A BIM-based cloud collaborative digital engineering management system

Abstract

The application discloses a cloud collaborative digital engineering management system based on BIM, and relates to the technical field of cloud collaboration, comprising the following steps: firstly, a dynamic digital twin body integrating BIM, GIS and sensor data is constructed, and then data anomalies are analyzed; if the anomalies are similar to historical conditions, a solution is directly pushed and permission is locked; if it is a new problem, permission is temporarily opened, and after the problem is solved, a new solution is saved and the system is calibrated, realizing safe closed-loop operation and maintenance. Through intelligent identification of data anomaly types, the application automatically pushes historical solutions or dynamically authorizes on-site troubleshooting, efficiently solves faults while continuously accumulating a knowledge base, and realizes effective inheritance of operation and maintenance experience and accurate balance of data security permissions.

Description

Technical Field

[0001] This invention relates to the field of cloud-based collaborative technology, specifically to a cloud-based collaborative digital engineering management system based on BIM. Background Technology

[0002] Addressing the challenges of complex natural environments, limited communication, and difficult on-site management in remote water conservancy projects, the digital engineering management system is based on building a high-precision digital twin. It integrates BIM micro-models, GIS macro-topography, and real-time data from drones, BeiDou, and IoT, calibrating and connecting multi-source heterogeneous data in the cloud. Leveraging an intelligent engine, it achieves 3D visualization under low bandwidth and automatically monitors key indicators such as slope displacement using AI risk identification, thus realizing data-driven intelligent control throughout the entire process. Regarding data security, the system implements strict hierarchical access control, restricting access to real-time project status data to core decision-makers to prevent information leakage and misinterpretation. This access system is dynamic and flexible: when the system triggers an alert, data access is temporarily granted to relevant maintenance personnel to guide accurate troubleshooting; access is automatically revoked after the incident, thus achieving a balance between safety and efficiency.

[0003] However, if the cause of the accident and the solution are not reported back to the system after maintenance, it will lead to the loss of handling experience. This will not only waste resources when troubleshooting similar problems in the future, but also require data access to be re-enabled every time the problem recurs, which will multiply the risk of data exposure. Summary of the Invention

[0004] Technical problems to be solved

[0005] To address the shortcomings of existing technologies, this invention provides a cloud-based collaborative digital engineering management system based on BIM. This system solves the problem that if maintenance personnel fail to report the cause and solution back to the system after troubleshooting, it will lead to the loss of organizational experience. This not only wastes the efficiency of subsequent troubleshooting of similar problems, but also causes the data security risks to accumulate due to the repeated granting of permissions.

[0006] To achieve the above objectives, this invention provides the following technical solution: a BIM-based cloud-based collaborative digital engineering management system, comprising the following specific modules: a twin component module: constructing a dynamic digital twin based on the BIM model, GIS map, and various sensors, including dynamic multi-source heterogeneous data; a data anomaly analysis module: performing comprehensive analysis on the stagnation and fluctuation of multi-source heterogeneous data to obtain a data anomaly model, analyzing whether the data is abnormal based on the data anomaly model; if the analysis finds it normal, then closing the login permissions of all construction personnel to the platform and ending the process; if the analysis finds anomalies, then executing the data similarity analysis module; and a data similarity analysis module: comparing the multi-source heterogeneous data with historically abnormal multi-source heterogeneous data. Similarity analysis: If similarity is found, the relevant issues and solutions are sent to the construction personnel's devices, but the construction personnel's login permissions are not unlocked, and the process returns to the twin component module. If dissimilarity is found, the construction personnel's login permissions are granted. After the construction personnel solve the problem, the new relevant issues and solutions are uploaded and saved to the cloud, and the issue similarity analysis module is executed. Issue similarity analysis module: Performs similarity analysis on the relevant issues and saved historical issues. If similarity is found, the accuracy of the similarity analysis between multi-source heterogeneous data and historically abnormal multi-source heterogeneous data is calibrated, and the relevant issues and solutions are deleted before returning to the twin component module. If dissimilarity is found, the process directly returns to the twin component module.

