Intelligent construction information abnormity identification method

Abstract

The invention relates to the technical field of anomaly recognition, and discloses an intelligent construction information anomaly recognition method, which comprises an information acquisition step, an evaluation step and an early warning step which are executed in sequence. According to the application, the information acquisition module is used for realizing the acquisition of the multi-source data related to the construction information, the evaluation module is used for realizing the information anomaly identification under the overall dimension and the individual dimension, and the identification of the construction link of the individual is corrected by using the overall information anomaly; the early warning module is utilized to realize visual display of the evaluation result of the evaluation module, and a comprehensive coping mode is provided for coping with the construction information abnormity, so that the score of the construction information abnormity of the construction site is comprehensively calculated to integrally grasp the condition of the construction information abnormity in the site; and meanwhile, construction information abnormal high-risk links can be accurately captured to clarify individual abnormal specific construction links, so that a construction administrator can know potential risks in advance and take corresponding measures, the possibility of occurrence of the construction information abnormal high-risk links in a construction site is effectively reduced, the construction management level is improved, and the construction efficiency is improved. And safety, high efficiency and quality of construction are guaranteed.

Description

Technical Field

[0001] The present application relates to the technical field of anomaly recognition, and specifically to an intelligent construction information anomaly recognition method. Background Technique

[0002] In the field of construction, it is crucial to ensure the smooth progress and construction safety of the construction process. In traditional construction management, manual inspections and experience-based judgments are often relied on to detect anomalies during construction. However, this approach has many limitations, specifically manifested in:

[0003] On the one hand, it is difficult for manual inspections to achieve full-scale and real-time monitoring of the construction process, and it is easy to miss some potential anomaly information. For example, minor faults in construction equipment may not be easily detected in the initial stage, and these faults may gradually develop into serious problems, affecting the construction progress and even triggering safety accidents.

[0004] On the other hand, it is difficult to accurately grasp the usage of construction materials solely through manual statistics, and situations such as material waste or insufficient supply are likely to occur, resulting in increased costs and project delays.

[0005] Furthermore, for the evaluation of the standardization of construction workers' operation behaviors, manual judgment is highly subjective and inefficient, and potential safety hazards caused by non-standard operations cannot be detected in a timely manner.

[0006] In addition, due to the complex and changeable construction environment, such as the impact of factors like noise and dust on construction, it is difficult for humans to comprehensively and accurately evaluate.

[0007] Chinese Patent No. CN202410082553.X discloses an anomaly recognition method and system for construction based on machine vision, but this invention fails to correct independent data under specific conditions (such as when the overall data has exceeded a reasonable range), and it is unable to capture anomaly trends.

[0008] In summary, the traditional construction management method lacks efficiency, accuracy, and comprehensiveness in the recognition of construction information anomalies, and there is an urgent need for a new technical solution for intelligent construction information anomaly recognition to improve the construction management level and ensure the safety, efficiency, and quality of construction. Summary of the Invention

[0009] The purpose of the present application is to provide an intelligent construction information anomaly recognition method to solve the technical problems raised in the above background technique.

[0010] To achieve the above purpose, the present application discloses the following technical solutions: An intelligent construction information anomaly recognition method, which includes an information collection step, an evaluation step, and an early warning step that are executed in sequence;

[0011] The information collection step is configured to collect multi-source data related to construction information; among them, the multi-source data at least includes construction equipment operation data, construction material usage data, construction personnel operation behavior data, and construction environment data;

[0012] The evaluation step is configured to: based on a preset construction information anomaly recognition model, and combine the multi-source data to identify construction information anomalies; among them, the construction information anomaly recognition model is used to calculate the construction information anomaly score of the construction site and / or capture high-risk links of construction information anomalies, the construction information anomaly score is used to characterize the anomaly situation of the overall construction information under the construction site, and the high-risk link of construction information anomaly is used to characterize the specific construction link that may have anomalies under the construction site;

[0013] When the construction information anomaly score evaluated by the evaluation step does not meet the preset construction information anomaly score threshold and / or a high-risk link of construction information anomaly is identified, the warning step gives a warning reminder to the construction administrator and provides countermeasures.

