Big data collaborative driving engineering cost checking system and method

Abstract

The application discloses a big data collaborative driving engineering cost checking system and method, relates to the technical field of engineering cost analysis, and constructs the cost checking object for the bill of quantities by uniformly marking and aligning the bill of quantities data, the measurement payment data and the cost release data in the engineering implementation period; on the basis, the quantity transmission characteristics and the structure constraint characteristics of the bill of quantities in the engineering quantity composition, the cost decomposition and the pricing structure are analyzed, the cost correlation relationship is established and the correlation strength factor is quantified; further, the analogy benchmark and the specification benchmark are collaboratively constructed by combining the historical engineering cost data and the current engineering cost specification, and the collaborative cost benchmark is formed according to the role difference of the bill of quantities in the cost correlation structure; finally, the structural checking, the abnormal identification and the benchmark dynamic updating of the engineering cost are realized through the deviation quantity calculation and the transmission consistency determination in the cost correlation relationship.

Description

Technical Field

[0001] This invention relates to the field of engineering cost analysis technology, specifically to an engineering cost verification system and method driven by big data collaboration. Background Technology

[0002] In the field of construction cost management, construction cost verification is a crucial step in ensuring the rationality of project investment and the compliance of costs. Existing construction cost verification methods often focus on individual bill of quantities items or partial costs, typically employing static comparisons, experience-based threshold judgments, or manual sampling. These methods struggle to systematically depict the inherent dependencies between bill of quantities items in terms of quantity composition, cost breakdown, and pricing structure. When construction cost deviations do not originate from a single bill of quantities item but are amplified step-by-step through quantity transfer or cost transmission between items, existing methods often fail to accurately identify their logical rationality.

[0003] On the other hand, while some cost verification methods incorporate historical project data or engineering cost standards as references, they typically treat these as isolated benchmarks, failing to consider the different roles of each item in the overall cost structure for coordinated constraints. This results in insufficient adaptability of the verification results to different project types, structural characteristics, and implementation cycles. Furthermore, existing methods lack a mechanism for dynamically correcting and continuously optimizing cost benchmarks based on verification results, making it difficult to form a closed-loop evolution system for engineering cost verification and hindering its refined application in complex engineering projects. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies by proposing a big data-driven collaborative engineering cost verification system and method.

[0005] The technical solution to achieve the purpose of this invention is as follows:

[0006] On the one hand, the big data-driven collaborative method for engineering cost verification includes the following steps:

[0007] Acquire bill of quantities data, measurement and payment data, and cost release data; apply unified identification and alignment to the bill of quantities item level for the three types of data to form cost verification objects oriented towards bill of quantities items;

[0008] Based on the cost verification object, a cost correlation table is constructed. The quantity transfer characteristics and structural constraint characteristics of the bill of quantities items in terms of quantity composition, cost decomposition and pricing structure are analyzed. Cost correlation relationships are established and correlation strength factors are quantified.

[0009] By combining historical engineering cost data to perform cost structure similarity analysis, an analogy benchmark is constructed. The constraints of engineering cost specification documents and cost release data are analyzed to construct a standard benchmark. The weights of the analogy benchmark and the standard benchmark are differentiated according to the structure of the list items and the correlation strength factor to form a collaborative cost benchmark.

[0010] Calculate the deviation of each item in the bill of quantities from the collaborative cost benchmark in terms of quantity, unit price, and cost composition. Combine this with the correlation strength factor to limit the transmission range and determine the transmission consistency to form an abnormal bill of quantities set.

[0011] Based on the anomaly list set, the historical engineering cost data is backfilled according to the anomaly type and cost structure characteristics, and the construction rules of the collaborative cost benchmark and the role weight of each list item are revised and updated.

[0012] Furthermore, cost correlations are established and quantified to obtain correlation strength factors, including:

[0013] Based on the cost association table, this study analyzes the dependencies between bill of quantities items in terms of quantity composition, cost breakdown, and pricing structure. Specifically, it sorts out the compositional relationships of bill of quantities items at the division, sub-item, and sub-item levels, identifies the dependency characteristics between bill of quantities items in terms of total quantity, proportional allocation, and structural constraints, and analyzes the cost transmission path between different bill of quantities items by combining the decomposition relationships of labor, materials, machinery, and fee elements in the comprehensive unit price composition. Furthermore, it identifies the linkage relationships between bill of quantities items at the pricing structure level by combining the consistency of quota sub-items, pricing basis, and fee rules, thus forming a set of cost association relationships that includes association type, association direction, and constraint conditions.

[0014] Based on cost correlation, a quantitative system is constructed to address the degree of dependence on changes in project quantity, the proportion of cost composition, and the coupling characteristics of pricing structure. The system quantifies the impact of changes in project quantity, cost proportion, and matching of pricing rules among the bill of quantities items. The quantitative results of each dimension are then standardized using a unified scale. Based on this, the multi-dimensional quantitative results are fused according to preset weights to obtain a correlation strength factor that characterizes the degree of correlation between cost elements among the bill of quantities items.

[0015] Furthermore, a collaborative cost benchmark will be established, including:

[0016] Based on cost correlation and correlation strength factors, cost structure similarity analysis is performed on historical engineering cost data. Reference engineering projects that match the current engineering project in terms of cost correlation structure are selected according to project type, cost structure characteristics and list item composition. The cost data of corresponding list items in the reference engineering projects are aggregated using the unified core identifier of the list items as an index to form an analogy benchmark.

[0017] Based on the project implementation cycle, we analyze the constraints of engineering cost specification documents and cost release data on pricing elements, time intervals, and applicable conditions. We then organize the applicable rules for pricing elements such as labor, materials, machinery, and fees in a structured manner to form a set of specification constraint rules. Finally, we encapsulate the rules according to the unified core identifier of the list items to form a specification benchmark.

[0018] Based on the structural role of the list items in the cost association structure, the roles of different list items in the overall cost system are distinguished, the weights of the analog benchmark and the standard benchmark on each list item are configured, and the weights are fine-tuned in combination with the association strength factor to obtain a collaborative cost benchmark for list items.

[0019] Furthermore, the deviation is calculated and the transmission consistency is determined, including:

[0020] Using the collaborative cost benchmark as a reference, the actual cost data of the bill of quantities items in terms of quantity, unit price, and cost composition are compared and analyzed. The deviation from the collaborative cost benchmark is calculated, and the direction of deviation, calculation basis, and benchmark value used are recorded simultaneously during the calculation process.

[0021] Based on the cost correlation, the transmission path of the deviation is sorted out, the transmission direction of the deviation under different correlation types is clarified, and the transmission range of the deviation is limited by the correlation strength factor. At the same time, the corresponding transmission attenuation coefficient is configured to correct the result of the deviation after transmission between list items, and the deviation after transmission is obtained.

[0022] Construct a criterion for determining the deviation threshold and the degree of matching in transmission, perform consistency determination on the adaptation relationship between the deviation amount after transmission and its transmission path in the cost association structure, and output a list of items that have inconsistencies in the cost association logic accordingly.

[0023] Furthermore, an anomaly list is created, including:

[0024] For items in the list that are determined to be logically inconsistent, they are classified and organized according to information such as anomaly type, related dimensions, deviation details and judgment basis to form a traceable anomaly list set;

[0025] In the anomaly list set, the unified core identifier of the list item, the list item name, structural role, association strength factor, anomaly dimension, deviation after propagation, anomaly description and associated list item identifier are recorded, and the anomaly information is sorted and encapsulated according to the unified core identifier.

[0026] Furthermore, construct analog benchmarks, including:

[0027] Historical project data is extracted from the historical engineering cost database, and the historical project data is processed by unified identification completion, list item-level alignment and data normalization to form a standardized historical project dataset with consistent structure and unified identification.

[0028] Based on this standardized historical project dataset, and combined with the cost correlation and correlation strength factors of current engineering projects, a similarity evaluation system is constructed to assess the degree of matching between historical projects and current projects at the cost structure level, and to select a set of reference engineering projects that meet the comprehensive similarity threshold requirements.

[0029] Based on the current project list items, the baseline quantities, unit prices, and cost composition data of the corresponding list items in the reference project set are aggregated, and statistical representative values ​​are calculated according to the unified core identifier of the list items, and encapsulated to form an analog benchmark.

[0030] Furthermore, establish normative benchmarks, including:

[0031] Based on the project implementation cycle, project type, and regional attributes, the applicable engineering cost specifications and cost data are systematically reviewed, and documents and data that are not applicable to the current project are removed.

[0032] Based on pricing elements, time intervals, and applicable conditions, the standard documents and cost release data are analyzed to form a structured set of standard constraint rules, and the rule associations are established using a unified core identifier for list items;

[0033] Based on the set of normative constraint rules, reasonable ranges for the quantity, unit price and cost composition of the bill of quantities items are defined, and they are encapsulated according to the unified core identifier of the bill of quantities items to form a normative benchmark.

[0034] Furthermore, the targets for cost verification include:

[0035] Perform original verification on bill of quantities data, measurement and payment data and cost release data, identify and record missing, duplicate and abnormal data, and form a data anomaly ledger;

[0036] Construct a hierarchical unified identification system, assign a unified core identifier to bill of quantities data, assign a core identifier and payment cycle identifier to bill of quantities items to measurement and payment data, assign a time dimension identifier and data type identifier to cost release data, and establish a unified identifier index table;

[0037] Using the bill of quantities data as the alignment benchmark, the bill of quantities item-level alignment process is performed according to the item code, project characteristic description, unit of measurement and pricing quota sub-item, and integrated into a cost verification object oriented towards the bill of quantities item.

[0038] Further, list item-level alignment processing includes:

[0039] The corresponding list items are located using a unified core identifier matching method, and consistency checks are performed on auxiliary elements such as list item names and units of measurement. For cases where there are differences in expression but are substantially consistent, a synonym mapping relationship is established and included in the alignment mapping ledger.

