A soft soil foundation pit risk assessment system and method

Abstract

The application discloses a soft soil foundation pit risk assessment system and method, relates to the technical field of geological assessment, and can quantitatively analyze different influence sensitivities of construction rhythm changes on responses such as ground surface settlement, enclosure structure lateral displacement, pore water pressure and support internal force by synchronizing and abnormally correcting a construction log sequence Log and a monitoring data sequence Mon, calculating a rhythm disturbance index Cpi, and then establishing a sensitivity mapping library Map from the rhythm disturbance index Cpi to a multi-source response increment Res. Trigger signals Trig and control instructions Cmd can be generated by calculating a non-stability risk index Rin based on the sensitivity mapping library Map and combining grading control standards, so that a safety management and control response can be automatically triggered in an early risk formation stage. On-site managers can adjust the construction rhythm or reinforce the support in advance during the soft soil foundation pit construction process, so that the whole-process intelligent perception and prospective control of the soft soil non-stability risk caused by the construction rhythm disturbance are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geological evaluation, in particular to a soft soil foundation pit risk assessment system and method. BACKGROUND

[0002] With the continuous acceleration of urbanization construction process, the proportion of large underground engineering, subway interval tunnel and urban pipe gallery projects in soft soil layer construction is increasing year by year. Due to the loose structure, high water content and complex stress history of soft soil foundation pit, it often shows significant deformation sensitivity and time characteristics during construction stage, and slight disturbance may cause uneven settlement or supporting structure instability.

[0003] At present, in the soft soil foundation pit engineering, the risk assessment method mainly takes monitoring data as the core, such as ground settlement, supporting structure internal force, pore water pressure and other parameters. These data can reflect the physical change trend of stratum, but lack of systematic quantification of human disturbance factors such as construction rhythm, operation density and process intervention. In reality, construction units often adjust the construction rhythm due to the construction period, resource scheduling or weather changes, resulting in insufficient stress recovery of soil body or incomplete dissipation of seepage field disturbance, thus producing hidden risk accumulation. Since the existing risk assessment model does not introduce the dynamic variable of "construction rhythm disturbance", when the construction rhythm appears unbalanced change, the system cannot reflect the short-term volatility of risk in real time, so that the monitoring and evaluation results lag behind the actual response on site, causing "blind area" in safety control. This reflects the key defect that the existing method ignores the influence of "human operation rhythm on soil body response". SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a soft soil foundation pit risk assessment system and method, which solves the problems mentioned in the background art.

[0005] To achieve the above purpose, the present application is realized by the following technical scheme: a soft soil foundation pit risk assessment method, comprising the following steps:

[0006] S1, collecting the construction log sequence Log and the monitoring data sequence Mon of the soft soil foundation pit area; and performing time synchronization, missing data compensation and abnormal correction to form an aligned data framework Frm;

[0007] S2, extracting rhythm change characteristics based on the aligned data framework Frm, constructing a rhythm feature vector Ryt, and generating a rhythm disturbance index Cpi in a preset sliding time window, taking the statistical mean and variance of the historical interval of the soft soil foundation pit as the dynamic reference;

[0008] S3, establishing a sensitivity mapping library Map from the rhythm disturbance index Cpi to the multi-source response increment Res through correlation and lag analysis;

[0009] S4, calculate the non-stable comprehensive index based on the sensitivity mapping library Map, generate the non-stability risk index Rin, input into the preset hierarchical control standard for processing, generate the trigger signal Trig and output the control instruction Cmd for issuing and notifying.

[0010] Preferably, the S1 comprises S11 and S12.

[0011] S11, based on the construction information automatic recording method, through the construction management system, the operation recording terminal and the construction scheduling system, the operation process, the equipment operation state, the number of personnel input and the shutdown period information of the construction site are collected in real time;

[0012] At the same time of real-time collection, time stamp automatic calibration technology is used to position the time of each operation process, and the structured construction log sequence Log is generated according to the unified data structure coding.

[0013] Each record in the construction log sequence Log contains operation start time, operation end time, process category, construction beat, shutdown duration and operation team number.

[0014] S12, through the distributed monitoring network arranged on site of the soft soil foundation pit, the real-time physical quantities of stratum and structure are obtained through multiple types of sensor nodes, including: settlement sensor, used for obtaining vertical settlement displacement of ground surface and pit bottom; inclinometer sensor, used for obtaining horizontal lateral displacement of enclosure structure; pore water pressure sensor, used for obtaining change of pore water pressure in soft soil layer; support internal force meter sensor, used for obtaining change of axial force of support member; seepage pressure sensor, used for obtaining seepage pressure of underground water at bottom of foundation pit and outside of enclosure body; inclination acceleration composite sensor, used for obtaining inclination change and instantaneous vibration response of enclosure structure; surface strain gauge sensor, used for obtaining surface strain change of enclosure wall or support structure.

[0015] All sensor nodes are connected with on-site data centralized module through wired data bus or wireless data acquisition unit, and automatically record measurement values according to set sampling period.

[0016] Further based on the distributed data aggregation method, the data stream output by each sensor node is time-calibrated and synchronously sampled to form the monitoring data sequence Mon.

[0017] After the collection of the construction log sequence Log and the monitoring data sequence Mon is completed, based on the unified time coding method, the time identification of the construction log sequence Log and the monitoring data sequence Mon is standardized, and the repeated and abnormal time records are eliminated to generate the multi-source structured raw data set Raw.

[0018] Preferably, the S1 further comprises S13 and S14.

[0019] S13, performing event interpolation expansion processing on the non-continuous construction event nodes based on a linear interpolation mapping method, with the timestamps in the construction log sequence Log as a main time axis;

[0020] The event interpolation expansion processing forms a continuous construction event sequence by inserting linear time points between adjacent work periods.

[0021] Resampling processing is performed on the monitoring data sequence Mon based on a sliding weighted interpolation method, with the timestamps in the construction log sequence Log as a main time axis alignment reference;

[0022] During the resampling processing, an equally spaced sampling window is established on the construction time axis, and the measurement data of each sensor in the window is weighted and averaged according to the sampling weight.

[0023] The processed construction log sequence Log and the monitoring data sequence Mon are time-mapped and integrated to generate a time-aligned data set Syn.

[0024] S14, performing data compensation processing on the short-time missing sections in the time-aligned data set Syn based on a local linear regression prediction method;

[0025] The data compensation processing takes the change trend of adjacent time slices as a fitting baseline.

[0026] After the data compensation processing is completed, the abrupt abnormal points in the time-aligned data set Syn are identified based on a sliding median deviation identification method, and the abnormal sections are smoothed and corrected using a local smoothing correction method.

[0027] After the data compensation processing and smoothing correction processing, the time-aligned data set Syn is renamed as the aligned data framework Frm.

[0028] Preferably, the S2 includes S21.

[0029] S21, using a segmented statistical method to segment the aligned data framework Frm based on the work start time and work end time recorded in the construction log sequence Log in the aligned data framework Frm, extracting rhythm-related features and constructing a rhythm feature vector Ryt.

[0030] The construction of the rhythm feature vector Ryt is completed through steps S211, S212 and S213.

[0031] S211, taking the work start time and work end time of the construction log sequence Log as the segmentation boundaries, establishing a continuous and non-overlapping segmentation set for the aligned data framework Frm.

