Power plant boiler combustion performance prediction method based on local knowledge fusion

Abstract

A power station boiler combustion performance prediction method and device based on local knowledge fusion and a storage medium, the method comprising the steps of: establishing an original sample set of a power station boiler; obtaining stable modal samples in the original sample set; obtaining characteristics of the power station boiler; obtaining an operating modal subspace of the power station boiler according to the stable modal samples and the characteristics; obtaining a monotonicity submodel of the power station boiler according to the operating modal subspace and the characteristics; obtaining a combustion performance prediction model of the power station boiler according to the operating modal subspace and the monotonicity submodel; and calculating the combustion performance of the power station boiler according to the characteristics and the combustion performance prediction model. The present application overcomes the problem of multimodality and local monotonicity between characteristics in the existing combustion optimization data, simultaneously considers the multimodality of the data and the prior knowledge followed by the sample points in different subspaces, thereby effectively monitoring the combustion performance of the power station boiler.

Description

Technical Field

[0001] This invention belongs to the field of thermal power boiler technology, specifically relating to a method, device, and storage medium for predicting the combustion performance of power plant boilers based on local knowledge fusion. Background Technology

[0002] According to the national "three reforms and linkage" policy, thermal power boilers need to implement flexible peak-shaving and pressure regulation, making the diagnosis of boiler combustion stability crucial. Boiler efficiency is generally calculated using the inverse balance method. The main parameters required for calculating boiler efficiency using this method are fly ash carbon content, slag carbon content, and CO content. However, the measurement of these parameters has significant errors and exhibits a lag. Therefore, researching the measurement methods for fly ash carbon content, slag carbon content, and carbon monoxide content is fundamental to solving these problems.

[0003] With the rapid development of artificial intelligence, researchers have used algorithms such as support vector machines, artificial neural networks, and extreme learning machines to model boilers. They then use these models to predict the carbon content of fly ash, carbon content of slag, and carbon monoxide content, finally inputting these three quantities into an inverse equilibrium model to calculate boiler efficiency. However, these methods do not consider the multimodal nature of the data itself or the prior knowledge followed by sample points in different subspaces, resulting in complex models with low accuracy that fail to adequately explain the nature of combustion. Summary of the Invention

[0004] In view of the above problems, the present invention provides a method, apparatus and storage medium for predicting the combustion performance of power plant boilers based on local knowledge fusion to overcome the above problems or at least partially solve the above problems.

[0005] To address the aforementioned technical problems, this invention provides a method for predicting the combustion performance of power plant boilers based on local knowledge fusion, the method comprising the following steps:

[0006] Establish the original sample set of power plant boilers;

[0007] Obtain stable modal samples from the original sample set;

[0008] Obtain the characteristics of the power plant boiler;

[0009] The operating mode subspace of the power plant boiler is obtained based on the stable mode samples and the features.

[0010] The monotonicity sub-model of the power plant boiler is obtained based on the operating mode subspace and the features.

[0011] The combustion performance prediction model of the power plant boiler is obtained based on the operating mode subspace and the monotonicity submodel.

[0012] The combustion performance of the power plant boiler is calculated based on the aforementioned characteristics and the combustion performance prediction model.

[0013] Preferably, establishing the original sample set of power plant boilers includes the following steps:

[0014] Query the power plant SIS system database and the power plant MIS system database;

[0015] Obtain power plant boiler operation data from the power plant SIS system database;

[0016] Obtain offline test data of the power plant boiler from the power plant MIS system database;

[0017] An original sample set was established based on the power plant boiler operation data and the power plant boiler offline test data.

[0018] Preferably, obtaining stable modal samples from the original sample set includes the following steps:

[0019] The original sample set is decomposed into transitional mode samples and stable mode samples:

[0020] Delete all the aforementioned transition mode samples.

[0021] Preferably, obtaining the characteristics of the power plant boiler includes the following steps:

[0022] Obtain the operating condition characteristics of the power plant boiler;

[0023] Obtain the target variables of the power plant boiler;

[0024] Obtain the characteristic variables of the power plant boiler.

