Turbine deviation intelligent identification and tracing method and system based on thermal performance index

Abstract

The invention relates to the technical field of operation monitoring of thermal power generation equipment, in particular to a steam turbine deviation intelligent identification and tracing method and system based on thermal performance indexes. The method comprises the following steps: collecting various thermal performance indexes in the running process of a steam turbine, and carrying out standardization processing on the collected data to form a unified feature vector; during anomaly detection, a data correction algorithm is adopted to compensate missing or abnormal data; based on the reference value after working condition adaptability adjustment, deviation detection is carried out on the standardized feature vector, and a deviation severity score is calculated; the deviation characteristics are input to a deviation mode recognition module, and possible deviation modes are matched; and inferring a fault reason through the causal inference rule base, and generating a health assessment report according to an inference result. The method can realize the intelligent identification and fault tracing of the operation performance deviation of the steam turbine, improves the diagnosis accuracy and the operation maintenance efficiency, and is suitable for the operation state evaluation and maintenance decision support of various steam turbine sets.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal power generation equipment operation monitoring and analysis, and in particular to a method and system for intelligently identifying and tracing steam turbine deviations based on thermal performance indicators. Background Art

[0002] As a key component in thermal power generation systems, the thermal performance of a steam turbine directly impacts the overall economic efficiency and operational safety of the unit. To understand the actual operating status of a steam turbine, regular or specific performance tests are typically conducted. By collecting various operating data, multiple thermal performance indicators such as the unit's total heat rate, cylinder efficiency, pipeline pressure loss, heater end differential, and extraction and exhaust parameters are calculated.

[0003] Existing performance test analysis methods primarily rely on manual data processing and performance indicator calculations, as well as manual comparisons to design parameters or historical benchmark data to determine whether the turbine is experiencing operational deviations. During this process, engineers must use their experience to determine the correlations between various indicators and comprehensively analyze potential causes of performance degradation, such as blade fouling, seal wear, increased flow resistance, and decreased heater efficiency. Due to the complexity of the thermal system, the numerous measurement points, and the strong correlations among parameters, this analysis process is not only time-consuming and labor-intensive, but also highly dependent on the engineer's experience, subjecting them to a degree of subjectivity and uncertainty.

[0004] On the other hand, as unit operation time increases and operating conditions change, the quality of data collected during performance tests is often affected by environmental factors, equipment aging, and insufficient measurement point accuracy. This can lead to abnormalities, missing data, or significant fluctuations in some data. Existing technologies lack effective mechanisms for automatically identifying and correcting data anomalies, which can easily introduce errors into subsequent analysis and affect the accuracy of deviation diagnosis.

[0005] Furthermore, most existing performance deviation analysis methods rely solely on preliminary inferences based on changes in a single indicator. They lack systematic exploration and comprehensive assessment of the complex relationships between multiple thermal performance indicators, making it difficult to intelligently reason and accurately trace potential turbine failure mechanisms. This limitation often necessitates multiple tests and extensive disassembly and inspections during actual operation and maintenance, increasing maintenance costs and the risk of unit downtime.

[0006] Therefore, there is an urgent need for a new method based on comprehensive analysis of multiple thermal performance indicators, which can operate robustly in data anomaly environments and has the capabilities of intelligent deviation identification, causal reasoning and fault tracing, so as to improve the accuracy and intelligence level of turbine performance diagnosis and meet the requirements of the modern power industry for efficient and reliable operation of equipment. Summary of the Invention

[0007] The purpose of the present invention is to overcome the shortcomings of the existing technology in which turbine performance tests rely on manual analysis, resulting in low diagnostic efficiency and poor accuracy, as well as insufficient data anomaly processing capabilities and a lack of a systematic deviation identification and tracing mechanism. A method and system for intelligent identification and tracing of turbine deviations based on thermal performance indicators are proposed, which is suitable for operating performance evaluation, deviation diagnosis and fault tracing analysis of turbine units in various thermal power plants.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] The intelligent identification and tracing method of steam turbine deviation based on thermal performance indicators includes the following steps:

