Gas turbine gas path system intelligent operation and maintenance platform based on digital twinning

Abstract

This invention discloses an intelligent operation and maintenance platform for a gas turbine gas circuit system based on digital twins, relating to the field of gas turbine technology. The platform includes: establishing a virtual model using digital twin technology; acquiring system operation data through a data acquisition and management module; performing online correction and dynamic evolution of the virtual model based on the operation data to achieve a mapping between the virtual model and the actual operating state; using the updated digital twin model to perform dynamic operation state prediction and multi-condition simulation analysis of the gas turbine gas circuit system, obtaining future operating trends and potential anomaly information; combining the prediction results to achieve fault diagnosis, fault warning, remaining life prediction, and intelligent operation and maintenance decision-making for the gas circuit system, providing quantitative and intelligent support for the operation and maintenance process; and displaying the real-time operating status of the gas turbine gas circuit system through a visual interface, providing intuitive and accurate operation and maintenance support, and ensuring the safe, stable, and efficient operation of the gas turbine.

Description

Technical Field

[0001] This invention relates to the field of gas turbine technology, and in particular to an intelligent operation and maintenance platform for gas turbine gas circuit systems based on digital twins. Background Technology

[0002] Gas turbines, as complex thermodynamic systems, are widely used in aerospace power and energy generation. Their operating performance, efficiency, and reliability largely depend on the working state of their gas circuit system. Changes in the operating state of the gas circuit system directly affect the output performance and operational safety of the gas turbine. Therefore, accurate monitoring, prediction, and management of the operating state of the gas turbine's gas circuit system are of great significance.

[0003] Existing analysis and simulation methods for gas turbine gas path systems are typically based on physical mechanism models to describe and calculate the system's operation. However, due to the complex flow characteristics within the gas path system and the variations in component performance over time, traditional methods often struggle to consistently and accurately reflect the true operating state of the gas turbine gas path system in practical applications. Furthermore, during long-term operation of a gas turbine, manufacturing errors, wear and tear, and changes in operating conditions lead to continuous variations in system operating characteristics, making analysis methods based on fixed model parameters insufficient for meeting the demands of high-precision monitoring and prediction.

[0004] With the development of sensing and data acquisition technologies, digital twin technology has been gradually introduced into the operation analysis and management of gas turbine gas circuit systems. By constructing a digital twin model corresponding to the gas turbine gas circuit system and integrating real-time operating data, dynamic mapping and predictive analysis of the gas circuit system's operating status can be achieved, providing a new technical approach for gas turbine operation management.

[0005] However, the existing digital twin operation and maintenance platforms for gas turbine gas circuit systems still have shortcomings. On the one hand, how to fully utilize real-time operational data to perform online correction and dynamic updates of the digital twin model to ensure consistency between the virtual model and the actual operating state still requires further research. On the other hand, how to achieve dynamic prediction of the gas circuit system's operating state and analysis of multiple operating conditions, and to visualize the analysis results in an intuitive and understandable way to support gas turbine fault diagnosis, health management, and intelligent operation and maintenance decisions, still lacks an integrated and systematic solution.

[0006] Therefore, there is an urgent need to establish an intelligent operation and maintenance platform for gas turbine gas circuit systems based on digital twins. By integrating multi-source operating data and model analysis results, the platform can achieve dynamic assessment and trend prediction of the gas circuit system's operating status, identify and warn of potential anomalies and degradation risks in advance, and present the analysis and prediction results in an intuitive way. This will provide effective support for fault diagnosis, health management, and intelligent operation and maintenance decisions for gas turbines. Summary of the Invention