[0040] A conflict classification and processing mechanism is implemented for conflicting data that occurs during the alignment process. For conflicts in units of measurement, conversion adjustments are performed and the calculation trajectory is retained. For conflicts in applicable scope, realignment and screening processing is performed, and the basis for correction, correction method and processing information are recorded.

[0041] Secondly, the big data collaborative engineering cost verification system includes a data acquisition module, an association construction module, a benchmark generation module, a deviation verification module, and a result update module.

[0042] The data acquisition module obtains bill of quantities data, measurement and payment data, and cost release data, and performs unified identification and alignment of the three types of data with the bill of quantities item level to form cost verification objects oriented towards bill of quantities items;

[0043] The association construction module constructs a cost association table based on the cost verification object, analyzes the quantity transfer characteristics and structural constraint characteristics between the items in the bill of quantities in terms of quantity composition, cost decomposition and pricing structure, establishes cost association relationships and quantifies the association strength factor;

[0044] The benchmark generation module combines historical engineering cost data to perform cost structure similarity analysis, constructs analog benchmarks, parses engineering cost specification documents and cost release data constraints, constructs standard benchmarks, and differentiates the role weights of analog benchmarks and standard benchmarks according to the list item structure and correlation strength factors to form collaborative cost benchmarks;

[0045] The deviation verification module calculates the deviation of each item in the bill of quantities from the collaborative cost benchmark in terms of quantity, unit price, and cost composition. It then combines the correlation strength factor to limit the transmission range and determine the transmission consistency, thus forming an abnormal list set.

[0046] The results update module, based on the anomaly list set, backfills historical engineering cost data according to anomaly type and cost structure characteristics, and corrects and updates the construction rules of the collaborative cost benchmark and the role weights of each list item.

[0047] Compared with the prior art, the significant advantages of this invention are:

[0048] 1. By constructing cost correlation relationships at the item level and quantifying correlation strength factors, the consistency judgment of cost deviations across the dimensions of project quantity, cost and pricing structure is achieved, and cost anomalies with logical inconsistencies are identified from the structural level, breaking through the limitations of traditional single-point verification.

[0049] 2. By introducing a collaborative cost benchmark construction mechanism that combines analogical and standard benchmarks, and by configuring the weights of different items based on their structure and correlation strength, a verification benchmark system that can be dynamically corrected and iteratively updated is formed, significantly improving the adaptability and stability of engineering cost verification. Attached Figure Description

[0050] Figure 1 Flowchart of a big data-driven collaborative method for engineering cost verification;

[0051] Figure 2 This is a flowchart illustrating the construction process of cost correlation and correlation strength factors in this invention;

[0052] Figure 3 This is a flowchart of the collaborative cost benchmark construction process in this invention;

[0053] Figure 4 This is a schematic diagram of the deviation amount ripple-like transmission and backfill update in this invention. Detailed Implementation

[0054] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0055] Example 1

[0056] like Figure 1 As shown, this invention discloses a big data collaborative-driven method for engineering cost verification, comprising the following steps:

[0057] S1: For completed projects, obtain bill of quantities data, measurement and payment data, and cost release data published by the government during the project implementation period. Apply unified identification and alignment to the bill of quantities item level to the obtained data to form cost verification objects oriented towards bill of quantities items.

[0058] S2: Based on the cost verification object, construct a cost correlation table according to the bill of quantities items. On the basis of the cost correlation table, analyze the quantity transmission characteristics and structural constraint characteristics between the bill of quantities items in terms of quantity composition, cost decomposition and pricing structure, and establish cost correlation relationships that reflect the dependence of cost elements between bill of quantities items. Based on the degree of transmission dependence of quantity changes between bill of quantities items, the proportion of cost composition and the coupling characteristics of pricing structure, perform quantitative calculation on the cost correlation relationship to obtain the correlation strength factor that characterizes the degree of correlation between cost elements between bill of quantities items.

[0059] S3: Based on the aforementioned cost correlation and correlation strength factors, perform cost structure similarity analysis on historical engineering cost data, screen reference engineering projects that match the current engineering project in terms of cost structure characteristics and engineering implementation cycle, and construct an analogy benchmark for horizontal benchmarking; simultaneously, based on the aforementioned cost correlation and engineering implementation cycle characteristics, analyze the constraint range of applicable engineering cost specification documents and engineering cost release data during the engineering implementation period in terms of pricing elements, time intervals, and applicable conditions, form a set of specification constraint rules consistent with the current engineering project cost structure, and construct a specification benchmark to limit the reasonable range of engineering costs; on this basis, combine the structural role of different bill of quantities items in the cost correlation and their corresponding correlation strength factors, differentiate the role weights of the analogy benchmark and specification benchmark on each bill of quantities item, and form a collaborative cost benchmark oriented towards bill of quantities items;

[0060] S4: Based on the cost association table, calculate the deviation of each bill of quantities item from the collaborative cost benchmark in terms of quantity, unit price and cost composition. Combine the characteristics of the transmission path of the deviation in the cost association relationship, and limit the range of transmission of the deviation between bill of quantities items according to the association strength factor. Determine the transmission consistency of the deviation in the cost element association structure. When the deviation cannot form a reasonable transmission relationship that matches the association strength factor in the cost association relationship, it is determined that the corresponding bill of quantities item has logical inconsistency and an abnormal bill of quantities set is formed.

[0061] S5: Based on the aforementioned abnormal list set, and combined with the cost correlation, correlation strength factor, and deviation propagation judgment results of the corresponding bill of quantities items, generate verification conclusion data for engineering cost verification. Fill the verification conclusion data back into the historical engineering cost data according to the abnormality type and cost structure characteristics. Correct and update the construction rules of the collaborative cost benchmark and its role weight on each bill of quantities item for use in subsequent engineering cost verification of engineering projects.

[0062] Step S1 involves acquiring bill of quantities data, measurement and payment data, and government cost data published during the implementation period. After unified identification and alignment with the bill of quantities item level, cost verification objects are formed for each bill of quantities item, including:

[0063] S101: Data Acquisition and Original Verification.

[0064] For completed engineering projects, complete bill of quantities data is collected through channels such as the project management system database, cost estimation software backup files, and construction unit archives. The bill of quantities data includes at least the bill of quantities item code, bill of quantities item name, project characteristic description, unit of measurement, quantity, comprehensive unit price composition, pricing details, and corresponding quota sub-items.

[0065] The measurement and payment data is collected through project payment management ledgers, financial settlement vouchers, and supervision review documents. The measurement and payment data includes at least the completed work quantity, paid amount, payable amount, deduction details, and payment approval basis for each payment cycle.

[0066] The cost data released during the project implementation period is collected through the official website of the government's engineering cost management department, official documents issued by the industry's competent authority, and authoritative cost information platforms. The cost data released includes at least the unit price of labor, material prices, unit price of machinery shifts, fee standards, fee rate adjustment documents, cost index, and quota revision notices.

[0067] Verification operations were performed on the three types of collected data. Integrity verification was carried out on the core fields of each data type to ensure that the bill of quantities data had no missing list item codes, project feature descriptions and other key information; the measurement and payment data had no missing payment cycles and corresponding list item relationships; and the cost release data had no missing key price adjustment documents and rate standards within the implementation period. Validity verification was carried out through cross-comparison and format verification to remove duplicate data, non-standard format data and invalid data, such as inconsistent units of measurement, incorrect numerical formats, cost release data that exceeded the project implementation period, and measurement and payment records without corresponding list items. For missing data found during verification, it was supplemented by collecting archived materials and verifying with relevant responsible parties. If it could not be supplemented, the reason for the missing data was marked and included in the data anomaly ledger.

[0068] For example:

[0069] With a building area of ​​120,000 square meters 2 Taking a residential construction project as an example, in the bill of quantities data collected through the project management system database, the bill of quantities item code "010502001" corresponds to the bill of quantities item name "rectangular column", the project characteristic description is "C30 commercial concrete, cross-sectional dimensions 600×600mm", and the unit of measurement is "m". 3 The project volume is 1200m. 3 The comprehensive unit price includes labor costs of 35 yuan / m. 3 Materials: 420 yuan / m 3 Mechanical equipment 18 yuan / m 3 According to the measurement and payment data collected through the project payment management ledger, the completed work volume of this item in the second payment cycle of 2024 was "300m". 3 The amount paid is "142,500 yuan", the amount due is "142,500 yuan", and there are no deduction details; according to the cost data collected from the local housing and construction bureau's official website, the price of C30 commercial concrete in the second quarter of 2024 was 415-430 yuan / m³. 3The labor cost is 85 yuan per man-day. During verification, it was found that the completed work volume record for this item was missing in the third payment cycle of 2024 in the measurement and payment data. This was supplemented by reviewing the supervision and audit documents, and the completed work volume "280m" was subsequently collected. 3 The data anomaly log will be marked as "Initially missing, supplemented by supervisory documents to complete".

[0070] S102: Assigning a unified data identifier.

[0071] A hierarchical unified identification system is constructed, adopting a 16-digit coding rule of "unique project code - sub-project code - item code - list item serial number". The first 6 digits are the unique project code, the 7th-9th digits are the sub-project code, the 10th-12th digits are the item code, and the 13th-16th digits are the list item serial number. For cost release data, an additional "time dimension identifier - data type identifier" is associated. The time dimension identifier uses an 8-digit code to represent the time interval for data release and application, and the data type identifier uses a 2-digit code to distinguish data types such as labor, materials, machinery, rates, and indices.

[0072] According to the aforementioned unified identification rules, each item in the bill of quantities data that has passed verification is assigned a unique core identifier to ensure that the same item is uniquely identified throughout the entire data process. For measurement and payment data, a link is established between the item name, project feature description, and bill of quantities data, and each measurement and payment record is assigned a core identifier and payment cycle identifier corresponding to the item. For cost release data, based on the matching relationship between its applicable time interval and the project implementation cycle, each cost release data is assigned a corresponding time dimension identifier and data type identifier, and linked to the corresponding sub-project code to achieve indirect identification association between cost release data and bill of quantities data. After the assignment is completed, a unified identifier index table is established to record the coding rules, corresponding data ranges, and association relationships of various identifiers to ensure that the identifiers are traceable.

