Boiler heating surface online fault diagnosis model construction method and system

Abstract

The invention discloses a boiler heating surface online fault diagnosis model construction method and system, and relates to the technical field of heat energy equipment fault diagnosis, and the method comprises the steps: constructing a boiler heating surface online fault diagnosis model through direct and indirect data paths, carrying out the boiler heating surface fault diagnosis, and constructing a feature library through principal component analysis; the model prediction result is compared with the actual fault condition of the boiler, and model parameters are optimized; and embedding the fault diagnosis model into a boiler operation management system, and automatically generating fault early warning information according to a fault diagnosis result. According to the method, through cooperative work of direct and indirect data paths, the boiler heating surface fault is accurately judged, and the potential trend of fault development is predicted; the boiler operation state key features are screened out through principal component analysis, redundant data are reduced, and the calculation efficiency of fault diagnosis is improved; through the dynamic relation matrix and the fault mode database optimization model, the sensitivity and accuracy of data analysis and fault detection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal energy equipment fault diagnosis, and specifically to a method and system for constructing an online fault diagnosis model for boiler heating surfaces. Background Technique

[0002] As a common thermal energy conversion equipment in industrial production, the operating state of the heating surface of a boiler is crucial to the overall performance and safety of the boiler. However, during long-term operation, the heating surface of the boiler is easily affected by various factors such as coking, corrosion, and temperature fluctuations, resulting in failures. If not detected and repaired in time, it may cause equipment damage, production stagnation, and even safety accidents. Therefore, the early diagnosis and online monitoring of boiler heating surface faults have become the key to ensuring the stable operation of the boiler and improving production efficiency.

[0003] Traditional methods for diagnosing boiler heating surface faults mostly rely on manual inspections and regular checks. This not only has diagnostic lag but also cannot reflect the operating state of the boiler in real time. In addition, existing diagnostic technologies mostly focus on the analysis of single data sources, ignoring the coupling relationship between multi-dimensional data and the potential of comprehensive analysis. With the development of intelligent technologies and sensor technologies, the acquisition and processing of real-time data have gradually become the main means of fault diagnosis. However, how to effectively integrate multi-source data for accurate fault diagnosis remains a technical problem to be solved urgently. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are: the existing methods for diagnosing boiler heating surface faults have insufficient multi-source data fusion capabilities, low efficiency in processing redundant data, lack of real-time warning mechanisms, and the problem of how to achieve accurate fault diagnosis.

[0006] To solve the above technical problems, the present invention provides the following technical solution: A method for constructing an online fault diagnosis model for boiler heating surfaces, including: collecting multi-dimensional data of the boiler heating surface in real time through a sensor network and performing standardized processing; constructing an online fault diagnosis model for the boiler heating surface through direct and indirect data paths and performing boiler heating surface fault diagnosis respectively, generating an optimal feature set through principal component analysis and constructing a feature library; comparing the model prediction results with the actual boiler fault conditions to optimize the model parameters; embedding the fault diagnosis model into the boiler operation management system and automatically generating fault warning information according to the fault diagnosis results.

[0007] As a preferred solution of the method for constructing an on-line fault diagnosis model of a boiler heating surface according to the present invention, wherein: the real-time acquisition and standardized processing of multi-dimensional data of the boiler heating surface includes arranging a sensor network for monitoring the boiler heating surface, the sensor network covering key parts of the boiler heating surface, and real-time acquisition of multi-dimensional data of key parts of the boiler heating surface through the sensor network; the key parts of the heating surface include the furnace inlet, the heating surface surface, and the furnace outlet; the multi-dimensional data includes boiler heating surface temperature parameters, boiler heating surface pressure parameters, boiler heating surface flow rate parameters, and boiler heating surface flue gas characteristic parameters; preprocess the acquired multi-dimensional data to eliminate noise and outliers, and normalize the multi-dimensional data; segment the multi-dimensional data based on time series and form a standardized data set.

