A project settlement management risk monitoring system

Abstract

The application discloses an engineering settlement management risk monitoring system, relates to management monitoring optimization technology, and is used for improving the data expansion or shortage problem caused by the data category difference of engineering data under fixed-point monitoring, and comprises the following steps: collecting the change conditions of various engineering data and recording the uploading and auditing frequency of different categories of engineering data, counting the historical error rate of various engineering data, formulating different risk braking values for different categories of engineering data, counting the revision conditions of various engineering data and setting different monitoring levels, comprehensively selecting engineering data risk monitoring methods for risk monitoring according to the monitoring levels and the historical error rates of various engineering data, obtaining monitoring results and inputting the monitoring results into a user end, obtaining the confirmation results of the user end within a test time, calculating the monitoring efficiency of various engineering data risk monitoring methods, calling the engineering completion progress, collecting different feedback data in combination with the monitoring efficiency, and judging whether the corresponding risk monitoring methods of different categories of engineering data are changed.

Description

Technical Field

[0001] This invention relates to the field of management, monitoring and optimization technology, and more specifically, to an engineering settlement management risk monitoring system. Background Technology

[0002] Management monitoring optimization technology is a technique and method used to improve management efficiency, enhance monitoring capabilities, and optimize decision-making processes. When applied to project settlement, management monitoring optimization technology can improve the security of project settlement and effectively regulate the monitoring and management of resources.

[0003] The existing technology has the following shortcomings:

[0004] In the past, when processing engineering data, we would periodically record each engineering data in the system and monitor it at fixed points until the project was completed. Finally, we would integrate and settle the engineering data. However, we did not take into account that different engineering resources were used in different construction periods. Periodic fixed-point monitoring could easily lead to problems such as data bloat or shortage, thereby increasing the risk of data omission. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an engineering settlement management risk monitoring system. This system calculates the historical error rate of each engineering data, and, in conjunction with changes and audits, formulates targeted risk mitigation measures for different categories of engineering data, sets different monitoring levels and methods, and finally analyzes the monitoring efficiency to update the monitoring methods for different categories of engineering data, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An engineering settlement management risk monitoring system includes a data acquisition module, a monitoring selection module, a risk processing module, and a feedback response module.

[0008] The data acquisition module is used to collect changes in various types of engineering data in the engineering management system and record the upload and review frequency of different categories of engineering data; it also calls historical data to calculate the historical error rate of various types of engineering data.

[0009] The monitoring selection module is used to receive changes to various types of engineering data and the frequency of upload and review. It sets risk braking values ​​for different categories of engineering data, counts the revisions of various types of engineering data, and sets different monitoring levels based on the risk braking values.

[0010] The risk processing module selects different engineering data risk monitoring methods based on the monitoring level and historical error rate of various types of engineering data. Each type of engineering data obtains monitoring results according to the corresponding risk monitoring method and sends them to the user terminal. The user terminal confirms the monitoring results and obtains the confirmation results.

[0011] The feedback response module sets the inspection time, obtains multiple confirmation results, calculates the monitoring efficiency of various engineering data risk monitoring methods, calls the project completion progress, collects different feedback data in combination with the monitoring efficiency, and determines whether to change the risk monitoring methods corresponding to different categories of engineering data.

[0012] In a preferred embodiment, the change in engineering data is referred to as the magnitude of data change. When each type of data changes at different points in time, the magnitude of data change is represented by a relative difference. In the formula, 'a' represents the magnitude of data change. The changed data value. The data value before the change;

[0013] Upload review frequency refers to the number of upload review events for each type of data within a preset time period. A time interval is selected as the time window, and the upload review frequency is obtained by dividing the number of upload review events for data within the time window by the time interval of the time window.

[0014] In a preferred embodiment, the historical error rate is the proportion of errors occurring in each type of data, obtained by comparing the number of erroneous changes to the total number of changes within a past time window interval.

[0015] The data acquisition module first compares each change record with a preset threshold. When the data change exceeds the preset threshold, the system marks it as an erroneous change. The ratio of the number of erroneous changes to the total number of changes is used as the historical error rate of various types of engineering data.