[0010] Step S1: Collecting steam turbine performance test data and calculating thermal performance indicators, including total heat rate, cylinder efficiency, pipeline pressure loss, heater terminal difference, and extraction and exhaust parameters;

[0011] Step S2: normalizing the thermal performance index to form a feature vector in a unified format, and applying a data anomaly correction algorithm to correct the abnormal data when anomalies are detected in the test data;

[0012] Step S3: Inputting the characteristic vector into the deviation detection module, identifying the deviation index and calculating the corresponding deviation severity score by comparing it with the baseline value adjusted for working condition adaptability, wherein the deviation severity score is used to characterize the degree of deviation;

[0013] Step S4: Inputting the identified deviation indicators and their deviation severity scores into a deviation pattern recognition module, which classifies the deviation feature vectors based on a pre-trained machine learning model, matches possible deviation patterns, and generates a description of the current deviation situation, including the deviation type, associated indicators, and severity score information;

[0014] Step S5: Input the deviation pattern into the causal reasoning module, perform fault source inference based on a preset causal reasoning rule library, and output an interpretable reasoning path, which is used to illustrate the logical chain from the deviation feature to the root cause of the fault;

[0015] Step S6: predicting the future thermal performance degradation trend of the unit based on the current deviation description and historical performance trend data, wherein the thermal performance trend data includes an archived historical thermal performance indicator sequence;

[0016] Step S7: Generate a unit health assessment report based on the future performance degradation trend. The health assessment report includes thermal performance deviation analysis results, suspected fault sources, reasoning paths, and performance trend prediction results.

[0017] In this embodiment, the data anomaly correction algorithm includes one or more of interpolation correction, correction based on thermodynamic balance estimation, or correction based on empirical rules.

[0018] In this embodiment, the operating condition adaptability adjustment is based on the reference range of the dynamic adjustment index of the actual load rate of the unit.

[0019] In this embodiment, the deviation detection module identifies abnormal indicators based on statistical methods and machine learning anomaly detection algorithms.

[0020] In this embodiment, the deviation severity score is obtained by comprehensively calculating the deviation amplitude and the historical stability parameter.

[0021] In this embodiment, the deviation pattern recognition module adopts an ensemble learning algorithm, including at least one of a random forest, a gradient boosting tree, and an extreme gradient boosting tree.

[0022] In this embodiment, the causal reasoning rule base includes a rule chain constructed based on expert knowledge and an automated reasoning chain based on data-driven mining.

[0023] In this embodiment, the performance trend prediction is based on a time series prediction model, including an autoregressive moving average model or a long short-term memory network.

[0024] The intelligent identification and tracing system for steam turbine deviations based on thermal performance indicators includes:

[0025] Data acquisition and calculation unit, which collects turbine performance test data and calculates thermal performance indicators, including total heat rate, cylinder efficiency, pipeline pressure loss, heater terminal difference, and extraction and exhaust parameters;

[0026] a standardization processing unit for performing standardization processing on the thermal performance index to form a feature vector in a unified format, and applying a data anomaly correction algorithm to correct the abnormal data when anomalies are detected in the test data;

[0027] a reference value comparison unit, which inputs the feature vector into a deviation detection module, compares it with the reference value after the working condition adaptability adjustment, identifies the deviation index and calculates the corresponding deviation severity score, wherein the deviation severity score is used to represent the degree of deviation;

[0028] A deviation situation description generation unit inputs the identified deviation indicators and their deviation severity scores into a deviation pattern recognition module. The deviation pattern recognition module classifies deviation feature vectors based on a pre-trained machine learning model, matches possible deviation patterns, and generates a current deviation situation description, which includes deviation type, associated indicators, and severity score information;

[0029] A fault source inference unit inputs the deviation pattern into a causal reasoning module, performs fault source inference based on a preset causal reasoning rule library, and outputs an interpretable reasoning path that illustrates the logical chain from the deviation feature to the root cause of the fault;

[0030] a prediction unit, configured to predict a future thermal performance degradation trend of the unit based on the current deviation description and historical performance trend data, wherein the thermal performance trend data includes an archived historical thermal performance indicator sequence;

[0031] The health assessment report generation unit generates a unit health assessment report based on future performance degradation trends. The health assessment report includes thermal performance deviation analysis results, suspected fault sources, reasoning paths, and performance trend prediction results.

