A building safety data analysis method and system based on a construction database

Abstract

The present application relates to the technical field of building safety data analysis based on construction database, and particularly relates to a building safety data analysis method and system based on construction database, which comprises the following steps: grouping displacement sensors in a deep foundation pit monitoring area, and presetting risk propagation correlation between each displacement sensor group; calculating the instantaneous change rate of displacement data; identifying that any displacement sensor group has entered an accelerated displacement trend, and marking the accelerated displacement trend as a local abnormal event; performing linkage search on displacement sensor groups associated with the local abnormal event in an incremental time window and spatial range; and if the linkage search finds that multiple displacement sensor groups with risk propagation correlation all present an accelerated displacement trend, generating a regional linkage instability early warning. The above can timely find a local abnormal event, thereby improving the accuracy of analysis.

Description

Technical Field

[0001] This invention relates to the technical field of building safety data analysis based on a building safety database, and specifically to a method and system for building safety data analysis based on a building safety database. Background Technology

[0002] In the safety management of construction projects, existing construction safety data analysis systems typically rely on construction safety databases to store and process safety data. The core of this system is a set of rules for identifying high-risk work behaviors, primarily derived from data summaries of traditional construction projects (such as steel structure installation and concrete pouring). Therefore, when handling routine construction activities, the system can accurately identify problems, such as insufficient protection during high-altitude operations, crane overloading, or abnormal concrete pouring speeds. The construction safety database is also designed around the characteristics of these traditional projects, facilitating the storage and retrieval of relevant risk data to assist the analysis system in making judgments.

[0003] Faced with changes in project types, existing high-risk operation identification rules are proving inadequate when dealing with data specific to deep foundation pit and underground structure construction. For example, when processing data on support structure deformation monitoring (such as retaining pile inclination and stress changes in support beams) and groundwater control parameters (such as dewatering well water levels and drainage volume), the system cannot identify or correlate these data for analysis. This is because the initial rules lack feature extraction and pattern matching logic for these new data types. The system cannot understand the significance of continuous local deformation of the retaining structure, nor can it link abnormal fluctuations in groundwater levels with potential soil collapse. This insufficient ability to handle new risk data results in existing analysis methods providing very incomplete coverage of the unique hazards of deep foundation pit and underground structure construction. Summary of the Invention

[0004] The purpose of this invention is to address the aforementioned shortcomings by proposing a building safety data analysis method and system based on a building safety database.

[0005] The present invention adopts the following technical solution:

[0006] A building safety data analysis method based on a building safety database, comprising the following steps:

[0007] Displacement sensors within the deep foundation pit monitoring area are grouped, and risk propagation relationships between each displacement sensor group are preset based on the geological conditions of the deep foundation pit.

[0008] Continuously acquire displacement data from each displacement sensor group and calculate the instantaneous rate of change of the displacement data;

[0009] Based on the instantaneous change rate of displacement data meeting the set change characteristics, it is identified that any group of displacement sensors has entered an accelerated displacement trend, and the accelerated displacement trend is marked as a local abnormal event.

[0010] Based on local anomalies, and according to the preset risk propagation correlation, a linkage search is performed on the displacement sensor group associated with the local anomalies with an increasing time window and spatial range. The linkage search is used to discover whether the associated displacement sensor group shows an accelerated displacement trend.

[0011] If the linked search finds that multiple displacement sensor groups with risk propagation correlations all show an accelerating displacement trend, then an aggregated regional linked instability warning will be generated.

[0012] Among them, when the preset risk propagation correlation fails to reflect the geological conditions of the deep foundation pit, the steps for generating a regional linkage instability early warning include:

[0013] Identify localized anomalous events;

[0014] Initiate an omnidirectional neighbor group search. The omnidirectional neighbor group search is used to send probe signals to all displacement sensor groups that are physically adjacent to the displacement sensor group where the local anomaly event occurred, in order to find whether a corresponding accelerated displacement trend also appears in the time period immediately following the time point of the local anomaly event, and mark it as a subsequent local anomaly event.

[0015] For any subsequent local anomaly event found in the omnidirectional nearest neighbor group search, the spatiotemporal correlation degree between the subsequent local anomaly event and the local anomaly event is calculated. The spatiotemporal correlation degree takes into account the time interval and spatial distance between the subsequent local anomaly event and the local anomaly event.

[0016] Based on the spatiotemporal correlation, candidate risk propagation paths are established, and these paths are used to infer the direction of risk propagation.

[0017] If the risk propagation direction corresponding to a certain candidate risk propagation path is repeatedly verified in actual data, the risk propagation direction will be marked as a dynamic candidate path, and the priority of the dynamic candidate path in subsequent linked scanning will be increased.

[0018] If, within a specific time and space range, multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend, then a regional linkage instability warning will be generated by aggregation.

[0019] Through this technical solution, this application can effectively identify early signs of structural instability in deep foundation pits. By grouping displacement sensor groups and pre-setting risk propagation correlations, combined with monitoring of instantaneous change rates, local abnormal events can be detected in a timely manner. Through a linkage search mechanism, when multiple related groups show an accelerating displacement trend, a regional linkage instability early warning is generated. This solves the problems of incomplete coverage of the unique hazards of deep foundation pit and underground structure construction and the loss of key dynamic displacement data leading to reduced early warning accuracy in existing technologies. It also achieves the unexpected technical effect of accurately capturing early signs of structural instability.

[0020] Furthermore, the method also includes:

[0021] Based on the aggregated regional linkage instability warning, corresponding regional linkage instability warning information is generated.

[0022] Real-time acquisition of dynamic environmental data at the deep foundation pit construction site, including real-time rainfall, surrounding construction vibration intensity, and changes in groundwater pumping rate;

[0023] Correlation analysis will be performed between dynamic environmental data and regional linkage instability early warning information.

[0024] Based on the correlation analysis results, the urgency of the regional linkage instability warning is revised or the potential impact range of the regional linkage instability warning is enhanced.

[0025] Furthermore, the regional linkage instability early warning information includes the unique identifier of the warning, the time of occurrence, the location of occurrence, the warning type, and the initial risk level, which is used to improve the standardization of subsequent correlation analysis.

[0026] Furthermore, the steps of grouping displacement sensors within the deep foundation pit monitoring area and pre-setting risk propagation relationships between each displacement sensor group based on the geological conditions of the deep foundation pit include:

[0027] Real-time acquisition of geological monitoring data at deep foundation pit construction sites; geological monitoring data indicates changes in geological conditions.

[0028] Identify geological condition change events based on geological monitoring data;

[0029] Based on geological condition change events, assess the impact of geological condition changes on the grouping of displacement sensors;

[0030] Based on the assessment results, the division of displacement sensor groups in the monitoring area of ​​the affected deep foundation pit will be adjusted.

[0031] Based on the adjusted displacement sensor group division and changes in geological conditions, the risk propagation correlation between each displacement sensor group is updated;

[0032] The updated displacement sensor grouping and its corresponding risk propagation relationships will be used in subsequent steps.

[0033] Furthermore, the steps of identifying any group of displacement sensors that has entered an accelerating displacement trend and marking the accelerating displacement trend as a local anomalous event include:

[0034] Real-time acquisition of local environmental data corresponding to the location of the displacement sensor group;

[0035] Match local environmental data with identified acceleration displacement trends;

[0036] If the matching results show that the accelerated displacement trend matches the non-structural displacement response characteristics caused by environmental disturbances, then the warning priority of the accelerated displacement trend should be reduced.

[0037] If the matching results show that the accelerated displacement trend has a low correlation with environmental disturbances, then the local abnormal event state of the accelerated displacement trend is maintained.

[0038] Furthermore, the steps of identifying any group of displacement sensors that has entered an accelerating displacement trend and marking the accelerating displacement trend as a local anomalous event include:

[0039] Identify the geological strata where the displacement sensor group is located;

[0040] Obtain information on the current construction phase;

[0041] Obtain the environmental conditions corresponding to the location of the displacement sensor group;

[0042] Based on geological strata, current construction stage information, and environmental conditions, adjust the judgment threshold and duration standard for identifying accelerated displacement trends;

[0043] When the displacement data of the displacement sensor group meets the adjusted judgment threshold and duration standard, the accelerated displacement trend will be marked as a local abnormal event.

[0044] Furthermore, the steps of identifying any group of displacement sensors that has entered an accelerating displacement trend and marking the accelerating displacement trend as a local anomalous event include:

[0045] Real-time monitoring of the continuity and stability of displacement data stream from the displacement sensor group;

[0046] When monitoring detects intermittent interruptions, sudden jumps in data values, or prolonged deviations from the normal operating range in the displacement sensor group's displacement data stream, the displacement sensor self-diagnosis program is activated to check the power supply, communication links, and internal status of the displacement sensor group, thereby identifying the abnormal displacement sensor group.

[0047] Cross-compare the displacement data of the abnormal displacement sensor group with those of the adjacent displacement sensor group;

[0048] If the cross-comparison results show that the displacement data of adjacent displacement sensor groups are normal and no corresponding acceleration displacement trend is displayed, the anomaly will be marked as displacement fluctuation caused by displacement sensor failure and excluded from local anomaly events.

