BIM-based dynamic pre-warning method for safety risk of deep foundation pit construction

Abstract

The application discloses a deep foundation pit construction safety risk dynamic early warning method based on BIM. The method relates to the technical field of building construction safety and comprises the following steps: data acquisition, model updating, early warning research and judgment, and effect evaluation. The matching degree is dynamically updated through the matching calculation of real-time monitoring data and the BIM model, and whether the optimization mechanism comprising response time and push time calibration is started is judged based on the matching degree. Then, the research and judgment efficiency is obtained through the early warning research and judgment process, whether the intelligent optimization mechanism comprising the dynamic calibration of the advance time threshold value and feedback iteration is started is judged based on the parameter, the double closed-loop optimization from the model updating to the early warning logic is realized, the deep foundation pit construction safety risk dynamic early warning reliability is improved, and the problem of low deep foundation pit construction safety risk dynamic early warning reliability in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of construction safety, and particularly relates to a deep foundation pit construction safety risk dynamic early warning method based on BIM. BACKGROUND

[0002] A three-dimensional visualization model containing geological survey data, supporting structure parameters, surrounding building and underground pipeline information is constructed through BIM (Building Information Modeling) technology, GIS (Geographic Information System) geographic information system and rock-soil engineering numerical simulation software (such as FLAC3D) are integrated into the model preprocessing stage to complete the preset and parameterization definition of risk sources (such as slope instability, supporting structure deformation and foundation pit water gushing);

[0003] During construction, real-time data such as supporting structure stress and strain, foundation pit settlement displacement and underground water level change are collected through sensors (such as strain gauges, inclinometers and osmotic pressure gauges) and technologies such as unmanned aerial vehicle oblique photography and laser scanning, and are transmitted to the BIM collaborative management platform through the Internet of Things. The platform combines preset thresholds and machine learning algorithms such as BP (Back Propagation Neural Network) neural networks to dynamically analyze the data, and synchronously calls numerical simulation tools to simulate the risk evolution trend. When the data exceeds the safety threshold or the simulated risk level rises, a hierarchical early warning (red / orange / yellow) is automatically triggered, the specific location of the risk is located through the BIM model and is visually presented, and the mobile terminal APP is linked to push early warning information and disposal suggestions to the management personnel. The management personnel make decisions in combination with the visual decision support of the BIM model, call the construction monitoring system to adjust the construction parameters (such as slowing down the excavation speed and adding temporary support), and record the adjusted construction data in the BIM model in reverse to update it.

[0004] The above technology at least has the following technical problems:

[0005] In the application scenario of deep foundation pit construction, the BIM model is mostly a static design model created based on design drawings. When dynamic changes such as over-excavation of earthwork and lagging of support installation occur on site, model updating completely depends on manual measurement and manual modification, which is tedious, time-consuming and costly, and the model is in a "outdated" state for a long time. Even if the geometric form is updated, the internal mechanical parameters (such as soil stress and supporting structure stress state) cannot be adjusted synchronously, which makes the safety analysis lose the accurate physical basis, and there is a contradiction between the model static and the site dynamic.

[0006] The core logic of the existing system still stays at the level of simple single-point threshold comparison, cannot analyze the spatio-temporal correlation and causal chain among multiple monitoring points and multiple types of data (such as displacement, axial force and water level), ignores dynamic trend precursors such as change rate, and can only perform post-alarming for the occurred over-limit, lacks prediction capability, the early warning logic is shallow and static, lacks systematic intelligent research and judgment, after the early warning is triggered, information is presented in the form of fragmented text list, lacks intuitive visual correlation with the spatial position, influence range and surrounding environment of the BIM model, the management personnel need to manually switch between the model, drawings and reports to locate the problem, leading to delayed emergency response, cannot automatically correlate the knowledge base to provide cause analysis and disposal scheme, the interface is rigid and cannot individualize the information display according to the roles, lacks bidirectional interaction function, and has the problem of low reliability of dynamic early warning of deep foundation pit construction safety risk. SUMMARY

[0007] In order to solve the technical problem of low reliability of dynamic early warning of deep foundation pit construction safety risk existing in the prior art, the embodiment of the present application provides a deep foundation pit construction safety risk dynamic early warning method based on BIM. The technical solution is as follows:

[0008] On the one hand, a deep foundation pit construction safety risk dynamic early warning method based on BIM is provided, which comprises: acquiring deep foundation pit construction monitoring data transmitted back from the deep foundation pit construction site in real time to match and update into the constructed BIM deep foundation pit model, obtaining an updated matching degree parameter, obtaining a BIM model dynamic update matching degree based on the updated matching degree parameter for quantifying the consistency degree of the BIM deep foundation pit model and the real-time working condition of the construction site in the deep foundation pit construction process; judging whether to perform dynamic update matching degree optimization based on the BIM model dynamic update matching degree, if yes, performing a deep foundation pit construction safety risk early warning process after the dynamic update matching degree optimization, if not, directly performing the deep foundation pit construction safety risk early warning process, and acquiring a research and judgment efficiency parameter in the deep foundation pit construction safety risk early warning process, the dynamic update matching degree optimization comprises a response time dynamic calibration mechanism and a push time dynamic calibration mechanism; obtaining an early warning logic intelligent research and judgment efficiency based on the research and judgment efficiency parameter for quantifying the accuracy of the BIM deep foundation pit model in construction safety risk identification, judging whether to perform intelligent research and judgment efficiency optimization based on the early warning logic intelligent research and judgment efficiency, if yes, performing deep foundation pit construction safety risk early warning verification after the intelligent research and judgment efficiency optimization, if not, directly performing the deep foundation pit construction safety risk early warning verification, the intelligent research and judgment efficiency optimization comprises an advance time threshold dynamic calibration mechanism and an early warning effect feedback iteration mechanism.

[0009] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0010] 1. Through the dynamic updating matching degree of BIM model, the intelligent research and judgment effective rate double core parameter dynamic linkage is constructed, combined with the updating matching degree optimization-intelligent research and judgment effective rate optimization two-level optimization mechanism, the complete closed loop of data updating-risk early warning-parameter optimization-verification iteration is constructed, and the deep foundation pit construction safety risk dynamic early warning reliability is improved.