[0014] Preferably, the information collection step includes an equipment data collection sub-step, which is used to collect the operation data of construction equipment based on a preset sensor; among them, the sensor is set on the construction equipment.

[0015] Preferably, the information collection step further includes a material data collection sub-step, which is used to collect construction material usage data from a preset material statistics terminal; among them, the material statistics terminal is used to record the usage of construction materials.

[0016] Preferably, the construction information anomaly recognition model is trained based on historical construction information anomaly records and combined with machine learning techniques;

[0017] The construction information anomaly recognition model stores construction information anomaly features corresponding to the multi-source data, which are used to identify construction information anomalies in the multi-source data;

[0018] The construction information anomaly recognition model also stores a construction information anomaly score calculation sub-model for calculating the construction information anomaly score and a construction information anomaly high-risk link capture sub-model for capturing the high-risk links of construction information anomalies;

[0019] The construction information anomaly score calculation sub-model evaluates the comparison situation between the multi-source data and the construction information anomaly features at a time node based on a time series, and calculates the construction information anomaly score;

[0020] The sub-model for capturing high-risk links of abnormal construction information is based on time series, continuously evaluates the comparison between multi-source data and the abnormal characteristics of construction information under this time series, and captures high-risk links of abnormal construction information in combination with the abnormal score of construction information.

[0021] Preferably, the operation of the sub-model for calculating the abnormal score of construction information includes the following steps:

[0022] A1: Obtain multi-source data within the construction site at a time node of the time series;

[0023] A2: Extract features from the multi-source data obtained in A1 to obtain corresponding real-time data features;

[0024] A3: Compare the similarity between the real-time data features and the abnormal characteristics of construction information, and calculate the abnormal score of construction information based on the result of the similarity comparison.

[0025] Preferably, the calculation of the abnormal score of construction information is specifically:

[0026] Count the number of the real-time data features, compare the real-time data features with the abnormal characteristics of construction information respectively, and calculate the abnormal score of construction information based on the obtained similarity, the number of the real-time data features and this similarity.

[0027] Preferably, the operation of the sub-model for capturing high-risk links of abnormal construction information includes the following steps:

[0028] B1: Continuously obtain multi-source data based on the time series;

[0029] B2: Extract features from the multi-source data obtained in B1 to obtain corresponding continuous data features;

[0030] B3: Based on the time series, compare the similarity between the continuous data features and the abnormal characteristics of construction information to obtain the corresponding relationship between the similarity and the time series, judge the change trend of the similarity based on this corresponding relationship, and when the change trend of the similarity does not meet the preset change threshold, capture the construction link corresponding to this change trend of the similarity and define it as a high-risk link of abnormal construction information.

[0031] Preferably, the corresponding relationship between the similarity and the time series is specifically:

[0032] Calculate the integral of the similarity between the continuous data features and the abnormal characteristics of construction information in the supervision period, calculate the maximum value and the minimum value of this integral within the supervision duration, and calculate the corresponding relationship between the similarity and the time series based on this maximum value and this minimum value; where the supervision duration is greater than the supervision period.

[0033] Preferably, the operation of the sub-model for capturing high-risk links of abnormal construction information further includes the following steps:

[0034] When the abnormal score of the construction information does not meet the threshold of the abnormal score of the construction information, update B3 to the following steps:

[0035] B31: Based on the time series, compare the similarity between the continuous data features and the abnormal features of the construction information to obtain the corresponding relationship between the similarity and the time series; obtain the abnormal score of the construction information, and correct the corresponding relationship between the similarity and the time series based on the abnormal score of the construction information to obtain a corrected corresponding relationship. Based on this corrected corresponding relationship, judge the change trend of the similarity. When the change trend of the similarity does not meet the preset change threshold, capture the construction link corresponding to the change trend of the similarity and define it as a high-risk link of abnormal construction information.