[0049] If the self-diagnostic results show that the displacement sensor group is working normally, and the cross-comparison results show that adjacent displacement sensor groups are associated with an accelerating displacement trend, then the local abnormal event state of the accelerating displacement trend is maintained.

[0050] Furthermore, the method includes:

[0051] The displacement data stream from the displacement sensor group is received in real time, and the displacement data stream from the displacement sensor group is segmented and buffered.

[0052] Calculate the packet loss rate and short-term fluctuation amplitude of the displacement data stream segments for each displacement sensor group;

[0053] Set short-time windows and long-time windows;

[0054] Determine whether the packet loss rate or short-term fluctuation within a short time window exceeds a preset first threshold, and mark the instantaneous anomaly based on the determination result;

[0055] Determine whether the frequency or duration of transient anomalies within a long time window exceeds a preset second threshold, and mark persistent anomalies based on the determination result;

[0056] Based on the marking results of instantaneous or persistent anomalies, a decision is made on whether to trigger a displacement sensor fault warning.

[0057] Furthermore, the steps of receiving the displacement data stream from the displacement sensor group in real time and segmenting and buffering the displacement data stream from the displacement sensor group include:

[0058] Real-time reception of displacement data streams from a group of displacement sensors;

[0059] Detect the wireless signal strength or multipath effect of the displacement data stream of a displacement sensor group and obtain the real-time signal quality;

[0060] When the wireless signal strength is lower than the preset strength threshold or the multipath effect is enhanced, the adaptive data reception mode is activated.

[0061] In adaptive data reception mode, the backup communication link is enabled or the data transmission protocol is adjusted to improve the stability of the displacement data stream of the displacement sensor group.

[0062] Based on the real-time signal quality, the segmentation and buffering strategy of the displacement data stream of the displacement sensor group is dynamically adjusted to improve the continuity of the displacement data stream of the displacement sensor group.

[0063] Perform integrity verification on the received data packets;

[0064] For incomplete or erroneous data packets, retransmission requests are made or local data interpolation is used to restore the data, thereby improving the continuity of the displacement data stream of the displacement sensor group.

[0065] The displacement data stream of the displacement sensor group is segmented and buffered.

[0066] This application also discloses a building safety data analysis system based on a building safety database, applicable to a building safety data analysis method based on a building safety database. The system includes:

[0067] The preset module groups the displacement sensors in the deep foundation pit monitoring area and presets the risk propagation correlation between each displacement sensor group based on the geological conditions of the deep foundation pit.

[0068] The processing module continuously acquires displacement data from each displacement sensor group and calculates the instantaneous rate of change of the displacement data.

[0069] The identification module identifies any group of displacement sensors that has entered an accelerating displacement trend based on the instantaneous change rate of displacement data meeting the set change characteristics, and marks the accelerating displacement trend as a local abnormal event.

[0070] The search module, based on local anomalies, performs a linked search on displacement sensor groups associated with local anomalies according to preset risk propagation relationships, with increasing time windows and spatial ranges. The linked search is used to discover whether the associated displacement sensor groups show an accelerating displacement trend.

[0071] If the linkage search finds that multiple displacement sensor groups with risk propagation correlations all show an accelerating displacement trend, the early warning generation module will aggregate and generate a regional linkage instability early warning.

[0072] Among them, when the preset risk propagation correlation fails to reflect the geological conditions of the deep foundation pit, the steps for generating a regional linkage instability early warning include:

[0073] Identify localized anomalous events;

[0074] Initiate an omnidirectional neighbor group search. The omnidirectional neighbor group search is used to send probe signals to all displacement sensor groups that are physically adjacent to the displacement sensor group where the local anomaly event occurred, in order to find whether a corresponding accelerated displacement trend also appears in the time period immediately following the time point of the local anomaly event, and mark it as a subsequent local anomaly event.

[0075] For any subsequent local anomaly event found in the omnidirectional nearest neighbor group search, the spatiotemporal correlation degree between the subsequent local anomaly event and the local anomaly event is calculated. The spatiotemporal correlation degree takes into account the time interval and spatial distance between the subsequent local anomaly event and the local anomaly event.

[0076] Based on the spatiotemporal correlation, candidate risk propagation paths are established, and these paths are used to infer the direction of risk propagation.

[0077] If the risk propagation direction corresponding to a certain candidate risk propagation path is repeatedly verified in actual data, the risk propagation direction will be marked as a dynamic candidate path, and the priority of the dynamic candidate path in subsequent linked scanning will be increased.

[0078] If, within a specific time and space range, multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend, then a regional linkage instability warning will be generated by aggregation.

[0079] This application provides a system-level solution through this technical solution. Through modular design, it realizes comprehensive analysis and early warning of safety data for deep foundation pits. It can effectively solve the problem of insufficient ability of existing systems to process data specific to deep foundation pits and underground structure construction, and achieves unexpected technical effects in accurately capturing early signs of structural instability.

[0080] This application enables accurate identification and early warning of early signs of structural instability, significantly improving the level of intelligence and risk prevention and control capabilities of building engineering safety management.

[0081] To further understand the features and technical content of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are for reference and illustration only and are not intended to limit the present invention. Attached Figure Description

[0082] Figure 1 This is a flowchart of a building safety data analysis method based on a building safety database according to the present invention;

[0083] Figure 2 This is a schematic diagram of the structure of a building safety data analysis system based on a building safety database according to the present invention. Detailed Implementation

[0084] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can understand the advantages and effects of the present invention from the content disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. Furthermore, the accompanying drawings of the present invention are for simple illustrative purposes only and are not depictions of actual dimensions; this is stated in advance. The following embodiments will further describe the relevant technical content of the present invention in detail, but the disclosed content is not intended to limit the scope of protection of the present invention.

[0085] This embodiment provides a building safety data analysis method and system based on a building safety database, combined with... Figure 1 and Figure 2 As shown.

[0086] refer to Figure 1 A building safety data analysis method based on a building safety database, comprising the following steps:

[0087] Displacement sensors within the deep foundation pit monitoring area are grouped, and risk propagation relationships between each displacement sensor group are preset based on the geological conditions of the deep foundation pit.

[0088] Continuously acquire displacement data from each displacement sensor group and calculate the instantaneous rate of change of the displacement data;

[0089] Based on the instantaneous change rate of displacement data meeting the set change characteristics, it is identified that any group of displacement sensors has entered an accelerated displacement trend, and the accelerated displacement trend is marked as a local abnormal event.

[0090] Based on local anomalies, and according to the preset risk propagation correlation, a linkage search is performed on the displacement sensor group associated with the local anomalies with an increasing time window and spatial range. The linkage search is used to discover whether the associated displacement sensor group shows an accelerated displacement trend.

[0091] If the linked search finds that multiple displacement sensor groups with risk propagation correlations all show an accelerating displacement trend, then an aggregated regional linked instability warning will be generated.

[0092] Among them, when the preset risk propagation correlation fails to reflect the geological conditions of the deep foundation pit, the steps for generating a regional linkage instability early warning include:

[0093] Identify localized anomalous events;

[0094] Initiate an omnidirectional neighbor group search. The omnidirectional neighbor group search is used to send probe signals to all displacement sensor groups that are physically adjacent to the displacement sensor group where the local anomaly event occurred, in order to find whether a corresponding accelerated displacement trend also appears in the time period immediately following the time point of the local anomaly event, and mark it as a subsequent local anomaly event.

[0095] For any subsequent local anomaly event found in the omnidirectional nearest neighbor group search, the spatiotemporal correlation degree between the subsequent local anomaly event and the local anomaly event is calculated. The spatiotemporal correlation degree takes into account the time interval and spatial distance between the subsequent local anomaly event and the local anomaly event.

[0096] Based on the spatiotemporal correlation, candidate risk propagation paths are established, and these paths are used to infer the direction of risk propagation.

[0097] If the risk propagation direction corresponding to a certain candidate risk propagation path is repeatedly verified in actual data, the risk propagation direction will be marked as a dynamic candidate path, and the priority of the dynamic candidate path in subsequent linked scanning will be increased.

[0098] If, within a specific time and space range, multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend, then a regional linkage instability warning will be generated by aggregation.

