Gas turbine performance analysis method and device based on AI large model

Abstract

The invention relates to a gas turbine performance analysis method and device based on an AI large model, and is applied to the technical field of gas turbines.The method comprises the steps that a user instruction is obtained, a gas turbine professional large language model is used for recognizing the user instruction, and whether the user instruction belongs to a performance analysis instruction of a gas turbine or not is judged; if not, outputting a corresponding query result; if yes, a corresponding performance analysis model is called, and a monitoring parameter list corresponding to each input physical quantity of the performance analysis model is extracted from a monitoring parameter set of the gas turbine; determining a final monitoring parameter list based on the Pearson correlation coefficients among all the monitoring parameters in the monitoring parameter list and the historical values of all the monitoring parameters; taking an average value of all monitoring parameters in the final monitoring parameter list as a target monitoring parameter of the input physical quantity; and inputting the target monitoring parameter corresponding to each input physical quantity into the performance analysis model, and outputting a performance analysis result. The efficiency and accuracy of performance analysis can be improved.

Description

Technical Field

[0001] The present application relates to the field of gas turbine technology, and in particular to a gas turbine performance analysis method and device based on an AI large model. Background Art

[0002] As an efficient and environmentally friendly power equipment, gas turbines play an important role in my country's energy, electricity and other fields. However, during the operation of gas turbines, due to various factors, performance degradation and frequent failures may occur. In order to improve the operating efficiency and safety of gas turbines, an efficient, accurate and intelligent performance analysis system is urgently needed.

[0003] Existing gas turbine performance analysis is mainly based on thermodynamic mechanism models. The gas turbine performance is calculated and evaluated according to thermodynamic principles, accurately describing the performance characteristics of components under different operating conditions, and providing reliable support for optimization strategies.

[0004] However, the gas turbine performance analysis method based on thermodynamic models can only solve specific problems of a certain component, and its versatility and integration are not high. When dealing with complex problems, users are required to develop or call relevant algorithms based on their knowledge and experience, which requires high professional knowledge and ability of users, and the efficiency and accuracy of gas turbine performance analysis are low. Summary of the invention

[0005] In order to solve the above technical problems, the present application provides a gas turbine performance analysis method, device, electronic device and storage medium based on an AI large model.

[0006] According to a first aspect of the present application, a gas turbine performance analysis method based on an AI large model is provided, comprising: Obtaining a user instruction, and using a pre-built large language model for gas turbine professionals to identify the user instruction, and determining whether the user instruction belongs to a performance analysis instruction for the gas turbine; If the user instruction does not belong to a performance analysis instruction of a gas turbine, outputting a query result corresponding to the user instruction; If the user instruction belongs to a performance analysis instruction of the gas turbine, calling a corresponding performance analysis model, and extracting a monitoring parameter list corresponding to each input physical quantity of the performance analysis model from a monitoring parameter set of the gas turbine; Determine a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter; Taking the average value of all monitoring parameters in the final monitoring parameter list as the target monitoring parameter of the input physical quantity; The target monitoring parameters corresponding to the various input physical quantities are input into the performance analysis model, and the corresponding performance analysis results are output.

[0007] Optionally, determining a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and historical values ​​of each monitoring parameter includes: For a single input physical quantity, calculating the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity; If there is a Pearson correlation coefficient less than a preset threshold, determining irrelevant monitoring parameters according to the historical values ​​of each monitoring parameter in the monitoring parameter list; Deleting irrelevant monitoring parameters from the monitoring parameter list, and returning to the step of calculating the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity, until there is no Pearson correlation coefficient less than a preset threshold; The monitoring parameter list without a Pearson correlation coefficient less than a preset threshold is used as the final monitoring parameter list corresponding to the input physical quantity.