[0036] Preferably, the countermeasures at least include a construction adjustment plan, an emergency resource allocation process, and an evacuation plan for on-site construction personnel.

[0037] Beneficial effects: The intelligent construction information anomaly recognition method of the present application uses an information collection module to collect multi-source data related to construction information, uses an evaluation module to realize information anomaly recognition in both the overall and individual dimensions, and uses the overall information anomaly to correct the recognition of individual construction links. The warning module realizes the visual display of the evaluation results of the evaluation module and provides a comprehensive response method for dealing with abnormal construction information. Thus, it realizes the comprehensive calculation of the abnormal score of the construction information in the construction site to overall grasp the abnormal situation of the construction information in the site, and at the same time can accurately capture the high-risk links of abnormal construction information to clarify the specific construction links of individual anomalies, which helps the construction administrator to know potential risks in advance, take corresponding measures, effectively reduce the possibility of high-risk links of abnormal construction information occurring in the construction site, improve the construction management level, and ensure the safety, efficiency and quality of construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application 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 following drawings are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0039] Figure 1 It is a flowchart of the intelligent construction information anomaly recognition method provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] The technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0041] In this article, the term "including" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without more limitations, the elements defined by the statement "including..." do not exclude the existence of additional identical elements in the process, method, article or device including the said elements.

[0042] This embodiment discloses Figure 1 an intelligent construction information anomaly recognition method as shown in

[0043] The information collection step is configured to collect multi-source data related to construction information; wherein, the multi-source data at least includes construction equipment operation data, construction material usage data, construction personnel operation behavior data, and construction environment data;

[0044] The evaluation step is configured to: based on a preset construction information anomaly recognition model, and in combination with multi-source data, perform construction information anomaly recognition; wherein, the construction information anomaly recognition model is used to calculate the construction information anomaly score of the construction site and / or capture the high-risk links of construction information anomalies, the construction information anomaly score is used to characterize the overall construction information anomaly situation under the construction site, and the high-risk links of construction information anomalies are used to characterize the specific construction links that may have anomalies under the construction site;

[0045] When the construction information anomaly score evaluated by the evaluation step does not meet the preset construction information anomaly score threshold and / or an anomaly high-risk link of construction information is identified, the warning step gives a warning reminder to the construction administrator and provides countermeasures.

[0046] Through the above, this embodiment uses the information collection module to collect multi-source data related to construction information, uses the evaluation module to identify information anomalies in both the overall and individual dimensions, and corrects the identification of individual construction links using the overall information anomalies. The warning module is used to visually display the evaluation results of the evaluation module and provides a comprehensive response method to address construction information anomalies. Thus, it realizes the comprehensive calculation of the construction information anomaly score of the construction site to overall grasp the situation of construction information anomalies in the site, and can also accurately capture the high-risk links of construction information anomalies to clarify the specific construction links of individual anomalies. This helps the construction administrator to be aware of potential risks in advance, take corresponding measures, effectively reduce the possibility of high-risk links of construction information anomalies occurring in the construction site, improve the construction management level, and ensure the safety, efficiency, and quality of construction.

[0047] In a simple example, the multi-source data of this embodiment can be, but is not limited to:

[0048] For the operation data of construction equipment, at least such as equipment rotation speed, temperature, vibration frequency, etc.;

[0049] For the usage data of construction materials, at least including such as material consumption speed, remaining quantity, etc.;

[0050] For the operation behavior data of construction personnel, at least including such as the standard degree of operation actions of construction personnel, operation speed, etc.;

[0051] For the construction environment data, at least including such as the noise intensity and dust concentration at the construction site.

[0052] Specifically, the information collection step includes an equipment data collection sub-step, which is used to collect the operation data of construction equipment based on the preset sensors; wherein, the sensors are set on the construction equipment.

[0053] It should be noted that this embodiment uses existing equipment monitoring technologies to collect and supervise the operation data of construction equipment.

[0054] Through the above, this embodiment uses the equipment data collection sub-step to collect the equipment data of construction information, providing a data basis for the early warning of construction information anomaly identification.