[0016] In a preferred embodiment, when setting the risk braking value, the mean and standard deviation of the data change range are first calculated, and the mean of the data change range is added to twice the standard deviation to obtain the preliminary risk braking value.

[0017] Adjusting the risk threshold based on upload review frequency; defining adjustment factors. The impact of upload review frequency on the risk braking value is indicated by the following formula: ,in, To adjust the frequency of uploads, f represents the upload review frequency. It is an adjustment factor;

[0018] Adjustment factor The initial risk braking value is used to derive the risk braking value, and the calculation formula is as follows: ,in, This is the risk braking value. This is the initial risk braking value. This is an adjustment factor.

[0019] In a preferred embodiment, the monitoring selection module sets a correction rate to correct the engineering data: Where L represents the correction rate, Indicates the number of corrected engineering data. This indicates the total number of historical erroneous data points.

[0020] After obtaining the revision rate for each type of data, the monitoring selection module sets the monitoring index based on the risk braking value. This is used to determine the monitoring level of the data.

[0021] In a preferred embodiment, the monitoring selection module establishes a linear regression model and fits the monitoring index using the least squares method. The relationship between revision rate and risk braking value is expressed by the following formula: ,in, It is the target variable, namely the monitoring index. It is a risk braking value. It's the correction rate. It is a constant term. , These are the preset regression coefficients, representing and The degree of influence on the target variable, It is an error term;

[0022] After the regression model is established, the monitoring selection module inputs the correction rate and risk braking value of various types of data into the regression equation to calculate the monitoring index of various types of data. .

[0023] In a preferred embodiment, the monitoring selection module calculates monitoring indices for different engineering data. Calculate multiple monitoring indices mean And Set as the threshold for level division.

[0024] when This indicates that the data changes are relatively small and the risk is low, thus it is classified as a low monitoring level.

[0025] when This indicates significant data fluctuations and a high risk level, thus classifying it as a high-level monitoring situation.

[0026] In a preferred embodiment, when the monitoring level of the engineering data is high and the historical error rate is greater than its average, the risk processing module will select the periodic risk review method; when the monitoring level of a certain type of data is low and the historical error rate is less than its average, the risk processing module will select the risk matrix method.

[0027] Whenever a data change operation is performed, the risk handling module will evaluate the magnitude of the data change. If the magnitude of the data change is greater than or equal to the risk braking value, it will be marked as risky data; if the magnitude of the data change is less than the risk braking value, it will be marked as non-risky data.

[0028] The risk handling module sends the values ​​of all changed data within the time window as monitoring results to the user terminal, and simultaneously obtains the audit results from the user terminal. The audit results include two options: "yes" or "no".

[0029] When the audit result is "No", the risk handling module will re-mark this changed data as non-risk data.

[0030] In a preferred embodiment, monitoring efficiency combines the proportion of risk data among all change data within the time window interval with the accuracy of risk data labeling: the calculation formula is as follows. ,in, For monitoring efficiency, The amount of unaudited risk data. For the total number of changed data, The number of risk data points for which the audit result is "yes". The total number of data points marked as risky.

[0031] In a preferred embodiment, the determination of the monitoring method is based on an adjustment threshold for monitoring efficiency. The adjustment threshold is used to divide the data into K clusters using a K-means clustering algorithm, and the Euclidean distance between the centers of adjacent clusters is calculated using the following formula:

[0032] ,

[0033] in, and It represents the progress and monitoring efficiency of the k-th cluster center. and It represents the progress and monitoring efficiency of the (k+1)th cluster center;

[0034] After calculating the distance between all adjacent cluster centers, the boundary monitoring efficiency value between the two clusters corresponding to the maximum distance is used as the adjustment threshold for adjusting the monitoring method;

[0035] If the monitoring efficiency is less than the adjustment threshold, continue using the current risk monitoring method. If the monitoring efficiency is greater than the adjustment threshold, switch the current risk monitoring method to the risk matrix method and push it to the risk processing module.