[0007] To address the limitations of existing intelligent operation and maintenance systems for gas turbine gas circuits, this invention provides an intelligent operation and maintenance platform for gas turbine gas circuit systems based on digital twins, comprising a data acquisition and management module, a digital twin modeling and updating module, a dynamic prediction and simulation analysis module, an intelligent diagnosis and operation and maintenance decision-making module, and a visualization display module connected in sequence. The data acquisition and management module is used to collect, preprocess, and store the operating data of various components in the gas turbine gas circuit system in real time. The digital twin modeling and updating module is used to construct a digital twin virtual model based on the structural information of the gas turbine gas circuit system, and to perform online correction and dynamic evolution of the virtual model according to the operating data, so as to realize the mapping between the virtual model and the actual operating state of the gas turbine gas circuit system; The dynamic prediction and simulation analysis module is used to perform dynamic operation state prediction and multi-condition simulation analysis of the gas turbine gas circuit system using the updated digital twin virtual model, and to obtain the future operation trend and potential anomaly information of the system. The intelligent diagnosis and operation and maintenance decision module is used to realize intelligent operation and maintenance decisions for fault diagnosis and early warning of gas turbine gas circuit system and prediction of remaining service life based on the dynamic prediction and simulation analysis results. The visualization module is used to provide real-time visualization of the operating status, prediction results, and maintenance decision information of the gas turbine gas circuit system.

[0008] Furthermore, before storing the data, the data acquisition and management module performs noise reduction, abnormal data identification and removal, missing data compensation and data standardization on the acquired operational data, and performs time synchronization processing on operational data with different sampling frequencies to construct a data sequence with a unified time axis.

[0009] Furthermore, the aforementioned time synchronization processing is applied to multi-source operating data with different sampling frequencies. Mapped to a unified time series through time resampling Its relational expression is: In the formula, For the first The class is mapped to runtime data in a unified time series. For the first Time alignment operator for class-based runtime data; At a certain moment on a unified timeline When no corresponding sample value is available, the runtime data at that moment is calculated using interpolation. : In the formula, , For adjacent times where sampled values ​​exist, ; , They are time points , The running data.

[0010] Furthermore, the conditional variance of the interpolated running data will be calculated to characterize the uncertainty. : In the formula, This is the function for calculating conditional variance. For a moment Operational data; A quality score is constructed based on the aforementioned uncertainty and alignment consistency. : In the formula: , These are the weighting coefficients. For reference operation data; The quality scores are normalized to obtain fusion weights, and these weights are then used to perform quality-gated fusion of various operational data to obtain a unified input for digital twin modeling and updating. In the formula, For the unified input after quality gating fusion, To normalize the fusion weights, For the number of data types running from multiple sources, For the first Class running data at time Quality rating It is a stable term.

[0011] Furthermore, the digital twin modeling and updating module performs online correction and dynamic evolution of the virtual model based on the operational data, wherein the online correction is based on a unified time series. On running data Calculate the digital twin model at time 1 Output and with actual operating data To make comparisons, construct a digital twin model based on parameters. An optimization problem with model bias as the objective function and independent variable as the objective variable. Represented as: In the formula, It is a digital twin model at any time And the parameters are The predicted output at that time It is a moment The actual running data output vector, For a unified time series set; The objective function is minimized using a Bayesian optimization algorithm. To enable online correction of parameters in digital twin models: In the formula, The weighting coefficients for physical consistency constraints. This is the physical consistency residual function; Online correction of synchronous output parameter uncertainty and recursive update of parameter covariance: In the formula, For a moment The parameter vector of the digital twin model, For a moment Update the model parameter vector. For the parameter gain matrix, For digital twin models with current parameters The predicted output is as follows. It is the identity matrix. For a moment The parameter covariance matrix, For a moment The updated parameter covariance matrix, For a moment The parameter covariance matrix, Let be the process noise covariance.

[0012] Furthermore, dynamic evolution, based on online correction of model parameters, models the performance degradation process of the gas turbine gas path system, assuming the key performance indicators of the gas path system are... Its dynamic evolution relationship is expressed as: In the formula, For degenerate evolution function, It is a forward time series; By combining the online correction process of model parameters with the dynamic evolution process of performance degradation, the overall state of the digital twin model is jointly updated, realizing the continuous dynamic evolution of the digital twin model as the gas turbine gas path system operates.