[0099] The term "deep foundation pit" as used in this application refers to foundation pit projects with an excavation depth exceeding a certain standard (e.g., 5 meters), characterized by poor soil stability, complex groundwater conditions, and high construction risks. A "displacement sensor" is a device used to measure changes in the position of an object; in this application, it is primarily used to monitor the displacement of the retaining structure, surrounding soil, or buildings during the construction process of deep foundation pits. A "displacement sensor group" refers to a logical grouping of multiple displacement sensors based on their geographical location, monitoring object, or function, forming a set for centralized management and analysis. "Risk propagation correlation" refers to the potential connection between different displacement sensor groups; due to geological conditions, structural connections, or construction activities, when one group experiences abnormal displacement, other groups may also experience displacement. "Instantaneous change rate" refers to the rate of change of displacement data within a very short time; it sensitively reflects the dynamic change trend of the structure or soil and is a key indicator for judging accelerated instability. A "local anomaly" refers to a situation where the instantaneous change rate of displacement data for a certain displacement sensor group meets the set change characteristics, indicating that the group has entered an accelerated displacement trend but has not yet reached the level of regional instability. "Linked search" is a process that systematically explores groups of displacement sensors associated with local anomalies based on pre-defined risk propagation relationships, using incremental time windows and spatial ranges, aiming to discover whether a chain reaction exists. "Regional linked instability early warning" refers to a high-level warning issued by the system when linked search finds that multiple groups of displacement sensors with risk propagation relationships are all showing an accelerated displacement trend, indicating that the deep foundation pit may experience regional overall instability.

[0100] When grouping displacement sensors within a deep foundation pit monitoring area and pre-setting risk propagation relationships between sensor groups based on the geological conditions of the pit, several methods can be employed. For example, adjacent displacement sensors can be grouped according to the planar layout and geological stratification of the deep foundation pit. Specifically, if a weak soil layer exists on one side of the deep foundation pit, all displacement sensors in that area can be grouped together, and a high risk propagation relationship can be pre-set between this group and adjacent groups. Another method is to group sensors based on the type of support structure and stress characteristics of the deep foundation pit. For instance, displacement sensors on the same support beam can be grouped together, and risk propagation relationships can be pre-set based on the force transmission path of the support beam. When pre-setting risk propagation relationships, they can be manually set based on historical monitoring data and expert experience, or initially set using geological survey reports and numerical simulation results.

[0101] The system continuously acquires displacement data from each group of displacement sensors and calculates the instantaneous rate of change of the displacement data. Displacement data can be acquired through wired or wireless communication, transmitting the data collected by the displacement sensors to the data processing center in real time. For example, the displacement sensors can upload displacement data to a cloud platform via a LoRaWAN network. The instantaneous rate of change of the displacement data can be calculated using the differential method. Specifically, the instantaneous rate of change can be obtained by subtracting the displacement data from the previous moment's displacement data at the current moment and then dividing by the time interval. For example, if displacement data is collected every 10 minutes, the instantaneous rate of change can be expressed as ΔL / Δt, where ΔL is the displacement change within 10 minutes, and Δt is 10 minutes.

[0102] Based on the instantaneous rate of change of displacement data meeting predefined characteristics, the system identifies any group of displacement sensors as having entered an accelerating displacement trend and marks this trend as a local anomaly. Predefined characteristics may include the absolute value of the instantaneous rate of change exceeding a certain threshold, or the instantaneous rate of change continuously increasing over a period of time. For example, if the instantaneous rate of change of a group of displacement sensors exceeds 5 mm / h for 30 consecutive minutes, the group can be considered to have entered an accelerating displacement trend. After identifying the accelerating displacement trend, the system marks it as a local anomaly and records the time, location, and displacement data of the event.

[0103] Based on local anomalies, and according to pre-defined risk propagation relationships, the system performs a linked search on displacement sensor groups associated with the local anomalies, using incremental time windows and spatial ranges. This linked search is used to determine whether associated displacement sensor groups exhibit an accelerating displacement trend. The linked search can begin with directly related groups to the group where the local anomaly occurred and gradually expand the search scope. For example, if a local anomaly occurs in group A, and group A has a risk propagation relationship with groups B and C, the system will first search groups B and C. The search time window can start from 1 hour after the event and gradually increase to 2 hours, 4 hours, etc. The spatial range can expand from directly adjacent groups to secondarily adjacent groups.

[0104] If a linked search reveals that multiple displacement sensor groups with a risk propagation relationship all exhibit an accelerating displacement trend, a regional linked instability warning will be generated. For example, if a linked search finds that groups A, B, and C all exhibit accelerating displacement trends and that there is a risk propagation relationship between them, the system will aggregate and generate a regional linked instability warning. The warning information may include the extent of the instability area, the warning level, and recommended emergency measures.

[0105] Specifically, identifying local anomalies means identifying any displacement sensor group that has entered an accelerating displacement trend based on the instantaneous rate of change of displacement data meeting the set change characteristics, and marking the accelerating displacement trend as a local anomaly.

[0106] The omnidirectional neighbor group search can be understood as an active detection mechanism. Its purpose is to break through the limitations of preset correlations by sending probing signals to physically adjacent groups of displacement sensors to discover potential risk propagation that is not covered by preset correlations. The probing signal aims to find whether these neighbor groups also show an accelerated displacement trend in the time period immediately following the occurrence of a local anomaly, and to mark them as subsequent local anomalies.

[0107] In practical applications, spatiotemporal correlation is a quantitative indicator that comprehensively considers the time interval and spatial distance between subsequent local abnormal events and local abnormal events. For example, weighted average or multidimensional distance calculation can be used. Its purpose is to assess the closeness of two abnormal events in time and space, so as to determine whether there is a causal or propagation relationship between them.

[0108] Furthermore, based on spatiotemporal correlation, candidate risk propagation paths can be established, which are used to infer the direction of risk propagation. For example, if a subsequent local anomaly occurs immediately after the local anomaly in time and is spatially close, it can be inferred that the risk may propagate along the direction between these two groups.

[0109] As a preferred implementation, if the risk propagation direction corresponding to a candidate risk propagation path is repeatedly verified in actual data, such as through historical data backtracking or multiple confirmations via real-time monitoring, then that risk propagation direction is marked as a dynamic candidate path, and its priority in subsequent linked scans is increased. The purpose is to enable the system to learn and adapt to actual risk propagation patterns through data verification, thereby improving the accuracy and efficiency of early warnings.

[0110] Therefore, if multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend within a specific spatiotemporal range, an early warning of regional linkage instability can be generated by aggregation.

[0111] In some preferred embodiments, it is assumed that during the construction of a deep foundation pit, due to sudden changes in the groundwater level or unknown anomalies in the local soil properties, the preset risk propagation correlation fails to accurately capture new potential slip surfaces. In this case, displacement sensor group A located in a certain area of ​​the deep foundation pit is first identified as entering an accelerated displacement trend and is marked as a local anomaly.

[0112] The system then initiates an omnidirectional neighbor group search, sending probe signals to displacement sensor groups B, C, and D, which are physically adjacent to group A. Shortly after the anomaly occurred in group A, groups B and C were also found to exhibit accelerated displacement trends and were marked as subsequent local anomalies. The system then calculates the spatiotemporal correlation between group B and group A, as well as the spatiotemporal correlation between group C and group A. If the spatiotemporal correlation between group B and group A is found to be significantly higher than that between group C and group A, the system establishes a candidate risk propagation path from group A to group B.

[0113] As monitoring continues, if the risk propagation direction from group A to group B is repeatedly verified in subsequent similar events—for example, if anomalies in group A always follow anomalies in group B at other times or under similar conditions—then this propagation direction will be marked as a dynamic candidate path and given higher priority in subsequent linked scans. Ultimately, if, within a specific spatiotemporal range, both group A and group B are found to exhibit an accelerating displacement trend along this dynamic candidate path, the system will aggregate and generate a regional linked instability warning. In this way, even if the preset correlation is imperfect, the system can dynamically learn and adapt to actual risk propagation patterns through data-driven methods, thereby providing more accurate and timely warnings.

[0114] This application further proposes a building safety data analysis method based on a building safety database, which also includes:

[0115] Based on the aggregated regional linkage instability warning, corresponding regional linkage instability warning information is generated.

[0116] Real-time acquisition of dynamic environmental data at the deep foundation pit construction site, including real-time rainfall, surrounding construction vibration intensity, and changes in groundwater pumping rate;

[0117] Correlation analysis will be performed between dynamic environmental data and regional linkage instability early warning information.

[0118] Based on the correlation analysis results, the urgency of the regional linkage instability warning is revised or the potential impact range of the regional linkage instability warning is enhanced.

[0119] Specifically, after aggregating and generating regional linkage instability warnings, the system will generate corresponding regional linkage instability warning information based on the warning.

[0120] Real-time acquisition of dynamic environmental data at deep foundation pit construction sites refers to the continuous collection of environmental parameters closely related to the stability of deep foundation pits through various environmental monitoring devices deployed at the construction site. This dynamic environmental data specifically includes real-time rainfall, surrounding construction vibration intensity, and changes in groundwater pumping rates. Real-time rainfall is used to assess the impact of surface water infiltration on soil moisture content and seepage pressure; surrounding construction vibration intensity is used to assess the disturbance of external construction activities to the structural stability of deep foundation pits; and changes in groundwater pumping rates are used to assess the impact of groundwater level fluctuations on the effective stress of the soil. The real-time nature of this data ensures timely response to environmental changes, aiming to provide the latest external influencing factors for the correction and enhancement of early warning systems.

