Construction cost data analysis method and system for building construction

Abstract

The invention relates to the technical field of data processing, and provides a building construction-oriented project cost data analysis method and system, and the method comprises the steps: obtaining an abnormal fluctuation index according to the data of each cost type in each related data type; obtaining a type combination according to the abnormal fluctuation index; according to the type combination of the related data type, obtaining a risk index, and screening to obtain a risk type; obtaining a risk degree according to the risk index and the abnormal fluctuation index; obtaining a scoring parameter according to the risk degree; obtaining a tampering risk of each risk type of each engineering project according to the scoring parameter and the risk degree; and obtaining a data exception item according to the tampering risk. By judging the tampering risk of each risk type of each engineering project, the accuracy of the analysis result of the engineering cost data of the building construction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a method and system for parsing project cost data for building construction. Background Art

[0002] Project cost refers to the cost price required to complete a project construction. The estimation of project cost involves multiple aspects of the project, and the parsing of cost is an effective means to judge the specific budget allocation in each aspect of the project. However, some people tamper with the cost amount to meet their own needs to profit from the cost. Their intentions are carried out in multiple project parts, such as modifying the calculation dimensions of the project quantity, exaggerating the material procurement cost, etc., to modify the cost data in a decentralized manner and try to pass the cost audit. Therefore, it is necessary to judge abnormal data based on the project cost data.

[0003] Existing methods mostly use the method of field threshold verification to screen abnormal data in the cost by auditing the cost. The sensitivity of the abnormal feedback of the level where the error caused by the tampering within different levels of the cost data in the cost structure is not high. Therefore, in the process of dividing the current level by the existing technology, the method of establishing a user index through the covariance of neighboring data is used. For the decentralized tampering of data by malicious people in the project cost data, the covariance established for abnormal data is not significantly different from the normal value in terms of the outlier situation, resulting in inaccurate parsing results of the project cost data for building construction. Summary of the Invention

[0004] The present invention provides a method and system for parsing project cost data for building construction to solve the problem of inaccurate parsing results of the existing project cost data for building construction. The specific technical solutions adopted are as follows: The present invention proposes a method for parsing project cost data for building construction, and the method includes the following steps: Obtain a number of relevant data types for each cost type of each engineering project; screen out deviation data from all relevant data types; Among the deviation data, according to the fluctuation situation of the data of each cost type in each relevant data type among engineering projects, obtain the abnormal fluctuation index of each cost type in each relevant data type; according to the abnormal fluctuation index of each cost type in each relevant data type, divide all cost types into several type combinations; According to the type combination where each cost type of each engineering project is located in each relevant data type, obtain the risk index of each cost type of each engineering project. Filter all cost types of all engineering projects according to the risk index to obtain several risk types of each engineering project; According to the risk index of each cost type of each engineering project and the abnormal fluctuation index in each relevant data type, obtain the risk level of each risk type of each engineering project. According to the risk levels of all risk types of each engineering project, obtain the scoring parameters of each engineering project relative to each other engineering project; According to the difference relationship between the scoring parameters of each engineering project relative to each other engineering project and the risk level, obtain the tampering risk of each risk type of each engineering project; According to the tampering risk of each risk type of all engineering projects, obtain the data abnormal projects.

[0005] Further, in the deviation data, according to the fluctuation situation of the data of each cost type in each relevant data type among engineering projects, obtain the abnormal fluctuation index of each cost type in each relevant data type. The specific method included is: In the deviation data, for any one relevant data type of any one cost type of any one engineering project, obtain the mean value of the data of this relevant data type of all cost types in this engineering project, and record the difference between the data of this cost type in this relevant data type and the mean value as the relative data of this cost type in this relevant data type. The th cost type in the th relevant data type, the calculation method of the abnormal fluctuation index is: In the formula, is the abnormal fluctuation index of the th cost type in the th relevant data type; is the number of engineering projects with the th cost type existing in the deviation data; is the th relative data of the th cost type of the th engineering project with the th cost type in the th relevant data type; is the th relative data of the th cost type of all engineering projects with the

