A method and system for constructing an engineering cost price database based on dynamic market data updates.

Abstract

This invention discloses a method and system for constructing an engineering cost price database based on dynamic market data updates, belonging to the field of engineering cost data processing technology. The method includes: establishing a standard price item system; collecting information price data, historical transaction price data, and real-time market price data, and mapping them to standard price items; validating the source price data and determining its reliability; determining the update trigger state based on the deviation of each source price data from the benchmark price; generating updated prices using information price data as benchmark anchor data, historical transaction price data as stability correction data, and real-time market price data as fluctuation correction data; selecting the corresponding update method to write the data into the price database, and simultaneously generating price version records. Through the above methods, the accuracy of the price database's response to market fluctuations, update stability, and traceability for subsequent verification are improved.

Description

Technical Field

[0001] This invention relates to the field of engineering cost data processing and engineering pricing information management, specifically to a method and system for constructing an engineering cost price database based on dynamic market data updates. Background Technology

[0002] The engineering cost price database is a crucial source of fundamental data for engineering investment estimation, design budget, construction drawing budget, maximum bid limit preparation, bid price analysis, and project settlement review. In current engineering cost estimation practices, it is typically necessary to establish corresponding price entries for building materials, equipment, components, and related professional sub-items, and to call upon the relevant price data during the cost estimation process to improve the standardization and efficiency of the pricing work.

[0003] In existing technologies, the price sources for engineering cost databases are mainly periodically published information prices, supplemented by procurement prices, contract prices, or manual inquiry results from historical projects. This approach can meet basic usage needs in early engineering cost management, but with the increasing frequency of material market fluctuations, greater differences in equipment selection, and increasingly significant differences in project implementation regions, times, and supply conditions, the traditional method of constructing price databases based primarily on static price information has gradually revealed its insufficient adaptability.

[0004] Specifically, existing engineering cost price databases typically suffer from the following shortcomings during the update process: First, price data from different sources lacks unified standards in terms of material names, specifications, units of measurement, regional affiliation, tax rates, and collection times, making it difficult to directly compare the same price item across different data sources, thus affecting the consistency and reusability of price entries. Second, existing price database update mechanisms mostly employ periodic overall updates or manual screening followed by updates, resulting in low update frequency and difficulty in timely reflecting market price changes, making it easy for discrepancies to arise between prices in the database and actual procurement prices. Third, when there are differences between real-time market inquiries, historical transaction prices, and information prices, existing technologies often lack the ability to accurately reflect these discrepancies. The lack of mechanisms for assigning roles and integrating different price sources can easily lead to situations where price updates rely excessively on a single source, result in large fluctuations in update results, or cause update delays. Fourth, for abnormal data such as short-term abnormal quotations, abnormal prices in local areas, and pseudo-similar prices caused by inconsistent specifications, the existing price database often lacks effective means of identification, limitation, and pending verification, which can easily lead to abnormal prices directly entering the price database, affecting the stability and reliability of subsequent pricing results. Fifth, the existing price database usually only retains the latest price after a price update, lacking a complete record of the update source, update trigger basis, update method, and historical versions, which is not conducive to subsequent cost audits, price reviews, and dispute tracing.

[0005] Therefore, how to establish a unified mapping, reliable evaluation, trigger judgment, integrated update, and traceable management mechanism for the construction and updating of engineering cost price database based on information price data, historical transaction price data, and real-time market price data, so as to ensure the timeliness of price updates while taking into account price stability and verifiability, has become an urgent technical problem to be solved in this field. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for constructing an engineering cost price database based on dynamic market data updates. This addresses the problems in existing engineering cost price databases, such as inconsistent definitions of price data from different sources, delayed price updates, difficulty in identifying and suppressing abnormal prices, and lack of version traceability for update results. The invention achieves unified mapping, reliable assessment, trigger determination, integrated updates, and version management of information price data, historical transaction price data, and real-time market price data. This improves the accuracy of the engineering cost price database's response to market price fluctuations, the stability of its updates, and the traceability of subsequent verification.

[0007] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution: A method for constructing an engineering cost price database based on dynamically updated market data, comprising the following steps: S1. Establish a standard price item system and uniformly define the material name, specifications, unit of measurement, applicable region, tax calculation caliber, and time granularity corresponding to the price items in the engineering cost price database; S2. Collect the corresponding information price data, historical transaction price data and real-time market price data for the target price item, and map the information price data, historical transaction price data and real-time market price data to the standard price item system; S3. Perform validity verification on each source price data after mapping to obtain the source credibility of each source price data. The source credibility is determined at least based on data integrity, time validity, regional matching degree, specification matching degree and degree of dispersion anomaly. S4. Using the current in-stock price of the target price item as the benchmark price, calculate the degree of deviation of the information price data, historical transaction price data and real-time market price data from the benchmark price, and determine the update trigger state by combining the source credibility corresponding to each source price data. S5. When the update trigger state meets the update conditions, the target price entry is fused and updated using the information price data as the benchmark anchor data, the historical transaction price data as the stability correction data, and the real-time market price data as the fluctuation correction data to generate an updated price. S6. Based on the update trigger status and the deviation consistency of each source price data, select the corresponding update method from the three update methods of direct update, limited update and pending update and write the updated price into the engineering cost price database. S7. Synchronously record the update time, data sources involved in the update, source credibility of each source price data, update method used, and price difference before and after the update for the target price entry after it is written, so as to generate the corresponding price version record. The fusion update includes: within the benchmark price range determined by the information price data, using historical transaction price data to suppress short-term abnormal fluctuations, and using real-time market price data to correct the current price deviation direction and deviation magnitude within the benchmark price range, thereby improving the accuracy of the price database's response to market fluctuations and maintaining the stability of price updates.

[0008] Optionally, the information price data is material or equipment price information released by the cost management agency, the historical transaction price data is the transaction price information of the corresponding price item in the completed project, the settled project or the signed purchase order, and the real-time market price data is at least one of supplier quotation data, inquiry feedback data or market-released price data.