[0025] Preferably, obtaining the operating mode subspace of the power plant boiler based on the stable mode samples and the features includes the following steps:

[0026] Obtain the stable modal samples and the runtime condition features from the features;

[0027] Based on the stable modal samples, the operating condition features are probabilistically clustered;

[0028] The operating mode subspace of the power plant boiler is obtained;

[0029] Calculate the probability that each running sample point belongs to each of the said running mode subspaces.

[0030] Preferably, obtaining the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features includes the following steps:

[0031] Obtain the target variable and feature variables from the features;

[0032] For each of the said operating mode subspaces, calculate the directional relationship between the target variable and the feature variable;

[0033] Extract the monotonicity patterns of the feature variables in each of the operating mode subspaces;

[0034] Using the monotonicity law as a constraint, a monotonicity sub-model of the target variable is established for each of the operating mode subspaces.

[0035] Preferably, obtaining the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel includes the following steps:

[0036] Obtain the prediction results of the monotonicity sub-model for the target variable;

[0037] Obtain the probability that each running sample point belongs to each of the said running mode subspaces;

[0038] The probabilities are used as weights to perform a weighted summation on each prediction result;

[0039] The combustion performance prediction model is obtained by summing the results.

[0040] This application also provides a power plant boiler combustion performance prediction device, comprising:

[0041] The original sample set creation module is used to create the original sample set for power plant boilers.

[0042] A stable mode sample acquisition module is used to acquire stable mode samples from the original sample set;

[0043] The feature acquisition module is used to acquire the features of the power plant boiler;

[0044] The operation mode subspace acquisition module is used to acquire the operation mode subspace of the power plant boiler based on the stable mode samples and the features;

[0045] The monotonicity sub-model acquisition module is used to acquire the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features.

[0046] The combustion performance prediction model acquisition module is used to acquire the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel.

[0047] The combustion performance calculation module is used to calculate the combustion performance of the power plant boiler based on the characteristics and the combustion performance prediction model.

[0048] This application also provides an electronic device, characterized in that the electronic device comprises:

[0049] At least one processor; and,

[0050] A memory communicatively connected to the at least one processor; wherein,

[0051] The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to execute any of the aforementioned methods for predicting the combustion performance of power plant boilers based on local knowledge fusion.

[0052] This application also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute any of the aforementioned methods for predicting the combustion performance of power plant boilers based on local knowledge fusion.

[0053] One or more technical solutions in the embodiments of the present invention have at least the following technical effects or advantages: The method, device and storage medium for predicting the combustion performance of power plant boilers based on local knowledge fusion provided in this application overcomes the problems of multimodality and local monotonicity between features in the combustion optimization data in the prior art. At the same time, it considers the multimodality of the data and the prior knowledge followed by the sample points in different subspaces, thereby effectively monitoring the combustion performance of power plant boilers. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a flowchart illustrating a method for predicting the combustion performance of a power plant boiler based on local knowledge fusion, provided in an embodiment of the present invention.

[0056] Figure 2 This is a schematic diagram of probabilistic clustering in a power plant boiler combustion performance prediction method based on local knowledge fusion provided in an embodiment of the present invention;

[0057] Figure 3 A schematic diagram illustrating the effect of oxygen under typical low-load conditions on the carbon content of fly ash and slag.

[0058] Figure 4 A schematic diagram illustrating the effect of oxygen on the carbon content of fly ash and slag under typical medium-load modes;

[0059] Figure 5 A schematic diagram illustrating the effect of oxygen under typical high-load modes on the carbon content of fly ash and slag;

[0060] Figure 6 A schematic diagram illustrating the effect of SOFA opening degree under typical low-load mode on carbon content in fly ash and slag;

[0061] Figure 7 A schematic diagram illustrating the effect of SOFA opening degree on the carbon content of fly ash and slag under typical medium-load mode;

[0062] Figure 8 A schematic diagram illustrating the effect of SOFA opening degree under typical high-load mode on carbon content in fly ash and slag;

[0063] Figure 9 This is a schematic diagram of the structure of a power plant boiler combustion performance prediction device provided in an embodiment of the present invention;

[0064] Figure 10 This is a schematic diagram of the structure of an electronic device provided by the present invention;

[0065] Figure 11 This is a schematic diagram of the structure of a non-transitory computer-readable storage medium provided by the present invention. Detailed Implementation

[0066] The present invention will be described in detail below with reference to specific embodiments and examples, thereby making the advantages and various effects of the present invention more clearly apparent. Those skilled in the art should understand that these specific embodiments and examples are for illustrative purposes only and are not intended to limit the present invention.

