A construction project compliance review method and system based on AI decision tracing

Abstract

The present invention discloses a construction project compliance review method and system based on AI decision tracing, which relates to the field of data processing technology. The method comprises determining standard data, annotating architectural drawing samples according to the standard data, storing the standard data and the annotated architectural drawing samples as an annotation rule library, setting a matching ring in the annotation rule library, dividing the standard data and the annotated architectural drawing samples into multiple data segments, storing the data segments in a matching sub-group, destroying the data segments in the matching sub-group when an anomaly occurs in the matching sub-group, and recovering the data segments of the abnormal matching sub-group based on a supplementary interval. The present invention can ensure the security and completeness of data through the matching sub-group, and when an anomaly occurs in the data, can help the abnormal matching sub-group to recover quickly through adjacent matching sub-groups, has a fast response and can stop the loss in time to avoid affecting more data, and has good data storage stability.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a construction project compliance review method and system based on AI decision tracing. Background Art

[0002] In the construction industry, traditional compliance review of construction drawings relies primarily on manual review, but this approach has numerous drawbacks. On the one hand, the manual review process is cumbersome and inefficient, requiring significant manpower and time, making it difficult to adapt to the fast-paced, high-efficiency demands of modern construction projects. On the other hand, manual review is susceptible to subjective interference, and different reviewers have varying understandings and implementations of standards. This directly leads to omissions or errors in the review results, making it impossible to effectively meet the increasingly complex design requirements and increasingly stringent standards of today's construction projects.

[0003] With the rapid development of artificial intelligence (AI) technology, its application in project compliance review is gradually gaining momentum. Training AI models to automatically identify violations in construction drawings has become a new approach to improving review efficiency and accuracy. However, traditional methods for construction project compliance review lack data security and integrity assurance mechanisms. Once data is tampered with, the lack of effective monitoring and early warning mechanisms makes it difficult for the system to quickly detect and respond, potentially affecting more related data and failing to effectively protect data. Summary of the Invention

[0004] The purpose of the present invention is to provide a construction project compliance review method and system based on AI decision tracing to address the shortcomings of the background technology.

[0005] In order to achieve the above objectives, the present invention provides the following technical solution: a construction project compliance review method based on AI decision tracing, comprising the following steps:

[0006] Determine the specification data, annotate the architectural drawing samples according to the specification data, and store the specification data and the annotated architectural drawing samples as an annotation rule library;

[0007] A matching ring is set in the annotation rule library, wherein the matching ring is composed of two sub-rings, and both sub-rings are composed of multiple matching sub-rings, and an open interval and a supplementary interval are set for each matching sub-ring;

[0008] Divide the specification data and the annotated architectural drawing samples into multiple data segments, store the data segments in the matching sub-groups, and when an anomaly occurs in the matching sub-groups, destroy the data segments in the matching sub-groups, and recover the data segments of the abnormal matching sub-groups based on the supplementary intervals;

[0009] Based on the open interval, the deep learning model is trained with the specification data of the annotation rule library and the annotated architectural drawing samples to obtain a trained violation recognition model, the architectural drawing data is determined, and the architectural drawing data is input into the trained violation recognition model to output violation information, wherein the violation information includes the illegal components and the corresponding violation degree;

[0010] Bind the violation information with the corresponding building drawing data as an analysis log, and store the analysis log.

[0011] In a preferred embodiment, the step of storing the specification data and the annotated architectural drawing samples as an annotation rule library includes:

[0012] A plurality of architectural drawings are obtained as architectural drawing samples, and component information in the architectural drawing samples is read, wherein the component information includes geometric properties and annotation information of the components.

[0013] Mark component information in architectural drawing samples according to specification data;

[0014] The specification data and the annotated architectural drawing samples are stored to obtain an annotation rule library.

[0015] In a preferred embodiment, the step of setting a matching ring in the annotation rule library includes:

[0016] Configure two sub-rings in the annotation rule base. Both sub-rings are composed of multiple identical matching sub-rings connected in a chain.