[0032] S212, in each segment in the segment set, four rhythm features are calculated based on time series statistical method: work duration feature Dur, process switching frequency feature Frq, downtime interval feature Int and downtime interval feature Int;

[0033] wherein, the work duration feature Dur is obtained by calculating the average value of work duration in the segment;

[0034] The process switching frequency feature Frq is obtained by counting the number of process category switching in the segment and converting it into frequency per unit time;

[0035] The downtime interval feature Int is obtained by calculating the average value of downtime between adjacent work in the segment;

[0036] The rhythm fluctuation amplitude feature Amp is obtained by calculating the standardized variation amplitude of construction activity intensity in the segment;

[0037] S213, according to the time sequence of the segment, the work duration feature Dur, the process switching frequency feature Frq, the downtime interval feature Int and the rhythm fluctuation amplitude feature Amp are sequentially combined to form the rhythm feature vector Ryt.

[0038] Preferably, the S2 further comprises S22;

[0039] S22, taking the rhythm feature vector Ryt as input, calculate the rhythm disturbance index Cpi representing the degree of rhythm instability, specifically through steps S221, S222, S223 and S224;

[0040] S221, based on the sliding time window difference method, perform adjacent time period difference calculation on each feature in the rhythm feature vector Ryt to obtain the feature change amount, record the rhythm change sequence set DltSet in time sequence, the rhythm change sequence set DltSet includes work duration feature change amount DurDlt, process switching frequency feature change amount FrqDlt, downtime interval feature change amount IntDlt and rhythm fluctuation amplitude feature change amount AmpDlt;

[0041] S222, based on the fluctuation degree calculation method, calculate the mean square fluctuation value in the sliding time window for the rhythm change sequence set DltSet, respectively obtain the fluctuation intensity of the four features: work duration feature fluctuation intensity DurWav, process switching frequency feature fluctuation intensity FrqWav, downtime interval feature fluctuation intensity IntWav and rhythm fluctuation amplitude feature fluctuation intensity AmpWav;

[0042] Then integrate the four fluctuation intensities to obtain the rhythm fluctuation intensity set WavSet;

[0043] S223, based on the normalized deviation method, each feature fluctuation intensity in the rhythm fluctuation intensity set WavSet is standardized, and the following is obtained: the operation duration feature normalized fluctuation value DurNor, the process switching frequency feature normalized fluctuation value FrqNor, the shutdown interval feature normalized fluctuation value IntNor and the rhythm fluctuation amplitude feature normalized fluctuation value AmpNor;

[0044] The four standardized values are integrated to form a normalized fluctuation value set NorSet in a unified dimension;

[0045] S224, based on the weighted average method, taking the normalized fluctuation value set NorSet as the weighted average basis, the preset weight value of the feature dimension is time weighted and synthesized, and the rhythm disturbance index Cpi is output.

[0046] Preferably, the S3 comprises S31;

[0047] S31, taking the aligned data framework Frm and the rhythm disturbance index Cpi as input, calculating the multi-source response increment Res of the soft soil foundation pit under different rhythm disturbances;

[0048] The multi-source response increment Res is specifically obtained by steps S311, S312 and S313;

[0049] S311, monitoring signal extraction: extracting the monitoring quantity related to the construction rhythm from the aligned data framework Frm, including the ground settlement amount Set, the enclosure structure lateral displacement amount Lat, the pore water pressure change amount Pre and the support internal force change amount For, and forming a time series set multi-source response sequence set SenSet;

[0050] S312, response change calculation: based on the time window matching method, taking the time slice corresponding to the rhythm disturbance index Cpi as the analysis window, performing adjacent time difference calculation on the multi-source response sequence set SenSet, and obtaining the response change amount: the ground settlement change amount SetRes, the enclosure structure lateral displacement change amount LatRes, the pore water pressure change amount PreRes and the support internal force change amount ForRes;

[0051] S313, normalization processing: based on the normalized deviation method, the above four response change amounts are dimensionally unified to obtain the standardized change amount: the ground settlement normalized response value SetRes, the enclosure structure lateral displacement normalized response value LatRes, the pore water pressure normalized response value PreRes and the support internal force normalized response value ForRes, and then integrated to form the multi-source response increment Res of the soft soil foundation pit under different rhythm disturbances.

[0052] Preferably, the S3 further comprises S32;

[0053] S32, based on the rhythm disturbance index Cpi and the multi-source response increment Res, a sensitivity mapping library Map of the nonlinear sensitivity mapping relationship library is established, which is used to describe the influence law of the rhythm disturbance on the multi-source response change, and is specifically completed through steps S321, S322 and S323;

[0054] S321, feature pairing: each response quantity in the rhythm disturbance index Cpi and the multi-source response increment Res is paired time by time to form a paired set of rhythm response pairing set PairSet;

[0055] The response quantities include a surface subsidence normalized response value SetRes, a building envelope lateral displacement normalized response value LatRes, a pore water pressure normalized response value PreRes, and a support internal force normalized response value ForRes;

[0056] S322, fitting modeling: based on the nonlinear least squares fitting method, taking the rhythm disturbance index Cpi as the independent variable and each response quantity as the dependent variable, a nonlinear sensitivity fitting equation is established to obtain: a surface subsidence sensitivity function SetFun, a building envelope lateral displacement sensitivity function LatFun, a pore water pressure sensitivity function PreFun, and a support internal force sensitivity function ForFun;

[0057] S323, mapping output: the four sensitivity function results are arranged into a unified structure to form a sensitivity mapping library Map.

[0058] Preferably, the S4 comprises S41;

[0059] S41, taking the rhythm disturbance index Cpi and the sensitivity mapping library Map as inputs, and combining each normalized response quantity in the multi-source response increment Res, a non-stability risk index Rin reflecting the construction non-stability risk degree is calculated, which is specifically completed through steps S411, S412 and S413;

[0060] S411, based on each sensitivity function in the sensitivity mapping library Map, including the surface subsidence sensitivity function SetFun, the building envelope lateral displacement sensitivity function LatFun, the pore water pressure sensitivity function PreFun, and the support internal force sensitivity function ForFun, the sensitivity response value at the corresponding time is calculated by taking the rhythm disturbance index Cpi as the independent variable, to obtain: a surface subsidence sensitivity response value SetSen, a building envelope lateral displacement sensitivity response value LatSen, a pore water pressure sensitivity response value PreSen, and a support internal force sensitivity response value ForSen;

[0061] And the four sensitivity response values are integrated and processed to obtain a sensitivity response set SenSet;

[0062] S412, based on the sensitivity response set SenSet and the multi-source response increment Res, the sensitivity response value is coupled and calculated with the corresponding response value based on the weighted aggregation method, and the risk aggregation value Rag is obtained;

[0063] S413, based on the interval mapping method, the risk aggregation value Rag is normalized to output the dimensionless risk representation value, marked as the non-stable risk index Rin, and the value range is [0, 1].

[0064] Preferably, the S4 further comprises S42;

[0065] S42, taking the non-stable risk index Rin as input, inputting into the preset hierarchical control standard, executing dynamic hierarchical control threshold judgment, outputting risk grade judgment result, marked as trigger signal Trig, and generating control instruction Cmd according to the trigger signal Trig and issuing notification;

[0066] The trigger signal Trig is obtained by the following determination method:

[0067] When the non-stable risk index Rin is less than or equal to 0.33, it is determined as low risk grade, and the trigger signal Trig=L1 is outputted;

[0068] When 0.33 is less than the non-stable risk index Rin and less than or equal to 0.67, it is determined as medium risk grade, and the trigger signal Trig=L2 is outputted;

[0069] When the non-stable risk index Rin is greater than 0.67, it is determined as high risk grade, and the trigger signal Trig=L3 is outputted;

[0070] The control instruction Cmd is generated by the following control measure matching method:

[0071] Trigger signal Trig=L1: maintain the current construction rhythm and monitoring frequency;

[0072] Trigger signal Trig=L2: prompt to reduce the construction time and double the monitoring frequency;

[0073] Trigger signal Trig=L3: suspend the disturbance construction source operation, and execute emergency unloading and reinforcement measures.