[0067] Throughout this specification, unless otherwise specified, the terminology used herein should be understood as having the meaning commonly used in the art. Therefore, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. In the event of any conflict, this specification shall prevail.

[0068] Unless otherwise specified, all raw materials, reagents, instruments and equipment used in this invention can be purchased from the market or prepared by existing methods.

[0069] like Figure 1 In this application embodiment, the present invention provides a method for predicting the combustion performance of a power plant boiler based on local knowledge fusion, the method comprising the following steps:

[0070] S1: Establish the original sample set of power plant boilers;

[0071] In this embodiment of the application, establishing the original sample set of power plant boilers includes the following steps:

[0072] Query the power plant SIS system database and the power plant MIS system database;

[0073] Obtain power plant boiler operation data from the power plant SIS system database;

[0074] Obtain offline test data of the power plant boiler from the power plant MIS system database;

[0075] An original sample set was established based on the power plant boiler operation data and the power plant boiler offline test data.

[0076] In this embodiment of the application, the original sample set of the power plant boiler includes the power plant boiler operation data in the power plant SIS system and the power plant boiler offline test data in the power plant MIS system. The power plant boiler operation data includes, but is not limited to, power plant boiler damper system data, pulverizing system data, flue gas system data, and steam-water system data; the power plant boiler offline test data includes, but is not limited to, coal quality test data, fly ash carbon content test data, and slag carbon content test data.

[0077] In this embodiment of the application, the original sample set X can be represented as:

[0078]

[0079] Where xi,j represents the data of the j-th feature of the i-th sample; the X dataset has n sample points and m dimensions, and Xj represents the vector composed of all sample data of the j-th feature.

[0080] S2: Obtain stable modal samples from the original sample set;

[0081] In this embodiment of the application, obtaining stable modal samples from the original sample set includes the following steps:

[0082] The original sample set is decomposed into transitional mode samples and stable mode samples:

[0083] Delete all the aforementioned transition mode samples.

[0084] In this embodiment, the transition mode identification method can decompose the original sample set into transition mode samples and stable mode samples. Then, all transition mode samples are deleted, and the remaining data are the stable mode samples. The transition mode identification method includes, but is not limited to, information entropy, T-test, R-test, etc.

[0085] S3: Obtain the characteristics of the power plant boiler;

[0086] In this embodiment of the application, obtaining the features of the power plant boiler includes the following steps:

[0087] Obtain the operating condition characteristics of the power plant boiler;

[0088] Obtain the target variables of the power plant boiler;

[0089] Obtain the characteristic variables of the power plant boiler.

[0090] In this embodiment of the application, the operating conditions of the power plant boiler include: unit load, coal ash content, and lower heating value of coal. The target variables of the power plant boiler include: CO concentration, fly ash carbon content, and slag carbon content. The characteristic variables of the power plant boiler include: damper opening, damper swing angle, and O2 content.

[0091] S4: Obtain the operating mode subspace of the power plant boiler based on the stable mode samples and the features;

[0092] In this embodiment of the application, obtaining the operating mode subspace of the power plant boiler based on the stable mode samples and the features includes the following steps:

[0093] Obtain the stable modal samples and the runtime condition features from the features;

[0094] Based on the stable modal samples, the operating condition features are probabilistically clustered;

[0095] The operating mode subspace of the power plant boiler is obtained;

[0096] Calculate the probability that each running sample point belongs to each of the said running mode subspaces.

[0097] In this embodiment of the application, initially, several sample points are randomly selected from the original sample set as cluster centers. The initial cluster centers are: l represents the maximum number of iterations, and V represents the distance from each sample to each cluster center. (0) The distance can be expressed as d ij Then update the membership matrix u. ij (l+1) This process continues until the iteration condition is met. Finally, the membership matrix uij for each sample data point with respect to each category is obtained, which is the probability score of each sample for each category. Then, the optimal number of categories p is determined.