[0008] As a preferred solution of the method for constructing an on-line fault diagnosis model of a boiler heating surface according to the present invention, wherein: the direct data path includes obtaining a standardized data set, constructing a dynamic relationship matrix for the multi-dimensional data at different time points in the standardized data set based on the physical coupling relationship between the multi-dimensional data, and for each new time point, recalculating the coupling relationship between the multi-dimensional data and updating the dynamic relationship matrix; analyzing the constructed dynamic relationship matrix and extracting the dynamic association characteristics between the multi-dimensional data; using the constructed dynamic relationship matrix and the extracted dynamic association characteristics, combining with real-time operation data, constructing an on-line fault diagnosis model of the boiler heating surface through a neural network, performing real-time analysis on the operation state of the boiler heating surface, comparing the characteristic differences between normal conditions and abnormal conditions, identifying the potential trend of the development of heating surface faults, and predicting the heating surface fault mode; the real-time analysis includes analyzing the abnormal characteristics under real-time conditions and calculating the deviation degree between the real-time conditions of the boiler heating surface and the historical normal conditions; the abnormal characteristics include abnormal fluctuations of operation parameters, abnormal changes in data association relationships, periodicity and trend anomalies of data, abnormal increases in noise and interference signals, and abnormal pattern deviations; the deviation degree includes a deviation degree calculation formula, expressed as:

[0009]

[0010] Wherein, represents the deviation degree between the boiler operation state and the normal condition at time, represents the number of boiler operation parameters, represents the parameter in the state deviation calculation, represents at time the th actual value of the boiler operation parameter, represents the parameter Reference values under normal operating conditions; the reference values are historical average values or set standard values; the predicted failure mode of the heating surface includes setting a failure prediction threshold, predicting the state changes at multiple future time points through LSTM, calculating the abnormal feature change rate, and judging whether the failure shows an intensifying trend; the abnormal feature change rate includes the abnormal feature change rate calculation formula, which is expressed as:

[0011]

[0012] Among them, represents the change amount of the boiler state deviation degree between time and ; represents the deviation degree of the boiler operating state from the normal operating condition at time ; if is greater than the failure prediction threshold, it indicates that the boiler operating state is deteriorating rapidly and there is a potential failure trend.

[0013] As a preferred solution of the method for constructing an online fault diagnosis model of the boiler heating surface according to the present invention, wherein: the indirect data path includes collecting the historical operation data of the boiler from the boiler operation management system, screening the data of the fault instances in the historical operation data, marking the fault types of the fault instances, the operating condition parameters at the time of the fault occurrence, and the changes in the operating parameters before and after the fault occurrence, and establishing a fault mode database based on the fault instances; the historical operation data includes historical heating surface temperature data, historical pressure data, historical flow rate data, and historical flue gas characteristic data; the fault instances include temperature anomalies, pressure fluctuations, and flow rate anomalies; each fault instance data in the fault mode database is labeled with the matching normal operating state data to form a training data set, and an online fault diagnosis model of the boiler heating surface is trained based on the training data set; an optimized loss function is constructed, and a dynamic adjustment term is introduced into the loss function; the dynamic adjustment term includes setting a weighting coefficient based on the change range of the real-time data and the steady-state interval of the historical data, and adjusting the weights in the training process of the online fault diagnosis model of the boiler heating surface; setting the weighting coefficient includes extracting key operating parameters from the historical data, calculating the historical mean and standard deviation of the parameters, setting the steady-state intervals of the respective key operating parameters, and assigning initial weights to different key operating parameters; collecting the boiler operation data at the current time point, calculating the deviation deviation of the current boiler operation data from the historical mean of the parameters, setting a deviation threshold, and judging whether the data deviates from the steady-state interval based on the deviation threshold; if the deviation deviation is greater than the deviation threshold, it is judged that there is an abnormal trend in the boiler operating state, and the weight is increased based on the initial weight of the current boiler operation data.

[0014] As a preferred embodiment of the method for constructing an online fault diagnosis model of a boiler heating surface according to the present invention, wherein: the construction of the feature library includes extracting key operating parameters of the operating state of the boiler heating surface from multi-dimensional data; the key operating parameters include heating surface temperature field parameters, pressure difference parameters, flow velocity field parameters, and flue gas characteristic parameters; principal component analysis is used to reduce the dimension of the key operating parameters, and by analyzing the variance contribution rate of the principal components, the principal components with a cumulative variance contribution rate greater than 90% are selected as the optimal feature set, and the optimal feature set is organized into a feature library; the feature library includes a detailed description of each optimal feature set; the detailed description includes the physical meaning, calculation method, and the role in boiler fault diagnosis of the features in the optimal feature set.

