Method and system for predicting real-time temperature of lubricating oil of steam turbine

Abstract

The invention relates to the technical field of equipment monitoring, and discloses a steam turbine lubricating oil real-time temperature prediction method and system.According to the steam turbine lubricating oil real-time temperature prediction method, lubricating oil measurement points are deployed by combining lubricating oil historical temperature data in target steam turbine associated data, and reasonable deployment of the lubricating oil measurement points is achieved; therefore, the temperature information of the lubricating oil can be obtained in real time, the temperature prediction model can more accurately capture temperature changes, abnormal conditions can be found in time, reliable data support is provided for follow-up analysis and decision making, and the defect that when the real-time operation data of the lubricating oil of the gas turbine is obtained through an existing method, the real-time operation data cannot be obtained is overcome. The problems that temperature monitoring points cannot be reasonably deployed based on the key position of a lubricating oil pipeline, the real-time operation data size of lubricating oil is too large, a model cannot accurately capture temperature changes, and the evaluation precision of the model is affected are solved.

Description

Technical Field

[0001] The invention belongs to the technical field of equipment monitoring, and in particular relates to a method and system for predicting the real-time temperature of steam turbine lubricating oil. Background Art

[0002] As the core equipment of the thermal power generation system, the steam turbine of a thermal power plant is a rotary steam power device that realizes energy conversion by driving the blades to rotate with high-temperature and high-pressure steam. Its efficient energy conversion capability and advanced manufacturing technology make it occupy an important position in the modern energy industry. The temperature monitoring of the lubricating oil of the steam turbine in a thermal power plant is of great significance for ensuring the normal operation of the equipment and extending the life of the equipment. By real-time monitoring of the temperature of the lubricating oil, problems such as insufficient lubrication or excessive oil temperature can be discovered in time, so that corresponding solutions can be taken to prevent equipment failures caused by poor lubrication.

[0003] Chinese patent CN117875173A discloses a method, device and equipment for predicting the lubricating oil drain temperature of a gas turbine, the method comprising: obtaining historical operation monitoring data of the lubricating oil of the gas turbine, and taking the historical operation monitoring data that meets preset requirements as a sample data set; establishing a support vector regression model, analyzing the historical operation monitoring data to determine input parameters, and taking the lubricating oil drain temperature as an output parameter; training the support vector regression model using the sample data set to obtain the trained support vector regression model as a lubricating oil drain temperature prediction model; obtaining the real-time operation data of the lubricating oil of the gas turbine and inputting it into the lubricating oil drain temperature prediction model to obtain the predicted lubricating oil drain temperature; however, when the existing method obtains the real-time operation data of the lubricating oil of the gas turbine, the temperature monitoring points cannot be reasonably deployed based on the key positions of the lubricating oil pipeline, resulting in an excessive amount of real-time operation data of the lubricating oil, making it impossible for the model to accurately capture the temperature changes, affecting the model evaluation accuracy. Summary of the invention

[0004] The purpose of the present invention is to overcome the above-mentioned problems and provide a method and system for predicting the real-time temperature of a steam turbine lubricating oil, which can solve the problem that when the existing method obtains the real-time operation data of the gas turbine lubricating oil, the temperature monitoring points cannot be reasonably deployed based on the key positions of the lubricating oil pipeline, resulting in an excessive amount of real-time operation data of the lubricating oil, making it impossible for the model to accurately capture the temperature changes, thus affecting the accuracy of the model evaluation.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for predicting the real-time temperature of a steam turbine lubricating oil, comprising the following steps: Acquire target steam turbine associated data, pre-process the target steam turbine associated data, and obtain an associated pre-processed set; Load the associated preprocessing set, and deploy lubricating oil measurement points based on the lubricating oil historical temperature data in the target steam turbine associated data; The real-time temperature of the steam turbine lubricating oil is collected in real time according to the lubricating oil measurement point, a Gaussian filter function is preset, and the real-time temperature of the steam turbine lubricating oil is converted into a visual temperature time-frequency diagram based on the Gaussian filter function; Obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs the temperature prediction result.