[0011] 2. Through multi-dimensional calculation of BIM model dynamic updating matching degree, the single-point deviation of foundation pit, response delay, updating cycle are fused, and the influence value coupling processing of spatial accuracy, time response and updating frequency is carried out, so as to ensure the consistency of model and field working condition, and based on BIM model dynamic updating matching degree, construction working condition-risk analysis response time difference, out-of-date model safety misjudgment frequency coupling calculation early warning logic intelligent research and judgment effective rate, combined with advance time threshold dynamic calibration, according to early warning information space visual matching rate adjustment threshold attenuation / gain factor and early warning effect feedback iteration, the risk early warning accuracy is improved, the data distortion and research deviation problem is solved, and the deep foundation pit construction safety risk dynamic early warning reliability is improved.

[0012] 3. According to the suddenness of deep foundation pit risk, by constructing the response time and push time double dynamic calibration mechanism, the response time calibration is carried out by collecting frequency adaptation, the data collection-model updating delay is compressed, the push time calibration is carried out by measuring point layout density adaptation, the risk expansion probability is significantly reduced, the early warning response timeliness is fully guaranteed, the key window for construction safety risk emergency disposal is strived for, and the deep foundation pit construction safety risk dynamic early warning reliability is improved. BRIEF DESCRIPTION OF DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0014] Figure 1 It is the BIM-based deep foundation pit construction safety risk dynamic early warning method flow chart provided by the embodiment of the present application;

[0015] Figure 2 It is the push time dynamic calibration mechanism flow chart of the BIM-based deep foundation pit construction safety risk dynamic early warning method provided by the embodiment of the present application;

[0016] Figure 3 It is the advance time threshold dynamic calibration mechanism flow chart of the BIM-based deep foundation pit construction safety risk dynamic early warning method provided by the embodiment of the present application. DETAILED DESCRIPTION

[0017] The technical solutions in the present application will be described below with reference to the drawings.

[0018] In the embodiments of the present application, the words such as "for example", "for instance", etc. are used to represent an example, illustration, or description. Any embodiment or design scheme described as "for example" in the present application should not be interpreted as more excellent or more advantageous than other embodiments or design schemes. Rather, the word "for example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0019] The embodiments of the present application provide a BIM-based deep foundation pit construction safety risk dynamic early warning method, as shown in a BIM-based deep foundation pit construction safety risk dynamic early warning method flow chart. Figure 1 The processing flow of the method can include the following steps:

[0020] The deep foundation pit construction monitoring data transmitted back from the deep foundation pit construction site in real time is acquired to be matched and updated to the constructed BIM deep foundation pit model, to obtain an updated matching degree parameter, and the BIM model dynamic update matching degree based on the updated matching degree parameter is obtained to quantify the consistency degree of the BIM deep foundation pit model and the real-time working condition of the construction site in the deep foundation pit construction process; it is judged whether to execute dynamic update matching degree optimization based on the BIM model dynamic update matching degree, if yes, after the dynamic update matching degree optimization, the deep foundation pit construction safety risk early warning process is performed, if not, the deep foundation pit construction safety risk early warning process is directly performed, and the research and judgment efficiency parameter in the deep foundation pit construction safety risk early warning process is acquired, the dynamic update matching degree optimization includes a response time dynamic calibration mechanism and a push time dynamic calibration mechanism; the early warning logic intelligent research and judgment efficiency based on the research and judgment efficiency parameter is obtained to quantify the accuracy of the BIM deep foundation pit model in the construction safety risk identification, it is judged whether to execute intelligent research and judgment efficiency optimization based on the early warning logic intelligent research and judgment efficiency, if yes, after the intelligent research and judgment efficiency optimization, the deep foundation pit construction safety risk early warning verification is performed, if not, the deep foundation pit construction safety risk early warning verification is directly performed, and the intelligent research and judgment efficiency optimization includes an advance time threshold dynamic calibration mechanism and an early warning effect feedback iteration mechanism.

[0021] In this embodiment, by constructing a three-level closed-loop optimization system of "model updating - early warning judgment - effect verification", the safety risk early warning of deep foundation pit is realized from passive response to active pre-control. First, based on the dynamic matching degree evaluation of real-time monitoring data and BIM model, a quantitative guarantee mechanism of the consistency of model and site working conditions is established. When the matching degree is insufficient, the double optimization including response time calibration and push time calibration is automatically triggered to ensure the transmission efficiency from monitoring data to early warning decision. Secondly, through the intelligent judgment of early warning logic, the self-improvement mechanism of early warning accuracy is constructed. When the judgment efficiency is lower than the benchmark value, the dynamic calibration of early warning effect feedback iteration and the advance time threshold are automatically started to reduce the false positive rate and the false negative rate tends to zero. Finally, an intelligent early warning system with self-learning ability is formed, which significantly improves the early warning accuracy and timeliness while ensuring the real-time of model data, and provides whole-process and self-adaptive high-precision risk pre-control guarantee for deep foundation pit construction safety.

[0022] It needs to be explained that before obtaining the dynamic updating matching degree of BIM model, it also includes:

[0023] The updating matching degree parameters include the maximum deviation of single point of foundation pit, dynamic construction working condition response delay and BIM updating average period. Among them, the maximum deviation of single point of foundation pit refers to the maximum value of the absolute value of the deviation between the measured single point position (coordinate / elevation) of the foundation pit on site and the position benchmark value of the corresponding single point of the foundation pit in the BIM model; the single point coordinate of the monitoring foundation pit is measured automatically and compared with the position benchmark value of the single point of the monitoring foundation pit in the BIM model in real time to obtain the maximum deviation of single point of foundation pit; the dynamic construction working condition response delay refers to the total time difference between the actual completion of a construction working condition (such as completing one layer of earth excavation, support installation in place, and dewatering well starting operation, etc.) on the construction site and the identification of the change and the issuance of the corresponding early warning information; the construction working condition change time stamp is obtained through the electronic signing of the construction log, and the time stamp of the early warning information generation is recorded through the log database record, and the deviation between the time stamp of the early warning information generation and the construction working condition change time stamp is recorded as the dynamic construction working condition response delay; the BIM updating average period refers to the average time interval between two consecutive effective BIM model version updates; the model file version history is recorded through the BIM management platform, including the creation time stamp of each new version, all model version records in a preset time period are extracted, the time difference between adjacent versions is calculated, and then the average value is obtained to get the BIM updating average period.