[0007] Furthermore, the specific method for obtaining the data anomaly model is as follows: when a certain value in multi-source heterogeneous data changes, the recording time is started to obtain the change time. Based on the change time and the interval time threshold, a stagnation value is obtained. At the same time, the amplitude of each change of this value is statistically analyzed to obtain the number of changes and the amplitude of each change. Based on the number of changes and the amplitude of each change, the standard deviation is calculated to obtain the change fluctuation. The change fluctuation is calculated with the amplitude change fluctuation threshold to obtain the fluctuation anomaly value. Based on the stagnation value and the fluctuation anomaly value, a comprehensive calculation is performed and standardized to obtain the data anomaly model. ;in, Indicates a data anomaly model. Indicates a stagnant value. Indicates outlier fluctuations. or All of them represent positive real numbers.

[0008] Furthermore, the specific method for obtaining the stagnation value is as follows: the difference between the change time and the interval time threshold is calculated to obtain the stagnation value.

[0009] Furthermore, the specific method for obtaining the fluctuation anomaly value is as follows: if the fluctuation anomaly value is greater than the upper limit of the amplitude change fluctuation threshold, the difference between the fluctuation anomaly value and the upper limit of the amplitude change fluctuation threshold is calculated; if the fluctuation anomaly value is less than the lower limit of the amplitude change fluctuation threshold, the difference between the fluctuation anomaly value and the lower limit of the amplitude change fluctuation threshold is calculated to obtain the fluctuation anomaly value.

[0010] Furthermore, the specific method for performing similarity analysis between multi-source heterogeneous data and historically abnormal multi-source heterogeneous data is as follows: set a truncation time period and a historical abnormal data segment; when the value of the data anomaly model is greater than the data anomaly threshold, the abnormal multi-source heterogeneous data is truncated within the truncation time period to obtain an abnormal data segment; the abnormal data segment is matched with the historical abnormal data segment; if the abnormal data segment matches the historical abnormal data segment, it indicates similarity; if the abnormal data segment does not match the historical abnormal data segment, it indicates dissimilarity.

[0011] Furthermore, the specific method for matching the abnormal data segment with historical abnormal data segments is as follows: based on the time series, the values ​​at each moment in the abnormal data segment are... The value at each time point in the historical outlier data segment Perform the difference calculation and take the absolute value to obtain the matching value. Set a matching threshold. Statistically count the number of matching values ​​within the matching threshold and the total number of matching values ​​based on the time series. Calculate the ratio of the number of matching values ​​within the matching threshold to the total number of matching values ​​to obtain the similarity ratio. Set a data similarity ratio threshold, and set the data similarity ratio threshold to be in the interval between zero and one. According to probability theory, compare the similarity ratio with the data similarity ratio threshold. If the similarity ratio is greater than the data similarity ratio threshold, it indicates a match; if the similarity ratio is less than or equal to the data similarity ratio threshold, it indicates a mismatch.

[0012] Furthermore, the specific method for performing similarity analysis between the relevant question and the saved historical question is as follows: both the relevant question and the saved historical question include n characters. Each character is assigned a corresponding character code, and the character code is unique. The characters of the relevant question are matched with the characters of the saved historical question. If the relevant question and the saved historical question match, they are considered similar; otherwise, they are considered dissimilar.