[0015] As a preferred embodiment of the method for constructing an online fault diagnosis model of a boiler heating surface according to the present invention, wherein: the optimization of the model parameters includes comparing the predicted results of the fault modes obtained through the direct data path and the indirect data path, and verifying whether the diagnostic results of the direct data path and the indirect data path are consistent with the actual boiler fault situation; the diagnostic results include the fault type, the time point of the fault occurrence, and the scope of the fault impact; if the diagnostic results are consistent, it indicates that the prediction of the online fault diagnosis model of the boiler heating surface is accurate; if the diagnostic results are inconsistent, the parameters of the online fault diagnosis model of the boiler heating surface are optimized; optimizing the parameters of the online fault diagnosis model of the boiler heating surface includes optimizing the dynamic relationship matrix in the direct data path and the fault mode database in the indirect data path; optimizing the dynamic relationship matrix in the direct data path includes optimizing by adding historical data sets under different working conditions, increasing the sampling frequency of sensor data, and adjusting the positioning of sensors; optimizing the fault mode database in the indirect data path includes optimizing by adding fault instances, fine-tuning the loss function, and introducing a dynamic adjustment term.

[0016] As a preferred embodiment of the method for constructing an online fault diagnosis model of a boiler heating surface according to the present invention, wherein: automatically generating a fault warning message according to the fault diagnosis result includes docking the online fault diagnosis model of the boiler heating surface with the boiler operation management system, sending the multi-dimensional data collected in real time into the online fault diagnosis model of the boiler heating surface for processing, judging whether there is a fault in the boiler heating surface, and real-time feedback of the fault diagnosis result to the boiler operation management system; setting and configuring a visualization interface; setting a real-time warning mechanism according to the fault diagnosis result; automatically generating a fault warning message according to the fault diagnosis result; the fault warning message includes the fault type, the specific location of the fault occurrence, the potential fault trend, and the fault handling measures.

[0017] Another object of the present invention is to provide a system for constructing an on-line fault diagnosis model of a boiler heating surface, which can evaluate the coupling relationship between various boiler parameters through a direct-path dynamic relationship matrix, and solves the problem of insufficient multi-source data fusion ability in the current technology.

[0018] As a preferred embodiment of the system for constructing an on-line fault diagnosis model of a boiler heating surface according to the present invention, it includes a data processing module, a fault diagnosis module, a parameter optimization module, and an automatic warning module; the data processing module is used to collect multi-dimensional data of the boiler heating surface in real time through a sensor network and perform standardized processing; the fault diagnosis module is used to construct an on-line fault diagnosis model of the boiler heating surface through direct and indirect data paths and perform fault diagnosis of the boiler heating surface respectively, generate an optimal feature set through principal component analysis and construct a feature library; the parameter optimization module is used to compare the model prediction result with the actual fault situation of the boiler and optimize the model parameters; the automatic warning module is used to embed the fault diagnosis model into the boiler operation management system and automatically generate fault warning information according to the fault diagnosis result.

[0019] A computer device includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the method for constructing an on-line fault diagnosis model of a boiler heating surface.

[0020] A computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the steps of the method for constructing an on-line fault diagnosis model of a boiler heating surface are implemented.

[0021] Advantages of the present invention: The method for constructing an online fault diagnosis model for the boiler heating surface provided by the present invention can not only accurately determine the fault type and occurrence time of the boiler heating surface through the collaborative work of the direct data path and the indirect data path, but also predict the potential trend of fault development; the direct data path evaluates the coupling relationship between boiler parameters through a dynamic relationship matrix, and the indirect data path predicts potential fault modes through the analysis of historical data and fault instances. The collaborative verification of the two ensures the high accuracy and credibility of the diagnosis results; through principal component analysis, dimensionality reduction processing is performed on multi-dimensional data, and key features most relevant to the boiler operating state are screened out, effectively reducing redundant data and improving the calculation efficiency of the fault diagnosis model; through the optimization of the online fault diagnosis model for the boiler heating surface using a dynamic relationship matrix and a fault mode database, the model can not only adapt to changes under different working conditions, but also improve the sensitivity and accuracy of data analysis and fault detection; through the integration of an intelligent boiler heating surface fault monitoring and warning system, the boiler operating state is tracked in real time, and early warnings are issued before faults occur. By displaying operating parameters and diagnosis results through a visual interface, not only can the boiler state be understood in real time, but effective measures can be taken before faults occur, improving the safety of boiler operation and reducing the potential impact of faults on the boiler system. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 FIG. is the overall flowchart of a method for constructing an online fault diagnosis model for a boiler heating surface provided by the first embodiment of the present invention.