[0121] In practical applications, correlation analysis between dynamic environmental data and regional linked instability early warning information refers to exploring the interrelationships and influence mechanisms between dynamic environmental factors and identified regional linked instability early warnings through data fusion and analysis algorithms. For example, statistical methods, machine learning models, or expert system rules can be used to analyze how the urgency or scope of impact of regional linked instability early warnings might change under specific conditions of rainfall, vibration intensity, or groundwater drainage variations. The aim is to reveal the potential aggravating or mitigating effects of environmental factors on the risk of deep foundation pit instability.

[0122] Furthermore, based on the correlation analysis results, the urgency of regional coordinated instability warnings can be revised, or the potential impact range of regional coordinated instability warnings can be enhanced. Revising the urgency means adjusting the warning priority or response time requirements based on the impact of environmental factors. For example, under extreme rainfall conditions, the warning level may be upgraded even if displacement data changes little. Enhancing the potential impact range means expanding the geographical area covered by the warning based on the potential risk diffusion caused by environmental factors. For example, in a strong vibration environment, instability risk may spread from a local area to the surrounding areas. The aim is to make the warning information more accurately reflect the actual risk situation and guide more targeted preventative measures on-site.

[0123] In some preferred embodiments, assuming that a regional linked instability warning has been generated within a deep foundation pit monitoring area, indicating a risk of linked instability in a certain area, the system will first generate regional linked instability warning information including the warning time, location, type, and initial severity. Subsequently, the system acquires dynamic environmental data of the deep foundation pit construction site in real time. For example, it may detect that real-time rainfall has reached 50 mm in the past 2 hours, the intensity of surrounding construction vibration has continuously exceeded the safety threshold near the warning area, and the groundwater pumping rate has increased significantly in a short period of time. The system performs correlation analysis on these dynamic environmental data with the generated regional linked instability warning information. The analysis results may show that high rainfall and continuous vibration will significantly accelerate soil softening and structural damage, while the increase in groundwater pumping rate may lead to changes in effective stress, further exacerbating the instability risk. Based on this correlation analysis results, the system will revise the urgency of the original regional linked instability warning, for example, by raising the warning level from "medium" to "high," and enhance the potential impact range of the warning, for example, by expanding the original impact range by 10 meters to cover areas that may be further affected by environmental factors. As a result, on-site management personnel can obtain more accurate and instructive early warning information, and thus take timely emergency measures such as evacuation, reinforcement or cessation of construction, effectively avoiding or mitigating potential accident losses.

[0124] This application further proposes that regional linkage instability early warning information include a unique identifier for the warning, the time of occurrence, the location of occurrence, the type of warning, and the initial risk level, in order to improve the standardization of subsequent correlation analysis.

[0125] Specifically, this early warning information can be understood as a structured description of the identified regional interconnected instability risks, with warning types such as soil sliding and localized collapse. Its purpose is to provide a standardized data foundation for subsequent analysis and decision-making.

[0126] This application further proposes the following steps for grouping displacement sensors within the deep foundation pit monitoring area and, based on the geological conditions of the deep foundation pit, pre-setting the risk propagation correlation between each displacement sensor group:

[0127] Real-time acquisition of geological monitoring data at deep foundation pit construction sites; geological monitoring data indicates changes in geological conditions.

[0128] Identify geological condition change events based on geological monitoring data;

[0129] Based on geological condition change events, assess the impact of geological condition changes on the grouping of displacement sensors;

[0130] Based on the assessment results, the division of displacement sensor groups in the monitoring area of ​​the affected deep foundation pit will be adjusted.

[0131] Based on the adjusted displacement sensor group division and changes in geological conditions, the risk propagation correlation between each displacement sensor group is updated;

[0132] The updated displacement sensor grouping and its corresponding risk propagation relationships will be used in subsequent steps.

[0133] Specifically, real-time acquisition of geological monitoring data at deep foundation pit construction sites refers to the continuous collection of various parameters reflecting changes in geological conditions through various geological monitoring equipment deployed around the deep foundation pit, such as pore water pressure gauges, earth pressure gauges, inclinometers, and settlement meters. This geological monitoring data can indicate changes in key geological parameters such as soil stress, deformation, groundwater level, and soil density.

[0134] In this context, identifying geological condition change events based on geological monitoring data can be understood as analyzing real-time acquired geological monitoring data. For example, by setting thresholds, trend analysis, or anomaly detection algorithms, it can be determined whether there are significant changes exceeding the normal fluctuation range. For instance, a sudden increase in pore water pressure, an abnormally accelerated soil settlement rate, or inclinometer data indicating significant displacement in the deeper layers of soil can all be identified as geological condition change events.

[0135] In practical applications, assessing the impact of geological condition changes on the classification of displacement sensor groups involves analyzing the potential effects of identified geological condition changes on the physical boundaries, geological continuity, or mechanical response characteristics of existing displacement sensor groups. For example, if a geological condition change indicates the emergence of new weak interlayers or a change in groundwater seepage paths in a certain area, it may be necessary to reconsider the classification method of displacement sensor groups in that area.

[0136] Furthermore, adjusting the displacement sensor group division in the affected deep foundation pit monitoring area, based on the assessment results, refers to reconfiguring the displacement sensor groups in specific areas affected by changes in geological conditions. This may include splitting an existing group into multiple subgroups, or merging multiple adjacent groups into a larger group, to better reflect the actual geological structure and potential instability modes.

[0137] Therefore, updating the risk propagation relationships between displacement sensor groups based on the adjusted group division and changes in geological conditions means reassessing and establishing the risk propagation paths and strengths between the new groups after the adjustment of the displacement sensor group division. For example, if changes in geological conditions weaken the soil connection between two groups, the risk propagation relationship between them may be strengthened; conversely, if a new barrier layer appears, the relationship may be weakened.

[0138] Finally, the updated displacement sensor group division and its corresponding risk propagation correlation were used in subsequent steps to ensure that the subsequent displacement data analysis, local anomaly event identification, and regional linkage instability early warning generation were all based on the latest and most accurate geological conditions and risk propagation models.

[0139] In some preferred embodiments, it is assumed that in the early stages of deep foundation pit construction, displacement sensor groups A and B are divided into two independent areas based on the survey report, and the risk propagation correlation between them is assumed to be weak because a stable rock stratum exists between them. However, during construction, due to continuous heavy rainfall, geological monitoring data shows a significant rise in the groundwater level, and inclinometer data indicates the appearance of new fissures in the rock stratum, indicating a change in geological conditions. After identifying this change in geological conditions, the system assesses the impact of this change on the displacement sensor group division. The assessment results show that the new fissures may lead to a decrease in rock stratum stability, weakening the geological continuity between group A and group B, and may even form a new potential slip surface. Based on this assessment result, the system adjusts the displacement sensor group division of the affected deep foundation pit monitoring area. For example, it may separate some sensors located near fissures from group A or group B to form a new subgroup C, or redefine the boundary between group A and group B. Simultaneously, based on the adjusted displacement sensor group division and the change in geological conditions, the system updates the risk propagation correlation between each displacement sensor group. For example, the presence of fractures may enhance the risk propagation relationship between group A and group B, while the relationship between group C and either group A or group B may be reassessed. Ultimately, the updated displacement sensor group division and its corresponding risk propagation relationships are used for subsequent linkage searches and early warning generation, thereby ensuring that the system can accurately identify potential regional linkage instability risks even under complex and variable geological conditions.

[0140] This application further proposes that the steps of identifying any group of displacement sensors as having entered an accelerating displacement trend and marking the accelerating displacement trend as a local anomalous event include:

[0141] Real-time acquisition of local environmental data corresponding to the location of the displacement sensor group;

[0142] Match local environmental data with identified acceleration displacement trends;

[0143] If the matching results show that the accelerated displacement trend matches the non-structural displacement response characteristics caused by environmental disturbances, then the warning priority of the accelerated displacement trend should be reduced.

[0144] If the matching results show that the accelerated displacement trend has a low correlation with environmental disturbances, then the local abnormal event state of the accelerated displacement trend is maintained.

[0145] Specifically, real-time acquisition of local environmental data corresponding to the location of the displacement sensor group refers to the continuous collection of environmental parameters related to the location of a specific displacement sensor group by using environmental sensors deployed within the deep foundation pit monitoring area, such as rain gauges, thermometers, anemometers, hygrometers, or vibration sensors. This data can include real-time rainfall, ambient temperature, wind speed, humidity, and the intensity of vibrations from surrounding construction. The purpose is to provide a basis for subsequent determination of whether displacement anomalies are caused by environmental factors.

[0146] Matching local environmental data with identified accelerated displacement trends can be understood as performing synchronous analysis of displacement and environmental data. For example, time series analysis, correlation analysis, or machine learning models can be used to assess whether there is a significant synchronicity or causal relationship between accelerated displacement trends and specific environmental events (such as heavy rainfall, severe vibration, or sudden temperature changes). The aim is to distinguish between structural and non-structural displacements.

