A method and apparatus for detecting the effect of a state reporting improvement measure

Abstract

The application discloses a state report improvement measure effect detection method and device, belongs to the nuclear power field, aims to solve the problem that the traditional detection method is strong in subjectivity, low in efficiency and insufficient in data utilization. The application contains five steps of data collection, preprocessing, index calculation, effect detection and report generation. First, multi-source data such as operation logs and fault records are collected, stored after cleaning and standardization; then, key indexes such as event frequency, severity, average repair time and average fault interval time are calculated; through statistical and machine learning algorithms such as multivariate regression model, the index changes before and after improvement are compared; finally, a detection report containing index comparison, effect analysis and improvement suggestions is generated. The application realizes quantitative detection, improves the objectivity and efficiency of detection, provides data support for nuclear power station operation decision, helps to improve the operation safety and efficiency, and has good flexibility and expansibility.

Description

Technical Field

[0001] This invention relates to the field of nuclear power, and in particular to a method and apparatus for detecting the effectiveness of status report improvement measures. Background Technology

[0002] Nuclear power plant operation, as a crucial component of the nuclear energy industry, is widely applied to improving equipment reliability and optimizing safety management. With the continuous development of nuclear power technology, the implementation of improvement measures has become a key means of ensuring system safety and economic efficiency. Among related technologies, an improvement effect monitoring system has been constructed through the collaborative work of manual statistics, experience-based judgment, and basic data analysis. Specifically, this system covers the entire process from event recording to indicator analysis, including event frequency, fault type identification, and repair time statistics. Based on this, existing technologies have formed a monitoring framework based on log analysis, fault records, and sensor data.

[0003] However, existing improvement measures and detection methods rely heavily on manual statistics and experience-based judgment, lacking a systematic quantitative analysis framework. This can lead to significant subjective biases in the results, making it difficult to accurately reflect actual effectiveness. Specifically, manual processing is inefficient, typically requiring weeks to complete data collection and analysis, impacting the timeliness of operational decisions. Furthermore, existing automated tools are mostly single-function modules, unable to effectively integrate multi-dimensional indicators such as event frequency, event severity, mean time to repair, and mean time between failures, resulting in insufficient integrity of the detection system. In addition, the lack of standardized processing for multi-source heterogeneous data (such as logs, sensor data, and reports) makes it difficult to guarantee data consistency, ultimately affecting the scientific validity and traceability of the detection results. Summary of the Invention

[0004] The present invention aims to at least partially solve one of the technical problems in the related art.

[0005] Therefore, the first objective of this invention is to provide a method for detecting the effectiveness of status report improvement measures.

[0006] Another objective of this invention is to provide a device for detecting the effectiveness of status report improvement measures.

[0007] The third objective of this invention is to provide a computer device.

[0008] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above objectives, a first aspect of the present invention provides a method for detecting the effectiveness of status report improvement measures, comprising: S1, acquire multi-source heterogeneous event data before and after the implementation of improvement measures. The multi-source heterogeneous event data includes event frequency, event severity, fault repair time and fault interval time. The event data is then cleaned and standardized. S2, calculate composite key performance indicators based on the processed event data. The composite key performance indicators include event frequency before improvement, event frequency after improvement, event severity before improvement, event severity after improvement, average repair time before improvement, average repair time after improvement, average failure interval before improvement, and average failure interval after improvement. S3. A multivariate regression model is used to compare and analyze the composite key performance indicators, separate the influence of improvement measures and external factors, and quantify the comprehensive effect of improvement measures on reducing event frequency, mitigating severity, improving repair efficiency and extending fault interval. S4. Generate a test report based on the comparative analysis results. The test report includes a comparison of indicators before and after the improvement, effect analysis conclusions, and improvement suggestions, which are used to guide the optimization of nuclear power plant operation decisions.

[0010] In one embodiment of the present invention, the step of acquiring multi-source heterogeneous event data before and after the implementation of the improvement measures and performing cleaning and standardization processing includes: S11, perform missing value imputation, duplicate data deletion and outlier detection on the multi-source heterogeneous event data, wherein the outlier detection adopts a statistical method based on the 3σ principle; S12 converts event data in different formats into a unified timestamp format and numerical encoding format, ensuring the standardized expression of event type and impact time fields.

[0011] In one embodiment of the present invention, the calculation of composite key performance indicators based on processed event data includes: S21, according to the formula Calculate the event frequency, where For the number of events, The length of the time period; S22, according to formula Calculate the severity of the event, where For the first The weight of each event, For the first The duration of the impact of each event.

[0012] In one embodiment of the present invention, the comparative analysis of the composite key performance indicators using a multivariate regression model includes: S31 will improve the frequency of previous events. and improved event frequency As an independent variable, input to a multivariate regression model Output detection indicators The changing trend; S32 will improve the severity of the previous event. and the severity of the improved incident As an independent variable, input to a multivariate regression model Output detection indicators The changing trend.

[0013] In one embodiment of the present invention, generating a detection report based on the comparative analysis results includes: S41 will display the calculated results of event frequency, event severity, MTTR, and MTBF before and after the improvement as percentage changes; S42, Based on the percentage change, generate improvement suggestions, including specific measures such as continuing to monitor the equipment's operating status and optimizing the cooling system.

[0014] To achieve the above objectives, a second aspect of the present invention provides a device for detecting the effectiveness of status report improvement measures, comprising: The data acquisition and preprocessing module is used to acquire multi-source heterogeneous event data before and after the implementation of improvement measures. The multi-source heterogeneous event data includes event frequency, event severity, fault repair time and fault interval time, and the event data is cleaned and standardized. The composite key performance indicator calculation module is used to calculate composite key performance indicators based on the processed event data. The composite key performance indicators include the event frequency before improvement, the event frequency after improvement, the event severity before improvement, the event severity after improvement, the average repair time before improvement, the average repair time after improvement, the average failure interval before improvement, and the average failure interval after improvement. The multivariate regression analysis module is used to compare and analyze the composite key performance indicators using a multivariate regression model, separate the influence of improvement measures and external factors, and quantify the comprehensive effect of improvement measures on reducing event frequency, mitigating severity, improving repair efficiency, and extending fault interval. The test report generation module is used to generate test reports based on the comparative analysis results. The test reports include a comparison of indicators before and after improvement, effect analysis conclusions, and improvement suggestions, which are used to guide the optimization of nuclear power plant operation decisions.