[0013] Furthermore, the specific method for performing similarity matching between the characters of related questions and the characters of saved historical questions is as follows: The character codes of related questions and the character codes of saved historical questions are stored in one-dimensional arrays respectively, resulting in a related array and a historical array. The character codes in each array are arranged according to their sequence numbers. In the first round, matching is performed starting from the character code with sequence number 0 in each of the two arrays, with one character code matched each time, until the sequence numbers of one of the arrays are completely matched. The round ends when either side is completely matched, and the number of identical character codes in the first round is obtained. In the second round, matching is performed starting from the character code with sequence number 0 in each of the two arrays, with one character code matched each time. The number of character codes is two. The process continues until the sequence numbers of one array are matched, yielding the number of identical character codes in the second round. This continues until the number of matched character codes equals the sequence length of a certain array, resulting in the number of identical character codes in the m-th round. The number of identical character codes in the first round, the second round, and so on, is then combined to calculate a character code similarity value. A character code similarity threshold is set, and the similarity value is compared to this threshold. If the similarity value is within the threshold, it indicates that the characters in the relevant question are similar to the characters in the saved historical questions; otherwise, it indicates that the characters in the relevant question are not similar to the characters in the saved historical questions.

[0014] Furthermore, the specific method for obtaining the character code similarity value is as follows: multiply the number of identical character codes in the first round by the corresponding round number to obtain the number of identical codes in the first round; multiply the number of identical character codes in the second round by the corresponding round number to obtain the number of identical codes in the second round; and so on, to obtain the number of identical codes in the m-th round; then sum the number of identical codes in the first round, the second round, ..., the m-th round to obtain the character code similarity value.

[0015] Furthermore, the specific method for calibrating the accuracy of the similarity analysis between multi-source heterogeneous data and historical anomalies is to adjust the data similarity ratio threshold towards a value approaching zero.

[0016] Beneficial effects

[0017] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0018] 1. By constructing data similarity analysis and problem similarity analysis modules, a closed-loop mechanism for automatic experience accumulation and reuse is established: When the system detects data anomalies similar to historical records, it can directly push archived relevant issues and solutions to on-site personnel, enabling them to quickly resolve issues without re-investigation; when encountering entirely new anomalies, the system will automatically save new issues and solutions after the problem is resolved, continuously expanding the knowledge base. This effectively overcomes the pain point of experience loss due to personnel turnover in traditional operations and maintenance, transforming personal experience into organizational assets, significantly improving the efficiency of handling similar issues subsequently, and avoiding the waste of time and resources from repeated investigations.

[0019] 2. Through a rigorous data anomaly model and multi-layer similarity analysis, precise control and intelligent decision-making regarding access permissions are achieved: login permissions for construction personnel are only temporarily granted when data anomalies are confirmed and there are no historical solutions available for reference; once the problem is resolved, the permissions are immediately revoked, ensuring that on-site personnel can obtain the necessary data to troubleshoot while minimizing the exposure time and scope of the platform's core data, fundamentally reducing the risk of data leakage caused by frequent access permissions.

[0020] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0021] Figure 1 This invention provides a flowchart of a cloud-based collaborative digital engineering management method based on BIM.

[0022] Figure 2 This invention relates to a cloud-based collaborative digital engineering management system based on BIM, and its structural diagram. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0025] Example 1:

[0026] like Figures 1-2 As shown, this embodiment of the invention provides a BIM-based cloud-based collaborative digital engineering management system, which includes the following specific modules:

[0027] Twin Component Module: First, the BIM model and GIS map are integrated in the cloud to assign unified geographic coordinates to all components. At the same time, various sensors deployed on site upload real-time data such as settlement, stress and seepage pressure to the platform via 4G or 5G networks. After the platform analyzes and cleans this batch of multi-source heterogeneous data, it attaches the multi-source heterogeneous data to the corresponding BIM model components based on the relationship between components and coordinates established by the aforementioned fusion model. This transforms the static BIM model into a dynamic digital twin that can monitor the status of the physical project in real time on the platform.