[0032] In this embodiment, the data anomaly correction algorithm in the standardization processing unit includes one or more of interpolation correction, correction based on thermodynamic balance estimation, or correction based on empirical rules.

[0033] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0034] The method and system for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators provided by the present invention can realize comprehensive deviation identification based on multiple thermal performance indicators, thereby improving the accuracy and sensitivity of unit performance deviation detection; in the case of abnormalities or missing collected data, an intelligent correction algorithm is used to ensure the continuity and reliability of the analysis process; a deviation severity scoring system is introduced to quantitatively evaluate the degree of performance deviation, providing a reference basis for maintenance decisions; through the combination of pattern recognition and causal reasoning, intelligent inference of the cause of the deviation is achieved, fault location time is shortened, and dependence on manual experience is reduced; health trend prediction based on historical operating data can assist in the early detection of potential performance degradation problems, thereby improving the safety and economy of unit operation; the method has good adaptability, and can dynamically adjust the index reference standard according to the changes in the operating conditions of different units, and is suitable for steam turbine generator sets of different types and sizes. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 Schematic diagram of the overall process of the method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators of the present invention;

[0037] Figure 2 This is a schematic diagram of the structure of the thermal performance index collection and standardization processing module in the present invention;

[0038] Figure 3 Schematic diagram of the structure of the deviation detection and deviation severity assessment module in the present invention;

[0039] Figure 4 This is a schematic diagram of the structure of the deviation pattern recognition and causal reasoning collaborative inference module of the present invention;

[0040] Figure 5 This is a schematic diagram of the structure of the health assessment report generation module in the present invention;

[0041] Figure 6 This is a flow chart showing an example of application of tracing the source of steam turbine shaft seal leakage deviation based on thermal performance indicators in the present invention;

[0042] Figure 7 This is a flowchart of an application example of tracing the source of turbine blade fouling deviation based on thermal performance indicators in the present invention;

[0043] Figure 8 This is a structural block diagram of the steam turbine deviation intelligent identification and tracing system based on thermal performance indicators in the present invention. DETAILED DESCRIPTION

[0044] Hereinafter, only certain exemplary embodiments are briefly described. As will be appreciated by those skilled in the art, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and description are to be considered as illustrative in nature and not restrictive.

[0045] In the description of the present invention, when used in this specification and the appended claims, the terms "include" and "comprising" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0046] It should also be understood that the terms used in the present specification are only for the purpose of describing particular embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0047] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0048] The accompanying drawings illustrate various schematic diagrams of structures according to embodiments disclosed herein. These figures are not drawn to scale; for clarity, some details are exaggerated and some details may be omitted. The shapes of the various regions and layers shown in the figures, as well as their relative sizes and positional relationships, are merely exemplary and may deviate in practice due to manufacturing tolerances or technical limitations. Those skilled in the art may design regions / layers with different shapes, sizes, and relative positions as needed.