[0009] Optionally, the mapping in step S2 includes: Standardize and merge the name fields of the price data from each source; Perform parameter splitting and standardized recombination on the specification and model field; Standardize the conversion of units of measurement; The prices including tax and prices excluding tax will be converted uniformly. Perform regional matching between the data collection area and the applicable area; After the mapping is completed, a corresponding standard price entry identifier is generated for each source price data.

[0010] Optionally, the source credibility in step S3 is determined according to the following rules: When the source price data has any of the following issues: missing fields, time limit exceeded, region mismatch, specification mismatch, or abnormal price distribution, the credibility of the corresponding source will be reduced. When the source price data has complete fields, valid time, matching within the same region, matching within the same specification, and the price dispersion is within the set range, the credibility of the corresponding source is increased.

[0011] Optionally, the update trigger state in step S4 is determined based on the following conditions: When the deviation of real-time market price data from the benchmark price reaches the first threshold, and at least two types of data, including information price data, historical transaction price data, or real-time market price data, support the same direction of deviation, it is determined to be in a direct update trigger state. When the deviation of real-time market price data from the benchmark price reaches the second threshold but there is insufficient supporting data, it is determined to be in a state of pending core update. When the deviation of the real-time market price data from the benchmark price does not reach the first threshold and the second threshold, the original in-stock price is maintained or a limited update is performed.

[0012] Optionally, the fusion update in step S5 includes: First, determine the basic price range for the target price item based on the aforementioned information price data; Then, based on the historical transaction price data, determine the stable correction range within the basic price range; Finally, the instantaneous correction value within the stable correction range is determined based on the real-time market price data. The updated price is within the stability correction range and does not exceed the allowable fluctuation boundary corresponding to the base price range.

[0013] Optionally, the limit update in step S6 is as follows: When the update trigger state meets the update conditions but the fluctuation of the real-time market price data exceeds the preset limit threshold, the price adjustment range is limited according to the preset limit threshold to avoid a sudden update of the price database caused by a single abnormal fluctuation.

[0014] Optionally, the kernel to be updated is: When the deviation between the real-time market price data of the target price item and the benchmark price reaches a preset abnormal threshold, and the historical transaction price data and the information price data do not form a consistent support, the original in-database price will be suspended from being covered, and a pending verification mark item will be generated. The entries to be verified include at least the anomaly source identifier, the anomaly price value, the data items to be supplemented, and the manual review entry.

[0015] Optionally, the price version record in step S7 further includes: The target price entry includes the update batch number, the effective time of the updated price, the expiration time, the update reason classification, and the price recovery path; The price recovery path is used to revert the target price item to the previous valid version when the current version is determined to be abnormal during subsequent review.

[0016] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution: A system for constructing an engineering cost price database, comprising: The standard item construction module is used to establish a standard price item system; The multi-source acquisition module is used to collect information price data, historical transaction price data, and real-time market price data; The standard mapping module is used to map price data from various sources to a standard price item system. The credibility assessment module is used to determine the source credibility of each source price data. The trigger determination module is used to determine the update trigger status based on the degree of deviation and consistency of deviation of each source price data relative to the benchmark price. The fusion update module is used to generate updated prices by using information price data as benchmark anchor data, historical transaction price data as stability correction data, and real-time market price data as fluctuation correction data. The update execution module is used to select the corresponding update method from three update methods: direct update, limited update, and pending update, and write it into the engineering cost price database. The version tracking module is used to generate price version records; The updated price generated by the fusion update module is constrained by the benchmark price range corresponding to the information price data and limited by the stable correction range corresponding to the historical transaction price data. The real-time market price data is only used to correct the current price offset direction and offset magnitude within the benchmark price range.

[0017] The main advantages of this invention compared to existing technologies are as follows: This invention establishes a standard price item system, collects and maps information price data, historical transaction price data, and real-time market price data respectively, performs validity verification, update trigger determination, and fusion update on price data from each source, and allows selection of three update methods: direct update, limited update, and pending verification update. This effectively solves the problems of inconsistent caliber of price data from different sources, delayed price updates, difficulty in identifying and suppressing abnormal prices, and lack of version traceability in existing engineering cost price databases. As a result, it improves the accuracy of the engineering cost price database in responding to market price fluctuations, the stability of updates, and the traceability of subsequent verification.

[0018] This invention improves the comparability and consistency of multi-source price data by uniformly defining material names, specifications, units of measurement, applicable regions, tax calculation methods, and time granularity, and mapping price data from different sources to a standard price item system. This avoids the problem of price items of the same type being unable to be directly compared or duplicate databases being built due to differences in data caliber.

[0019] This invention assigns different data roles to information price data, historical transaction price data, and real-time market price data. Information price data is used for benchmark anchoring, historical transaction price data is used for stabilization correction, and real-time market price data is used for fluctuation correction. This can suppress price abrupt changes caused by a single data source while taking into account the responsiveness to market changes, thereby improving the rationality and stability of price update results.

[0020] This invention determines the source credibility of each source price data based on data integrity, time validity, regional matching degree, specification matching degree, and degree of dispersion anomaly, and determines the update trigger state by combining the deviation of each source price data from the benchmark price. This can improve the pertinence of price update judgment and avoid low-quality data, expired data, or pseudo-similar data from directly participating in price updates.

[0021] This invention sets up three update methods: direct update, limited update, and pending update. It can adopt differentiated processing strategies for price changes with different degrees of deviation and different support conditions. It can adjust the price database in a timely manner when the market price fluctuates in a real way, and retain the space for manual review when abnormal quotations or insufficient support occur, thereby reducing the risk of abnormal price propagation.