[0015] The method and apparatus of this invention can achieve objective quantitative detection of the effects of nuclear power plant improvement measures through multivariate regression models and composite KPI index systems, thereby improving the accuracy, comprehensiveness and real-time nature of the detection and providing scientific data support for operational decisions.

[0016] To achieve the above objectives, a third aspect of this application provides a computer device, including a processor and a memory; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, for implementing a method for detecting the effect of a status report improvement measure as described in the first aspect embodiment.

[0017] To achieve the above objectives, a fourth aspect of this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for detecting the effectiveness of a status report improvement measure as described in the first aspect embodiment.

[0018] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0019] Figure 1 This is a flowchart of a method for detecting the effectiveness of status report improvement measures according to an embodiment of the present invention; Figure 2 This is a structural diagram of a status report improvement measure effect detection device according to an embodiment of the present invention; Figure 3 It is a computer device according to an embodiment of the present invention. Detailed Implementation

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0022] The following description, with reference to the accompanying drawings, describes a method and apparatus for detecting the effectiveness of status reporting improvement measures according to an embodiment of the present invention.

[0023] Example 1 Figure 1 This is a flowchart of a method for detecting the effectiveness of status report improvement measures according to an embodiment of the present invention, such as... Figure 1 As shown, it includes: S1, acquire multi-source heterogeneous event data before and after the implementation of improvement measures. The multi-source heterogeneous event data includes event frequency, event severity, fault repair time and fault interval time. The multi-source heterogeneous event data is then cleaned and standardized.

[0024] Specifically, in some implementations, the technical principle of the present invention’s step of “acquiring multi-source heterogeneous event data before and after the implementation of improvement measures, and cleaning and standardizing the event data” is based on a systematic process of data acquisition, data cleaning and standardization, which aims to provide high-quality and consistent data input for subsequent calculation of key performance indicators (KPIs) and effect detection.

[0025] Furthermore, this step first extracts event data from heterogeneous data sources such as the nuclear power plant's operation logs, event reports, fault records, and sensor data through a multi-source data acquisition mechanism. Event data includes, but is not limited to, event occurrence time, event type, event frequency, event severity, fault repair time, and fault interval. Due to the diverse data sources and inconsistent formats, data cleaning and standardization are necessary. Data cleaning includes missing value handling, outlier detection and removal, and duplicate data identification and deletion. For example, missing event frequency data can be filled using interpolation or a sliding window mean based on time series; outliers can be identified and removed using the 3σ principle or box plot method.

[0026] Furthermore, data standardization includes unifying time formats (e.g., YYYY-MM-DD), event type encoding (e.g., using a unified fault classification coding standard), and numerical units (e.g., unifying repair time to hours). In some implementations, ETL (Extract, Transform, Load) tools can be optionally introduced to automate data extraction, transformation, and loading, ensuring the efficiency and consistency of data processing.

[0027] Furthermore, event frequency From the formula Definition, where The number of times the event occurred. For statistical time period length (in months or days). Event severity. The formula is calculated using weighted influence time. ,in Assign event weights (e.g., set to levels 1-5 based on fault severity). This refers to the duration of the event's impact (in hours or days). After standardization, all event data must meet uniform data quality standards, such as timestamp accuracy in seconds, no null values ​​in numeric fields, and categorization fields conforming to preset coding specifications.

[0028] Furthermore, this step is widely applied in scenarios such as nuclear power plant equipment improvement, system optimization, and safety measure implementation. For example, after replacing pump bearings or optimizing electrical systems, the system automatically collects event data before and after the improvement, and through data cleaning and standardization, provides structured input for subsequent KPI calculations and regression analysis. This step ensures the comparability and analyzability of the data, which is a prerequisite for achieving scientific monitoring.

[0029] Furthermore, this step significantly improves data accuracy and consistency by eliminating data noise and format differences, providing a reliable data foundation for subsequent quantitative testing. Simultaneously, standardization enables the integration of data from different devices and time periods, enhancing the algorithm's versatility and scalability, and providing strong support for the intelligent operation and maintenance and continuous improvement of nuclear power plants.

[0030] Furthermore, S1 includes: S11, perform missing value imputation, duplicate data deletion and outlier detection on the multi-source heterogeneous event data, wherein the outlier detection adopts a statistical method based on the 3σ principle.

[0031] Specifically, in the data preprocessing stage, this invention employs a systematic approach to clean the event data, including missing value imputation, deduplication, and outlier detection. This step affects the accuracy of subsequent Key Performance Indicators (KPIs) calculations and the reliability of the detection results.

[0032] Furthermore, the system first identifies missing fields in the dataset, such as event date, impact time, or repair time. For handling missing values, options include mean imputation, median imputation, or time-series-based interpolation, with the specific choice depending on the data distribution characteristics and the proportion of missing values. For example, if the missing proportion of the impact time field is less than 5%, the historical mean of that field can be used for imputation; if the missing proportion is high, linear interpolation of time series data or a KNN-based imputation strategy can be used to maintain data continuity and representativeness.

[0033] Furthermore, regarding deduplication, the system uses unique identifiers (such as a combination of event ID, timestamp, and device number) for deduplication. Optionally, a hash algorithm is used to encode the characteristics of each record, and the existence of duplicate records is determined by comparing hash values. Furthermore, the system supports a dynamic deduplication mechanism based on time windows; for example, identical events reported repeatedly within 15 minutes will be automatically merged or removed to avoid data redundancy interfering with the detection results.

[0034] Furthermore, in terms of outlier detection, this invention employs a statistical method based on the 3σ principle. Specifically, the system first calculates the mean of a numerical field (such as impact time or repair time) in the event data. and standard deviation Then, the range will be exceeded. Data points are identified as outliers. This method is based on the normal distribution assumption and is suitable for scenarios where event data has stable statistical characteristics. In practical applications, the system can set dynamic thresholds, such as calculating the mean and standard deviation based on a sliding window of historical data, to adapt to the trend of data changes over time.

[0035] Furthermore, this step is of significant value in the nuclear power plant operating environment. For example, before and after improvements to the cooling or electrical systems, the system needs to preprocess a large amount of event data to ensure that the calculated KPIs such as event frequency, event severity, MTTR, and MTBF are statistically significant. Through this step, the system can effectively improve the integrity and consistency of the data, providing high-quality input data for subsequent multivariate regression models, thereby enabling the scientific evaluation of the effectiveness of improvement measures.

