the characteristic efficiency data within the standard efficiency data interval where the v1-th historical monitoring data falls. The weighting coefficient for the v1th historical monitoring data;

[0127] The formula for calculating the historical emission coefficient is as follows:

[0128] ;

[0129] Where L2 is the historical emission coefficient, u1 is the first emission conversion coefficient, u2 is the second emission conversion coefficient, w3 is the number of historical monitoring data points within the corresponding standard emission data range in the same historical emission data subset, and w4 is the total number of historical monitoring data points in the same historical efficiency data subset. This is the v2nd historical monitoring data. The characteristic emission data within the standard emission data range where the v2th historical monitoring data falls. The weighting coefficient for the v2th historical monitoring data.

[0130] In this embodiment, characteristic efficiency data refers to the efficiency data of the maximum boiler combustion efficiency mapped in the standard efficiency data range, and characteristic emission data refers to the minimum emission data in the standard emission data range.

[0131] In this embodiment, the first efficiency conversion coefficient and the first emission conversion coefficient refer to converting the ratio of the number of items into values ​​with the same dimension as the efficiency coefficient and the emission coefficient. When the ratio of the number of items is larger, the efficiency coefficient and the emission coefficient are larger, and vice versa. The second efficiency conversion coefficient and the second emission conversion coefficient refer to converting the absolute value of the difference between efficiency data and the absolute value of the difference between emission data into values ​​with the same dimension as the efficiency coefficient and the emission coefficient. When the absolute value of the data difference is smaller, the efficiency coefficient and the emission coefficient are larger, and vice versa.

[0132] In this embodiment, by calculating the historical efficiency coefficient and historical emission coefficient, the combustion efficiency performance, pollutant emissions and overall situation of each dataset under the corresponding operating conditions and the corresponding historical control data subset can be accurately quantified. This lays the foundation for determining the adjustment mapping table and building the adjustment model, thereby realizing the adaptive optimization adjustment of the boiler combustion process and improving the overall performance and stability of boiler operation.

[0133] In some embodiments of this application, each dataset-control group is analyzed to obtain a first adjustment mapping table, a second adjustment mapping table, and a third adjustment mapping table for each operating condition characteristic, including:

[0134] Each dataset in the same first dataset sequence is compared sequentially with the historical control data subset and historical efficiency data subset of other datasets to obtain the first control data difference of each dataset and the efficiency data difference of the historical control data subset of other datasets, and the first correlation matrix of each dataset is generated by combining the sequence order.

[0135] The initial adjustment mapping relationship between historical control data and historical efficiency data is determined based on the first association matrix of each dataset.

[0136] The initial adjustment mapping relationship between the historical control data and historical efficiency data determined by all datasets in the same first dataset sequence is summarized and integrated to obtain the first adjustment mapping relationship and generate the first adjustment mapping table under the working condition characteristics of the corresponding dataset-control group.

[0137] The first adjustment mapping table includes a first adjustment amount of several control data, each first adjustment amount is mapped to several efficiency data, the change in efficiency data and the increase in efficiency coefficient.

[0138] Each dataset in the same second dataset sequence is compared sequentially with the historical control data subset and historical emission data subset of other datasets to obtain the second control data difference of each dataset with the historical control data subset and the emission data difference of the historical emission data subset of other datasets. The second correlation matrix of each dataset is then generated by combining the sequence order.

[0139] The initial adjustment mapping relationship between historical control data and historical emission data is determined based on the second correlation matrix of each dataset;

[0140] The initial adjustment mapping relationship between the historical control data and historical emission data determined by all datasets in the same second dataset sequence is summarized and integrated to obtain the second adjustment mapping relationship and generate the second adjustment mapping table under the operating condition characteristics of the corresponding dataset-control group.

[0141] The second adjustment mapping table includes a second adjustment amount of several control data, each second adjustment amount is mapped to several emission data, the amount of change in emission data and the amount of increase in emission coefficient;

[0142] Each dataset in the same third dataset sequence is compared sequentially with the historical control data subset, historical efficiency data subset, and historical emission data subset of other datasets to obtain the third control data difference, efficiency data difference, and emission data difference of each dataset with the historical control data subset, historical efficiency data subset, and historical emission data subset of other datasets. The third correlation matrix of each dataset is then generated by combining the sequence order.