[0098] In the embodiments of this application, the probabilistic clustering methods include, but are not limited to, FCM, GMM and other methods, and the methods for determining the optimal number of clusters include, but are not limited to, the elbow method and other methods.

[0099] S5: Obtain the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features;

[0100] In this embodiment of the application, obtaining the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features includes the following steps:

[0101] Obtain the target variable and feature variables from the features;

[0102] For each of the said operating mode subspaces, calculate the directional relationship between the target variable and the feature variable;

[0103] Extract the monotonicity patterns of the feature variables in each of the operating mode subspaces;

[0104] Using the monotonicity law as a constraint, a monotonicity sub-model of the target variable is established for each of the operating mode subspaces.

[0105] In this embodiment, for each operational modality subspace, the directional relationships between target variables such as CO concentration, fly ash carbon content, and slag carbon content and feature variables such as damper opening, damper swing angle, and O2 content are sequentially mined. The monotonicity patterns of the feature variables in each operational modality subspace are extracted and summarized into a list. Then, using local monotonicity as a constraint, a prior machine learning model is adopted to model the monotonicity sub-model of the target variables, which can be obtained.

[0106] S6: Obtain the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel;

[0107] In this embodiment of the application, obtaining the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel includes the following steps:

[0108] Obtain the prediction results of the monotonicity sub-model for the target variable;

[0109] Obtain the probability that each running sample point belongs to each of the said running mode subspaces;

[0110] The probabilities are used as weights to perform a weighted summation on each prediction result;

[0111] The combustion performance prediction model is obtained by summing the results.

[0112] In this embodiment, a priori machine learning model is used to model each category of power plant operation data. The monotonic machine learning algorithm includes, but is not limited to, monotonic support vector machines and monotonic neural networks. The probability of each sample in each category is used as a weight, and then the prediction results of each sub-model are weighted and summed to predict the CO concentration, fly ash carbon content, and slag carbon content.

[0113] In this embodiment of the application, the membership matrix corresponding to each category for each sample is divided into classes. The CO concentration, fly ash carbon content, and slag carbon content can be expressed as follows:

[0114]

[0115]

[0116]

[0117] Here, represents the output of machine learning based on prior knowledge.

[0118] S7: Calculate the combustion performance of the power plant boiler based on the characteristics and the combustion performance prediction model.

[0119] In this embodiment, CO concentration, fly ash carbon content, and slag carbon content are input into the boiler combustion performance calculation model to obtain the boiler combustion performance corresponding to the respective operating conditions. Methods for calculating boiler combustion performance include, but are not limited to, the positive balance method, the inverse balance method, or machine learning.

[0120] The following uses a power plant boiler as an example to introduce the specific implementation steps of this method. The entire process is divided into five core parts. Figure 1 The process is as follows:

[0121] (1) Acquire power plant operation data, perform steady-state identification, and remove data.

[0122] Historical operational data obtained from the power plant includes numerous operating modes. Each operating mode has a transition period, known as a transition mode. Since stable mode data is required for modeling, a recursive local outlier factor algorithm is used for mode identification. When a transition mode is detected, it is removed. This process continues until the sliding window has processed all data, removing all transition modes. The remaining data represents the stable modes.

[0123] (2) Cluster the data, determine the optimal number of clusters, and derive the probability of each cluster.

[0124] Because the data distribution differs under different operating conditions, fuzzy clustering (FCM) is used to cluster the steady-state samples obtained in step (1) to distinguish these data distributions. Initially, p sample points are randomly selected from X as cluster centers, with the initial cluster centers being... l represents the maximum number of iterations, and V represents the distance from each sample to each cluster center. (0) The distance can be expressed as d ij Then update the membership matrix u. ij (l+1) The calculation formula is as follows:

[0125]

[0126] Update cluster center v i (l+1) The calculation formula is as follows:

[0127]

[0128] The algorithm iterates continuously until the difference between two consecutive objective functions is less than a predetermined error, at which point it stops. The objective function J is calculated using the following formula:

[0129]

[0130] The membership matrix u for each sample data for each category was obtained. ij This refers to the probability score of each sample for each category. Then, the elbow method is used to determine the optimal number of categories, p. The clustering results are as follows: Figure 2 As shown.