[0073] S103: List item-level alignment processing.

[0074] Using the bill of quantities items in the bill of quantities data as the benchmark for alignment, the core elements for alignment of each bill of quantities item are clarified, including four core dimensions: bill of quantities item code, project characteristic description, unit of measurement, and pricing quota sub-item. This ensures the uniqueness and consistency of the bill of quantities items during the alignment process. Among them, the pricing quota sub-item is the pricing details and the corresponding quota sub-item.

[0075] For measurement and payment data, alignment is achieved through a combination of core identifier matching and auxiliary element verification. First, the corresponding list item is matched using a unified identifier index table. Then, the consistency between the list item name and unit of measurement in the measurement and payment data and the benchmark list item is verified. If there are differences in name but consistency in project description and unit of measurement, a synonym mapping relationship is established after manual verification and confirmation, and the data is included in the alignment mapping ledger. For cost release data, alignment is achieved with the pricing details of the corresponding list item according to the sub-project code, data type identifier, and applicable time interval. For example, labor unit price data is aligned to the labor pricing section of the corresponding list item, and material price data is aligned to the material pricing details of the corresponding list item, ensuring accurate matching between the cost release data and the pricing elements of the list item.

[0076] For conflicting data discovered during the alignment process, such as inconsistencies between the units of measurement for items in the bill of quantities and the benchmark bill of quantities, or mismatches between the scope of application of the cost release data and the pricing requirements of the bill of quantities items, a conflict classification and handling mechanism is established: For conflicts in units of measurement, the quantities and amounts in the bill of quantities data are converted and adjusted based on the units of measurement in the bill of quantities data, and the calculation process is recorded; for conflicts in scope of application, the cost release data for the corresponding time interval and the corresponding sub-project within the project implementation period are screened and re-aligned, and data that exceeds the scope of application is removed; all conflict handling is recorded with the basis for correction, the method of correction, and the person in charge, to ensure that the alignment results are traceable.

[0077] S104: Data integration and verification object generation.

[0078] Using a unified core identifier for each item in the bill of quantities as a link, the corresponding measurement and payment data and cost release data are integrated to form a dataset based on the bill of quantities item. The measurement and payment data includes the quantity, amount and approval information of each payment period, while the cost release data includes the price, rate and adjustment documents of the corresponding pricing elements. Each dataset contains three major modules: benchmark bill of quantities information, associated measurement and payment data and associated cost release data. The modules are linked through a unified identifier.

[0079] The integrated list item-level dataset is standardized to unify the data format, numerical precision, and field naming rules. Field names are uniformly formatted as "list item code-core identifier-data type-field name". After encapsulation, it forms a cost verification object for the bill of quantities items.

[0080] like Figure 2 As shown, in step S2, a cost correlation table is constructed, the composition and structural constraint features between the list items are analyzed to establish cost correlation relationships, and the correlation strength factor is calculated based on multiple quantitative characteristics.

[0081] It is understandable that during the process of forming project costs, the quantities, costs and pricing rules of the items in the bill of quantities often interact with each other through decomposition, summarization and linkage adjustment. If only a single item in the bill of quantities is checked independently, it is easy to find that each item in the bill of quantities is within a reasonable range when judged individually, but the overall cost structure is unbalanced and difficult to identify.

[0082] Establishing cost relationships between bill of quantities items based on cost association tables can clarify the transmission path and constraint boundaries of quantities, costs, and pricing rules among bill of quantities items, preventing cost deviations from being dispersed or masked among associated bill of quantities items, and providing a traceable analytical basis for subsequent consistency determination of deviations.

[0083] The steps include:

[0084] S201: Construction of the cost association table.

[0085] Using the bill of quantities items in the cost verification object as the core carrier, a two-dimensional cost association table is constructed. The core fields of the table include basic information fields for the bill of quantities items, association dimension fields, and data traceability fields. Among them, the basic information fields for the bill of quantities items cover the unified core identifier, code, name, unit of measurement, project characteristic description, and baseline quantity of the bill of quantities item; the association dimension fields reserve three types of fields for quantity composition association, cost breakdown association, and pricing structure association, which are used to record the association information between bill of quantities items in subsequent analysis; the data traceability fields record the corresponding data source and data generation time to ensure the traceability of data in the association table.

[0086] Completely populate the basic information of each item in the bill of quantities corresponding to each cost verification object into the cost association table, ensuring that each item corresponds to a unique data row and that the unified core identifier remains consistent with the previous text; for the completed work quantity and cost composition data of the bill of quantities items in the measurement and payment data, and the pricing basis data in the cost release data, associate and populate the corresponding association dimension field's preparatory area according to the unified core identifier of the bill of quantities item; after the population is completed, perform normalization processing on the association table data, unify the field data format and numerical precision, and remove duplicate data and data with abnormal format to ensure the accuracy and consistency of the association table data, providing data support for subsequent feature analysis.

[0087] S202: Feature analysis and cost correlation establishment among list items.

[0088] Based on the cost association table and the pricing details and quota sub-item data in the cost verification object, we analyze the quantity transfer characteristics of each item in the bill of quantities in terms of the composition of the engineering quantity, and clarify the direction of engineering quantity dependence and transfer rules between the items in the bill of quantities.

[0089] For example, there is a positive transmission characteristic between sectional works and individual works in the bill of quantities: "the quantity of sectional works equals the sum of the quantities of all individual works." Conversely, there is a negative transmission characteristic between related work process items: "deviation in the quantity of the preceding work process → adjustment in the quantity of the subsequent work process." The transmission ratio threshold and the scope of influence are recorded simultaneously. At the same time, structural constraint characteristics are extracted to clarify the rigid rules governing the composition of quantities, such as the quantity of the main structure item in the bill of quantities not being less than a reasonable proportion of the quantity of the corresponding auxiliary structure item, and the sum of the quantities of all sub-items under the same sectional work process needing to be consistent with the quantity of the individual works in the bill of quantities.

[0090] By combining the comprehensive unit price composition, fee standards and measurement and payment data of the cost verification objects, we analyze the quantity transmission characteristics of cost decomposition between list items and sort out the allocation and transmission paths of direct costs and indirect costs. Direct costs include labor, materials and machinery costs, while indirect costs include management fees, profits and regulatory fees.

[0091] For example, the costs of general contracting items are broken down into various specialized subcontracting items according to a predetermined ratio, resulting in a transmission relationship where "the general contracting cost equals the sum of the costs of all subcontracting items plus the general contracting management fee." For items with a high proportion of material costs, fluctuations in material prices will be transmitted to the comprehensive unit price of related items according to the cost composition ratio. Simultaneously, structural constraints are clearly defined, including hierarchical rules for cost breakdown (e.g., total cost → segment cost → item cost → sub-item cost) and constraints on the cost base (e.g., management fees are calculated based on labor + machinery costs, and the base range cannot exceed the cost boundary of the corresponding item).

[0092] Based on the pricing basis, quota items, and rate standards in the cost release data of the cost verification objects, we analyze the coupling characteristics of the pricing structure between the list items, that is, the extension of the quantity transfer characteristics, and clarify the linkage relationship of the pricing rules.

[0093] For example, if the same professional category of list items adopts a unified quota sub-item pricing model, there is a coupling characteristic of "quota sub-item adjustment → synchronous adjustment of the pricing structure of related list items"; for list items that require the calculation of special fees (such as large machinery entry and exit fees), their pricing structure is rigidly coupled with the pricing structure of the corresponding main project list items, and the calculation of special fees must be based on the quantity of the main project. At the same time, structural constraint characteristics are extracted, including the consistency constraint of pricing basis (such as the same type of list items in the same project must use the same version of quota and fee standard) and the hierarchical matching constraint of pricing structure (such as the sub-item pricing structure must be compatible with the pricing structure of the itemized list items), etc.

[0094] Based on the feature analysis results of the above three dimensions, the dependencies between the items in the bill of quantities are classified and sorted to form a structured cost relationship. According to the relationship type, they are divided into three categories: composition dependency, cost dependency, and pricing dependency. Composition dependency refers to the dependency at the level of quantity composition, cost dependency refers to the dependency at the level of cost decomposition, and pricing dependency refers to the dependency at the level of pricing structure.

[0095] For each type of relationship, label the relationship direction (one-way / two-way), dependent carrier (work quantity / cost / pricing rules) and constraints, and record them in the format of "List Item A (Unified Core Identifier) ​​- Relationship Type - Relationship Direction - Constraint Condition - List Item B (Unified Core Identifier)" to form a complete set of cost relationship relationships, and update the relationship dimension fields of the cost relationship table synchronously.

[0096] For example:

[0097] In the analysis of the composition characteristics of the project quantity, the "Concrete and Reinforced Concrete Works" sub-item (core identifier 320102-0105-0000-0001) exhibits a positive transmission characteristic with the "Rectangular Column" item (320102-0105-0200-0001) and the "Rectangular Beam" item (320102-0105-0200-0002), that is, "Sub-item project quantity = Rectangular column project quantity + Rectangular beam project quantity + Other sub-item project quantity". The transmission ratio threshold is "The total quantity of each sub-item project accounts for 95%-105% of the sub-item project quantity". The structural constraint characteristic is "The rectangular column project quantity shall not be less than 3 times the corresponding column reinforcement sub-item project quantity (320102-0105-0300-0001) (set according to the conventional ratio of reinforcement content)".