[0036] The technical effects and advantages of the engineering settlement management risk monitoring system of the present invention are as follows:

[0037] This invention improves the efficiency and accuracy of engineering data monitoring by collecting data on changes in various engineering data and recording the upload and review frequency of different categories of engineering data. It also involves analyzing historical error rates for various types of engineering data based on changes and upload and review frequencies. Different risk control values ​​are applied to different categories of engineering data for targeted analysis, reducing the impact of variations. The invention further analyzes revisions of various types of engineering data and sets different monitoring levels based on the risk control values. By combining the monitoring levels and historical error rates of various types of engineering data, different risk monitoring methods are selected to avoid the limitations of a uniform monitoring method. Monitoring results for each type of engineering data are obtained according to the corresponding risk monitoring method and transmitted to the user terminal. An inspection time is set, and user confirmation results are obtained within the inspection time. The monitoring efficiency of various engineering data risk monitoring methods is calculated, and the project completion progress is monitored. Based on the monitoring efficiency, different feedback data is collected to determine whether to change the corresponding risk monitoring methods for different categories of engineering data, thereby improving the efficiency and accuracy of engineering data monitoring. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of an engineering settlement management risk monitoring system according to the present invention.

[0039] Figure 2 This is a flowchart of an engineering settlement management risk monitoring system according to the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0041] This invention improves the monitoring efficiency and accuracy of engineering data by collecting data on changes in various types of engineering data and recording the upload and review frequency of different categories of engineering data. It also involves analyzing historical data to calculate the historical error rate of various types of engineering data, setting different risk control values ​​for different categories of engineering data based on changes and upload and review frequencies, statistically analyzing revisions of various types of engineering data, setting different monitoring levels based on the risk control values, and selecting different risk monitoring methods for various types of engineering data by combining the monitoring levels and historical error rates. The invention then transmits the monitoring results of each type of engineering data to the user terminal according to the corresponding risk monitoring method, sets a verification time, obtains user confirmation results within the verification time, calculates the monitoring efficiency of various engineering data risk monitoring methods, and retrieves project completion progress data. Finally, it collects different feedback data based on the monitoring efficiency and determines whether to change the corresponding risk monitoring methods for different categories of engineering data.

[0042] Example: An engineering settlement management risk monitoring system, such as... Figure 1 and Figure 2 As shown, it includes a data acquisition module, a monitoring selection module, a risk handling module, and a feedback response module:

[0043] The data acquisition module is used to collect changes in various types of engineering data in the engineering management system and record the upload and review frequency of different categories of engineering data, which is then passed to the monitoring selection module; historical data is retrieved to calculate the historical error rate of various types of engineering data and sent to the risk processing module.

[0044] The monitoring selection module is used to receive changes to various types of engineering data and the frequency of upload review. It sets risk braking values ​​for different categories of engineering data, counts the revisions of various types of engineering data, and sets different monitoring levels based on the risk braking values. The monitoring levels of various types of engineering data are then passed to the risk processing module.

[0045] The risk processing module selects different engineering data risk monitoring methods, such as risk matrix or scenario simulation, based on the monitoring level and historical error rate of various types of engineering data. For each type of engineering data, the monitoring results are obtained according to the corresponding risk monitoring method and sent to the user terminal. After the user terminal confirms the monitoring results, it obtains the confirmation results and sends them to the feedback response module.

[0046] The feedback response module sets the inspection time, acquires multiple confirmation results within the inspection time, calculates the monitoring efficiency of various engineering data risk monitoring methods, calls the project completion progress, collects different feedback data based on the comprehensive monitoring efficiency and project completion progress, and determines whether to change the risk monitoring methods corresponding to different categories of engineering data.

[0047] The specific implementation is as follows:

[0048] The data acquisition module monitors the business database, records data snapshots before and after changes, captures data change events and upload review events and records timestamps, and calculates the magnitude of data changes and the frequency of upload review for various types of data.

[0049] The magnitude of data change refers to the extent of change in each type of data at different points in time. It is represented by the relative difference between the data values ​​before and after the change, and the specific formula is as follows: In the formula, 'a' represents the magnitude of data change. The changed data value. The data value before the change;

[0050] Upload review frequency refers to the number of upload review events for each type of data within a specific time period. It is expressed as a percentage of the number of upload review events for data within a certain time window to the duration of the time window. The calculation first selects a time interval as the time window, and then calculates the upload review frequency by dividing the number of upload review events within the time window by the time interval. The specific formula is as follows: In the formula, f represents the upload review frequency. This refers to the number of times data is uploaded and reviewed within a day. The time interval for a certain time window.