[0074] A soft soil foundation pit risk assessment system comprises a foundation pit data acquisition and processing module, a foundation pit feature generation module, a foundation pit risk analysis module and a foundation pit state triggering module.

[0075] The foundation pit data acquisition and processing module acquires the construction log sequence Log and the monitoring data sequence Mon of the soft soil foundation pit area, and performs time synchronization, missing compensation and abnormal correction to form an aligned data framework Frm.

[0076] The foundation pit feature generation module extracts rhythm change features based on the alignment data framework Frm, and constructs a rhythm feature vector Ryt;In a preset sliding time window, the statistical mean and variance of the soft soil foundation pit history interval are used as the dynamic reference to generate the rhythm disturbance index Cpi;

[0077] The foundation pit risk analysis module establishes a sensitivity mapping library Map of the rhythm disturbance index Cpi to the multi-source response increment Res through correlation and lag analysis;

[0078] The foundation pit state triggering module calculates the non-stability comprehensive index based on the sensitivity mapping library Map, generates a non-stability risk index Rin, inputs it into the preset hierarchical control standard for processing, generates a trigger signal Trig and outputs a control instruction Cmd for issuing a notification.

[0079] The present application provides a kind of soft soil foundation pit risk assessment system and method, with the following beneficial effects:

[0080] (1) By synchronizing and abnormal correction of construction log sequence Log and monitoring data sequence Mon, the rhythm disturbance index Cpi is calculated, which can quantify the disturbance degree of different construction shifts, equipment operation rhythm and intermittent shutdown on the mechanical state of soil, and by establishing the sensitivity mapping library Map of the rhythm disturbance index Cpi to the multi-source response increment Res, the different influence sensitivity of construction rhythm change on surface subsidence, enclosure structure lateral displacement, pore water pressure and support internal force response can be quantitatively analyzed, and the mapping relationship between risk source and receptor is realized;By calculating the non-stability risk index Rin based on the sensitivity mapping library Map and generating the trigger signal Trig and control instruction Cmd in combination with the hierarchical control standard, the safety control response can be automatically triggered in the early stage of risk formation, so that the construction rhythm or reinforcement support can be adjusted in advance by the site management personnel during the construction process of soft soil foundation pit, thereby realizing the whole process intelligent perception and prospective control of soft soil non-stability risk caused by construction rhythm disturbance.

[0081] (2) By establishing the nonlinear sensitivity mapping relationship between the rhythm disturbance index Cpi and the multi-source response increment Res of soft soil foundation pit, the sensitivity mapping library Map is generated to realize the quantitative description of soft soil system response law under different rhythm disturbance, and the instantaneous change trend of surface subsidence sensitivity function SetFun and pore water pressure sensitivity function PreFun corresponding to the rhythm disturbance index Cpi is calculated, if the curve appears significantly steep or inflection point, it means that the soil pore pressure response amplification effect is occurring. Prompt the site monitoring personnel to strengthen the support axial force observation or prolong the precipitation interval, so as to intervene in advance in the early stage of disturbance leading to response amplification, and realize the active risk intervention based on sensitivity.

[0082] (3) The risk aggregation value Rag which comprehensively reflects the joint effect of disturbance intensity and response amplitude is obtained. After interval mapping and normalization processing, the non-stability risk index Rin with a value range of [0, 1] is output, which is used to quantitatively represent the overall non-stability risk level of the system, and the risk grade division and control triggering are realized through the threshold interval. This grading response mechanism changes the risk control measures from passive manual response to automatic determination and adaptive intervention, significantly shortens the time lag from risk identification to control implementation, and improves the safety management efficiency and response accuracy of the foundation pit construction stage. BRIEF DESCRIPTION OF DRAWINGS

[0083] Figure 1 A soft soil foundation pit risk assessment method step schematic diagram is provided in the present application.

[0084] Figure 2 A soft soil foundation pit risk assessment system block diagram schematic diagram is provided in the present application. DETAILED DESCRIPTION

[0085] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0086] Embodiment 1: The present application provides a soft soil foundation pit risk assessment method, please refer to Figure 1 , comprising the following steps:

[0087] S1, collecting the construction log sequence Log and the monitoring data sequence Mon of the soft soil foundation pit area; and performing time synchronization, missing data compensation and abnormal correction to form an aligned data framework Frm;

[0088] S2, extracting rhythm change features based on the aligned data framework Frm, constructing a rhythm feature vector Ryt; in a preset sliding time window, taking the statistical mean and variance of the historical interval of the soft soil foundation pit as a dynamic reference, generating a rhythm disturbance index Cpi;

[0089] S3, establishing a sensitivity mapping library Map from the rhythm disturbance index Cpi to the multi-source response increment Res through correlation and lag analysis;

[0090] S4, calculating a non-stability comprehensive index based on the sensitivity mapping library Map, generating a non-stability risk index Rin, inputting it into a preset grading control standard for processing, generating a trigger signal Trig and outputting a control instruction Cmd for notification.

[0091] In this embodiment, by synchronizing and correcting the abnormality of the construction log sequence Log and the monitoring data sequence Mon in step S1, the unstructured construction behavior data and the quantitative monitoring data are unified in the time dimension for the first time, so that the construction event and the soft soil response process can be accurately corresponded; in step S2, by constructing the rhythm feature vector Ryt based on the rhythm features extracted from the aligned data framework Frm, and calculating the rhythm disturbance index Cpi, the disturbance degree of the rhythm change of different construction shifts, equipment operation rhythm and intermittent shutdown on the mechanical state of the soil can be quantified, so that the “construction rhythm non-stable effect” which is difficult to capture by traditional monitoring methods can be revealed.

[0092] In step S3, by establishing the sensitivity mapping library Map of the rhythm disturbance index Cpi to the multi-source response increment Res, the different influence sensitivities of the construction rhythm change on the responses of the ground surface settlement, the enclosure structure lateral displacement, the pore water pressure and the support internal force can be quantitatively analyzed, and the mapping relationship description between the risk source and the receptor is realized;

[0093] And in step S4, by calculating the non-stability risk index Rin based on the sensitivity mapping library Map and generating the trigger signal Trig and the control instruction Cmd combined with the grading control standard, the safety management and control response can be automatically triggered in the early stage of risk formation, so that the construction rhythm or the reinforcement support can be adjusted in advance by the site management personnel in the soft soil foundation pit construction process.

[0094] Therefore, the initiative and real-time performance of the soft soil foundation pit risk assessment can be significantly improved. Especially in typical subway section foundation pits, deep underground garages or high water level soft soil layer excavation engineering, when the rhythm disturbance index Cpi suddenly increases due to the switching of night construction shifts or the intermittent pumping operation, the system can judge the risk rising trend in real time through the non-stability risk index Rin and automatically generate the control instruction Cmd for risk warning, so as to avoid the safety hazards such as abnormal rise of pore pressure or excessive lateral displacement of enclosure structure caused by sudden change of manual operation rhythm, and realize the whole process intelligent perception and forward control of the soft soil non-stable risk caused by construction rhythm disturbance.