[0028] Data Anomaly Analysis Module: Based on the stagnation and fluctuation of multi-source heterogeneous data in the dynamic digital twin, a data anomaly model is obtained. The data anomaly model is used to analyze whether the data is abnormal. A data anomaly threshold is set through historical experiments. The value of the data anomaly model is compared with the data anomaly threshold. If the value of the data anomaly model is less than or equal to the data anomaly threshold, it is considered normal. If the analysis shows that it is normal, the login permissions of all construction personnel are closed to prevent construction personnel from leaking information, and the process ends. If the value of the data anomaly model is greater than the data anomaly threshold, it is considered abnormal. If the analysis shows that it is abnormal, the data similarity analysis module is executed.

[0029] The data similarity analysis module performs similarity analysis between multi-source heterogeneous data and historically abnormal multi-source heterogeneous data. If similarity is found, relevant issues and solutions are sent to the construction personnel's devices, but their access to the platform is not unlocked. That is, although the data is abnormal, the abnormal data state is similar to the previously recorded abnormal data state. The relevant issues that caused the data abnormality and the solutions to the data abnormality have been uploaded and saved to the cloud. Therefore, the construction personnel can directly solve the data abnormality based on the relevant issues and solutions before, without having to start from scratch to find solutions. The restriction on the construction personnel's access to the platform is also to prevent the risk of platform information leakage. The module then returns to the twin component module. If dissimilarity is found, the construction personnel's access to the platform is granted, allowing them to investigate relevant issues and find solutions based on the multi-source heterogeneous data in the dynamic digital twin. After the construction personnel solve the problem, the new relevant issues and solutions are uploaded and saved to the cloud, and the issue similarity analysis module is executed.

[0030] Problem Similarity Analysis Module: This module performs similarity analysis between relevant problems and saved historical problems. If similarity is found, it indicates that the accuracy of similarity analysis between multi-source heterogeneous data and historically abnormal multi-source heterogeneous data is low. In other words, the original multi-source heterogeneous data and historically abnormal multi-source heterogeneous data were analyzed as dissimilar, necessitating granting permissions to construction personnel. Since the problems uploaded by construction personnel are similar to saved historical problems, permissions that were not originally necessary have been granted to them. This not only wastes the efficiency of subsequent investigations into similar problems but also continuously increases data security risks due to the repeated granting of permissions. Therefore, the accuracy of the similarity analysis between multi-source heterogeneous data and historically abnormal multi-source heterogeneous data is calibrated, and the relevant problems and solutions are deleted to improve storage space before returning to the twin component module. If dissimilarity is found, the module directly returns to the twin component module.

[0031] Example 2 differs from Example 1 in that:

[0032] The specific methods for obtaining the data anomaly model are as follows:

[0033] Since multi-source heterogeneous data is dynamic, the average change interval of each value in normal multi-source heterogeneous data is recorded to obtain the interval time threshold. When a value does not change within the interval time threshold, it indicates that the change of this value is stagnant. At the same time, the average amplitude change fluctuation of each value in normal multi-source heterogeneous data is recorded to obtain the amplitude change fluctuation threshold, which is in the form of an interval. When a value is not within the amplitude change fluctuation threshold, it indicates that the amplitude fluctuation of this value is abnormal.

[0034] Therefore, when a value in multi-source heterogeneous data changes, the recording time is started to obtain the change time. The stagnation value is obtained by calculating the change time and the interval time threshold. At the same time, the amplitude of each change of this value is statistically analyzed to obtain the number of changes and the amplitude of each change. The standard deviation is calculated based on the number of changes and the amplitude of each change to obtain the change fluctuation. The change fluctuation is calculated with the amplitude change fluctuation threshold to obtain the fluctuation anomaly value.

[0035] The data anomaly model is obtained by comprehensively calculating the stagnant values ​​and fluctuation anomalies, and by standardizing the data to eliminate dimensional differences and converting values ​​of different orders of magnitude into a unified numerical range.