[0017] Divide the matching sub-interval into multiple parallel sub-intervals, connect the multiple sub-intervals to each other, set corresponding matching codes for the multiple sub-intervals, and bind the matching codes to the corresponding sub-intervals;

[0018] Connect the two sub-chains to the matching sub-chains one-to-one, and use the matching codes corresponding to the matching sub-chains as the connection pairing conditions;

[0019] The subinterval matching the preset data in the middle part of the subinterval is selected as the open interval, and the subintervals on both sides of the open interval are respectively made to correspond to the supplementary intervals of the adjacent subintervals.

[0020] In a preferred embodiment, the step of recovering data segments of abnormal matching sub-groups based on the supplementary interval includes:

[0021] Divide the specification data and the annotated architectural drawing samples into a number of data segments equal to the number of matching sub-groups;

[0022] A one-to-one correspondence is established between the data segment and the matching sub-item, and the data segment is randomly divided into different data volumes according to the number of open intervals in the corresponding matching sub-item to obtain multiple sub-data blocks, and the multiple sub-data blocks are stored in the open intervals of the matching sub-item;

[0023] The data segment is copied simultaneously and divided into two parts with different data amounts to obtain two supplementary data, and the supplementary data are respectively stored in the supplementary intervals of adjacent matching sub-parts;

[0024] Sort the matching codes of the matchers from large to small according to the amount of data in the corresponding subintervals, and use the matching codes of the two subchains after sorting them from large to small as the pairing condition;

[0025] The matching sub-items with changed matching codes are regarded as abnormal matching sub-items, the data segments in the open intervals of the abnormal matching sub-items are destroyed, and the data segments of the abnormal matching sub-items are recovered based on the supplementary intervals in the adjacent matching sub-items.

[0026] In a preferred embodiment, the step of storing the supplementary data in the supplementary intervals of adjacent matching sub-elements respectively includes:

[0027] The supplementary data are divided into different data amounts according to the number of supplementary intervals in adjacent matching sub-groups;

[0028] The divided supplementary data are stored in corresponding supplementary intervals respectively.

[0029] In a preferred embodiment, the step of inputting the architectural drawing data into a trained violation recognition model to output violation information includes:

[0030] Obtain data segments through the open interval, and obtain the standard data of the annotation rule library and the architectural drawing samples after annotation through the data segments in the entire matching loop;

[0031] The violation recognition model is obtained by training the deep learning model through the normative data of the annotation rule library and the annotated architectural drawing samples;

[0032] The architectural drawing data that needs to be analyzed is input into the violation identification model to output the violation components and the corresponding violation degree.

[0033] In a preferred embodiment, the step of binding the violation information with the corresponding building drawing data as an analysis log and storing the analysis log includes:

[0034] Obtain analysis time for architectural drawing data;

[0035] The violation information and the corresponding building drawing data are bound and stored as analysis logs according to the analysis time.

[0036] The present invention also provides a construction project compliance review system based on AI decision tracing, comprising:

[0037] A storage module is used to determine specification data, annotate architectural drawing samples according to the specification data, and store the specification data and the annotated architectural drawing samples as an annotation rule library;

[0038] A setting module is connected to the storage module and is used to set a matching ring in the annotation rule library, wherein the matching ring is composed of two sub-rings, and both sub-rings are composed of multiple matching sub-rings, and an open interval and a supplementary interval are set for each matching sub-ring;

[0039] A data processing module, connected to the setting module, is used to divide the specification data and the annotated architectural drawing samples into multiple data segments, store the data segments in the matching sub-modules, and when an abnormality occurs in the matching sub-modules, destroy the data segments in the matching sub-modules and recover the data segments of the abnormal matching sub-modules based on the supplementary interval;

[0040] An analysis module, connected to the data processing module, is configured to train a deep learning model based on the open interval using the specification data of the annotation rule library and the annotated architectural drawing samples to obtain a trained violation recognition model, determine architectural drawing data, input the architectural drawing data into the trained violation recognition model, and output violation information, wherein the violation information includes the illegal component and the corresponding violation degree;

[0041] The recording module is connected to the analysis module and is used to bind the violation information with the corresponding building drawing data as an analysis log and store the analysis log.