[0006] A further improvement of the present invention is that the step of preprocessing the target steam turbine associated data to obtain an associated preprocessing set specifically includes: Loading target steam turbine associated data, processing missing values ​​and duplicate values ​​of the target steam turbine associated data, and obtaining a first preprocessing set; Acquire a first preprocessing set, and mark the first preprocessing set based on the internal lubricating oil pipeline coordinate information of the target steam turbine and the lubricating oil pipeline grade parameter to obtain a second preprocessing set containing the lubricating oil pipeline mark; Traversing the second preprocessing set, three-dimensionally modeling the lubricating oil pipeline inside the target steam turbine based on a three-dimensional visualization tool to obtain a three-dimensional model of the pipeline; Nodes and edges of the pipeline three-dimensional model are extracted, and the nodes and edges are connected in sequence based on the flow direction of the lubricating oil to generate a pipeline topological connection diagram. The pipeline topological connection diagram and the second preprocessing set are integrated to obtain an associated preprocessing set.

[0007] A further improvement of the present invention is that the step of deploying lubricating oil measurement points in combination with lubricating oil historical temperature data in target steam turbine associated data specifically includes: Obtain lubricating oil historical temperature data and associated preprocessing sets, and calculate the internal lubricating oil pipeline weight score based on the random forest decision tree model:

[0008] in, is the current pipeline weight score, is the number of lubricating oil pipe edges inside the target steam turbine, is the number of nodes of the lubricating oil pipeline inside the target steam turbine, is the number of nodes associated with the internal lubricating oil pipeline, is the number of decision trees in the random forest decision tree model, is the temperature transition value of the pipeline node within a unit acquisition cycle;

[0009] in, is the temperature transition value of the pipeline node within the unit acquisition cycle, is the number of acquisition cycles, is the highest temperature during the acquisition period, is the average temperature during the acquisition period, is the expansion factor of pipeline lubricating oil; Based on the current pipeline weight score combined with the convolutional neural network, the internal lubricating oil pipeline is weighted and fused to calculate the pipeline independence corresponding to the internal lubricating oil pipeline;

[0010] in, The pipeline independence corresponding to the internal lubricating oil pipeline; Using pipeline independence and pipeline weight scores as measurement point deployment evaluation indicators, the importance of internal lubricating oil pipelines is ranked, and internal lubricating oil pipelines with a value greater than a preset importance threshold are output as lubricating oil measurement points;

[0011] in, is the pipeline importance, is the importance weighting parameter;

[0012] in, is the current pipeline weight score, is the number of nodes of the lubricating oil pipeline inside the target steam turbine, is the number of nodes associated with the internal lubricating oil pipeline.

[0013] A further improvement of the present invention is that the step of converting the real-time temperature of the steam turbine lubricating oil into a visualized temperature time-frequency diagram based on the Gaussian filter function specifically includes: A Gaussian filter function is preset, and standard wavelet basis conversion is performed on the real-time temperature of the lubricating oil in combination with the Gaussian filter function; The standard wavelet basis transformation is expressed as:

[0014] in, is the Gaussian filter function, is the lubricating oil real-time temperature sequence output vector, are the wavelet transform coefficients, They represent the high-frequency oscillation component and the low-frequency oscillation component of the real-time temperature series of the lubricating oil respectively;

[0015] in, is the scale parameter of the Gaussian filter function, is the lubricating oil real-time temperature sequence output vector, is the average temperature of the lubricating oil real-time temperature series, is the normalized coefficient of the Gaussian filter function; The inner product of the standard wavelet basis transformation and the real-time temperature series of the lubricating oil is performed to obtain the temperature frequency coefficient; Load the temperature frequency coefficient and draw a visual temperature time-frequency graph based on the temperature frequency coefficient and temperature acquisition time.

[0016] A further improvement of the present invention is that the step of constructing a training temperature prediction model based on a training set and a test set specifically includes: Obtain a training set, preprocess the training set, and detect temperature mutation points in the training set; Construct a temperature prediction model. The temperature prediction model uses the grey prediction model as the initial model. The preprocessed training set is input into the initial model for unsupervised prediction. The number of iterative training times is set to obtain the temperature prediction result. Get the temperature prediction result, use the cross entropy loss function to calculate the difference between the temperature prediction result and the actual temperature result, and use the stochastic gradient descent algorithm to iteratively initialize the hyperparameters of the initial model along the negative gradient direction of the cross entropy loss function until the model converges; A validation set is obtained, and the accuracy of the model is verified using the validation set to determine whether the accuracy of the temperature prediction result and the actual temperature result exceeds the preset accuracy threshold. If the accuracy exceeds the preset accuracy threshold, the trained temperature prediction model is output.