[0147] In practical applications, if the matching results show that the accelerated displacement trend matches the non-structural displacement response characteristics caused by environmental disturbances, the warning priority of the accelerated displacement trend will be reduced. This means that when the analysis indicates that the observed accelerated displacement trend is likely due to environmental factors (e.g., sensor thermal expansion, slight swaying caused by wind load, short-term surface soil displacement caused by surface water infiltration due to rainfall, etc.) rather than instability of the deep foundation pit structure itself, the system will not immediately trigger a high-level warning, but will instead mark it as a lower-level event of concern, or simply record and continuously observe it. This helps reduce false alarms and avoid unnecessary resource investment and panic.

[0148] Furthermore, if the matching results show a low correlation between the accelerated displacement trend and environmental disturbances, the local anomaly event status of the accelerated displacement trend will be maintained. This means that when there is no significant correlation between the accelerated displacement trend and known environmental disturbance factors, or when its displacement characteristics (such as duration, amplitude, direction, etc.) do not match the non-structural response caused by environmental disturbances, the system will consider the accelerated displacement trend to be more likely a precursor to structural problems in the deep foundation pit, and thus maintain its early warning status as a local anomaly event for further linkage search and risk assessment.

[0149] In some preferred embodiments, assuming that a group of displacement sensors detects a rate of change in displacement data exceeding a preset threshold within a short period, it is initially judged as an accelerating displacement trend. At this time, the system acquires local environmental data corresponding to the location of the displacement sensor group in real time. For example, if the area is simultaneously experiencing continuous heavy rainfall, and historical data shows that this type of rainfall often causes short-term, small-amplitude non-structural displacement of the surface soil, the system will match the current accelerating displacement trend with the heavy rainfall data. If the matching result shows that the characteristics of the accelerating displacement trend (such as displacement direction, duration, amplitude, etc.) highly match the non-structural displacement response characteristics caused by heavy rainfall, the system will lower the warning priority of the accelerating displacement trend, marking it as "displacement fluctuation under environmental influence," rather than immediately triggering a regional coordinated instability warning. Conversely, if the accelerating displacement trend occurs without significant environmental interference, or if its characteristics do not match the response caused by environmental interference, the system will maintain its local abnormal event state and initiate subsequent coordinated search procedures to assess whether there is a real risk of structural instability.

[0150] This application further proposes a step of identifying that any group of displacement sensors has entered an accelerating displacement trend and marking the accelerating displacement trend as a local anomalous event, the step of which includes:

[0151] Identify the geological strata where the displacement sensor group is located;

[0152] Obtain information on the current construction phase;

[0153] Obtain the environmental conditions corresponding to the location of the displacement sensor group;

[0154] Based on geological strata, current construction stage information, and environmental conditions, adjust the judgment threshold and duration standard for identifying accelerated displacement trends;

[0155] When the displacement data of the displacement sensor group meets the adjusted judgment threshold and duration standard, the accelerated displacement trend will be marked as a local abnormal event.

[0156] Specifically, identifying the geological strata of a displacement sensor group refers to determining the specific geological composition of the soil or rock mass beneath the area monitored by each sensor group. For example, based on geological survey reports, borehole data, or geophysical exploration results, the monitoring area of ​​a deep foundation pit can be divided into different geological strata, such as soft soil, sand, clay, and weathered rock. Soil in different geological strata has different mechanical properties and deformation responses, and its displacement rate and pattern under the same load or disturbance may differ significantly.

[0157] Obtaining information on the current construction stage refers to acquiring real-time information on the specific construction phase of the deep foundation pit project. For example, construction stage information may include the depth of earthwork excavation, the progress of support structure installation, groundwater drainage, and changes in surrounding loads. Different construction stages introduce different stress states and disturbance sources, thus affecting the overall stability and local displacement response of the deep foundation pit.

[0158] In practical applications, obtaining environmental conditions corresponding to the location of a displacement sensor group refers to collecting external environmental data related to the location of the displacement sensor group. For example, environmental conditions may include real-time rainfall, ambient temperature, groundwater level changes, and the intensity of vibrations from surrounding construction. These environmental factors, such as soil softening caused by rainfall, thermal expansion and contraction of materials due to temperature changes, and changes in effective stress caused by fluctuations in groundwater level, can all affect the trend of displacement data.

[0159] Furthermore, adjusting the threshold and duration criteria for identifying accelerated displacement trends based on geological strata, current construction stage information, and environmental conditions means dynamically revising the critical values ​​and duration requirements for determining whether a group of displacement sensors has entered an accelerated displacement trend, taking into account the aforementioned dynamic factors. For example, in soft soil layers, deep foundation pits excavated to critical depths, and accompanied by continuous heavy rainfall, even small changes in displacement rate or short durations may be considered accelerated displacement trends; while in hard rock layers, with stable construction loads and favorable environmental conditions, higher displacement rate changes or longer durations may be required to trigger an early warning. The aim is to make the identification of accelerated displacement trends more consistent with actual working conditions and improve the sensitivity and accuracy of early warnings.

[0160] Therefore, when the displacement data from the displacement sensor group meets the adjusted judgment threshold and duration criteria, the accelerated displacement trend is marked as a local anomaly. This means that only when the displacement data takes into account dynamic criteria under specific geological, construction, and environmental conditions is it confirmed as a true accelerated displacement trend, thus avoiding misjudgments caused by improper static threshold settings.

[0161] In some preferred embodiments, it is assumed that in a deep foundation pit project, a group of displacement sensors is deployed at different geological strata, for example, some in the upper soft clay layer and others in the lower hard rock layer. In the early stages of construction, when the excavation depth is shallow and the weather is sunny and the groundwater level is stable, the system may use a relatively lenient judgment threshold. However, as construction progresses into the deep excavation stage, especially when the displacement sensor group is located in the soft clay layer and encounters continuous heavy rainfall, the system dynamically tightens the judgment threshold for identifying accelerated displacement trends and shortens the duration standard based on the acquired geological strata (soft clay), construction stage (deep excavation), and environmental conditions (heavy rainfall). For example, under soft clay layer and heavy rainfall conditions, even a slight increase in displacement rate and a relatively short duration may be immediately flagged as a local anomaly. Conversely, if the displacement sensor group is located in the hard rock layer, under the same construction stage and environmental conditions, the system may use a relatively high judgment threshold and a long duration standard to avoid misjudging minor displacements caused by normal stress release in the rock mass as accelerated displacement trends. Through this dynamic adjustment mechanism, the system can provide more refined and intelligent safety warnings based on the actual complexities of deep foundation pit projects, ensuring the accuracy and practicality of the warnings.

[0162] This application further proposes that the steps of identifying any group of displacement sensors as having entered an accelerating displacement trend and marking the accelerating displacement trend as a local anomalous event include:

[0163] Real-time monitoring of the continuity and stability of displacement data stream from the displacement sensor group;

[0164] When monitoring detects intermittent interruptions, sudden jumps in data values, or prolonged deviations from the normal operating range in the displacement sensor group's displacement data stream, the displacement sensor self-diagnosis program is activated to check the power supply, communication links, and internal status of the displacement sensor group, thereby identifying the abnormal displacement sensor group.

[0165] Cross-compare the displacement data of the abnormal displacement sensor group with those of the adjacent displacement sensor group;

[0166] If the cross-comparison results show that the displacement data of adjacent displacement sensor groups are normal and no corresponding acceleration displacement trend is displayed, the anomaly will be marked as displacement fluctuation caused by displacement sensor failure and excluded from local anomaly events.

[0167] If the self-diagnostic results show that the displacement sensor group is working normally, and the cross-comparison results show that adjacent displacement sensor groups are associated with an accelerating displacement trend, then the local abnormal event state of the accelerating displacement trend is maintained.

[0168] Specifically, real-time monitoring of the continuity and stability of displacement data streams from a group of displacement sensors refers to the system continuously monitoring indicators such as the integrity of data packets received from each displacement sensor group, transmission delay, fluctuation range of data values, and data update frequency. Its purpose is to promptly detect any anomalies in data transmission or the sensors themselves. When monitoring detects intermittent interruptions, sudden jumps in data values, or prolonged deviations from the normal operating range in the displacement data stream of a displacement sensor group, a self-diagnostic program for the displacement sensors will be initiated. This self-diagnostic program aims to conduct an in-depth examination of the displacement sensor group exhibiting abnormal data streams, specifically evaluating its power supply status, communication link connection quality, and the operating status of its internal sensing elements, thereby determining whether the displacement sensor group is abnormal. Its purpose is to preliminarily determine whether the data anomaly stems from a hardware or software fault in the sensor itself. In practical applications, cross-comparison of displacement data from abnormal displacement sensor groups with that of adjacent displacement sensor groups refers to comparing and analyzing the displacement data reported by the abnormal displacement sensor group with the displacement data of other displacement sensor groups physically adjacent to it. Its purpose is to utilize spatial correlation to verify the authenticity of the abnormal data. Furthermore, if the cross-comparison results show that the displacement data of adjacent displacement sensor groups are normal and do not show a corresponding accelerating displacement trend, the anomaly is marked as a displacement fluctuation caused by a displacement sensor malfunction and excluded from the list of local anomaly events. This means that if the abnormal data is not corroborated by surrounding sensors, it is tended to be considered a problem with the sensor itself. Conversely, if the self-diagnostic results show that the displacement sensor group is working normally, and the cross-comparison results show that adjacent displacement sensor groups also show a related accelerating displacement trend, the local anomaly event status of accelerating displacement trend is maintained. This indicates that when the sensor itself is not faulty and surrounding sensors also show a similar accelerating trend, the accelerating displacement trend is confirmed as a genuine local anomaly event.