[0024] Figure 2 FIG. is the overall flowchart of a system for constructing an online fault diagnosis model for a boiler heating surface provided by the third embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] To make the above objects, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0026] Example 1, referring to Figure 1, which is an embodiment of the present invention, provides a method for constructing an online fault diagnosis model for boiler heating surfaces, including:

[0027] S1: Real-time collect multi-dimensional data of the boiler heating surface through a sensor network and perform standardization processing.

[0028] Furthermore, real-time collecting multi-dimensional data of the boiler heating surface and performing standardization processing includes deploying a sensor network for monitoring the boiler heating surface. The sensor network covers key parts of the boiler heating surface, and multi-dimensional data of key parts of the boiler heating surface are collected in real time through the sensor network; the key parts of the heating surface include the furnace inlet, the surface of the heating surface, and the furnace outlet; the multi-dimensional data includes boiler heating surface temperature parameters, boiler heating surface pressure parameters, boiler heating surface flow rate parameters, and boiler heating surface flue gas characteristic parameters; preprocess the collected multi-dimensional data to eliminate noise and outliers, and normalize the multi-dimensional data; segment the multi-dimensional data based on time series and form a standardized data set.

[0029] It should be noted that by accurately deploying the sensor network and collecting multi-dimensional data in real time, it is ensured that multi-dimensional data of each key part of the boiler heating surface can be accurately collected, providing rich and comprehensive data support for the fault diagnosis model; by preprocessing the data to eliminate noise and outliers, the accuracy and reliability of the input data in the subsequent diagnosis process are guaranteed. Normalizing the multi-dimensional data ensures that the data input into the model has high quality and small fluctuations, thus avoiding diagnostic errors caused by poor data quality and ensuring the efficiency and accuracy of fault diagnosis.

[0030] S2: Construct an online fault diagnosis model for the boiler heating surface through direct data paths and indirect data paths, and perform boiler heating surface fault diagnosis respectively. Generate an optimal feature set through principal component analysis and construct a feature library.

[0031] Furthermore, the direct data path includes obtaining a standardized data set. For the multi-dimensional data at different time points in the standardized data set, based on the physical coupling relationship between the multi-dimensional data, a dynamic relationship matrix is constructed. For each new time point, the coupling relationship between the multi-dimensional data is recalculated, and the dynamic relationship matrix is updated; the constructed dynamic relationship matrix is analyzed, and the dynamic association features between the multi-dimensional data are extracted; using the constructed dynamic relationship matrix and the extracted dynamic association features, combined with real-time operation data, an on-line fault diagnosis model of the boiler heating surface is constructed through a neural network to perform real-time analysis on the operation state of the boiler heating surface, compare the characteristic differences between normal conditions and abnormal conditions, identify the potential trend of the development of the heating surface fault, and predict the heating surface fault mode; the real-time analysis includes analyzing the abnormal features under real-time conditions and calculating the deviation degree between the real-time conditions of the boiler heating surface and the historical normal conditions; the abnormal features include abnormal fluctuations in operation parameters, abnormal changes in data association relationships, periodicity and trend anomalies in data, abnormal increases in noise and interference signals, and abnormal pattern deviations; the deviation degree includes a deviation degree calculation formula, which is expressed as:

[0032]

[0033] Among them, represents the deviation degree between the boiler operation state and the normal condition at time, represents the number of boiler operation parameters, represents the parameter in the state deviation calculation weight, represents at time the th actual value of the boiler operation parameter, represents the parameter reference value under normal conditions; the reference value is the historical average value or the set standard value; predicting the heating surface fault mode includes setting a fault prediction threshold, predicting the state changes at multiple future time points through LSTM, calculating the abnormal feature change rate, and judging whether the fault shows an aggravating trend; the abnormal feature change rate includes an abnormal feature change rate calculation formula, which is expressed as:

[0034]

[0035] Among them, represents the change amount of the boiler state deviation degree between time and , represents the deviation degree between the boiler operation state and the normal condition at time ; if is greater than the fault prediction threshold, it indicates that the boiler operation state is deteriorating rapidly and there is a potential fault trend.