[0036] ;

[0037] in, This represents a data anomaly model, reflecting whether the data is abnormal. This indicates a stagnant value, reflecting whether the change of a certain value in multi-source heterogeneous data has stagnated. This indicates outlier values, reflecting whether the fluctuation of a certain value in multi-source heterogeneous data is abnormal. The absolute value of the outlier value indicates that the outlier value may be less than the lower limit of the amplitude change threshold or greater than the upper limit of the amplitude change threshold. or All numbers represent positive real numbers to avoid the fact that logarithms do not hold true.

[0038] The specific method for obtaining the stagnation value is as follows:

[0039] The difference between the change time and the interval time threshold is calculated to obtain the stagnation value. The longer the change time, the larger the stagnation value.

[0040] The specific methods for obtaining fluctuation outliers are as follows:

[0041] If the fluctuation anomaly is greater than the upper limit of the amplitude change fluctuation threshold, the difference between the fluctuation anomaly and the upper limit of the amplitude change fluctuation threshold is calculated. If the fluctuation anomaly is less than the lower limit of the amplitude change fluctuation threshold, the difference between the fluctuation anomaly and the lower limit of the amplitude change fluctuation threshold is calculated to obtain the fluctuation anomaly.

[0042] The specific method for performing similarity analysis between multi-source heterogeneous data and historically anomalous multi-source heterogeneous data is as follows:

[0043] Set a time frame and historical abnormal data segments. When the value of the data anomaly model is greater than the data anomaly threshold, the abnormal multi-source heterogeneous data within the time frame is truncated to obtain an abnormal data segment. The abnormal data segment is then matched with historical abnormal data segments. The historical abnormal data segments are also multi-source heterogeneous data segments that were truncated within the time frame when the value of the data anomaly model is greater than the data anomaly threshold. If the abnormal data segment matches the historical abnormal data segment, it is considered similar; if the abnormal data segment does not match the historical abnormal data segment, it is considered dissimilar.

[0044] The specific method for matching abnormal data segments with historical abnormal data segments is as follows:

[0045] Based on the time series, the values ​​at each moment in the outlier data segment are... The value at each time point in the historical outlier data segment Perform a difference calculation and take the absolute value to obtain the matching value. Set a matching threshold. Based on the time series, statistically analyze the number of matching values ​​within the matching threshold and the total number of matching values. Calculate the similarity ratio by dividing the number of matching values ​​within the matching threshold by the total number of matching values. ;in, Represents the similarity ratio. This indicates the number of matching values ​​within the matching threshold. This represents the total number of matched values, and is not equal to zero.

[0046] Let there be a data similarity ratio threshold, and the data similarity ratio threshold is in the range of zero to one. According to probability theory, the similarity ratio value is compared with the data similarity ratio threshold. If the similarity ratio value is greater than the data similarity ratio threshold, it means that there is a match, that is, the closer it is to one, the more matched. If the similarity ratio value is less than or equal to the data similarity ratio threshold, it means that there is no match, that is, the closer it is to zero, the less matched.

[0047] The specific method for conducting similarity analysis between related issues and preserved historical issues is as follows:

[0048] Both the related questions and the saved historical questions consist of n characters. Each character is assigned a unique character code, indicating the uniqueness of the character. The characters in the related questions are matched with the characters in the saved historical questions based on similarity. If the related questions and the saved historical questions match, they are considered similar; otherwise, they are considered dissimilar.

[0049] The specific method for performing similarity matching between the characters of relevant questions and the characters of saved historical questions is as follows:

[0050] The character codes of the relevant questions and the saved character codes of the historical questions are stored in one-dimensional arrays to obtain the relevant array and the historical array. The character codes in each array are arranged according to their sequence numbers. In the first round, the same character codes are matched starting from the character codes with sequence number 0 in each array. The number of character codes matched each time is one, until the sequence numbers of one set of arrays are matched. That is, the lengths of the two sets of arrays are not necessarily the same. The round ends when either set of arrays is matched. The number of the same character codes in the first round is obtained. In the second round, the same character codes are matched starting from the character codes with sequence number 0 in each array. The number of character codes matched each time is two, until the sequence numbers of one set of arrays are matched. The number of the same character codes in the second round is obtained. This process continues until the number of character codes matched each time is equal to the sequence length of a certain array. The number of the same character codes in the m-th round is obtained.