[0017] A further improvement of the present invention is that the step of the temperature prediction model identifying the visualized temperature time-frequency diagram and outputting the temperature prediction result specifically includes: Obtain a visualized temperature time-frequency graph, and use the temperature prediction model to extract the mutation factor of the visualized temperature time-frequency graph; Loading the mutation factor and the visualized temperature time-frequency diagram, the temperature prediction model performs parameter fitting on the mutation factor and the visualized temperature time-frequency diagram to obtain a parameter fitting set of at least one order; Obtain a parameter fitting set, use the parameter fitting set as input, execute a temperature prediction model, and the temperature prediction model outputs a temperature prediction result corresponding to a visualized temperature time-frequency graph; The output of the parameter fit set is expressed as:

[0018] in, Output signal for parameter fitting set, is the input vector for visualizing the temperature time-frequency diagram, is the mutation factor corresponding to the visualized temperature time-frequency diagram.

[0019] In a second aspect, the present invention further provides a real-time temperature prediction system for lubricating oil of a steam turbine, comprising the following modules: A preprocessing module is used to obtain target steam turbine associated data, preprocess the target steam turbine associated data, and obtain an associated preprocessing set; The measurement point deployment module is used to load the associated preprocessing set and deploy the lubricating oil measurement points in combination with the target steam turbine associated data; The time-frequency graph conversion module is used to collect the real-time temperature of the steam turbine lubricating oil in real time according to the lubricating oil measurement point, preset a Gaussian filter function, and convert the real-time temperature of the steam turbine lubricating oil into a visual temperature time-frequency graph based on the Gaussian filter function; The temperature prediction module is used to obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a training temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs the temperature prediction result.

[0020] A further improvement of the present invention is that the preprocessing module specifically comprises: A first preprocessing unit is used to load target steam turbine associated data, perform missing value and duplicate value processing on the target steam turbine associated data, and obtain a first preprocessing set; A second preprocessing unit is used to obtain the first preprocessing set, and mark the first preprocessing set based on the internal lubricating oil pipeline coordinate information of the target steam turbine and the lubricating oil pipeline grade parameter to obtain a second preprocessing set containing the lubricating oil pipeline mark; A three-dimensional modeling unit is used to traverse the second preprocessing set, and to perform three-dimensional modeling of the lubricating oil pipeline inside the target steam turbine based on a three-dimensional visualization tool to obtain a three-dimensional model of the pipeline; The preprocessing set generation unit is used to extract nodes and edges of the pipeline three-dimensional model, connect the nodes and edges in sequence based on the flow direction of the lubricating oil, generate a pipeline topology connection diagram, integrate the pipeline topology connection diagram and the second preprocessing set, and obtain an associated preprocessing set.

[0021] In a third aspect, the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned method for real-time temperature prediction of turbine lubricating oil when executing the computer program.

[0022] In a fourth aspect, the present invention further provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-mentioned method for predicting the real-time temperature of the turbine lubricating oil.

[0023] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a real-time temperature prediction method for steam turbine lubricating oil. By deploying lubricating oil measurement points in combination with historical temperature data of lubricating oil in target steam turbine associated data, reasonable deployment of lubricating oil measurement points is achieved, and then the temperature information of lubricating oil can be obtained in real time, so that the temperature prediction model can more accurately capture temperature changes, which is helpful to timely discover abnormal conditions and provide reliable data support for subsequent analysis and decision-making. The method overcomes the problem that when the existing method obtains the real-time operation data of gas turbine lubricating oil, the temperature monitoring points cannot be reasonably deployed based on the key positions of the lubricating oil pipeline, resulting in an excessive amount of real-time operation data of the lubricating oil, making it impossible for the model to accurately capture the temperature changes, thereby affecting the accuracy of model evaluation.

[0024] Furthermore, when deploying lubricating oil measurement points, the weight score of the internal lubricating oil pipeline is calculated based on the random forest decision tree model, thereby ensuring the accurate evaluation of the internal lubricating oil pipeline. At the same time, the internal lubricating oil pipeline is weightedly fused based on the weight score combined with the convolutional neural network to calculate the pipeline independence corresponding to the internal lubricating oil pipeline. Finally, the pipeline independence and pipeline weight score are used as measurement point deployment evaluation indicators to rank the importance of the internal lubricating oil pipeline, thereby ensuring the accurate and scientific deployment of the internal lubricating oil pipeline measurement points, and then the overheating problem of the lubricating oil pipeline can be discovered in time, so that measures can be taken to avoid equipment damage and extend the service life of the equipment.