[0095] Embodiment 2: Specifically, the S1 includes S11 and S12;

[0096] S11, based on the construction information automatic recording method, the operation process, the equipment operation state, the number of personnel input and the shutdown period information of the construction site are collected in real time through the construction management system, the operation recording terminal and the construction scheduling system;

[0097] At the same time of real-time collection, the time stamp automatic calibration technology is used to position the time of each operation process, and the structured construction log sequence Log is generated according to the unified data structure coding;

[0098] Each record in the construction log sequence Log contains the start time of the work, the end time of the work, the process category, the construction rhythm, the duration of the stop work and the work team number;

[0099] S12, through the distributed monitoring network arranged on site of the soft soil foundation pit, real-time physical quantities of the stratum and structure are acquired through multiple types of sensor nodes, including: a settlement sensor for acquiring vertical settlement displacement of the ground surface and the pit bottom; an inclinometer sensor for acquiring horizontal lateral displacement of the enclosure structure; a pore water pressure sensor for acquiring changes in pore water pressure in the soft soil layer; a support internal force meter sensor for acquiring changes in the axial force of the support member; a seepage pressure sensor for acquiring seepage pressure of underground water at the bottom of the foundation pit and outside the enclosure; an inclination acceleration composite sensor for acquiring inclination changes and instantaneous vibration responses of the enclosure structure; and a surface strain gauge sensor for acquiring surface strain changes of the enclosure wall or the support structure;

[0100] All sensor nodes are connected with the on-site data centralized module through a wired data bus or a wireless data acquisition unit and automatically record the measured values according to the set sampling period;

[0101] Based on the distributed data aggregation method, the data stream output by each sensor node is time-calibrated and synchronously sampled to form a monitoring data sequence Mon.

[0102] After the collection of the construction log sequence Log and the monitoring data sequence Mon is completed, based on the unified time coding method, the time identifiers of the construction log sequence Log and the monitoring data sequence Mon are standardized in format, and repeated and abnormal time records are eliminated to generate a multi-source structured raw data set Raw.

[0103] The S1 further includes S13 and S14;

[0104] S13, based on a linear interpolation mapping method, with the timestamp in the construction log sequence Log as the main time axis, event interpolation expansion processing is performed on non-continuous construction event nodes;

[0105] In the event interpolation expansion processing, linear time points are inserted between adjacent work periods to form a continuous construction event sequence, which is used to reflect the continuity of the construction rhythm in the time dimension.

[0106] With the timestamp in the construction log sequence Log as the main time axis alignment reference, the monitoring data sequence Mon is resampled based on a sliding weighted interpolation method;

[0107] In the resampling process, an equally spaced sampling window is established on the construction time axis, and the measurement data of each sensor in the window is weighted and averaged according to the sampling weight, so that the monitoring quantity and the construction rhythm are consistent in time.

[0108] The processed construction log sequence Log is time-mapped and integrated with the monitoring data sequence Mon to generate a time-aligned data set Syn with uniform time granularity;

[0109] S14, based on a local linear regression prediction method, data compensation processing is performed on the short-time missing section in the time-aligned data set Syn;

[0110] The data compensation processing takes the change trend of adjacent time slices as the fitting baseline, predicts and fills in the missing data points, and is used to restore the integrity of the time series;

[0111] After the data compensation processing is completed, the mutation abnormal points of the time-aligned data set Syn are identified based on a sliding median deviation identification method, and a local smoothing correction method is used to perform smoothing correction on the abnormal section, so that the data change curve conforms to the actual physical trend, and the influence of sensor drift and transient noise is reduced;

[0112] After the data compensation processing and smoothing correction processing, the time-aligned data set Syn is renamed as an aligned data framework Frm;

[0113] It should be noted that:

[0114] The aligned data framework Frm: based on the time-aligned data set Syn, after local linear regression compensation and sliding median deviation correction, a time-continuous, value-stable and physically logical data framework is formed, which is the input basis for the calculation of the rhythm feature vector Ryt and the rhythm disturbance index Cpi;

[0115] By taking the construction log sequence Log as the time alignment main shaft, and performing resampling, missing compensation and anomaly correction on the monitoring data sequence Mon, high-precision matching of construction process data and stratum response data on a unified time scale is realized, providing a stable and reliable input basis for subsequent calculation of the rhythm disturbance index Cpi.

[0116] In this embodiment, the construction log sequence Log collected based on the construction information automatic recording method and the monitoring data sequence Mon obtained by the distributed sensor network are processed by linear interpolation mapping, sliding weighted resampling and local linear regression compensation to generate the aligned data framework Frm, which effectively solves the core problem of "construction behavior and monitoring quantity are out of synchronization, time drift and sampling discontinuity exist between data" in traditional foundation pit risk assessment.

[0117] The aligned data framework Frm maintains second-level consistency in time granularity, guarantees continuous and stable characteristics in numerical level, and can truly reflect the time corresponding relationship between construction disturbance action and stratum response.

[0118] In actual engineering applications, for example, in the monitoring process of subway section foundation pit or deep foundation pit enclosure structure, when there is frequent intersection and intermittent of concrete pouring, support installation, dewatering operation and other processes, the traditional manual record often appears time dislocation or data dislocation, making it difficult to correspond the monitoring values such as ground settlement and pore water pressure to specific construction events.

[0119] After the alignment data framework Frm formed by the present step, the physical response section corresponding to each construction beat can be accurately identified. For example, in the support installation process, the system can automatically associate the "installation start-tensioning completion" interval in the construction log sequence Log with the real-time record of the "support internal force meter sensor" in the monitoring data sequence Mon, ensuring that the response relationship of construction disturbance to the mechanical state of the stratum has continuous traceability.

[0120] This not only improves the calculation accuracy of the rhythm feature vector Ryt and the rhythm disturbance index Cpi, but also significantly reduces the risk of misjudgment probability caused by data mismatch, thereby realizing high timeliness and high confidence of soft soil foundation pit construction stage risk assessment, and providing high-quality input data basis for subsequent non-stability risk index calculation.

[0121] In a specific embodiment, the S2 comprises S21;

[0122] S21, based on the work start time and work end time recorded in the construction log sequence Log in the alignment data framework Frm, using a segmented statistical method to segment the alignment data framework Frm, extracting rhythm-related features and constructing a rhythm feature vector Ryt;

[0123] The construction of the rhythm feature vector Ryt is specifically completed by steps S211, S212 and S213;

[0124] S211, taking the work start time and work end time of the construction log sequence Log as the segmentation boundary, a continuous and non-overlapping segmentation set is established for the alignment data framework Frm;

[0125] S212, in each segmentation in the segmentation set, four rhythm features are calculated based on time series statistical method: work duration feature Dur, process switching frequency feature Frq, work stop interval feature Int and work stop interval feature Int;

[0126] Wherein, the work duration feature Dur is obtained by calculating the average value of the work duration in the segmentation;

[0127] The process switching frequency feature Frq is obtained by counting the number of process category switching in the segmentation and converting it into a frequency per unit time;

[0128] The downtime interval feature Int is obtained by calculating the average of the downtime between adjacent operations in each segment;

[0129] The rhythm fluctuation amplitude feature Amp is obtained by calculating the standardized fluctuation amplitude of the construction activity intensity in each segment;

[0130] S213, in time sequence, the operation duration feature Dur, the process switching frequency feature Frq, the downtime interval feature Int and the rhythm fluctuation amplitude feature Amp are sequentially combined to form a rhythm feature vector Ryt;

[0131] It should be noted that:

[0132] The operation duration feature Dur: the operation duration in each segment is averaged to reflect the construction continuity;

[0133] The process switching frequency feature Frq: the number of process category switching times in each segment is counted and converted into a frequency per unit time to reflect the rhythm change speed;

[0134] The downtime interval feature Int: the downtime between adjacent operations in each segment is averaged to reflect the rhythm intermittence;

[0135] The rhythm fluctuation amplitude feature Amp: the standardized fluctuation amplitude of the construction activity intensity in each segment is calculated to reflect the rhythm fluctuation.