[0169] In some preferred embodiments, suppose a displacement sensor group A within the deep foundation pit monitoring area suddenly reports a significant accelerating displacement trend at a certain point in time. The system first monitors the continuity and stability of the displacement data stream of displacement sensor group A in real time. If the monitoring detects intermittent interruptions in its data stream or sudden jumps in data values, the system will immediately initiate a displacement sensor self-diagnostic program to check the power supply, communication link, and internal status of displacement sensor group A. Specifically, if the self-diagnostic program reports a fault in the communication link of displacement sensor group A, and simultaneously, through cross-comparison, it is found that the displacement data of adjacent displacement sensor groups B and C both show normal values ​​and do not exhibit an accelerating displacement trend, then the system will determine that the abnormal data of displacement sensor group A is a displacement fluctuation caused by a sensor fault and exclude it from local anomaly events, thereby avoiding false alarms. Conversely, if the self-diagnosis results of displacement sensor group A show that it is working normally, and the cross-comparison results show that the adjacent displacement sensor groups B and C also show an accelerated displacement trend associated with displacement sensor group A, then the system will maintain the accelerated displacement trend of displacement sensor group A as a local abnormal event state, and continue to carry out subsequent linkage search and early warning generation to ensure timely response to the real deep foundation pit instability risk.

[0170] This application further proposes a building safety data analysis method based on a building safety database, the method comprising:

[0171] The displacement data stream from the displacement sensor group is received in real time, and the displacement data stream from the displacement sensor group is segmented and buffered.

[0172] Calculate the packet loss rate and short-term fluctuation amplitude of the displacement data stream segments for each displacement sensor group;

[0173] Set short-time windows and long-time windows;

[0174] Determine whether the packet loss rate or short-term fluctuation within a short time window exceeds a preset first threshold, and mark the instantaneous anomaly based on the determination result;

[0175] Determine whether the frequency or duration of transient anomalies within a long time window exceeds a preset second threshold, and mark persistent anomalies based on the determination result;

[0176] Based on the marking results of instantaneous or persistent anomalies, a decision is made on whether to trigger a displacement sensor fault warning.

[0177] Specifically, real-time reception of displacement data streams from a group of displacement sensors refers to the system continuously acquiring displacement data generated by each group of displacement sensors. Segmenting and buffering the displacement data stream from the sensor group can be understood as dividing the continuously arriving data stream into segments according to a preset time length or data volume and temporarily storing them to facilitate subsequent data processing and analysis. For example, the data can be buffered and segmented every minute or every 100 data packets collected.

[0178] The packet loss rate for each displacement sensor group's displacement data stream segment is calculated as a percentage, representing the difference between the expected number of received packets and the actual number of received packets within a specific data stream segment. A high packet loss rate typically indicates a problem with the communication link. Short-term fluctuations can be understood as the range or dispersion of displacement data values ​​within a specific data stream segment over a short period. This can be measured, for example, by calculating the standard deviation, mean absolute deviation, or the difference between the maximum and minimum values. Excessive short-term fluctuations may indicate noise or instability within the sensor itself.

[0179] In practical applications, setting short and long time windows aims to simultaneously capture both transient and persistent data stream anomalies. Short time windows are typically short, such as a few seconds to a few minutes, used for rapid response to sudden data quality issues. Long time windows are relatively long, such as a few minutes to a few hours, used to accumulate and assess the persistence and frequency of data quality problems.

[0180] Furthermore, it is determined whether the packet loss rate or short-term fluctuation amplitude within a short time window exceeds a preset first threshold, and a transient anomaly is marked based on the determination result. The first threshold is a critical value set for the packet loss rate and short-term fluctuation amplitude. Once either indicator exceeds this threshold within the short time window, it is judged as a transient anomaly. For example, if the packet loss rate exceeds 5% or the short-term fluctuation amplitude exceeds 0.1 mm within the short time window, it is marked as a transient anomaly.

[0181] Based on this, it is determined whether the frequency or duration of transient anomalies within a long time window exceeds a preset second threshold, and persistent anomalies are marked according to the determination result. The second threshold is set for the cumulative effect of transient anomalies, aiming to identify those anomalies that, although not severe individually, occur frequently or last for a long time. For example, if a transient anomaly occurs more than 3 times or its cumulative duration exceeds 10 minutes within a long time window, it is marked as a persistent anomaly.

[0182] Ultimately, based on the flagging results of transient or persistent anomalies, a decision is made as to whether to trigger a displacement sensor fault warning. This means that if a transient or persistent anomaly is detected, the system will issue a warning, indicating a possible displacement sensor fault, thereby prompting maintenance personnel to conduct inspections and maintenance.

[0183] In some preferred embodiments, it is assumed that a group of displacement sensors within a deep foundation pit monitoring area is continuously transmitting displacement data. The system receives the displacement data stream from the group of displacement sensors in real time and buffers it in segments of one minute.

[0184] At a certain moment, the system calculates that the packet loss rate within the current one-minute segment is 8%, while the preset first threshold is 5%. Since 8% exceeds 5%, the system immediately marks it as a transient anomaly. Simultaneously, the system also calculates that the short-term fluctuation amplitude of this segment is 0.2 mm, while the preset first threshold is 0.15 mm. Since 0.2 mm exceeds 0.15 mm, this further confirms the occurrence of the transient anomaly.

[0185] The system continuously monitored and detected five similar transient anomalies occurring in different minute segments within the next hour (long-term window), exceeding the preset second threshold of three occurrences. Furthermore, the cumulative duration of these transient anomalies reached 20 minutes, while the preset second threshold was 10 minutes. Given that both the frequency and duration of the transient anomalies exceeded the second threshold, the system determined this to be a persistent anomaly.

[0186] Therefore, based on the persistent anomaly marker, the system immediately triggered a displacement sensor fault warning and notified maintenance personnel that the displacement sensor group might be malfunctioning. Upon receiving the warning, maintenance personnel can quickly inspect the displacement sensor group, checking its power supply, communication links, or internal status, thereby promptly troubleshooting the fault and ensuring the accuracy and reliability of subsequent displacement data. This prevents sensor malfunctions from affecting the judgment of the actual displacement trend of the deep foundation pit.

[0187] This application further proposes the following steps for receiving displacement data streams from a group of displacement sensors in real time and for segmenting and buffering the displacement data streams from the group of displacement sensors:

[0188] Real-time reception of displacement data streams from a group of displacement sensors;

[0189] Detect the wireless signal strength or multipath effect of the displacement data stream of a displacement sensor group and obtain the real-time signal quality;

[0190] When the wireless signal strength is lower than the preset strength threshold or the multipath effect is enhanced, the adaptive data reception mode is activated.

[0191] In adaptive data reception mode, the backup communication link is enabled or the data transmission protocol is adjusted to improve the stability of the displacement data stream of the displacement sensor group.

[0192] Based on the real-time signal quality, the segmentation and buffering strategy of the displacement data stream of the displacement sensor group is dynamically adjusted to improve the continuity of the displacement data stream of the displacement sensor group.

[0193] Perform integrity verification on the received data packets;

[0194] For incomplete or erroneous data packets, retransmission requests are made or local data interpolation is used to restore the data, thereby improving the continuity of the displacement data stream of the displacement sensor group.

[0195] The displacement data stream of the displacement sensor group is segmented and buffered.

[0196] Specifically, real-time reception of displacement data streams from a group of displacement sensors refers to the system continuously receiving displacement data generated by various displacement sensor groups within the deep foundation pit monitoring area. This data is typically transmitted via a wireless communication network in the form of data packets. Detecting the wireless signal strength or multipath effect of the displacement data stream from the sensor group and determining the real-time signal quality can be understood as the system simultaneously monitoring the quality parameters of the wireless communication link during data reception, such as received signal strength indication, signal-to-noise ratio, and the degree of signal fading and distortion caused by multipath propagation. These parameters comprehensively reflect the real-time signal quality of the current data transmission. In practical applications, when the wireless signal strength falls below a preset strength threshold or the multipath effect intensifies, the system automatically activates an adaptive data reception mode. This mode is designed to cope with harsh communication environments and ensure the reliability of data transmission.

[0197] In adaptive data reception mode, backup communication links can be enabled, such as switching from the primary wireless channel to an auxiliary wireless channel, or switching from wireless communication to a wired backup link (if available). Simultaneously, the data transmission protocol can be adjusted, for example, by reducing the data transmission rate, increasing error correction coding redundancy, or adopting a more robust modulation and demodulation method, to improve the stability of the displacement data stream from the displacement sensor group. Furthermore, based on real-time signal quality, the system dynamically adjusts the segmented buffering strategy of the displacement data stream from the displacement sensor group. For example, when signal quality is poor, the size of the buffer segment can be appropriately increased or the buffering time extended to cope with data transmission delays or intermittent interruptions, thereby improving the continuity of the displacement data stream from the displacement sensor group.