[0036] S12 converts event data in different formats into a unified timestamp format and numerical encoding format, ensuring the standardized expression of event type and impact time fields.

[0037] Specifically, in the data preprocessing stage, this invention converts event data of different formats into a unified timestamp format and numerical encoding format to ensure the standardized expression of event type and impact time fields. This step is crucial for achieving data consistency and computability in the entire algorithm for detecting the effectiveness of improvement measures.

[0038] Furthermore, the raw event data may originate from multiple heterogeneous systems, such as operational logs, event reports, and sensor data. Their time fields may use different representations, such as "2024-06-01", "2024 / 06 / 01", or "June 1, 2024". To achieve unified processing, the system first converts all time fields to the ISO8601 standard timestamp format, i.e., "YYYY-MM-DDHH:MM:SS", and performs unified calibration using UTC time or the local time zone. The timestamp conversion process typically involves matching the original time format using regular expressions and calling a time parsing library (such as Python's `datetime` module) for standardization.

[0039] Furthermore, the event type field is typically unstructured text, such as "PumpFailure" or "TransformerOverheating." To facilitate subsequent quantitative analysis, the system uses numerical encoding to map event types. Optionally, the encoding method includes one-hot encoding or integer encoding, depending on the number of event types and the complexity of the classification. For the time-related field, the system uniformly converts it to a numerical format in hours; for example, "120 minutes" is converted to "2 hours," while "4 hours" remains unchanged. This conversion process must adhere to the time unit definition in nuclear power plant operation standards to ensure data consistency across different systems.

[0040] Furthermore, the timestamp conversion requires second-level precision to support subsequent MTBF calculations. Event type encoding must guarantee uniqueness and reversibility for semantic restoration during report generation. Numerical representations of time fields must retain one decimal place to improve calculation accuracy.

[0041] Furthermore, this step primarily involves the integrated processing of nuclear power plant operational data. For instance, after implementing improvements such as "replacing pump bearings" or "optimizing the electrical system," the system needs to standardize the time-series event data before and after the improvements to calculate key performance indicators (KPIs) such as event frequency, event severity, MTTR, and MTBF. By standardizing timestamps and numerical encoding, the system can eliminate interference from differences in data sources, improving the input quality for subsequent statistical analysis and machine learning models.

[0042] Furthermore, this step achieves the structuring and standardization of event data, providing a unified data foundation for subsequent KPI calculation and performance evaluation. By standardizing timestamps, the system can accurately calculate event intervals, thereby improving the accuracy of MTBF. Through numerical encoding of event types, the system can transform unstructured information into computable variables, enhancing the model's input dimensionality and detection capabilities. This step significantly improves the automation level of data processing and the objectivity of detection results, serving as a crucial prerequisite for data-driven decision support.

[0043] S2, calculate composite key performance indicators based on the processed event data. The composite key performance indicators include event frequency before improvement, event frequency after improvement, event severity before improvement, event severity after improvement, average repair time before improvement, average repair time after improvement, average failure interval before improvement, and average failure interval after improvement.

[0044] Specifically, in the "calculating composite key performance indicators based on processed event data" step of this invention, the system calculates multiple key performance indicators (KPIs) using structured and standardized event data to quantify the effects before and after the implementation of improvement measures. This step, based on mathematical modeling and statistical analysis of indicators such as event frequency, event severity, mean time to repair (MTTR), and mean time between failures (MTBF), provides a data foundation for subsequent effect detection and decision support.

[0045] Furthermore, event frequency (EventFrequency) The calculation uses the formula. ,in This represents the total number of events that occurred within a specific time period. This indicates the length of the time period (usually in months or days). Event frequency measures the number of failures or abnormal events that occur in the system per unit of time and is an important indicator for detecting system stability.

[0046] Furthermore, the severity of the event (EventSeverity, The formula is then used to calculate the weighted summation. ,in For the first The weight of each event (e.g., set to level 1-5 based on the event type). This refers to the duration of the event's impact (such as downtime, repair time, etc.). This metric comprehensively reflects the overall impact of an event on system operation and is particularly suitable for complex scenarios where multiple types of events coexist.

[0047] Furthermore, Mean Time to Repair (MTTR) and Mean Time Between Failures (MTBF) are used to detect the system's response efficiency and reliability in the event of a fault, respectively. The formula for calculating MTTR is as follows: ,in Indicates the first The repair time for the second fault; the formula for calculating MTBF is: ,in Indicates the first The timestamp of the incident. These metrics are important in nuclear power plant equipment maintenance, for example, for assessing the stability and responsiveness of cooling or electrical systems.

[0048] Furthermore, this step is typically deployed within the central monitoring and detection system of a nuclear power plant, processing structured event data from multiple data sources (such as sensors, log systems, and maintenance records). The system performs statistical analysis on events within preset time windows (such as 30 days or 90 days) and calculates indicators based on fields such as event type, impact time, and repair duration. During the calculation process, the system supports weighted processing of event data to reflect the differentiated impacts of different events on system operation.

[0049] Furthermore, the technical advantage of this step lies in transforming the originally scattered and unstructured event data into quantifiable performance indicators through standardized indicator calculation methods, providing input variables for subsequent multivariate regression models. Its innovation lies in unifying indicators from multiple dimensions, such as event frequency, severity, MTTR, and MTBF, into the detection system, enabling multi-dimensional and systematic detection of improvement measures, thereby enhancing the scientific rigor and accuracy of nuclear power plant operation decisions.

[0050] Furthermore, S2 includes: S21, according to the formula Calculate the event frequency, where For the number of events, This represents the length of the time period.

[0051] Specifically, in the improved measure effect detection algorithm of the present invention, the calculation of event frequency is performed using the formula... Implementation, in which This represents the total number of events that occurred within a specific time period. This indicates the length of the time period, typically measured in hours, days, or months, depending on the granularity of data collection and the detection requirements. Event frequency. The calculation aims to quantify the rate of failure or abnormal events in a system or device per unit time, thereby providing key input parameters for subsequent performance testing.