[0025] Furthermore, using Gaussian filter function to perform standard wavelet basis transformation on the real-time temperature of lubricating oil can more effectively remove the noise in the real-time temperature signal of lubricating oil and retain useful information. This combined method can identify and eliminate high-frequency noise while retaining important low-frequency features, thereby improving the quality of the signal. This combination makes the processing of lubricating oil temperature data more efficient, saves computing resources, and speeds up the analysis.

[0026] Furthermore, drawing a visualized temperature time-frequency diagram based on the temperature frequency coefficient and the temperature acquisition time can simultaneously display the signal information in the time domain and the frequency domain, making the complex temperature change trend clear at a glance. By observing the time-frequency diagram, you can intuitively see the change of temperature over time and the distribution of different frequency components, which helps the temperature prediction model to quickly understand the data characteristics.

[0027] Furthermore, a temperature prediction model is provided. The temperature prediction model uses the gray prediction model as the initial model, so that the trained temperature prediction model can combine the advantages of the gray prediction model and the visualized temperature time-frequency diagram, significantly improving the prediction ability and robustness of the model. The temperature prediction model combined with the visualized temperature time-frequency diagram can more quickly identify key features in the temperature signal, such as periodic fluctuations, abnormal points, etc. These features are crucial for subsequent model construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The accompanying drawings described herein are only for explanation purposes and are not intended to limit the scope of the present invention in any way. In addition, the shapes and proportional dimensions of the various components in the drawings are only schematic, used to help understand the present invention, and are not intended to specifically limit the shapes and proportional dimensions of the various components of the present invention.

[0029] Figure 1 It is a schematic flow chart of the method for predicting the real-time temperature of the lubricating oil of a steam turbine according to the present invention; Figure 2 It is a flow chart of a method for preprocessing target steam turbine associated data according to the present invention; Figure 3 A schematic flow chart of a method for deploying lubricating oil measurement points in combination with lubricating oil historical temperature data in target steam turbine associated data according to the present invention; Figure 4 It is a schematic flow chart of a method for converting the real-time temperature of lubricating oil into a visualized temperature time-frequency diagram based on Gaussian filtering transformation according to the present invention; Figure 5 A schematic diagram of a process for constructing a training temperature prediction model based on a training set and a test set according to the present invention; Figure 6 A schematic flow chart of a method for the temperature prediction model of the present invention to identify a visualized temperature time-frequency graph and output a temperature prediction result; Figure 7 It is a schematic diagram of the flow chart of the real-time temperature prediction system of the steam turbine lubricating oil of the present invention; Figure 8 The figure is a schematic diagram of an electronic device for predicting the real-time temperature of lubricating oil of a steam turbine according to the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0031] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] The present invention is further described in detail below in conjunction with the accompanying drawings: like Figure 1As shown, an embodiment of the present invention provides a method for predicting the real-time temperature of a steam turbine lubricating oil, comprising the following steps: S1, obtaining target steam turbine associated data, preprocessing the target steam turbine associated data, and obtaining an associated preprocessing set; In this embodiment, the target steam turbine may be a single-stage steam turbine, a multi-stage steam turbine, a single-cylinder steam turbine, or a multi-cylinder steam turbine; S2, load the associated preprocessing set, and deploy the lubricating oil measurement points in combination with the lubricating oil historical temperature data in the target steam turbine associated data; S3, collecting the real-time temperature of the steam turbine lubricating oil in real time according to the lubricating oil measurement point, presetting a Gaussian filter function, and converting the real-time temperature of the steam turbine lubricating oil into a visualized temperature time-frequency graph based on the Gaussian filter function; S4, obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs a temperature prediction result.

[0033] like Figure 2 As shown, the steps of preprocessing the target steam turbine associated data to obtain an associated preprocessing set specifically include: S11, loading target steam turbine associated data, processing missing values ​​and duplicate values ​​of the target steam turbine associated data, and obtaining a first preprocessing set; In this embodiment, the target steam turbine associated data includes but is not limited to steam turbine foundation structure parameters, module operation parameters, load parameters, lubricating oil pipeline distribution information and lubricating oil historical temperature data, and when processing missing values ​​and duplicate values ​​of the target steam turbine associated data, interpolation filling, multiple interpolation and model prediction can be used; S12, obtaining a first preprocessing set, and marking the first preprocessing set based on the internal lubricating oil pipeline coordinate information of the target steam turbine and the lubricating oil pipeline grade parameter to obtain a second preprocessing set containing the lubricating oil pipeline mark; In this embodiment, the lubricating oil pipeline inside the target steam turbine can be divided into primary, secondary, tertiary, quaternary and fifth levels, wherein primary and secondary levels are main pipelines, and tertiary, quaternary and fifth levels are branch pipelines, and the ratio of main pipelines to branch pipelines is 2:5; S13, traversing the second preprocessing set, and performing three-dimensional modeling of the lubricating oil pipeline inside the target steam turbine based on a three-dimensional visualization tool to obtain a three-dimensional model of the pipeline; In this embodiment, when the three-dimensional modeling of the lubricating oil pipeline inside the target steam turbine is performed based on the three-dimensional visualization tool, the three-dimensional visualization tool can be Maya, Houdini and SketchUp to realize the three-dimensional modeling of the pipeline; S14, extracting nodes and edges of the pipeline three-dimensional model, connecting the nodes and edges in sequence based on the flow direction of the lubricating oil, generating a pipeline topology connection diagram, integrating the pipeline topology connection diagram and the second preprocessing set, and obtaining an associated preprocessing set.