[0042] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0043] The present invention can ensure the security and completeness of data through matching sub-sub-subs. When data anomalies occur, adjacent matching sub-sub-subs can help the abnormal matching sub-subs to recover quickly. It has a fast response and can stop the loss in time to avoid affecting more data, and has good data storage stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0045] Figure 1 Flow chart of the method of the present invention.

[0046] Figure 2 This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0048] Example 1, please refer to Figure 1 As shown, the construction project compliance review method based on AI decision tracing described in this embodiment includes the following steps:

[0049] S1. Determine specification data, annotate architectural drawing samples according to the specification data, and store the specification data and the annotated architectural drawing samples as an annotation rule library;

[0050] S2. Set a matching ring in the annotation rule library, where the matching ring consists of two sub-rings, both of which are composed of multiple matching sub-rings, and set open intervals and supplementary intervals for the corresponding matching sub-rings;

[0051] S3. Divide the specification data and the annotated architectural drawing samples into multiple data segments, store the data segments in matching sub-groups, and when an abnormality occurs in a matching sub-group, destroy the data segments in the matching sub-group and recover the data segments of the abnormal matching sub-group based on the supplementary interval;

[0052] S4. Based on the open interval, the deep learning model is trained using the standard data of the annotation rule library and the annotated architectural drawing samples to obtain a trained violation recognition model. The architectural drawing data is determined, and the architectural drawing data is input into the trained violation recognition model to output violation information, wherein the violation information includes the illegal component and the corresponding violation degree.

[0053] S5. Bind the violation information with the corresponding building drawing data as an analysis log, and store the analysis log.

[0054] In one embodiment, the step S1 of storing the specification data and the annotated architectural drawing samples as an annotation rule library includes:

[0055] S11. Obtain a plurality of architectural drawings as architectural drawing samples, and read component information in the architectural drawing samples, wherein the component information includes geometric properties and annotation information of the components.

[0056] S12. Mark the component information in the architectural drawing sample according to the specification data;

[0057] S13. Store the specification data and the annotated architectural drawing samples to obtain an annotation rule library.

[0058] As described in steps S11-S13 above, multiple architectural drawings (CAD drawings) are obtained as architectural drawing samples, and component information in the architectural drawing samples is read. The drawing files are read through the CAD software interface, and the layer information is used to distinguish different components (such as beams, columns, and walls). The geometric properties (size, position) and annotation information (material grade, load parameters) of each component are extracted. The component information in the architectural drawing samples is then annotated according to the specification data. The annotation here indicates whether the component information meets the specification data. For example, if the clear width of a stairwell is marked as ≥1.1m, if the actual value is 1.05m, it is marked as a violation. The annotation process distinguishes between serious violations (such as structural safety) and minor violations (such as non-critical dimensional deviations). The specification data here refers to the architectural design specification version. The specification data and the annotated architectural drawing samples are stored to obtain a annotation rule library, which has good data preparation.

[0059] In one embodiment, the step S2 of setting a matching ring in the annotation rule library includes:

[0060] S21. Configure two sub-rings in the annotation rule library. Both sub-rings are composed of multiple identical matching sub-rings connected in a chain-like manner.

[0061] S22: Divide the matching sub-interval into multiple parallel sub-intervals, connect the multiple sub-intervals to each other, set corresponding matching codes for each of the multiple sub-intervals, and bind the matching codes to the corresponding sub-intervals;

[0062] S23, connecting the two sub-chains to the matching sub-chains one-to-one, using the matching codes corresponding to the matching sub-chains as the connection pairing conditions;

[0063] S24 , selecting a subinterval that matches the preset data in the middle part of the subinterval as an open interval, and arranging the subintervals on both sides of the open interval to correspond to the supplementary intervals of the adjacent subintervals.