[0143] The initial adjustment mapping relationship between historical control data, historical efficiency data, and historical emission data is determined based on the third correlation matrix of each dataset;

[0144] The initial regulation mapping relationship of all datasets in the same third dataset sequence is summarized and integrated to obtain the third regulation mapping relationship and generate the third regulation mapping table under the operating condition characteristics of the corresponding dataset-control group.

[0145] The third adjustment mapping table includes a third adjustment amount of several control data, each third adjustment amount is mapped to several efficiency data and emission data, the change amount of efficiency data and emission data and the increase amount of comprehensive control coefficient.

[0146] In this embodiment, the first correlation matrix includes several triple correlation information of first control data difference - efficiency data difference - efficiency coefficient difference between each dataset and other datasets; the second correlation matrix includes several triple correlation information of second control data difference - emission data difference - emission coefficient difference; and the third correlation matrix includes several triple correlation information of third control data difference - (efficiency data difference and emission data difference) - comprehensive control coefficient difference.

[0147] In this embodiment, the first control data difference refers to the data and specific numerical differences (i.e., adjustment amount) of the same historical monitoring data belonging to the control category in different datasets. The efficiency data difference refers to the data and specific numerical differences (i.e., change amount) of the same historical monitoring data belonging to the efficiency category in different datasets. The efficiency coefficient difference refers to the difference in the historical efficiency coefficients corresponding to different datasets. The first correlation matrix directly reflects the impact of different control data on efficiency data and combustion efficiency. The first adjustment mapping table includes several efficiency data affected by each control data, the change amount of the adjustment amount of the corresponding control data for each affected efficiency data, and the increase amount of the efficiency coefficient.

[0148] In this embodiment, the meanings of the second control data difference, emission data difference, and emission coefficient difference are similar to those of the first control data difference, efficiency data difference, and efficiency coefficient difference. However, the focus is on historical monitoring data and historical emission coefficients belonging to the emission category, which will not be elaborated here. The corresponding initial adjustment mapping relationship includes several emission data affected by each control data, the change of the adjustment amount of the corresponding control data for each affected emission data, and the increase of the emission coefficient.

[0149] In this embodiment, the third control data difference is still the data with significant differences in historical monitoring data and the specific difference values ​​between different datasets. The efficiency data difference and emission data difference correspond to the data and numerical differences with significant differences in combustion efficiency and pollutant emissions, respectively. The comprehensive control coefficient difference is the difference in the historical comprehensive control coefficients of different datasets. The difference is that the historical monitoring data in the third control data difference can affect both efficiency data and emission data. The corresponding initial adjustment mapping relationship includes the efficiency data and emission data affected by each control data, the change of the adjustment amount of the corresponding control data on each affected efficiency data and emission data, and the increase of the comprehensive control coefficient.

[0150] In this embodiment, the first adjustment mapping table mainly records the adjustment mapping relationship between historical control data and historical efficiency data, providing a basis for adjustment when only combustion efficiency optimization is needed; the second adjustment mapping table mainly records the adjustment mapping relationship between historical control data and historical emission data, used for adjustment operations with the goal of reducing pollutant emissions only; the third adjustment mapping table integrates the relationship between historical control data, historical efficiency data, and historical emission data, and is suitable for comprehensive optimization adjustment scenarios that need to take into account both combustion efficiency and pollutant emissions.

[0151] In this embodiment, by constructing several adjustment mapping tables, a key data foundation and mapping rules are provided for the subsequent construction of the adjustment model. This enables the adjustment model to accurately adjust the control parameters of boiler combustion according to different operating conditions and objectives, thereby improving the overall performance and stability of boiler operation.

[0152] In some embodiments of this application, a first regulation model, a second regulation model, and a third regulation model are constructed, including:

[0153] A first training set is constructed based on the first adjustment mapping table of all working conditions, and the first adjustment model is obtained by training the model based on the first training set.