[0052] Furthermore, the calculation of event frequency needs to be based on a preprocessed event dataset. The data preprocessing stage includes cleaning, deduplication, anomaly detection, and timestamp standardization of the raw event records to ensure... The statistical results are accurate and representative. For example, if the event log contains multiple duplicate reports of the same event, the system will perform deduplication using a unique identifier (such as device number + event type + timestamp) to avoid overestimating the event frequency. Furthermore, timestamps must be uniformly converted to a standard time format (such as `YYYY-MM-DDHH:MM:SS`) to facilitate time interval division and calculation.

[0053] Furthermore, time period The settings should be adjusted according to the actual application scenario. In nuclear power plant operation, inspections are typically conducted on a monthly basis. Tian or Monthly. In scenarios requiring higher temporal resolution, weekly (7 days) or daily (24 hours) periods can also be used as the detection cycle. The calculated event frequency will directly affect the subsequent input to the multivariate regression model, used to detect the impact of improvement measures on system stability.

[0054] Furthermore, from a technical perspective, accurate calculation of event frequency can provide an objective basis for the quantitative evaluation of improvement measures. For example, after implementing equipment replacement or process optimization, if the event frequency increases from before the improvement... The number of times / months decreased to The frequency of occurrences per month indicates that the improvement measures have a significant effect on reducing the failure rate. This step, as one of the core indicators in the entire detection algorithm, provides a time benchmark and event quantity basis for the subsequent calculation of indicators such as event severity, MTTR, and MTBF, thereby achieving multi-dimensional and systematic detection of the improvement measures.

[0055] S22, according to formula Calculate the severity of the event, where For the first The weight of each event, For the first The duration of the impact of each event.

[0056] Specifically, in the improved measure effect detection algorithm of this invention, the calculation of event severity is an important component of key performance indicator (KPI) detection, and its technical implementation is based on formulas. ,in Indicates the first The weight of each event, This indicates the duration of the event's impact. This step quantifies the overall impact of the event on system operation, providing multi-dimensional numerical data for detecting improvement measures.

[0057] Furthermore, the severity of the event is calculated using a weighted summation method, which multiplies the duration of each event's impact by its corresponding weight and then sums the results to obtain the overall severity score. Weight The settings need to be standardized based on factors such as event type, fault level, and impact on system safety or operational efficiency. For example, in a nuclear power plant, pump failures may be assigned a higher weight (e.g., ), while slight overheating of the transformer may be assigned a lower weight (e.g. ). Duration of impact It is usually expressed in hours or days, indicating the duration of interruption or degrade in normal system operation caused by the event.

[0058] Furthermore, weight The value range is typically between [1, 10], and the specific value can be set according to the fault classification standards established internally by the nuclear power plant. Impact Time Data collection must meet timestamp accuracy requirements, generally using the ISO 8601 standard time format to ensure accurate time calculations. In practical applications, this step is typically deployed in the central monitoring and detection system of a nuclear power plant to quantitatively compare various events occurring before and after the implementation of improvement measures, thereby determining whether the measures have effectively reduced system operational risks.

[0059] Furthermore, the technical advantage of this step lies in its ability to more reasonably reflect the differentiated impacts of various events on system operation by introducing a weighting mechanism, avoiding the biases caused by simply detecting the number of events or the sum of their durations. Simultaneously, this method supports joint analysis with other KPIs such as event frequency, MTTR, and MTBF, providing input variables for multivariate regression models and thus enabling comprehensive evaluation of the effectiveness of improvement measures. In the complex systems of nuclear power plants, this step provides operators with intuitive and quantifiable decision-making support, contributing to improved system safety and operational efficiency.

[0060] S3. A multivariate regression model is used to compare and analyze the composite key performance indicators, separate the influence of improvement measures and external factors, and quantify the comprehensive effect of improvement measures on reducing event frequency, mitigating severity, improving repair efficiency, and extending fault interval.

[0061] Specifically, in the effect detection step, the present invention uses a multivariate regression model to compare and analyze composite key performance indicators in order to separate the influence of improvement measures and external factors, thereby achieving comprehensive quantitative detection of the improvement measures in terms of reduced event frequency, reduced severity, improved repair efficiency and extended fault interval.

[0062] Furthermore, the model introduces multiple independent variables, including the frequency of events before and after the improvement. Severity of the incident Average repair time Mean Time Between Failures We constructed a regression equation to examine the contribution of the improvement measures to the system performance.

[0063] Furthermore, the mathematical expression of the multivariate regression model is as follows:

[0064] in, Indicates the detection indicators (such as event frequency or severity). The original observation values. For the improved observations, For regression coefficients, This represents the error term. By fitting the model parameters using the least squares method or maximum likelihood estimation, the significant impact of the improvement measures on the changes in the indicators can be further identified.

[0065] Furthermore, models typically use standardized data input to eliminate the influence of different units of measurement on the regression results. For example, event frequency. ,in For the number of events, The duration of the event (in months or days); the severity of the event. ,in Assign event weights (e.g., set to levels 1-5 based on fault severity). The duration of the event's impact (in hours or days).

[0066] Furthermore, in practical applications of nuclear power plants, this step is typically deployed during the performance testing phase after the implementation of improvement measures. For example, after replacing critical equipment or optimizing operating procedures, the system automatically collects event data before and after the improvement, performs data cleaning and format standardization through a preprocessing module, and then inputs it into a multivariate regression model for analysis. By comparing the regression coefficients... and The difference can be used to determine whether the improvement measures have significantly reduced the frequency or severity of events, while eliminating the interference of external variables such as ambient temperature and equipment aging.

[0067] Furthermore, the technical advantage of this step lies in its ability to provide objective and quantifiable detection results, avoiding biases caused by subjective judgment in traditional methods. Through the output of the regression model, operators can clearly define the specific contributions of improvement measures in enhancing system stability, shortening repair time, and extending intervals between failures, thus providing a scientific basis for subsequent maintenance strategies and resource allocation. This method has broad application value in equipment maintenance, system optimization, and safety management in nuclear power plants, and is particularly suitable for complex systems requiring long-term tracking of improvement effects.

[0068] Furthermore, S3 includes: S31 will improve the frequency of previous events. and improved event frequency As an independent variable, input to a multivariate regression model Output detection indicators The changing trend.