[0034] like Figure 3 As shown, the steps of deploying lubricating oil measurement points in combination with lubricating oil historical temperature data in target steam turbine associated data specifically include: S21, obtaining lubricating oil historical temperature data and associated preprocessing sets, and calculating the internal lubricating oil pipeline weight score based on a random forest decision tree model;

[0035] in, is the current pipeline weight score, is the number of lubricating oil pipe edges inside the target steam turbine, is the number of nodes of the lubricating oil pipeline inside the target steam turbine, is the number of nodes associated with the internal lubricating oil pipeline, is the number of decision trees in the random forest decision tree model, is the temperature transition value of the pipeline node within a unit acquisition cycle;

[0036] in, is the temperature transition value of the pipeline node within the unit acquisition cycle, is the number of acquisition cycles, is the highest temperature during the acquisition period, is the average temperature during the acquisition period, is the expansion factor of pipeline lubricating oil; S22, weighted fusion of the internal lubricating oil pipeline based on the current pipeline weight score combined with the convolutional neural network, and calculating the pipeline independence corresponding to the internal lubricating oil pipeline;

[0037] in, The pipeline independence corresponding to the internal lubricating oil pipeline; S23, using pipeline independence and pipeline weight score as measurement point deployment evaluation indicators, sorting the importance of internal lubricating oil pipelines, and outputting internal lubricating oil pipelines with an importance greater than a preset threshold as lubricating oil measurement points;

[0038] in, is the pipeline importance, is the importance weighting parameter;

[0039] in, is the current pipeline weight score, is the number of nodes of the lubricating oil pipeline inside the target steam turbine, is the number of nodes associated with the internal lubricating oil pipeline.

[0040] like Figure 4 As shown in FIG. 1 , the steps of converting the real-time temperature of the steam turbine lubricating oil into a visualized temperature time-frequency diagram based on the Gaussian filter function specifically include: S31, presetting a Gaussian filter function, and performing a standard wavelet basis conversion on the real-time temperature of the lubricating oil in combination with the Gaussian filter function; The standard wavelet basis transformation is expressed as:

[0041] in, is the Gaussian filter function, is the lubricating oil real-time temperature sequence output vector, are the wavelet transform coefficients, They represent the high-frequency oscillation component and the low-frequency oscillation component of the real-time temperature series of the lubricating oil respectively;

[0042] in, is the scale parameter of the Gaussian filter function, is the lubricating oil real-time temperature sequence output vector, is the average temperature of the lubricating oil real-time temperature series, is the normalized coefficient of the Gaussian filter function; S32, performing inner product on the standard wavelet basis conversion and the lubricating oil real-time temperature sequence to obtain the temperature frequency coefficient; S33, loading the temperature frequency coefficient, and drawing a visualized temperature time-frequency diagram based on the temperature frequency coefficient and the temperature acquisition time.

[0043] like Figure 5 As shown in FIG. 1 , the steps of constructing a training temperature prediction model based on a training set and a test set specifically include: S41, obtaining a training set, preprocessing the training set, and detecting a temperature mutation point in the training set; S42, constructing a temperature prediction model, the temperature prediction model uses a grey prediction model as an initial model, inputs the preprocessed training set into the initial model for unsupervised prediction, sets the number of iterative training times, and obtains a temperature prediction result; S43, obtaining a temperature prediction result, using a cross entropy loss function to calculate the difference between the temperature prediction result and the actual temperature result, and using a stochastic gradient descent algorithm to iteratively initialize the hyperparameters of the initial model along the negative gradient direction of the cross entropy loss function until the model converges; S44, obtaining a validation set, using the validation set to verify the accuracy of the model, and determining whether the accuracy of the temperature prediction result and the actual temperature result exceeds a preset accuracy threshold; S45, if it exceeds the preset accuracy threshold, output the trained temperature prediction model; if it does not exceed the preset accuracy threshold, return to step S42.