[0064] As described in the above steps S21-S23, the matching ring is set in the annotation rule library, and the two sub-rings are composed of the same multiple matching sub-rings connected in a chain. The matching sub-ring is a virtual machine, which is equivalent to a hand-in-hand connection between multiple virtual machines. Since the matching sub-ring is a virtual machine for storage and connection, the matching sub-ring can be divided into multiple parallel sub-intervals. The multiple sub-intervals are connected to each other, and corresponding matching codes are set for the corresponding sub-intervals. The matching codes are bound to the corresponding sub-intervals. The divided sub-intervals are used for the planning of subsequent open intervals and supplementary intervals. At the same time, the corresponding matching codes can also be bound and set. One sub-interval corresponds to one matching code, and the code corresponding to one matching sub-interval is multiple. The matching codes are sorted according to the amount of data in the sub-intervals. The matching codes are sorted when the specification data and the annotated architectural drawing samples are stored. The current matching ring is in the initial connection state (the two groups of sub-chains connect the matching sub-children one-to-one, and the matching codes corresponding to the matching sub-children are used as the connection pairing conditions). The sub-interval with preset data in the middle part of the matching sub-children is selected as the open interval. The open interval is open to the outside. The sub-intervals on both sides of the open interval correspond to the supplementary intervals of the adjacent sub-intervals respectively. The supplementary interval is closed to the outside and is only used for subsequent supplementation of adjacent matching sub-children, which can ensure the accuracy of the data.

[0065] In one embodiment, the step S3 of recovering the data segment of the abnormal matching sub-item based on the supplementary interval includes:

[0066] S31, dividing the specification data and the annotated architectural drawing sample into a number of data segments equal to the number of matching sub-segments;

[0067] S32, establishing a one-to-one correspondence between the data segment and the matching element, randomly dividing the data segment into different data sizes according to the number of open intervals in the corresponding matching element to obtain multiple sub-data blocks, and storing the multiple sub-data blocks in the corresponding open intervals of the matching element;

[0068] S33, the data segment is copied simultaneously and divided into two parts with different data amounts to obtain two parts of supplementary data, and the supplementary data are respectively stored in supplementary intervals in adjacent matching sub-parts;

[0069] S34. Sort the matching codes of the matching sub-elements from large to small according to the amount of data in the corresponding sub-intervals, and use the matching codes of the matching sub-elements after sorting the two sub-chains from large to small as the pairing condition;

[0070] S35: The matching element with a changed matching code is regarded as an abnormal matching element, the data segment in the open interval of the abnormal matching element is destroyed, and the data segment of the abnormal matching element is recovered based on the supplementary interval in the adjacent matching element.

[0071] In one embodiment, the step S32 of storing the supplementary data in the supplementary intervals of adjacent matching sub-elements respectively includes:

[0072] 321. Divide the supplementary data into different data amounts according to the number of supplementary intervals in adjacent matching sub-groups;

[0073] 322. Store the divided supplementary data in corresponding supplementary intervals respectively.