[0006] Further, the method of dividing all cost types into several type combinations according to the abnormal fluctuation index of each cost type in each relevant data type is as follows: Step 1: In the deviation data, for any relevant data type, sort the abnormal fluctuation indices of all cost types in this relevant data type in descending order to obtain a fluctuation index sequence. Step 2: Denote any abnormal fluctuation index in any fluctuation index sequence as the target abnormal fluctuation index; denote the difference between the target abnormal fluctuation index and its next abnormal fluctuation index as the change degree of the target abnormal fluctuation index; denote the abnormal fluctuation index with the largest change degree in this fluctuation index sequence as the segmentation fluctuation index of this fluctuation index sequence. Step 3: For any fluctuation index sequence, break the fluctuation index sequence after its segmentation fluctuation index to obtain two updated fluctuation index sequences. Step 4: Use the updated fluctuation index sequence as the fluctuation index sequence and repeat Step 2 and Step 3 until the number of abnormal fluctuation indices in an updated fluctuation index sequence is less than the average of the number of cost types in all engineering projects, then stop repeating. Step 5: The cost types corresponding to all abnormal fluctuation indices in any updated fluctuation index sequence form a type combination.

[0007] Further, the method of obtaining the risk index of each cost type of each engineering project according to the type combination where each cost type of each engineering project is located in each relevant data type, and screening several risk types of each engineering project according to the risk index is as follows: In the deviation data, for any relevant data type, obtain the type combination with the largest number of all cost types of all engineering projects belonging to it under this relevant data type, and denote it as the normal combination of this relevant data type; for any cost type of any engineering project, if this cost type does not belong to the normal combination under this relevant data type, then mark this cost type as an abnormal type under this relevant data type; denote the ratio of the number of times this cost type is marked as an abnormal type under all relevant data types to the number of all relevant data types as the risk index of this cost type; if the risk index of this cost type is greater than the preset risk threshold, then mark this cost type as a risk type of this engineering project.

[0008] Further, the method of obtaining the risk degree of each risk type of each engineering project according to the risk index of each cost type of each engineering project and the abnormal fluctuation index in each relevant data type is as follows: In the deviation data, for any cost type of any engineering project, the sum value of the abnormal fluctuation indexes of this cost type in all relevant data types is recorded as the overall abnormal degree of this cost type; In the deviation data, for any risk type of any engineering project, the average value of the overall abnormal degrees of all cost types except this risk type within this engineering project is denoted as the standard abnormal degree of this risk type; The ratio of the overall abnormal degree of this risk type to the standard abnormal degree is denoted as the abnormal index of this risk type; The product of the risk index and the abnormal index of this risk type is denoted as the risk degree of this risk type.

[0009] Furthermore, the specific method for obtaining the scoring parameter of each engineering project relative to each other engineering project according to the risk degrees of all risk types of each engineering project includes: In the deviation data, for any engineering project, the column vector composed of the risk degrees of all risk types of this engineering project is recorded as the attribute vector of this engineering project; The th engineering project relative to the th engineering project, the calculation method of the scoring parameter is: In the formula, is the scoring parameter of the th engineering project relative to the th engineering project; is the standard deviation of the risk degrees of all risk types of the th engineering project and the th engineering project; is the cosine similarity of the attribute vectors of the th engineering project and the th engineering project; is a hyperparameter.

[0010] Furthermore, the specific method for obtaining the tampering risk of each risk type of each engineering project according to the scoring parameter of each engineering project relative to each other engineering project and the difference relationship of the risk degree includes: In the deviation data, the calculation method of the tampering index of the th engineering project relative to the th engineering project for the th risk type is: In the formula, is the th engineering project relative to the Tampering index of the th risk type for a construction project; is the scoring parameter of the th construction project relative to the th construction project; is the th risk level of the th risk type for a construction project; is the th risk level of the th risk type for a construction project; is the absolute value function; According to the tampering index of each risk type of each construction project relative to all other construction projects, the tampering risk of each risk type of each construction project is obtained.