[0008] Optionally, before determining the final monitoring parameter list based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list and the historical values ​​of each monitoring parameter, the method further includes: If there are monitoring parameters in the monitoring parameter list that are inconsistent with the target unit of the input physical quantity, the monitoring parameters are converted into units to obtain the monitoring parameters in the target unit. Optionally, determining irrelevant monitoring parameters according to historical values ​​of each monitoring parameter in the monitoring parameter list includes: Calculate the average value and variance of each monitoring parameter in the monitoring parameter list at multiple times; Determine the mean and variance of the anomaly based on the mean and variance of each monitoring parameter; The monitoring parameters corresponding to the abnormal mean and variance are determined as irrelevant monitoring parameters.

[0009] Optionally, the gas turbine performance analysis method based on the AI ​​large model further includes: Obtaining gas turbine professional corpus, and converting the gas turbine professional corpus into question-answer pair data; A general large language model is obtained, and the general large language model is fine-tuned using the question-answer pair data to obtain the gas turbine professional large language model.

[0010] According to a second aspect of the present application, a gas turbine performance analysis device based on an AI large model is provided, comprising: A user instruction recognition module is used to obtain user instructions, identify the user instructions using a pre-built gas turbine professional large language model, and determine whether the user instructions belong to a gas turbine performance analysis instruction; a query result output module, configured to output a query result corresponding to the user instruction if the user instruction recognition module determines that the user instruction does not belong to a performance analysis instruction of a gas turbine; a monitoring parameter list determination module, configured to call a corresponding performance analysis model if the user instruction recognition module determines that the user instruction belongs to a performance analysis instruction of a gas turbine, and extract a monitoring parameter list corresponding to each input physical quantity of the performance analysis model from a monitoring parameter set of the gas turbine; A monitoring parameter list updating module, used for determining a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter; A target monitoring parameter determination module, used to take the average value of all monitoring parameters in the final monitoring parameter list as the target monitoring parameter of the input physical quantity; The performance analysis result output module is used to input the target monitoring parameters corresponding to each input physical quantity into the performance analysis model and output the corresponding performance analysis results.

[0011] Optionally, the monitoring parameter list updating module is specifically used to calculate, for a single input physical quantity, the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity; if there is a Pearson correlation coefficient less than a preset threshold, determine irrelevant monitoring parameters based on the historical values ​​of each monitoring parameter in the monitoring parameter list; delete the irrelevant monitoring parameters from the monitoring parameter list, and return to the step of calculating the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity until there is no Pearson correlation coefficient less than the preset threshold; and use the monitoring parameter list in which there is no Pearson correlation coefficient less than the preset threshold as the final monitoring parameter list corresponding to the input physical quantity.

[0012] Optionally, the gas turbine performance analysis device based on the AI ​​large model further includes: The unit conversion module is used to convert the monitoring parameters in the monitoring parameter list to obtain the monitoring parameters in the target unit if there are monitoring parameters that are inconsistent with the target unit of the input physical quantity.

[0013] Optionally, the monitoring parameter list updating module is specifically configured to determine irrelevant monitoring parameters according to historical values ​​of each monitoring parameter in the monitoring parameter list through the following steps: Calculate the average value and variance of each monitoring parameter in the monitoring parameter list at multiple times; Determine the mean and variance of the anomaly based on the mean and variance of each monitoring parameter; The monitoring parameters corresponding to the abnormal mean and variance are determined as irrelevant monitoring parameters.

[0014] Optionally, the gas turbine performance analysis device based on the AI ​​large model further includes: The gas turbine professional large language model construction module is used to obtain gas turbine professional corpus and convert the gas turbine professional corpus into question-answer pair data; obtain a general large language model and use the question-answer pair data to fine-tune the general large language model to obtain the gas turbine professional large language model.

[0015] According to a third aspect of the present application, an electronic device is provided, comprising: a processor, wherein the processor is configured to execute a computer program stored in a memory, wherein the computer program implements the method described in the first aspect when executed by the processor.

[0016] According to a fourth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method described in the first aspect is implemented.

[0017] According to a fifth aspect of the present application, a computer program product is provided. When the computer program product is run on a computer, the computer is enabled to execute the method described in the first aspect.