[0036] It should be noted that the indirect data path includes collecting the historical operation data of the boiler from the boiler operation management system, screening the data for fault instances in the historical operation data, labeling the fault types of the fault instances, the operating condition parameters at the time of fault occurrence, and the changes in the operating parameters before and after the fault occurrence, and establishing a fault mode database based on the fault instances; the historical operation data includes historical heating surface temperature data, historical pressure data, historical flow rate data, and historical flue gas characteristic data; the fault instances include temperature anomalies, pressure fluctuations, and flow rate anomalies; each fault instance data in the fault mode database is labeled with the matching normal operating state data to form a training data set, and the online fault diagnosis model of the boiler heating surface is trained based on the training data set; an optimized loss function is constructed, and a dynamic adjustment term is introduced into the loss function; the dynamic adjustment term includes setting a weighting coefficient based on the change amplitude of the real-time data and the steady-state interval of the historical data, and adjusting the weights during the training process of the online fault diagnosis model of the boiler heating surface; setting the weighting coefficient includes extracting key operating parameters from the historical data, calculating the historical mean and standard deviation of the parameters, setting the steady-state interval of each key operating parameter, and assigning initial weights to different key operating parameters; collecting the boiler operation data at the current time point, calculating the deviation of each current boiler operation data from the historical mean of the parameters, setting a deviation threshold, and judging whether the data deviates from the steady-state interval based on the deviation threshold; if the deviation is greater than the deviation threshold, it is judged that there is an abnormal trend in the boiler operation state, and the weight is increased based on the initial weight of the current boiler operation data.

[0037] It should also be noted that constructing the feature library includes extracting the key operating parameters of the operating state of the boiler heating surface from multi-dimensional data; the key operating parameters include heating surface temperature field parameters, pressure difference parameters, flow rate field parameters, and flue gas characteristic parameters; principal component analysis is used to perform dimensionality reduction on the key operating parameters, and by analyzing the variance contribution rate of the principal components, the principal components with a cumulative variance contribution rate greater than 90% are selected as the optimal feature set, and the optimal feature set is organized into a feature library; the feature library includes a detailed description of each optimal feature set; the detailed description includes the physical meaning, calculation method, and the role in boiler fault diagnosis of the features in the optimal feature set.

[0038] It should also be noted that by constructing a dynamic relationship matrix, the dynamic coupling relationship between multi-dimensional data can be deeply explored; through the relationship matrix based on physical coupling, not only the internal connection between the operating parameters of the boiler heating surface can be revealed, but also the accurate prediction of the fault trend can be further realized through the neural network model; by comparing the characteristic differences between normal conditions and abnormal conditions, potential fault hazards can be detected in a timely manner to avoid accidents; by analyzing the dynamic relationship between the parameters of the boiler heating surface, the model can be helped to deeply understand the change law of the operating state of the boiler heating surface; this not only provides a solid data basis for real-time diagnosis, but also realizes the accurate prediction of the boiler state through feature extraction, enhancing the real-time performance of fault detection.

[0039] It should also be noted that by collecting and analyzing the historical operation data and fault cases of the boiler, a fault mode database is constructed, which enhances the adaptability and prediction ability of the model to various fault modes, ensures that the fault diagnosis model can adapt to a variety of complex working conditions, and enhances the stability and robustness of the model in practical applications; through the labeling process of fault cases, the model can quickly match the corresponding fault modes in real-time data, so as to make an accurate fault diagnosis.

[0040] It should also be noted that by performing dimensionality reduction processing on multi-dimensional data through principal component analysis, the most representative characteristic information is retained, and the key characteristic parameters closely related to boiler fault diagnosis are effectively screened out; by selecting the principal components with higher variance contribution rate, the redundancy of the data is reduced, while the meaningful information in the data is retained, which not only improves the data processing efficiency, but also reduces the calculation burden, enhances the real-time response ability of the diagnosis model, and ensures the high efficiency of the diagnosis model in practical applications.

[0041] S3: Compare the model prediction results with the actual boiler fault conditions and optimize the model parameters.

