An engineering cost data management method and system

Abstract

The application relates to the field of engineering cost, and discloses an engineering cost data management method and system, which comprises the following steps: acquiring real-time collected cost data; if the cost data exceeds the monitoring threshold range of the corresponding project type, marking the cost data as a potential abnormal point and obtaining an abnormal data list; judging whether the abnormal point is related to the project type or the construction scale size based on the abnormal data list, regional market fluctuation data and material supply change characteristics, and obtaining an abnormal reason classification result; constructing a warning rule customization mechanism for regional market fluctuation to generate and verify a warning rule set suitable for the regional characteristics; acquiring new real-time cost data; and if the new real-time cost data triggers the conditions in the warning rule set, generating a corresponding warning signal. The method has the following effects: effectively identifying and analyzing cost data abnormalities.

Description

Technical Field

[0001] This invention relates to the field of engineering cost, and in particular to a method and system for managing engineering cost data. Background Technology

[0002] Construction cost data management is a crucial area for cost control and resource optimization in the construction industry, directly impacting project economic benefits and sustainable development. As the construction industry expands and project complexity increases, effective cost data management becomes key to ensuring project success.

[0003] However, current cost management methods have significant limitations in addressing diverse project needs. Many systems rely on fixed price monitoring standards, making it difficult to adapt to differences in project type, regional characteristics, or construction scale. This results in insufficient precision in cost control, and early warning mechanisms often fail or become overly generalized, failing to provide targeted guidance for specific projects. These limitations prevent cost management from reaching its maximum effectiveness in complex and ever-changing project environments.

[0004] Solving the above-mentioned technical problems is a technical challenge that needs to be overcome by those skilled in the art. Summary of the Invention

[0005] This invention provides a method and system for managing engineering cost data, which at least partially solves the above-mentioned technical problems.

[0006] Firstly, in order to solve the above-mentioned technical problems, the present invention provides a method for managing engineering cost data, comprising:

[0007] Acquire real-time cost data; if the cost data exceeds the monitoring threshold range for the corresponding project type, mark it as a potential anomaly and obtain a list of abnormal data;

[0008] Based on the list of abnormal data, regional market fluctuation data and material supply change characteristics, it is determined whether the abnormal points are related to the project type or construction scale, and the abnormal cause classification results are obtained.

[0009] Based on the classification results of the abnormal causes, a mechanism for customizing early warning rules for regional market fluctuations is constructed to generate and verify an early warning rule set that is adapted to regional characteristics;

[0010] New real-time cost data is acquired. If the new real-time cost data triggers the conditions in the early warning rule set, a corresponding early warning signal is generated.

[0011] In one optional implementation, the monitoring threshold range is obtained through the following steps:

[0012] Historical data is extracted from a pre-defined project characteristic database; based on the extracted historical data, data on project type, regional market, material supply, and construction scale are obtained to get an initial dataset;

[0013] Based on the initial dataset, clustering analysis is used to group regional market fluctuations and material supply changes to obtain feature data groups. Based on each group of feature data, the price fluctuation range is determined to obtain a price benchmark interval associated with project type and construction scale.

[0014] A calculation model for monitoring thresholds is constructed based on the price benchmark range and project demand matching rules, and a monitoring threshold range applicable to different construction scales is determined.

[0015] In one optional implementation, cluster analysis is used to group regional market fluctuations and material supply changes based on the initial dataset to obtain feature data groups. Based on each group of feature data, a price fluctuation range is determined to obtain a price benchmark interval associated with project type and construction scale, including:

[0016] Feature data groups are obtained by grouping regional market fluctuations and changes in material supply.

[0017] Calculate the price fluctuation range for each group of characteristic data;

[0018] Analyze the correlation between feature data grouping and project type to establish a mapping relationship between project type and price fluctuation range;

[0019] Calculate the weighting coefficient of construction scale level on price fluctuation range;

[0020] The price benchmark range is generated by combining the price fluctuation range affected by the construction scale with the price fluctuation range associated with the project type and construction scale.

[0021] In one optional implementation, a calculation model for the monitoring threshold is constructed based on the price benchmark range and project demand matching rules, and a monitoring threshold range applicable to different construction scales is determined, including:

[0022] An initial price data set is extracted from the price benchmark range associated with project type and construction scale, and the mean and standard deviation of price fluctuations are calculated using statistical methods to obtain the benchmark price characteristics.