[0024] The bias correction factor is interactively processed with the spatial accuracy component to obtain a spatial accuracy influence value, the spatial accuracy component representing the analysis results of the single-point maximum deviation threshold and the single-point maximum deviation proportion of the foundation pit; the response delay correction factor is interactively processed with the time response component to obtain a time response influence value, the time response influence value representing the analysis results of the response delay threshold and the response delay proportion of the dynamic construction condition; the average period correction factor is interactively processed with the update frequency component to obtain an update frequency influence value, the update frequency component representing the analysis results of the average period threshold and the average period proportion of the BIM update; the spatial scale influence value, the time response influence value and the update frequency influence value are coupled to obtain the dynamic update matching degree of the BIM model. The interactive processing refers to multiplication operation, the proportion analysis refers to division operation, and the limit expression of the dynamic update matching degree of the BIM model is:

[0025] ;

[0026] In the formula, E represents the dynamic update matching degree of the BIM model; w1 represents the bias correction factor obtained from the dynamic early warning database; w2 represents the response delay correction factor obtained from the dynamic early warning database; w3 represents the average period correction factor obtained from the dynamic early warning database; Q0 represents the single-point maximum deviation threshold obtained from the dynamic early warning database; R0 represents the response delay threshold obtained from the dynamic early warning database; P0 represents the average period threshold obtained from the dynamic early warning database; Q represents the single-point maximum deviation of the foundation pit; R represents the response delay of the dynamic construction condition; and P represents the average period of the BIM update.

[0027] It needs to be understood that the longer the average period of BIM update, the longer the BIM model is in an outdated state and cannot reflect the dynamic changes on site, and the larger the maximum deviation of the single point of the foundation pit; the longer the average period of BIM update, the more outdated the data basis for risk analysis itself, even if the early warning algorithm is advanced, it cannot make timely and correct judgments based on outdated data, and the dynamic construction condition response delay is higher; the larger the maximum deviation of the single point of the foundation pit, the more difficult it is to identify the current real risk based on the old BIM model, so as to issue a warning, resulting in an increase in the dynamic construction condition response delay. At the same time, the maximum deviation of the single point of the foundation pit and the dynamic update matching degree of the BIM model have a negative correlation, the larger the maximum deviation of the single point of the foundation pit, the more serious the disconnection between the model and the physical form of reality in a local position of the deep foundation pit, and the smaller the dynamic update matching degree of the BIM model; the dynamic construction condition response delay and the dynamic update matching degree of the BIM model have a negative correlation, the larger the dynamic construction condition response delay, the larger the time gap between the change on site and the system identifying the risk, and the smaller the dynamic update matching degree of the BIM model; the average period of BIM update and the dynamic update matching degree of the BIM model have a negative correlation, the longer the average period of BIM update, the longer the model is in a static outdated state and cannot keep up with the dynamic changes on the construction site, and the smaller the dynamic update matching degree of the BIM model.

[0028] Further, the specific steps of judging whether to perform dynamic update matching degree optimization are: if the dynamic update matching degree of the BIM model is higher than or equal to the matching degree reference value, dynamic update matching degree optimization is not performed, otherwise, if the dynamic update matching degree of the BIM model is not higher than the matching degree reference value, based on the matching degree offset, the response time dynamic calibration mechanism and the push time dynamic calibration mechanism are performed, and the matching degree offset represents the negative difference between the dynamic update matching degree of the BIM model and the matching degree reference value.

[0029] It needs to be further explained that the specific process of performing the response time dynamic calibration mechanism is:

[0030] If the deep foundation pit data collection frequency is greater than the upper limit of the frequency reference, it is determined that it is in the super-frequency collection-emergency response state, and the batch deep foundation pit data processing mode is switched to the streaming deep foundation pit data processing mode. Specifically, based on the matching degree offset and the collection frequency offset, the streaming deep foundation pit data processing window length reference value and the construction condition recognition result confidence threshold are obtained by querying the pre-defined collection frequency-identification time mapping table. The streaming deep foundation pit data processing window length reference value represents the length of the data time window relied on by the system for single condition recognition calculation in the streaming processing mode. This value is directly obtained from the mapping table as a reference for subsequent adjustment. The construction condition recognition result confidence threshold represents the minimum confidence required by the system to determine the validity of a single condition recognition result in the streaming processing mode. This value is directly obtained from the mapping table as a basis for judgment. The collection frequency offset represents the positive difference between the deep foundation pit data collection frequency and the upper limit of the frequency reference. This ensures that the optimal processing parameters can be quickly established in an emergency state. The setting of the processing window length reference value shortens the data time window for single recognition calculation while ensuring data integrity, and the dynamic adjustment of the confidence threshold provides quantitative guarantee for the accuracy of risk identification.

[0031] If the streaming deep foundation pit data processing window length reference value is greater than the window length reference value, the window length offset and the construction condition recognition result confidence threshold are input into the collection frequency-identification time mapping table to obtain the identification response time decay factor. The streaming deep foundation pit data processing window length reference value and the identification response time decay factor are interactively processed to obtain the construction condition recognition response time target value. The window length offset represents the positive difference between the streaming deep foundation pit data processing window length reference value and the window length reference value. By dynamically adjusting the response time of the identification algorithm, the identification response time of high-risk conditions is compressed under the premise of ensuring identification reliability, which saves valuable time window for emergency disposal of deep foundation pit construction sites.

[0032] In the embodiment, when the deep foundation pit construction monitoring enters the emergency response state of high-frequency collection, the synergistic optimization of "data processing efficiency-risk identification accuracy-response time efficiency" in the emergency scenario is realized through "precise switching from batch processing mode to streaming processing mode" and "response time dynamic calibration based on multi-parameter linkage". On the one hand, by linking the matching degree offset and the collection frequency offset to query the pre-defined mapping table, the streaming processing window length reference value and the construction condition identification result confidence threshold value are accurately obtained, which not only avoids the data accumulation problem caused by the "fixed processing mode cannot adapt to emergency high-frequency data" in the prior art (such as analysis delay of high-frequency data for more than 15 minutes in batch processing mode), but also ensures the effectiveness of the identification result in streaming processing through the confidence threshold constraint. On the other hand, by judging the size relationship between the window length reference value and the reference value, the decay factor is linked with the window length offset and the confidence threshold to query, and then the response time target value is obtained by the interaction between the window length reference value and the decay factor, which can shorten the construction condition identification response time in the emergency scenario, and finally realize the technical goal of "real-time data processing without accumulation, accurate risk identification without deviation, and response time efficiency adapting to emergency demand" in the emergency state, greatly improving the reliability and timeliness of deep foundation pit emergency risk control.