[0051] The number of identical character codes in the first round, the second round, and so on, up to the m-th round, is combined to calculate the character code similarity value. A character code similarity threshold is set based on historical experiments. The character code similarity value is compared with the character code similarity threshold. If the character code similarity value is within the character code similarity threshold, it means that the characters of the relevant question are similar to the characters of the saved historical questions; otherwise, it means that the characters of the relevant question are not similar to the characters of the saved historical questions.

[0052] The specific method for obtaining character code similarity values ​​is as follows:

[0053] The number of identical character codes in the first round is multiplied by the corresponding round number to obtain the number of identical codes in the first round. The number of identical character codes in the second round is multiplied by the corresponding round number to obtain the number of identical codes in the second round. This process is repeated until the number of identical codes in the m-th round is obtained. The number of identical codes in the first round, the second round, ..., the m-th round are then summed to obtain the character code similarity value.

[0054] ;

[0055] in, Indicates character code similarity value, Indicates the total number of rounds. Indicates the first wheel, Indicates the first The number of identical character codes in the round.

[0056] The specific method for calibrating the accuracy of similarity analysis between multi-source heterogeneous data and historical anomalies is as follows:

[0057] Adjusting the data similarity threshold towards a value approaching zero expands the range of the data similarity threshold to one, thereby improving the accuracy of data similarity.

[0058] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A BIM-based cloud collaborative digital engineering management system, characterized in that: The specific modules include the following: Twin component module: constructing a dynamic digital twin according to a BIM model, a GIS map, and various sensors, including dynamic multi-source heterogeneous data; Data anomaly analysis module: comprehensively analyzing stagnation and fluctuations of multi-source heterogeneous data to obtain a data anomaly model, analyzing whether the data is abnormal according to the data anomaly model, if normal, closing the permission and ending, if abnormal, executing a data similarity analysis module; Data similarity analysis module: performing similarity analysis on multi-source heterogeneous data and historical multi-source heterogeneous data of anomalies, if similar, sending related problems and solutions and returning, if not similar, opening the permission, after the construction personnel solve the problem, uploading new related problems and solutions to the cloud and executing a problem similarity analysis module; Problem similarity analysis module: performing similarity analysis on related problems and saved historical problems, if similar, calibrating the accuracy of data similarity analysis, deleting the related problems and solutions, and returning, if not similar, directly returning; The specific way of performing similarity analysis on related problems and saved historical problems is as follows: The related problems and the saved historical problems each include n characters, each character is set to correspond to a character code, and the character codes are unique, the characters of the related problems are matched with the characters of the saved historical problems in similarity, if the related problems and the saved historical problems are matched, it means similar, otherwise, it means not similar; The specific way of matching the characters of the related problems with the characters of the saved historical problems in similarity is as follows: The character codes of the related problems and the character codes of the saved historical problems are respectively stored in one-dimensional arrays to obtain related arrays and historical arrays, the character codes in each array are arranged according to sequence numbers, in the first round, the character codes with sequence number 0 in the two arrays are respectively matched, the number of character codes matched each time is one, until the sequence numbers of one of the arrays are matched, when any one is matched, it is complete, the number of the same character codes in the first round is obtained, in the second round, the character codes with sequence number 0 in the two arrays are respectively matched, the number of character codes matched each time is two, until the sequence numbers of one of the arrays are matched, the number of the same character codes in the second round is obtained, and so on, until the number of character codes matched each time is equal to the sequence length of a certain array, the number of the same character codes in the mth round is obtained; the number of the same character codes in the first round, the number of the same character codes in the second round,..., and the number of the same character codes in the mth round are comprehensively calculated to obtain a character code similarity value, a character code similarity threshold value is set, the character code similarity value is compared with the character code similarity threshold value, if the character code similarity value is within the character code similarity threshold value, it means that the characters of the related problems and the characters of the saved historical problems are similar, otherwise, it means that the characters of the related problems and the characters of the saved historical problems are not similar; The specific acquisition method of the character code similarity value is as follows: The first round same character code number is multiplied by the corresponding round number to obtain a first round same number, the second round same character code number is multiplied by the corresponding round number to obtain a second round same number, and the same is done to obtain an mth round same number.