[0069] Specifically, in the effect detection step, the frequency of events before improvement will be... and improved event frequency As an independent variable, input to a multivariate regression model To output detection indicators The trend of change is analyzed. This step is based on the principle of statistical modeling. By constructing a linear relationship model, the changes in key performance indicators (KPIs) before and after the implementation of improvement measures are fitted and predicted, thereby detecting their comprehensive impact on the system's operating status.

[0070] Furthermore, the regression model employs Ordinary Least Squares (OLS) for parameter estimation, optimizing the objective function. Solving for regression coefficients ,in These are the model's predicted values. These are the actual observed values ​​of the detection indicators. (Model input) and These represent the event frequencies before and after the improvement, respectively. They are usually standardized to the number of events occurring per unit time (e.g., times / month) to eliminate the impact of time span differences on the model.

[0071] Furthermore, the regression coefficients in the model and These represent the impact of event frequencies before and after the improvement on the detection index, respectively. The degree of impact. If This indicates that the decrease in event frequency after the implementation of the improvement measures had a positive impact on the detection indicators. Error Term To capture unexplained random disturbances in the model, its variance should be controlled within a reasonable range to ensure the model's fitting accuracy and predictive stability.

[0072] Furthermore, this step is widely applied in various operation and maintenance scenarios at nuclear power plants, including equipment upgrades, system optimizations, and process adjustments. For example, after replacing critical equipment components or optimizing electrical systems, by collecting event frequency data for several consecutive months before and after the improvement, and inputting it into a regression model for fitting, the long-term impact of the improvement measures on indicators such as system stability and failure rate can be intuitively reflected. This model can also be combined with other KPIs (such as event severity). Multi-dimensional detection is performed to improve the comprehensiveness of the analysis.

[0073] Furthermore, this step enables quantitative analysis of the effectiveness of improvement measures, avoiding the limitations of traditional methods that rely on subjective experience. The output of the regression model identifies the significant impact of improvement measures on system performance, providing a scientific basis for subsequent decision-making. In addition, this method has good interpretability, facilitating operators' understanding of the causal relationship between improvement measures and system performance, thereby promoting the intelligent and data-driven transformation of nuclear power plant operation and maintenance management.

[0074] S32 will improve the severity of the previous event. and the severity of the improved incident As an independent variable, input to a multivariate regression model Output detection indicators The changing trend.

[0075] Specifically, in the effectiveness evaluation step, the severity of the event before improvement will be assessed. and the severity of the improved incident As an independent variable, input to a multivariate regression model To output detection indicators The changing trend is based on the principle of statistical modeling. By constructing a linear relationship model, mathematical modeling and trend prediction are performed on the changes in the severity of events before and after the implementation of improvement measures, thereby achieving an objective and quantitative analysis of the improvement effect.

[0076] Furthermore, this step first extracts preprocessed event data from the central database, including event severity values ​​for both the pre- and post-improvement phases. and Severity of the incident Usually through formulas Perform calculations, where For the first The weight of each event, The duration of the event's impact. Weight. The settings can be configured based on factors such as event type, scope of impact, and economic losses. For example, a major malfunction can be set as... Medium fault is set as Minor faults are set as In practical applications, and It can represent the weighted sum of all events within a certain period before and after the improvement, respectively, and is used to reflect the changes in the overall system operating status.

[0077] Furthermore, this multivariate regression model middle, Indicates the changing trend of the detection indicators. For the model intercept term, and These are the regression coefficients representing the severity of the events before and after the improvement, respectively. This is the error term, used to capture random fluctuations not explained by the model. By fitting the model parameters using methods such as least squares (OLS) or maximum likelihood estimation (MLE), the quantitative impact of improvements on dimensions such as system stability and security can be obtained.

[0078] Furthermore, this step is widely used for post-implementation monitoring of improvements such as nuclear power plant equipment upgrades, system optimizations, and maintenance strategy adjustments. For example, after replacing bearings in critical equipment or optimizing electrical systems, this model can detect a downward trend in the severity of the event, thereby determining whether the improvements have effectively reduced potential risks. In addition, the model can also be used to predict changes in system operating status over a future period, providing forward-looking support for operational decisions.

[0079] Furthermore, the technical advantage of this step lies in the fact that by introducing a multivariate regression model, it can comprehensively consider the changes in event severity before and after the improvement, avoiding bias caused by single-indicator detection and improving the accuracy and scientific rigor of the detection results. At the same time, the model has good interpretability, making it easier for managers to understand the specific impact of improvement measures on system performance, thereby achieving data-driven management optimization.

[0080] S4. Generate a test report based on the comparative analysis results. The test report includes a comparison of indicators before and after improvement, effect analysis conclusions, and improvement suggestions, which are used to guide the optimization of nuclear power plant operation decisions.

[0081] Specifically, in some implementations, the step of generating an inspection report based on the comparative analysis results involves systematically integrating key performance indicators (KPIs) before and after the implementation of improvement measures, and combining the output results of statistical analysis and machine learning models to generate a structured and traceable inspection report, which is used to guide the operation optimization and risk control of nuclear power plants.

[0082] Furthermore, this step first retrieves preprocessed event data from the central database, including event frequency. Severity of the incident Average repair time and mean time between failures These indicators quantify the improvement effect from four dimensions: event frequency, impact, response efficiency, and system stability. Subsequently, the system compares and analyzes these indicators according to time series, identifies the trends before and after the improvement, and uses a multivariate regression model to further analyze the results. The overall impact of the improvement measures is modeled, among which... This represents the index value before the improvement. This indicates the improved indicator value. To evaluate the overall performance score output by the detection model, For regression coefficients, This represents the error term. The model output can be used to determine whether the improvement measures have achieved their intended goals and to predict their long-term impact.

[0083] Furthermore, the test report should clearly list the percentage change in each KPI, such as a 66.7% reduction in event frequency, a 50% reduction in event severity, a 25% shortening of MTTR, and a 50% increase in MTBF. In addition, the report should include statistical indicators such as the goodness of fit of the regression model (R² value) and significance test (p-value) to verify the reliability of the test results. In practical applications, the system can set thresholds, such as an event frequency decrease of more than 30% and an MTBF increase of more than 20%, as standards for judging the effectiveness of improvement measures.