[0074] As described in steps S31-S35 above, the specification data and the annotated architectural drawing samples are divided into multiple data segments equal to the number of matching sub-groups. The data segments can be stored one by one in the matching sub-groups. The data segments are randomly divided into different data volumes according to the number of open intervals in the corresponding matching sub-groups. The different data volumes represent different sorting of matching codes, thereby obtaining multiple sub-data blocks. The multiple sub-data blocks are stored in the open intervals of the matching sub-groups. The open intervals are open to the outside world. The data can be exported for training deep learning models. The data segments are simultaneously copied and divided into two copies with different data volumes. Get two sets of supplementary data, store them in the supplementary intervals of adjacent matching sub-groups respectively. The supplementary intervals cannot be accessed, and data recovery can only be performed on adjacent matching sub-groups in the annotation rule library. Specifically, the supplementary data are divided into different data volumes according to the number of supplementary intervals in the adjacent matching sub-groups, and the divided supplementary data are stored in the corresponding supplementary intervals respectively. The matching codes of the supplementary intervals in the matching sub-groups and the matching codes of the open intervals are combined into codes, and the matching codes of the matching sub-groups are sorted from large to small according to the data volume in the corresponding sub-intervals (here, the sorted matching codes are the codes). The matching codes of the two subchains after sorting from large to small are used as the pairing condition. The two subchains can be matched and connected through the matching codes of the matching codes after sorting from large to small. In this way, the matching codes with changes are regarded as abnormal matching codes (matching code changes mean that the matching codes have been accessed and tampered by external networks, resulting in changes in the size of the data. Even a short change time will cause a short change in the sorting of the matching codes). The abnormal matching codes on the subchain can be quickly disconnected from the corresponding matching codes on the other subchain, with a quick response. Destroying the data segment can prevent unsafe data from affecting the data in more matching sub-chains. The data segment of the abnormal matching sub-chain is restored based on the supplementary interval in the adjacent matching sub-chains. After the abnormal destruction of the data segment, the data segment in the abnormal matching sub-chain can be restored through the adjacent matching sub-chains. The restored matching sub-chain is connected to the corresponding matching sub-chain in another sub-chain through the matching code of the matching sub-chain sorted from largest to smallest, which can ensure the safe storage of data. For example, there are two sub-chains with four matching sub-chains each, and the matching sub-chains in one of the sub-chains are recorded as a1, a2, a3 and a4 respectively.The matching sub-chains in another sub-chain are denoted as b1, b2, b3 and b4 respectively. For example, each matching sub-chain has four sub-intervals denoted as c1, c2, c3 and c4. c2 and c3 can be used as open intervals, c1 and c4 as supplementary intervals, c2 and c3 are used to store the first data segment of different data amounts, c1 is used to store a data segment of a part of the data amount in one of the adjacent matching sub-chains, and c4 is used to store a data segment of a part of the data amount in another adjacent matching sub-chain. a1 and b1 are connected by the matching codes of the matching sub-chains after sorting from large to small, a2 and b2 are connected by the matching codes of the matching sub-chains after sorting from large to small, and a2 and b2 are connected by the matching codes of the matching sub-chains after sorting from large to small. In the process of storing data, the storage safety index can be evaluated. The calculation formula of the storage safety index is: ,in, is the storage safety index, The number of times the matching code of the matcher changes. is the frequency of changes in the matching code of the matcher, is the number of matches, This is a constant greater than zero. It should be noted that a larger value for the storage security index indicates a greater risk to data storage security. Matchers ensure data security and integrity. When data anomalies occur, adjacent matchers can help the abnormal matcher recover quickly. This provides a fast response and timely stop-loss to avoid impacting more data. This provides good data storage stability, ensuring accurate data can be used to train accurate models.

[0075] In one embodiment, the step S4 of inputting the architectural drawing data into the trained violation recognition model to output violation information includes:

[0076] S41, obtaining data segments through the open interval, and obtaining the standard data of the annotation rule library and the annotated architectural drawing sample through the data segments in the entire matching loop;

[0077] S42, training the deep learning model by using the normative data of the annotation rule library and the annotated architectural drawing samples to obtain a violation recognition model;

[0078] S43. Input the architectural drawing data that needs to be analyzed into the violation identification model to output the violation components and the corresponding violation degrees.

[0079] As described in the above steps S41-S43, the open interval can export data for model training. The deep learning model here is a convolutional neural network, which can obtain the standard data of the annotation rule library and the architectural drawing samples after annotation for the data segments in the entire matching loop. The deep learning model is trained by the standard data of the annotation rule library and the architectural drawing samples after annotation to obtain a violation recognition model. After that, the model can be put into use, and the architectural drawing data that currently needs to be analyzed is input into the violation recognition model to output the illegal components and the corresponding violation degree as violation information, which can complete the compliance review of the construction project.

[0080] In one embodiment, the step S5 of binding the violation information with the corresponding building drawing data as an analysis log and storing the analysis log includes:

[0081] S51, obtaining analysis time of architectural drawing data;

[0082] S52. Bind the violation information and the corresponding building drawing data according to the analysis time and store them as an analysis log.