[0042] Further, optimizing the model parameters includes comparing the fault mode prediction results obtained through the direct data path and the indirect data path, and verifying whether the diagnostic results of the direct data path and the indirect data path are consistent with the actual boiler fault conditions; the diagnostic results include the fault type, the time point when the fault occurs, and the scope of the fault impact; if the diagnostic results are consistent, it indicates that the prediction of the on-line boiler heating surface fault diagnosis model is accurate; if the diagnostic results are inconsistent, the parameters of the on-line boiler heating surface fault diagnosis model are optimized; optimizing the parameters of the on-line boiler heating surface fault diagnosis model includes optimizing the dynamic relationship matrix in the direct data path and optimizing the fault mode database in the indirect data path; optimizing the dynamic relationship matrix in the direct data path includes optimizing by adding historical data sets under different working conditions, increasing the sampling frequency of sensor data, and adjusting the positioning of sensors; optimizing the fault mode database in the indirect data path includes optimizing by adding fault instances, fine-tuning the loss function, and introducing a dynamic adjustment term.

[0043] It should be noted that by comparing the diagnostic results of the direct data path and the indirect data path, it is possible to verify whether the prediction result of the model is consistent with the actual situation; in practical applications, by dynamically optimizing the model parameters of the two paths, the diagnostic accuracy is further improved; by adding historical data sets, adjusting sensor parameters, etc., the stability of the data path can be further improved, so as to ensure that the real-time fault diagnosis during the boiler operation is always accurate and effective.

[0044] S4: Embed the fault diagnosis model into the boiler operation management system, and automatically generate a fault warning message according to the fault diagnosis result.

[0045] Further, automatically generating a fault warning message according to the fault diagnosis result includes docking the on-line boiler heating surface fault diagnosis model with the boiler operation management system, sending the multi-dimensional data collected in real time into the on-line boiler heating surface fault diagnosis model for processing, judging whether there is a fault in the boiler heating surface, and real-time feedback the fault diagnosis result to the boiler operation management system; setting and configuring a visualization interface; setting a real-time warning mechanism according to the fault diagnosis result; automatically generating a fault warning message according to the fault diagnosis result; the fault warning message includes the fault type, the specific location where the fault occurs, the potential fault trend, and the fault handling measures.

[0046] It should be noted that by deeply integrating the on-line boiler heating surface fault diagnosis model with the boiler operation management system, real-time monitoring and intelligent warning are realized; by dynamically visualizing and displaying the boiler operation parameters and the fault diagnosis result, the operation state of the boiler can be grasped in the first time; when a potential fault occurs, the system can automatically trigger a warning and generate a fault message to remind the personnel to handle it in time, improving the safety of the boiler operation and reducing the accidents caused by faults.

[0047] Embodiment 2, the second embodiment of the present invention, which is different from the previous two embodiments in that:

[0048] If the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that makes a contribution to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, etc., all of which can store program codes.

[0049] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a predefined sequence of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in conjunction with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0050] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other appropriate processing as necessary, and then stored in a computer memory.

[0051] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0052] Embodiment 3, referring to Figure 2 , which is the third embodiment of the present invention. This embodiment provides a system for a load balancing method of a computing platform based on a particle swarm genetic algorithm, including a data processing module, a fault diagnosis module, a parameter optimization module, and an automatic warning module.

[0053] Among them, the data processing module is used to collect multi-dimensional data of the boiler heating surface in real time through a sensor network and perform standardized processing; the fault diagnosis module is used to construct an online fault diagnosis model for the boiler heating surface through direct and indirect data paths and perform boiler heating surface fault diagnosis respectively, generate an optimal feature set through principal component analysis and construct a feature library; the parameter optimization module is used to compare the model prediction results with the actual boiler fault conditions and optimize the model parameters; the automatic warning module is used to embed the fault diagnosis model into the boiler operation management system and automatically generate fault warning information according to the fault diagnosis results.

[0054] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for constructing an on-line fault diagnosis model of a boiler heating surface, characterized in that Including: Collecting multi-dimensional data of the boiler heating surface in real time through a sensor network and performing standardized processing; Constructing an online fault diagnosis model for the boiler heating surface through direct and indirect data paths, respectively diagnosing faults in the boiler heating surface, generating an optimal feature set through principal component analysis, and constructing a feature library; Comparing the model prediction results with the actual fault conditions of the boiler to optimize the model parameters; Embedding the fault diagnosis model into the boiler operation management system to automatically generate fault warning information according to the fault diagnosis results.