[0011] Furthermore, the method for obtaining the tampering risk of each risk type of each construction project according to the tampering index of each risk type of each construction project relative to all other construction projects includes the following specific steps: In the deviation data, for any construction project and any risk type, the linearly normalized result of the mean of the tampering index of this risk type of this construction project relative to all construction projects is denoted as the tampering risk of this risk type of this construction project.

[0012] Furthermore, the method for obtaining data abnormal projects according to the tampering risk of each risk type of all construction projects includes the following specific steps: Mark the construction project where the risk type with a tampering risk greater than the preset tampering threshold is located as a data abnormal project.

[0013] The present invention also proposes a project cost data analysis system for building construction, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the above method are implemented.

[0014] The beneficial effects of the present invention are as follows: When analyzing the project cost data of building construction, since there are fluctuations in individual cost types among engineering projects, the present invention obtains the abnormal fluctuation index of each cost type in each relevant data type based on the fluctuation of each cost type in the data of each relevant data type among engineering projects, and obtains several type combinations through the abnormal fluctuation index to distinguish the cost types that may be tampered with from other cost types; Since most of the data of cost types are normal data when analyzing the project cost data of building construction, for the same relevant data type, most of the type combinations where cost types are located are the same. The present invention obtains the risk index of each cost type of each engineering project based on the type combination where each cost type of each engineering project is located in each relevant data type, and screens all cost types of all engineering projects according to the risk index to obtain several risk types of each engineering project, and then combines the abnormal fluctuation index to obtain the risk level of each risk type of each engineering project, and judges the possibility of each risk type of each engineering project being tampered with; Since the scope of data classification modified by the tamperer is restricted by their permissions, there are differences in the risk levels of the data tampered with by the tamperer compared to other data. The present invention obtains the tampering risk of each risk type of each engineering project based on the difference relationship between the scoring parameters of each engineering project relative to each other engineering project and the risk level, and then obtains the data abnormal projects. Thus, the present invention obtains the data abnormal projects through the tampering risk of each risk type of each engineering project, improving the accuracy of the analysis result of the project cost data of building construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings without creative efforts based on these drawings.

[0016] Figure 1 It is a schematic flowchart of a method for parsing project cost data for building construction provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0018] Please refer toFigure 1 , which shows a flowchart of a method for parsing project cost data for building construction provided by an embodiment of the present invention. The method includes the following steps: Step S001: Obtain a number of relevant data types for each cost type of each engineering project; screen out deviation data from all relevant data types.

[0019] The purpose of this embodiment is to analyze the project cost data, judge the tampering risk of the project cost data, and thus obtain abnormal data. Therefore, it is first necessary to obtain the project cost data.

[0020] Specifically, for a number of engineering projects to be analyzed, obtain a number of relevant data types for each cost type of each engineering project; among them, the cost types include but are not limited to civil engineering costs, installation costs, material costs, and labor costs; among them, the relevant data types include but are not limited to the quantity, unit price, and total price of each cost type.

[0021] It should be noted that generally, the project cost data conforms to the mathematical model of unit price multiplied by quantity equals total price. However, when some prices fluctuate due to market fluctuations, construction losses, etc., it may cause the project cost data to deviate from this mathematical model. This kind of deviation phenomenon is a normal phenomenon. However, generally, the deviation data may contain project cost data tampered by some personnel. Therefore, it is first necessary to obtain the project cost data that deviates from this mathematical model, and then analyze these project cost data.

[0022] Specifically, use the mathematical model of unit price multiplied by quantity equals total price to traverse and check all relevant data types of each cost type of each engineering project, and screen out the cost types that do not conform to this mathematical model among all cost types of each engineering project as deviation data.

[0023] Step S002: In the deviation data, obtain the abnormal fluctuation index of each cost type in each relevant data type according to the fluctuation situation of the data of each cost type in each relevant data type among engineering projects; divide all cost types into several type combinations according to the abnormal fluctuation index of each cost type in each relevant data type.

[0024] It should be noted that since the difference levels of the same relevant data type of the same cost type of different engineering projects are similar, when some personnel tamper with the cost, in order to reflect the difference between normal data and tampered data, it is first necessary to clean the relevant data types.