[0023] Based on benchmark price characteristics and project demand matching rules, a linear regression algorithm is used to predict price change trends and determine the initial monitoring threshold range.

[0024] The initial monitoring threshold range is adjusted based on the sensitivity parameter; the sensitivity parameter is used to characterize the impact of different sensitivity levels on the monitoring threshold.

[0025] The threshold adaptability for different construction scales is verified through simulation analysis, and monitoring threshold ranges adapted to different construction scales are generated based on the simulation results.

[0026] In one optional implementation, based on the abnormal data list, regional market fluctuation data, and material supply change characteristics, it is determined whether the anomalies are related to the project type or construction scale, resulting in an anomaly cause classification result, including:

[0027] Key feature values ​​were extracted from the list of abnormal data and combined with time-series data on regional market fluctuations and changes in material supply to form a preliminary abnormal dataset.

[0028] A data matching method is used to associate abnormal datasets with material supply change records, and a feature set of associations between abnormal datasets and supply changes is constructed.

[0029] A multivariate linear regression model is constructed based on the associated feature set. Abnormal data is used as the dependent variable, and regional market volatility, material supply change rate, project type and construction scale are used as independent variables. The regression coefficient and significance test P value are calculated to determine whether the abnormality is related to the project type.

[0030] If the regression coefficient of the project type is significant, the construction scale data is extracted, and the distribution characteristics of the anomaly under different scales are calculated and analyzed by scatter plot and correlation coefficient. If the correlation coefficient between the construction scale and the anomaly is greater than the preset value, it is determined that there is a direct relationship between the anomaly and the construction scale.

[0031] By combining regional market fluctuation data and material supply change characteristics, the causes of anomalies are classified through multi-dimensional statistical analysis.

[0032] An anomaly cause classification dataset is generated based on the classification results.

[0033] In one optional implementation, a mechanism for customizing early warning rules for regional market fluctuations is constructed based on the anomaly cause classification results to generate and verify an early warning rule set adapted to regional characteristics, including:

[0034] Extract feature data related to regional market fluctuations from the classification results of abnormal causes. If the classification results contain regional market fluctuation features, extract the corresponding regional feature data.

[0035] By acquiring market dynamic data of the target area through external data interfaces, analyzing the correlation between regional characteristic data and market dynamic data, and calculating fluctuation correlation parameters;

[0036] Adjust the trigger condition thresholds in the early warning rule set based on fluctuation correlation parameters;

[0037] The updated trigger condition thresholds are combined with the early warning rule set logic to generate early warning rules bound to regional characteristics;

[0038] Extract rule adjustment parameters from the generated early warning rule set, determine whether they meet the preset verification threshold requirements, and output the verified early warning rule set if they do.

[0039] In one optional implementation, new real-time cost data is acquired. If the new real-time cost data triggers a condition in the early warning rule set, a corresponding early warning signal is generated, including:

[0040] Real-time cost data streams, including material costs, labor costs, and equipment depreciation, are obtained from the cost project characteristics database.

[0041] Load a predefined alert rule set, which contains conditional expressions and trigger actions;

[0042] Analyze the cost data stream, calculate the total cost, and compare each item with the conditional expressions in the early warning rule set;

[0043] If the field values ​​in the cost data stream meet the rule conditions, they are determined to be abnormal data, and a corresponding warning signal is generated. The warning signal includes a timestamp, rule ID, and warning description.

[0044] The warning signal is pushed to the monitoring platform through a message queue.

[0045] Secondly, the present invention provides an engineering cost data management system, comprising:

[0046] The first processing module is used to: acquire real-time cost data; if the cost data exceeds the monitoring threshold range of the corresponding project type, mark it as a potential anomaly and obtain an anomaly data list;

[0047] The second processing module is used to: determine whether the anomalies are related to the project type or construction scale based on the anomaly data list, regional market fluctuation data and material supply change characteristics, and obtain the anomaly cause classification result;

[0048] The third processing module is used to: construct a customized early warning rule mechanism for regional market fluctuations based on the classification results of the abnormal causes, and generate and verify an early warning rule set that is adapted to the regional characteristics;

[0049] The fourth processing module is used to: acquire new real-time cost data; and generate a corresponding early warning signal if the new real-time cost data triggers the conditions in the early warning rule set.