[0051] For example, if the time window interval is set to 24 hours and the number of data uploads and reviews per day is 15, then the upload review frequency obtained by the above formula is 62.5%. The time window interval setting is not unique and can be adjusted according to the actual situation, which will not be analyzed here.

[0052] The upload review frequency data and the magnitude of data changes are recorded together and used as parameters input to the monitoring selection module.

[0053] It should be noted that the business database refers to the database that stores the main business data of the engineering settlement management risk monitoring system, the data snapshot refers to a static copy of the data at a specific point in time, and the timestamp is an identifier used by those skilled in the art to record the time when a specific event or operation occurs, which will not be elaborated here.

[0054] In addition, the data acquisition module retrieves historical data from the business database within a certain time window to calculate the historical error rate for various data types. After calculation, the historical error rate is transmitted to the risk processing module. The historical error rate refers to the proportion of errors occurring for each data type within a certain time window, calculated by comparing the number of erroneous changes to the total number of changes. During calculation, the data acquisition module first compares each change record with a preset threshold. When the data change exceeds the preset threshold, the system marks it as an "erroneous change."

[0055] The preset threshold is based on the 90th quantile of historical data. The data acquisition module sorts all calculated data variation ranges in ascending order to obtain an ordered list of variation ranges. ,in Next, determine the position of the quantile using the formula: If k is an integer, then a preset threshold is used. If k is a decimal, that is, the position is Where d is the decimal part, the preset threshold is calculated using interpolation. .

[0056] The data acquisition module uses the ratio of past erroneous changes to the total number of changes to represent the historical error rate. The specific calculation formula is as follows: In this formula, This refers to the number of data changes that were previously marked as "errors". This represents the total number of data changes within a specific time window in the past. For example, if there were 100 data change records of a certain type in the past 24 hours, and 5 of them were judged to be incorrect, then the historical error rate of that type of data is 5%.

[0057] After receiving the data change magnitude and upload review frequency, the monitoring selection module sets a corresponding risk braking value for each data category. The risk braking value refers to the threshold at which the data change magnitude of a certain type of data triggers a risk warning or other risk response measures. When the data change magnitude is lower than the risk braking value, it is considered a low value, indicating that the data is relatively stable and the monitoring requirements are low. When the data change magnitude is higher than the risk braking value, it is considered a high value, indicating that the risk is higher and sensitive monitoring is required.

[0058] The risk braking value is calculated based on twice the standard deviation of the data change magnitude and the upload review frequency. First, the mean and standard deviation of the data change magnitude are calculated; the formula for the mean is as follows: In the formula, This represents the average of the data change rates. For each data change range, where n is the total number of historical data change ranges, its standard deviation is calculated using the following formula: Where σ is the standard deviation of the data change magnitude. This represents the average of the data change rates. Let n represent the magnitude of each data change, and n be the total number of historical data change magnitudes. The initial risk braking value is obtained by adding twice the standard deviation to the mean of the data change magnitudes, calculated using the following formula: ,in, This is the initial risk braking value. Let σ be the mean of the data change magnitude, and σ be the standard deviation of the data change magnitude. After calculating the initial risk braking value, adjust the risk braking value based on the upload review frequency, defining an adjustment factor. The impact of upload review frequency on the risk braking value is indicated by the following formula: ,in, To adjust the frequency of uploads, f represents the upload review frequency. This is an adjustment factor, derived from experiments by professionals in the field, and will not be elaborated upon here. As the upload review frequency (f) increases, the adjustment factor... A value approaching 0 indicates lower risk, and the initial risk braking value decreases accordingly; as the upload review frequency f decreases, the adjustment factor... A value close to 1 indicates a higher risk, and the initial risk braking value should be increased accordingly. Adjustment factors will be adjusted accordingly. The initial risk braking value is used to derive the risk braking value, and the calculation formula is as follows: ,in, This is the risk braking value. This is the initial risk braking value. This is an adjustment factor.

[0059] Additionally, the monitoring selection module reviews erroneous change records within a certain time window in the past, identifies and corrects them. Correction refers to modifying data marked as "erroneous changes" to data that no longer exceeds a preset threshold. The correction rate reflects the extent of data correction; its value is the proportion of corrected data in the historical erroneous data. The specific calculation formula is... Where L represents the correction rate, Indicates the number of data points to be corrected. This indicates the total number of historical erroneous data points.