[0013] Furthermore, the dynamic prediction and simulation analysis module's dynamic operating state prediction and multi-condition simulation analysis process lies in the fact that, based on the digital twin model, the state variables at the current moment... Model parameters Key performance indicators By extrapolating the system state, future moments can be obtained. Predicted running status : In the formula To predict the time step, For running state prediction function; Furthermore, by changing the model input parameters or operating boundary conditions under different operating conditions, multi-condition simulation analysis was conducted on the digital twin model to obtain the predicted operating state results under different operating conditions. : In the formula, The number of simulation conditions. For the first Simulation results under various operating conditions; Weights are assigned to each operating condition and updated recursively to obtain the probability weights of the operating conditions at future times. : In the formula, For working conditions To working conditions The transition probability, For the future The Each working condition probability weight; Weighted fusion of multi-condition prediction results based on operating condition probability weights: Propagating the parameter uncertainty to the prediction output yields the prediction covariance and confidence interval: In the formula, For a moment The predicted covariance matrix, For the covariance of operating condition uncertainties, For a moment Jacobian matrix, For a moment At confidence level The confidence interval below, , is the confidence coefficient This indicates taking the diagonal elements of the covariance matrix; Based on the predicted operating status at future moments and the results of uncertainty characterization, the operating trend and potential abnormal states of the gas turbine gas circuit system are analyzed.

[0014] Furthermore, the intelligent diagnosis and operation and maintenance decision-making module compares and analyzes the dynamic operating data of the digital twin model with the health status to identify the fault type and location of the gas turbine gas circuit system; and based on the predicted operating status change trend and fault degree, it assesses the operating risk of the gas turbine gas circuit system, determines the corresponding risk level, and generates fault warning information when the risk level exceeds a preset threshold.

[0015] Furthermore, the remaining service life prediction combines the dynamic evolution trend of key performance indicators with the prediction results of dynamic operating status at future moments to determine the time corresponding to the key performance indicators reaching the preset failure threshold, thereby obtaining the remaining service life of the key components.

[0016] Furthermore, the intelligent diagnosis and operation and maintenance decision module comprehensively evaluates the operating status of the gas turbine gas circuit system based on fault diagnosis results, fault early warning information, and remaining service life prediction results, and generates corresponding intelligent operation and maintenance decisions, including: operating parameter adjustment suggestions generated according to fault type and risk level, maintenance timing suggestions determined according to remaining service life prediction results, maintenance strategy suggestions generated according to the degree of degradation of key components, and component replacement suggestions generated according to failure risk assessment results.

[0017] Compared with the prior art, the significant advantages of this invention are as follows: 1) By integrating multi-source operating data with a digital twin model, this invention achieves real-time mapping and dynamic updating of the operating status of the gas turbine gas circuit system. Compared with traditional analysis methods based on fixed parameters or offline models, it can more accurately reflect the state changes of the gas turbine during actual operation, and improve the accuracy and reliability of operating status assessment and prediction.

[0018] 2) This invention, through dynamic operation state prediction and multi-condition simulation analysis using a digital twin model, can comprehensively evaluate the operating trend of the gas turbine gas circuit system under different operating conditions, providing reliable data support for operating condition optimization, maintenance strategy formulation and operation and maintenance decision-making, and enhancing the scientific nature of system operation management.

[0019] 3) This invention achieves adaptive suppression of low-confidence data through a fusion mechanism of interpolation uncertainty quantification and quality gating, significantly improving model update stability. By introducing physical consistency constraints and recursively updating parameter covariance, parameter correction satisfies both data fitting and physical feasibility, avoiding model drift. Furthermore, by propagating parameter uncertainty to the prediction output and combining it with the working condition probability to recursively calculate the confidence interval, an upgrade from single-value prediction to probabilistic prediction is achieved, significantly enhancing risk assessment capabilities. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 The flowchart of the intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin of this invention; Figure 2 A schematic diagram of the interactive interface of the digital twin platform implemented in this invention; Figure 3 A schematic diagram of the user interface for adjusting simulation parameters implemented in this invention. Detailed Implementation