[0131] Compare the univariate variation plots of the corresponding features for each category to identify the local monotonicity of the corresponding features.

[0132] like Figure 2 As shown, it was found that the group formed in step (2) was most affected by the unit load. Therefore, the patterns of changes in CO concentration, fly ash carbon content and slag carbon content caused by the changes in various characteristics under different unit loads were identified.

[0133] 420MW operating conditions

[0134]

[0135] 550MW operating conditions

[0136]

[0137] 660MW operating conditions

[0138]

[0139] (—Indicates no trend is observed)

[0140] The local monotonicity under different unit loads can be obtained. Under operating conditions of 420MW, 550MW, and 660MW:

[0141] For oxygen concentration such as Figure 3 , 4 5. As we know, the carbon and CO content of fly ash decreases with increasing oxygen content, while the carbon and oxygen content of slag do not have a monotonic relationship.

[0142] For SOFA opening such as Figure 6 , 78. As known: The carbon content of fly ash increases with increasing SOFA (Solar-Air Fuel Fiber) opening at 550MW, but does not show monotonicity at 420MW and 660MW; the carbon content of slag does not show monotonicity at the three different loads; the CO content increases with increasing SOFA opening at 420MW and 660MW, but at 550MW, the CO content does not show monotonicity with SOFA opening. In summary, the carbon content of fly ash and SOFA opening show local monotonicity at a unit load of 550MW; the CO content and SOFA opening show local monotonicity at unit loads of 420MW and 660MW.

[0143] Regarding the SOFA swing angle: the carbon content of fly ash increases with the increase of SOFA swing angle at 440MW, but does not show monotonicity at 550MW and 660MW; the carbon content of slag does not show monotonicity at the three different loads; the CO content increases with the increase of SOFA swing angle at 440MW, but does not show monotonicity at 550MW and 660MW; in summary, the carbon content of fly ash shows local monotonicity with SOFA swing angle and CO content at a unit load of 420MW.

[0144] Regarding primary air pressure: the carbon content of fly ash and slag increases with the increase of primary air pressure when the unit load is 660MW. However, no monotonicity is observed at 550MW, indicating that the carbon content of fly ash and slag exhibits local monotonicity when the unit load is 660MW.

[0145] Using local monotonicity as a constraint, a priori support vector machine (SVM) model is adopted, incorporating the constraints obtained in step (3) into the additive model of the SVM. (Additive model of support vector machine)

[0146]

[0147] Where is the input sample matrix, and is the i-th input variable of the i-th sample.

[0148] The empirical knowledge of combustion with local monotonicity is incorporated as a monotonic constraint into the additive model of the support vector machine. Its key feature is that, for a continuously integrable function, monotonically increasing means that the derivative of the parameter is greater than or equal to 0. In the additive model of the support vector machine, this can be transformed into the following inequality:

[0149]

[0150] Similarly:

[0151]

[0152] Since there may be k variables that have a monotonic relationship with the output variable, the regression problem can be transformed into the following inequality optimization problem:

[0153]

[0154]

[0155] (5) Substitute the CO concentration, fly ash carbon content, and slag carbon content into the inverse equilibrium model for calculating boiler efficiency to obtain the boiler efficiency corresponding to the respective operating conditions.

[0156] According to the boiler unit's heat balance, we can obtain:

[0157] Q r =Q1+Q2+Q3+Q4+Q5+Q6;

[0158] Among them, Q r The heat input to the boiler is Q1, the heat effectively utilized by the boiler is Q2, the flue gas heat loss is Q3, the heat loss due to incomplete chemical combustion is Q4, the heat loss due to incomplete mechanical combustion is Q5, the heat dissipation loss is Q5, and other heat losses are Q6. Transforming the formula yields:

[0159] 100 = q1 + q2 + q3 + q4 + q5 + q6;

[0160] Wherein, q1 is the percentage of effective heat utilization relative to the input heat, q2 is the percentage of flue gas heat loss relative to the input heat, q3 is the percentage of incomplete chemical combustion heat loss relative to the input heat, q4 is the percentage of incomplete mechanical combustion heat loss relative to the input heat, q5 is the percentage of heat dissipation loss relative to the input heat, and q6 is the percentage of other heat losses relative to the input heat.