[0049] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] Example 1

[0051] The method for intelligently identifying and tracing steam turbine deviations based on thermal performance indicators provided by the present invention comprises the following steps:

[0052] Step S1: Collecting steam turbine performance test data and calculating thermal performance indicators, including but not limited to total heat rate, cylinder efficiency, pipeline pressure loss, heater terminal difference, and extraction and exhaust parameters;

[0053] Step S2: normalizing the thermal performance index to form a feature vector in a unified format, and applying a data anomaly correction algorithm to correct the abnormal data when anomalies are detected in the test data;

[0054] Step S3: Inputting the characteristic vector into the deviation detection module, identifying the deviation index and calculating the corresponding deviation severity score by comparing it with the baseline value adjusted for working condition adaptability, wherein the deviation severity score is used to characterize the degree of deviation;

[0055] Step S4: Inputting the identified deviation indicators and their deviation severity scores into a deviation pattern recognition module, which classifies the deviation feature vectors based on a pre-trained machine learning model and matches possible deviation patterns;

[0056] Step S5: Input the deviation pattern into the causal reasoning module, perform fault source inference based on a preset causal reasoning rule library, and output an interpretable reasoning path, which is used to illustrate the logical chain from the deviation feature to the root cause of the fault;

[0057] Step S6: predicting the future thermal performance degradation trend of the unit based on the current deviation description and historical performance trend data, wherein the performance trend data includes archived historical thermal performance indicator sequences;

[0058] Step S7: Generate a unit health assessment report, which includes thermal performance deviation analysis results, suspected fault sources, reasoning paths, and performance trend prediction results.

[0059] In this embodiment, the data anomaly correction algorithm includes one or more of interpolation correction, correction based on thermodynamic balance estimation, or correction based on empirical rules.

[0060] In this embodiment, the operating condition adaptability adjustment is based on the reference range of the dynamic adjustment index of the actual load rate of the unit.

[0061] In this embodiment, the deviation detection module identifies abnormal indicators based on statistical methods and machine learning anomaly detection algorithms.

[0062] In this embodiment, the deviation severity score is obtained by comprehensively calculating the deviation amplitude and the historical stability parameter.

[0063] In this embodiment, the deviation pattern recognition module adopts an ensemble learning algorithm, including at least one of a random forest, a gradient boosting tree, and an extreme gradient boosting tree.

[0064] In this embodiment, the causal reasoning rule base includes a rule chain constructed based on expert knowledge and an automated reasoning chain based on data-driven mining.

[0065] In this embodiment, the performance trend prediction is based on a time series prediction model, including an autoregressive moving average model (ARIMA) or a long short-term memory network (LSTM).

[0066] Example 2

[0067] This embodiment provides a method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators, which analyzes and processes thermal performance data during the operation of steam turbines. This method is carried out in accordance with the steps of data collection, indicator standardization, deviation detection, pattern recognition, causal reasoning and report generation. Figure 1 As shown in the figure, the intelligent identification and tracing method of steam turbine deviation based on thermal performance indicators proposed in the present invention includes key steps such as data acquisition, indicator standardization, deviation detection, pattern recognition, causal reasoning and report generation, forming a complete deviation diagnosis and trend assessment process.

[0068] (1) Data collection

[0069] Key steam turbine operating parameters are collected, including main steam pressure, main steam temperature, reheat steam pressure, exhaust pressure, exhaust flow, condenser vacuum, generator output power, and extraction flow. These operating parameters, along with environmental conditions (such as ambient temperature and humidity), are entered into a database to generate a rich historical operating data set, providing the raw data foundation for subsequent analysis.

[0070] (2) Standardization of indicators

[0071] The collected raw data is pre-processed, including denoising, filtering, missing value interpolation and outlier removal. Then the key thermal performance indicators are standardized. Figure 2 As shown in Figure 1, the thermal performance index collection and standardization processing module mainly includes a data preprocessing unit, an abnormality correction unit, and a standardization calculation unit. It can convert multi-source heterogeneous raw data into a unified standard feature vector and provide input for subsequent deviation detection. Taking unit heat consumption HR as an example, it can be calculated based on the main steam and re-steam flow rates F ms 、F crh , main steam and feed water enthalpy h ms 、h fw , cold reheat and hot reheat steam enthalpy h crh 、h hrh , and reheating water flow F rhsp and enthalpy h rhsp And the generator output P is used to calculate the unit heat consumption:

[0072]

[0073] The historical mean μ and standard deviation σ of this indicator and other indicators (such as thermal efficiency η, vacuum degree V, etc.) are calculated, and z-score standardization is performed using the following formula:

[0074]

[0075] Where I is the original indicator value, and I' is the normalized indicator value. Normalization eliminates dimensionality effects and facilitates comparison of multiple indicators. For indicators significantly affected by load and environment, a load-based regression correction method can be used to convert the indicator value under actual operating conditions to rated operating conditions or standard reference conditions to obtain an equivalent indicator value.