[0060] After obtaining the revision rate for each type of data, the monitoring selection module sets the monitoring index based on the previously calculated risk braking value. This is used to determine the monitoring level of the data.

[0061] The monitoring selection module uses statistical analysis tools to build a linear regression model and fits the monitoring index using the least squares method. The relationship between revision rate and risk braking value is expressed by the following formula: ,in, It is the target variable, namely the monitoring index. It is a risk braking value. It's the correction rate. It is a constant term (intercept). , It is the regression coefficient, representing and The degree of influence on the target variable, This is the error term, used to reduce error. After the regression model is established, the monitoring selection module will train the model using data within a certain time window, adjust the regression coefficients using the training data, and then input the correction rate and risk braking value of each type of data into the regression equation to calculate the monitoring index of each type of data. .

[0062] It should be noted that the statistical analysis tools are tools used within the monitoring selection module to collect, organize, analyze, and interpret data.

[0063] The monitoring selection module calculates the monitoring index for different engineering data. Calculate multiple monitoring indices mean And Set as the threshold for level division.

[0064] when This indicates that the data changes are relatively small and the risk is low, thus it is classified as a low monitoring level.

[0065] when This indicates significant data fluctuations and a high risk level, thus classifying it as a high-level monitoring situation.

[0066] Based on the analysis results, the monitoring selection module transmits the monitoring level of various types of data to the risk processing module.

[0067] The risk handling module selects a risk monitoring method based on the monitoring level and historical error rate of various data types. The threshold for determining the historical error rate is also taken as the average of historical error rates over multiple identical time intervals within different time windows. When a data type has a high monitoring level and its historical error rate is greater than its average, the risk handling module will select the periodic risk review method. When a data type has a low monitoring level and its historical error rate is less than its average, the risk handling module will select the risk matrix method.

[0068] After each type of data is processed by the risk processing module and the corresponding risk monitoring method is applied, whenever a data change operation is performed, the risk processing module will evaluate the magnitude of the data change. If the magnitude of the data change is greater than or equal to the risk braking value, it will be marked as risky data; if the magnitude of the data change is less than the risk braking value, it will be marked as non-risky data.

[0069] When the risk handling module adopts the periodic risk review method, the risk handling module periodically sends the values ​​of all changed data and their marking status (i.e., risk data or non-risk data) within the time window as monitoring results to the user terminal, and simultaneously obtains multiple review results from the user terminal. The review results include two options: "yes" or "no".

[0070] If the audit result is "Yes", it means the risk labeling is correct; if the audit result is "No", it means the risk labeling is incorrect. When the audit result is "No", the risk processing module will modify the labeling of the corresponding changed data, that is, relabel the data that was originally labeled as risk data as non-risk data, and relabel the data that was originally labeled as non-risk data as risk data.

[0071] When the risk processing module adopts the periodic risk review method, the risk matrix classifies the data change range of various types of data according to the probability of risk occurrence, forming a two-dimensional matrix. Under this method, when the risk processing module marks the changed data as risk data, it sends the data and the risk marking status (i.e., risk data) as monitoring results to the user terminal and obtains an audit result from the user terminal. The audit result includes two options: "yes" or "no".

[0072] If the audit result is "Yes", it means the risk labeling is correct; if the audit result is "No", it means the risk labeling is incorrect. When the audit result is "No", the risk processing module will re-label this changed data as non-risk data.

[0073] After the user reviews the monitoring results, the risk handling module captures the review results and sends them to the feedback response module.

[0074] It should be noted that periodic risk review is a risk management method based on periodic assessment. It has high monitoring costs but high monitoring accuracy and is suitable for monitoring high-risk engineering data. Risk matrix is ​​a quantitative analysis tool based on probability and impact measurement. It has low monitoring costs but low monitoring accuracy and is suitable for monitoring low-risk engineering data. A two-dimensional matrix is ​​a data structure, usually used to represent a two-dimensional table or array with rows and columns. The difference between periodic risk review and risk matrix methods is that periodic review requires periodic review of all data, while risk matrix methods only review data marked as risk in a single instance.