[0050] Compared with existing technologies, the present invention has at least the following beneficial effects: effectively identifying and analyzing cost data anomalies, and improving the accuracy and timeliness of cost management. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating an engineering cost data management method provided in the first embodiment of the present invention;

[0052] Figure 2 This is a block diagram of an engineering cost data management system provided in the second embodiment of the present invention. Detailed Implementation

[0053] 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.

[0054] Reference Figure 1 The first embodiment of the present invention provides a method for managing engineering cost data, including the following steps:

[0055] S101, acquire real-time cost data; if the cost data exceeds the monitoring threshold range of the corresponding project type, mark it as a potential anomaly and obtain an anomaly data list;

[0056] S102, based on the abnormal data list, regional market fluctuation data and material supply change characteristics, determine whether the abnormal points are related to the project type or construction scale, and obtain the abnormal cause classification results;

[0057] S103, Based on the classification results of the abnormal causes, construct a mechanism for customizing early warning rules for regional market fluctuations, generate and verify an early warning rule set adapted to regional characteristics;

[0058] S104: Obtain new real-time cost data. If the new real-time cost data triggers the conditions in the early warning rule set, generate the corresponding early warning signal.

[0059] In one implementation, the monitoring threshold range is obtained through the following steps:

[0060] Historical data is extracted from a pre-defined project characteristic database; based on the extracted historical data, data on project type, regional market, material supply, and construction scale are obtained to get an initial dataset;

[0061] Based on the initial dataset, clustering analysis is used to group regional market fluctuations and material supply changes to obtain feature data groups. Based on each group of feature data, the price fluctuation range is determined to obtain a price benchmark interval associated with project type and construction scale.

[0062] A calculation model for monitoring thresholds is constructed based on the price benchmark range and project demand matching rules, and a monitoring threshold range applicable to different construction scales is determined.

[0063] Specifically, historical data is extracted from a pre-defined project characteristic database. This historical data includes project type, regional market fluctuation data, material supply change characteristics, and construction scale data (such as building area in tens of thousands of square meters). Through data cleaning and standardization, an initial dataset is formed. The project types mentioned above include residential and commercial; the regional market fluctuation data includes regional economic indices and price volatility; the material supply change characteristics include building material price indices and supply change rates; and the construction scale data can be understood as building area.

[0064] Cluster analysis was performed on the regional market volatility and material supply change characteristics in the initial dataset, using the K-means algorithm to divide the data into several feature groups. For example, based on the similarity between regional market volatility and material supply change rates, the data was divided into categories such as "high volatility - low supply" and "low volatility - high supply." Cluster analysis is an unsupervised learning method that divides data into clusters by calculating the similarity between data points, ensuring high similarity within the same cluster and significant differences between different clusters. Through cluster analysis, market behavior patterns in different regions or time periods can be identified.

[0065] For each group of characteristic data, calculate its price fluctuation range, such as maximum, minimum, mean, and standard deviation. Combine this with project type and construction scale to establish a mapping relationship between project type and construction scale. Generally, residential projects typically have a higher proportion of material costs, while large projects have a lower tolerance for price fluctuations. For example: Residential projects: material cost fluctuation range is large, with a baseline range of ±15%; Commercial projects: labor cost fluctuation is significant, with a baseline range of ±10%; Large projects: construction scale coefficient weight is 0.8, with a baseline range of ±8%; Small projects: construction scale coefficient weight is 0.5, with a baseline range of ±12%.

[0066] By combining the price fluctuation range affected by construction scale with the price fluctuation range associated with project type, a price benchmark range is generated. This benchmark range is a reasonable price fluctuation range calculated based on historical data and project characteristics, serving as a benchmark for monitoring thresholds. For example, the price benchmark range for residential projects might be ±15%, while for commercial projects it might be ±10%.

[0067] Based on a price benchmark range, the mean and standard deviation of price fluctuations are calculated using statistical methods to serve as benchmark price characteristics. For example, if the historical average price of residential projects in a certain area is 1 million yuan and the standard deviation is 100,000 yuan, then the benchmark price characteristic is (1 million ± 100,000 yuan).

[0068] Based on benchmark price characteristics and project demand matching rules, such as "residential projects need to strictly control material cost fluctuations", a linear regression algorithm is used to predict price change trends and determine the initial monitoring threshold range. For example, if the linear regression model predicts that future price fluctuations may expand to ±18%, the initial threshold range is set to ±18%.