[0076] (3) Deviation detection

[0077] On the basis of standardized indicators, deviation analysis is performed on each performance indicator. The standardized indicator value I' actual Compared with the reference value (design value or historical average value) I' ref The difference between ΔI=I' actual -I' ref Calculate. Using the statistical threshold judgment method, when ΔI exceeds the set threshold (such as the 3σ principle), it is determined that the indicator has a significant deviation. You can also define the comprehensive deviation ΔI=(ΔI1, ΔI2, ..., ΔI n ), calculate its Euclidean norm And compare with the threshold to determine the overall deviation. Figure 3As shown, the deviation detection and assessment module includes an indicator deviation calculation unit, an operating condition adjustment module, and a severity score calculation unit. It can perform real-time comparisons of standardized indicators and quantify the degree of deviation, providing basic input for deviation pattern identification. If one or more key indicator deviations are detected outside the normal range, the system is considered to have a performance deviation and enters the fault pattern identification phase.

[0078] (4) Pattern recognition

[0079] According to the detected index deviation characteristics, matching analysis is performed in the pre-established fault mode library. First, the index deviations ΔI1, ΔI2, ..., ΔI n Combined into a deviation feature vector Δ=(ΔI1, ΔI2, ..., ΔI n The fault mode library stores typical deviation signatures M corresponding to common faults (such as shaft seal leakage, blade fouling, valve leakage, etc.). j =(M j,1 ,M j,2 ,...,M j,n ). By calculating Δ and each mode M j The similarity between them can be used for classification and identification, for example, the Euclidean distance metric can be used: The fault type corresponding to the pattern with the smallest distance is taken as the current recognition result. During the recognition process, machine learning algorithms such as support vector machines, decision trees, and neural networks can also be used to classify the deviation vectors to improve recognition accuracy.

[0080] (5) Causal reasoning

[0081] The specific cause of the identified fault pattern can be further inferred. Probability-based reasoning methods or expert rules can be used for judgment. For example, a Bayesian network of faults and causes can be established to model the conditional probability relationship between the fault pattern and possible causes. The posterior probability of a cause c when a deviation Δ is observed can be calculated using Bayes' theorem:

[0082]

[0083] Where P(Δ|c) can be estimated using historical data or models, P(c) is the prior probability of cause c, and P(Δ) is a normalization constant. The most likely cause is determined by comparing the posterior probabilities of candidate causes. Furthermore, expert experience rules based on failure modes can be incorporated into the judgment. For example, shaft seal leakage is often accompanied by increased vacuum and decreased unit power; while blade fouling typically leads to decreased thermal efficiency while maintaining minimal changes in condenser vacuum. By comparing these characteristics, the inference results can be further verified and confirmed.

[0084] like Figure 4As shown in the figure, the deviation pattern recognition and causal reasoning modules work together. The deviation recognition model first matches the fault pattern, then infers the possible fault source through the causal reasoning rule library, and generates an explainable reasoning path to achieve traceability analysis of complex deviation problems.

[0085] (6) Report generation

[0086] A diagnostic report is automatically generated based on the above analysis process. The report lists the current value, reference value, and deviation of each performance indicator, and explains the matched failure mode and the most likely cause inferred. The report uses a standardized template and includes deviation detection results, identified failure types, cause inference conclusions, and maintenance recommendations, providing a reference for decision-making by operation and maintenance personnel. Figure 5 As shown in the figure, the health assessment report generation module integrates performance deviation results, fault source inference conclusions and trend prediction results, and generates a health report using a structured template, which makes it easy for power plant operation and maintenance personnel to quickly obtain the equipment operation status.