[0098] In the cost breakdown feature analysis, there is a cost transfer relationship between the "General Contracting Project" item (320102-0000-0000-0001) and the "Rectangular Column" item. The general contracting cost is decomposed to this item according to "Rectangular Column Cost × 1.03 (including 3% general contracting management fee)". The structural constraint feature is "Management fee base = labor + machinery cost, and the base shall not exceed 80% of the direct cost of rectangular column". In the pricing structure feature analysis, the "Rectangular Column" and "Rectangular Beam" items are priced using the same version of the "Standard for Calculation of Quantities of Building Construction and Decoration Engineering". There is a coupling feature of "Adjustment of quota item A4-005 (Rectangular Column Concrete Pouring) → Synchronous Adjustment of the Corresponding Quota Item of Rectangular Beam". The structural constraint feature is "Both of the fee standards are implemented according to the local 2023 version of the building engineering fee quota".

[0099] The final cost relationship is "320102-0105-0000-0001-Composition Dependency-One-Way-Sub-Project Quantity = Total of All Items - 320102-0105-0200-0001".

[0100] S203: Quantitative calculation of correlation strength factor.

[0101] To address the correlation of costs, a three-dimensional quantitative system is constructed, corresponding to three dimensions: the degree of dependence on changes in project quantity, the proportional relationship of cost composition, and the coupling characteristics of the pricing structure. Each dimension adopts... For interval standardization, the closer the value is to 1, the stronger the correlation.

[0102] The degree of dependency function for the propagation of changes in the quantity of work is denoted as: This characterizes the degree of impact of changes in the quantity of item A on the quantity of item B; the cost composition ratio function is denoted as... This characterizes the proportion and influence weight of the cost of item A in the associated cost of item B; the pricing structure coupling characteristic function is denoted as... This indicates the degree of linkage between pricing rules and quota sub-items and other pricing elements among the list items.

[0103] In calculating the degree of dependency propagation of changes in engineering quantities To accurately quantify the sensitivity of changes in quantities among bill of quantities items, an elasticity coefficient method is used to construct a function, and the calculation formula is shown below:

[0104] ,

[0105] in, This refers to the change in quantity of work for item A in the bill of quantities. The baseline quantity of work for item A in the bill of quantities. For the recipient The change in the quantity of work affecting item B in the bill of quantities. This is the baseline quantity of work for item B in the bill of quantities; if it is a positive structural relationship (such as a section and an item), then... Take the calculated value directly. If it's a reverse influence relationship (such as inter-process constraints), then... Take the absolute value of the calculated value, and finally standardize the calculation result to... interval, For strong dependence, It is a moderate dependency. It is a weak dependency.

[0106] In calculating the proportional relationship function of cost composition When considering the difference in the weighting of direct and indirect costs in the cost correlation, a weighted summation method is used. Direct and indirect costs are calculated separately and then summed using a weighted method. The calculation formula is as follows:

[0107] ,

[0108] in, , These are the direct costs and indirect costs of item A in the list. , These are the direct and indirect costs of related item B, respectively. , The weights for direct costs and indirect costs are respectively determined and can be dynamically adjusted based on industry cost structure conventions for project types such as building construction, installation engineering, and municipal engineering, according to the actual proportion of direct costs in the total project cost. A higher proportion of direct costs results in higher weightings. The larger the value, Perform reverse adaptation adjustments. ,default , If the item in the bill of quantities involves only a single cost type, such as only direct costs, then only the corresponding item is calculated, and the final result is standardized to... Interval.

[0109] In calculating the coupling characteristic function of the pricing structure When using the matching degree analysis method, the matching degree is calculated from three sub-dimensions: pricing basis, quota item, and fee standard, and then the average value is taken. The calculation formula is as follows:

[0110] ,

[0111] in, The matching degree of the pricing basis between the two list items is 1 if the basis is the same and 0 if the basis is different. The matching degree of quota sub-items between the two lists is 1 for complete consistency of core sub-items, 0.5 for partial consistency, and 0 for complete inconsistency. To determine the matching degree of fee standards between two list items, a value of 1 is assigned if the fee base and rate rules are completely identical, 0.5 is assigned if they are partially identical, and 0 is assigned if they are completely inconsistent. The result naturally falls into the range of 1. Interval.

[0112] The weighted summation method is used to integrate the three-dimensional quantization functions to obtain the final correlation strength factor. The calculation formula is as follows:

[0113] ,

[0114] in, , , These are the weights of the three quantitative functions, dynamically adjusted according to the key points of project cost control. By default, the quantity of work dimension has the highest priority. , , ,and After the calculation is completed, for The value is standardized and validated to ensure it falls within the range of 100%. interval, The cost factors among the items in the characterization list are highly correlated. The correlation strength is moderate. The correlation strength is low. The correlation strength factor corresponding to each cost correlation relationship is used to characterize the low correlation strength. The cost correlation table is entered synchronously to provide a quantitative basis for the subsequent collaborative cost benchmark construction and deviation determination.

[0115] For example:

[0116] Taking "Concrete and Reinforced Concrete Works" item (A) and "Rectangular Columns" item (B) as examples, calculate the associated strength factor. :

[0117] Dependence on the transmission of changes in project quantity Item A in the bill of quantities has a base quantity of 5000m³. 3 Item B in the bill of quantities has a base quantity of 1200m³. 3 If the quantity of project A increases by 500m 3 ,Right now This resulted in an increase of 120m in the amount of work in Project B. 3 ,Right now ,but After standardization, it is 1.0, which indicates a strong dependency;

[0118] Cost composition ratio Item A has direct costs of 2.8 million yuan and indirect costs of 420,000 yuan; Item B has direct costs of 588,000 yuan and indirect costs of 84,000 yuan. (The last part is a partial translation and doesn't need a direct translation.) , ,but =0.6×(280 / 58.8)+0.4×(42 / 8.4)=0.6×4.7619+0.4×5=2.8571+2=4.8571; The calculation result here exceeds the [0,1] interval. After standardization as required, Taking the upper limit of 1.0 indicates that the cost ratio and influence weight of list item A to B have reached a strong correlation level;

[0119] Coupling characteristics of pricing structure Both are based on GB50500-2013 ( The core quota items all belong to the concrete pouring category ( The fee standards are all based on the 2023 fixed rates. ),but =(1+0.5+1) / 3≈0.833;

[0120] Correlation strength factor :Pick , , ,but =0.4×1.0+0.3×1.0+0.3×0.833≈0.4+0.3+0.2499≈0.9499, which, after standardization, is still in the [0,1] interval. The cost elements of items A and B in the characterization list have a high degree of correlation.

[0121] like Figure 3 As shown, step S3 involves analyzing historical data and current standards to construct analogy benchmarks and standard benchmarks, and then assigning weights based on the structural roles and correlation strength of the list items to ultimately form a collaborative cost benchmark, including:

[0122] S301: Analog Benchmark Construction.

[0123] Extract project data of the same type and scale as the current project from the historical engineering cost database. The screening scope covers the entire process of project cost data from project initiation to final settlement, including bill of quantities data, measurement and payment data, cost release data and corresponding cost correlation tables and correlation strength factors for each project.

[0124] Preprocessing operations are performed on the extracted historical data, and unified identification completion, list item alignment and data standardization are completed according to the S101-S104 standards. Abnormal data with missing core fields, incomplete correlations and exceeding the reasonable cost fluctuation range are removed to form a standardized historical project dataset.

[0125] Based on the cost correlation and correlation strength factors of current engineering projects, a cost structure similarity analysis system is constructed. Constraint analysis is conducted from three dimensions: similarity of bill of quantities items, similarity of cost breakdown ratios, and similarity of pricing structure coupling. The similarity of each dimension is assessed using... The interval is standardized, and the closer the value is to 1, the higher the similarity.

[0126] The specific construction content is as follows:

[0127] Similarity in constructing the list items At that time, based on the unified core identifier of the list items in the current project cost association table, the percentage of list items that overlap between historical projects and the current project is calculated, and the association strength factor corresponding to the overlapping list items is combined. Weighted calculation, the formula is as follows:

[0128] ,

[0129] in, The number of overlapping items in the list. For the first The correlation strength factor of each overlapping list item. This represents the total number of items in the current project list.

[0130] Similarity of cost breakdown ratios in construction When comparing the proportions of direct and indirect costs between historical projects and current projects, and combining this with a weighted adjustment based on the correlation strength factor corresponding to the cost dependency relationship, the formula is as follows:

[0131] ,

[0132] in, For the number of fee types, , These are the current project and the historical project, respectively. The proportion of such expenses For the first The association strength factor of the dependency relationship corresponding to the class cost.

[0133] Constructing the similarity of pricing structure coupling At that time, based on the pricing dependency relationship, the matching degree between historical projects and current projects in terms of quota item selection, fee standards, and pricing basis is compared, and the pricing structure coupling characteristic function is referenced. The calculation logic uses the weighted average of the three matching degrees as the evaluation result, and the formula is:

[0134] ,

[0135] in, To determine the matching degree of quota sub-items between historical and current project list items, To determine the matching degree between historical and current project list items' fee standards, The definition of the matching degree between the pricing basis of historical and current project list items is consistent with that in S203. , , These are the correlation strength factors for the corresponding pricing dependencies.

[0136] A comprehensive similarity threshold is set, with a default value of 0.7. This threshold can be dynamically adjusted according to project type, such as different types of projects like building construction, municipal engineering, and installation engineering. The threshold is adjusted to suit different project types based on factors like process complexity and cost structure differences. For specialized projects with complex processes and diverse cost structures, the threshold is raised to 0.75-0.8, while for ordinary projects with simple processes and a single cost structure, the threshold can be lowered to 0.65-0.7. The comprehensive similarity score for historical projects is then calculated. Its weight and correlation strength factor The calculation weights are consistent, highlighting the priority of the engineering quantity dimension, and historical projects with a comprehensive score not lower than the similarity comprehensive threshold are selected as the reference engineering project set.

[0137] Based on the current project bill of quantities items, the benchmark quantities, unit prices, cost composition and related strength factors of the corresponding bill of quantities items in the reference project set are integrated, and the average values ​​are calculated by grouping according to the unified core identifier of the bill of quantities items to form a horizontal benchmarking data set with bill of quantities items as the unit. After encapsulation, the analogy benchmark is obtained, and the reasonable cost range of each bill of quantities item is clarified.