[0022] This invention records the update time, data sources involved in the update, source credibility of each price data source, update method used, and price difference before and after each price update, thus forming a price version record. This provides clear evidence for subsequent cost audits, price reviews, dispute resolution, and historical retrospective processes, improving the management standardization and traceability of the engineering cost price database. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the steps of the method for constructing an engineering cost price database based on dynamically updated market data according to the present invention. Figure 2 This is a structural diagram of the engineering cost price database construction system of the present invention. Detailed Implementation

[0024] The following is a detailed implementation method and system for constructing an engineering cost price database based on dynamically updated market data. See Figure 2 As shown, the system architecture for constructing the engineering cost price database of the present invention includes: The standard item construction module is the foundational framework of the entire system. It is responsible for uniformly defining the key attributes related to price items in the engineering cost price database, covering aspects such as material name, specifications, unit of measurement, applicable region, tax calculation method, and time granularity.

[0025] When defining material names, a strict naming standard is established. For example, for steel, it is clearly and uniformly named as "[material][specification] steel", such as "HRB400EΦ12 steel", to avoid non-standard or easily confused names such as "rebar 12" or "HRB400 round steel (actually rebar)".

[0026] For specifications and models, the method and order of representing each parameter will be specified in detail. Taking a certain brand of air conditioner as an example, the specification is "[Brand][Cooling Capacity (kW)]-[Heating Capacity (kW)]-[Energy Efficiency Rating]-[Horsepower] Air Conditioner". This can clearly and accurately represent the product specifications and avoid confusion caused by different manufacturers or sources of data due to different representation methods.

[0027] Regarding units of measurement, a unified conversion standard will be established, such as converting length to "m" and weight to "kg". Regardless of whether the original data is presented in "cm", "t" or other units, it must be converted to the standard unit.

[0028] The applicable regions will be specified down to the specific administrative divisions, such as provinces, cities, and districts, clearly defining the scope of application for different price items and avoiding the misapplication of price data in inapplicable regions.

[0029] The tax calculation method will be uniformly determined based on relevant national tax policies and industry practices, including whether it is a price including tax or excluding tax, as well as the standard for determining the tax rate.

[0030] The time granularity specifies the time period for price data, such as whether it is updated daily, weekly, monthly, or quarterly. Through these unified definitions, a clear, standardized, and unified system of standard price entries is established, providing a foundation for the subsequent integration and processing of multi-source price data.

[0031] The multi-source acquisition module plays a crucial role in collecting various types of price data. It gathers information price data, historical transaction price data, and real-time market price data corresponding to the target price item from multiple channels.

[0032] The price information comes from material and equipment price information regularly released by local cost management agencies. This information is usually presented in the form of documents, website announcements, etc. The data collection module uses web crawling technology and data interface integration to periodically obtain price information data from the official website of the cost management agency or designated data sources. For example, at the beginning of each month, it automatically downloads the latest price information document from the local engineering cost information website and extracts the price information of relevant materials and equipment.

[0033] Historical transaction price data is obtained from project documents of completed and settled projects, as well as signed purchase orders. This requires data integration with the company's project management system and contract management system. By setting query conditions, such as project name, material or equipment name, and contract signing time, the transaction price information of the corresponding price items can be extracted from these systems. For example, the transaction price records of "HRB400EΦ12 steel" in all completed construction projects within the past year can be retrieved from the company's project management database.

[0034] Real-time market price data is collected from supplier quotations, inquiry feedback, or market-published price data. This can be achieved by establishing data interfaces with suppliers to obtain their latest quotations in real time; by using online inquiry platforms to send inquiry requests to multiple suppliers and collect feedback data; or by collecting relevant data from professional market price publishing platforms. For example, by connecting with the systems of major steel suppliers, the latest quotations for "HRB400EΦ12 steel" can be obtained in real time; simultaneously, inquiries can be sent to multiple suppliers on building materials inquiry platforms to collect feedback prices from different suppliers for that steel.

[0035] The standard mapping module receives price data from various sources acquired by the multi-source acquisition module and maps it to the standard price item system established by the standard item construction module.

[0036] For the name field, the material names from different sources of price data are uniformly converted to standard names according to the standard name merging rules. For example, non-standard names such as "rebar 12" and "HRB400 round 12" are uniformly merged into "HRB400EΦ12 steel".

[0037] In processing the specification and model field, the original data is first split into parameters. For example, for "a certain brand of 3 horsepower air conditioner", it is split into parameters such as brand, cooling capacity, heating capacity, energy efficiency rating, and horsepower. Then, it is standardized and reorganized according to the standard format, and converted into the standard specification and model representation of "[brand][cooling capacity (kW)]-[heating capacity (kW)]-[energy efficiency rating]-[horsepower] air conditioner".

[0038] Standardized conversion of units of measurement, such as converting steel weight data in "t" to "kg", and plate area data in "cm²" to "m²", ensures consistency of all data in terms of units of measurement.

[0039] For prices including tax and prices excluding tax, a unified conversion rule is applied. If the standard tax calculation method specifies a price excluding tax, then the price including tax is converted to a price excluding tax according to the corresponding tax rate; conversely, the same applies.

[0040] In terms of regional matching, the data collection area will be matched with the applicable area. If the data collected is from City A in a certain province, but the standard price item applies to City B in the same province, it is necessary to determine whether the price data is applicable to City B. If it is applicable, the mapping will be performed; if it is not applicable, it will be marked as an anomaly.

[0041] After completing the above processing, corresponding standard price entry identifiers are generated for each source price data to facilitate subsequent data identification and management, ensuring that each price data can accurately correspond to the corresponding position in the standard price entry system.

[0042] The credibility assessment module verifies the validity of each source price data after it has been processed by the standard mapping module, thereby determining the source credibility of each source price data.

[0043] When source price data has missing fields, such as price or specifications, its reliability is reduced. For example, if a steel price record only provides the price and not the specifications, its reliability will be significantly lower.

[0044] Exceeding time limits is also an important factor affecting credibility. If the information price data has exceeded its validity period, such as an information price document that was published 3 months ago but has a stipulated validity period of 1 month, then the credibility of the data will be reduced.