[0075] The feedback response module sets an inspection time window and statistically analyzes the confirmation results within that window. Based on the collected review results, it calculates the monitoring efficiency of various engineering data risk monitoring methods. Monitoring efficiency combines the proportion of risk data among all change data within the time window with the risk data labeling accuracy rate (i.e., the proportion of "yes" reviews), multiplying these two factors during the calculation. The calculation formula is as follows: ,in, For monitoring efficiency, The amount of unaudited risk data. For the total number of changed data, The number of risk data points for which the audit result is "yes". The total number of data points marked as risky, including risky data with an audit result of "yes" and non-risky data with an audit result of "no".

[0076] While evaluating monitoring efficiency, the feedback response module retrieves the project completion progress from the business database. The formula for calculating the project completion progress is as follows: ,in, To ensure the project is completed on schedule, This represents the amount of work already completed. This represents the total workload.

[0077] The feedback response module combines monitoring efficiency and project completion progress feedback data to determine whether various project data require adjustments to the risk monitoring method. In the early stages of project progress, due to high uncertainty, frequent changes, and potential design and planning flaws, data risk is high, requiring detailed monitoring through regular risk reviews. In the later stages, as the project gradually enters its final stages, most risks have been resolved or foreseen, and data changes are stable. Therefore, the monitoring method needs to be adjusted to a risk matrix approach for lightweight monitoring, avoiding excessive intervention.

[0078] The determination of the monitoring method adjustment is based on the adjustment threshold of monitoring efficiency. The calculation of the adjustment threshold is based on the K-means clustering algorithm. First, a clustering model is established, and the data is divided into K clusters using the K-means clustering algorithm. The Euclidean distance between the centers of adjacent clusters is then calculated, and the formula is as follows:

[0079] ,

[0080] in, and It represents the progress and monitoring efficiency of the k-th cluster center. and It represents the progress and monitoring efficiency of the (k+1)th cluster center.

[0081] After calculating the distance between all adjacent cluster centers, the boundary monitoring efficiency value between the two clusters corresponding to the maximum distance is used as the adjustment threshold for adjusting the monitoring method;

[0082] If the monitoring efficiency is less than the adjustment threshold, it means that the monitoring method matches the project completion progress, and the current risk monitoring method continues to be used. If the monitoring efficiency is greater than the adjustment threshold, it means that the monitoring method does not match the project completion progress, and the feedback response module switches the current risk monitoring method to the risk matrix method and pushes it to the risk processing module.

[0083] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0084] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0085] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0086] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0087] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A project settlement management risk monitoring system, characterized in that, It includes a data acquisition module, a monitoring selection module, a risk handling module, and a feedback response module: The data acquisition module is used to collect changes in various types of engineering data in the engineering management system and record the upload and review frequency of different categories of engineering data; it also calls historical data to calculate the historical error rate of various types of engineering data. Changes in engineering data are described as data change magnitudes. When each type of data changes at different points in time, the relative difference is used to represent the data change magnitude. In the formula, 'a' represents the magnitude of data change. The changed data value. The data value before the change; Upload review frequency refers to the number of upload review events for each type of data within a preset time period. A time interval is selected as the time window, and the upload review frequency is obtained by dividing the number of upload review events for data within the time window by the time interval of the time window. The historical error rate is the proportion of errors in each data category, calculated by comparing the number of erroneous changes to the total number of changes within a past time window. The data acquisition module first compares each change record with a preset threshold. When the data change exceeds the preset threshold, the system marks it as an erroneous change. The ratio of the number of erroneous changes to the total number of changes is used as the historical error rate for various types of engineering data; The monitoring selection module is used to receive changes to various types of engineering data and the frequency of upload and review. It sets risk braking values ​​for different categories of engineering data, counts the revisions of various types of engineering data, and sets different monitoring levels based on the risk braking values. When setting a risk braking value, first calculate the mean and standard deviation of the data change range, and add twice the standard deviation to the mean of the data change range to obtain the preliminary risk braking value. Adjusting the risk threshold based on upload review frequency; defining adjustment factors. The impact of upload review frequency on the risk braking value is indicated by the following formula: ,in, To adjust the frequency of uploads, f represents the upload review frequency. It is an adjustment factor; Adjustment factor The initial risk braking value is used to derive the risk braking value, and the calculation formula is as follows: ,in, This is the risk braking value. This is the initial risk braking value. For adjustment factors; The risk processing module selects different engineering data risk monitoring methods based on the monitoring level and historical error rate of various types of engineering data. Each type of engineering data obtains monitoring results according to the corresponding risk monitoring method and sends them to the user terminal. The user terminal confirms the monitoring results and obtains the confirmation results. The feedback response module sets the inspection time, obtains multiple confirmation results, calculates the monitoring efficiency of various engineering data risk monitoring methods, calls the project completion progress, collects different feedback data in combination with the monitoring efficiency, and determines whether to change the risk monitoring methods corresponding to different categories of engineering data.