[0022] The following is in conjunction with the instruction manual appendix. Figure 1-3 This paper provides a more detailed description of the intelligent operation and maintenance platform for gas turbine gas circuit systems based on digital twins provided by this invention. It includes, in sequence, a data acquisition and management module, a digital twin modeling and updating module, a dynamic prediction and simulation analysis module, an intelligent diagnosis and operation and maintenance decision-making module, and a visualization module. The data acquisition and management module is used to collect, preprocess, and store the operating data of each component in the gas turbine gas circuit system in real time. Sensing devices are installed on each key component in the gas turbine gas circuit system to collect multi-source operating data generated during system operation in real time. The multi-source operating data includes operating parameters such as temperature, pressure, speed, and flow rate related to the compressor, combustion chamber, and turbine. Before data storage, the data acquisition and management module performs noise reduction, abnormal data identification and removal, missing data compensation, and data standardization on the acquired operational data. It also performs time synchronization processing on operational data with different sampling frequencies to construct a data sequence with a unified time axis, providing a high-quality data foundation for subsequent digital twin model updates and dynamic predictive analysis. The time synchronization processing mentioned above is for multi-source running data with different sampling frequencies. Mapped to a unified time series through time resampling Its relational expression is: In the formula, For the first The class is mapped to runtime data in a unified time series. For the first Time alignment operator for class-based runtime data; At a certain moment on a unified timeline When no corresponding sample value is available, the runtime data at that moment is calculated using interpolation. : In the formula, , For adjacent times where sampled values ​​exist, ; , They are time points , The running data.

[0023] The conditional variance of the interpolated running data is calculated to characterize the uncertainty. : In the formula, This is the function for calculating conditional variance. For a moment Operational data; A quality score is constructed based on the aforementioned uncertainty and alignment consistency. : In the formula: , These are the weighting coefficients. For reference operation data; The quality scores are normalized to obtain fusion weights, and these weights are then used to perform quality-gated fusion of various operational data to obtain a unified input for digital twin modeling and updating. In the formula, For the unified input after quality gating fusion, To normalize the fusion weights, For the number of data types running from multiple sources, For the first Class running data at time Quality rating It is a stable term.

[0024] By combining the online correction process of model parameters with the dynamic evolution process of performance degradation, the overall state of the digital twin model is jointly updated to achieve the continuous dynamic evolution of the digital twin model as the gas turbine gas circuit system operates. The digital twin modeling and updating module is used to construct a digital twin virtual model based on the structural information of the gas turbine gas circuit system, and to perform online correction and dynamic evolution of the virtual model according to the operating data, so as to realize the mapping between the virtual model and the actual operating state of the gas turbine gas circuit system; The digital twin modeling and updating module performs online correction and dynamic evolution of the virtual model based on the operational data, wherein the online correction is based on a unified time series. On running data Calculate the digital twin model at time 1 Output and with actual operating data To make comparisons, construct a digital twin model based on parameters. An optimization problem with model bias as the objective function and independent variable as the objective variable. Represented as: In the formula, For the parameter vector of the digital twin model, It is a digital twin model at any time And the parameters are The predicted output at that time It is a moment The actual running data output vector, For a unified time series set; Considering physical consistency constraints, the objective function is minimized using a Bayesian optimization algorithm. To enable online correction of parameters in digital twin models: In the formula, The weighting coefficients for physical consistency constraints. This is the physical consistency residual function; Online correction of synchronous output parameter uncertainty and recursive update of parameter covariance: In the formula, For a moment The parameter vector of the digital twin model, For a moment Update the model parameter vector. For the parameter gain matrix, For digital twin models with current parameters The predicted output is as follows. It is the identity matrix. For a moment The parameter covariance matrix, For a moment The updated parameter covariance matrix, For a moment The parameter covariance matrix, For process noise covariance; Dynamic evolution, based on online correction of model parameters, models the performance degradation process of the gas turbine gas path system. The key performance indicators of the gas path system are defined as follows: Its dynamic evolution relationship is expressed as: In the formula, For degenerate evolution function, It is a forward time series; By combining the online correction process of model parameters with the dynamic evolution process of performance degradation, the overall state of the digital twin model is jointly updated, realizing the continuous dynamic evolution of the digital twin model as the gas turbine gas path system operates.