[0154] The first training set uses operating condition characteristics, the increase in efficiency coefficient, and the change in several efficiency data as training input data, and the corresponding control data and the first adjustment amount of the control data as training output data.

[0155] A second training set is constructed based on the second adjustment mapping table of all working conditions, and the second adjustment model is obtained by training the model based on the second training set.

[0156] The second training set uses operating condition characteristics, the increase in emission coefficients, and the change in several emission data as training input data, and the corresponding control data and the second adjustment of the control data as training output data.

[0157] A third training set is constructed based on the third adjustment mapping table of all working conditions, and the third adjustment model is obtained by training the model based on the third training set.

[0158] The third training set uses operating condition characteristics, the increase in the comprehensive control coefficient, and changes in several emission and efficiency data as training input data, and the corresponding control data and the third adjustment of the control data as training output data.

[0159] In this embodiment, by constructing multiple adjustment models, precise parameter adjustments can be made for different adjustment objectives (such as only improving combustion efficiency, only reducing pollutant emissions, or both). Together, these models constitute the core of the boiler combustion adaptive adjustment system, providing a strong guarantee for the efficient, stable, and environmentally friendly operation of the boiler.

[0160] In some embodiments of this application, the preset adjustment conditions include:

[0161] The preset adjustment conditions include a first adjustment condition, a second adjustment condition, and a third adjustment condition;

[0162] Pre-set efficiency coefficient thresholds and emission coefficient thresholds;

[0163] When the real-time efficiency coefficient is less than the efficiency coefficient threshold and the real-time emission coefficient is not less than the emission coefficient threshold, the first adjustment condition is triggered and the first adjustment model is selected for adaptive adjustment.

[0164] When the real-time efficiency coefficient is not less than the efficiency coefficient threshold and the real-time emission coefficient is less than the emission coefficient threshold, the second adjustment condition is triggered and the second adjustment model is selected for adaptive adjustment.

[0165] When the real-time efficiency coefficient is less than the efficiency coefficient threshold and the real-time emission coefficient is less than the emission coefficient threshold, the third adjustment condition is triggered and the third adjustment model is selected for adaptive adjustment.

[0166] In this embodiment, the efficiency coefficient threshold refers to the minimum efficiency coefficient that meets the boiler combustion efficiency requirements under the corresponding operating conditions, and the emission coefficient threshold refers to the minimum emission coefficient that meets the pollutant emission requirements under the corresponding operating conditions.

[0167] In this embodiment, by setting the adjustment conditions for each adjustment model, a mechanism for automatically selecting the adjustment model is formed, enabling the boiler combustion system to automatically and accurately select the most suitable adjustment model for adaptive adjustment based on the comparison results of the efficiency coefficient and emission coefficient monitored in real time with the preset threshold, thereby improving the adjustment efficiency.

[0168] In some embodiments of this application, if so, acquiring real-time monitoring data and determining whether a preset adjustment condition is triggered, selecting a corresponding adjustment model, generating an adjustment strategy, and issuing an adjustment command includes:

[0169] Acquire real-time monitoring data and determine real-time operating condition characteristics;

[0170] The real-time efficiency coefficient is calculated based on the real-time monitoring data belonging to the efficiency category, and the real-time emission coefficient is calculated based on the real-time monitoring data belonging to the emission category.

[0171] The real-time efficiency coefficient and real-time emission coefficient are compared with the efficiency coefficient threshold and emission coefficient threshold corresponding to the real-time operating condition characteristics, respectively, and the preset adjustment conditions are determined based on the comparison results.

[0172] If the first adjustment condition is triggered, the first adjustment model is selected, and the real-time operating condition characteristics, real-time efficiency coefficient difference, efficiency data to be adjusted and the corresponding expected change are input into the first adjustment model to obtain the data to be controlled and the corresponding first adjustment amount and generate the adjustment strategy.