[0044] In this embodiment, the preset accuracy threshold can be set to 0.95-0.98.

[0045] like Figure 6 As shown in FIG. 1 , the steps of the temperature prediction model identifying the visualized temperature time-frequency graph and outputting the temperature prediction result specifically include: S51, obtaining a visualized temperature time-frequency graph, and extracting a mutation factor of the visualized temperature time-frequency graph by a temperature prediction model; S52, loading the mutation factor and the visualized temperature time-frequency diagram, and the temperature prediction model performs parameter fitting on the mutation factor and the visualized temperature time-frequency diagram to obtain a parameter fitting set of at least one order; S53, obtaining a parameter fitting set, taking the parameter fitting set as input, executing a temperature prediction model, and the temperature prediction model outputs a temperature prediction result corresponding to a visualized temperature time-frequency graph; The output of the parameter fit set is expressed as:

[0046] in, Output signal for parameter fitting set, is the input vector for visualizing the temperature time-frequency diagram, is the mutation factor corresponding to the visualized temperature time-frequency diagram.

[0047] like Figure 7 As shown, the embodiment of the present invention also provides a real-time temperature prediction system for steam turbine lubricating oil, comprising the following modules: A preprocessing module is used to obtain target steam turbine associated data, preprocess the target steam turbine associated data, and obtain an associated preprocessing set; The measurement point deployment module is used to load the associated preprocessing set and deploy the lubricating oil measurement points in combination with the target steam turbine associated data; The time-frequency graph conversion module is used to collect the real-time temperature of the steam turbine lubricating oil in real time according to the lubricating oil measurement point, preset a Gaussian filter function, and convert the real-time temperature of the steam turbine lubricating oil into a visual temperature time-frequency graph based on the Gaussian filter function; The temperature prediction module is used to obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a training temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs the temperature prediction result.

[0048] In this embodiment, the preprocessing module specifically includes: A first preprocessing unit is used to load target steam turbine associated data, process missing values ​​and duplicate values ​​of the target steam turbine associated data, and obtain a first preprocessing set; A second preprocessing unit is used to obtain the first preprocessing set, and mark the first preprocessing set based on the internal lubricating oil pipeline coordinate information of the target steam turbine and the lubricating oil pipeline grade parameter to obtain a second preprocessing set containing the lubricating oil pipeline mark; A three-dimensional modeling unit is used to traverse the second preprocessing set, and to perform three-dimensional modeling of the lubricating oil pipeline inside the target steam turbine based on a three-dimensional visualization tool to obtain a three-dimensional model of the pipeline; The preprocessing set generation unit is used to extract nodes and edges of the pipeline three-dimensional model, connect the nodes and edges in sequence based on the flow direction of the lubricating oil, generate a pipeline topology connection diagram, integrate the pipeline topology connection diagram and the second preprocessing set, and obtain an associated preprocessing set.

[0049] Example 1 like Figure 8 As shown, the present invention also provides an electronic device 100 for a real-time temperature prediction method of a turbine lubricating oil; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0050] The memory 101 can be used to store the computer program 103, and the processor 102 implements the steps of the real-time temperature prediction method of the steam turbine lubricating oil described in the embodiment by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data (such as audio data) created according to the use of the electronic device 100, etc. In addition, the memory 101 can include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0051] The at least one processor 102 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and uses various interfaces and lines to connect various parts of the entire electronic device 100.

[0052] The memory 101 in the electronic device 100 stores a plurality of instructions to implement a method for predicting the real-time temperature of a steam turbine lubricating oil. The processor 102 can execute the plurality of instructions to implement: Acquire target steam turbine associated data, pre-process the target steam turbine associated data, and obtain an associated pre-processed set; Load the associated preprocessing set, and deploy lubricating oil measurement points based on the lubricating oil historical temperature data in the target steam turbine associated data; The real-time temperature of the steam turbine lubricating oil is collected in real time according to the lubricating oil measurement point, a Gaussian filter function is preset, and the real-time temperature of the steam turbine lubricating oil is converted into a visual temperature time-frequency diagram based on the Gaussian filter function; Obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs the temperature prediction result.

[0053] Example 2 If the module / unit integrated in the electronic device 100 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 this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory and read-only memory (ROM, Read-Only Memory).