[0025] The dynamic prediction and simulation analysis module is used to perform dynamic operation state prediction and multi-condition simulation analysis of the gas turbine gas circuit system using the updated digital twin virtual model, and to obtain the future operation trend and potential anomaly information of the system. The dynamic prediction and simulation analysis module's dynamic operating state prediction and multi-condition simulation analysis process lies in the fact that, based on the digital twin model, the state variables at the current moment... Model parameters Key performance indicators By extrapolating the system state, future moments can be obtained. Predicted running status : In the formula To predict the time step, For running state prediction function; Furthermore, by changing the model input parameters or operating boundary conditions under different operating conditions, multi-condition simulation analysis was conducted on the digital twin model to obtain the predicted operating state results under different operating conditions. : In the formula, The number of simulation conditions. For the first Simulation results under various operating conditions; Weights are assigned to each operating condition and updated recursively to obtain the probability weights of the operating conditions at future times. : In the formula, For working conditions To working conditions The transition probability, For the future The Each working condition probability weight; Weighted fusion of multi-condition prediction results based on operating condition probability weights: Propagating the parameter uncertainty to the prediction output yields the prediction covariance and confidence interval: In the formula, For a moment The predicted covariance matrix, For the covariance of operating condition uncertainties, For a moment Jacobian matrix, For a moment At confidence level The confidence interval below, , is the confidence coefficient This indicates taking the diagonal elements of the covariance matrix; Based on the predicted operating status at future moments and the results of uncertainty characterization, the operating trend and potential abnormal states of the gas turbine gas circuit system are analyzed.

[0026] The intelligent diagnosis and operation and maintenance decision module is used to realize intelligent operation and maintenance decisions for fault diagnosis and early warning of gas turbine gas circuit system and prediction of remaining service life based on the dynamic prediction and simulation analysis results. The intelligent diagnosis and maintenance decision-making module identifies the fault type and location of the gas turbine gas circuit system by comparing and analyzing the dynamic operating data of the digital twin model with the health status. Based on the predicted operating status change trend and fault severity, it assesses the operational risk of the gas turbine gas circuit system, determines the corresponding risk level, and generates a fault warning when the risk level exceeds a preset threshold. The remaining service life prediction combines the dynamic evolution trend of key performance indicators with the predicted dynamic operating status at future times to determine the time when the key performance indicators reach the preset failure threshold, thereby obtaining the remaining service life of the key components. Based on the fault diagnosis results, fault warning information, and remaining service life prediction results, the intelligent diagnosis and maintenance decision-making module comprehensively evaluates the operating status of the gas turbine gas circuit system and generates corresponding intelligent maintenance decisions, including: operating parameter adjustment suggestions based on fault type and risk level; maintenance timing suggestions determined by the remaining service life prediction results; maintenance strategy suggestions based on the degree of degradation of key components; and component replacement suggestions based on the failure risk assessment results.

[0027] The visualization module is used to provide real-time visualization of the operating status, prediction results, and maintenance decision information of the gas turbine gas path system. Specifically, based on real-time operating data acquired by the data acquisition and management module, and analysis results output by the digital twin modeling and updating module, dynamic prediction and simulation analysis module, and intelligent diagnosis and maintenance decision module, the overall structure of the gas turbine gas path system and the operating status of each key gas path component are visualized. The visualization module centrally displays the real-time operating parameters, predicted operating status, fault diagnosis and early warning information, remaining service life prediction results, and intelligent operation and maintenance decision information of the gas turbine gas circuit system in a graphical manner. It intuitively reflects the operating status and health level of the gas circuit system in the form of curves, dashboards, status indicators, or three-dimensional structural diagrams, so that operation and maintenance personnel can quickly and comprehensively grasp the operating status of the gas turbine gas circuit system, providing intuitive and effective support for operation monitoring, fault analysis, and operation and maintenance decision-making.