[0173] If the second adjustment condition is triggered, the second adjustment model is selected, and the real-time operating condition characteristics, real-time emission coefficient difference, emission data to be adjusted and the corresponding expected change are input into the second adjustment model to obtain the data to be controlled and the corresponding second adjustment amount and generate the adjustment strategy.

[0174] If the third adjustment condition is triggered, the third adjustment model is selected, and the real-time operating condition characteristics, real-time comprehensive control coefficient difference, efficiency data to be adjusted, emission data to be adjusted, and the corresponding expected change are input into the third adjustment model to obtain the data to be controlled and the corresponding third adjustment quantity and generate the adjustment strategy.

[0175] The adjustment strategy is converted into adjustment commands and sent to the boiler combustion control system.

[0176] In this embodiment, by acquiring monitoring data in real time and accurately determining triggering conditions, it can be ensured that the boiler combustion system can quickly and accurately select the most suitable adjustment model when facing different operating conditions and performance requirements.

[0177] In this embodiment, the method for determining real-time operating characteristics, the calculation method for real-time efficiency coefficient and real-time emission coefficient are the same as above, and will not be repeated here. The efficiency data to be adjusted refers to the efficiency data that is not in the corresponding standard efficiency data range. The expected change is obtained by subtracting the efficiency data to be adjusted from the corresponding standard efficiency data range. The same applies to the emission data to be adjusted and the corresponding expected change, and will not be repeated here.

[0178] In this embodiment, by constructing multiple adjustment models and determining whether adjustment conditions are triggered, a reasonable adjustment model is selected to issue adjustment commands, thereby achieving precise and adaptive adjustment of the boiler combustion process. This not only improves the response speed of adjustment but also significantly enhances the overall performance and stability of boiler operation, providing solid technical support for the efficient and environmentally friendly operation of the boiler.

[0179] In some embodiments of this application, a boiler combustion adaptive control system is also included:

[0180] The construction module is used to construct several datasets based on historical monitoring data of the boiler combustion process. Each dataset includes a subset of historical operating condition data, a subset of historical control data, a subset of historical efficiency data, and a subset of historical emission data.

[0181] The classification model is used to determine the working condition characteristics based on a subset of historical working condition data in each dataset, calculate the working condition similarity coefficient between different datasets, classify several datasets according to the working condition similarity coefficient, and obtain multiple datasets-control groups.

[0182] The analysis module is used to analyze each dataset-control group to obtain the first regulation mapping table, the second regulation mapping table and the third regulation mapping table for each working condition feature, and to construct the first regulation model, the second regulation model and the third regulation model respectively.

[0183] The adjustment module is used to acquire real-time monitoring data and determine whether preset adjustment conditions are triggered. If so, it selects the corresponding adjustment model, generates an adjustment strategy, and issues adjustment instructions.

[0184] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of this application, and these improvements and substitutions should also be considered within the scope of protection of this application.

Claims

1. A method of adaptive regulation of a boiler combustion, characterized by, The method comprises the following steps: According to the historical monitoring data of the boiler combustion process, a plurality of data sets are constructed, each of which comprises a historical working condition data subset, a historical control data subset, a historical efficiency data subset, and a historical emission data subset; According to the historical working condition data subset in each data set, the working condition characteristics are determined, the working condition similarity coefficients between different data sets are calculated, and the plurality of data sets are classified according to the working condition similarity coefficients to obtain a plurality of data set-control groups; Each data set-control group is analyzed to obtain a first adjustment mapping table, a second adjustment mapping table, and a third adjustment mapping table of each working condition characteristic, and a first adjustment model, a second adjustment model, and a third adjustment model are constructed respectively; Real-time monitoring data are acquired, and it is judged whether a preset adjustment condition is triggered. If yes, a corresponding adjustment model is selected, an adjustment strategy is generated, and an adjustment instruction is issued.