[0055] Specifically, the information collection step further includes a material data collection sub-step, which is used to collect the usage data of construction materials from the preset material statistics terminal; wherein, the material statistics terminal is used to record the usage situation of construction materials.

[0056] It should be noted that this embodiment uses existing material management technologies to statistically collect and supervise the usage situation of construction materials.

[0057] Through the above, the present embodiment uses the device data collection sub-step to achieve the collection of material data of construction information, providing a data basis for the early warning of abnormal construction information identification.

[0058] Specifically, the construction information anomaly recognition model is trained based on historical construction information anomaly records and in combination with machine learning techniques;

[0059] The construction information anomaly recognition model stores construction information anomaly features corresponding to multi-source data, and this construction information anomaly feature is used for the recognition of construction information anomalies in multi-source data;

[0060] The construction information anomaly recognition model also stores a construction information anomaly score calculation sub-model for calculating the construction information anomaly score and a construction information anomaly high-risk link capture sub-model for capturing high-risk links of construction information anomalies;

[0061] The construction information anomaly score calculation sub-model evaluates the comparison situation between multi-source data and construction information anomaly features at a time node based on a time series, and calculates the construction information anomaly score;

[0062] The construction information anomaly high-risk link capture sub-model continuously evaluates the comparison situation between multi-source data and construction information anomaly features under this time series based on the time series, and combines the construction information anomaly score to capture high-risk links of construction information anomalies.

[0063] It should be noted that the present embodiment uses existing machine learning techniques to implement the construction of the construction information anomaly recognition model, the construction information anomaly score calculation sub-model, and the construction information anomaly high-risk link capture sub-model. Exemplarily, such as a deep learning model. The construction information anomaly features in the present embodiment are the abnormal features of construction information known to those skilled in the art. Exemplarily, such as abnormal equipment operation, abnormal material use, and abnormal personnel behavior (such as violent behavior, etc.), and the construction information anomaly features in the present embodiment can be obtained based on an open-source construction information feature database, and the present embodiment does not limit this here.

[0064] Through the above, the present embodiment realizes the early warning of construction information anomaly recognition in two dimensions of the whole and the individual based on the construction information anomaly score calculation sub-model and the construction information anomaly high-risk link capture sub-model, thereby realizing the supervision of the overall construction information anomaly of the construction site and the continuous monitoring of the individual construction information, providing a data basis and technical guarantee for the early warning of construction information anomalies.

[0065] Specifically, the operation of the construction information anomaly score calculation sub-model includes the following steps:

[0066] A1: Obtain multi-source data within a construction site at a time node of a time series;

[0067] A2: Extract features from the multi-source data obtained in A1 to obtain corresponding real-time data features;

[0068] A3: Compare the similarity between the real-time data features and the abnormal features of construction information, and calculate the abnormal score of construction information based on the result of the similarity comparison.

[0069] It can be understood that in an actual construction site, most construction information is not generated in a separate form, and the occurrence of abnormal construction information is not without precursors. The purpose of designing the sub-model for calculating the abnormal score of construction information in this embodiment is to comprehensively evaluate the situation of the overall multi-source data of a construction site at a time node of a time series through the index of the abnormal score of construction information, and to judge whether the overall situation is likely to lead to the occurrence of abnormal construction information, that is, to judge whether there are precursors to the occurrence of abnormal construction information.

[0070] Specifically, the calculation of the abnormal score of construction information is specifically as follows:

[0071] Count the number of real-time data features, compare the real-time data features with the abnormal features of construction information respectively, and calculate the abnormal score of construction information based on the obtained similarity, the number of real-time data features and this similarity.