[0069] An initial monitoring threshold range is adjusted based on a sensitivity parameter to balance the false alarm rate and the false negative rate. The sensitivity parameter is a dynamic coefficient used to adjust the monitoring threshold, typically ranging from 0.8 to 1.2. A higher parameter results in a more lenient threshold, tolerating greater fluctuations, but may increase the false alarm rate; a lower parameter results in a more stringent threshold, but may increase the false negative rate. For example, if the sensitivity parameter is 1.2, the threshold range is adjusted to ±21.6%.

[0070] The suitability of thresholds for different construction scales is verified through simulation analysis. For example, simulation tests are conducted on large projects with a construction scale of >100,000 square meters and small projects with a construction scale of <20,000 square meters. If the simulation results show that the threshold for large projects needs to be relaxed to ±20%, while the threshold for small projects needs to be tightened to ±10%, then the monitoring threshold range suitable for different construction scales is finally generated.

[0071] In one implementation, based on the initial dataset, cluster analysis is used to group regional market fluctuations and material supply changes to obtain feature data groups. Based on each group of feature data, a price fluctuation range is determined to obtain a price benchmark interval associated with project type and construction scale, including:

[0072] Feature data groups are obtained by grouping regional market fluctuations and changes in material supply.

[0073] Calculate the price fluctuation range for each group of characteristic data;

[0074] Analyze the correlation between feature data grouping and project type to establish a mapping relationship between project type and price fluctuation range;

[0075] Calculate the weighting coefficient of construction scale level on price fluctuation range;

[0076] The price benchmark range is generated by combining the price fluctuation range affected by the construction scale with the price fluctuation range associated with the project type and construction scale.

[0077] Specifically, the regional market fluctuations in the initial dataset, such as regional economic indices and price volatility, and material supply changes, such as building material price indices and supply change rates, are grouped. The K-means algorithm is used to divide the data into several feature data groups. By calculating the similarity between data points, the data is divided into several clusters, ensuring high similarity within the same cluster and significant differences between different clusters. For example, based on the similarity between regional market volatility and material supply change rates, the data is divided into categories such as "high volatility - low supply" and "low volatility - high supply".

[0078] For each data feature group, such as the "high volatility - low supply" group, calculate its price fluctuation range, which includes the maximum, minimum, mean, and standard deviation of the price.

[0079] Analyze the correlation between characteristic data groupings and project types to establish a mapping relationship. For example, residential projects typically have a high proportion of material costs and are strongly correlated with the "high volatility - low supply" group; commercial projects, on the other hand, have significant fluctuations in labor costs and are more strongly correlated with the "low volatility - high supply" group. Verify the significance of the relationship between characteristic data groupings and project types through statistical tests, and establish a mapping relationship between project types and price fluctuation ranges. For example, the price fluctuation range for residential projects is ±15%, while the price fluctuation range for commercial projects is ±10%.

[0080] The weighting coefficient for price fluctuation range is calculated based on the construction scale level. The construction scale level is a category divided according to the project's building area or investment amount. For example: large projects with 100,000 square meters: construction scale coefficient weight is 0.8; medium-sized projects with 30,000 to 100,000 square meters: construction scale coefficient weight is 0.5; small projects with less than 30,000 square meters: construction scale coefficient weight is 0.3.

[0081] The weighting coefficient reflects the impact of construction scale on the tolerance for price fluctuations. Large projects, due to their high resource concentration, have a lower tolerance for price fluctuations and therefore a smaller weighting coefficient; while small projects, due to their greater flexibility, have a higher tolerance and therefore a larger weighting coefficient.

[0082] The price fluctuation range influenced by construction scale is combined with the price fluctuation range associated with project type to generate the final price benchmark range. For example: for large residential projects, the construction scale coefficient has a weight of 0.8, the project type benchmark range is ±15%, and the comprehensive benchmark range is ±12%; for small commercial projects, the construction scale coefficient has a weight of 0.3, the project type benchmark range is ±10%, and the comprehensive benchmark range is ±3%. Through weighted calculation, the price benchmark range is ensured to reflect the impact of both project type characteristics and construction scale, improving the adaptability of monitoring thresholds.