[0138] For example:

[0139] Total number of items in the current residential project list Three similar residential projects (Project 1, Project 2, and Project 3) were extracted from the historical database and a similarity evaluation system was constructed after preprocessing.

[0140] List item composition similarity Number of items in the list that overlap between the three historical projects and the current project The corresponding correlation strength factors are 65, 68, and 72, respectively, for overlapping list items. The sums are 45, 48, and 52 respectively, then the sum of item 1 is... =45 / 80=0.5625, Item 2 =48 / 80=0.6, Item 3 =52 / 80=0.65;

[0141] Cost breakdown ratio similarity Fee type Currently, direct costs account for 68% of project costs and indirect costs for 32%. For project 3, direct costs account for 66% and indirect costs for 34%. This corresponds to the correlation strength factor under the cost dependency dimension. The total is 55, of which the cost dependency strength factor corresponding to direct costs is 28, and the cost dependency strength factor corresponding to indirect costs is 27. =1 - (|68% - 66%| × 28 + |32% - 34%| × 27) / 55 = 1 - (0.02 × 28 + 0.02 × 27) / 55 = 1 - 1.1 / 55 = 0.98;

[0142] Pricing structure coupling similarity : The degree of matching between Project 3 and the quota sub-items in the current project list Fee standard matching degree Matching degree of pricing basis Corresponding correlation strength factor , , ,but = (0.8×30+1×15+1×10) / (30+15+10)=(24+15+10) / 55≈0.8909;

[0143] Overall similarity score =0.4×0.65+0.3×0.98+0.3×0.8909≈0.26+0.294+0.2673≈0.8213≥0.7, therefore, item 3 is selected as the reference item, and its "rectangular column" bill of quantities has a base quantity of 1180m. 3 The comprehensive unit price is 473 yuan / m. 3 After calculating the mean, it is used as a benchmark for comparison, and the reasonable range is 1150-1210m. 3 (Quantity of work) 460-485 yuan / m 3 (Comprehensive unit price).

[0144] S302: Standard benchmark construction.

[0145] Based on the current project implementation cycle, project type, and regional attributes, we systematically compiled applicable engineering cost standards and published cost data. The standards include current national and industry pricing standards, quota standards, fee regulations, and quality acceptance standards. The published cost data includes labor unit prices, material prices, machinery shift unit prices, rate adjustment documents, and cost indices collected in section S101 above. Standards and data that are outdated, inapplicable to certain regions, or have been repealed have been removed.

[0146] Based on three dimensions—pricing elements, time intervals, and applicable conditions—the standardized documents and cost release data are analyzed and sorted to form a set of structured standardized constraint rules.

[0147] The details are as follows:

[0148] Analysis of pricing element constraints: For core pricing elements such as labor, materials, machinery, management fees, profit, and regulatory fees, the calculation methods, fee bases, and fee rate ranges are clearly defined and standardized. Combined with the cost dependence rules in the cost relationship, the scope of influence of each pricing element on the corresponding list item is marked.

[0149] For example, according to the "Construction Project Quantity List Pricing Specification", the management fee is based on the labor + machinery cost, with a rate range of 3%-8%, and is linked to the corresponding sub-project list item.

[0150] Time interval constraint analysis: Based on the time nodes within the project implementation cycle (such as quarters or months), the applicable time period for cost release data is broken down, and the triggering conditions and adjustment range for price adjustments are clarified.

[0151] For example, material prices are updated quarterly. When the price of a certain type of main material fluctuates by more than ±5%, the comprehensive unit price is adjusted according to the material cost ratio of the corresponding list item, and is simultaneously linked to the pricing rules of the quota sub-items in the pricing dependency relationship.

[0152] Analysis of applicable conditions and constraints: Clarify the applicable scenarios for each specification clause and cost data, including constraints such as project scale, structural type, and construction technology, and eliminate rules that do not match the current project conditions.

[0153] For example, the regulations for calculating the entry and exit fees for large machinery only apply to single buildings with a floor area of ​​≥20,000 m². 2 The project corresponds to the pricing dependency of the main structure list items.

[0154] The above analysis results are organized in the format of "unified core identifier of list item - constraint type - constraint content - scope of application - source of basis" to form a set of standardized constraint rules consistent with the current project cost structure, ensuring that each rule can be accurately associated with the corresponding list item.

[0155] Based on a set of normative constraint rules, a reasonable cost range is defined for each item in the bill of quantities. For quantities, the minimum and maximum reasonable quantity ranges are determined according to the calculation rules and measurement standards defined in the norms. For unit prices, the upper and lower limits are calculated by combining the price ranges in the published cost data and the fee standards. For cost composition, the proportion boundaries of each cost type are defined according to the constraint rules. The above range data are integrated, associated, and encapsulated according to a unified core identifier for each item in the bill of quantities, forming a normative benchmark for vertical constraints, thus clarifying the rigid constraint range of the cost for each item in the bill of quantities.

[0156] For example:

[0157] The current project applies the "Standard for Measurement and Pricing of Construction Engineering Quantities" GB50500-2013 and local cost data released in 2024, and the resulting standard constraint rules are analyzed as follows:

[0158] Pricing element constraints: The management fee base for the rectangular column list item = labor cost + machinery cost, with a rate range of 3%-8%, associated with the core identifier 320102-0105-0200-0001, and its scope of influence is the calculation of management fees in the direct costs of this list item;

[0159] Time interval constraint: When the price of C30 ready-mixed concrete fluctuates by more than ±5% in the second quarter of 2024, the comprehensive unit price will be adjusted. The material cost of this item accounts for 89% of the total cost. The adjustment formula is "New comprehensive unit price = Original unit price + (Current material price - Benchmark price) × Material usage". Related quota item A4-005;

[0160] Applicable conditions and constraints: The calculation of entry and exit fees for large machinery does not apply to this project (single building area of ​​120,000 m²).2 <20,000 m 2 Eliminate the corresponding constraint rules. Based on the rule-defined "rectangular column" item specification benchmark: quantity 1100-1250m. 3 (Calculated according to measurement standards and rules), the comprehensive unit price is 455-490 yuan / m. 3 (Including material price range + 3%-8% management fee), direct costs account for ≥85%.

[0161] S303: Formation of collaborative cost benchmarks.

[0162] Based on cost correlation and correlation strength factor The current project's bill of quantities items are divided into three structural roles: core bill of quantities items, related bill of quantities items, and subsidiary bill of quantities items. Among them, core bill of quantities items are related strength factors. Items in the bill of quantities that have a weight of 10% or more on the overall cost (such as the concrete pouring item in the main structure) constitute the dominant cost component; related items are... Items in the bill of quantities that have an impact weight of 3%-10% (such as template items corresponding to the main structure) carry over the cost transfer from the core items in the bill of quantities; supplementary items are... Items with an impact weight of less than 3% (such as miscellaneous parts installation items) are less affected by fluctuations in the overall cost.

[0163] Combining the list item structure and correlation strength factor Differentiated action weights are assigned to the analog benchmark and the normative benchmark to ensure that the sum of the two action weights is 1. The action weights are dynamically adjusted according to the structural roles, as follows:

[0164] Core list items: Prioritize compliance with normative constraints and configure initial weights for normative benchmarks. Analogous to the initial weights of the benchmark Because the cost of core items on the bill of quantities has a significant impact, rigid specifications should be the primary basis, and analogous benchmarks should only be used as supplementary references.

[0165] Related list items: Emphasis on both specifications and analogies, configuration , It takes into account both the rigid constraints of regulations and the actual cost experience of historical projects, balancing rationality and practicality.

[0166] Additional list items: Focus on analogy and reference, configuration , Because the pricing of supplementary items is more flexible, the cost data of similar items in historical projects are more valuable for reference.

[0167] The above configuration values ​​can be adjusted according to the actual application scenario, and should also be combined with the correlation strength factor. The weights are fine-tuned, and the calculation formula is shown below:

[0168]

[0169] ,

[0170] in, To incorporate the standardized baseline weights after fine-tuning the correlation strength factor, To incorporate the analogy benchmark weights after fine-tuning the correlation strength factor, the numerical settings in the formula are all aligned with the cost control logic and correlation strength characteristics. Here, 0.5 represents the correlation strength factor. The median value of the interval, corresponding to the previous text The three-level division standard (0.3, 0.7 boundary) can achieve precise control by increasing the normative weight for strong correlations and decreasing the normative weight for weak correlations. The weights are kept at their initial configuration, aligning with the control requirements for moderately related items in the cost estimate list. A fine-tuning coefficient of 0.2 limits the weight adjustments to within ±0.1, highlighting the weight differences between items with varying degrees of correlation while preventing excessive deviation from the initial weight configuration, thus ensuring benchmark stability. This ensures that for items with higher correlation strength, the regulatory benchmark's constraint weight is more prominent, adapting to the transmission characteristics of cost-related relationships.

[0171] For each item in the bill of quantities, a weighted fusion algorithm is used to integrate the cost ranges of the analogy benchmark and the standard benchmark to obtain the collaborative benchmark fusion formula, as shown in the following formula:

[0172] ,

[0173] in, For the first The benchmark value for collaborative cost of each item on the list. , These are the analogy benchmark and the normative benchmark, respectively. The baseline value for each item in the bill of quantities is established. After integration, the consistency of the collaborative cost baseline is verified by combining the structural constraint characteristics in the cost relationship (such as the rules for the composition of engineering quantities and the cost decomposition ratio), and the interval conflicts caused by the weight configuration are corrected (such as when the collaborative interval of the core bill of quantities item exceeds the upper limit of the specification, the upper limit of the specification baseline shall prevail). After the verification is passed, the cost baseline is organized and packaged according to the unified core identifier of the bill of quantities item to form a collaborative cost baseline for the bill of quantities item, providing a unified reference standard for subsequent deviation calculation.