2. The BIM-based cloud collaborative digital engineering management system according to claim 1, characterized in that: The specific acquisition method of the data anomaly model is as follows: When a value in the multi-source heterogeneous data changes, the change time is recorded to obtain a change time, the change time and the interval time threshold are calculated to obtain a stagnation value, at the same time, the amplitude of each change of the value is counted to obtain the number of changes and the amplitude of each change, the number of changes and the amplitude of each change are calculated to obtain a change fluctuation, the change fluctuation and the amplitude change fluctuation threshold are calculated to obtain a fluctuation anomaly value; According to the stagnation value and the fluctuation abnormal value, a data abnormal model is obtained through comprehensive calculation and standardization processing. ; wherein, represents a data anomaly model, represents a stagnation value, represents a volatility anomaly value, or each represents a positive real number.

3. The BIM-based cloud collaborative digital engineering management system according to claim 2, characterized in that: The specific acquisition method of the stagnation value is as follows: The difference between the change time and the interval time threshold is calculated to obtain the stagnation value.

4. The BIM-based cloud collaborative digital engineering management system according to claim 2, characterized in that: The specific acquisition method of the fluctuation anomaly value is as follows: If the fluctuation anomaly value is greater than the upper limit of the amplitude change fluctuation threshold, the difference between the fluctuation anomaly value and the upper limit of the amplitude change fluctuation threshold is calculated, and if the fluctuation anomaly value is less than the lower limit of the amplitude change fluctuation threshold, the difference between the fluctuation anomaly value and the lower limit of the amplitude change fluctuation threshold is calculated to obtain the fluctuation anomaly value.

5. The BIM-based cloud collaborative digital engineering management system according to claim 1, characterized in that: The specific method of similarity analysis of the multi-source heterogeneous data and the multi-source heterogeneous data of historical anomalies is as follows: The interception time period and the historical abnormal data segment are set, when the value of the data anomaly model is greater than the data anomaly threshold, the abnormal multi-source heterogeneous data in the interception time period is intercepted to obtain an abnormal data segment, the abnormal data segment and the historical abnormal data segment are matched, if the abnormal data segment and the historical abnormal data segment are matched, it means similar, if the abnormal data segment and the historical abnormal data segment are not matched, it means not similar.

6. The BIM-based cloud collaborative digital engineering management system according to claim 5, characterized in that: The specific method of matching the abnormal data segment and the historical abnormal data segment is as follows: According to the time sequence, the value of each time point in the abnormal data segment The value of each time point in the historical abnormal data segment Difference calculation is performed, and an absolute value is taken to obtain a matching value, a matching threshold is set, the number of matching values within the matching threshold and the total number of matching values are counted according to the time sequence, and the number of matching values within the matching threshold and the total number of matching values are calculated by ratio to obtain a similarity ratio; The data similarity ratio threshold is set, and the data similarity ratio threshold is in the interval of zero to one, the similarity ratio value is compared with the data similarity ratio threshold according to probability theory, if the similarity ratio value is greater than the data similarity ratio threshold, it means matching, if the similarity ratio value is less than or equal to the data similarity ratio threshold, it means not matching. 7.The BIM-based cloud collaborative digital engineering management system according to any one of claims 1 and 6, characterized in that: The specific method of calibrating the accuracy of data similarity analysis is as follows: The data similarity ratio threshold is adjusted to a value close to zero.

Images