[0018] Compared with the prior art, the technical solution provided by the embodiments of the present application has the following advantages: By pre-building a large language model for gas turbines, the gas turbine performance analysis model based on thermodynamic mechanisms is intelligently integrated, which can understand user instructions in language form and understand them independently, and automatically make decisions and plans. The performance analysis model is called according to demand to complete the complex tasks contained in the user instructions, with higher integration and user-friendliness. When calling the corresponding performance analysis model, the corresponding monitoring parameter list is extracted based on the input physical quantity of the performance analysis model. In order to avoid the hallucination phenomenon that may occur in the large language model, the Pearson correlation coefficient between the monitoring parameters is used to update the monitoring parameter list to ensure the correctness of the final monitoring parameter list, and then ensure the correctness of the target monitoring parameters. The target monitoring parameters of each input physical quantity are input into the performance analysis model, and the performance analysis results are output, which can ultimately improve the accuracy of the performance analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0021] Figure 1 This is a system architecture diagram applicable to the gas turbine performance analysis method based on the AI ​​large model in the embodiment of the present application; Figure 2 A schematic diagram of the construction and calling of a gas turbine knowledge base in an embodiment of the present application; Figure 3 A flow chart of a gas turbine performance analysis method based on an AI large model in an embodiment of the present application; Figure 4 A schematic diagram of the structure of a gas turbine performance analysis device based on an AI large model in an embodiment of the present application; Figure 5 A schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to more clearly understand the above-mentioned purposes, features and advantages of the present application, the scheme of the present application will be further described below. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0023] In the following description, many specific details are set forth to facilitate a full understanding of the present application, but the present application may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only part of the embodiments of the present application, rather than all of the embodiments.

[0024] With the continuous advancement of science and technology, artificial intelligence technology has ushered in significant development. In particular, in recent years, the large language model with the Transformer model architecture as the core has made significant breakthroughs. Representative artificial intelligence large language models such as BERT, GPT, and GLM continue to emerge. With the continuous improvement of large language models, their application in various fields has gradually expanded. From text generation, machine translation, sentiment analysis to information extraction, large language models provide strong support for solving complex problems and have triggered revolutionary changes in many industries.

[0025] The current capabilities of the big model are mainly focused on solving general language problems. It lacks industry-related knowledge accumulation and has weak understanding of professional scenarios and question-answering capabilities. In addition, the big language model can only process text and has no processing capabilities for thermodynamic models, operating data, and images that are widely present in the gas turbine field.

[0026] Based on this, the embodiment of the present application constructs a professional large language model for gas turbines, which has a strong understanding of professional knowledge. It can combine professional data and documents to deeply understand relevant issues in the field of gas turbines and provide professional answers, and specifically improve the professional capabilities of the general large language model in the field of gas turbine operation and maintenance, identify and respond to professional issues more quickly, reduce error rates, and improve work efficiency.

[0027] Figure 1 The system architecture diagram of the gas turbine performance analysis method based on the AI ​​big model in the embodiment of the present application is shown in FIG. The system architecture includes a gas turbine knowledge base and a performance analysis model, and the two are integrated through the gas turbine big language model intelligent agent technology to form a gas turbine professional big language model platform, which is responsible for interacting with users based on natural language.

[0028] Optionally, the method for constructing a large language model for gas turbines includes: Obtain professional corpus of gas turbines, for example, collect corpus from books, technical documents or other sources related to gas turbines. Perform preprocessing, data cleaning and corpus quality assessment on the collected corpus to ensure the accuracy, comprehensiveness and representativeness of the corpus. Convert the professional corpus of gas turbines into question-answer pair data, which is suitable for fine-tuning the general large language model.

[0029] Obtain a general language model, use the general language model as the base, and use the question-answer data to fine-tune the general language model to obtain a gas turbine professional language model. The gas turbine professional language model has the ability to understand and reason about the gas turbine professional knowledge involved in the corpus.