2. The boiler combustion adaptive tuning method of claim 1, wherein, According to the historical monitoring data of the boiler combustion process, a plurality of data sets are constructed, comprising: A plurality of historical monitoring logs of the boiler combustion process are acquired, and a plurality of collection time nodes of each historical monitoring log are set according to a preset monitoring time interval; According to the plurality of collection time nodes, the historical monitoring data in each historical monitoring log are collected, and the historical monitoring data difference value at the previous adjacent collection time node is calculated; The historical monitoring data with a historical monitoring data difference value greater than a preset difference threshold and belonging to a working condition category are screened out, and an initial monitoring data set of the working condition category is constructed according to the screened historical monitoring data; The change coefficient of the initial monitoring data set at the current collection time node is calculated; If the change coefficient is greater than a preset change coefficient threshold, the initial monitoring data set at the current collection time node is set as a historical working condition data subset; A historical control data subset is constructed according to the historical monitoring data belonging to a control category at the current collection time node; A historical efficiency data subset and a historical emission data subset are respectively constructed according to the historical monitoring data belonging to an efficiency category and the historical monitoring data belonging to an emission category at the historical feedback node corresponding to the current collection time node, and a data set is constructed by combining the historical working condition data subset and the historical control data subset; A plurality of data sets are sequentially generated.

3. The boiler combustion adaptive tuning method of claim 2, wherein, The working condition similarity coefficients between different data sets are calculated, and the plurality of data sets are classified according to the working condition similarity coefficients to obtain a plurality of data set-control groups, comprising: The historical working condition data subset in each data set is subjected to feature extraction to obtain corresponding working condition characteristics, and each working condition characteristic is mapped with a corresponding weight coefficient; The working condition characteristics corresponding to different data sets are subjected to similarity analysis to obtain the number of working condition characteristics with similarity in different data sets and the corresponding similarity coefficients, and the working condition similarity coefficients between different data sets are calculated by combining the weight coefficients of the corresponding working condition characteristics; A working condition similarity coefficient threshold is preset; The data sets with a working condition similarity coefficient greater than the working condition similarity coefficient threshold are classified into the same category, and a data set-control group is constructed based on all the data sets of the category; A plurality of data set-control groups are sequentially constructed.

4. The boiler combustion adaptive tuning method of claim 3, wherein, Each of the data set-control groups comprises a first data set sequence, a second data set sequence, and a third data set sequence, specifically: comparing the historical monitoring data in the historical efficiency data subset of each data set in the same data set-contrast group with the standard efficiency data interval corresponding to the working condition feature, and calculating a historical efficiency coefficient of each data set according to a comparison result; The standard efficiency data interval includes a plurality of standard efficiency intervals, and each standard efficiency interval is mapped with a corresponding preset combustion efficiency. comparing the historical monitoring data in the historical emission data subset of each data set in the same data set-contrast group with the standard emission data interval corresponding to the working condition feature, and calculating a historical emission coefficient of each data set according to a comparison result; According to the historical efficiency coefficient and the historical emission coefficient of each data set, a corresponding historical comprehensive control coefficient is calculated. According to the historical efficiency coefficient, all data sets in the same data set-contrast group are sorted to obtain a first data set sequence of the data set-contrast group. According to the historical emission coefficient, all data sets in the same data set-contrast group are sorted to obtain a second data set sequence of the data set-contrast group. According to the historical comprehensive control coefficient, all data sets in the same data set-contrast group are sorted to obtain a third data set sequence of the data set-contrast group.

5. The boiler combustion adaptive adjustment method of claim 4, wherein, The calculation formula of the historical efficiency coefficient is: ； Wherein, L1 is a historical efficiency coefficient, g1 is a first efficiency conversion coefficient, g2 is a second efficiency conversion coefficient, w1 is the number of historical monitoring data in the corresponding standard efficiency data interval in the same historical efficiency data subset, w2 is the total number of historical monitoring data in the same historical efficiency data subset, is the v1th historical monitoring data, is the feature efficiency data in the standard efficiency data interval where the v1th historical monitoring data is located, is the weight coefficient of the v1th historical monitoring data; The calculation formula of the historical emission coefficient is: ； Wherein, L2 is a historical emission coefficient, u1 is a first emission conversion coefficient, u2 is a second emission conversion coefficient, w3 is the number of historical monitoring data in the corresponding standard emission data interval in the same historical emission data subset, w4 is the total number of historical monitoring data in the same historical efficiency data subset, is the v2th historical monitoring data, is the feature emission data in the standard emission data interval where the v2th historical monitoring data is located, is the weight coefficient of the v2th historical monitoring data.