[0025] Specifically, in the deviation data, for any relevant data type of any cost type of any engineering project, the mean value of the data of that relevant data type for all cost types in the engineering project is obtained, and the difference between the data of that cost type in that relevant data type and the mean value is denoted as the relative data of that cost type in that relevant data type.

[0026] It should be noted that the larger the relative data of a certain cost type in a certain relevant data type, the greater the deviation of that cost type from the average level in that relevant data type.

[0027] It should be noted that since there are fluctuations in individual cost types among engineering projects, and the tampered abnormal cost types are mainly manifested as a greater fluctuation deviation of the cost types of this engineering project compared to the same cost types of other engineering projects, that is, the relative data is larger. Therefore, the cost types are distinguished based on the abnormal fluctuations of the relative data under each cost type.

[0028] Specifically, in the deviation data, the th cost type in the th relevant data type has the following calculation method for the abnormal fluctuation index: In the formula, is the abnormal fluctuation index of the th cost type in the th relevant data type; is the number of engineering projects with the th cost type in the deviation data; is the relative data of the th cost type in the th engineering project with the th cost type in the th relevant data type; is the standard deviation of the relative data of the th cost type in the th cost type in all engineering projects with the th relevant data type in the deviation data.

[0029] It should be noted that the larger the abnormal fluctuation index of the th cost type in the th relevant data type, the greater the abnormal fluctuation in the fluctuations among engineering projects of the th cost type in the th relevant data type, that is, the th cost type may be abnormal in some engineering projects compared to other engineering projects.

[0030] It should be noted that after obtaining the abnormal fluctuation index of each cost type in each relevant data type, when a certain person tampers with the cost, the person will tamper with the indicators in the engineering field under his responsibility, so the data access rights he has are concentrated in the attributes under his jurisdiction. For example, the installation person in charge of Project A can only modify the installation quantity, installation unit price, etc. under his jurisdiction, and cannot tamper with the other items such as civil engineering quantity and civil engineering unit price. Therefore, the tampered data is mainly concentrated in certain cost types. Therefore, it is necessary to group the cost types according to the abnormal fluctuation index to obtain a type combination, and then judge the possibility of being tampered with for each type combination.

[0031] Specifically, the specific method for obtaining the type combination is as follows: Step 1. In the deviation data, for any one of the relevant data types, sort the abnormal fluctuation indexes of all cost types in this relevant data type in descending order to obtain a fluctuation index sequence. Step 2. Denote any abnormal fluctuation index in any one of the fluctuation index sequences as the target abnormal fluctuation index; denote the difference between the target abnormal fluctuation index and its next abnormal fluctuation index as the change degree of the target abnormal fluctuation index; denote the abnormal fluctuation index with the largest change degree in this fluctuation index sequence as the segmentation fluctuation index of this fluctuation index sequence. Step 3. For any one of the fluctuation index sequences, disconnect the fluctuation index sequence after its segmentation fluctuation index to obtain two updated fluctuation index sequences. Step 4. Use the updated fluctuation index sequence as the fluctuation index sequence and repeat Step 2 and Step 3 until the number of abnormal fluctuation indexes in an updated fluctuation index sequence is less than the average value of the number of cost types in all engineering projects, and then stop repeating. Step 5. The cost types corresponding to all abnormal fluctuation indexes in any one of the updated fluctuation index sequences form a type combination.

[0032] Step S003. According to the type combination where each cost type of each engineering project is located in each relevant data type, obtain the risk index of each cost type of each engineering project. Screen all cost types of all engineering projects according to the risk index to obtain several risk types of each engineering project; according to the risk index of each cost type of each engineering project and the abnormal fluctuation index in each relevant data type, obtain the risk degree of each risk type of each engineering project.

[0033] It should be noted that for any relevant data type, if all expense types of all engineering projects are within this relevant data type, the type combination in which most expense types are located is the type combination of normal data under this relevant data type. If a certain expense type of a certain engineering project does not conform to the type combination of normal data under most relevant data types, then this expense type of this engineering project may be abnormal.