[0033] It should be further pointed out that the response time dynamic calibration mechanism also includes:

[0034] If not, the window length correction amount and the construction condition identification result confidence threshold value are input into the collection frequency-identification time mapping table for querying to obtain an identification response time gain factor, the streaming deep foundation pit data processing window length reference value and the identification response time gain factor are interactively processed to obtain a construction condition identification response time target value, and the window length correction amount represents the negative difference between the streaming deep foundation pit data processing window length reference value and the window length reference value. The identification reliability problem caused by insufficient data samples in a short time window is effectively solved, and the response time is controlled within a reasonable range while ensuring identification accuracy through dynamic compensation of the gain factor.

[0035] If the deep foundation pit data collection frequency is within the frequency reference interval, the response time dynamic calibration mechanism is not executed, and the frequency reference interval represents a closed interval formed by the frequency reference lower limit and the frequency reference upper limit. This avoids unnecessary parameter adjustment, ensures the operation stability of the system under normal conditions, and reduces the consumption of computing resources, so that the system can invest more resources into the core risk identification task.

[0036] If the deep foundation data collection frequency is less than the lower limit of the frequency reference, it is determined that it is in a low-frequency collection-delayed response state, and the batch deep foundation data processing mode is maintained. Specifically, based on the collection frequency correction amount and the matching degree offset amount, a batch processing interval reference value and a construction condition recognition result confidence threshold are obtained from a predefined collection frequency-recognition time mapping table. The batch processing interval reference value represents the length of the time interval between the start of two construction condition recognition calculations in the batch processing mode, and the batch processing interval reference value is set as the construction condition recognition response time target value. The collection frequency correction amount represents the negative difference between the deep foundation data collection frequency and the lower limit of the frequency reference. Through intelligent calibration of the batch processing interval, the construction condition recognition accuracy is still maintained in a low data density environment, the system energy consumption is reduced, and the best balance between resource utilization efficiency and risk monitoring effect is achieved.

[0037] In the embodiment, through the establishment of a complete adaptive response mechanism, intelligent processing of the deep foundation safety monitoring system in different data collection states is achieved. When the system is in the short window processing mode, through the cooperative query of the window length correction amount and the confidence threshold, the recognition response time gain factor is obtained, and coupled calculation is performed with the stream processing window length reference value, and finally the optimized response time target value is generated. This mechanism effectively solves the contradiction between data processing depth and response speed in the emergency state. Under the normal data collection frequency, the system maintains a stable running state, does not start the dynamic calibration mechanism, ensures the continuity of the monitoring process, and avoids unnecessary resource consumption. When entering the low-frequency collection state, the system intelligently switches to the batch processing mode, obtains the batch processing interval reference value based on the collection frequency correction amount and the matching degree offset amount, and directly sets it as the response time target value. This design enables the system to maintain reliable monitoring capability in a low data density environment. Through the three-state adaptive mechanism of "short window compensation-normal stability-low frequency optimization", intelligent regulation and control of the whole process from data collection to risk identification is achieved, which ensures the rapid response requirement in high-risk working conditions, takes into account the running efficiency in normal working conditions, and optimizes the resource allocation in low-risk stages, providing all-round and intelligent monitoring protection for deep foundation construction safety.

[0038] It should be understood that, as Figure 2It is shown that the push time dynamic calibration mechanism flow chart of the BIM-based deep foundation pit construction safety risk dynamic early warning method provided by the embodiment of the application, the specific logic is: if the deep foundation pit edge measurement point arrangement density is greater than the arrangement density reference upper limit, the ultra-fast push early warning information mode is carried out, based on the matching degree offset and the arrangement density offset, the early warning confidence threshold scaling factor is obtained from the pre-defined arrangement density-push time mapping table, the early warning confidence threshold scaling factor and the current early warning information push response time are interactively processed and the integral result is taken as the early warning information push response time target value, if the deep foundation pit edge measurement point arrangement density is within the arrangement density reference interval, the push time dynamic calibration mechanism is not executed, if the deep foundation pit edge measurement point arrangement density is less than the arrangement density reference lower limit, the delayed push early warning information mode is carried out, based on the matching degree offset and the arrangement density correction amount, the early warning confidence threshold multiplication factor is obtained from the pre-defined arrangement density-push time mapping table, the early warning confidence threshold multiplication factor and the current early warning information push response time are interactively processed and the integral result is taken as the early warning information push response time target value.

[0039] It needs to be further explained that the specific process of the push time dynamic calibration mechanism is:

[0040] If the deep foundation pit edge measurement point arrangement density is greater than the arrangement density reference upper limit, the ultra-fast push early warning information mode is carried out, specifically: based on the matching degree offset and the arrangement density offset, the early warning confidence threshold scaling factor is obtained from the pre-defined arrangement density-push time mapping table, the early warning confidence threshold scaling factor and the current early warning information push response time are interactively processed and the integral result is taken as the early warning information push response time target value, and the arrangement density offset is used to reflect the positive difference between the deep foundation pit edge measurement point arrangement density and the arrangement density reference upper limit. The data redundancy advantage of high-density measurement points is utilized, the confidence threshold is appropriately reduced, the early warning push response time is shortened under the premise of ensuring the accuracy, and valuable time is won for emergency disposal of high-risk working conditions.

[0041] The push time dynamic calibration mechanism also includes:

[0042] If the deep foundation pit edge measurement point arrangement density is within the arrangement density reference interval, the push time dynamic calibration mechanism is not executed, and the arrangement density reference interval represents a closed interval formed by the arrangement density reference lower limit and the arrangement density reference upper limit. The reliability of the early warning information push is ensured, unnecessary system adjustment is avoided, and the early warning system maintains the best operating state under normal working conditions.

[0043] If the measurement point density of the deep foundation pit edge is less than the lower limit of the reference density, a delayed push warning information mode is performed, specifically: based on the matching degree offset and the density correction amount, a pre-defined density-pushing time mapping table is queried to obtain a warning confidence threshold multiplication factor, the warning confidence threshold multiplication factor and the current warning information push response time are interactively processed and the result is rounded to obtain a warning information push response time target value, and the density correction amount is used to reflect the negative difference between the deep foundation pit edge measurement point density and the lower limit of the reference density. By increasing the confidence threshold and appropriately prolonging the analysis time, the defect of insufficient data support of low-density measurement points is effectively overcome, and the accuracy of the warning information is ensured by adding a data verification link.