6. The boiler combustion adaptive tuning method of claim 5, wherein, Each data set-contrast group is analyzed to obtain a first adjustment mapping table, a second adjustment mapping table and a third adjustment mapping table of each working condition feature, including: comparing each data set in the same first data set sequence with the historical control data subset and the historical efficiency data subset of other data sets in turn to obtain a first control data difference of the historical control data subset and an efficiency data difference of the historical efficiency data subset of each data set and other data sets, and generating a first correlation matrix of each data set in combination with the sequence order; determining an initial adjustment mapping relationship of the historical control data and the historical efficiency data according to the first correlation matrix of each data set; The initial adjustment mapping relationship of the historical control data and the historical efficiency data determined by all data sets in the same first data set sequence is integrated to obtain a first adjustment mapping relationship and generate a first adjustment mapping table of the working condition feature under the corresponding data set-contrast group; The first adjustment mapping table includes a plurality of first adjustment amounts of control data, each first adjustment amount is mapped with a plurality of efficiency data, a change amount of the efficiency data and an improvement amount of the efficiency coefficient; comparing each data set in the same second data set sequence with the historical control data subset and the historical emission data subset of other data sets in turn to obtain a second control data difference of the historical control data subset and an emission data difference of the historical emission data subset of each data set and other data sets, and generating a second correlation matrix of each data set in combination with the sequence order; determining an initial adjustment mapping relationship of the historical control data and the historical emission data according to the second correlation matrix of each data set; The initial adjustment mapping relationship of the historical control data and the historical emission data of all the data sets in the same second data set sequence is integrated to obtain a second adjustment mapping relationship and generate a second adjustment mapping table corresponding to the data set-contrast group under the working condition characteristics; The second adjustment mapping table includes second adjustment amounts of the control data, each second adjustment amount is mapped with a plurality of emission data, a change amount of the emission data, and an improvement amount of the emission coefficient; Each data set in the same third data set sequence is compared with the historical control data subset, the historical efficiency data subset, and the historical emission data subset of other data sets in sequence to obtain third control data differences of each data set and the historical control data subset of other data sets, efficiency data differences of the historical efficiency data subset, and emission data differences of the historical emission data subset, and a third correlation matrix of each data set is generated in combination with the sequence order; The initial adjustment mapping relationship of the historical control data and the historical efficiency data and the historical emission data is determined according to the third correlation matrix of each data set; The initial adjustment mapping relationship of the historical control data and the historical efficiency data and the historical emission data of all the data sets in the same third data set sequence is integrated to obtain a third adjustment mapping relationship and generate a third adjustment mapping table corresponding to the data set-contrast group under the working condition characteristics; The third adjustment mapping table includes third adjustment amounts of the control data, each third adjustment amount is mapped with a plurality of efficiency data and emission data, a change amount of the efficiency data and the emission data, and an improvement amount of the comprehensive control coefficient.

7. A method of boiler combustion adaptive regulation as claimed in claim 6, characterised by, The first adjustment model, the second adjustment model, and the third adjustment model are respectively constructed, including: A first training set is constructed according to the first adjustment mapping table of all working condition characteristics, and a first adjustment model is obtained by model training according to the first training set; The first training set takes the working condition characteristics, the improvement amount of the efficiency coefficient, and the change amount of the plurality of efficiency data as training input data, and takes the corresponding control data and the first adjustment amount of the control data as training output data; A second training set is constructed according to the second adjustment mapping table of all working condition characteristics, and a second adjustment model is obtained by model training according to the second training set; The second training set takes the working condition characteristics, the improvement amount of the emission coefficient, and the change amount of the plurality of emission data as training input data, and takes the corresponding control data and the second adjustment amount of the control data as training output data; A third training set is constructed according to the third adjustment mapping table of all working condition characteristics, and a third adjustment model is obtained by model training according to the third training set; The third training set takes the working condition characteristics, the improvement amount of the comprehensive control coefficient, and the change amount of the plurality of emission data and efficiency data as training input data, and takes the corresponding control data and the third adjustment amount of the control data as training output data.