[0054] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0055] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0056] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0057] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0058] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, a person skilled in the art can still modify or make equivalent substitutions to the specific implementations of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention are within the scope of protection of the claims of the present invention to be approved.

Claims

1. A method for predicting the real-time temperature of a steam turbine lubricating oil, characterized in that: The following steps are involved: Acquire target steam turbine associated data, pre-process the target steam turbine associated data, and obtain an associated pre-processed set; Load the associated preprocessing set, and deploy lubricating oil measurement points based on the lubricating oil historical temperature data in the target steam turbine associated data; The real-time temperature of the steam turbine lubricating oil is collected in real time according to the lubricating oil measurement point, a Gaussian filter function is preset, and the real-time temperature of the steam turbine lubricating oil is converted into a visual temperature time-frequency diagram based on the Gaussian filter function; Obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs the temperature prediction result.

2. A method for predicting the real-time temperature of a steam turbine lubricating oil according to claim 1, characterized in that: The step of preprocessing the target steam turbine associated data to obtain an associated preprocessing set specifically includes: Loading target steam turbine associated data, processing missing values ​​and duplicate values ​​of the target steam turbine associated data, and obtaining a first preprocessing set; Acquire a first preprocessing set, and mark the first preprocessing set based on the internal lubricating oil pipeline coordinate information of the target steam turbine and the lubricating oil pipeline grade parameter to obtain a second preprocessing set containing the lubricating oil pipeline mark; Traversing the second preprocessing set, three-dimensionally modeling the lubricating oil pipeline inside the target steam turbine based on a three-dimensional visualization tool to obtain a three-dimensional model of the pipeline; Nodes and edges of the pipeline three-dimensional model are extracted, and the nodes and edges are connected in sequence based on the flow direction of the lubricating oil to generate a pipeline topological connection diagram. The pipeline topological connection diagram and the second preprocessing set are integrated to obtain an associated preprocessing set.

3. A method for predicting the real-time temperature of a steam turbine lubricating oil according to claim 1, characterized in that: The step of deploying lubricating oil measurement points in combination with lubricating oil historical temperature data in target steam turbine associated data specifically includes: Obtain lubricating oil historical temperature data and associated preprocessing sets, and calculate the internal lubricating oil pipeline weight score based on the random forest decision tree model; in, is the current pipeline weight score, is the number of lubricating oil pipe edges inside the target steam turbine, is the number of nodes of the lubricating oil pipeline inside the target steam turbine, is the number of nodes associated with the internal lubricating oil pipeline, is the number of decision trees in the random forest decision tree model, is the temperature transition value of the pipeline node within a unit acquisition cycle; in, is the temperature transition value of the pipeline node within the unit acquisition cycle, is the number of acquisition cycles, is the highest temperature during the acquisition period, is the average temperature during the acquisition period, is the expansion factor of pipeline lubricating oil; Based on the current pipeline weight score combined with the convolutional neural network, the internal lubricating oil pipeline is weighted and fused to calculate the pipeline independence corresponding to the internal lubricating oil pipeline; in, The pipeline independence corresponding to the internal lubricating oil pipeline; Using pipeline independence and pipeline weight scores as measurement point deployment evaluation indicators, the importance of internal lubricating oil pipelines is ranked, and internal lubricating oil pipelines with a value greater than a preset importance threshold are output as lubricating oil measurement points; in, is the pipeline importance, is the importance weighting parameter; in, is the current pipeline weight score, is the number of nodes of the lubricating oil pipeline inside the target steam turbine, is the number of nodes associated with the internal lubricating oil pipeline.

4. A method for predicting the real-time temperature of a steam turbine lubricating oil according to claim 1, characterized in that: The step of converting the real-time temperature of the steam turbine lubricating oil into a visualized temperature time-frequency graph based on a Gaussian filter function specifically includes: A Gaussian filter function is preset, and standard wavelet basis conversion is performed on the real-time temperature of the lubricating oil in combination with the Gaussian filter function; The standard wavelet basis transformation is expressed as: in, is the Gaussian filter function, is the lubricating oil real-time temperature sequence output vector, are the wavelet transform coefficients, They represent the high-frequency oscillation component and the low-frequency oscillation component of the real-time temperature series of the lubricating oil respectively; in, is the scale parameter of the Gaussian filter function, is the lubricating oil real-time temperature sequence output vector, is the average temperature of the lubricating oil real-time temperature series, is the normalized coefficient of the Gaussian filter function; The inner product of the standard wavelet basis transformation and the real-time temperature series of the lubricating oil is performed to obtain the temperature frequency coefficient; Load the temperature frequency coefficient and draw a visual temperature time-frequency graph based on the temperature frequency coefficient and temperature acquisition time.