[0198] Furthermore, integrity verification of received data packets is a crucial step in ensuring data accuracy. This is typically achieved through Cyclic Redundancy Check (CRC) or other checksum algorithms to detect errors that may have occurred during transmission. For incomplete or erroneous data packets, the system will request a retransmission, requiring the sender to resend the lost or corrupted packets. If retransmission is not feasible or inefficient, local data interpolation recovery techniques can be used to estimate and fill in missing data based on preceding and following normal data points, thereby improving the continuity of the displacement data stream from the displacement sensor group. Finally, the displacement data stream processed in the above manner will be segmented and buffered, providing a stable, continuous, and high-quality data foundation for subsequent fault early warning analysis.

[0199] In some preferred embodiments, it is assumed that the displacement data stream of a displacement sensor group is transmitted via a wireless network within the deep foundation pit monitoring area. When the monitoring system detects a sudden drop in the wireless signal strength of the displacement sensor group from -60dBm to -85dBm, accompanied by a significant increase in multipath effect, the system immediately activates an adaptive data reception mode. In this mode, the system first attempts to switch the communication link of the displacement sensor group from the primary 2.4GHz band to the backup 5.8GHz band to avoid interference from the current band. If the signal quality is still unsatisfactory after the switch, the system further adjusts the data transmission protocol, for example, reducing the data packet transmission rate from 100 packets per second to 50 packets per second and increasing the redundancy of the forward error correction code to improve the data packet loss resistance. Simultaneously, the segmented buffering strategy is dynamically adjusted, extending the duration of each buffer segment from 5 seconds to 10 seconds to tolerate longer transient network interruptions. During data reception, each data packet undergoes CRC verification. If a data packet fails verification, the system immediately initiates a retransmission request to the sender. If a retransmission request fails within the specified time, the system will estimate and recover the missing displacement data using linear or polynomial interpolation methods based on the data points successfully received before and after the data packet. Through this series of adaptive reception, transmission, and recovery mechanisms, the displacement data stream of the displacement sensor group can maintain high stability and continuity even under harsh communication conditions, ensuring the accuracy of subsequent fault warning analysis.

[0200] refer to Figure 2 This application further proposes a building safety data analysis system based on a building safety database, applied to a building safety data analysis method based on a building safety database. The system includes:

[0201] The preset module is used to group displacement sensors in the deep foundation pit monitoring area and preset the risk propagation correlation between each displacement sensor group according to the geological conditions of the deep foundation pit.

[0202] The processing module is used to continuously acquire displacement data from each displacement sensor group and calculate the instantaneous rate of change of the displacement data.

[0203] The identification module is used to identify any group of displacement sensors that has entered an accelerating displacement trend based on the instantaneous change rate of displacement data meeting the set change characteristics, and to mark the accelerating displacement trend as a local abnormal event.

[0204] The search module is used to perform a linked search on displacement sensor groups associated with local abnormal events based on preset risk propagation relationships and with increasing time windows and spatial ranges. The linked search is used to discover whether the associated displacement sensor groups show an accelerating displacement trend.

[0205] The early warning generation module is used to generate a regional linkage instability early warning if the linkage search finds that multiple displacement sensor groups with risk propagation correlations all show an accelerating displacement trend.

[0206] Among them, when the preset risk propagation correlation fails to reflect the geological conditions of the deep foundation pit, the steps for generating a regional linkage instability early warning include:

[0207] Identify localized anomalous events;

[0208] Initiate an omnidirectional neighbor group search. The omnidirectional neighbor group search is used to send probe signals to all displacement sensor groups that are physically adjacent to the displacement sensor group where the local anomaly event occurred, in order to find whether a corresponding accelerated displacement trend also appears in the time period immediately following the time point of the local anomaly event, and mark it as a subsequent local anomaly event.

[0209] For any subsequent local anomaly event found in the omnidirectional nearest neighbor group search, the spatiotemporal correlation degree between the subsequent local anomaly event and the local anomaly event is calculated. The spatiotemporal correlation degree takes into account the time interval and spatial distance between the subsequent local anomaly event and the local anomaly event.

[0210] Based on the spatiotemporal correlation, candidate risk propagation paths are established, and these paths are used to infer the direction of risk propagation.

[0211] If the risk propagation direction corresponding to a certain candidate risk propagation path is repeatedly verified in actual data, the risk propagation direction will be marked as a dynamic candidate path, and the priority of the dynamic candidate path in subsequent linked scanning will be increased.

[0212] If, within a specific time and space range, multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend, then a regional linkage instability warning will be generated by aggregation.

[0213] Specifically, the preset module can be a configuration management unit responsible for receiving geological survey data, historical settlement data, and expert experience; logically dividing the deep foundation pit monitoring area; and classifying physically adjacent or geologically similar displacement sensors into displacement sensor groups. Simultaneously, this module is also responsible for establishing and maintaining risk propagation models between groups, such as using graph theory or probabilistic models to represent the probability and direction of displacement anomaly propagation between different groups.

[0214] The processing module can be a data acquisition and processing unit that receives raw displacement data from various displacement sensor groups in real time via wireless or wired networks. This module preprocesses the received data, including data cleaning, noise reduction, and format conversion, and uses algorithms such as differential and moving average to calculate the displacement change rate, i.e., the instantaneous rate of change, for each displacement sensor group over a short period of time.

[0215] In practical applications, the identification module can be an anomaly detection engine that compares the instantaneous rate of change calculated by the processing module with a preset threshold or pattern. When the instantaneous rate of change continuously exceeds a certain critical value within a certain period of time, or exhibits a specific nonlinear growth pattern, the identification module will determine that the displacement sensor group has entered an accelerated displacement trend and mark it as a local anomaly event that requires attention.

[0216] Furthermore, the search module can be a risk diffusion analyzer, activated when the identification module reports a local anomaly. It utilizes the risk propagation relationships established by pre-defined modules, starting with the group where the local anomaly occurred, and gradually expands the search scope (spatially) and extends the observation time (temporally) to detect whether related groups also show an accelerated displacement trend, thereby determining whether the risk is spreading.

[0217] The early warning generation module can be a decision support system that receives the linked search results from the search module. When multiple displacement sensor groups along the risk propagation path are confirmed to simultaneously show an accelerating displacement trend, the early warning generation module will integrate this information, assess the risk of regional instability, and aggregate it to generate a regional linked instability early warning. This early warning may include information such as the risk level, the possible scope of impact, and recommended countermeasures.

[0218] The content disclosed above is only a preferred and feasible embodiment of the present invention, and is not intended to limit the scope of protection of the present invention. Therefore, all equivalent technical changes made based on the content of the present invention specification and drawings are included within the scope of protection of the present invention. Furthermore, the elements therein can be updated as technology develops.

Claims

1. A method for analyzing building safety data based on a building safety database, characterized in that, The method includes the following steps: Displacement sensors within the monitoring area of ​​the deep foundation pit are grouped, and risk propagation relationships between each displacement sensor group are preset based on the geological conditions of the deep foundation pit. The risk propagation relationship represents the potential connection between each displacement sensor group that, when an abnormal displacement occurs in one displacement sensor group, other displacement sensor groups will also experience displacement. This relationship is used to characterize the structural instability risk of the deep foundation pit. Continuously acquire displacement data from each displacement sensor group and calculate the instantaneous rate of change of the displacement data; Based on the instantaneous change rate of displacement data meeting the set change characteristics, it is identified that any group of displacement sensors has entered an accelerated displacement trend, and the accelerated displacement trend is marked as a local abnormal event. Based on local anomalies, and according to the preset risk propagation correlation, a linkage search is performed on the displacement sensor group associated with the local anomalies with an increasing time window and spatial range. The linkage search is used to discover whether the associated displacement sensor group shows an accelerated displacement trend. If the linked search finds that multiple displacement sensor groups with risk propagation correlations all show an accelerating displacement trend, then an aggregated regional linked instability warning will be generated. Among them, when the preset risk propagation correlation fails to reflect the geological conditions of the deep foundation pit, the steps for generating a regional linkage instability early warning include: Identify localized anomalous events; Initiate an omnidirectional neighbor group search. The omnidirectional neighbor group search is used to send probe signals to all displacement sensor groups that are physically adjacent to the displacement sensor group where the local anomaly event occurred, in order to find whether a corresponding accelerated displacement trend also appears in the time period immediately following the time point of the local anomaly event, and mark it as a subsequent local anomaly event. For any subsequent local anomaly event found in the omnidirectional nearest neighbor group search, the spatiotemporal correlation degree between the subsequent local anomaly event and the local anomaly event is calculated. The spatiotemporal correlation degree takes into account the time interval and spatial distance between the subsequent local anomaly event and the local anomaly event. Based on the spatiotemporal correlation, candidate risk propagation paths are established, and these paths are used to infer the direction of risk propagation. If the risk propagation direction corresponding to a certain candidate risk propagation path is repeatedly verified in actual data, the risk propagation direction will be marked as a dynamic candidate path, and the priority of the dynamic candidate path in subsequent linked scanning will be increased. If, within a specific time and space range, multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend, then an early warning of regional linkage instability is generated by aggregation.