8. A method of boiler combustion adaptive regulation as claimed in claim 7, characterised by, The preset adjustment condition includes: The preset adjustment condition includes a first adjustment condition, a second adjustment condition, and a third adjustment condition; An efficiency coefficient threshold and an emission coefficient threshold are preset; When the real-time efficiency coefficient is less than the efficiency coefficient threshold and the real-time emission coefficient is not less than the emission coefficient threshold, a first adjustment condition is triggered and a first adjustment model is selected for adaptive adjustment; When the real-time efficiency coefficient is not less than the efficiency coefficient threshold and the real-time emission coefficient is less than the emission coefficient threshold, a second adjustment condition is triggered and a second adjustment model is selected for adaptive adjustment; When the real-time efficiency coefficient is less than the efficiency coefficient threshold and the real-time emission coefficient is less than the emission coefficient threshold, a third adjustment condition is triggered and a third adjustment model is selected for adaptive adjustment.

9. A method of boiler combustion adaptive regulation as claimed in claim 8, characterised by, If yes, real-time monitoring data is acquired and it is determined whether a preset adjustment condition is triggered, a corresponding adjustment model is selected, an adjustment strategy is generated and an adjustment instruction is issued, including: Real-time monitoring data is acquired and real-time working condition characteristics are determined; A real-time efficiency coefficient is calculated according to real-time monitoring data belonging to an efficiency category and a real-time emission coefficient is calculated according to real-time monitoring data belonging to an emission category; The real-time efficiency coefficient and the real-time emission coefficient are compared with an efficiency coefficient threshold and an emission coefficient threshold corresponding to the real-time working condition characteristics respectively, and it is determined whether a preset adjustment condition is triggered according to a comparison result; If the first adjustment condition is triggered, a first adjustment model is selected, the real-time working condition characteristics, a real-time efficiency coefficient difference value, to-be-adjusted efficiency data and a corresponding expected change amount are input into the first adjustment model, to-be-controlled data and a corresponding first adjustment amount are obtained and an adjustment strategy is generated; If the second adjustment condition is triggered, a second adjustment model is selected, the real-time working condition characteristics, a real-time emission coefficient difference value, to-be-adjusted emission data and a corresponding expected change amount are input into the second adjustment model, to-be-controlled data and a corresponding second adjustment amount are obtained and an adjustment strategy is generated; If the third adjustment condition is triggered, a third adjustment model is selected, the real-time working condition characteristics, a real-time comprehensive control coefficient difference value, to-be-adjusted efficiency data, to-be-adjusted emission data and a corresponding expected change amount are input into the third adjustment model, to-be-controlled data and a corresponding third adjustment amount are obtained and an adjustment strategy is generated; The adjustment strategy is converted into an adjustment instruction and is issued to a boiler combustion control system.

10. A boiler combustion adaptive tuning system, characterized by, Including: A construction module is configured to construct a plurality of data sets according to historical monitoring data of a boiler combustion process, each of the data sets including a historical working condition data subset, a historical control data subset, a historical efficiency data subset and a historical emission data subset; A classification model is configured to determine working condition characteristics according to the historical working condition data subset in each data set, calculate working condition similarity coefficients between different data sets, classify the plurality of data sets according to the working condition similarity coefficients and obtain a plurality of data set-contrast groups; An analysis module is configured to analyze each data set-contrast group, obtain a first adjustment mapping table, a second adjustment mapping table and a third adjustment mapping table of each working condition characteristic and construct a first adjustment model, a second adjustment model and a third adjustment model respectively; An adjustment module is configured to acquire real-time monitoring data and determine whether a preset adjustment condition is triggered, and if yes, select a corresponding adjustment model, generate an adjustment strategy and issue an adjustment instruction.

Images