2. The project settlement management risk monitoring system according to claim 1, characterized in that: The monitoring selection module sets the correction rate to correct the engineering data. Where L represents the correction rate, Indicates the number of corrected engineering data. This indicates the total number of historical erroneous data points. After obtaining the revision rate for each type of data, the monitoring selection module sets the monitoring index based on the risk braking value. This is used to determine the monitoring level of the data.

3. The project settlement management risk monitoring system according to claim 2, characterized in that: The monitoring selection module establishes a linear regression model and fits the monitoring index using the least squares method. The relationship between revision rate and risk braking value is expressed by the following formula: ,in, It is the target variable, namely the monitoring index. It is a risk braking value. It's the correction rate. It is a constant term. , These are the preset regression coefficients, representing and The degree of influence on the target variable, It is an error term; After the regression model is established, the monitoring selection module inputs the correction rate and risk braking value of various types of data into the regression equation to calculate the monitoring index of various types of data. .

4. The project settlement management risk monitoring system according to claim 3, characterized in that: The monitoring selection module calculates the monitoring index for different engineering data. Calculate multiple monitoring indices mean And Set as the threshold for level division. when This indicates that the data changes are relatively small and the risk is low, thus it is classified as a low monitoring level. when This indicates significant data fluctuations and a high risk level, thus classifying it as a high-level monitoring situation.

5. The project settlement management risk monitoring system according to claim 4, characterized in that: When the monitoring level of engineering data is high and the historical error rate is greater than its average, the risk processing module will select the periodic risk review method; when the monitoring level of a certain type of data is low and the historical error rate is less than its average, the risk processing module will select the risk matrix method. Whenever a data change operation is performed, the risk handling module will evaluate the magnitude of the data change. If the magnitude of the data change is greater than or equal to the risk braking value, it will be marked as risky data; if the magnitude of the data change is less than the risk braking value, it will be marked as non-risky data. The risk handling module sends the values ​​of all changed data within the time window as monitoring results to the user terminal, and simultaneously obtains the audit results from the user terminal. The audit results include two options: "yes" or "no". When the audit result is "No", the risk handling module will re-mark this changed data as non-risk data.

6. The project settlement management risk monitoring system according to claim 5, characterized in that: Monitoring efficiency is calculated by combining the percentage of unaudited risk data in all change data within that time window and the accuracy of risk data labeling: the formula is as follows. ,in, For monitoring efficiency, The amount of unaudited risk data. For the total number of changed data, The number of risk data points for which the audit result is "yes". The total number of data points marked as risky.

7. The project settlement management risk monitoring system according to claim 6, characterized in that: The determination of the monitoring method adjustment is based on an adjustment threshold for monitoring efficiency. This threshold is used to divide the data into K clusters using a K-means clustering algorithm, and the Euclidean distance between the centers of adjacent clusters is calculated. The formula is as follows: ， in, and It represents the progress and monitoring efficiency of the k-th cluster center. and It represents the progress and monitoring efficiency of the (k+1)th cluster center; After calculating the distance between all adjacent cluster centers, the boundary monitoring efficiency value between the two clusters corresponding to the maximum distance is used as the adjustment threshold for adjusting the monitoring method; If the monitoring efficiency is less than the adjustment threshold, continue using the current risk monitoring method. If the monitoring efficiency is greater than the adjustment threshold, switch the current risk monitoring method to the risk matrix method and push it to the risk processing module.

Images