[0028] After the actual operating data of the gas turbine gas path system is connected to the intelligent operation and maintenance platform described in this invention, the data acquisition and management module, the digital twin modeling and updating module, the dynamic prediction and simulation analysis module, and the intelligent diagnosis and operation and maintenance decision-making module are activated sequentially. This allows each functional module to operate collaboratively, achieving real-time monitoring and analysis of the gas turbine gas path system's operating status. The interactive visualization interface of the digital twin platform is as follows: Figure 2 As shown, this interface provides a clear view of the overall structure, real-time operating status, and related analysis results of the gas turbine gas circuit system.

[0029] The platform also provides simulation parameter adjustment functionality. Users can set simulation conditions and model input parameters through the parameter adjustment interface to conduct simulation analysis and prediction under different operating conditions. A diagram of the simulation parameter adjustment interface is shown below. Figure 3 As shown above, users can intuitively observe the impact of different parameter settings on the operating status and prediction results of the gas circuit system, thereby providing intuitive and effective technical support for the operation management and maintenance decisions of the gas turbine gas circuit system.

[0030] In this specification, the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the descriptions of the embodiments described later are relatively simple, and relevant parts can be referred to the descriptions of the foregoing embodiments.

[0031] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A smart operation and maintenance platform for gas turbine gas circuit systems based on digital twins, characterized in that, It includes a data acquisition and management module, a digital twin modeling and updating module, a dynamic prediction and simulation analysis module, an intelligent diagnosis and operation and maintenance decision-making module, and a visualization display module, which are connected in sequence. The data acquisition and management module is used to collect, preprocess, and store the operating data of various components in the gas turbine gas circuit system in real time. The digital twin modeling and updating module is used to construct a digital twin virtual model based on the structural information of the gas turbine gas circuit system, and to perform online correction and dynamic evolution of the virtual model according to the operating data, so as to realize the mapping between the virtual model and the actual operating state of the gas turbine gas circuit system; The dynamic prediction and simulation analysis module is used to perform dynamic operation state prediction and multi-condition simulation analysis of the gas turbine gas circuit system using the updated digital twin virtual model, and to obtain the future operation trend and potential anomaly information of the system. The intelligent diagnosis and operation and maintenance decision module is used to realize intelligent operation and maintenance decisions for fault diagnosis and early warning of gas turbine gas circuit system and prediction of remaining service life based on the dynamic prediction and simulation analysis results. The visualization module is used to provide real-time visualization of the operating status, prediction results, and maintenance decision information of the gas turbine gas circuit system.

2. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 1, characterized in that, Before storing the data, the data acquisition and management module performs noise reduction, abnormal data identification and removal, missing data compensation and data standardization on the acquired operational data, and performs time synchronization processing on operational data with different sampling frequencies to construct a data sequence with a unified time axis.

3. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 2, characterized in that, The time synchronization processing mentioned above is for multi-source running data with different sampling frequencies. Mapped to a unified time series through time resampling Its relational expression is: In the formula, For the first The class is mapped to runtime data in a unified time series. For the first Time alignment operator for class-based runtime data; At a certain moment on a unified timeline When no corresponding sample value is available, the runtime data at that moment is calculated using interpolation. : In the formula, , For adjacent times where sampled values ​​exist, ; , They are time points , The running data.

4. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 3, characterized in that, The conditional variance of the interpolated running data is calculated to characterize the uncertainty. : In the formula, This is the function for calculating conditional variance. For a moment Operational data; A quality score is constructed based on the aforementioned uncertainty and alignment consistency. : In the formula: , These are the weighting coefficients. For reference operation data; The quality scores are normalized to obtain fusion weights, and these weights are then used to perform quality-gated fusion of various operational data to obtain a unified input for digital twin modeling and updating. In the formula, For the unified input after quality gating fusion, To normalize the fusion weights, For the number of data types running from multiple sources, For the first Class running data at time Quality rating It is a stable term.

5. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 4, characterized in that, The digital twin modeling and updating module performs online correction and dynamic evolution of the virtual model based on the operational data, wherein the online correction is based on a unified time series. On running data Calculate the digital twin model at time 1 Output and with actual operating data To make comparisons, construct a digital twin model based on parameters. An optimization problem with model bias as the objective function and independent variable as the objective variable. Represented as: In the formula, It is a digital twin model at any time And the parameters are The predicted output at that time It is a moment The actual running data output vector, For a unified time series set; The objective function is minimized using a Bayesian optimization algorithm. To enable online correction of parameters in digital twin models: In the formula, The weighting coefficients for physical consistency constraints. This is the physical consistency residual function; Online correction of synchronous output parameter uncertainty and recursive update of parameter covariance: In the formula, For a moment The parameter vector of the digital twin model, For a moment Update the model parameter vector. For the parameter gain matrix, For digital twin models with current parameters The predicted output is as follows. It is the identity matrix. For a moment The parameter covariance matrix, For a moment The updated parameter covariance matrix, For a moment The parameter covariance matrix, Let be the process noise covariance.

6. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 5, characterized in that, Dynamic evolution, based on online correction of model parameters, models the performance degradation process of the gas turbine gas path system. The key performance indicators of the gas path system are defined as follows: Its dynamic evolution relationship is expressed as: In the formula, For degenerate evolution function, It is a forward time series; By combining the online correction process of model parameters with the dynamic evolution process of performance degradation, the overall state of the digital twin model is jointly updated, realizing the continuous dynamic evolution of the digital twin model as the gas turbine gas path system operates.

7. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 6, wherein the dynamic operation state prediction and multi-condition simulation analysis process of the dynamic prediction and simulation analysis module is based on the state variables of the digital twin model at the current moment. Model parameters Key performance indicators By extrapolating the system state, future moments can be obtained. Predicted running status : In the formula To predict the time step, For running state prediction function; Furthermore, by changing the model input parameters or operating boundary conditions under different operating conditions, multi-condition simulation analysis was conducted on the digital twin model to obtain the predicted operating state results under different operating conditions. : In the formula, The number of simulation conditions. For the first Simulation results under various operating conditions; Weights are assigned to each operating condition and updated recursively to obtain the probability weights of the operating conditions at future times. : In the formula, For working conditions To working conditions The transition probability, For the future The Each working condition probability weight; Weighted fusion of multi-condition prediction results based on operating condition probability weights: Propagating the parameter uncertainty to the prediction output yields the prediction covariance and confidence interval: In the formula, For a moment The predicted covariance matrix, For the covariance of operating condition uncertainties, For a moment Jacobian matrix, For a moment At confidence level The confidence interval below, Here is the confidence coefficient. This indicates taking the diagonal elements of the covariance matrix; Based on the predicted operating status at future moments and the results of uncertainty characterization, the operating trend and potential abnormal states of the gas turbine gas circuit system are analyzed.

8. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 7, characterized in that, The intelligent diagnosis and operation and maintenance decision-making module identifies the fault type and location of the gas turbine gas circuit system by comparing and analyzing the dynamic operating data of the digital twin model with the health status; and assesses the operating risk of the gas turbine gas circuit system based on the predicted operating status change trend and fault degree, determines the corresponding risk level, and generates fault warning information when the risk level exceeds a preset threshold.

9. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 8, characterized in that, The remaining service life prediction is obtained by combining the dynamic evolution trend of key performance indicators with the prediction results of dynamic operating status at future times to determine the time when the key performance indicators reach the preset failure threshold, thereby obtaining the remaining service life of the key components.

10. The intelligent operation and maintenance platform for gas turbine gas circuit system based on digital twin as described in claim 9, characterized in that, The intelligent diagnosis and operation and maintenance decision module comprehensively evaluates the operating status of the gas turbine gas circuit system based on fault diagnosis results, fault early warning information, and remaining service life prediction results, and generates corresponding intelligent operation and maintenance decisions, including: operating parameter adjustment suggestions generated according to fault type and risk level, maintenance timing suggestions determined according to remaining service life prediction results, maintenance strategy suggestions generated according to the degree of degradation of key components, and component replacement suggestions generated according to failure risk assessment results.

Images