[0045] Regional mismatch can also lead to decreased credibility. For example, if price data is collected from a northern province, but the target price item applies to a southern province, its credibility will decrease if there is insufficient evidence to show that the price applies to the southern province.

[0046] Specification mismatch should not be ignored. If the specifications of an air conditioner price data are significantly different from those specified in the standard entry, such as inconsistent parameters such as horsepower or energy efficiency rating, the reliability of the data is reduced.

[0047] Abnormal price distribution can also affect credibility. If the price of a material deviates greatly or slightly from the price of similar materials without a reasonable explanation, such as the price of a certain brand of ordinary steel suddenly being several times higher than that of similar products on the market, the credibility of the data will decrease.

[0048] Conversely, source price data with complete fields, a valid publication date, matching collection and application regions, consistent specifications and standards, and price dispersion within a set range will have increased source credibility. For example, if a steel price data point has all fields complete, was published last week, is in the same application and collection region, conforms to standards, and its price is similar to recent prices of similar products, its credibility will be higher. In this way, each source price data point is assigned a source credibility value that reflects its reliability, providing crucial information for subsequent update trigger determination.

[0049] The trigger determination module uses the current in-stock price of the target price entry as the benchmark price, calculates the deviation of the information price data, historical transaction price data, and real-time market price data from the benchmark price, and combines the source credibility of each source price data to determine the update trigger status.

[0050] When the deviation of real-time market price data from the benchmark price reaches a first threshold (e.g., set to 20%), and at least two types of data—information price data, historical transaction price data, or real-time market price data—support the same direction of deviation, a direct update trigger state is established. For example, if real-time market price data shows that the price of "HRB400EΦ12 steel" has increased by 25% compared to the price in inventory, and information price data also shows a certain increase in price, while historical transaction price data, although showing a smaller increase, also exhibits an upward trend, then the direct update trigger state is met.

[0051] When the deviation of real-time market price data from the benchmark price reaches the second threshold (e.g., set to 30%) but there is insufficient supporting data, it is determined to be in a pending update trigger state. For example, real-time market price data shows that the price has fallen by 35%, but the information price data and historical transaction price data do not change significantly and do not form a corresponding support, thus entering the pending update trigger state.

[0052] When the deviation of real-time market price data from the benchmark price does not reach the first or second threshold, the original price in the database is maintained or a limited update is performed. For example, if the real-time market price data fluctuates within 10%, the original price in the database can be maintained; if the fluctuation is between 10% and 20%, a limited update can be performed to restrict the price adjustment range and avoid excessive price fluctuations. Through this judgment mechanism, it is possible to accurately determine whether the price database needs to be updated and what update method should be adopted based on different market price changes.

[0053] When the update trigger state meets the update conditions, the fusion update module uses information price data as the benchmark anchor data, historical transaction price data as the stability correction data, and real-time market price data as the fluctuation correction data to perform a fusion update on the target price item and generate an updated price.

[0054] First, determine the base price range for the target price item based on the information price data. For example, if the information price data shows that the price range for "HRB400EΦ12 steel" is 4000-4500 yuan / ton, this range provides a general framework for price updates.

[0055] Next, a stable correction range within the base price range is determined based on historical transaction price data. Assuming historical transaction price data indicates that the reasonable transaction price range for this steel product over a period of time was 4100-4400 yuan / ton, this range constitutes the stable correction range within the base price range of the information price. It can suppress excessive price deviations caused by short-term market fluctuations or abnormal pricing.

[0056] Finally, an immediate correction value within the stable correction range is determined based on real-time market price data. If the real-time market price is 4300 yuan / ton and within the stable correction range, this price can be used as the immediate correction value to comprehensively determine the updated price. The updated price is within the stable correction range and does not exceed the allowable fluctuation boundary corresponding to the base price range, thus ensuring both price responsiveness to market fluctuations and maintaining price stability.

[0057] The update execution module selects the corresponding update method from three update methods—direct update, limited update, and pending update—based on the update trigger status determined by the trigger judgment module and the deviation consistency of each source price data, and writes the updated price into the engineering cost price database.

[0058] When the direct update is triggered, the updated price generated by the fusion update module is directly written into the price database. For example, if the direct update is triggered as described above, and the price after fusion update is 4350 yuan / ton, this price is directly updated to the price entry for "HRB400EΦ12 steel".

[0059] For price limit updates, when the update trigger status meets the update conditions but the fluctuation range of the real-time market price data exceeds a preset price limit threshold (e.g., set to 25%), the price adjustment range is limited according to the preset price limit threshold. For example, if the real-time market price data shows a 30% price increase, but the price limit threshold is 25%, the updated price is calculated based on the price in the database, with a 25% increase, and written to the price database. This prevents a single abnormal fluctuation from causing a sudden change in the price database update.

[0060] When in the pending update trigger state, the pending update is as follows: when the deviation between the real-time market price data of the target price item and the benchmark price reaches a preset anomaly threshold (e.g., set to 30%), and the historical transaction price data and the information price data do not form a consistent support, the overwriting of the original price in the database is suspended, and a pending update mark item is generated. A pending update mark item includes at least an anomaly source identifier (e.g., an abnormal quote from a supplier), an abnormal price value (e.g., a price 35% higher than the benchmark price), data items to be supplemented (e.g., quotes from other suppliers, market supply and demand, etc.), and a manual review entry. Through manual review, the reasonableness of the abnormal price is determined, and then a decision is made on whether to update the price database, reducing the risk of abnormal price propagation.

[0061] The version tracking module synchronously records the update time, data sources involved in the update, source credibility of each source price data, update method used, and price difference before and after the update for each target price entry after it is written, in order to generate the corresponding price version record.