[0044] In the embodiment, by establishing a three-level push calibration mechanism based on the measurement point density, the intelligent optimization of the response speed of the deep foundation pit safety warning system is realized under the premise of ensuring accuracy. When the measurement point density is greater than the upper limit of the reference, the system queries the warning confidence threshold scaling factor based on the matching degree offset and the density offset, and obtains the target value by interacting with the current push response time. This fast push mode fully utilizes the data redundancy advantage of high-density measurement points, reduces the warning push response time while maintaining accuracy. When the density is in the normal reference interval, the system maintains a stable operating state to avoid unnecessary parameter adjustment. When the density is lower than the lower limit of the reference, the system queries the warning confidence threshold multiplication factor based on the matching degree offset and the density correction amount, and obtains the target value by interacting with the current response time. This delayed push mode reduces the false positive rate in low-density areas by increasing the confidence threshold and prolonging the analysis time. This complete dynamic calibration mechanism enables the warning system to adapt to different monitoring environment characteristics, meets the rapid response needs of high-risk areas, and guarantees the warning reliability of data sparse areas, improves the overall warning accuracy, and forms a precise and efficient intelligent warning solution.

[0045] It should be noted that the research and judgment efficiency parameter includes the BIM model dynamic update matching degree, the construction working condition-risk analysis response time difference and the obsolete model safety misjudgment frequency. Among them, the construction working condition-risk analysis response time difference refers to the total time difference between the actual occurrence or completion of an important dynamic working condition (such as completing one layer of earth excavation, support installation in place, and dewatering well starting operation) in the construction site and the risk brought by the dynamic working condition change being identified by the BIM-based early warning and the corresponding early warning information being generated and sent; the difference value between the risk early warning time stamp and the construction working condition change time stamp is recorded as the construction working condition-risk analysis response time difference through the electronic record of the construction log / supervision log construction working condition change timestamp and the risk early warning timestamp recorded by the background log database of the early warning system platform; the ratio of the total number of misjudgment events confirmed within the preset statistical period to the preset statistical period is recorded as the obsolete model safety misjudgment frequency.

[0046] The matching degree correction factor and the update matching degree component are interactively processed to obtain an update matching degree influence value, and the update matching degree component represents the result of the BIM model dynamic update matching degree and the matching degree threshold proportion analysis; the response time difference correction factor and the response time difference component are interactively processed to obtain a response time difference influence value, and the response time difference component represents the result of the response time difference threshold and the construction working condition-risk analysis response time difference proportion analysis; the misjudgment frequency correction factor and the misjudgment frequency component are interactively processed to obtain a misjudgment frequency influence value, and the misjudgment frequency component represents the result of the misjudgment frequency threshold and the obsolete model safety misjudgment frequency proportion analysis; the update matching degree influence value, the response time difference influence value and the misjudgment frequency influence value are coupled to obtain the early warning logic intelligent research and judgment efficiency. The specific limit expression of the early warning logic intelligent research and judgment efficiency is:

[0047] ;

[0048] In the formula, N represents the early warning logic intelligent research and judgment efficiency; v1 represents the matching degree correction factor obtained from the dynamic early warning database; v2 represents the response time difference correction factor obtained from the dynamic early warning database; v3 represents the misjudgment frequency correction factor obtained from the dynamic early warning database; M0 represents the matching degree threshold value obtained from the dynamic early warning database; Y0 represents the response time difference threshold value obtained from the dynamic early warning database; H0 represents the misjudgment frequency threshold value obtained from the dynamic early warning database; M represents the BIM model dynamic update matching degree; Y represents the construction working condition-risk analysis response time difference; and H represents the obsolete model safety misjudgment frequency.

[0049] It should be understood that the higher the BIM model dynamic updating matching degree, the stronger the consistency of the model with the actual construction conditions on site, and the less time needed to correct the model deviation during risk analysis, and the smaller the delay of dynamic construction condition response. The higher the BIM model dynamic updating matching degree, the more real the model can reflect the safety hazards on site (such as missing edge protection and component collision risk), and the lower the frequency of safety misjudgment of the outdated model. The longer the construction condition-risk analysis response time difference, the greater the deviation between the model and the site, the longer the correction time, and the higher the frequency of safety misjudgment of the outdated model. Meanwhile, the BIM model dynamic updating matching degree and the intelligent research and judgment efficiency of the early warning logic have a positive correlation. The higher the BIM model dynamic updating matching degree, the more the early warning system can judge whether the template support is overloaded and whether the pipeline collision exists based on the real data, the higher the effective early warning proportion, and the greater the intelligent research and judgment efficiency of the early warning logic. The construction condition-risk analysis response time difference and the intelligent research and judgment efficiency of the early warning logic have a negative correlation. The greater the construction condition-risk analysis response time difference, the more the site conditions may have been advanced (such as the concrete has been poured and completed, and the template overload risk no longer exists), the less the early warning has guiding significance, and the smaller the intelligent research and judgment efficiency of the early warning logic. The higher the frequency of safety misjudgment of the outdated model, the more the early warning system will frequently appear "missed judgment" or "false report", for example, the model has not been updated to add the scaffold hidden danger (missed judgment), or the model shows that the temporary structure still has a risk after being removed (false report), and the smaller the intelligent research and judgment efficiency of the early warning logic.

[0050] Further, the specific steps of judging whether to perform intelligent research and judgment efficiency optimization are as follows: if the intelligent research and judgment efficiency of the early warning logic is greater than or equal to the efficiency reference value, intelligent research and judgment efficiency optimization is not performed, otherwise, if the intelligent research and judgment efficiency of the early warning logic is less than the efficiency reference value, based on the efficiency offset, the early warning effect feedback iteration mechanism and the early warning effect feedback iteration mechanism are performed, and the efficiency offset represents the negative difference between the intelligent research and judgment efficiency of the early warning logic and the efficiency reference value.

[0051] It should be noted that, as Figure 3It is shown that the early time threshold dynamic calibration mechanism flow chart of the BIM-based deep foundation pit construction safety risk dynamic early warning method provided by the embodiment of the application, the specific logic is: if the early warning information space visualization matching rate is higher than the matching rate reference upper limit, based on the effective rate offset and the matching rate offset, the matching rate-threshold adjustment mapping table is queried to obtain the early time threshold decay factor, the current early time threshold reference value and the early time threshold decay factor are interactively processed to obtain the target early warning average early time threshold, if the early warning information space visualization matching rate is within the matching rate reference interval, the early time threshold dynamic calibration mechanism is not executed, if the early warning information space visualization matching rate is lower than the matching rate reference lower limit, based on the effective rate offset and the matching rate correction amount, the matching rate-threshold adjustment mapping table is queried to obtain the early time threshold gain factor, the current early time threshold reference value and the early time threshold gain factor are interactively processed to obtain the target early warning average early time threshold.