[0034] Specifically, in the deviation data, for any relevant data type, obtain the type combination with the largest number of all expense types of all engineering projects belonging to under this relevant data type, and record it as the normal combination of this relevant data type; for any expense type of any engineering project, if this expense type does not belong to the normal combination under this relevant data type, then mark this expense type as an abnormal type under this relevant data type; record the ratio of the number of times this expense type is marked as an abnormal type under all relevant data types to the number of all relevant data types as the risk index of this expense type; if the risk index of this expense type is greater than the preset risk threshold, then record this expense type as the risk type of this engineering project; where the preset risk threshold is one-half, and this embodiment is described by taking this as an example.

[0035] Furthermore, in the deviation data, for any expense type of any engineering project, record the sum value of the abnormal fluctuation indices of this expense type under all relevant data types as the overall abnormal degree of this expense type; In the deviation data, for any risk type of any engineering project, record the average value of the overall abnormal degrees of all expense types other than this risk type within this engineering project as the standard abnormal degree of this risk type; Record the ratio of the overall abnormal degree of this risk type to the standard abnormal degree as the abnormal index of this risk type; Record the product of the risk index and the abnormal index of this risk type as the risk degree of this risk type.

[0036] It should be noted that the higher the abnormal index of a risk type, the greater the overall abnormal degree of this risk type compared to other expense types within its engineering project; the greater the risk degree of a risk type, the more likely the data of the relevant data type of this risk type has been tampered with artificially.

[0037] Step S004: Obtain the scoring parameters of each engineering project relative to each other engineering project according to the risk degrees of all risk types of each engineering project; obtain the tampering risk of each risk type of each engineering project according to the difference relationship between the scoring parameters of each engineering project relative to each other engineering project and the risk degree; obtain the data abnormal projects according to the tampering risks of each risk type of all engineering projects.

[0038] It should be noted that the calculation of the safety of engineering projects often first conducts an inventory and liquidation of data within the engineering project. Due to the permission restrictions of the tamperer on the scope of data classification to be modified, there are differences in the risk levels of the data tampered by the tamperer compared to other data. Therefore, it is necessary to conduct a comparative analysis of the risk levels between different risk types of different engineering projects.

[0039] Specifically, among the deviated data, for any engineering project, the column vector composed of the risk levels of all risk types of this engineering project is denoted as the attribute vector of this engineering project; among them, the positions of the risk levels of the same risk type of different engineering projects in the column vector should be the same. If an engineering project does not have a risk type owned by other engineering projects, it should be filled with 0 at the corresponding position; The th engineering project relative to the th engineering project, the calculation method of the scoring parameter is as follows: In the formula, is the scoring parameter of the th engineering project relative to the th engineering project; is the standard deviation of the risk levels of all risk types of the th engineering project and the th engineering project; is the cosine similarity of the attribute vectors of the th engineering project and the th engineering project; is a hyperparameter to prevent the denominator from being 0, , and this embodiment is described by taking this as an example.

[0040] It should be noted that the larger the scoring parameter of the th engineering project relative to the th engineering project, the more similar the overall distribution of the engineering cost data of the th engineering project and the th engineering project, which is beneficial to identifying the risk types that have been tampered with as a whole.

[0041] It should be noted that when comparing the risk levels of the th risk type of the th engineering project and the th engineering project, when the scoring parameter of the th engineering project relative to the th engineering project is larger, it indicates that the th engineering project and the The comparison of multiple engineering projects is beneficial for identifying the risk types of overall tampering. That is, the comparison result between the th engineering project and the th engineering project has a higher confidence level. At the same time, since the tampered data is a minority in the engineering project data, that is, the risk levels of the risk types of most engineering projects are relatively low. When the risk level difference of the th engineering project is relatively large compared to the th risk type of most engineering projects, it indicates that the possibility of tampered data in the th risk type of the th engineering project is greater.

[0042] Furthermore, in the deviated data, the calculation method of the tampering index of the th engineering project with respect to the th engineering project for the th risk type is as follows: In the formula, is the tampering index of the th engineering project with respect to the th engineering project for the th risk type; is the scoring parameter of the th engineering project with respect to the th engineering project; is the risk level of the th engineering project for the th risk type; is the risk level of the th engineering project for the th risk type; is the absolute value function.