2. The building safety data analysis method based on a building safety database as described in claim 1, characterized in that, The method also includes: Based on the aggregated regional linkage instability warning, corresponding regional linkage instability warning information is generated. Real-time acquisition of dynamic environmental data at the deep foundation pit construction site, including real-time rainfall, surrounding construction vibration intensity, and changes in groundwater pumping rate; Correlation analysis will be performed between dynamic environmental data and regional linkage instability early warning information. Based on the correlation analysis results, the urgency of the regional linkage instability warning is revised or the potential impact range of the regional linkage instability warning is enhanced.

3. The building safety data analysis method based on a building safety database as described in claim 2, characterized in that, Regional linkage instability early warning information includes a unique identifier for the warning, the time of occurrence, the location of occurrence, the warning type, and the initial risk level, which is used to improve the standardization of subsequent correlation analysis.

4. The building safety data analysis method based on a building safety database as described in claim 1, characterized in that, The steps for grouping displacement sensors within the deep foundation pit monitoring area and pre-setting risk propagation relationships between each displacement sensor group based on the geological conditions of the deep foundation pit include: Real-time acquisition of geological monitoring data at deep foundation pit construction sites; geological monitoring data indicates changes in geological conditions. Identify geological condition change events based on geological monitoring data; Based on geological condition change events, assess the impact of geological condition changes on the grouping of displacement sensors; Based on the assessment results, the division of displacement sensor groups in the monitoring area of ​​the affected deep foundation pit will be adjusted. Based on the adjusted displacement sensor group division and changes in geological conditions, the risk propagation correlation between each displacement sensor group is updated; The updated displacement sensor grouping and its corresponding risk propagation relationships will be used in subsequent steps.

5. The building safety data analysis method based on a building safety database as described in claim 1, characterized in that, The steps to identify any group of displacement sensors that has entered an accelerating displacement trend and to mark the accelerating displacement trend as a local anomalous event include: Real-time acquisition of local environmental data corresponding to the location of the displacement sensor group; Match local environmental data with identified acceleration displacement trends; If the matching results show that the accelerated displacement trend matches the non-structural displacement response characteristics caused by environmental disturbances, then the warning priority of the accelerated displacement trend should be reduced. If the matching results show that the accelerated displacement trend has a low correlation with environmental disturbances, then the local abnormal event state of the accelerated displacement trend is maintained.

6. The building safety data analysis method based on a building safety database as described in claim 1, characterized in that, The steps to identify any group of displacement sensors that has entered an accelerating displacement trend and to mark the accelerating displacement trend as a local anomalous event include: Identify the geological strata where the displacement sensor group is located; Obtain information on the current construction phase; Obtain the environmental conditions corresponding to the location of the displacement sensor group; Based on geological strata, current construction stage information, and environmental conditions, adjust the judgment threshold and duration standard for identifying accelerated displacement trends; When the displacement data of the displacement sensor group meets the adjusted judgment threshold and duration standard, the accelerated displacement trend will be marked as a local abnormal event.

7. The building safety data analysis method based on a building safety database as described in claim 1, characterized in that, The steps to identify any group of displacement sensors that has entered an accelerating displacement trend and to mark the accelerating displacement trend as a local anomalous event include: Real-time monitoring of the continuity and stability of displacement data stream from the displacement sensor group; When monitoring detects intermittent interruptions, sudden jumps in data values, or prolonged deviations from the normal operating range in the displacement sensor group's displacement data stream, the displacement sensor self-diagnosis program is activated to check the power supply, communication links, and internal status of the displacement sensor group, thereby identifying the abnormal displacement sensor group. Cross-compare the displacement data of the abnormal displacement sensor group with those of the adjacent displacement sensor group; If the cross-comparison results show that the displacement data of adjacent displacement sensor groups are normal and no corresponding acceleration displacement trend is displayed, the anomaly will be marked as displacement fluctuation caused by displacement sensor failure and excluded from local anomaly events. If the self-diagnostic results show that the displacement sensor group is working normally, and the cross-comparison results show that adjacent displacement sensor groups are associated with an accelerating displacement trend, then the local abnormal event state of the accelerating displacement trend is maintained.

8. The building safety data analysis method based on a building safety database as described in claim 7, characterized in that, The method includes: The displacement data stream from the displacement sensor group is received in real time, and the displacement data stream from the displacement sensor group is segmented and buffered. Calculate the packet loss rate and short-term fluctuation amplitude of the displacement data stream segments for each displacement sensor group; Set short-time windows and long-time windows; Determine whether the packet loss rate or short-term fluctuation within a short time window exceeds a preset first threshold, and mark the instantaneous anomaly based on the determination result; Determine whether the frequency or duration of transient anomalies within a long time window exceeds a preset second threshold, and mark persistent anomalies based on the determination result; Based on the marking results of instantaneous or persistent anomalies, a decision is made on whether to trigger a displacement sensor fault warning.

9. A building safety data analysis method based on a building safety database as described in claim 8, characterized in that, The steps of receiving displacement data streams from a group of displacement sensors in real time and segmenting and buffering the displacement data streams from the group of displacement sensors include: Real-time reception of displacement data streams from a group of displacement sensors; Detect the wireless signal strength or multipath effect of the displacement data stream of a displacement sensor group and obtain the real-time signal quality; When the wireless signal strength is lower than the preset strength threshold or the multipath effect is enhanced, the adaptive data reception mode is activated. In adaptive data reception mode, the backup communication link is enabled or the data transmission protocol is adjusted to improve the stability of the displacement data stream of the displacement sensor group. Based on the real-time signal quality, the segmentation and buffering strategy of the displacement data stream of the displacement sensor group is dynamically adjusted to improve the continuity of the displacement data stream of the displacement sensor group. Perform integrity verification on the received data packets; For incomplete or erroneous data packets, retransmission requests are made or local data interpolation is used to restore the data, thereby improving the continuity of the displacement data stream of the displacement sensor group. The displacement data stream of the displacement sensor group is segmented and buffered.

10. A building safety data analysis system based on a building safety database, applied to the building safety data analysis method based on a building safety database as described in claim 1, characterized in that, The system includes: The preset module groups the displacement sensors in the deep foundation pit monitoring area and presets the risk propagation correlation between each displacement sensor group according to the geological conditions of the deep foundation pit. The risk propagation correlation represents the potential connection between each displacement sensor group that when one displacement sensor group experiences abnormal displacement, it will cause other displacement sensor groups to also experience displacement. This is used to characterize the structural instability risk of the deep foundation pit. The processing module continuously acquires displacement data from each displacement sensor group and calculates the instantaneous rate of change of the displacement data. The identification module identifies any group of displacement sensors that has entered an accelerating displacement trend based on the instantaneous change rate of displacement data meeting the set change characteristics, and marks the accelerating displacement trend as a local abnormal event. The search module, based on local anomalies, performs a linked search on displacement sensor groups associated with local anomalies according to preset risk propagation relationships, with increasing time windows and spatial ranges. The linked search is used to discover whether the associated displacement sensor groups show an accelerating displacement trend. If the linkage search finds that multiple displacement sensor groups with risk propagation correlations all show an accelerating displacement trend, the early warning generation module will aggregate and generate a regional linkage instability early warning. Among them, when the preset risk propagation correlation fails to reflect the geological conditions of the deep foundation pit, the steps for generating a regional linkage instability early warning include: Identify localized anomalous events; Initiate an omnidirectional neighbor group search. The omnidirectional neighbor group search is used to send probe signals to all displacement sensor groups that are physically adjacent to the displacement sensor group where the local anomaly event occurred, in order to find whether a corresponding accelerated displacement trend also appears in the time period immediately following the time point of the local anomaly event, and mark it as a subsequent local anomaly event. For any subsequent local anomaly event found in the omnidirectional nearest neighbor group search, the spatiotemporal correlation degree between the subsequent local anomaly event and the local anomaly event is calculated. The spatiotemporal correlation degree takes into account the time interval and spatial distance between the subsequent local anomaly event and the local anomaly event. Based on the spatiotemporal correlation, candidate risk propagation paths are established, and these paths are used to infer the direction of risk propagation. If the risk propagation direction corresponding to a certain candidate risk propagation path is repeatedly verified in actual data, the risk propagation direction will be marked as a dynamic candidate path, and the priority of the dynamic candidate path in subsequent linked scanning will be increased. If, within a specific time and space range, multiple groups of displacement sensors associated along dynamic candidate paths are found to exhibit an accelerating displacement trend, then an early warning of regional linkage instability is generated by aggregation.

Images