[0136] The S2 further comprises S22;

[0137] S22, taking the rhythm feature vector Ryt as input, calculates a rhythm disturbance index Cpi representing the degree of rhythm instability, which is specifically obtained through steps S221, S222, S223 and S224;

[0138] S221, based on the sliding time window difference method, performs adjacent time period difference calculation on each feature in the rhythm feature vector Ryt to obtain a feature change amount, which is recorded in time sequence to form a rhythm change sequence set DltSet, the rhythm change sequence set DltSet comprising an operation duration feature change amount DurDlt, a process switching frequency feature change amount FrqDlt, a downtime interval feature change amount IntDlt and a rhythm fluctuation amplitude feature change amount AmpDlt;

[0139] The operation duration feature change amount DurDlt is used to reflect the change rate of the operation duration in adjacent time periods;

[0140] The process switching frequency feature change amount FrqDlt is used to reflect the change amplitude of the process switching frequency over time;

[0141] The interval downtime feature change amount IntDlt is used to reflect the dynamic change of the construction intermittent time;

[0142] The rhythm fluctuation amplitude feature change amount AmpDlt is used to reflect the time change trend of the construction intensity fluctuation amplitude;

[0143] In S222, the mean square fluctuation value of the rhythm change sequence set DltSet in the sliding time window is calculated based on the fluctuation degree calculation method, and four feature corresponding fluctuation intensities are obtained: the work duration feature fluctuation intensity DurWav, the process switching frequency feature fluctuation intensity FrqWav, the downtime interval feature fluctuation intensity IntWav, and the rhythm fluctuation amplitude feature fluctuation intensity AmpWav.

[0144] The four fluctuation intensities are integrated again to obtain the rhythm fluctuation intensity set WavSet, which is used to describe the overall fluctuation energy of the rhythm in the time dimension.

[0145] In S223, the feature fluctuation intensities in the rhythm fluctuation intensity set WavSet are standardized based on the normalized deviation method, and the work duration feature normalized fluctuation value DurNor, the process switching frequency feature normalized fluctuation value FrqNor, the downtime interval feature normalized fluctuation value IntNor, and the rhythm fluctuation amplitude feature normalized fluctuation value AmpNor are obtained.

[0146] The four standardized values are integrated to form the normalized fluctuation value set NorSet in a unified dimension, which is used to ensure the comparability of different feature dimensions.

[0147] In S224, the normalized fluctuation value set NorSet is used as the weighted average basis, and the preset weight value of the feature dimension is used for time weighted synthesis to output the rhythm disturbance index Cpi based on the weighted average method.

[0148] The rhythm disturbance index Cpi comprehensively reflects the fluctuation intensity, stability level and change energy of the construction rhythm in the whole time domain, which is used for subsequent risk sensitivity mapping and non-stability risk assessment.

[0149] In this embodiment, the work duration feature Dur, the process switching frequency feature Frq, the downtime interval feature Int, and the rhythm fluctuation amplitude feature Amp are extracted by segmenting statistics, and the rhythm feature vector Ryt is formed, so that the originally discrete and empirical construction rhythm information is converted into a quantitative feature set with continuous time attribute.

[0150] Further, by the sliding time window difference calculation and fluctuation analysis method, the system automatically generates the rhythm change sequence set DltSet and the rhythm fluctuation intensity set WavSet, and finally obtains the rhythm disturbance index Cpi which comprehensively reflects the rhythm stability, thereby establishing the "disturbance energy channel" between the construction rhythm change and the soft soil response at the data level.

[0151] Compared with the traditional method of relying only on the experience of engineers to judge whether the construction is smooth, after adopting this step, the strength and duration of rhythm fluctuation can be quantitatively identified through the real-time calculation result of the rhythm disturbance index Cpi, and then the potential triggering trend of the risk can be judged.

[0152] For example, in the continuous support construction of urban subway interval foundation pit, when the phenomenon of "short-time multi-process alternating operation" or "frequent intermittent start of dewatering" occurs on site, the operation duration characteristic change amount DurDlt and the process switching frequency characteristic change amount FrqDlt will significantly increase, resulting in peak fluctuation of the rhythm disturbance index Cpi; the system can issue a rhythm abnormality signal in this stage to prompt the construction management personnel to reduce the rhythm intensity or adjust the operation timing.

[0153] This identification mechanism based on statistical feature difference and fluctuation energy evaluation changes the risk perception of soft soil foundation pit from "passive response" to "rhythm-driven active detection", and is particularly suitable for working conditions with frequent rhythm changes such as continuous dewatering, cyclic support, and night alternating construction.

[0154] Therefore, not only the sensitivity of the rhythm disturbance index Cpi to the non-stability trend is significantly improved, but also the influence mechanism of the construction rhythm change is quantitatively expressed, so that the risk assessment model has the ability to dynamically adapt to the change of construction behavior, and provides a clear disturbance input basis for the establishment of the subsequent sensitivity mapping library Map.

[0155] In a specific embodiment, the S3 comprises S31;

[0156] S31, taking the aligned data framework Frm and the rhythm disturbance index Cpi as inputs, calculating the multi-source response increment Res of the soft soil foundation pit under different rhythm disturbances;

[0157] The multi-source response increment Res is specifically obtained by steps S311, S312 and S313;

[0158] S311, monitoring signal extraction: extracting the monitoring quantities related to the construction rhythm from the aligned data framework Frm, including the ground settlement amount Set, the enclosure structure lateral displacement amount Lat, the pore water pressure change amount Pre, and the support internal force change amount For, and forming the time sequence set multi-source response sequence set SenSet;

[0159] S312, response change calculation: based on the time window matching method, taking the time slice corresponding to the rhythm disturbance index Cpi as the analysis window, performing adjacent time difference calculation on the multi-source response sequence set SenSet to obtain the response change quantity: surface subsidence change quantity SetRes, enclosure structure lateral displacement change quantity LatRes, pore water pressure change quantity PreRes and support internal force change quantity ForRes;

[0160] S313, normalization processing: based on the normalization deviation method, the four response change quantities are dimensionally unified to obtain the standardized change quantity: surface subsidence normalized response value SetRes, enclosure structure lateral displacement normalized response value LatRes, pore water pressure normalized response value PreRes and support internal force normalized response value ForRes, and then integrated to form the multi-source response increment Res of the soft soil foundation pit under different rhythm disturbances.