[0072] As a preferred implementation manner of this embodiment, the abnormal score of construction information is calculated using the formula for calculating the abnormal score of construction information. Among them, the formula for calculating the abnormal score of construction information is specifically as follows:

[0073]

[0074] Among them, m is the number of real-time data features, SR (i,τ) is the similarity between the real-time data feature i and the corresponding abnormal feature τ of construction information, VS total is the calculated abnormal score of construction information, and this embodiment uses any existing similarity analysis method to achieve the acquisition of SR (i,τ) The abnormal score threshold of construction information in this embodiment is an empirical value set based on the common knowledge known to those skilled in the art. When the calculated abnormal score of construction information is greater than or equal to the abnormal score threshold of construction information, it is determined that the overall data situation is likely to lead to the occurrence of abnormal construction information, that is, there are precursors to the occurrence of abnormal construction information.

[0075] Specifically, the operation of the sub-model for capturing high-risk links of abnormal construction information includes the following steps:

[0076] B1: Continuously obtain multi-source data based on the time series;

[0077] B2: Extract features from the multi-source data obtained by B1 to obtain corresponding continuous data features;

[0078] B3: Based on the time series, compare the similarity between the continuous data features and the abnormal construction information features to obtain the corresponding relationship between the similarity and the time series. Based on this corresponding relationship, judge the change trend of the similarity. When the change trend of the similarity does not meet the preset change threshold, capture the construction link corresponding to the change trend of the similarity and define it as the high-risk link of abnormal construction information.

[0079] Specifically, the corresponding relationship between the similarity and the time series is specifically:

[0080] Calculate the integral of the similarity between the continuous data features and the abnormal construction information features during the supervision period, and calculate the maximum and minimum values of this integral within the supervision duration. Based on the maximum value and the minimum value, calculate the corresponding relationship between the similarity and the time series.

[0081] It can be understood that in the actual construction process, the generation of abnormal construction information is a process, and based on the analysis of historical abnormal construction information, during the generation process of abnormal construction information, there are fluctuations (that is, the similarity values between the characteristics of individual data and the abnormal construction information characteristics fluctuate). The trend corresponding to this fluctuation reflects that there may be problems in the construction, and further reflects the precursor before the occurrence of abnormal construction information. Therefore, in this embodiment, a sub-model for capturing high-risk links of abnormal construction information is designed to realize continuous analysis of individual data in the construction, so as to capture this fluctuation (that is, capture the high-risk links of abnormal construction information).

[0082] As a preferred implementation manner of this embodiment, use the formula for capturing high-risk links of abnormal construction information to capture high-risk links of abnormal construction information. Among them, the formula for capturing high-risk links of abnormal construction information is specifically:

[0083]

[0084] Among them, t n ∈T and T = [t1, t2, t3,... t n , T is the preset supervision duration, divide T into n equal time periods, t n is the nth time period within T, represents the integral of the similarity between the real-time data feature i and the corresponding abnormal construction information feature τ in the nth time period within T, max is the maximum value calculation operator, min is the minimum value calculation operator, VS single_iFor the fluctuation of the calculated real-time data feature i, when the fluctuation is greater than or equal to the change threshold (the change threshold is an empirical value set based on the common knowledge known to those skilled in the art), it is determined that there is a fluctuation in the real-time data feature i, that is, there is a precursor before the occurrence of abnormal construction information, and the real-time data feature i is defined as a high-risk link for abnormal construction information.

[0085] Specifically, the operation of the high-risk link capture sub-model for abnormal construction information further includes the following steps:

[0086] When the abnormal construction information score does not meet the abnormal construction information score threshold, update B3 to the following steps:

[0087] B31: Based on the time series, compare the similarity between the continuous data feature and the abnormal construction information feature to obtain the corresponding relationship between the similarity and the time series; obtain the abnormal construction information score, and correct the corresponding relationship between the similarity and the time series based on the abnormal construction information score to obtain the corrected corresponding relationship. Based on the corrected corresponding relationship, judge the change trend of the similarity. When the change trend of the similarity does not meet the preset change threshold, capture the construction link corresponding to the change trend of the similarity and define it as a high-risk link for abnormal construction information.