[0084] Furthermore, this step is widely applied in scenarios such as nuclear power plant equipment upgrades, system optimization, and maintenance strategy adjustments. For example, after replacing pump bearings or optimizing electrical systems, the system will automatically generate an inspection report for reference by operations engineers, maintenance personnel, and management. The report can be integrated into the nuclear power plant's decision support system (DSS) to achieve real-time feedback and long-term tracking of the effectiveness of improvement measures.

[0085] Furthermore, this step, through structured report output, significantly enhances the objectivity and traceability of improvement measure detection, reducing bias from human judgment. Simultaneously, it supports multi-dimensional indicator comparison and model-assisted analysis, providing a scientific basis for operational decisions and contributing to the refined management and continuous improvement of nuclear power plants.

[0086] Furthermore, S4 includes: S41 displays the calculated results of event frequency, event severity, MTTR, and MTBF before and after the improvement as percentage changes.

[0087] Specifically, in some implementations, the calculated results of event frequency, event severity, MTTR (mean time to repair), and MTBF (mean time between failures) before and after the improvement are presented as percentage changes. This step provides nuclear power plant operators with intuitive and comparable test results by quantifying the performance differences before and after the implementation of improvement measures, thereby supporting subsequent decision optimization and system improvement.

[0088] Furthermore, this step first calculates key performance indicators (KPIs) based on the preprocessed event data, both before and after the improvement. For example, event frequency. ,in The number of times the event occurred. For statistical time period length (in months or days); event severity ,in For the first The weight of each event, This represents the duration of the event's impact. MTTR and MTBF are calculated using the following formulas: and ,in For the first Repair time for this fault For the first The time point when the fault occurred.

[0089] Furthermore, the formula for calculating the percentage change is: ,in and These represent the indicator values ​​before and after the improvement, respectively. For example, if the event frequency was 1.5 times / month before the improvement and 0.5 times / month after the improvement, the percentage change is -66.7%, indicating a significant decrease in event frequency. This calculation method complies with industrial equipment reliability testing standards such as ISO 14224, ensuring the comparability and standardization of test results.

[0090] Furthermore, this step is widely applied in scenarios such as nuclear power plant equipment improvement, system optimization, and maintenance strategy adjustment. For example, after replacing pump bearings or optimizing electrical systems, this step can visually display the changing trends of various indicators before and after the improvement, assisting engineers in determining whether the improvement measures have achieved the expected goals. In actual deployment, this step is usually integrated into the condition monitoring and detection system, supporting multi-dimensional and multi-time window comparative analysis, and is suitable for monthly, quarterly, or annual inspection cycles.

[0091] Furthermore, this step, through the intuitive display of percentage changes, effectively improves the readability and operability of the detection results. It not only reflects the impact of improvement measures on a single indicator but also reveals the overall performance improvement or degradation trend of the system through the comprehensive changes of multiple indicators. Moreover, this step provides input data for subsequent multivariate regression models, enhancing the interpretability and predictive power of the detection model, thereby realizing a data-driven continuous improvement mechanism.

[0092] S42, Based on the percentage change, generate improvement suggestions, including specific measures such as continuing to monitor the equipment's operating status and optimizing the cooling system.

[0093] Specifically, in some implementations, the step of generating improvement suggestions based on the percentage change is the output of the effect detection algorithm in this invention. Its technical implementation principle is based on the interpretation of quantitative analysis and statistical modeling results of key performance indicators (KPIs). This step compares the percentage changes in indicators such as event frequency, event severity, mean time to repair (MTTR), and mean time between failures (MTBF) before and after the implementation of improvement measures. Combined with preset thresholds and detection models, it generates actionable improvement suggestions to support the continuous optimization and safety management of nuclear power plants.

[0094] Furthermore, this step is widely applied to equipment maintenance, system optimization, and safety management in nuclear power plants. For example, after cooling system optimization, if the MTBF increases from 30 days to 45 days and the MTTR decreases from 4 hours to 3 hours, the system will generate recommendations to "continue monitoring the cooling system's operating status" and "optimize the cooling system's heat dissipation mechanism." These recommendations can be integrated into the operations management system for maintenance and management personnel to reference in developing subsequent maintenance plans and system upgrade strategies.

[0095] Furthermore, this step achieves a closed-loop feedback from data to decision-making, ensuring that improvement recommendations are targeted and feasible. By quantifying changes in indicators, the system can identify the long-term stability and potential risks of improvement measures, thereby enhancing the safety and efficiency of nuclear power plant operation. In addition, this step supports multi-dimensional, multi-device parallel analysis, providing technical support for the management optimization of large-scale equipment groups.

[0096] The method for quantitatively detecting the effects of nuclear power plant improvement measures in this invention can achieve quantitative detection of the effects of nuclear power plant improvement measures. Through automated data processing and multivariate regression analysis, it improves the objectivity, accuracy, and real-time performance of the detection, providing scientific decision support for operation and management.

[0097] Example 2 This invention provides an algorithm for detecting the effectiveness of improvement measures, aiming to provide scientific data support for the operation and management of nuclear power plants by quantifying and detecting the effects before and after the implementation of improvement measures. The algorithm comprises five main steps: data collection, data preprocessing, indicator calculation, effect detection, and report generation. The main workflow of the system is as follows: Data Collection: Collect various event data before and after the implementation of improvement measures, including event frequency, event severity, and fault type. Data sources include operation logs, event reports, fault records, and sensor data. Data Preprocessing: Clean the collected data, removing noise and erroneous data. Perform format conversion and standardization to ensure data consistency. The processed data is stored in a central database for use by subsequent modules. Indicator Calculation: Calculate key performance indicators (KPIs), such as event frequency, event severity, mean time to repair (MTTR), and mean time between failures (MTBF). These indicators are used to quantify the effects before and after the implementation of improvement measures. Effect Detection: Detect the effectiveness of improvement measures based on the calculated key performance indicators. Use statistical analysis and machine learning algorithms to compare the changes in indicators before and after the implementation of improvement measures to determine the actual effect of the improvement measures. Report Generation: Generates detailed monitoring reports, including comparisons of indicators before and after the implementation of improvement measures, effect analysis results, and improvement suggestions. The reports are used to support operational decisions and optimize management strategies.