[0030] The main function of the gas turbine knowledge base is to store and manage professional books, documents and other text materials related to gas turbines, and to provide basic or timely professional knowledge reference materials. Figure 2 A schematic diagram of the construction and call of the gas turbine knowledge base in the embodiment of the present application. The steps of constructing the gas turbine knowledge base include: collecting books and documents related to gas turbines, and storing and managing the collected text materials. Using text recognition technology, extracting text from the document to meet the input requirements of the large language model. Segmenting the text in the document to form text blocks, and converting the text blocks into vectors through the language embedding model to form a vector database, namely the gas turbine knowledge base.

[0031] When the gas turbine professional large language model interacts with the user based on natural language, it can also call the gas turbine knowledge base on the basis of its own professional knowledge to output content related to the user's instructions. Specifically, for any user instruction, the steps of calling the gas turbine knowledge base include: converting the user instruction into a vector through an embedding model, calculating the similarity between the user instruction and the vector of each text block in the gas turbine knowledge base, retrieving the text block with the highest similarity to the user instruction vector, and constructing the prompt word of the gas turbine professional large language model. The prompt word is a relatively standardized word that can be used as the input of the gas turbine professional large language model. The prompt word can include the user instruction and the text block with the highest similarity. The prompt word is input into the gas turbine professional large language model, and the gas turbine professional large language model outputs the knowledge base search content related to the user instruction.

[0032] The performance analysis model includes performance analysis models of multiple components. For example, the compressor efficiency model can be expressed as the following formula: ; Among them, the input physical quantities include: inlet temperature , inlet pressure , outlet temperature and outlet pressure , the output physical quantity is the compressor efficiency ; It represents the adiabatic index of the working fluid and is selected according to the properties of the working fluid.

[0033] The turbine efficiency model can be expressed as follows: ; Among them, the input physical quantities include: inlet temperature , inlet pressure , outlet temperature and outlet pressure ; The output physical quantity is the compressor efficiency .

[0034] The combustion chamber temperature and pressure model can be expressed as follows: ; Among them, the input physical quantities include: combustion chamber inlet temperature , Combustion chamber inlet pressure , fuel flow and working fluid flow ; The output physical quantity is the combustion chamber outlet temperature and combustion chamber outlet pressure . is the calorific value of fuel, is the specific heat capacity of the working fluid at constant pressure.

[0035] The nozzle flow model can be expressed as follows: ; Among them, the input physical quantities include: nozzle inlet temperature , nozzle inlet pressure , nozzle outlet pressure and nozzle throat area ; The intermediate variable is the exit Mach number , the output physical quantity is the nozzle flow rate , R is the gas constant of the working fluid.

[0036] The models of each component are encapsulated to form a performance analysis model that can be used by the gas turbine professional large language model. For the model of each component, a tool declaration can be given, including the data type and physical meaning of its input and output, as well as the overall functional description of the model.

[0037] See also Figure 3 , Figure 3 A flow chart of a gas turbine performance analysis method based on an AI large model in an embodiment of the present application may include the following steps: Step S302, obtaining user instructions.

[0038] Step S304, using the gas turbine professional large language model, the user instruction is identified to determine whether the user instruction belongs to the gas turbine performance analysis instruction.

[0039] The gas turbine professional language model can identify the user instruction based on its own understanding ability and determine whether the user instruction belongs to the gas turbine performance analysis instruction. If the user instruction does not belong to the gas turbine performance analysis instruction, execute step S306; if the user instruction belongs to the gas turbine performance analysis instruction, execute step S308.

[0040] For example, if the user's instruction is "What is a compressor?", the gas turbine professional language model can recognize that the user's instruction is a general knowledge answer instruction and can use Figure 2 The process shown outputs the query result corresponding to the user instruction. If the user instruction is "What is the efficiency of the compressor?", the gas turbine professional language model can recognize that the user instruction is a gas turbine performance analysis instruction, and execute step S308.

[0041] Step S306: output the query result corresponding to the user instruction.