[0087] like Figure 6 As shown in the figure, in a certain test, abnormalities in the exhaust vacuum and low-pressure cylinder efficiency were detected. Through deviation identification and causal reasoning, the typical shaft seal leakage pattern was identified, and the fault location and reasoning path analysis were completed, verifying the effectiveness of this method.

[0088] like Figure 7 As shown in the figure, in another set of data, the increase in heat rate and the change in final stage vacuum were not significant. The system identified a suspected blade fouling deviation pattern and located the fault as a decrease in heat transfer efficiency caused by increased flow resistance through model matching and causal reasoning.

[0089] Example 3

[0090] like Figure 8 As shown, the steam turbine deviation intelligent identification and tracing system based on thermal performance indicators provided by the present invention includes:

[0091] Data acquisition and calculation unit, which collects turbine performance test data and calculates thermal performance indicators, including total heat rate, cylinder efficiency, pipeline pressure loss, heater terminal difference, and extraction and exhaust parameters;

[0092] a standardization processing unit for performing standardization processing on the thermal performance index to form a feature vector in a unified format, and applying a data anomaly correction algorithm to correct the abnormal data when anomalies are detected in the test data;

[0093] a reference value comparison unit, which inputs the feature vector into a deviation detection module, compares it with the reference value after the working condition adaptability adjustment, identifies the deviation index and calculates the corresponding deviation severity score, wherein the deviation severity score is used to represent the degree of deviation;

[0094] A deviation situation description generation unit inputs the identified deviation indicators and their deviation severity scores into a deviation pattern recognition module. The deviation pattern recognition module classifies deviation feature vectors based on a pre-trained machine learning model, matches possible deviation patterns, and generates a current deviation situation description, which includes deviation type, associated indicators, and severity score information;

[0095] A fault source inference unit inputs the deviation pattern into a causal reasoning module, performs fault source inference based on a preset causal reasoning rule library, and outputs an interpretable reasoning path that illustrates the logical chain from the deviation feature to the root cause of the fault;

[0096] a prediction unit, configured to predict a future thermal performance degradation trend of the unit based on the current deviation description and historical performance trend data, wherein the thermal performance trend data includes an archived historical thermal performance indicator sequence;

[0097] The health assessment report generation unit generates a unit health assessment report based on future performance degradation trends. The health assessment report includes thermal performance deviation analysis results, suspected fault sources, reasoning paths, and performance trend prediction results.

[0098] In this embodiment, the data anomaly correction algorithm in the standardization processing unit includes one or more of interpolation correction, correction based on thermodynamic balance estimation, or correction based on empirical rules.

[0099] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, from all points of view, the embodiments should be regarded as illustrative and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes that fall within the meaning and range of equivalents of the claims are included in the present invention. Any reference signs in the claims should not be construed as limiting the claim to which they relate.

[0100] In addition, it should be understood that although this specification describes the embodiments, not every embodiment contains only one independent technical solution. This description is for clarity only. Those skilled in the art should consider the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art. The above content is only for the purpose of illustrating the technical concept of the present invention and cannot be used to limit the scope of protection of the present invention. Any changes made based on the technical solution in accordance with the technical concept proposed by the present invention fall within the scope of protection of the claims of the present invention.