[0161] The S3 further includes S32;

[0162] S32, based on the rhythm disturbance index Cpi and the multi-source response increment Res, a sensitivity mapping library Map of the nonlinear sensitivity mapping relationship library is established to describe the influence law of the rhythm disturbance on the multi-source response change, which is specifically completed through steps S321, S322 and S323;

[0163] S321, feature pairing: each response quantity in the rhythm disturbance index Cpi and the multi-source response increment Res is paired time slice by time slice to form a paired set rhythm response paired set PairSet;

[0164] The response quantities include the surface subsidence normalized response value SetRes, the enclosure structure lateral displacement normalized response value LatRes, the pore water pressure normalized response value PreRes and the support internal force normalized response value ForRes;

[0165] S322, fitting modeling: based on the nonlinear least squares fitting method, taking the rhythm disturbance index Cpi as the independent variable and each response quantity as the dependent variable, nonlinear sensitivity fitting equations are established to obtain: surface subsidence sensitivity function SetFun, enclosure structure lateral displacement sensitivity function LatFun, pore water pressure sensitivity function PreFun and support internal force sensitivity function ForFun;

[0166] S323, mapping output: the four sensitivity function results are arranged into a unified structure to form the sensitivity mapping library Map;

[0167] It should be noted that:

[0168] The surface subsidence sensitivity function SetFun, the enclosure structure lateral displacement sensitivity function LatFun, the pore water pressure sensitivity function PreFun, and the support internal force sensitivity function ForFun respectively describe the sensitivity curves of the corresponding response quantities with respect to the rhythm disturbance index Cpi;

[0169] By establishing the Cpi-Res pairing relationship and constructing the sensitivity mapping library Map, the dynamic quantitative description of the influence of the rhythm disturbance change on the multi-source response of the foundation pit system is realized, and a mathematical basis is provided for subsequent non-stability risk assessment.

[0170] In this embodiment, by implementing step S3, the nonlinear sensitivity mapping relationship between the rhythm disturbance index Cpi and the multi-source response increment Res of the soft soil foundation pit is established, the quantitative modeling from the construction rhythm change to the stratum response effect is realized, and the risk formation process is changed from empirical speculation to calculable causal correlation.

[0171] The sensitivity mapping library Map realizes the quantitative description of the response law of the soft soil system under different rhythm disturbances, so that the relationship between the rhythm disturbance change and the structural response amplitude can be intuitively expressed in the form of a curve, and the problem that the influence of the disturbance intensity change on the nonlinear law of the response amplitude cannot be described in the traditional risk assessment is solved.

[0172] In practical applications, for example, during the deep foundation pit excavation construction phase, when the pumping dewatering rate, the support member prestress adjustment frequency, or the night work switching rhythm changes, the system can calculate the instantaneous change trend of the corresponding surface subsidence sensitivity function SetFun and pore water pressure sensitivity function PreFun according to the rhythm disturbance index Cpi. If the curve appears a significant steep rise or inflection point, it means that the soil pore pressure response amplification effect is occurring.

[0173] At this time, the system can automatically mark the high sensitivity area and prompt the on-site monitoring personnel to strengthen the support axial force observation or prolong the dewatering interval, so as to intervene in advance in the early stage of the disturbance leading to the response amplification, and realize the active risk intervention based on sensitivity.

[0174] Therefore, the technical implementation of this step not only makes the relationship between the rhythm disturbance index Cpi and the multi-source response increment Res calculable and interpretable, but also establishes the response mechanism model of the non-stability risk of the soft soil foundation pit, so that the risk prediction is changed from "result-oriented" to "process-driven".

[0175] The sensitivity mapping library Map generated through this step provides accurate response sensitivity parameters support for the subsequent non-stability risk index Rin calculation, and realizes the quantification, interpretation, and early warning of the soft soil foundation pit risk assessment.

[0176] Embodiment 5: Specifically, the S4 comprises S41;

[0177] S41, taking the rhythm disturbance index Cpi and the sensitivity mapping library Map as inputs, combining each normalized response quantity in the multi-source response increment Res, calculating a non-stability risk index Rin that comprehensively reflects the construction non-stable risk degree, and the calculation is completed through steps S411, S412 and S413;

[0178] S411, based on each sensitivity function in the sensitivity mapping library Map, including the surface subsidence sensitivity function SetFun, the enclosure structure lateral displacement sensitivity function LatFun, the pore water pressure sensitivity function PreFun and the support internal force sensitivity function ForFun, taking the rhythm disturbance index Cpi as the independent variable, calculating the sensitivity response value at the corresponding time, and obtaining: the surface subsidence sensitivity response value SetSen, the enclosure structure lateral displacement sensitivity response value LatSen, the pore water pressure sensitivity response value PreSen and the support internal force sensitivity response value ForSen;

[0179] And the four sensitivity response values are integrated and processed to obtain a sensitivity response set SenSet;

[0180] S412, based on the sensitivity response set SenSet and the multi-source response increment Res, based on the weighted aggregation method, the sensitivity response value and the corresponding response quantity are weighted and coupled to obtain a risk aggregation value Rag;

[0181] The risk aggregation value Rag comprehensively considers the dual influence of rhythm disturbance sensitivity and response amplitude, and reflects the comprehensive effect of construction disturbance on soil stability;

[0182] The specific calculation formula is as follows:

[0183] ;

[0184] In the formula, r1, r2, r3 and r4 respectively represent the weight coefficients of surface subsidence, lateral displacement, pore pressure and support internal force, and the specific values are set by the user, and r1+r2+r3+r4=1;

[0185] S413, based on the interval mapping method, the risk aggregation value Rag is normalized to output a dimensionless risk representation value, marked as a non-stability risk index Rin, and the value range is [0, 1], and the larger the value, the higher the system non-stability risk;

[0186] It should be noted that:

[0187] Sensitivity response set SenSet: a set composed of surface settlement sensitivity response value SetSen, enclosure lateral displacement sensitivity response value LatSen, pore water pressure sensitivity response value PreSen, and support internal force sensitivity response value ForSen, used to reflect the real-time sensitivity of each response to the rhythm disturbance;

[0188] Risk aggregation value Rag: an intermediate result calculated by the weighted coupling of sensitivity response value and response amplitude, used to reflect the coupling risk intensity between disturbance and response;

[0189] Unstable risk index Rin: a normalized comprehensive risk index, with a value range of [0, 1], used to quantitatively describe the overall instability of the foundation pit system.

[0190] The S4 further comprises S42;

[0191] S42, taking the unstable risk index Rin as input, inputting into the preset hierarchical control standard, performing dynamic hierarchical and control threshold determination, outputting risk grade determination results, marked as trigger signal Trig, and generating control instruction Cmd according to the trigger signal Trig and issuing notification;

[0192] The trigger signal Trig is specifically obtained by the following determination method:

[0193] When the unstable risk index Rin is less than or equal to 0.33, it is determined as a low risk grade, and the trigger signal Trig=L1 is outputted;

[0194] When 0.33 < unstable risk index Rin ≤ 0.67, it is determined as a medium risk grade, and the trigger signal Trig=L2 is outputted;

[0195] When the unstable risk index Rin is greater than 0.67, it is determined as a high risk grade, and the trigger signal Trig=L3 is outputted;

[0196] The control instruction Cmd is generated by the following control measure matching method:

[0197] Trigger signal Trig=L1: maintain the current construction rhythm and monitoring frequency;

[0198] Trigger signal Trig=L2: prompt to reduce the construction time and double the monitoring frequency;

[0199] Trigger signal Trig=L3: suspend the disturbance construction source operation, and execute emergency unloading and reinforcement measures.

[0200] In this embodiment, through the implementation of step S4, the whole process closed loop from rhythm disturbance identification to risk grading control is completed, the real-time quantitative determination and dynamic response control of the soft soil foundation pit unsteady risk are realized, and the risk aggregation value Rag reflecting the combined effect of disturbance intensity and response amplitude is obtained. After interval mapping and normalization processing, the unsteady risk index Rin is output, which has a value range of [0, 1] and is used to quantitatively represent the overall unsteady risk level of the system, and the risk level division and control triggering are realized through the threshold interval.

[0201] Unlike traditional manual interpretation or single monitoring index early warning, this step realizes the quantitative grading- automatic triggering- instruction linkage mechanism of foundation pit risk.