2. The method for constructing an on-line fault diagnosis model of a boiler heating surface according to claim 1, characterized in that: The real-time collection of multi-dimensional data of the boiler heating surface and the standardized processing include deploying a sensor network for monitoring the boiler heating surface, the sensor network covering key parts of the boiler heating surface, and collecting multi-dimensional data of key parts of the boiler heating surface in real time through the sensor network; The key parts of the heating surface include the furnace inlet, the surface of the heating surface, and the furnace outlet; The multi-dimensional data includes the temperature parameters of the boiler heating surface, the pressure parameters of the boiler heating surface, the flow velocity parameters of the boiler heating surface, and the flue gas characteristic parameters of the boiler heating surface; Preprocessing the collected multi-dimensional data, eliminating noise and outliers, and normalizing the multi-dimensional data; Segmenting the multi-dimensional data based on time series and forming a standardized data set.

3. The method for constructing an on-line fault diagnosis model of a boiler heating surface according to claim 2, characterized in that: The direct data path includes obtaining the standardized data set, constructing a dynamic relationship matrix based on the physical coupling relationship between multi-dimensional data for different time points in the standardized data set, and for each new time point, recalculating the coupling relationship between multi-dimensional data and updating the dynamic relationship matrix; Analyzing the constructed dynamic relationship matrix and extracting the dynamic correlation features between multi-dimensional data; Using the constructed dynamic relationship matrix and the extracted dynamic correlation features, combined with real-time operation data, constructing an online fault diagnosis model for the boiler heating surface through a neural network, analyzing the operating state of the boiler heating surface in real time, comparing the characteristic differences between normal and abnormal conditions, identifying the potential trend of the development of heating surface faults, and predicting the heating surface fault mode; Real-time analysis includes analyzing the abnormal features under real-time conditions and calculating the deviation degree between the real-time conditions of the boiler heating surface and the historical normal conditions; The abnormal features include abnormal fluctuations in operating parameters, abnormal changes in data correlation relationships, periodicity and trend anomalies in data, abnormal increases in noise and interference signals, and abnormal pattern deviations; The deviation degree includes a deviation degree calculation formula, expressed as: Among them, D(t) represents the deviation degree of the boiler operation state from the normal condition at time t, n represents the number of boiler operation parameters, w i represents the parameter X i in the weight calculation of the state deviation, X i (t) represents the actual value of the i-th boiler operation parameter at time t, represents the parameter X i reference value under normal conditions; The reference value is the historical average value or a set standard value; Predicting the heating surface fault mode includes setting a fault prediction threshold, predicting the state changes at multiple future time points through LSTM, calculating the abnormal feature change rate, and judging whether the fault shows an aggravating trend; The abnormal feature change rate includes an abnormal feature change rate calculation formula, expressed as: ΔD(t) = D(t) - D(t - 1) Where, ΔD(t) represents the change amount of the boiler state deviation degree between time t and t - 1, and ΔD(t - 1) represents the deviation degree of the boiler operating state from the normal condition at time t - 1; If ΔD(t) is greater than the fault prediction threshold, it indicates that the boiler operating state is deteriorating rapidly and there is a potential fault trend.

4. The method for constructing an on-line fault diagnosis model of a boiler heating surface according to claim 3, characterized in that: The indirect data path includes collecting the historical operation data of the boiler from the boiler operation management system, screening the data for fault instances in the historical operation data, annotating the fault types of the fault instances, the operating condition parameters at the time of fault occurrence, and the changes in the operation parameters before and after the fault occurrence, and establishing a fault mode database based on the fault instances; The historical operation data includes historical heating surface temperature data, historical pressure data, historical flow rate data, and historical flue gas characteristic data; The fault instances include temperature anomalies, pressure fluctuations, and flow rate anomalies; Tag each fault instance data in the fault mode database with the matching normal operation state data to form a training data set, and train the online fault diagnosis model for the boiler heating surface based on the training data set; Construct an optimized loss function and introduce a dynamic adjustment term into the loss function; The dynamic adjustment term includes setting a weighting coefficient based on the change amplitude of the real-time data and the steady-state interval of the historical data, and adjusting the weights during the training process of the online fault diagnosis model for the boiler heating surface; Setting the weighting coefficient includes extracting key operation parameters from the historical data, calculating the historical mean and standard deviation of the parameters, setting the steady-state intervals of each key operation parameter, and assigning initial weights to different key operation parameters; Collect the boiler operation data at the current time point, calculate the deviation of the current boiler operation data from the historical mean of the parameters, set a deviation threshold, and judge whether the data deviates from the steady-state interval based on the deviation threshold; If the deviation is greater than the deviation threshold, it is judged that there is an abnormal trend in the boiler operation state, and the weight is increased based on the initial weight of the current boiler operation data.