[0161] The boiler efficiency calculated using the reverse balance method is:

[0162] η = 100 - (q² + g³ + q⁴ + q⁵ + q⁶);

[0163] (5a) Input heat

[0164] in, It is the low calorific value of fuel.

[0165] (5b) Flue gas heat loss

[0166] in, The heat carried away by the dry flue gas. The heat carried away by the water vapor contained in the flue gas.

[0167] (5c) Heat loss due to incomplete combustion of chemicals

[0168] Where, α CO This represents the volume percentage of CO in dry flue gas. The weight of carbon completely burned per kilogram of fuel fed into the furnace. This represents the volume percentage of CO2 in the dry flue gas.

[0169] (5d) Heat loss due to incomplete combustion in machinery

[0170] Among them, A y The ash content is the mass content of the fuel as received. This refers to the carbon content per kilogram of ash.

[0171] (5e) Heat loss

[0172] Where P max P represents the maximum unit load of the boiler. real This represents the actual unit load.

[0173] (5f) Other heat losses q6=α;

[0174] Where α is the uncalculated loss coefficient.

[0175] CO2 content is an important parameter for calculating the heat loss of dry flue gas. The formula for calculating CO2 is as follows:

[0176]

[0177] Where β is the excess air coefficient.

[0178]

[0179]

[0180] The input parameters for the boiler efficiency mechanism model are: low calorific value, flue gas volume percentage, flue gas temperature, reference temperature, carbon content in fly ash and slag, ash content on the received basis, correction coefficient, S, C, N, H, and O content on the received basis, and total water content on the received basis. The input parameter is the boiler combustion efficiency.

[0181] like Figure 9 In this embodiment of the application, a power plant boiler combustion performance prediction device is also provided, comprising:

[0182] The original sample set establishment module 10 is used to establish the original sample set of the power plant boiler;

[0183] Stable mode sample acquisition module 20 is used to acquire stable mode samples in the original sample set;

[0184] Feature acquisition module 30 is used to acquire features of the power plant boiler;

[0185] The operating mode subspace acquisition module 40 is used to acquire the operating mode subspace of the power plant boiler based on the stable mode samples and the features.

[0186] The monotonicity sub-model acquisition module 50 is used to acquire the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features.

[0187] Combustion performance prediction model acquisition module 60 is used to acquire the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel.

[0188] Combustion performance calculation module 70 is used to calculate the combustion performance of the power plant boiler based on the characteristics and the combustion performance prediction model.

[0189] The power plant boiler combustion performance prediction device provided in this application can perform the power plant boiler combustion performance prediction method based on local knowledge fusion provided in the above steps.

[0190] It should be understood that the specific embodiments described above are merely illustrative or explanatory of the principles of the invention and do not constitute a limitation thereof. Therefore, any modifications, equivalent substitutions, improvements, etc., made without departing from the spirit and scope of the invention should be included within the protection scope of the invention. Furthermore, the appended claims are intended to cover all variations and modifications falling within the scope and boundaries of the appended claims, or equivalent forms of such scope and boundaries.

[0191] The following is for reference. Figure 10 The diagram illustrates a structural schematic of an electronic device 100 suitable for implementing embodiments of the present disclosure. The electronic devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 10 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0192] like Figure 10As shown, the electronic device 100 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 101, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 102 or a program loaded from a storage device 108 into a random access memory (RAM) 103. The RAM 103 also stores various programs and data required for the operation of the electronic device 100. The processing unit 101, ROM 102, and RAM 103 are interconnected via a bus 104. An input / output (I / O) interface 105 is also connected to the bus 104.

[0193] Typically, the following devices can be connected to I / O interface 105: input devices 106 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 107 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 108 including, for example, magnetic tapes, hard disks, etc.; and communication devices 109. Communication device 109 allows electronic device 100 to communicate wirelessly or wiredly with other devices to exchange data. Although electronic device 100 with various devices is shown in the figure, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0194] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 109, or installed from storage device 108, or installed from ROM 102. When the computer program is executed by processing device 101, it performs the functions defined in the methods of embodiments of this disclosure.