[0202] In engineering practice, for example, when continuous dewatering operation and support tensioning process cause simultaneous increase of ground surface settlement and pore water pressure in deep foundation pit support construction in soft soil layer, the risk aggregation value Rag calculated by the system will increase significantly, and then the unsteady risk index Rin with high value is output.

[0203] If the value exceeds the grading threshold value 0.67, the trigger signal Trig=L3 is output, and the system can automatically issue the control instruction Cmd, requiring to suspend the dewatering operation and execute unloading or reinforcement measures; if it is in the medium risk level (0.33

[0204] This grading response mechanism changes the risk control measures from passive manual response to automatic determination and adaptive intervention, significantly shortens the time lag from risk identification to control implementation, and improves the safety management efficiency and response accuracy of foundation pit construction stage.

[0205] Therefore, through the closed loop linkage of the unsteady risk index Rin and the trigger signal Trig-control instruction Cmd established by this step, the real-time perception, quantitative determination and automatic disposal of the unsteady risk of soft soil foundation pit are realized.

[0206] This mechanism can still maintain high-precision identification and interpretable output under the coupling working condition of multiple disturbance sources and multiple response dimensions, and provides an executable and traceable risk control means for the construction process of soft soil foundation pit, which significantly improves the active prevention and control ability and intelligent level of the overall engineering safety.

[0207] Embodiment 6: A soft soil foundation pit risk assessment system, please refer to Figure 2 , specifically: including a foundation pit data acquisition and processing module, a foundation pit feature generation module, a foundation pit risk analysis module, and a foundation pit state triggering module;

[0208] The foundation pit data acquisition processing module acquires a construction log sequence Log and a monitoring data sequence Mon of a soft soil foundation pit area; and performs time synchronization, missing data compensation and abnormal correction to form an aligned data framework Frm;

[0209] The foundation pit feature generation module extracts rhythm change features based on the aligned data framework Frm, constructs a rhythm feature vector Ryt, and generates a rhythm disturbance index Cpi based on a statistical mean and variance of a historical interval of the soft soil foundation pit as a dynamic reference within a preset sliding time window.

[0210] The foundation pit risk analysis module establishes a sensitivity mapping library Map from the rhythm disturbance index Cpi to a multi-source response increment Res through correlation and lag analysis;

[0211] The foundation pit state triggering module calculates a non-stability comprehensive index based on the sensitivity mapping library Map, generates a non-stability risk index Rin, inputs it into a preset hierarchical control standard for processing, generates a triggering signal Trig and outputs a control instruction Cmd for issuing a notification.

[0212] Although embodiments of the present application have been shown and described, it is to be understood that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for risk assessment of soft soil foundation pits, characterized in that: Includes the following steps: S1. Collect the construction log sequence Log and monitoring data sequence Mon of the soft soil foundation pit area; and perform time synchronization, missing measurement compensation and anomaly correction to form an aligned data framework Frm. S2. Based on the aligned data framework Frm, extract rhythm change features and construct rhythm feature vector Ryt; within a preset sliding time window, use the statistical mean and variance of the historical interval of soft soil foundation pit as a dynamic benchmark to generate rhythm disturbance index Cpi. S3. Establish a sensitivity mapping library Map from the rhythm disturbance index Cpi to the multi-source response increment Res through correlation and hysteresis analysis; S3 includes S31; S31. Using the aligned data frame Frm and the rhythm disturbance index Cpi as inputs, calculate the multi-source response increment Res of the soft soil foundation pit under different rhythm disturbances. The multi-source response increment Res is specifically obtained through steps S311, S312 and S313; S311. Monitoring signal extraction: Extract monitoring quantities related to the construction rhythm from the aligned data frame Frm, including surface settlement Set, lateral displacement of retaining structure Lat, pore water pressure Pre, and internal force of support For, and form a time series set multi-source response sequence set SenSet; S312. Response Change Calculation: Based on the time window matching method, the time slice corresponding to the rhythm disturbance index Cpi is used as the analysis window. Adjacent time difference calculation is performed on the multi-source response sequence set SenSet to obtain the response change: surface settlement change SetRes, retaining structure lateral displacement change LatRes, pore water pressure change PreRes, and support internal force change ForRes. S313. Normalization Processing: Based on the normalization deviation method, the four response change quantities are dimensionlessly unified to obtain standardized change quantities: normalized response value SetRes' of surface settlement, normalized response value LatRes' of retaining structure lateral displacement, normalized response value PreRes' of pore water pressure, and normalized response value ForRes' of support internal force. These are then integrated to form the multi-source response increment Res of soft soil foundation pit under different rhythm disturbances. S3 further includes S32; S32. Based on the rhythm perturbation index Cpi and the multi-source response increment Res, establish a sensitivity mapping library Map, which is used to describe the influence of rhythm perturbation on the changes in multi-source response. This is specifically accomplished through steps S321, S322 and S323. S321. Feature pairing: Pair the rhythm perturbation index Cpi with each response quantity in the multi-source response increment Res on a time-slice basis to form a pairing set, rhythm response pairing set PairSet; The response quantities include the normalized response value SetRes' of surface settlement, the normalized response value LatRes' of lateral displacement of retaining structure, the normalized response value PreRes' of pore water pressure, and the normalized response value ForRes' of support internal force; S322. Fitting Modeling: Based on the nonlinear least squares fitting method, with the rhythm disturbance index Cpi as the independent variable and each response quantity as the dependent variable, nonlinear sensitivity fitting equations are established to obtain: surface settlement sensitivity function SetFun, retaining structure lateral displacement sensitivity function LatFun, pore water pressure sensitivity function PreFun, and support internal force sensitivity function ForFun. S323. Mapping Output: Organize the results of the four sensitivity functions into a unified structure to form a sensitivity mapping library (Map); S4. Calculate the unstable comprehensive index based on the sensitivity mapping library Map, generate the unstable risk index Rin, input it into the preset graded control standard for processing, generate the trigger signal Trig, and output the control command Cmd to issue a notification.

2. The method for risk assessment of soft soil foundation pits according to claim 1, characterized in that: S1 includes S11 and S12; S11. Based on the automatic recording method of construction information, the construction management system, work record terminal and construction scheduling system are used to collect information on the work procedures, equipment operation status, number of personnel and downtime at the construction site in real time. While collecting data in real time, the system uses automatic timestamp calibration technology to locate the time of each work process and encodes it according to a unified data structure to generate a structured construction log sequence. Each record in the construction log sequence Log includes the start time of the operation, the end time of the operation, the type of operation, the construction rhythm, the duration of the downtime, and the work team number; S12. A distributed monitoring network deployed on-site in the soft soil foundation pit acquires real-time physical quantities of the strata and structure through multiple types of sensor nodes, including: settlement sensors for acquiring vertical settlement displacement at the surface and bottom of the pit; inclinometer sensors for acquiring horizontal lateral displacement of the retaining structure; pore water pressure sensors for acquiring changes in pore water pressure in the soft soil layer; support internal force gauge sensors for acquiring changes in axial force of support components; seepage pressure sensors for acquiring groundwater seepage pressure at the bottom of the foundation pit and outside the retaining structure; tilt acceleration composite sensors for acquiring tilt changes and instantaneous vibration response of the retaining structure; and surface strain gauge sensors for acquiring surface strain changes of the retaining wall or support structure. All sensor nodes are connected to the field data centralization module via a wired data bus or wireless data acquisition unit, and automatically record the measured values ​​according to the set sampling period. Then, based on the distributed data aggregation method, the data streams output by each sensor node are time-calibrated and synchronously sampled to form the monitoring data sequence Mon; After the construction log sequence Log and monitoring data sequence Mon are collected, the time identifiers of the construction log sequence Log and monitoring data sequence Mon are standardized according to the unified time coding method, and duplicate and abnormal time records are removed to generate a multi-source structured raw dataset Raw.