[0083] As described in the above steps S51 and S52, if there is any non-compliance in the architectural drawing data after analysis, it needs to be modified, and then the architectural drawing data is analyzed again until it is compliant. There is analysis time in the process of compliance review of the architectural drawing data. Therefore, for subsequent data tracing and division of responsibilities for architectural drawing data, it is necessary to bind the violation information and the corresponding architectural drawing data according to the analysis time and store them as an analysis log.

[0084] Example 2, please refer to Figure 2 As shown, the construction project compliance review system based on AI decision tracing described in this embodiment includes:

[0085] A storage module is used to determine specification data, annotate architectural drawing samples according to the specification data, and store the specification data and the annotated architectural drawing samples as an annotation rule library;

[0086] A setting module is connected to the storage module and is used to set a matching ring in the annotation rule library, wherein the matching ring is composed of two sub-rings, and both sub-rings are composed of multiple matching sub-rings, and an open interval and a supplementary interval are set for each matching sub-ring;

[0087] A data processing module, connected to the setting module, is used to divide the specification data and the annotated architectural drawing samples into multiple data segments, store the data segments in the matching sub-modules, and when an abnormality occurs in the matching sub-modules, destroy the data segments in the matching sub-modules and recover the data segments of the abnormal matching sub-modules based on the supplementary interval;

[0088] An analysis module, connected to the data processing module, is configured to train a deep learning model based on the open interval using the specification data of the annotation rule library and the annotated architectural drawing samples to obtain a trained violation recognition model, determine architectural drawing data, input the architectural drawing data into the trained violation recognition model, and output violation information, wherein the violation information includes the illegal component and the corresponding violation degree;

[0089] The recording module is connected to the analysis module and is used to bind the violation information with the corresponding building drawing data as an analysis log and store the analysis log.

[0090] It should be noted that the security and completeness of the data can be guaranteed through the matcher. When data anomalies occur, the adjacent matchers can help the abnormal matcher to recover quickly. It has a fast response and can stop the loss in time to avoid affecting more data, and has good data storage stability.

[0091] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A construction project compliance review method based on AI decision tracing, characterized by: The following steps are involved: Determine the specification data, annotate the architectural drawing samples according to the specification data, and store the specification data and the annotated architectural drawing samples as an annotation rule library; A matching ring is set in the annotation rule library, wherein the matching ring is composed of two sub-rings, and both sub-rings are composed of multiple matching sub-rings, and an open interval and a supplementary interval are set for each matching sub-ring; Divide the specification data and the annotated architectural drawing samples into multiple data segments, store the data segments in the matching sub-items, and when an anomaly occurs in the matching sub-items, destroy the data segments in the matching sub-items and recover the data segments of the abnormal matching sub-items based on the supplementary interval; Based on the open interval, the deep learning model is trained with the specification data of the annotation rule library and the annotated architectural drawing samples to obtain a trained violation recognition model, the architectural drawing data is determined, and the architectural drawing data is input into the trained violation recognition model to output violation information, wherein the violation information includes the illegal components and the corresponding violation degree; Bind the violation information with the corresponding building drawing data as an analysis log, and store the analysis log; The step of setting a matching ring in the annotation rule library includes: Configure two sub-rings in the annotation rule base. Both sub-rings are composed of multiple identical matching sub-rings connected in a chain. Divide the matching sub-interval into multiple parallel sub-intervals, connect the multiple sub-intervals to each other, set corresponding matching codes for the multiple sub-intervals, and bind the matching codes to the corresponding sub-intervals; Connect the two sub-chains to the matching sub-chains one-to-one, and use the matching codes corresponding to the matching sub-chains as the connection pairing conditions; Select the subinterval that matches the preset data in the middle of the subinterval as the open interval, and arrange the subintervals on both sides of the open interval to correspond to the supplementary intervals of the adjacent subintervals respectively; The step of recovering the data segment of the abnormal matching sub-item based on the supplementary interval includes: Divide the specification data and the annotated architectural drawing samples into a number of data segments equal to the number of matching sub-groups; A one-to-one correspondence is established between the data segment and the matching sub-item, and the data segment is randomly divided into different data volumes according to the number of open intervals in the corresponding matching sub-item to obtain multiple sub-data blocks, and the multiple sub-data blocks are stored in the open intervals of the matching sub-item; The data segment is copied simultaneously and divided into two parts with different data amounts to obtain two supplementary data, and the supplementary data are respectively stored in the supplementary intervals of adjacent matching sub-parts; Sort the matching codes of the matchers from large to small according to the amount of data in the corresponding subintervals, and use the matching codes of the two subchains after sorting them from large to small as the pairing condition; The matching sub-items with changed matching codes are regarded as abnormal matching sub-items, the data segments in the open intervals of the abnormal matching sub-items are destroyed, and the data segments of the abnormal matching sub-items are recovered based on the supplementary intervals in the adjacent matching sub-items; The step of storing the supplementary data in the supplementary intervals of adjacent matching sub-elements respectively includes: The supplementary data are divided into different data amounts according to the number of supplementary intervals in adjacent matching sub-groups; The divided supplementary data are stored in corresponding supplementary intervals respectively.