Claims

1. The intelligent identification and tracing method of steam turbine deviation based on thermal performance indicators is characterized by: The steps include: Step S1: Collecting steam turbine performance test data and calculating thermal performance indicators, including total heat rate, cylinder efficiency, pipeline pressure loss, heater terminal difference, and extraction and exhaust parameters; Step S2: normalizing the thermal performance index to form a feature vector in a unified format, and applying a data anomaly correction algorithm to correct the abnormal data when anomalies are detected in the test data; Step S3: Inputting the characteristic vector into the deviation detection module, identifying the deviation index and calculating the corresponding deviation severity score by comparing it with the baseline value adjusted for working condition adaptability, wherein the deviation severity score is used to characterize the degree of deviation; Step S4: Inputting the identified deviation indicators and their deviation severity scores into a deviation pattern recognition module, which classifies the deviation feature vectors based on a pre-trained machine learning model, matches possible deviation patterns, and generates a description of the current deviation situation, including the deviation type, associated indicators, and severity score information; Step S5: Input the deviation pattern into the causal reasoning module, perform fault source inference based on a preset causal reasoning rule library, and output an interpretable reasoning path, which is used to illustrate the logical chain from the deviation feature to the root cause of the fault; Step S6: predicting the future thermal performance degradation trend of the unit based on the current deviation description and historical performance trend data, wherein the thermal performance trend data includes an archived historical thermal performance indicator sequence; Step S7: Generate a unit health assessment report based on the future performance degradation trend. The health assessment report includes thermal performance deviation analysis results, suspected fault sources, reasoning paths, and performance trend prediction results.

2. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: The data anomaly correction algorithm includes one or more of interpolation correction, correction based on thermodynamic balance estimation, or correction based on empirical rules.

3. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: The operating condition adaptability adjustment is based on the reference range of the dynamic adjustment index of the actual load rate of the unit.

4. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: The deviation detection module identifies anomaly indicators based on a combination of statistical methods and machine learning anomaly detection algorithms.

5. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: The deviation severity score is calculated by combining the deviation magnitude and the historical stability parameter.

6. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: The deviation pattern recognition module adopts an ensemble learning algorithm, including at least one of a random forest, a gradient boosting tree, and an extreme gradient boosting tree.

7. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: The causal reasoning rule base includes rule chains built based on expert knowledge and automated reasoning chains based on data-driven mining.

8. The method for intelligent identification and tracing of steam turbine deviations based on thermal performance indicators according to claim 1 is characterized in that: Performance trend prediction is based on time series forecasting models, including autoregressive moving average models or long short-term memory networks.

9. The intelligent identification and tracing system of steam turbine deviation based on thermal performance indicators is characterized by: include: Data acquisition and calculation unit, which collects turbine performance test data and calculates thermal performance indicators, including total heat rate, cylinder efficiency, pipeline pressure loss, heater terminal difference, and extraction and exhaust parameters; a standardization processing unit for performing standardization processing on the thermal performance index to form a feature vector in a unified format, and applying a data anomaly correction algorithm to correct the abnormal data when anomalies are detected in the test data; a reference value comparison unit, which inputs the feature vector into a deviation detection module, compares it with the reference value after the working condition adaptability adjustment, identifies the deviation index and calculates the corresponding deviation severity score, wherein the deviation severity score is used to represent the degree of deviation; A deviation situation description generation unit inputs the identified deviation indicators and their deviation severity scores into a deviation pattern recognition module. The deviation pattern recognition module classifies deviation feature vectors based on a pre-trained machine learning model, matches possible deviation patterns, and generates a current deviation situation description, which includes deviation type, associated indicators, and severity score information; A fault source inference unit inputs the deviation pattern into a causal reasoning module, performs fault source inference based on a preset causal reasoning rule library, and outputs an interpretable reasoning path that illustrates the logical chain from the deviation feature to the root cause of the fault; a prediction unit, configured to predict a future thermal performance degradation trend of the unit based on the current deviation description and historical performance trend data, wherein the thermal performance trend data includes an archived historical thermal performance indicator sequence; The health assessment report generation unit generates a unit health assessment report based on future performance degradation trends. The health assessment report includes thermal performance deviation analysis results, suspected fault sources, reasoning paths, and performance trend prediction results.

10. The steam turbine deviation intelligent identification and tracing system based on thermal performance indicators according to claim 9 is characterized in that: The data anomaly correction algorithm in the standardization processing unit includes one or more of interpolation correction, correction based on thermodynamic balance estimation, or correction based on empirical rules.

Images