[0042] Step S308: calling the corresponding performance analysis model, and extracting a monitoring parameter list corresponding to each input physical quantity of the performance analysis model from the monitoring parameter set of the gas turbine.

[0043] As mentioned above, different components can correspond to different models, and the gas turbine professional language model can call the corresponding performance analysis model according to the recognized content. The input physical quantity of each performance analysis model is different. Therefore, based on the input physical quantity of the performance analysis model, the monitoring parameter list corresponding to each input physical quantity can be extracted from the monitoring parameter set of the gas turbine.

[0044] For example, the input physical quantities of the compressor efficiency model include: inlet temperature, inlet pressure, outlet temperature and outlet pressure. Therefore, the gas turbine professional large language model can extract the monitoring parameter list corresponding to the inlet temperature, the monitoring parameter list corresponding to the inlet pressure, the monitoring parameter list corresponding to the outlet temperature and the monitoring parameter list corresponding to the outlet pressure from the monitoring parameter set of the gas turbine. It should be noted that there are many monitoring parameters of the gas turbine system, and the same physical quantity often has multiple redundant quantities. For example, the inlet temperature may have more than two monitoring parameters. Therefore, extracting the monitoring parameter list corresponding to each input physical quantity can extract more complete monitoring parameters.

[0045] Step S310 , determining a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter.

[0046] In some embodiments, the units of monitoring parameters of different units may also be different. If there are monitoring parameters in the monitoring parameter list that are inconsistent with the target unit of the input physical quantity, the monitoring parameters are converted to obtain the monitoring parameters under the target unit. The target unit of the input physical quantity is the unit that should be used for the input physical quantity in the performance analysis model.

[0047] If the gas turbine professional large language model has hallucinations, the output monitoring parameter list will be inaccurate. In order to avoid this problem, the gas turbine professional large language model can use the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list and the historical values ​​of each monitoring parameter to update the monitoring parameter list. In other words, the monitoring parameters with low accuracy in the monitoring parameter list are eliminated.

[0048] Optionally, for a single input physical quantity, the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity can be calculated. If there is a Pearson correlation coefficient less than a preset threshold, it means that there are monitoring parameters with low accuracy in the monitoring parameter list, and irrelevant monitoring parameters are determined based on the historical values ​​of each monitoring parameter in the monitoring parameter list. Irrelevant monitoring parameters are monitoring parameters with low accuracy. Delete the irrelevant monitoring parameters from the monitoring parameter list, and return to the step of calculating the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity until there is no Pearson correlation coefficient less than the preset threshold. The monitoring parameter list in which there is no Pearson correlation coefficient less than the preset threshold is used as the final monitoring parameter list corresponding to the input physical quantity. In this way, by repeatedly confirming and iterating the monitoring parameter list, inaccurate monitoring parameters are eliminated to ensure the correctness of the final monitoring parameter list.

[0049] Optionally, the method for determining irrelevant monitoring parameters may be: calculating the average value and variance of each monitoring parameter in the monitoring parameter list at multiple moments (e.g., the current moment and multiple historical moments before the current moment). Determine the abnormal average value and variance based on the average value and variance of each monitoring parameter. Determine the monitoring parameter corresponding to the abnormal average value and variance as the irrelevant monitoring parameter.

[0050] For example, the average values ​​of multiple monitoring parameters are compared, and the obviously abnormal monitoring parameters are determined as irrelevant monitoring parameters. Alternatively, if the average values ​​of the various monitoring parameters are consistent, the monitoring parameters with larger variances are determined as irrelevant monitoring parameters.

[0051] Step S312: taking the average value of all monitoring parameters in the final monitoring parameter list as the target monitoring parameter of the input physical quantity.

[0052] Step S314: input the target monitoring parameters corresponding to each input physical quantity into the performance analysis model, and output the corresponding performance analysis results.