2. The construction project compliance review method based on AI decision tracing according to claim 1 is characterized by: The step of storing the specification data and the annotated architectural drawing samples as an annotation rule library includes: Acquiring a plurality of architectural drawings as architectural drawing samples, and reading component information from the architectural drawing samples, wherein the component information includes geometric properties and annotation information of the components; Mark component information in architectural drawing samples according to specification data; The specification data and the annotated architectural drawing samples are stored to obtain an annotation rule library.

3. The construction project compliance review method based on AI decision tracing according to claim 1 is characterized by: The step of inputting the architectural drawing data into the trained violation recognition model to output violation information includes: Obtain data segments through the open interval, and obtain the standard data of the annotation rule library and the architectural drawing samples after annotation through the data segments in the entire matching loop; The violation recognition model is obtained by training the deep learning model through the normative data of the annotation rule library and the annotated architectural drawing samples; The architectural drawing data that needs to be analyzed is input into the violation identification model to output the violation components and the corresponding violation degree.

4. The construction project compliance review method based on AI decision tracing according to claim 1 is characterized by: The step of binding the violation information with the corresponding building drawing data as an analysis log and storing the analysis log includes: Obtain analysis time for architectural drawing data; The violation information and the corresponding building drawing data are bound and stored as analysis logs according to the analysis time.

5. A construction project compliance review system based on AI decision tracing, used to implement a construction project compliance review method based on AI decision tracing as described in any one of claims 1 to 4, characterized in that: include: A storage module is used to determine specification data, annotate architectural drawing samples according to the specification data, and store the specification data and the annotated architectural drawing samples as an annotation rule library; A setting module is connected to the storage module and is used to set a matching ring in the annotation rule library, wherein the matching ring is composed of two sub-rings, and both sub-rings are composed of multiple matching sub-rings, and an open interval and a supplementary interval are set for each matching sub-ring; A data processing module, connected to the setting module, is used to divide the specification data and the annotated architectural drawing samples into multiple data segments, store the data segments in the matching sub-modules, and when an abnormality occurs in the matching sub-modules, destroy the data segments in the matching sub-modules and recover the data segments of the abnormal matching sub-modules based on the supplementary interval; An analysis module, connected to the data processing module, is configured to train a deep learning model based on the open interval using the specification data of the annotation rule library and the annotated architectural drawing samples to obtain a trained violation recognition model, determine architectural drawing data, input the architectural drawing data into the trained violation recognition model, and output violation information, wherein the violation information includes the illegal component and the corresponding violation degree; The recording module is connected to the analysis module and is used to bind the violation information with the corresponding building drawing data as an analysis log and store the analysis log.

Images