[0052] It needs to be further explained that the specific process of executing the early time threshold dynamic calibration mechanism is:

[0053] If the early warning information space visualization matching rate is higher than the matching rate reference upper limit, based on the effective rate offset and the matching rate offset, the matching rate-threshold adjustment mapping table is queried to obtain the early time threshold decay factor, the current early time threshold reference value and the early time threshold decay factor are interactively processed to obtain the target early warning average early time threshold, and the matching rate offset is used to reflect the positive difference between the early warning information space visualization matching rate and the matching rate reference upper limit. By moderately reducing the early threshold, the system can issue an early warning at the risk germination stage, extend the average early warning time, and win a valuable time window for on-site risk disposal.

[0054] If the early warning information space visualization matching rate is within the matching rate reference interval, the early time threshold dynamic calibration mechanism is not executed, and the matching rate reference interval represents the closed interval formed by the matching rate reference lower limit and the matching rate reference upper limit. Avoid unnecessary parameter fluctuations and maintain the running stability of the early warning system.

[0055] If the matching rate of the early warning information space visualization is lower than the lower limit of the matching rate reference, a gain factor of the early warning time threshold is obtained from a predefined matching rate-threshold adjustment mapping table based on the effective rate offset and a matching rate correction amount, the current early warning time threshold reference value is interactively processed with the gain factor of the early warning time threshold to obtain a target early warning average early warning time threshold, and the matching rate correction amount is used to reflect the negative difference between the early warning information space visualization matching rate and the upper limit of the matching rate reference. By appropriately increasing the early warning threshold and introducing an additional verification link, unreliable early warning signals are effectively filtered out, the false positive rate of the system is reduced, and the reliability of the early warning decision is improved by prolonging the analysis time.

[0056] In the embodiment, by constructing a dynamic threshold adjustment mechanism based on the early warning information space visualization matching rate, an intelligent balance between the timeliness and accuracy of the deep foundation pit safety early warning system is achieved. When the matching rate is higher than the upper limit of the reference, a decay factor of the early warning time threshold is obtained from the effective rate offset and the matching rate offset, and a new target threshold is obtained by interactive processing with the reference threshold. This mechanism can timely warn at the risk of germination stage by appropriately reducing the early warning threshold, significantly prolonging the average early warning time, and winning the key time window for on-site risk disposal. When the matching rate is in the normal range, the system maintains stable parameters to ensure continuous reliable early warning performance. When the matching rate is lower than the lower limit of the reference, a gain factor is obtained from the effective rate offset and the matching rate correction amount, the early warning threshold is increased, and the verification link is introduced to effectively filter out unreliable early warning signals, and the decision reliability is improved by prolonging the analysis time. This complete dynamic calibration system enables the early warning system to intelligently adjust the early warning strategy according to the actual operation effect, forming an intelligent early warning solution with foresight and reliability.

[0057] It should be further pointed out that the specific process of the early warning effect feedback iteration mechanism is as follows:

[0058] If the construction safety risk early warning threshold is greater than the first risk threshold and less than the second risk threshold, it is determined that the construction safety state is medium risk, and a medium speed high quality iteration mode is performed, specifically: the early warning threshold offset and the effective rate offset are obtained from the predefined early warning threshold-iteration frequency mapping table to obtain an early warning threshold dynamic adjustment frequency coefficient, and the early warning threshold offset is used to reflect the negative difference between the construction safety risk early warning threshold and the first risk threshold.

[0059] If the early warning threshold dynamic adjustment frequency coefficient is greater than or equal to the frequency coefficient reference coefficient, the early warning threshold dynamic adjustment frequency coefficient is input to the predefined early warning threshold-iteration frequency mapping table for query to obtain an iteration period gain amount, and the result of coupling processing of the current early warning effect feedback iteration period and the iteration period gain amount is taken as an integer to obtain a target early warning effect feedback iteration period. First, by dynamically adjusting the iteration frequency, the system maintains the optimal parameter update speed in the medium risk state, avoiding both resource waste caused by excessive iteration and early warning lag caused by insufficient iteration. Second, the collaborative analysis based on the double offset ensures the scientificity of the iteration decision, enabling the system to accurately adjust according to the actual risk situation and the early warning effect. Finally, by coupling the gain amount and the iteration period, the traditional fixed cycle mode is improved to an adaptive cycle mode, improving the iteration efficiency of the system in the medium risk state while keeping the early warning accuracy stable.

[0060] In the embodiment, by establishing an intelligent iteration optimization mechanism in the medium risk state, the adaptive performance of the deep foundation pit safety early warning system is improved. When the construction safety risk early warning threshold is between the first and second risk thresholds, the system automatically enters the medium-speed high-quality iteration mode, and based on the early warning threshold offset and the efficiency offset, the early warning threshold dynamic adjustment frequency coefficient is obtained from the predefined early warning threshold-iteration frequency mapping table, accurately reflecting the state relationship between the current risk level and the system performance. When the coefficient reaches or exceeds the reference value, the system further queries to obtain the iteration period gain amount, and through coupling operation and integer processing of the current early warning effect feedback iteration period, the optimal target iteration period is generated. This innovative mechanism enables the system to automatically adjust the parameter update frequency according to the real-time risk situation, improves the iteration efficiency in the medium risk state, shortens the early warning response time, and reduces the false positive rate, avoiding both resource waste caused by excessive iteration and early warning lag caused by insufficient iteration, and achieving precise matching of the early warning system performance and the risk level, providing more intelligent and reliable technical support for deep foundation pit construction safety.

[0061] It should be further noted that the early warning effect feedback iteration mechanism further includes:

[0062] If the early warning threshold dynamic adjustment frequency coefficient is less than the frequency coefficient reference coefficient, the early warning threshold dynamic adjustment frequency coefficient is input to the predefined early warning threshold-iteration frequency mapping table for query to obtain an iteration period reduction amount, and the result of difference processing of the current early warning effect feedback iteration period and the iteration period reduction amount is taken as an integer to obtain a target early warning effect feedback iteration period. Under the premise of ensuring system stability, the iteration frequency in the low risk state is reduced, effectively saving system resources.

[0063] If the construction safety risk early warning threshold is less than or equal to the first risk threshold, it is determined that the construction safety state is low risk, and the early warning effect feedback iteration mechanism is not performed. Unnecessary parameter adjustment is avoided, and system resources can be concentrated on core monitoring tasks, while maintaining the operation stability of the early warning system.