[0083] In one implementation, a calculation model for the monitoring threshold is constructed based on the price benchmark range and project demand matching rules, and a monitoring threshold range applicable to different construction scales is determined, including:

[0084] An initial price data set is extracted from the price benchmark range associated with project type and construction scale, and the mean and standard deviation of price fluctuations are calculated using statistical methods to obtain the benchmark price characteristics.

[0085] Based on benchmark price characteristics and project demand matching rules, a linear regression algorithm is used to predict price change trends and determine the initial monitoring threshold range.

[0086] The initial monitoring threshold range is adjusted based on the sensitivity parameter; the sensitivity parameter is used to characterize the impact of different sensitivity levels on the monitoring threshold.

[0087] The threshold adaptability for different construction scales is verified through simulation analysis, and monitoring threshold ranges adapted to different construction scales are generated based on the simulation results.

[0088] Specifically, an initial price data set is extracted from the price benchmark range, and statistical methods, such as mean calculation and standard deviation analysis, are used to quantify the price fluctuation characteristics. For example, if the price benchmark range for residential projects in a certain area is ±15%, the historical average price is 1 million yuan, and the standard deviation is 100,000 yuan, then the benchmark price characteristic can be expressed as 1,000,000 ± 100,000 yuan. The benchmark price characteristic is calculated through statistical methods and reflects the core trend and dispersion of price fluctuations.

[0089] Combining benchmark price characteristics and project demand matching rules, such as "residential projects must strictly control material cost fluctuations," a linear regression algorithm is used to predict price change trends. For example, if the linear regression model predicts that future price fluctuations may widen to ±18%, then the initial monitoring threshold range is set to ±18%.

[0090] An approach based on sensitivity parameters is introduced to adjust the initial monitoring threshold range in order to balance the false alarm rate and the false negative rate.

[0091] The adaptability of thresholds for different construction scales is verified by simulation analysis. For example, simulation tests are conducted on large projects and small projects respectively. If the simulation results show that the threshold for large projects needs to be relaxed to ±20%, while the threshold for small projects needs to be tightened to ±10%, then the monitoring threshold range adapted to different construction scales is finally generated.

[0092] In one implementation, based on the abnormal data list, regional market fluctuation data, and material supply change characteristics, it is determined whether the anomalies are related to the project type or construction scale, resulting in an anomaly cause classification result, including:

[0093] Key feature values ​​were extracted from the list of abnormal data and combined with time-series data on regional market fluctuations and changes in material supply to form a preliminary abnormal dataset.

[0094] A data matching method is used to associate abnormal datasets with material supply change records, and a feature set of associations between abnormal datasets and supply changes is constructed.

[0095] A multivariate linear regression model is constructed based on the associated feature set. Abnormal data is used as the dependent variable, and regional market volatility, material supply change rate, project type and construction scale are used as independent variables. The regression coefficient and significance test P value are calculated to determine whether the abnormality is related to the project type.

[0096] If the regression coefficient of the project type is significant, the construction scale data is extracted, and the distribution characteristics of the anomaly under different scales are calculated and analyzed by scatter plot and correlation coefficient. If the correlation coefficient between the construction scale and the anomaly is greater than the preset value, it is determined that there is a direct relationship between the anomaly and the construction scale.

[0097] By combining regional market fluctuation data and material supply change characteristics, the causes of anomalies are classified through multi-dimensional statistical analysis.

[0098] An anomaly cause classification dataset is generated based on the classification results.

[0099] Specifically, key characteristic values, such as price volatility and material supply change rate, are extracted from the list of abnormal data. Combined with regional market fluctuation data, such as regional economic indices, price volatility, and time-series data on material supply changes, such as building material price indices and supply change rates, a preliminary abnormal dataset is formed.

[0100] A data matching method is used to associate abnormal datasets with material supply change records, and a feature set of associations between abnormal datasets and supply changes is constructed.

[0101] A multivariate linear regression model was constructed based on the associated feature set, using outlier data as the dependent variable and regional market volatility, material supply change rate, project type, and construction scale as independent variables. Regression coefficients and significance tests (P-values) were calculated. If the P-value corresponding to the regression coefficient for project type is less than 0.05, the outlier is considered significantly correlated with the project type.

[0102] If the regression coefficient of the project type is significant, further extract the construction scale data, and calculate and analyze the distribution characteristics of anomalies under different scales through scatter plots and correlation coefficients.