[0062] The price version record further includes the update batch number of the target price item, the effective time of the updated price, the expiration time, the update reason classification, and the price recovery path. For example, for the price item "HRB400EΦ12 steel", the update batch number is 005, the update time is May 10, 2024, and the data sources involved in the update include the information price released by a local cost management agency, the historical transaction prices of three completed projects, and the real-time market prices of five suppliers. The credibility of each source is 0.9 for the information price, 0.85 for the historical transaction price, and 0.8 for the real-time market price. The update method used is direct update, the price difference before and after the update is +350 yuan / ton, the effective time of the updated price is May 10, 2024, the expiration time is not yet set, and the update reason classification is normal market price fluctuation. The price recovery path is used to revert the target price item to the previous valid version when the current version is determined to be abnormal during subsequent review. For example, by recording the price data and update operation records of the previous version, a one-click rollback can be achieved, providing a clear basis for subsequent cost audits, price reviews, dispute resolution, and historical retrospective processes, and improving the management standardization and traceability of the engineering cost price database.

[0063] like Figure 1 As shown, the method for constructing an engineering cost price database based on dynamic market data updates according to the present invention includes: Step S1: Establish a standard price item system In practice, professional cost estimators and data managers first jointly develop standard definition rules for the attributes related to price items. For material names, a unified naming convention is determined by referring to industry standards, authoritative cost data, and commonly used market names. For example, for building bricks, based on their material, specifications, and uses, they are uniformly named "[Material][Specifications (mm)][Use] Brick", such as "Shale 240×115×53 Load-bearing Brick".

[0064] Regarding specifications and models, we communicated with manufacturers of various materials and equipment, as well as industry associations, to clarify the accurate representation methods and order of each parameter. Taking elevators as an example, the standard is "[Brand][Rated Load Capacity (kg)]-[Rated Speed ​​(m / s)]-[Number of Floors] Elevator", ensuring that elevator specifications and models from different sources can be accurately standardized.

[0065] The unified conversion rules for units of measurement refer to national measurement standards. For example, length units are uniformly converted to "m", area units to "m²", and volume units to "m³". For units of measurement of special materials or equipment, such as cables which use "m" as the basic unit, but may be sold in "reel" in some cases, the conversion relationship between "reel" and "m" needs to be clarified, such as each reel of cable being 100m long.

[0066] The applicable regions are defined according to national administrative division standards, accurate to the city and district level. Factors such as economic development levels and market price differences in different regions are also considered to reasonably define the scope of application for each price item. For example, for certain materials significantly affected by regional transportation costs, separate price items will be established for different regions.

[0067] The tax calculation method is determined based on national tax policies and common practices in the construction cost industry, establishing a unified standard for whether the price includes tax or excludes tax, as well as the tax rate. For example, since the construction industry currently widely uses value-added tax (VAT) pricing, the prices in the price database are explicitly defined as VAT-exclusive, and the tax rate is determined according to the VAT rates stipulated by the state for the construction industry.

[0068] The time granularity is determined based on market price fluctuations and the needs of engineering cost estimation. For materials with frequent price fluctuations, such as steel and cement, the time granularity is set to be updated weekly; for materials with relatively stable prices, such as some decorative materials, the time granularity is set to be updated monthly. Through the formulation of these detailed rules, a comprehensive, standardized, and unified standard price item system is established, laying a solid foundation for subsequent data collection and processing.

[0069] Step S2: Collect and map multi-source price data The multi-source acquisition module obtains price data from various channels according to a predetermined acquisition strategy. For information price data, a web crawler program is written to periodically download the latest price information files, such as XML and Excel formats, from the official websites of local cost management agencies. After downloading, data parsing tools are used to extract price information for relevant materials and equipment, including fields such as material name, specifications, price, and release date. For example, downloading the monthly information price file from a provincial engineering cost information website reveals that the information price for "HRB400EΦ12 steel" is 4200 yuan / ton, and the release date is April 30, 2024.

[0070] Historical transaction price data is collected through data integration with the company's project management system and contract management system. Query conditions are set within the system, such as querying the transaction price records of "HRB400EΦ12 steel" in all completed construction projects within the past two years. Contract documents and settlement documents for relevant projects are retrieved from these systems, and the transaction price information is extracted, while also recording relevant information such as project name, contract signing date, and supplier. For example, the contract documents of a completed project show a transaction price of 4150 yuan / ton for this steel, a contract signing date of November 2023, and the supplier as XX Steel Co., Ltd.

[0071] Real-time market price data is collected through multiple methods. On one hand, data interfaces are established with major suppliers to obtain their latest quotations in real time. On the other hand, inquiry requests are sent to multiple suppliers through an online inquiry platform, and feedback data is collected. Simultaneously, relevant data is collected from professional market price publishing platforms. For example, through data interfaces with three steel suppliers, real-time quotations for "HRB400EΦ12 steel" were obtained at 4250 yuan / ton, 4280 yuan / ton, and 4230 yuan / ton, respectively; inquiries were sent to ten suppliers on a building materials inquiry platform, and feedback was received from six suppliers, with prices ranging from 4200 to 4300 yuan / ton; the average market price for this steel was obtained from a certain market price publishing platform as 4240 yuan / ton.

[0072] After collecting price data from various sources, the standard mapping module processes it. For the name field, it uses a pre-established standard name library for matching and merging. For example, non-standard names such as "12mm rebar" and "HRB40012 specification steel" are uniformly merged into "HRB400EΦ12 steel".

[0073] When processing the specification / model field, the original data is first split into parameters. Taking an air conditioner specification "Brand X 1.5 HP Level 3 Energy Efficiency Cooling and Heating Air Conditioner" as an example, it is split into parameters such as brand, horsepower, energy efficiency level, and cooling type. Then, it is standardized and recombined according to the standard format "[Brand][Cooling Capacity (kW)]-[Heating Capacity (kW)]-[Energy Efficiency Level]-[Horsepower] Air Conditioner". Assuming that the air conditioner has a cooling capacity of 3.5kW and a heating capacity of 4.0kW, the recombined specification / model is "Brand X 3.5-4.0-Level 3-1.5 HP Air Conditioner".