[0043] Furthermore, in the deviated data, for any engineering project and any risk type, the linearly normalized result of the mean of the tampering indexes of the engineering project with respect to the risk type of all engineering projects is denoted as the tampering risk of the engineering project for the risk type. Among them, the normalization object is the mean of the tampering indexes of each engineering project with respect to the risk type of all engineering projects; The engineering project where the risk type with a tampering risk greater than the preset tampering threshold is marked as a data abnormal project. Among them, the preset tampering threshold is 0.6, and this embodiment is described by taking this as an example; All relevant data types of all cost types of the data abnormal project are manually calculated by professional financial personnel to verify the authenticity of the project cost data.

[0044] Another embodiment of the present invention provides a project cost data analysis system for building construction. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the above method steps S001 to S004 are implemented.

[0045] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for analyzing project cost data for building construction, characterized in that, The method includes the following steps: Obtain a number of relevant data types for each cost type of each engineering project; screen out deviation data from all relevant data types; Among the deviation data, according to the fluctuation of the data of each cost type in each relevant data type among engineering projects, obtain the abnormal fluctuation index of each cost type in each relevant data type; according to the abnormal fluctuation index of each cost type in each relevant data type, divide all cost types into several type combinations; According to the type combination where each cost type of each engineering project is located in each relevant data type, obtain the risk index of each cost type of each engineering project, and screen out several risk types of each engineering project according to the risk index; according to the risk index of each cost type of each engineering project and the abnormal fluctuation index in each relevant data type, obtain the risk level of each risk type of each engineering project; According to the risk levels of all risk types of each engineering project, obtain the scoring parameters of each engineering project relative to each other engineering project; according to the difference relationship between the scoring parameters of each engineering project relative to each other engineering project and the risk level, obtain the tampering risk of each risk type of each engineering project; according to the tampering risk of each risk type of all engineering projects, obtain the data abnormal projects.

2. The engineering cost data analysis method for building construction according to claim 1, wherein The specific method included in obtaining the abnormal fluctuation index of each cost type in each relevant data type according to the fluctuation of the data of each cost type in each relevant data type among engineering projects in the deviation data is as follows: In the deviation data, for any one relevant data type of any one cost type of any one engineering project, obtain the mean value of the data of this relevant data type of all cost types in this engineering project, and record the difference between the data of this cost type in this relevant data type and the mean value as the relative data of this cost type in this relevant data type; The th expense type calculates the abnormal fluctuation index of the th related data type as follows: Wherein, is the abnormal fluctuation index of the th cost type in the th related data type; is the number of engineering projects with the th cost type in the deviation data; is the relative data of the th cost type in the th engineering project with the th cost type in the th related data type; is the standard deviation of the relative data of the th cost type in all engineering projects with the th cost type in the th related data type in the deviation data.

3. The engineering cost data parsing method for building construction according to claim 1, characterized in that The specific method included in dividing all cost types into several type combinations according to the abnormal fluctuation index of each cost type in each relevant data type is as follows: Step 1. In the deviation data, for any one relevant data type, arrange the abnormal fluctuation indexes of all cost types in this relevant data type in descending order to obtain a fluctuation index sequence; Step 2. Denote any one abnormal fluctuation index in any one fluctuation index sequence as the target abnormal fluctuation index; record the difference between the target abnormal fluctuation index and its next abnormal fluctuation index as the change degree of the target abnormal fluctuation index; Denote the abnormal fluctuation index with the largest change degree in this fluctuation index sequence as the segmentation fluctuation index of this fluctuation index sequence; Step 3. For any one fluctuation index sequence, disconnect the fluctuation index sequence after its segmentation fluctuation index to obtain two updated fluctuation index sequences; Step 4: Use the updated volatility index sequence as the volatility index sequence and repeat Steps 2 and 3 until the number of abnormal volatility indices in an updated volatility index sequence is less than the mean of the number of cost types in all engineering projects, then stop repeating; Step 5: The cost types corresponding to all abnormal volatility indices in any one updated volatility index sequence form a type combination.