[0103] Combining regional market fluctuation data and material supply change characteristics, the causes of anomalies are classified through multi-dimensional statistical analysis: market-driven anomalies: anomalies mainly caused by regional market fluctuations; supply-driven anomalies: anomalies mainly caused by changes in material supply; scale-driven anomalies: anomalies strongly correlated with construction scale.

[0104] By verifying the independence between different dimensions through cross-analysis, an anomaly cause classification dataset is finally generated.

[0105] Based on the classification results, a structured anomaly cause classification dataset is generated. The anomaly cause classification dataset includes anomaly types, such as market-driven, supply-driven, and scale-driven, as well as related factors, such as project type, construction scale, and confidence level.

[0106] In one implementation, a mechanism for customizing early warning rules for regional market fluctuations is constructed based on the classification results of the abnormal causes, generating and validating an early warning rule set adapted to regional characteristics, including:

[0107] Extract feature data related to regional market fluctuations from the classification results of abnormal causes. If the classification results contain regional market fluctuation features, extract the corresponding regional feature data.

[0108] By acquiring market dynamic data of the target area through external data interfaces, analyzing the correlation between regional characteristic data and market dynamic data, and calculating fluctuation correlation parameters;

[0109] Adjust the trigger condition thresholds in the early warning rule set based on fluctuation correlation parameters;

[0110] The updated trigger condition thresholds are combined with the early warning rule set logic to generate early warning rules bound to regional characteristics;

[0111] Extract rule adjustment parameters from the generated early warning rule set, determine whether they meet the preset verification threshold requirements, and output the verified early warning rule set if they do.

[0112] Specifically, feature data related to regional market fluctuations, such as regional economic indices and price volatility, are extracted from the classification results of abnormal causes. If the classification results include regional market fluctuation characteristics, the corresponding regional feature data is extracted. Regional feature data is a core indicator reflecting the market dynamics of a specific region, such as the monthly volatility of the building materials price index in a certain city.

[0113] Market dynamic data for the target region, such as real-time building material price indices and regional economic activity indices, is obtained through external data interfaces. The correlation between regional characteristic data and market dynamic data is analyzed, and volatility correlation parameters are calculated. If the correlation coefficient between the volatility of a region's building material price index and its regional economic index is 0.75, it indicates a strong positive correlation between the two. This correlation coefficient can be calculated using the Pearson correlation coefficient. ;in, and These represent a pair of observations, one representing the volatility of the building materials price index and the other representing the regional economic index. and These are the average values ​​of the building materials price index volatility and the regional economic index, respectively. The correlation coefficient is calculated, and its value ranges from -1 to +1. Data on the volatility of the building materials price index and the regional economic index for a given period are collected. The sum of the products of the deviations of each pair of observed values ​​from their respective averages, as well as the square root of the sum of the squares of the deviations of each variable, are calculated using the formula above. The final result is the Pearson correlation coefficient, r. When r is close to +1 or -1, it indicates a strong linear relationship between the two variables; a positive value indicates a positive correlation, and a negative value indicates a negative correlation. When the absolute value of r is close to 0, it means that there is almost no linear relationship between the two variables.

[0114] Adjust the trigger condition thresholds of the early warning rules based on the volatility correlation parameters. For example: in high-correlation areas (correlation coefficient > 0.7): relax the trigger condition thresholds (e.g., price fluctuation ± 10%) to avoid oversensitivity; in low-correlation areas (correlation coefficient < 0.3): tighten the trigger condition thresholds (e.g., price fluctuation ± 5%) to improve the accuracy of early warnings.

[0115] The updated trigger thresholds are combined with the early warning rule set logic to generate early warning rules bound to regional characteristics. For example: Rule 1: If the volatility of the building materials price index in a certain region is >10% and the economic index decreases by >5%, then the "market volatility early warning" is triggered; Rule 2: If the construction scale in a certain region is >100,000 square meters and the material supply decreases by >15%, then the "supply shortage early warning" is triggered.

[0116] Extract rule adjustment parameters from the generated warning rule set, such as trigger condition thresholds and correlation coefficients, and determine whether they meet the preset verification threshold requirements. If they do, output the verified rule set; otherwise, return to step three to readjust the thresholds.

[0117] In one implementation, new real-time cost data is acquired. If the new real-time cost data triggers a condition in the early warning rule set, a corresponding early warning signal is generated, including:

[0118] Real-time cost data streams, including material costs, labor costs, and equipment depreciation, are obtained from the cost project characteristics database.