[0053] For example, in response to a user instruction of “What is the efficiency of the compressor?”, after obtaining the target monitoring parameters of inlet temperature, inlet pressure, outlet temperature and outlet pressure respectively, the compressor efficiency model is input to obtain the compressor efficiency, and finally the compressor efficiency is output.

[0054] The gas turbine performance analysis method based on the AI ​​large model of the embodiment of the present application, by pre-building a professional large language model of the gas turbine, intelligently integrates the gas turbine performance analysis model based on the thermodynamic mechanism, can understand the user instructions in the form of language and understand them independently, and automatically make decisions and plans. The performance analysis model is called according to demand to complete the complex tasks contained in the user instructions, with higher integration and user-friendliness. When calling the corresponding performance analysis model, the corresponding monitoring parameter list is extracted based on the input physical quantity of the performance analysis model. In order to avoid the hallucination phenomenon that may occur in the large language model, the Pearson correlation coefficient between the monitoring parameters is used to update the monitoring parameter list to ensure the correctness of the final monitoring parameter list, and then ensure the correctness of the target monitoring parameters. The target monitoring parameters of each input physical quantity are input into the performance analysis model, and the performance analysis results are output, which can ultimately improve the accuracy of the performance analysis results.

[0055] Corresponding to the above method embodiment, the present application embodiment also provides a gas turbine performance analysis device based on an AI large model, see Figure 4 The gas turbine performance analysis device 400 based on the AI ​​large model includes: The user instruction recognition module 402 is used to obtain the user instruction, identify the user instruction by using the pre-built gas turbine professional large language model, and determine whether the user instruction belongs to the gas turbine performance analysis instruction; The query result output module 404 is used to output the query result corresponding to the user instruction if the user instruction recognition module determines that the user instruction does not belong to the performance analysis instruction of the gas turbine; The monitoring parameter list determination module 406 is used to call the corresponding performance analysis model and extract the monitoring parameter list corresponding to each input physical quantity of the performance analysis model from the monitoring parameter set of the gas turbine if the user instruction recognition module determines that the user instruction belongs to the performance analysis instruction of the gas turbine; A monitoring parameter list updating module 408 is used to determine a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter; A target monitoring parameter determination module 410 is used to use the average value of all monitoring parameters in the final monitoring parameter list as the target monitoring parameter of the input physical quantity; The performance analysis result output module 412 is used to input the target monitoring parameters corresponding to each input physical quantity into the performance analysis model and output the corresponding performance analysis results.

[0056] Optionally, the monitoring parameter list update module 408 is specifically used to calculate the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity for a single input physical quantity; if there is a Pearson correlation coefficient less than a preset threshold, determine the irrelevant monitoring parameters based on the historical values ​​of each monitoring parameter in the monitoring parameter list; delete the irrelevant monitoring parameters from the monitoring parameter list, and return to the step of calculating the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity until there is no Pearson correlation coefficient less than the preset threshold; and use the monitoring parameter list in which there is no Pearson correlation coefficient less than the preset threshold as the final monitoring parameter list corresponding to the input physical quantity.

[0057] Optionally, the gas turbine performance analysis device 400 based on the AI ​​large model further includes: The unit conversion module is used to convert the monitoring parameters in the target unit if there are monitoring parameters in the monitoring parameter list that are inconsistent with the target unit of the input physical quantity, so as to obtain the monitoring parameters in the target unit.

[0058] Optionally, the monitoring parameter list updating module 408 is specifically configured to determine irrelevant monitoring parameters according to historical values ​​of each monitoring parameter in the monitoring parameter list through the following steps: Calculate the average value and variance of each monitoring parameter in the monitoring parameter list at multiple times; Determine the mean and variance of the anomaly based on the mean and variance of each monitoring parameter; The monitoring parameters corresponding to the abnormal mean and variance are determined as irrelevant monitoring parameters.

[0059] Optionally, the gas turbine performance analysis device 400 based on the AI ​​large model further includes: The gas turbine professional large language model construction module is used to obtain gas turbine professional corpus and convert the gas turbine professional corpus into question-answer pair data; obtain a general large language model and use the question-answer pair data to fine-tune the general large language model to obtain a gas turbine professional large language model.