[0064] If the construction safety risk early warning threshold is greater than or equal to the second risk threshold, it is determined that the construction safety state is high risk, and the preset relevant personnel are reminded to check the deep foundation pit construction site. Through the establishment of a "system early warning-artificial verification" linkage response system, the traditional passive response is changed to active prevention and control, the efficiency of on-site risk identification is improved, and the risk can be found and handled in time.

[0065] In the embodiment, by establishing a complete risk grading response system, the optimal allocation and precise prevention and control of deep foundation pit safety early warning system resources are realized. When the early warning threshold dynamic adjustment frequency coefficient is lower than the benchmark value, the system intelligently queries the iteration period reduction amount, and through difference processing and integer operation of the current iteration period, the iteration frequency in the low risk state is optimized to a reasonable interval, while ensuring the stability of the system, the operation and maintenance cost is significantly reduced. In the low risk construction safety state, the system suspends the early warning effect feedback iteration mechanism, avoids unnecessary parameter adjustment, and concentrates computing resources on core monitoring tasks. When entering the high risk state, the system immediately triggers the emergency response mechanism, automatically notifies the preset person in charge to carry out on-site verification, and forms a "system early warning-artificial intervention" rapid disposal closed loop. This three-level response mechanism adjusts the iteration frequency dynamically, ensures the timely response capability in high risk working conditions, and realizes resource saving in stable working conditions, so that the system reduces the overall operation and maintenance cost while maintaining the early warning accuracy, effectively improves the intelligent level and economic benefit of deep foundation pit safety monitoring.

[0066] The above describes only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A BIM-based deep foundation pit construction safety risk dynamic early warning method, characterized in that, Comprise the following steps: Obtain the deep foundation pit construction monitoring data transmitted back from the deep foundation pit construction site in real time to match and update to the built BIM deep foundation pit model, obtain the update matching degree parameter, and obtain the BIM model dynamic update matching degree based on the update matching degree parameter to quantify the consistency degree of the BIM deep foundation pit model and the real-time working condition of the construction site in the deep foundation pit construction process; Based on the BIM model dynamic update matching degree, it is judged whether to execute the dynamic update matching degree optimization, if yes, after the dynamic update matching degree optimization, the deep foundation pit construction safety risk early warning process is carried out, if not, the deep foundation pit construction safety risk early warning process is directly carried out, and the research and judgment efficiency parameter in the deep foundation pit construction safety risk early warning process is obtained, the dynamic update matching degree optimization includes response time dynamic calibration mechanism and push time dynamic calibration mechanism; Based on the research and judgment efficiency parameter, the early warning logic intelligent research and judgment efficiency is obtained to quantify the accuracy of the BIM deep foundation pit model in the construction safety risk identification, and it is judged whether to execute the intelligent research and judgment efficiency optimization based on the early warning logic intelligent research and judgment efficiency, if yes, after the intelligent research and judgment efficiency optimization, the deep foundation pit construction safety risk early warning verification is carried out, if not, the deep foundation pit construction safety risk early warning verification is directly carried out, the intelligent research and judgment efficiency optimization includes advance time threshold dynamic calibration mechanism and early warning effect feedback iteration mechanism.

2. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 1, characterized in that, Before obtaining the BIM model dynamic update matching degree, it further comprises: The update matching degree parameter comprises the maximum deviation of the foundation pit single point, the dynamic construction working condition response delay and the BIM update average period; The deviation correction factor and the space precision component are interactively processed to obtain the space precision influence value, and the space precision component represents the result of single point maximum deviation threshold and foundation pit single point maximum deviation ratio analysis; The response delay correction factor and the time response component are interactively processed to obtain the time response influence value, and the time response influence value represents the result of response delay threshold and dynamic construction working condition response delay ratio analysis; The average period correction factor and the update frequency component are interactively processed to obtain the update frequency influence value, and the update frequency component represents the result of average period threshold and BIM update average period ratio analysis; The space precision influence value, the time response influence value and the update frequency influence value are coupled to obtain the BIM model dynamic update matching degree; The specific steps of judging whether to execute the dynamic update matching degree optimization are: If the BIM model dynamic update matching degree is higher than or equal to the matching degree reference value, the dynamic update matching degree optimization is not executed, otherwise, if the BIM model dynamic update matching degree is not higher than the matching degree reference value, the response time dynamic calibration mechanism and the push time dynamic calibration mechanism are executed based on the matching degree offset, and the matching degree offset is used to reflect the negative deviation of the BIM model dynamic update matching degree and the matching degree reference value.

3. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 2, characterized in that, The specific process of executing the response time dynamic calibration mechanism is: If the deep foundation data collection frequency is greater than the upper limit of the frequency reference, it is determined that it is in the super-frequency collection-emergency response state, and the batch deep foundation data processing mode is switched to the streaming deep foundation data processing mode. Specifically, based on the matching degree offset and the collection frequency offset, the streaming deep foundation data processing window duration reference value and the construction condition recognition result confidence threshold are obtained by querying the pre-defined collection frequency-recognition time mapping table. The collection frequency offset represents the positive deviation of the deep foundation data collection frequency from the upper limit of the frequency reference. If the streaming deep foundation data processing window duration reference value is greater than the window duration reference value, the window duration offset and the construction condition recognition result confidence threshold are input into the collection frequency-recognition time mapping table for querying to obtain the recognition response time decay factor. The construction condition recognition response time target value is obtained by interacting the streaming deep foundation data processing window duration reference value and the recognition response time decay factor. The window duration offset is used to reflect the positive deviation of the streaming deep foundation data processing window duration reference value from the window duration reference value.

4. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 3, characterized in that, The execution of the response time dynamic calibration mechanism further includes: If not, the window duration correction amount and the construction condition recognition result confidence threshold are input into the collection frequency-recognition time mapping table for querying to obtain the recognition response time gain factor. The construction condition recognition response time target value is obtained by interacting the streaming deep foundation data processing window duration reference value and the recognition response time gain factor. The window duration correction amount is used to reflect the negative deviation of the streaming deep foundation data processing window duration reference value from the window duration reference value. If the deep foundation data collection frequency is within the frequency reference interval, the response time dynamic calibration mechanism is not executed. The frequency reference interval represents the closed interval formed by the lower limit of the frequency reference and the upper limit of the frequency reference. If the deep foundation data collection frequency is less than the lower limit of the frequency reference, it is determined that it is in the low-frequency collection-delayed response state, and the batch deep foundation data processing mode is maintained. Specifically, based on the collection frequency correction amount and the matching degree offset, the batch processing interval reference value and the construction condition recognition result confidence threshold are obtained by querying the pre-defined collection frequency-recognition time mapping table. The batch processing interval reference value is set as the construction condition recognition response time target value. The collection frequency correction amount represents the negative deviation of the deep foundation data collection frequency from the lower limit of the frequency reference.

5. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 2, characterized in that, The specific process of the push time dynamic calibration mechanism is as follows: If the deep foundation edge measurement point arrangement density is greater than the upper limit of the arrangement density reference, the ultra-fast push early warning information mode is performed. Specifically, based on the matching degree offset and the arrangement density offset, the early warning confidence threshold scaling factor is obtained by querying the pre-defined arrangement density-push time mapping table. The early warning information push response time target value is obtained by interacting the early warning confidence threshold scaling factor and the current early warning information push response time and taking the integer result. The arrangement density offset is used to reflect the positive deviation of the deep foundation edge measurement point arrangement density from the upper limit of the arrangement density reference.

6. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 5, characterized in that, The push time dynamic calibration mechanism further includes: If the deep foundation pit edge measurement point layout density is within the layout density reference interval, the push time dynamic calibration mechanism is not executed, and the layout density reference interval represents a closed interval formed by the layout density reference lower limit and the layout density reference upper limit; If the deep foundation pit edge measurement point layout density is less than the layout density reference lower limit, a delayed push early warning information mode is performed, specifically: based on the matching degree offset and the layout density correction amount, the early warning confidence threshold multiplier is obtained by querying the pre-defined layout density-push time mapping table, and the early warning confidence threshold multiplier and the current early warning information push response time are interactively processed to obtain the early warning information push response time target value, and the layout density correction amount is used to reflect the negative deviation of the deep foundation pit edge measurement point layout density and the layout density reference lower limit.

7. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 1, characterized in that, The research and judgment efficiency parameter includes a BIM model dynamic update matching degree, a construction condition-risk analysis response time difference, and an out-of-date model safety misjudgment frequency; The matching degree correction factor and the update matching degree component are interactively processed to obtain an update matching degree influence value, and the update matching degree component represents the result of the BIM model dynamic update matching degree and the matching degree threshold proportion analysis; The response time difference correction factor and the response time difference component are interactively processed to obtain a response time difference influence value, and the response time difference component represents the result of the response time difference threshold and the construction condition-risk analysis response time difference proportion analysis; The misjudgment frequency correction factor and the misjudgment frequency component are interactively processed to obtain a misjudgment frequency influence value, and the misjudgment frequency component represents the result of the misjudgment frequency threshold and the out-of-date model safety misjudgment frequency proportion analysis; The update matching degree influence value, the response time difference influence value, and the misjudgment frequency influence value are coupled to obtain an early warning logic intelligent research and judgment efficiency. The specific steps of determining whether to execute intelligent research and judgment efficiency optimization are: If the early warning logic intelligent research and judgment efficiency is greater than or equal to the efficiency reference value, intelligent research and judgment efficiency optimization is not executed, otherwise, if the early warning logic intelligent research and judgment efficiency is less than the efficiency reference value, based on the efficiency offset, a pre-warning time threshold dynamic calibration mechanism and an early warning effect feedback iteration mechanism are executed, and the efficiency offset represents the negative deviation of the early warning logic intelligent research and judgment efficiency and the efficiency reference value.

8. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 7, characterized in that, The specific process of executing the pre-warning time threshold dynamic calibration mechanism is: If the early warning information space visualization matching rate is higher than the matching rate reference upper limit, based on the efficiency offset and the matching rate offset, the pre-warning time threshold decay factor is obtained by querying the pre-defined matching rate-threshold adjustment mapping table, and the current pre-warning time threshold reference value and the pre-warning time threshold decay factor are interactively processed to obtain the target pre-warning average pre-warning time threshold, and the matching rate offset is used to reflect the positive deviation of the early warning information space visualization matching rate and the matching rate reference upper limit; If the early warning information space visualization matching rate is within the matching rate reference interval, the pre-warning time threshold dynamic calibration mechanism is not executed, and the matching rate reference interval represents a closed interval formed by the matching rate reference lower limit and the matching rate reference upper limit; If the matching rate of the early warning information space visualization is lower than the lower limit of the matching rate reference, a pre-defined matching rate-threshold adjustment mapping table is queried based on the effective rate offset and the matching rate correction amount to obtain a lead time threshold gain factor, and the current lead time threshold reference value is interactively processed with the lead time threshold gain factor to obtain a target early warning average lead time threshold, wherein the matching rate correction amount is used to reflect the negative deviation of the early warning information space visualization matching rate from the upper limit of the matching rate reference.

9. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 8, characterized in that, The specific process of the early warning effect feedback iteration mechanism is as follows: If the construction safety risk early warning threshold is greater than the first risk threshold and less than the second risk threshold, it is determined that the construction safety state is medium risk, and a medium-speed high-quality iteration mode is performed, specifically: the early warning threshold offset and the effective rate offset are queried from a pre-defined early warning threshold-iteration frequency mapping table to obtain an early warning threshold dynamic adjustment frequency coefficient, wherein the early warning threshold offset is used to reflect the negative deviation of the construction safety risk early warning threshold from the first risk threshold; If the early warning threshold dynamic adjustment frequency coefficient is greater than or equal to the frequency coefficient reference coefficient, the early warning threshold dynamic adjustment frequency coefficient is input into the pre-defined early warning threshold-iteration frequency mapping table to query an iteration cycle gain amount, and the current early warning effect feedback iteration cycle is coupled with the iteration cycle gain amount to obtain the target early warning effect feedback iteration cycle.

10. The BIM-based deep foundation pit construction safety risk dynamic early warning method according to claim 9, characterized in that, The early warning effect feedback iteration mechanism further comprises: If the early warning threshold dynamic adjustment frequency coefficient is less than the frequency coefficient reference coefficient, the early warning threshold dynamic adjustment frequency coefficient is input into the pre-defined early warning threshold-iteration frequency mapping table to query an iteration cycle reduction amount, and the current early warning effect feedback iteration cycle is coupled with the iteration cycle reduction amount to obtain the target early warning effect feedback iteration cycle; If the construction safety risk early warning threshold is less than or equal to the first risk threshold, it is determined that the construction safety state is low risk, and the early warning effect feedback iteration mechanism is not performed; If the construction safety risk early warning threshold is greater than or equal to the second risk threshold, it is determined that the construction safety state is high risk, and the pre-set relevant personnel are reminded to check the deep foundation pit construction site.

Images