[0119] Load a predefined alert rule set, which contains conditional expressions and trigger actions;

[0120] Analyze the cost data stream, calculate the total cost, and compare each item with the conditional expressions in the early warning rule set;

[0121] If the field values ​​in the cost data stream meet the rule conditions, they are determined to be abnormal data, and a corresponding warning signal is generated. The warning signal includes a timestamp, rule ID, and warning description.

[0122] The warning signal is pushed to the monitoring platform through a message queue.

[0123] Specifically, the system extracts data streams from the cost database in real time via an API interface. These data streams include material costs, labor costs, and equipment depreciation. For example, a project might currently have material costs of 1 million yuan, labor costs of 500,000 yuan, and equipment depreciation of 200,000 yuan, totaling 1.7 million yuan. The system predefines a series of warning rules based on different business logics. For example: Rule 1: If material costs exceed 60% of the total cost, a high material cost warning is triggered; Rule 2: If the total cost exceeds the budget by 1.5 million yuan, an over-budget warning is triggered. These rules are stored in XML or similar formats, containing specific conditional expressions and corresponding action instructions. The system parses the received data streams and compares each item against the specific conditions of each rule. In the example above, the calculated material cost percentage is 100 / 170 = 0.588, which does not trigger Rule 1, but the total cost of 1.7 million yuan does exceed the budget of 1.5 million yuan, triggering Rule 2. Once a field value meets any of the abnormal conditions set by a rule, it is marked as abnormal data and added to the abnormal data set. As shown in the example above, since the total cost exceeds the budget, this data will be considered abnormal data. For abnormal data, the system will obtain the relevant early warning rules and associate the abnormal data with the early warning rules through a rule matching process to form a rule trigger list. Based on the rule trigger list, the system generates early warning signals containing timestamps, rule IDs, and early warning descriptions. These early warning signals are transformed into visual information through the information output channel and pushed to the target system or platform for relevant personnel to view and take measures.

[0124] Reference Figure 2 The second embodiment of the present invention provides an engineering cost data management system, comprising:

[0125] The first processing module is used to: acquire real-time cost data; if the cost data exceeds the monitoring threshold range of the corresponding project type, mark it as a potential anomaly and obtain an anomaly data list;

[0126] The second processing module is used to: determine whether the anomalies are related to the project type or construction scale based on the anomaly data list, regional market fluctuation data and material supply change characteristics, and obtain the anomaly cause classification result;

[0127] The third processing module is used to: construct a customized early warning rule mechanism for regional market fluctuations based on the classification results of the abnormal causes, and generate and verify an early warning rule set that is adapted to the regional characteristics;

[0128] The fourth processing module is used to: acquire new real-time cost data; and generate a corresponding early warning signal if the new real-time cost data triggers the conditions in the early warning rule set.

[0129] It should be noted that the engineering cost data management system provided in this embodiment of the invention is used to execute all the process steps of the engineering cost data management method in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0130] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for managing engineering cost data, characterized in that, include: Obtain real-time cost data; If the cost data exceeds the monitoring threshold range for the corresponding project type, it is marked as a potential anomaly and a list of abnormal data is obtained. Based on the list of abnormal data, regional market fluctuation data and material supply change characteristics, it is determined whether the abnormal points are related to the project type or construction scale, and the abnormal cause classification results are obtained. Based on the anomaly cause classification results, a mechanism for customizing early warning rules for regional market fluctuations is constructed to generate and verify an early warning rule set adapted to regional characteristics. Specifically, this includes: extracting feature data related to regional market fluctuations from the anomaly cause classification results; if the classification results contain regional market fluctuation characteristics, then extracting the corresponding regional feature data; obtaining market dynamic data of the target region through an external data interface, analyzing the correlation between regional feature data and market dynamic data, and calculating fluctuation correlation parameters; adjusting the trigger condition thresholds in the early warning rule set based on the fluctuation correlation parameters; logically combining the updated trigger condition thresholds with the early warning rule set to generate early warning rules bound to regional characteristics; extracting rule adjustment parameters from the generated early warning rule set, determining whether they meet the preset verification threshold requirements, and if so, outputting the verified early warning rule set. New real-time cost data is acquired. If the new real-time cost data triggers the conditions in the early warning rule set, a corresponding early warning signal is generated. The monitoring threshold range is obtained through the following steps: Historical data is extracted from a pre-defined project characteristic database; based on the extracted historical data, data on project type, regional market, material supply, and construction scale are obtained to get an initial dataset; Based on the initial dataset, cluster analysis is used to group regional market fluctuations and material supply changes to obtain feature data groups. For each group of feature data, a price fluctuation range is determined to obtain a price benchmark interval associated with project type and construction scale. Specifically, this includes: grouping regional market fluctuations and material supply changes to obtain feature data groups; calculating the price fluctuation range for each feature data group; analyzing the correlation between feature data groups and project type to establish a mapping relationship between project type and price fluctuation range; calculating the weighting coefficient of construction scale level on price fluctuation range; and combining the price fluctuation range affected by construction scale with the price fluctuation range associated with project type and construction scale to generate a price benchmark interval. A calculation model for monitoring thresholds is constructed based on the price benchmark range and project demand matching rules, and monitoring threshold ranges applicable to different construction scales are determined. Specifically, this includes: extracting an initial price data set from the price benchmark range, calculating the mean and standard deviation of price fluctuations using statistical methods to obtain benchmark price characteristics; predicting price change trends using a linear regression algorithm based on the benchmark price characteristics and project demand matching rules to determine the initial monitoring threshold range; adjusting the initial monitoring threshold range based on sensitivity parameters; the sensitivity parameters are used to characterize the impact of different sensitivity levels on the monitoring thresholds; verifying the threshold adaptability for different construction scales through simulation analysis, and generating monitoring threshold ranges suitable for different construction scales based on the simulation results.