[0088] Based on B3, it can be understood that compared with the construction sites that meet the abnormal construction information score threshold, the probability of abnormal construction information occurring at the construction sites that do not meet the abnormal construction information score threshold will increase significantly. Therefore, this embodiment designs B31 to capture the high-risk link for abnormal construction information that comprehensively considers the abnormal construction information score. Specifically, the high-risk link for abnormal construction information is captured urgently using the emergency high-risk link capture formula for abnormal construction information. Among them, the emergency high-risk link capture formula for abnormal construction information is specifically:

[0089]

[0090] Among them, VS total_τ is the aforementioned abnormal construction information score threshold. Based on the emergency high-risk link capture formula for abnormal construction information, when the abnormal construction information score does not meet the abnormal construction information score threshold, use the operator to amplify VS in B3 single_i , calculate to obtain VS single_i_E . When the fluctuation is greater than or equal to the change threshold, it is determined that there is a fluctuation in the real-time data feature i, that is, there is a precursor before the occurrence of abnormal construction information, and the real-time data feature i is defined as a high-risk link for abnormal construction information, so as to realize the capture of the high-risk link for abnormal construction information that comprehensively considers the abnormal construction information score.

[0091] Specifically, the countermeasures at least include the construction adjustment plan, the emergency resource allocation process, and the evacuation plan for the on-site construction personnel.

[0092] It should be noted that based on the historical processing method of abnormal construction information in this embodiment, corresponding countermeasures are refined, and a corresponding early warning plan is designed based on the actual construction situation, so as to output corresponding response methods for different situations of abnormal construction information.

[0093] Through the above, this embodiment realizes the targeted processing of abnormal construction information, effectively reduces the possibility of occurrence of high-risk links of abnormal construction information in the construction site, improves the construction management level, and ensures the safety, efficiency and quality of construction.

[0094] In summary, the intelligent construction information anomaly recognition method of this embodiment uses the information acquisition module to realize the acquisition of multi-source data related to construction information, uses the evaluation module to realize the information anomaly recognition in both the overall and individual dimensions, and corrects the recognition of individual construction links by using the overall information anomaly. The early warning module realizes the visual display of the evaluation results of the evaluation module, and provides a comprehensive response method for dealing with abnormal construction information, so as to realize the comprehensive calculation of the abnormal construction information score of the construction site to overall grasp the situation of abnormal construction information in the site, and at the same time can accurately capture the high-risk links of abnormal construction information to clarify the specific construction links of individual anomalies, which helps the construction administrator to know the potential risks in advance, take corresponding measures, effectively reduce the possibility of occurrence of high-risk links of abnormal construction information in the construction site, improve the construction management level, and ensure the safety, efficiency and quality of construction.

[0095] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. The computer-readable storage medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transmission of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. The computer-readable storage medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM, or other optical disk storage, magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0096] Finally, it should be noted that the above are only the preferred embodiments of the present application and are not used to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for identifying abnormalities in intelligent construction information, characterized in that: The method comprises an information collection step, an evaluation step and an early warning step which are performed in sequence; The information collection step is configured to collect multi-source data related to construction information; wherein the multi-source data at least includes construction equipment operation data, construction material usage data, construction personnel operation behavior data and construction environment data; The evaluation step is configured as follows: based on a preset construction information anomaly identification model, construction information anomaly identification is performed in combination with the multi-source data; wherein the construction information anomaly identification model is used to calculate the construction information anomaly score of the construction site and / or capture the high-risk links of construction information anomaly, the construction information anomaly score is used to characterize the abnormal situation of the overall construction information at the construction site, and the high-risk links of construction information anomaly are used to characterize the specific construction links at the construction site that may have abnormalities; When the construction information anomaly score obtained by the evaluation step does not meet a preset construction information anomaly score threshold and / or a high-risk link of construction information anomaly is identified, the early warning step issues an early warning reminder to the construction manager and provides countermeasures.

2. The method for identifying abnormality of intelligent construction information according to claim 1, characterized in that: The information collection step includes an equipment data collection sub-step, which is used to collect operation data of the construction equipment based on preset sensors; wherein the sensors are arranged on the construction equipment.