[0074] Units of measurement are uniformly converted according to pre-set conversion rules. For example, steel weight data in "t" is converted to "kg". If a steel price is in "t" and the price is 4000 yuan / t, the converted price is 4 yuan / kg. Prices including tax and prices excluding tax are converted according to uniform conversion rules. If the standard tax calculation caliber is excluding tax and the tax rate is 13%, a material with a tax-inclusive price of 113 yuan will have a tax-exclusive price of 100 yuan after conversion.

[0075] In terms of regional matching, the data collection area is compared with the applicable area. If the data collection area is City A in a certain province, and the applicable area is City B in the same province, the correlation between market prices and transportation costs between the two cities are analyzed to determine whether the price data is applicable to City B. If applicable, the data is mapped; if not, it is marked as an anomaly. For example, if the market price difference of a certain material in City A and City B is mainly due to transportation costs, after calculating the impact of transportation costs, it is determined that the price data is applicable to City B, and the data is mapped.

[0076] After completing the above processing, a unique standard price entry identifier is generated for each source price data. The identifier contains key information such as material name, specifications, and applicable region, which facilitates subsequent data identification and management. For example, the identifier generated for "HRB400EΦ12 steel, applicable to XX province XX city, standard specifications" is "HRB400E_Φ12_XX province_XX city".

[0077] Step S3: Determine the source credibility of each source price data The credibility assessment module verifies the validity of each source price data after mapping, and determines the source credibility based on factors such as data integrity, time validity, regional matching degree, specification matching degree, and degree of dispersion anomaly.

[0078] During data integrity verification, it is checked whether the price data contains all necessary fields, such as material name, specifications, and price. If a steel price record lacks the specifications field and only records the price as 4200 yuan / ton, the data is considered incomplete, and its source credibility is reduced. For example, if the baseline credibility value for complete data is set to 0.8, the credibility value drops to 0.6 when key fields are missing.

[0079] Time validity verification compares the data's publication or collection time with the specified valid time range. For information price data, if the publication time exceeds the specified validity period (e.g., the specified validity period is one month, but the publication time is two months ago), the time is invalid, reducing credibility. Assuming a normal time validity period has a credibility of 0.8, the credibility drops to 0.5 after exceeding the time limit.

[0080] Regional matching verification confirms whether the data collection region and the applicable region are consistent or related. If the data collection region is a northern province, and the applicable region is a southern province, and there is no evidence to suggest that the prices are universal, a regional mismatch is determined, and the credibility is reduced. For example, the credibility of regional matching is 0.8, but it drops to 0.4 when there is a mismatch.

[0081] Specification matching verification checks whether the specifications of the price data are consistent with the standard specifications. If the specifications of an air conditioner's price data are inconsistent with the standard specifications in key parameters such as horsepower and energy efficiency rating, it is judged as a specification mismatch, reducing the reliability. Assume the reliability is 0.8 when specifications match, and drops to 0.5 when they don't.

[0082] Discreteness anomaly verification analyzes the degree of dispersion of price data compared to the prices of similar materials. If the price of a material deviates significantly or minimally from the prices of similar products in the market without a reasonable explanation, such as a sudden 50% increase in the price of a certain brand of ordinary steel compared to similar products in the market, the price distribution is judged to be abnormal, and the reliability is reduced. For example, the reliability is 0.8 when the price dispersion is normal, but drops to 0.3 when it is abnormal.

[0083] Conversely, when all conditions of the source price data are met—namely, complete fields, valid time, matching region, matching specification, and price dispersion within a set range—its source credibility is increased. For example, for a steel price data point that is complete, valid time, matches region and specification, and has normal price dispersion, its credibility can be increased from the base value of 0.8 to 0.9. This verification method assigns accurate source credibility values ​​to each source price data point, providing a reliable basis for subsequent update triggering decisions.

[0084] Step S4: Determine the update trigger status The trigger determination module uses the current in-stock price of the target price entry as the benchmark price, calculates the deviation of the information price data, historical transaction price data and real-time market price data from the benchmark price, and determines the update trigger status by combining the source credibility of each source price data.

[0085] The following formula is used to calculate the degree of deviation: Degree of deviation = (Current price - Benchmark price) / Benchmark price × 100%. For example, the current in-stock price of "HRB400EΦ12 steel" is 4200 yuan / ton, and a certain information price is 4300 yuan / ton. Then, the degree of deviation of the information price relative to the benchmark price = (4300-4200) / 4200×100%≈2.38%.

[0086] When the deviation of real-time market price data from the benchmark price reaches the first threshold (set at 20%), and at least two types of data—information price data, historical transaction price data, or real-time market price data—support the same direction of deviation, the direct update trigger state is determined. For example, if real-time market price data shows the price has risen to 5040 yuan / ton, with a deviation of (5040-4200) / 4200×100%=20%, while information price data shows the price has risen to 4350 yuan / ton, and historical transaction price data, although showing a smaller increase, also shows an upward trend, then the direct update trigger state is met.

[0087] When the deviation of real-time market price data from the benchmark price reaches the second threshold (set at 30%) but there is insufficient supporting data, it is determined to be in a pending update trigger state. For example, real-time market price data shows that the price has fallen to 2940 yuan / ton, with a deviation of (2940-4200) / 4200×100%=-30%, but the information price data and historical transaction price data do not change significantly and do not form a consistent support, thus entering the pending update trigger state.

[0088] When the deviation of real-time market price data from the benchmark price does not reach the first or second threshold, the original warehouse price is maintained or a limited update is performed. For example, if the real-time market price data fluctuates within 15%, the original warehouse price can be maintained; if the fluctuation is between 15% and 20%, a limited update can be performed to restrict the price adjustment range and avoid excessive price fluctuations. Assuming the real-time market price data rises to 4830 yuan / ton, the deviation is (4830-4200) / 4200×100%=15%. A limited update is performed, restricting the price adjustment range according to certain rules. For example, if the limit for the increase is 10%, then based on the warehouse price of 4200 yuan / ton, the updated price is 4200×(1+10%)=4620 yuan / ton.