5. A method for predicting the real-time temperature of a steam turbine lubricating oil according to claim 1, characterized in that: The step of constructing a training temperature prediction model based on the training set and the test set specifically includes: Obtain a training set, preprocess the training set, and detect temperature mutation points in the training set; Construct a temperature prediction model. The temperature prediction model uses the grey prediction model as the initial model. The preprocessed training set is input into the initial model for unsupervised prediction. The number of iterative training times is set to obtain the temperature prediction result. Get the temperature prediction result, use the cross entropy loss function to calculate the difference between the temperature prediction result and the actual temperature result, and use the stochastic gradient descent algorithm to iteratively initialize the hyperparameters of the initial model along the negative gradient direction of the cross entropy loss function until the model converges; A validation set is obtained, and the accuracy of the model is verified using the validation set to determine whether the accuracy of the temperature prediction result and the actual temperature result exceeds the preset accuracy threshold. If the accuracy exceeds the preset accuracy threshold, the trained temperature prediction model is output.

6. A method for predicting the real-time temperature of a steam turbine lubricating oil according to claim 1, characterized in that: The step of the temperature prediction model identifying the visualized temperature time-frequency graph and outputting the temperature prediction result specifically includes: Obtain a visualized temperature time-frequency graph, and use the temperature prediction model to extract the mutation factor of the visualized temperature time-frequency graph; Loading the mutation factor and the visualized temperature time-frequency diagram, the temperature prediction model performs parameter fitting on the mutation factor and the visualized temperature time-frequency diagram to obtain a parameter fitting set of at least one order; Obtain a parameter fitting set, use the parameter fitting set as input, execute a temperature prediction model, and the temperature prediction model outputs a temperature prediction result corresponding to a visualized temperature time-frequency graph; The output of the parameter fit set is expressed as: in, Output signal for parameter fitting set, is the input vector for visualizing the temperature time-frequency diagram, is the mutation factor corresponding to the visualized temperature time-frequency diagram.

7. A real-time temperature prediction system for steam turbine lubricating oil, characterized in that: Includes the following modules: A preprocessing module is used to obtain target steam turbine associated data, preprocess the target steam turbine associated data, and obtain an associated preprocessing set; The measurement point deployment module is used to load the associated preprocessing set and deploy the lubricating oil measurement points in combination with the target steam turbine associated data; The time-frequency graph conversion module is used to collect the real-time temperature of the steam turbine lubricating oil in real time according to the lubricating oil measurement point, preset a Gaussian filter function, and convert the real-time temperature of the steam turbine lubricating oil into a visual temperature time-frequency graph based on the Gaussian filter function; The temperature prediction module is used to obtain a visualized temperature time-frequency graph, divide the visualized temperature time-frequency graph into a training set and a test set, build a training temperature prediction model based on the training set and the test set, execute the temperature prediction model, and the temperature prediction model recognizes the visualized temperature time-frequency graph and outputs the temperature prediction result.

8. A real-time temperature prediction system for lubricating oil of a steam turbine according to claim 7, characterized in that: The preprocessing module specifically includes: A first preprocessing unit is used to load target steam turbine associated data, process missing values ​​and duplicate values ​​of the target steam turbine associated data, and obtain a first preprocessing set; A second preprocessing unit is used to obtain the first preprocessing set, and mark the first preprocessing set based on the internal lubricating oil pipeline coordinate information of the target steam turbine and the lubricating oil pipeline grade parameter to obtain a second preprocessing set containing the lubricating oil pipeline mark; A three-dimensional modeling unit is used to traverse the second preprocessing set, and to perform three-dimensional modeling of the lubricating oil pipeline inside the target steam turbine based on a three-dimensional visualization tool to obtain a three-dimensional model of the pipeline; The preprocessing set generation unit is used to extract nodes and edges of the pipeline three-dimensional model, connect the nodes and edges in sequence based on the flow direction of the lubricating oil, generate a pipeline topology connection diagram, integrate the pipeline topology connection diagram and the second preprocessing set, and obtain an associated preprocessing set.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for predicting the real-time temperature of the steam turbine lubricating oil according to any one of claims 1 to 6 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for predicting the real-time temperature of the steam turbine lubricating oil according to any one of claims 1 to 6 are implemented.

Images