[0174] For example:

[0175] "Rectangular Column" list item associated strength factor It has a 12% weighting on the overall cost and is classified as a core item in the cost list. , ; combination Fine-tuning the weights: =0.7×(1+0.2×(0.7117-0.5))=0.7×(1+0.2×0.2117)=0.7×1.0423≈0.7296, =1-0.7296≈0.2704; the quantity of this item in the benchmark is 1180m. 3 The comprehensive unit price is 473 yuan / m. 3 The standard engineering quantity is 1175m. 3 The comprehensive unit price is 472.5 yuan / m². 3 Then the benchmark value for collaborative construction costs is:

[0176] Quantity of work = 0.2704 × 1180 + 0.7296 × 1175 ≈ 319.07 + 857.28 ≈ 1176.35 m 3 ;

[0177] The comprehensive unit price = 0.2704 × 473 + 0.7296 × 472.5 ≈ 127.90 + 344.79 ≈ 472.69 yuan / m 3 After consistency verification, the coordination range did not exceed the upper limit of the specification, and the final coordination benchmark was determined as: 1170-1182m of work area. 3 The comprehensive unit price is 470-475 yuan / m. 3 .

[0178] Step S4 involves calculating the deviation of each item in the cost list from the collaborative cost benchmark, analyzing its transmission consistency in cost relationships, and summarizing logically inconsistent items into an anomaly list set, including:

[0179] S401: Multidimensional deviation calculation.

[0180] Using the cost correlation table as a carrier, the actual cost data of each item in the bill of quantities is extracted, including the quantity of work, the comprehensive unit price, and the cost composition. The deviation is calculated in three dimensions by comparing it with the S303 collaborative cost benchmark value, and all deviations are standardized to [specific values]. The interval, the absolute value represents the degree of deviation, and the positive and negative values ​​distinguish the direction of deviation, with positive values ​​being higher than the benchmark and negative values ​​being lower than the benchmark.

[0181] The deviation of the project quantity is calculated using the relative deviation method. The calculation formula is as follows:

[0182] ,

[0183] in, For the first The deviation of the quantity of each item in the bill of quantities is standardized to... interval; For the first The actual quantities of each item in the bill of quantities are taken from the cost-related table. For the first The collaborative cost benchmark quantity for each item in the bill of quantities. If If there is no baseline quantity for a miscellaneous item in the bill of quantities, it is directly judged as abnormal and the deviation is recorded as 1. Here, 1 is a positive extreme deviation indicator, which is set manually to distinguish the scenario without a baseline.

[0184] Overall unit price deviation Excluding the reasonable price fluctuation range defined in the S302 standard, the reasonable price fluctuation range refers to the price fluctuation range of the pricing elements constituting the comprehensive unit price, such as labor, materials, and machinery shifts, within the project implementation period, which does not exceed the general threshold of the cost data published by the cost management department of the project location. For example, the price fluctuation range of main materials... For cases involving auxiliary materials ±8%, labor / machinery shifts ±3%, or fluctuation thresholds stipulated in the project construction contract or cost management documents, the calculation formula is as follows:

[0185] ,

[0186] in, For the first Deviation of the overall unit price for each item in the list; For the first The actual comprehensive unit price of each item in the bill of quantities is taken from the cost correlation table; For the first The benchmark unit price for collaborative cost estimation of each item in the list; This is a price fluctuation correction factor, within a reasonable price fluctuation range. If it exceeds This only reflects irrational fluctuations.

[0187] Deviation in cost composition Weighted aggregation based on direct / indirect costs, the formula is as follows:

[0188] ,

[0189] in, For the first The cost of each item on the list constitutes a deviation; The weights for direct costs and indirect costs are set as follows: This highlights the dominance of direct costs; For the first The actual direct / indirect costs of each item on the list; The first The collaborative benchmark direct costs and collaborative benchmark indirect costs for each item on the list.

[0190] The three-dimensional deviation is entered into the cost correlation table, and the deviation direction, calculation basis and benchmark value are marked to provide data support for the transmission analysis.

[0191] For example:

[0192] The actual data for the "rectangular column" item in the bill of quantities is: actual project quantity 1300m. 3 The actual comprehensive unit price is 490 yuan / m². 3 The actual direct costs were 637,000 yuan, and the actual indirect costs were 83,000 yuan. The collaborative benchmark data is: the collaborative cost benchmark quantity is 1176.35m. 3 The benchmark unit price for collaborative construction is 472.69 yuan / m². 3 The direct cost of the collaborative benchmark for this item is 595,000 yuan, and the indirect cost of the collaborative benchmark is 78,000 yuan.

[0193] Deviation in quantity of work: (1300-1176.35) / 1176.35≈123.65 / 1176.35≈0.1051, which is 0.1051 after standardization, indicating a positive deviation;

[0194] Overall unit price deviation: The price of C30 ready-mixed concrete fluctuates by 3% (within a reasonable range of ±5%). ),but =(490-472.69) / 472.69×0=0;

[0195] Deviation in cost composition : , ,but =0.6×(63.7-59.5) / 59.5+0.4×(8.3-7.8) / 7.8≈0.6×4.2 / 59.5+0.4×0.5 / 7.8≈0.0423+0.0256≈0.0679.

[0196] S402: Deviation Quantity Transmission Path Characteristic Analysis and Range Limitation.

[0197] Based on the cost correlation, the deviation transmission path is identified, and the correlation strength factor is used to analyze it. Define the transmission boundaries to achieve adaptation between the transmission range and the association strength.

[0198] Clarify the path attributes and directions according to the type of association: the dependency relationship is passed up along "sub-item → section → total item"; the cost dependency relationship is passed along "core cost → related cost"; the pricing dependency relationship is passed along "quota sub-item → pricing list item".

[0199] With correlation strength factor Based on the three-level classification standard, differentiated transmission rules are set to avoid the disorderly diffusion of deviations. Among them, strong correlation ( This can be passed to all related items, passing the attenuation coefficient. for Medium correlation ( Only directly related core items and items at the same level are allowed to pass on the attenuation coefficient. for Weak association ( (No cross-level transmission, transmission attenuation coefficient) for The formula for correcting deviation after transmission is shown below:

[0200] ,

[0201] in, For the first The list item to the first Correction deviation after passing on each associated list item; For the first The deviation of each list item, corresponding to ; To transmit the attenuation coefficient, based on the correlation strength factor Set levels and establish strong associations. Medium-term correlation weak association ; For list items and The correlation strength factor is obtained through S203 quantization calculation and takes the following value. .

[0202] For example:

[0203] "Rectangular Column" list item Related to the "Concrete and Reinforced Concrete Works" sub-item list item Correlation strength (Strong correlation), transmission attenuation coefficient Deviation of "rectangular column" project quantity Then the deviation after transmission =0.1051×0.8×0.7117≈0.1051×0.5694≈0.0598; "Rectangular Column" item With the "Installation of Miscellaneous Parts" list item Correlation strength (Weak correlation), propagation attenuation coefficient Correction deviation =0.1051×0×0.25=0, no cross-level transmission.

[0204] S403: Determination of consistency in deviation transmission.

[0205] Construct a two-dimensional judgment system of "deviation threshold + transmission matching degree", combining deviation amount, transmission path and Values ​​are used to determine consistency across dimensions.

[0206] Set a judgment threshold, where the absolute value threshold of the deviation is... Based on industry cost deviation control standards, high-precision projects can be adjusted to... ; Transmit logic matching threshold The value is manually set to ensure the adaptability of the transmission path and the association strength. .

[0207] For consistency assessment across different scenarios, only the absolute value of the deviation is checked for individual list items. The associated list items must simultaneously satisfy: the absolute value of the original and the transferred deviation ≤ ; Transmission path and Value matching degree ≥ Otherwise, it will be judged as a logical inconsistency.

[0208] Abnormal situation: The absolute value of the deviation exceeds Furthermore, it cannot be offset by transmission, and the transmission range is different. Value mismatch, contradictory transmission directions, abnormally amplified correction values ​​without standard basis.

[0209] S404: Exception list set generation and encapsulation.

[0210] Summarize the list of items with inconsistent logic, categorize them by "anomaly type - related dimension - deviation details - judgment basis", and build a standardized and traceable anomaly list set to adapt to subsequent data backfilling needs.

[0211] The core fields include the unified core identifier, name, structural role, and association strength factor of the list item. Anomaly dimension, deviation after transmission, anomaly description, judgment threshold and basis, and associated list item identifier.

[0212] Sort by a unified core identifier, cross-validate the associated cost tables, remove duplicate records, correct errors, mark the anomaly level, and finally form an anomaly list set.

[0213] For example:

[0214] The core record corresponding to the "Rectangular Column" item in the anomaly list set is: "Unified Core Identifier: 320102-0105-0200-0001; Item Name: Rectangular Column; Structural Role: Core List Item; Association Strength Factor". 0.7117; Anomaly Dimension: Quantity of Work; Deviation: 0.1051; Post-Transfer Deviation: 0.0598; Anomaly Description: Actual quantity of work is 1300m. 3 Exceeding the coordinated reference range (1170-1182m) 3 The transmission path does not match the strength of the strong correlation; the judgment threshold is: , Judgment criteria: The absolute value of the deviation exceeds the threshold and the transmission matching degree does not meet the standard; Related list item identifier: 320102-0105-0000-0001 (Concrete and Reinforced Concrete Engineering Division)

[0215] Step S5, based on the abnormal list set, combines the cost correlation, correlation strength factor, and deviation propagation judgment results of the corresponding bill of quantities items to generate verification conclusion data, and then backfills it into the historical project cost data to correct and update the collaborative cost benchmark, including:

[0216] S501: Verification conclusion data generation.

[0217] Using the anomaly list as the core data source, the cost correlation and correlation strength factors in the cost correlation table are linked. The S403 deviation transmission judgment results are used to generate standardized verification conclusion data according to the three-layer structure of "abnormal attribution - impact assessment - rectification suggestions" to ensure that the conclusions are traceable and implementable, while maintaining consistency with the data dimensions mentioned above.