[0060] The specific details of each module or unit in the above device have been described in detail in the corresponding method, so they will not be repeated here.

[0061] It should be noted that, although several modules or units of the equipment for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0062] An embodiment of the present application also provides an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the above-mentioned gas turbine performance analysis method based on the AI ​​large model in this example implementation.

[0063] Reference Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device in an embodiment of the present application. The specific embodiment of the present application does not limit the specific implementation of the electronic device.

[0064] like Figure 5 As shown, the electronic device may include: a processor 502 , a communication interface 504 , a memory 506 , and a communication bus 508 .

[0065] The processor 502 , the communication interface 504 , and the memory 506 communicate with each other via a communication bus 508 .

[0066] The communication interface 504 is used to communicate with other electronic devices or servers.

[0067] The processor 502 is used to execute the program 510, and specifically can execute the relevant steps in the above method embodiment.

[0068] Specifically, the program 510 may include program codes, which include computer operation instructions.

[0069] The processor 502 may be a central processing unit, or a specific integrated circuit, or one or more integrated circuits configured to implement the embodiments of the present application. The one or more processors included in the smart device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0070] The memory 506 is used to store the program 510. The memory 506 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0071] The program 510 can be specifically used to enable the processor 502 to execute the steps in the above-mentioned embodiment of the gas turbine performance analysis method based on the AI ​​large model.

[0072] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and modules can refer to the corresponding process descriptions in the aforementioned method embodiments and will not be repeated here.

[0073] A computer-readable storage medium is also provided in an embodiment of the present application, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned gas turbine performance analysis method based on the AI ​​large model is implemented.

[0074] It should be noted that the computer-readable storage medium shown in the present application may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, radio frequency, etc., or any suitable combination of the above.

[0075] In an embodiment of the present application, a computer program product is also provided. When the computer program product runs on a computer, it enables the computer to execute the above-mentioned gas turbine performance analysis method based on the AI ​​large model.

[0076] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0077] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A gas turbine performance analysis method based on AI large model, characterized in that: include: Obtaining a user instruction, and using a pre-built large language model for gas turbine professionals to identify the user instruction, and determining whether the user instruction belongs to a performance analysis instruction for the gas turbine; If the user instruction does not belong to a performance analysis instruction of a gas turbine, outputting a query result corresponding to the user instruction; If the user instruction belongs to a performance analysis instruction of the gas turbine, calling a corresponding performance analysis model, and extracting a monitoring parameter list corresponding to each input physical quantity of the performance analysis model from a monitoring parameter set of the gas turbine; Determine a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter; Taking the average value of all monitoring parameters in the final monitoring parameter list as the target monitoring parameter of the input physical quantity; The target monitoring parameters corresponding to the various input physical quantities are input into the performance analysis model, and the corresponding performance analysis results are output.

2. The gas turbine performance analysis method based on AI large model according to claim 1 is characterized in that: The method of determining a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter comprises: For a single input physical quantity, calculating the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity; If there is a Pearson correlation coefficient less than a preset threshold, determining irrelevant monitoring parameters according to the historical values ​​of each monitoring parameter in the monitoring parameter list; Deleting irrelevant monitoring parameters from the monitoring parameter list, and returning to the step of calculating the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity, until there is no Pearson correlation coefficient less than a preset threshold; The monitoring parameter list without a Pearson correlation coefficient less than a preset threshold is used as the final monitoring parameter list corresponding to the input physical quantity.

3. The gas turbine performance analysis method based on AI large model according to claim 1 is characterized in that: Before determining the final monitoring parameter list based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list and the historical values ​​of each monitoring parameter, the method further includes: If there are monitoring parameters in the monitoring parameter list that are inconsistent with the target unit of the input physical quantity, the monitoring parameters are converted into units to obtain the monitoring parameters in the target unit.