[0195] The following is for reference. Figure 11 It illustrates a schematic diagram of a computer-readable storage medium suitable for implementing embodiments of the present disclosure, the computer-readable storage medium storing a computer program that, when executed by a processor, can implement the multi-scale component model finite element mesh generation method as described above.

[0196] This application provides a method, device, and storage medium for predicting the combustion performance of power plant boilers based on local knowledge fusion. This overcomes the problems of multimodality and local monotonicity between features in the combustion optimization data of the prior art. At the same time, it takes into account the multimodality of the data and the prior knowledge followed by the sample points in different subspaces, thereby effectively monitoring the combustion performance of power plant boilers.

[0197] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. The above descriptions are merely specific embodiments of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

[0198] In summary, the above description is merely a preferred embodiment of the technical solution of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting the combustion performance of power plant boilers based on local knowledge fusion, characterized in that, The method includes the following steps: Establish the original sample set of power plant boilers; Obtain stable modal samples from the original sample set; Obtain the characteristics of the power plant boiler; The operating mode subspace of the power plant boiler is obtained based on the stable mode samples and the features. The monotonicity sub-model of the power plant boiler is obtained based on the operating mode subspace and the features. The combustion performance prediction model of the power plant boiler is obtained based on the operating mode subspace and the monotonicity submodel. The combustion performance of the power plant boiler is calculated based on the aforementioned characteristics and the combustion performance prediction model. The steps for establishing the original sample set of power plant boilers include: Query the power plant SIS system database and the power plant MIS system database; Obtain power plant boiler operation data from the power plant SIS system database; Obtain offline test data of the power plant boiler from the power plant MIS system database; An original sample set was established based on the power plant boiler operation data and the power plant boiler offline test data. The steps of obtaining stable modal samples from the original sample set include: The original sample set is decomposed into transitional mode samples and stable mode samples: Delete all of the aforementioned transition mode samples; The steps of obtaining the characteristics of the power plant boiler include: Obtain the operating condition characteristics of the power plant boiler; Obtain the target variables of the power plant boiler; Obtain the characteristic variables of the power plant boiler; The step of obtaining the operating mode subspace of the power plant boiler based on the stable mode samples and the features includes the following steps: Obtain the stable modal samples and the runtime condition features from the features; Based on the stable modal samples, the operating condition features are probabilistically clustered; The operating mode subspace of the power plant boiler is obtained; Calculate the probability that each running sample point belongs to each of the said running mode subspaces; The step of obtaining the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features includes the following steps: Obtain the target variable and feature variables from the features; For each of the said operating mode subspaces, calculate the directional relationship between the target variable and the feature variable; Extract the monotonicity patterns of the feature variables in each of the operating mode subspaces; Using the monotonicity law as a constraint, a monotonicity sub-model of the target variable is established for each of the operating mode subspaces; The step of obtaining the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel includes the following steps: Obtain the prediction results of the monotonicity sub-model for the target variable; Obtain the probability that each running sample point belongs to each of the said running mode subspaces; The probabilities are used as weights to perform a weighted summation on each prediction result; The combustion performance prediction model is obtained by summing the results.

2. A power plant boiler combustion performance prediction device, applied to the method of claim 1, characterized in that, include: The original sample set creation module is used to create the original sample set for power plant boilers. A stable mode sample acquisition module is used to acquire stable mode samples from the original sample set; The feature acquisition module is used to acquire the features of the power plant boiler; The operation mode subspace acquisition module is used to acquire the operation mode subspace of the power plant boiler based on the stable mode samples and the features; The monotonicity sub-model acquisition module is used to acquire the monotonicity sub-model of the power plant boiler based on the operating mode subspace and the features. The combustion performance prediction model acquisition module is used to acquire the combustion performance prediction model of the power plant boiler based on the operating mode subspace and the monotonicity submodel. The combustion performance calculation module is used to calculate the combustion performance of the power plant boiler based on the characteristics and the combustion performance prediction model.

3. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the power plant boiler combustion performance prediction method based on local knowledge fusion as described in claim 1.

4. A non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the power plant boiler combustion performance prediction method based on local knowledge fusion as described in claim 1.

Images