[0098] Furthermore, data collection and preprocessing include: Data Collection: Data sources: operational logs, event reports, fault records, sensor data, etc. Data types: event date, event type, impact time, description, fault repair time, etc. Data Preprocessing: Data Cleaning: Removing missing values, duplicate data, and outliers. Data Standardization: Converting data of different formats to a standard format. Data Storage: Storing preprocessed data in a central database. Example Data: Event Log 1: 2024-08-01, Cooling System Failure, 90, Minor failure, 2; Event Log 2: 2024-09-10, Electrical Short, 240, Major failure, 5. Key Performance Indicator (KPI) Calculation:

[0099]

[0100]

[0101] EventSeverity(S): The degree of impact of an event, which can be calculated based on the duration of the event's impact or other weights.

[0102]

[0103]

[0104] Mean Time to Repair (MTTR): The average time required to repair a fault.

[0105] Where (Ri) is the repair time for the (i)th fault.

[0106] Mean Time Between Failures (MTBF): The average time interval between two failures.

[0107]

[0108] Where (Ti) is the time when the (i)th fault occurs.

[0109] Example calculation: Event Log 1: 2024-08-01, Cooling System Failure, 90, Minor Failure, 2; Event Log 2: 2024-09-10, Electrical Short, 240, Major Failure, 5; Event Frequency (f): 2 times / 2 months = 1 time / month; Event Severity (S): 90 + 240 = 330; Mean Time To Repair (MTTR): (2 + 5) / 2 = 3.5 hours; Mean Time Between Failures (MTBF): (September 10th - August 1st) = 40 days Furthermore, the effectiveness evaluation method involves comparing key performance indicators (KPIs) before and after the implementation of improvement measures. Statistical analysis and machine learning algorithms are used to evaluate the effectiveness of the improvement measures. The effectiveness evaluation algorithm uses a multivariate regression model to detect changes in event frequency and severity.

[0110]

[0111] Where (y) is the detection index, (x1) is the data before improvement, (x2) is the data after improvement, and (β0,β1,β2) are the regression coefficients. The error term is represented by ). Example detection: Assume the event frequency before improvement was 1.8 times / month, and the event frequency after improvement is 0.6 times / month. Assume the event severity before improvement was 350, and the event severity after improvement is 150.

[0112] Using a multivariate regression model to detect: Changes in event frequency: y = β0 + β1 1.8+β2 0.6 Change in event severity: y = β0 + β1 350+β2 150. Report Content: Comparison of indicators before and after the implementation of improvement measures. Results of the effectiveness analysis.

[0113] Further improvements are recommended: Continue to monitor the operating status of the electrical system and perform regular maintenance. Further optimize the cooling mechanism of the electrical system to prevent overheating-related malfunctions. Unique algorithms and data processing include: Multivariate regression model: By introducing a multivariate regression model, various influencing factors are comprehensively considered to detect the overall effectiveness of the improvement measures. Formula:

[0114] Where (y) is the detection index, (x1) is the data before improvement, (x2) is the data after improvement, and (β0,β1,β2) are the regression coefficients. The error term is denoted as ). Key Performance Indicators (KPIs) Calculation: Accurately calculate multiple key performance indicators such as event frequency, event severity, mean time to repair (MTTR), and mean time between failures (MTBF).

[0115]

[0116]

[0117]

[0118]

[0119] Through the detailed implementation details and examples above, the algorithm for detecting the effectiveness of improvement measures of the present invention can scientifically and systematically detect the actual effect of improvement measures, providing data support and decision-making basis for the operation and management of nuclear power plants.

[0120] By accurately calculating key performance indicators (KPIs) such as event frequency, event severity, mean time to repair (MTTR), and mean time between failures (MTBF), the effectiveness of improvement measures is quantified, providing scientific data support. Automated data processing: Automatically collects and preprocesses various event data before and after the implementation of improvement measures, reducing manual operations, ensuring data consistency and integrity, and improving work efficiency. Multivariate regression model: Introduces a multivariate regression model to comprehensively consider multiple influencing factors, detecting the overall effect of improvement measures and providing more comprehensive and accurate detection results. Data-driven decision support: Through systematic data analysis and detection, provides scientific data support for operation management, helping managers make more informed decisions and improving the operational efficiency and safety of nuclear power plants. Real-time monitoring and feedback: The system can analyze and provide feedback on new data in real time, providing timely detection results and improvement suggestions for the effectiveness of improvement measures, ensuring continuous operational optimization. Detailed detection reports: Generates detailed detection reports, including comparisons of indicators before and after the implementation of improvement measures, effect analysis results, and improvement suggestions, supporting the optimization and improvement of operation management.

[0121] Furthermore, the system integrates multiple detection indicators: It innovatively combines several key performance indicators such as event frequency, event severity, mean time to repair (MTTR), and mean time between failures (MTBF) to comprehensively assess the effectiveness of improvement measures, providing a more scientific and systematic detection method. Multivariate regression analysis: Utilizing a multivariate regression model, the system innovatively detects the overall effectiveness of improvement measures, considering the impact of multiple factors on the improvement effect, thus improving the accuracy and scientific rigor of the detection. Automation and real-time performance: The system automatically collects, preprocesses, and detects data, providing real-time detection results and feedback, ensuring efficient and real-time data processing, and improving the intelligence level of nuclear power plant operation and management. Data-driven management optimization: Through data-driven methods, the system systematically detects the effectiveness of improvement measures, supporting the scientific and accurate nature of management decisions and promoting continuous optimization and improvement of nuclear power plant operation and management. Flexible scalability: The system architecture features a modular design with good flexibility and scalability, allowing for the addition of new detection indicators and algorithms as needed to adapt to the different operation and management requirements of nuclear power plants.

[0122] The algorithm for detecting the effectiveness of improvement measures in this invention provides a scientific, systematic, and efficient method for detecting the effectiveness of improvement measures through several innovative technologies, including quantitative detection, automated data processing, multivariate regression analysis, and real-time monitoring feedback. This method not only improves the operational efficiency and safety of nuclear power plants but also provides strong data support for management decisions, promoting the intelligent and continuous optimization of nuclear power plant operation and management.