2. The method for managing engineering cost data according to claim 1, characterized in that, Based on the aforementioned list of abnormal data, regional market fluctuation data, and material supply change characteristics, it is determined whether the anomalies are related to the project type or construction scale, resulting in an anomaly cause classification, including: Key feature values ​​were extracted from the list of abnormal data and combined with time-series data on regional market fluctuations and changes in material supply to form a preliminary abnormal dataset. A data matching method is used to associate abnormal datasets with material supply change records, and a feature set of associations between abnormal datasets and supply changes is constructed. A multivariate linear regression model is constructed based on the associated feature set. Abnormal data is used as the dependent variable, and regional market volatility, material supply change rate, project type and construction scale are used as independent variables. The regression coefficient and significance test P value are calculated to determine whether the abnormality is related to the project type. If the regression coefficient of the project type is significant, the construction scale data is extracted, and the distribution characteristics of the anomaly under different scales are calculated and analyzed by scatter plot and correlation coefficient. If the correlation coefficient between the construction scale and the anomaly is greater than the preset value, it is determined that there is a direct relationship between the anomaly and the construction scale. By combining regional market fluctuation data and material supply change characteristics, the causes of anomalies are classified through multi-dimensional statistical analysis. An anomaly cause classification dataset is generated based on the classification results.

3. The engineering cost data management method according to claim 1, characterized in that, New real-time cost data is acquired. If the new real-time cost data triggers the conditions in the early warning rule set, a corresponding early warning signal is generated, including: Real-time cost data streams, including material costs, labor costs, and equipment depreciation, are obtained from the cost project characteristics database. Load a predefined alert rule set, which contains conditional expressions and trigger actions; Analyze the cost data stream, calculate the total cost, and compare each item with the conditional expressions in the early warning rule set; If the field values ​​in the cost data stream meet the rule conditions, they are determined to be abnormal data, and a corresponding warning signal is generated. The warning signal includes a timestamp, rule ID, and warning description. The warning signal is pushed to the monitoring platform through a message queue.

4. An engineering cost data management system, characterized in that, The method for implementing the engineering cost data management method as described in any one of claims 1-3 includes: The first processing module is used to: acquire real-time cost data; if the cost data exceeds the monitoring threshold range of the corresponding project type, mark it as a potential anomaly and obtain an anomaly data list; The second processing module is used to: determine whether the anomalies are related to the project type or construction scale based on the anomaly data list, regional market fluctuation data and material supply change characteristics, and obtain the anomaly cause classification result; The third processing module is used to: construct a customized early warning rule mechanism for regional market fluctuations based on the classification results of the abnormal causes, and generate and verify an early warning rule set that is adapted to the regional characteristics; The fourth processing module is used to: acquire new real-time cost data; and generate a corresponding early warning signal if the new real-time cost data triggers the conditions in the early warning rule set.

Images