[0089] Step S5: Perform merge update When the update trigger state meets the update conditions, the fusion update module uses information price data as the benchmark anchor data, historical transaction price data as the stability correction data, and real-time market price data as the fluctuation correction data to perform a fusion update on the target price item and generate an updated price.

[0090] First, determine the base price range for the target price item based on the information price data. For example, if the information price data shows that the price range for "HRB400EΦ12 steel" is 4000-4500 yuan / ton, this range provides a general range for price updates, indicating that under the current market conditions, the price of this steel should fluctuate within this range.

[0091] Next, the stable correction range within the basic price range is determined based on historical transaction price data. Assuming that historical transaction price data indicates that the reasonable transaction price range for this steel product over a period of time is 4100-4400 yuan / ton, this range is the stable correction range within the basic price range of the information price. It reflects the common price range of this steel product in actual transactions and can suppress excessive price deviations caused by short-term market fluctuations or abnormal quotations.

[0092] Finally, an immediate correction value within the stable correction range is determined based on real-time market price data. If the real-time market price is 4300 yuan / ton and within the stable correction range, then this price can be used as the immediate correction value. When comprehensively determining the updated price, it is ensured that the updated price is within the stable correction range and does not exceed the allowable fluctuation boundary corresponding to the base price range. For example, an updated price of 4300 yuan / ton falls within the stable correction range of 4100-4400 yuan / ton and does not exceed the base price range of 4000-4500 yuan / ton. This integrated update method considers both the overall trend of market prices (the base price range determined by the information price) and actual transaction conditions (the stable correction range determined by historical transaction prices), while also using real-time market prices for fine-tuning (the immediate correction value determined by the real-time market price). This improves the accuracy of the price database's response to market fluctuations and maintains the stability of price updates.

[0093] Step S6: Select update method and write to price database The update execution module selects the corresponding update method from three update methods—direct update, limited update, and pending update—based on the update trigger status determined by the trigger judgment module and the deviation consistency of each source price data, and writes the updated price into the engineering cost price database.

[0094] When the direct update is triggered, the updated price generated by the fusion update module is directly written into the price database. For example, if the direct update is triggered as described above, and the price after fusion update is 4350 yuan / ton, this price is directly updated to the price entry for "HRB400EΦ12 steel", replacing the original price in the database, and the price update operation is completed.

[0095] For price limit updates, when the update trigger condition is met but the fluctuation range of the real-time market price data exceeds a preset limit threshold (set at 25%), the price adjustment range is limited according to the preset limit threshold. For example, if the real-time market price data shows a 30% price increase, but the limit threshold is 25%, the updated price is calculated based on the price in the warehouse, with a 25% increase, and then written into the price database. For instance, if the price in the warehouse is 4200 yuan / ton, the updated price = 4200 × (1 + 25%) = 5250 yuan / ton. This 5250 yuan / ton is written into the price database to prevent a sudden change in the price database due to a single abnormal fluctuation, thus ensuring the relative stability of the price in the database.

[0096] When in the pending update trigger state, the pending update is as follows: when the deviation between the real-time market price data of the target price item and the benchmark price reaches a preset abnormal threshold (set to 30%), and the historical transaction price data and the information price data do not form a consistent support, the overwriting of the original price in the database is suspended, and a pending update mark item is generated. The pending update mark item includes at least an abnormal source identifier (such as an abnormal quotation from a supplier), an abnormal price value (such as a price that is 35% higher than the benchmark price), data items to be supplemented (such as quotations from other suppliers, market supply and demand, etc.), and a manual review entry. For example, the real-time market price data for "HRB400EΦ12 steel" shows a price of 5670 yuan / ton, with a deviation of (5670-4200) / 4200×100%=35%. However, the information price data and historical transaction price data show little change and do not support each other. In this case, a "pending verification" entry is generated, with the source of the anomaly identified as "XX supplier quotation," the abnormal price value as 5670 yuan / ton, and the "data to be supplemented" message indicating "more supplier quotations and market supply and demand information need to be collected." A manual review entry is provided to cost estimators for further review. Through manual review, the reasonableness of the abnormal price is determined, and a decision is made on whether to update the price database, reducing the risk of abnormal price propagation.

[0097] Step S7: Generate price version record The version tracking module synchronously records the update time, data sources involved in the update, source credibility of each source price data, update method used, and price difference before and after the update for each target price entry after it is written, in order to generate the corresponding price version record.

[0098] The price version record further includes the update batch number of the target price item, the effective time of the updated price, the expiration time, the update reason classification, and the price recovery path. For example, for the price item "HRB400EΦ12 steel", the update batch number is 005, the update time is May 10, 2024, and the data sources involved in the update include the information price released by a local cost management agency, the historical transaction prices of three completed projects, and the real-time market prices of five suppliers. The credibility of each source is 0.9 for the information price, 0.85 for the historical transaction price, and 0.8 for the real-time market price. The update method used is direct update, the price difference before and after the update is +350 yuan / ton, the effective time of the updated price is May 10, 2024, the expiration time is not yet set, and the update reason classification is normal market price fluctuation. The price restoration path is implemented by recording the price data of the previous version and update operation records. If a subsequent review determines that the current version is abnormal, the target price item can be rolled back to the previous valid version through the price restoration path. For example, by clicking a specific button or performing a specific operation, the previous version's price data of 4200 yuan / ton can be retrieved to replace the current abnormal price. This provides a clear basis for subsequent cost audits, price reviews, dispute resolution, and historical retrospective processes, improving the management standardization and traceability of the engineering cost price database. By fully recording price version information, cost estimators can clearly understand the ins and outs of price changes. When encountering disputes or needing to retrospectively examine price history, they can quickly and accurately obtain relevant information, ensuring the accuracy and reliability of engineering cost work.