[0123] Example 3 To achieve the above embodiments, such as Figure 2 As shown, this embodiment also provides a status report improvement measure effect detection device 10, including: The data acquisition and preprocessing module 100 is used to acquire multi-source heterogeneous event data before and after the implementation of improvement measures. The multi-source heterogeneous event data includes event frequency, event severity, fault repair time and fault interval time, and the event data is cleaned and standardized. The composite key performance indicator calculation module 200 is used to calculate composite key performance indicators based on the processed event data. The composite key performance indicators include the event frequency before improvement, the event frequency after improvement, the event severity before improvement, the event severity after improvement, the average repair time before improvement, the average repair time after improvement, the average failure interval before improvement, and the average failure interval after improvement. The multivariate regression analysis module 300 is used to compare and analyze the composite key performance indicators using a multivariate regression model, separate the influence of improvement measures and external factors, and quantify the comprehensive effect of improvement measures on reducing event frequency, mitigating severity, improving repair efficiency, and extending fault interval. The test report generation module 400 is used to generate a test report based on the comparative analysis results. The test report includes a comparison of indicators before and after improvement, effect analysis conclusions, and improvement suggestions, which are used to guide the optimization of nuclear power plant operation decisions.

[0124] Furthermore, the data acquisition and preprocessing module 100 is also used for: The multi-source heterogeneous event data is subjected to missing value imputation, deduplication, and outlier detection, wherein outlier detection is based on... Statistical methods based on principles; Convert event data in different formats into a unified timestamp format and numerical encoding format to ensure standardized expression of event type and impact time fields.

[0125] Furthermore, the composite key performance indicator calculation module 200 is also used for: According to the formula Calculate the event frequency, where For the number of events, The length of the time period; According to the formula Calculate the severity of the event, where For the first The weight of each event, For the first The duration of the impact of each event.

[0126] Example 4 To implement the methods of the above embodiments, the present invention also provides a computer device, such as... Figure 3 As shown, the computer device 600 includes a memory 601 and a processor 602; wherein, the processor 602 reads the executable program code stored in the memory 601 to run a program corresponding to the executable program code, so as to implement the various steps of the status report improvement measure effect detection method described above.

[0127] Example 5 To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for detecting the effect of status report improvement measures as described in the foregoing embodiments.

[0128] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0129] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for detecting the effectiveness of status report improvement measures, characterized in that, include: S1, acquire multi-source heterogeneous event data before and after the implementation of improvement measures. The multi-source heterogeneous event data includes event frequency, event severity, fault repair time and fault interval time. The multi-source heterogeneous event data is then cleaned and standardized. S2, calculate composite key performance indicators based on the processed event data. The composite key performance indicators include event frequency before improvement, event frequency after improvement, event severity before improvement, event severity after improvement, average repair time before improvement, average repair time after improvement, average failure interval before improvement, and average failure interval after improvement. S3. A multivariate regression model is used to compare and analyze the composite key performance indicators, separate the influence of improvement measures and external factors, and quantify the comprehensive effect of improvement measures on reducing event frequency, mitigating severity, improving repair efficiency and extending fault interval. S4. Generate a test report based on the comparative analysis results. The test report includes a comparison of indicators before and after the improvement, effect analysis conclusions, and improvement suggestions, which are used to guide the optimization of nuclear power plant operation decisions.

2. The method as described in claim 1, characterized in that, The process of acquiring multi-source heterogeneous event data before and after the implementation of improvement measures, and performing cleaning and standardization processing, includes: S11, perform missing value imputation, duplicate data deletion and outlier detection on the multi-source heterogeneous event data, wherein the outlier detection adopts a statistical method based on the 3σ principle; S12 converts event data in different formats into a unified timestamp format and numerical encoding format, ensuring the standardized expression of event type and impact time fields.

3. The method as described in claim 1, characterized in that, The calculation of composite key performance indicators based on processed event data includes: S21, according to the formula Calculate the event frequency, where For the number of events, The length of the time period; S22, according to formula Calculate the severity of the event, where For the first The weight of each event, For the first The duration of the impact of each event.

4. The method as described in claim 1, characterized in that, The comparative analysis of the composite key performance indicators using a multivariate regression model includes: S31 will improve the frequency of previous events. and improved event frequency As an independent variable, input to a multivariate regression model Output detection indicators The changing trend; S32 will improve the severity of the previous event. and the severity of the improved incident As an independent variable, input to a multivariate regression model Output detection indicators The changing trend.

5. The method as described in claim 1, characterized in that, The step of generating a test report based on the comparative analysis results includes: S41 will display the calculated results of event frequency, event severity, MTTR, and MTBF before and after the improvement as percentage changes; S42, Based on the percentage change, generate improvement suggestions, including specific measures such as continuing to monitor the equipment's operating status and optimizing the cooling system.

6. A device for detecting the effectiveness of status report improvement measures, characterized in that, include: The data acquisition and preprocessing module is used to acquire multi-source heterogeneous event data before and after the implementation of improvement measures. The multi-source heterogeneous event data includes event frequency, event severity, fault repair time and fault interval time, and the event data is cleaned and standardized. The composite key performance indicator calculation module is used to calculate composite key performance indicators based on the processed event data. The composite key performance indicators include the event frequency before improvement, the event frequency after improvement, the event severity before improvement, the event severity after improvement, the average repair time before improvement, the average repair time after improvement, the average failure interval before improvement, and the average failure interval after improvement. The multivariate regression analysis module is used to compare and analyze the composite key performance indicators using a multivariate regression model, separate the influence of improvement measures and external factors, and quantify the comprehensive effect of improvement measures on reducing event frequency, mitigating severity, improving repair efficiency, and extending fault interval. The test report generation module is used to generate test reports based on the comparative analysis results. The test reports include a comparison of indicators before and after improvement, effect analysis conclusions, and improvement suggestions, which are used to guide the optimization of nuclear power plant operation decisions.

7. The apparatus as claimed in claim 6, characterized in that, The data acquisition and preprocessing module is also used for: The multi-source heterogeneous event data is subjected to missing value imputation, deduplication, and outlier detection, wherein outlier detection is based on... Statistical methods based on principles; Convert event data in different formats into a unified timestamp format and numerical encoding format to ensure standardized expression of event type and impact time fields.

8. The apparatus as claimed in claim 6, characterized in that, The composite key performance index calculation module is also used for: According to the formula Calculate the event frequency, where For the number of events, The length of the time period; According to the formula Calculate the severity of the event, where For the first The weight of each event, For the first The duration of the impact of each event.

9. A computer device, characterized in that, Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement a method for detecting the effect of status report improvement measures as described in any one of claims 1-5.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements a method for detecting the effect of status report improvement measures as described in any one of claims 1-5.

Images