[0218] Extract the unified core identifier, structural role, anomaly dimension (quantity / unit price / cost composition), deviation after transmission, and correlation strength factor from the anomaly list. The abnormal situations and judgment criteria are supplemented in the related cost table, which includes the dependency relationship types, transmission paths and constraints of the corresponding list items, forming a basic information matrix.

[0219] By combining the basic information matrix to classify and locate the causes of anomalies, the responsible parties and triggering factors are identified, and correlation strength factors are used to determine the causes. Quantify the impact of anomalies on the overall cost, including core list items ( The impact is assessed based on the proportion of abnormal deviations to the total cost of the corresponding sub-project. A proportion of ≥5% indicates a significant impact, 2%~5% indicates a moderate impact, and <2% indicates a minor impact; related list items ( Only assess the transitive impact on directly related core list items; the absolute value of the deviation after transit is ≥ Significant impact, otherwise controllable impact; Items on the supplementary list ( Only the impact of a single item is recorded; the assessment is not disseminated.

[0220] The conclusion data is packaged in the format of "list item identifier - anomaly details - cause analysis - impact assessment - rectification suggestions - responsible entity - completion deadline" to generate a structured verification conclusion report, and simultaneously outputs machine-readable data files and visualization reports.

[0221] S502: Backfilling of verification conclusion data.

[0222] Based on the anomaly type and cost structure characteristics, the verification conclusion data is classified and backfilled into the historical engineering cost database to establish a full-link data association of "anomaly-conclusion-rectification". At the same time, the database index is updated to ensure the efficiency of subsequent queries and reuse. The anomaly types include data type, rule type, and execution type, and the cost structure characteristics include composition dependency, cost dependency, and pricing dependency.

[0223] Specifically as follows:

[0224] Data-related anomaly conclusions are backfilled to the "Data Quality Control Table" in the historical database, and associated with the original data records of the corresponding list items; rule-related anomaly conclusions are backfilled to the "Benchmark Construction Parameter Table", and associated with the S303 collaborative benchmark configuration records; execution-related anomaly conclusions are backfilled to the "Pricing Operation Log Table", and associated with the corresponding operation process records.

[0225] The conclusions of dependency-related anomalies are associated with the "Quantity Relationship Table", the cost-related anomalies are associated with the "Cost Decomposition Rule Table", and the pricing-related anomalies are associated with the "Pricing Structure Configuration Table", ensuring alignment with the historical data structure.

[0226] The "incremental backfilling + cross-validation" model is adopted, which only supplements the newly added verification conclusion data and does not cover the historical original data. After backfilling, the unified core identifier of the list item is used for linkage verification to ensure the consistency between the conclusion data and the corresponding historical cost data and cost correlation, eliminate duplicate records, correct field matching errors, and record the backfilling time, operator and data source to form a backfilling log archive.

[0227] For historical items with abnormal records, add labels such as "Abnormality Identifier", "Impact Level", and "Rectification Status" to create an abnormal data archive, which will facilitate subsequent project screening and reference, such as avoiding similar abnormal pricing models.

[0228] S503: Collaborative cost benchmark revision and update.

[0229] Based on the verification conclusions after backfilling, we focus on the construction rules of the collaborative cost benchmark and the role weight of the list items. Combining the causes of anomalies and the rectification effects, we conduct dynamic correction and iterative updates to ensure that the benchmark is adapted to the cost control needs of subsequent engineering projects and improve the accuracy of verification.

[0230] The specific modifications to the construction rules are as follows:

[0231] For rule-based anomalies, based on the threshold suitability analysis in the verification conclusion, adjust the S301 similarity comprehensive threshold and the S403 deviation absolute value threshold. and the threshold for logical matching To ensure that the corrected threshold aligns with actual cost deviation patterns, a threshold correction formula is constructed based on statistical data of similar abnormal deviations. The formula is as follows:

[0232] ,

[0233] in, This is the threshold for the absolute value of the deviation after correction; The threshold for the absolute value of the original deviation is set with reference to industry cost deviation control standards. This is the sum of the absolute values ​​of the deviations of similar anomalies, taken from similar anomaly records in the anomaly list set; This refers to the number of anomalies of the same type, i.e., the number of anomaly entries of the corresponding type in the statistical anomaly list set.

[0234] To address anomalies such as range mismatch and direction conflict, optimize the differentiated transmission rules and transmission attenuation coefficient of S402. For example, the transmission attenuation coefficient for strongly related list items has been adjusted from 0.8 to 0.75 to reduce excessive transmission loss, and a "special scenario cross-level transmission approval mechanism" has been added for weakly related list items.

[0235] Supplement the auxiliary indicators for the structural division of the bill of quantities items, such as cost ratio and construction complexity, and revise the division criteria for core, related, and auxiliary bill of quantities items in S303 to avoid relying solely on the correlation strength factor. This leads to deviations in weight allocation.

[0236] Adjustment of the weight of list item effects:

[0237] Based on the weight adaptability analysis of the abnormality list items in the verification conclusion, the baseline initial weights of the list items for different structural roles were adjusted. For scenarios where abnormalities still occur after weight fine-tuning, the S303 weight fine-tuning formula was optimized by changing the fixed coefficient 0.2 to a dynamic value, i.e., a dynamic fine-tuning coefficient. To ensure that the adjustment range matches the correlation strength of the corresponding structural role list item, the dynamic coefficient formula is as follows:

[0238] ,

[0239] in, 0.2 is the dynamic fine-tuning coefficient; 0.2 is the preset basic fine-tuning amplitude coefficient, used to control the weight adjustment range. This represents the average association strength factor for each item in the list of corresponding structural roles, and is used to statistically analyze all items in the list under this structural role. The values ​​are averaged; 0.5 is the correlation strength factor. The median value of the interval, corresponding to The three-level classification standard boundary.

[0240] To verify the rationality of the weight adjustment, the adjusted weights are substituted into the S303 collaborative benchmark fusion formula to verify whether the collaborative benchmark value of the core list item conforms to the upper limit of the specification, and whether the related / subsidiary list items balance the analogy and specification requirements. A deviation of ≤3% is considered a pass.

[0241] The revised construction rules and their weights are synchronously updated to the cost association table and the historical database "Collaborative Benchmark Configuration Table," with the updated version, the basis for the revision, and the applicable project types noted. At the same time, a benchmark update documentation is generated to clarify the differences between the old and new parameters and their impact on subsequent projects, ensuring consistency in application.

[0242] S504: Modified Benchmark Verification and Application.

[0243] Cost data from similar completed projects were selected as the test set. The modified collaborative cost benchmark was used to perform a simulation verification to verify the accuracy of anomaly identification, the precision of deviation calculation, and the rationality of consistency judgment, ensuring that the correction effect met the standards.

[0244] Its verification indicators can be adaptively set according to the actual application scenario. If the indicators are not met, the system will return to S503 for correction. After the verification is passed, the corrected collaborative cost benchmark and update instructions will be pushed to the subsequent project cost verification system to clarify the effective time and scope of application. A benchmark regular update mechanism will be established and continuously optimized in combination with the newly added verification conclusion data to form a closed-loop iterative system of "verification-conclusion-backfilling-correction-verification".

[0245] refer to Figure 4 , Figure 4 Corresponding to steps S4 and S5 in this embodiment, this is a schematic diagram of deviation ripple transmission and backfill update.

[0246] Figure 4 This paper demonstrates how, with a collaborative cost benchmark as a reference, deviations are calculated for each item in the bill of quantities in terms of quantity, unit price, and cost composition. The deviations are then extrapolated along the transmission path of cost relationships. The transmission range of the deviations is limited based on the correlation strength factor, and a transmission consistency judgment is performed. When the deviations cannot form a reasonable transmission relationship that matches the correlation strength factor, they are judged as logically inconsistent and an abnormal bill of quantities set is formed. Based on the abnormal bill of quantities set, verification conclusion data is generated and backfilled into the historical engineering cost database. The construction rules of the collaborative cost benchmark and the weight of the bill of quantities items are then corrected and updated.

[0247] Example 2

[0248] This invention discloses a big data collaborative-driven engineering cost verification system, including a data acquisition module, an association construction module, a benchmark generation module, a deviation verification module, and a result update module:

[0249] The data acquisition module acquires bill of quantities data, measurement and payment data, and cost release data published by the government during the project implementation period for completed projects. It performs unified identification and alignment with the bill of quantities item level on the acquired data to form cost verification objects oriented towards the bill of quantities items.

[0250] The association construction module constructs a cost association table based on the cost verification object and according to the bill of quantities items. Based on the cost association table, it analyzes the quantity transfer characteristics and structural constraint characteristics between the bill of quantities items in terms of quantity composition, cost decomposition, and pricing structure, and establishes cost association relationships that reflect the dependence of cost elements between bill of quantities items. Based on the degree of transfer dependence of quantity changes, cost composition ratio, and coupling characteristics of pricing structure between bill of quantities items, the module performs quantitative calculations on the cost association relationships to obtain the association strength factor that characterizes the degree of association between cost elements between bill of quantities items.

[0251] The benchmark generation module performs cost structure similarity analysis on historical engineering cost data based on the cost correlation and correlation strength factor, and selects reference engineering projects that match the current engineering project in terms of cost structure characteristics and engineering implementation cycle, thus constructing an analogy benchmark for horizontal benchmarking. Simultaneously, based on the cost correlation and engineering implementation cycle characteristics, it analyzes the constraints of applicable engineering cost specifications and engineering cost release data during the engineering implementation period in terms of pricing elements, time intervals, and applicable conditions, forming a set of specification constraint rules consistent with the current engineering project's cost structure, and constructing a specification benchmark to limit the reasonable range of engineering costs. Furthermore, combining the structural roles of different bill of quantities items in the cost correlation and their corresponding correlation strength factors, it differentiates the weighting of the analogy benchmark and specification benchmark on each bill of quantities item, forming a collaborative cost benchmark oriented towards bill of quantities items.