3. The method for identifying abnormality of intelligent construction information according to claim 1, characterized in that: The information collection step also includes a material data collection sub-step, which is used to collect construction material usage data from a preset material statistics terminal; wherein the material statistics terminal is used to record the usage of construction materials.

4. The method for identifying abnormality of intelligent construction information according to claim 1, characterized in that: The construction information anomaly recognition model is based on historical construction information anomaly records and is trained in combination with machine learning technology; The construction information anomaly recognition model stores construction information anomaly features corresponding to the multi-source data, and the construction information anomaly features are used to recognize construction information anomalies in the multi-source data; The construction information anomaly identification model also stores a construction information anomaly score calculation sub-model for calculating the construction information anomaly score and a construction information anomaly high-risk link capture sub-model for capturing the construction information anomaly high-risk link; The construction information anomaly score calculation submodel is based on a time node of the time series, evaluates the comparison between the multi-source data and the construction information anomaly characteristics at the time node, and calculates the construction information anomaly score; The abnormal high-risk link capture sub-model of construction information is based on time series, continuously evaluates the comparison between multi-source data and the abnormal characteristics of construction information under the time series, and captures the abnormal high-risk link of construction information in combination with the abnormal score of construction information.

5. The method for identifying abnormality of intelligent construction information according to claim 4, characterized in that: The operation of the construction information anomaly score calculation sub-model includes the following steps: A1: Obtain multi-source data at a construction site at a time node in the time series; A2: Extract features from the multi-source data obtained in A1 to obtain corresponding real-time data features; A3: Performing a similarity comparison between the real-time data feature and the abnormal construction information feature, and calculating the abnormal construction information score based on the result of the similarity comparison.

6. The method for identifying abnormality in intelligent construction information according to claim 5, characterized in that: The calculation of the construction information abnormality score is specifically as follows: The number of the real-time data features is counted, and the real-time data features are respectively compared with the construction information abnormal features for similarity, and based on the obtained similarity, the construction information abnormality score is calculated based on the number of the real-time data features and the similarity.

7. The method for identifying abnormality of intelligent construction information according to claim 4, characterized in that: The operation of the sub-model for capturing abnormal high-risk links of construction information includes the following steps: B1: Based on time series, continuously obtain multi-source data; B2: Extract features from the multi-source data obtained in B1 to obtain corresponding continuous data features; B3: Based on the time series, the continuous data features are compared with the abnormal construction information features to obtain the corresponding relationship between the similarity and the time series, and the similarity change trend is judged based on the corresponding relationship. When the similarity change trend does not meet the preset change threshold, the construction link corresponding to the similarity change trend is captured and defined as a high-risk link for construction information abnormality.

8. The method for identifying abnormality in intelligent construction information according to claim 7, characterized in that: The corresponding relationship between the similarity and the time series is specifically: Calculate the integral of the similarity between the continuous data features and the abnormal construction information features during the supervision period, and calculate the maximum and minimum values ​​of the integral within the supervision time, and calculate the corresponding relationship between the similarity and the time series based on the maximum and minimum values; wherein the supervision time is greater than the supervision period.

9. The method for identifying abnormality in intelligent construction information according to claim 7, characterized in that: The operation of the sub-model for capturing abnormal high-risk links of construction information also includes the following steps: When the construction information anomaly score does not meet the construction information anomaly score threshold, update B3 to the following steps: B31: Based on the time series, the continuous data features are compared with the abnormal features of the construction information to obtain the corresponding relationship between the similarity and the time series; the construction information abnormality score is obtained, and the corresponding relationship between the similarity and the time series is corrected based on the construction information abnormality score to obtain the corrected corresponding relationship, and the similarity change trend is judged based on the corrected corresponding relationship. When the similarity change trend does not meet the preset change threshold, the construction link corresponding to the similarity change trend is captured and defined as a high-risk link for construction information abnormality.

10. The method for identifying abnormality in intelligent construction information according to claim 1, characterized in that: The response measures shall at least include a construction adjustment plan, an emergency resource deployment process and a construction site personnel evacuation plan.

Images