4. The engineering cost data analysis method for building construction according to claim 1, wherein The method for obtaining the risk index of each cost type of each engineering project based on the type combination where each cost type of each engineering project is located in each relevant data type, and screening several risk types of each engineering project according to the risk index, specifically includes: In the deviation data, for any one relevant data type, obtain the type combination with the largest number of all cost types of all engineering projects belonging to it under this relevant data type, denoted as the normal combination of this relevant data type; for any one cost type of any one engineering project, if this cost type does not belong to the normal combination under this relevant data type, then mark this cost type as an abnormal type under this relevant data type; The ratio of the number of times this cost type is marked as an abnormal type under all relevant data types to the number of all relevant data types is denoted as the risk index of this cost type; If the risk index of this cost type is greater than the preset risk threshold, then mark this cost type as a risk type of this engineering project.

5. The engineering cost data analysis method for building construction according to claim 1, characterized in that The method for obtaining the risk level of each risk type of each engineering project based on the risk index of each cost type of each engineering project and the abnormal volatility index in each relevant data type, specifically includes: In the deviation data, for any one cost type of any one engineering project, denote the sum value of the abnormal volatility indices of this cost type in all relevant data types as the overall abnormal degree of this cost type; In the deviation data, for any one risk type of any one engineering project, denote the mean value of the overall abnormal degrees of all cost types other than this risk type in this engineering project as the standard abnormal degree of this risk type; The ratio of the overall abnormal degree of this risk type to the standard abnormal degree is denoted as the abnormal index of this risk type; The product of the risk index and the abnormal index of this risk type is denoted as the risk level of this risk type.

6. The method for parsing engineering cost data for building construction according to claim 1, wherein The method for obtaining the scoring parameter of each engineering project relative to each other engineering project based on the risk levels of all risk types of each engineering project, specifically includes: In the deviation data, for any one engineering project, denote the column vector formed by the risk levels of all risk types of this engineering project as the attribute vector of this engineering project; The calculation method of the scoring parameters of the th engineering project relative to the th engineering project is as follows: Wherein, is the scoring parameter of the th engineering project relative to the th engineering project; is the standard deviation of the risk levels of all risk types of the th engineering project and the th engineering project; is the cosine similarity of the attribute vectors of the th engineering project and the th engineering project; is a hyperparameter.

7. The engineering cost data analysis method for building construction according to claim 1, characterized in that The method for obtaining the tampering risk of each risk type of each engineering project based on the scoring parameter of each engineering project relative to each other engineering project and the difference relationship of the risk levels, specifically includes: In the deviation data, for the th engineering project relative to the th engineering project, the calculation method of the tampering index for the th risk type is as follows: Wherein, is the tampering index of the th engineering project with respect to the th engineering project for the th risk type; is the scoring parameter of the th engineering project with respect to the th engineering project; is the risk level of the th engineering project for the th risk type; is the risk level of the th engineering project for the th risk type; is the absolute value function; Obtain the tampering risk of each risk type of each engineering project according to the tampering index of each risk type of each engineering project relative to all other engineering projects.

8. The engineering cost data analysis method for building construction according to claim 7, characterized in that Obtaining the tampering risk of each risk type of each engineering project according to the tampering index of each risk type of each engineering project relative to all other engineering projects, including the specific method as follows: In the deviation data, for any one engineering project and any one risk type, the linearly normalized result of the mean value of the tampering index of this engineering project relative to the risk type of all engineering projects is recorded as the tampering risk of this engineering project for this risk type.

9. The method for parsing project cost data for building construction according to claim 1, characterized in that, Obtaining the data abnormal projects according to the tampering risk of each risk type of all engineering projects, including the specific method as follows: Mark the engineering projects where the risk types with tampering risks greater than the preset tampering threshold are located as data abnormal projects.

10. A project cost data analysis system for building construction, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the engineering cost data parsing method for building construction as described in any one of claims 1-9.

Images