[0252] Based on the cost association table, the deviation verification module calculates the deviation of each bill of quantities item from the collaborative cost benchmark in terms of quantity, unit price, and cost composition. Combining the characteristics of the transmission path of the deviation in the cost association relationship, and limiting the range of transmission of the deviation between bill of quantities items according to the association strength factor, the module judges the consistency of the transmission of the deviation in the cost element association structure. When the deviation cannot form a reasonable transmission relationship that matches the association strength factor in the cost association relationship, it is determined that the corresponding bill of quantities item has logical inconsistency, and an abnormal bill of quantities set is formed.

[0253] The result update module, based on the anomaly list set, combines the cost correlation, correlation strength factor, and deviation propagation judgment results of the corresponding bill of quantities items to generate verification conclusion data for engineering cost verification. The verification conclusion data is then backfilled into the historical engineering cost data according to the anomaly type and cost structure characteristics. The construction rules of the collaborative cost benchmark and its role weight on each bill of quantities item are corrected and updated for use in subsequent engineering cost verification of engineering projects.

[0254] This invention discloses a big data-driven collaborative engineering cost verification system and method. By uniformly identifying and aligning bill of quantities (BOQ) data, measurement and payment data, and cost publication data at the BOQ item level, a cost verification object is formed for each BOQ item. Based on this, the system analyzes the dependencies between BOQ items in terms of quantity composition, cost decomposition, and pricing structure, constructing cost correlation relationships and calculating correlation strength factors to characterize the transmission characteristics of cost elements between BOQ items. Furthermore, it constructs an analogy benchmark by combining historical engineering cost data and a standard benchmark by combining applicable cost specifications within the project implementation cycle, collaboratively forming a cost benchmark under the constraints of BOQ item structure and correlation strength. By calculating the deviation of each BOQ item from the collaborative cost benchmark and determining its transmission consistency in cost correlation relationships, it identifies logically inconsistent abnormal BOQ items. Finally, the verification conclusions are backfilled and used for benchmark correction and updating, thereby achieving structured, interpretable, and continuously optimized engineering cost verification.

[0255] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A big data-driven collaborative method for engineering cost verification, characterized in that, Includes the following steps: Acquire bill of quantities data, measurement and payment data, and cost release data; apply unified identification and alignment to the bill of quantities item level for the three types of data to form cost verification objects oriented towards bill of quantities items; Based on the cost verification object, a cost correlation table is constructed. The quantity transfer characteristics and structural constraint characteristics of the bill of quantities items in terms of quantity composition, cost decomposition and pricing structure are analyzed. Cost correlation relationships are established and correlation strength factors are quantified. By combining historical engineering cost data to perform cost structure similarity analysis, an analogy benchmark is constructed. The constraints of engineering cost specification documents and cost release data are analyzed to construct a standard benchmark. The weights of the analogy benchmark and the standard benchmark are differentiated according to the structure of the list items and the correlation strength factor to form a collaborative cost benchmark. Calculate the deviation of each item in the bill of quantities from the collaborative cost benchmark in terms of quantity, unit price, and cost composition. Combine this with the correlation strength factor to limit the transmission range and determine the transmission consistency to form an abnormal bill of quantities set. Based on the anomaly list set, the historical engineering cost data is backfilled according to the anomaly type and cost structure characteristics, and the construction rules of the collaborative cost benchmark and the role weight of each list item are corrected and updated. Among them, establishing cost correlation and quantifying the correlation strength factor includes: Based on the cost association table, dependency parsing is performed around the composition of engineering quantities, cost decomposition, and pricing structure among the bill of quantities items to form a set of cost association relationships that include association type, association direction, and constraint conditions; Based on cost correlation, a quantitative system is constructed that considers the degree of dependence on changes in project quantity, the proportion of cost composition, and the coupling characteristics of pricing structure. The quantitative results of each dimension are standardized and fused according to preset weights to obtain the correlation strength factor. Calculate the deviation and determine the transmission consistency, including: Using the collaborative cost benchmark as a reference, calculate the deviation of each item in the bill of quantities in terms of quantity, unit price, and cost composition, and record the direction of deviation, calculation basis, and benchmark value. Based on the cost correlation, the deviation transmission path is sorted out, the transmission direction is defined according to the correlation type, the transmission range is set based on the correlation strength factor, and the transmission attenuation coefficient is configured to obtain the deviation after transmission. Construct a criterion for determining deviation threshold and transmission matching degree, perform consistency determination on the deviation amount after transmission and the adaptation relationship of transmission path, and output a list of logically inconsistent items; The formula for correcting deviation after transmission is shown below: ， in, For the first The list item to the first Correction deviation after passing on each associated list item; For the first Deviation of each item in the list; To transmit the attenuation coefficient, based on the correlation strength factor Set levels; For list items and The correlation strength factor.

2. The big data collaborative-driven engineering cost verification method as described in claim 1, characterized in that, Establishing a collaborative cost benchmark includes: Based on cost correlation and correlation strength factors, cost structure similarity analysis is performed on historical engineering cost data to screen reference engineering projects and aggregate them into an analogy benchmark by unified core identifiers of the list items; Based on the project implementation cycle, we analyze the constraints of engineering cost specification documents and cost release data on pricing elements, time intervals, and applicable conditions, form a set of specification constraint rules, and encapsulate them into a specification benchmark by uniform core identifiers of list items. Based on the structural role of the list items in the cost association structure, the weights of the analogy benchmark and the standard benchmark on the list items are configured, and the weights are fine-tuned by combining the association strength factor to obtain the collaborative cost benchmark for the list items.

3. The big data collaborative-driven engineering cost verification method as described in claim 1, characterized in that, Create an exception list set, including: Items in the list that are determined to be logically inconsistent are categorized and organized according to anomaly type, related dimension, deviation details, and judgment basis to form an anomaly list set; The exception list set records the unified core identifier of the list item, the list item name, the structural role, the association strength factor, the exception dimension, the deviation after propagation, the exception description, and the associated list item identifier, and is sorted and encapsulated according to the unified core identifier.

4. The big data collaborative-driven engineering cost verification method as described in claim 2, characterized in that, Constructing analog benchmarks includes: Historical project data is extracted from the historical engineering cost database, and the historical project data is processed by unified identification and completion, list item alignment and data standardization to form a standardized historical project dataset; Based on the cost correlation and correlation strength factor between this dataset and the current engineering project, a similarity evaluation system is constructed to screen a set of reference engineering projects that meet the comprehensive similarity threshold. Based on the current project list items, the benchmark quantities, unit prices, and cost composition data of the corresponding list items in the reference project set are collected, and the statistical representative values ​​are calculated according to the unified core identifier of the list items, and then packaged to form an analog benchmark.

5. The big data collaborative-driven engineering cost verification method as described in claim 2, characterized in that, Establish standard benchmarks, including: Organize the applicable engineering cost standards and cost data according to the project implementation cycle, project type and regional attributes, and eliminate inapplicable documents and data; The standard documents and cost release data are analyzed according to pricing elements, time intervals, and applicable conditions to form a set of structured standard constraint rules, and the rule associations are established by using a unified core identifier for list items; Based on a set of normative constraint rules, reasonable ranges are defined for the quantity, unit price, and cost composition of each item in the bill of quantities, and these ranges are encapsulated according to the unified core identifier of each item to form a normative benchmark.

6. The big data collaborative-driven engineering cost verification method as described in claim 1, characterized in that, The objects to be included in the cost verification include: Perform original verification on bill of quantities data, measurement and payment data, and cost release data to create a data anomaly ledger; Construct a hierarchical unified identification system, assign a unified core identifier to bill of quantities data, assign a core identifier and payment cycle identifier to bill of quantities items to measurement and payment data, assign a time dimension identifier and data type identifier to cost release data, and establish a unified identifier index table; Using the bill of quantities data as the alignment benchmark, the bill of quantities item-level alignment process is performed according to the item code, project characteristic description, unit of measurement, and pricing quota sub-item, and integrated to form a cost verification object oriented towards the bill of quantities item.

7. The big data collaborative-driven engineering cost verification method as described in claim 6, characterized in that, List item-level alignment processing includes: The corresponding list item is located by using a unified core identifier matching method, and auxiliary element verification of the list item name and unit of measurement is performed to establish synonym mapping relationship and include it in the alignment mapping ledger; A conflict classification and processing mechanism is implemented for conflicting data during the alignment process. For conflicts in units of measurement, conversion adjustments are performed and the calculation trajectory is retained. For conflicts in applicable scope, realignment and screening processing is performed, and the basis for correction, correction method, and processing information are recorded.

8. A big data collaborative engineering cost verification system, used to implement the big data collaborative engineering cost verification method according to any one of claims 1-7, characterized in that, It includes a data acquisition module, a correlation construction module, a benchmark generation module, a deviation verification module, and a result update module: The data acquisition module obtains bill of quantities data, measurement and payment data, and cost release data, and performs unified identification and alignment of the three types of data with the bill of quantities item level to form cost verification objects oriented towards bill of quantities items; The association construction module constructs a cost association table based on the cost verification object, analyzes the quantity transfer characteristics and structural constraint characteristics between the items in the bill of quantities in terms of quantity composition, cost decomposition and pricing structure, establishes cost association relationships and quantifies the association strength factor; The benchmark generation module combines historical engineering cost data to perform cost structure similarity analysis, constructs analog benchmarks, parses engineering cost specification documents and cost release data constraints, constructs standard benchmarks, and differentiates the role weights of analog benchmarks and standard benchmarks according to the list item structure and correlation strength factors to form collaborative cost benchmarks; The deviation verification module calculates the deviation of each item in the bill of quantities from the collaborative cost benchmark in terms of quantity, unit price, and cost composition. It then combines the correlation strength factor to limit the transmission range and determine the transmission consistency, thus forming an abnormal list set. The results update module, based on the anomaly list set, backfills historical engineering cost data according to anomaly type and cost structure characteristics, and corrects and updates the construction rules of the collaborative cost benchmark and the role weights of each list item.

Images