[0099] The above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for constructing an engineering cost price database based on dynamically updated market data, characterized in that, Includes the following steps: S1. Establish a standard price item system and uniformly define the material name, specifications, unit of measurement, applicable region, tax calculation caliber, and time granularity corresponding to the price items in the engineering cost price database; S2. Collect the corresponding information price data, historical transaction price data and real-time market price data for the target price item, and map the information price data, historical transaction price data and real-time market price data to the standard price item system; S3. Perform validity verification on each source price data after mapping to obtain the source credibility of each source price data. The source credibility is determined at least based on data integrity, time validity, regional matching degree, specification matching degree and degree of dispersion anomaly. S4. Using the current in-stock price of the target price item as the benchmark price, calculate the degree of deviation of the information price data, historical transaction price data and real-time market price data from the benchmark price, and determine the update trigger state by combining the source credibility corresponding to each source price data. S5. When the update trigger state meets the update conditions, the target price entry is fused and updated using the information price data as the benchmark anchor data, the historical transaction price data as the stability correction data, and the real-time market price data as the fluctuation correction data to generate an updated price. S6. Based on the update trigger status and the deviation consistency of each source price data, select the corresponding update method from the three update methods of direct update, limited update and pending update and write the updated price into the engineering cost price database. S7. Synchronously record the update time, data sources involved in the update, source credibility of each source price data, update method used, and price difference before and after the update for the target price entry after it is written, so as to generate the corresponding price version record. The fusion update includes: within the benchmark price range determined by the information price data, using historical transaction price data to suppress short-term abnormal fluctuations, and using real-time market price data to correct the current price deviation direction and deviation magnitude within the benchmark price range, thereby improving the accuracy of the price database's response to market fluctuations and maintaining the stability of price updates.

2. The method for constructing an engineering cost price database according to claim 1, characterized in that, The information price data refers to the material or equipment price information released by the cost management agency; the historical transaction price data refers to the transaction price information of the corresponding price item in the completed project, the settled project, or the signed purchase order; and the real-time market price data refers to at least one of the supplier quotation data, inquiry feedback data, or market-released price data.

3. The method for constructing an engineering cost price database according to claim 1, characterized in that, The mapping in step S2 includes: Standardize and merge the name fields of the price data from each source; Perform parameter splitting and standardized recombination on the specification and model field; Standardize the conversion of units of measurement; The prices including tax and prices excluding tax will be converted uniformly. Perform regional matching between the data collection area and the applicable area; After the mapping is completed, a corresponding standard price entry identifier is generated for each source price data.

4. The method for constructing an engineering cost price database according to claim 1, characterized in that, The source credibility in step S3 is determined according to the following rules: When the source price data has any of the following issues: missing fields, time limit exceeded, region mismatch, specification mismatch, or abnormal price distribution, the credibility of the corresponding source will be reduced. When the source price data has complete fields, valid time, matching within the same region, matching within the same specification, and the price dispersion is within the set range, the credibility of the corresponding source is increased.

5. The method for constructing an engineering cost price database according to claim 1, characterized in that, The update trigger state in step S4 is determined based on the following conditions: When the deviation of real-time market price data from the benchmark price reaches the first threshold, and at least two types of data, including information price data, historical transaction price data, or real-time market price data, support the same direction of deviation, it is determined to be in a direct update trigger state. When the deviation of real-time market price data from the benchmark price reaches the second threshold but there is insufficient supporting data, it is determined to be in a state of pending core update. When the deviation of the real-time market price data from the benchmark price does not reach the first threshold and the second threshold, the original in-stock price is maintained or a limited update is performed.

6. The method for constructing an engineering cost price database according to claim 1, characterized in that, The fusion update in step S5 includes: First, determine the basic price range for the target price item based on the aforementioned information price data; Then, based on the historical transaction price data, determine the stable correction range within the basic price range; Finally, the instantaneous correction value within the stable correction range is determined based on the real-time market price data. The updated price is within the stability correction range and does not exceed the allowable fluctuation boundary corresponding to the base price range.

7. The method for constructing an engineering cost price database according to claim 1, characterized in that, The amplitude limit update in step S6 is as follows: When the update trigger state meets the update conditions but the fluctuation of the real-time market price data exceeds the preset limit threshold, the price adjustment range is limited according to the preset limit threshold to avoid a sudden update of the price database caused by a single abnormal fluctuation.

8. The method for constructing an engineering cost price database according to claim 1, characterized in that, The update to be performed is: When the deviation between the real-time market price data of the target price item and the benchmark price reaches a preset abnormal threshold, and the historical transaction price data and the information price data do not form a consistent support, the original in-database price will be suspended from being covered, and a pending verification mark item will be generated. The entries to be verified include at least the anomaly source identifier, the anomaly price value, the data items to be supplemented, and the manual review entry.

9. The method for constructing an engineering cost price database according to claim 1, characterized in that, The price version record in step S7 further includes: The target price entry includes the update batch number, the effective time of the updated price, the expiration time, the update reason classification, and the price recovery path; The price recovery path is used to revert the target price item to the previous valid version when the current version is determined to be abnormal during subsequent review.

10. A system for constructing an engineering cost price database, characterized in that, include: The standard item construction module is used to establish a standard price item system; The multi-source acquisition module is used to collect information price data, historical transaction price data, and real-time market price data; The standard mapping module is used to map price data from various sources to a standard price item system. The credibility assessment module is used to determine the source credibility of each source price data. The trigger determination module is used to determine the update trigger status based on the degree of deviation and consistency of deviation of each source price data relative to the benchmark price. The fusion update module is used to generate updated prices by using information price data as benchmark anchor data, historical transaction price data as stability correction data, and real-time market price data as fluctuation correction data. The update execution module is used to select the corresponding update method from three update methods: direct update, limited update, and pending update, and write it into the engineering cost price database. The version tracking module is used to generate price version records; The updated price generated by the fusion update module is constrained by the benchmark price range corresponding to the information price data and limited by the stable correction range corresponding to the historical transaction price data. The real-time market price data is only used to correct the current price offset direction and offset magnitude within the benchmark price range.

Images