4. The gas turbine performance analysis method based on AI large model according to claim 2 is characterized in that: The determining of irrelevant monitoring parameters according to the historical values ​​of each monitoring parameter in the monitoring parameter list includes: Calculate the average value and variance of each monitoring parameter in the monitoring parameter list at multiple times; Determine the mean and variance of the anomaly based on the mean and variance of each monitoring parameter; The monitoring parameters corresponding to the abnormal mean and variance are determined as irrelevant monitoring parameters.

5. The gas turbine performance analysis method based on AI large model according to claim 1 is characterized in that: The method further comprises: Obtaining professional corpus of gas turbines, and converting the professional corpus of gas turbines into question-answer pair data; A general large language model is obtained, and the general large language model is fine-tuned using the question-answer pair data to obtain the gas turbine professional large language model.

6. A gas turbine performance analysis device based on AI large model, characterized in that: The device comprises: A user instruction recognition module is used to obtain user instructions, identify the user instructions using a pre-built gas turbine professional large language model, and determine whether the user instructions belong to a gas turbine performance analysis instruction; a query result output module, configured to output a query result corresponding to the user instruction if the user instruction recognition module determines that the user instruction does not belong to a performance analysis instruction of a gas turbine; a monitoring parameter list determination module, configured to call a corresponding performance analysis model if the user instruction recognition module determines that the user instruction belongs to a performance analysis instruction of a gas turbine, and extract a monitoring parameter list corresponding to each input physical quantity of the performance analysis model from a monitoring parameter set of the gas turbine; A monitoring parameter list updating module, used for determining a final monitoring parameter list corresponding to each input physical quantity based on the Pearson correlation coefficients between all monitoring parameters in the monitoring parameter list corresponding to each input physical quantity and the historical values ​​of each monitoring parameter; A target monitoring parameter determination module, used to take the average value of all monitoring parameters in the final monitoring parameter list as the target monitoring parameter of the input physical quantity; The performance analysis result output module is used to input the target monitoring parameters corresponding to each input physical quantity into the performance analysis model and output the corresponding performance analysis results.

7. The gas turbine performance analysis device based on AI large model according to claim 6 is characterized in that: The monitoring parameter list updating module is specifically used to calculate the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity for a single input physical quantity; if there is a Pearson correlation coefficient less than a preset threshold, determine irrelevant monitoring parameters based on the historical values ​​of each monitoring parameter in the monitoring parameter list; delete the irrelevant monitoring parameters from the monitoring parameter list, and return to the step of calculating the Pearson correlation coefficient between all monitoring parameters in the monitoring parameter list corresponding to the input physical quantity until there is no Pearson correlation coefficient less than the preset threshold; and use the monitoring parameter list without a Pearson correlation coefficient less than the preset threshold as the final monitoring parameter list corresponding to the input physical quantity.

8. The gas turbine performance analysis device based on AI large model according to claim 6 is characterized in that: The device also includes: The unit conversion module is used to convert the monitoring parameters in the monitoring parameter list to obtain the monitoring parameters in the target unit if there are monitoring parameters that are inconsistent with the target unit of the input physical quantity.

9. The gas turbine performance analysis device based on AI large model according to claim 7 is characterized in that: The monitoring parameter list updating module is specifically used to determine irrelevant monitoring parameters according to the historical values ​​of each monitoring parameter in the monitoring parameter list through the following steps: Calculate the average value and variance of each monitoring parameter in the monitoring parameter list at multiple times; Determine the mean and variance of the anomaly based on the mean and variance of each monitoring parameter; The monitoring parameters corresponding to the abnormal mean and variance are determined as irrelevant monitoring parameters.

10. The gas turbine performance analysis device based on AI large model according to claim 6, characterized in that: The device also includes: The gas turbine professional large language model construction module is used to obtain gas turbine professional corpus and convert the gas turbine professional corpus into question-answer pair data; obtain a general large language model and use the question-answer pair data to fine-tune the general large language model to obtain the gas turbine professional large language model.

Images