5. The method for constructing an on-line fault diagnosis model of a boiler heating surface according to claim 4, characterized in that: The construction of the feature library includes extracting key operation parameters of the operation state of the boiler heating surface from multi-dimensional data; The key operation parameters include heating surface temperature field parameters, pressure difference parameters, flow rate field parameters, and flue gas characteristic parameters; Use principal component analysis to reduce the dimension of the key operation parameters. By analyzing the variance contribution rate of the principal components, select the principal components with a cumulative variance contribution rate greater than 90% as the optimal feature set, and organize the optimal feature set into a feature library; The feature library includes a detailed description of each optimal feature set; The detailed description includes the physical meaning, calculation method, and the role in boiler fault diagnosis of the features in the optimal feature set.

6. The method for constructing an on-line fault diagnosis model of a boiler heating surface according to claim 5, characterized in that: The optimization of the model parameters includes comparing the fault mode prediction results obtained through the direct data path and the indirect data path, and checking whether the diagnostic results of the direct data path and the indirect data path are consistent with the actual boiler fault situation; The diagnostic results include the fault type, the time point of fault occurrence, and the fault impact range; If the diagnostic results are consistent, it indicates that the prediction of the online fault diagnosis model for the boiler heating surface is accurate; If the diagnostic results are inconsistent, optimize the parameters of the online fault diagnosis model for the boiler heating surface; Optimizing the parameters of the online fault diagnosis model for the boiler heating surface includes optimizing the dynamic relationship matrix in the direct data path and the fault mode database in the indirect data path; Optimizing the dynamic relationship matrix in the direct data path includes optimizing by increasing the historical data sets under different working conditions, increasing the sampling frequency of sensor data, and adjusting the positioning of sensors; Optimizing the fault mode database in the indirect data path includes optimization by increasing fault instances, fine-tuning the loss function, and introducing dynamic adjustment terms.

7. The method for constructing an on-line fault diagnosis model of a boiler heating surface according to claim 6, characterized in that: Automatically generating fault warning information according to the fault diagnosis result includes docking the on-line fault diagnosis model of the boiler heating surface with the boiler operation management system, sending the multi-dimensional data collected in real time into the on-line fault diagnosis model of the boiler heating surface for processing, judging whether there is a fault in the boiler heating surface, and feeding back the fault diagnosis result to the boiler operation management system in real time; Set and configure the visualization interface; Set the real-time warning mechanism according to the fault diagnosis result; Automatically generate fault warning information according to the fault diagnosis result; The fault warning information includes the fault type, the specific location where the fault occurs, the potential fault trend, and the fault handling measures.

8. A system adopting the method for constructing an on-line fault diagnosis model of a boiler heating surface according to any one of claims 1 to 7, characterized in that: It includes a data processing module, a fault diagnosis module, a parameter optimization module, and an automatic warning module; The data processing module is used to collect multi-dimensional data of the boiler heating surface in real time through the sensor network and perform standardized processing; The fault diagnosis module is used to construct an on-line fault diagnosis model of the boiler heating surface through the direct data path and the indirect data path and perform boiler heating surface fault diagnosis respectively, generate an optimal feature set through principal component analysis and construct a feature library; The parameter optimization module is used to compare the model prediction result with the actual fault situation of the boiler and optimize the model parameters; The automatic warning module is used to embed the fault diagnosis model into the boiler operation management system and automatically generate fault warning information according to the fault diagnosis result.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method for constructing an on-line fault diagnosis model of the boiler heating surface according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for constructing an on-line fault diagnosis model of the boiler heating surface according to any one of claims 1 to 7.

Images