3. The method for risk assessment of soft soil foundation pits according to claim 2, characterized in that: S1 also includes S13 and S14; S13. Based on the linear interpolation mapping method, the timestamps in the construction log sequence Log are used as the main time axis to perform event interpolation expansion processing on non-continuous construction event nodes. Among them, the event interpolation extension process forms a continuous sequence of construction events by inserting linear time points between adjacent work periods; Using the timestamps in the construction log sequence Log as the main time axis alignment reference, the monitoring data sequence Mon is resampled based on the sliding weighted interpolation method; In the resampling process, equally spaced sampling windows are established on the construction time axis, and the measurement data of each sensor within the window are weighted and averaged according to the sampling weight. The processed construction log sequence Log and the monitoring data sequence Mon are integrated by time mapping to generate a time-aligned dataset Syn. S14. Based on the local linear regression prediction method, perform data compensation processing on the short-time missing test segments in the time-aligned dataset Syn. The data compensation process uses the changing trend of adjacent time slices as the fitting baseline; After the data compensation process is completed, the abrupt change outliers in the time-aligned dataset Syn are identified based on the sliding median deviation identification method, and the outlier segments are smoothed using the local smoothing correction method. After data compensation and smoothing correction, the time-aligned dataset Syn was renamed the aligned data frame Frm.

4. The method for risk assessment of soft soil foundation pits according to claim 3, characterized in that: S2 includes S21; S21. Based on the start time and end time of the operation recorded in the construction log sequence Log in the alignment data framework Frm, the alignment data framework Frm is segmented using a segmented statistical method to extract rhythm-related features and construct a rhythm feature vector Ryt. The construction of the rhythm feature vector Ryt is specifically completed through steps S211, S212 and S213; S211. Using the start time and end time of the work in the construction log sequence Log as the segmentation boundary, establish a continuous and non-overlapping set of segments for the aligned data frame Frm. S212. Within each segment of the segmented set, calculate four rhythmic features based on time series statistical methods: Dur of operation duration, Frq of process switching frequency, Int of downtime interval, and Amp of rhythmic fluctuation. Among them, the job duration feature Dur is obtained by calculating the average duration of the job within the segment; The process switching frequency characteristic Frq is obtained by statistically analyzing the number of process category switching times within a segment and converting it into a unit time frequency. The downtime interval feature Int is obtained by calculating the average downtime between adjacent operations within a segment. The rhythm fluctuation amplitude characteristic (Amp) is obtained by calculating the standardized variation amplitude of the intensity of construction activities within the segment; S213. According to the segmented time sequence, combine the operation duration feature Dur, process switching frequency feature Frq, downtime interval feature Int, and rhythm fluctuation amplitude feature Amp in sequence to form the rhythm feature vector Ryt.

5. The method for risk assessment of soft soil foundation pits according to claim 4, characterized in that: S2 further includes S22; S22. Using the rhythm feature vector Ryt as input, calculate the rhythm disturbance index Cpi, which represents the degree of rhythm instability. Specifically, this is obtained through steps S221, S222, S223, and S224. S221. Based on the sliding time window difference method, perform adjacent time period difference calculation on each feature in the rhythm feature vector Ryt to obtain the feature change amount, and record it in time order to form a rhythm change sequence set DltSet. The rhythm change sequence set DltSet includes the change amount of operation duration feature DurDlt, the change amount of process switching frequency feature FrqDlt, the change amount of stop interval feature IntDlt, and the change amount of rhythm fluctuation amplitude feature AmpDlt. S222. Based on the volatility calculation method, the mean square volatility value of the rhythm change sequence set DltSet is calculated within a sliding time window, and the volatility intensity corresponding to the four features is obtained respectively: the volatility intensity of the operation duration feature DurWav, the volatility intensity of the process switching frequency feature FrqWav, the volatility intensity of the downtime feature IntWav, and the volatility intensity of the rhythm volatility amplitude feature AmpWav. Then, the four wave intensities are integrated to obtain the rhythm wave intensity set WavSet; S223. Based on the normalized deviation method, the intensity of each characteristic fluctuation in the rhythm fluctuation intensity set WavSet is standardized to obtain: the normalized fluctuation value of the operation duration feature DurNor, the normalized fluctuation value of the process switching frequency feature FrqNor, the normalized fluctuation value of the downtime interval feature IntNor, and the normalized fluctuation value of the rhythm fluctuation amplitude feature AmpNor. The four standardized values ​​are integrated and processed to form a normalized fluctuation value set NorSet under a unified dimension; S224. Based on the weighted average method, using the normalized fluctuation value set NorSet as the basis for weighted average, time-weighted synthesis is performed according to the preset weight values ​​of the feature dimensions to output the rhythm disturbance index Cpi.

6. The method for risk assessment of soft soil foundation pits according to claim 1, characterized in that: S4 includes S41; S41. Using the rhythm disturbance index Cpi and the sensitivity mapping library Map as inputs, and combining the normalized response quantities in the multi-source response increment Res, calculate the instability risk index Rin, which comprehensively reflects the degree of construction instability risk. This is done through steps S411, S412 and S413. S411. Based on the sensitivity functions in the sensitivity mapping library Map, including the surface settlement sensitivity function SetFun, the retaining structure lateral displacement sensitivity function LatFun, the pore water pressure sensitivity function PreFun, and the support internal force sensitivity function ForFun, calculate the sensitivity response values ​​at the corresponding time with the rhythm disturbance index Cpi as the independent variable, and obtain: the surface settlement sensitivity response value SetSen, the retaining structure lateral displacement sensitivity response value LatSen, the pore water pressure sensitivity response value PreSen, and the support internal force sensitivity response value ForSen; The four sensitivity response values ​​are then integrated to obtain the sensitivity response set SenSet'. S412. Based on the sensitivity response set SenSet' and the multi-source response increment Res, and using a weighted aggregation method, the sensitivity response values ​​and their corresponding response quantities are weighted and coupled to obtain the risk aggregation value Rag. S413. Based on the interval mapping method, the risk aggregation value Rag is normalized to output a dimensionless risk characterization value, which is labeled as the instability risk index Rin, with a value range of [0,1].

7. The method for risk assessment of soft soil foundation pits according to claim 6, characterized in that: S4 further includes S42; S42. Using the instability risk index Rin as input, input it to the preset graded control standard, perform dynamic grading and control threshold determination, output the risk level determination result, mark it as the trigger signal Trig, and generate the control command Cmd to issue a notification based on the trigger signal Trig. The trigger signal Trig is obtained through the following determination method: When the instability risk index Rin≤0.33, it is judged as a low risk level, and the trigger signal Trig=L1 is output; When 0.33 < instability risk index Rin ≤ 0.67, it is judged as medium risk level, and the trigger signal Trig = L2 is output; When the instability risk index Rin > 0.67, it is judged as a high-risk level, and the trigger signal Trig = L3 is output. The control command Cmd is generated through the following control measure matching method: Trigger signal Trig=L1: Maintain the current construction pace and monitoring frequency; Trigger signal Trig=L2: This indicates a reduction in construction time and a doubling of monitoring frequency; Trigger signal Trig=L3: Suspend the work on the disturbance source and implement emergency unloading and reinforcement measures.

8. A soft soil foundation pit risk assessment system, applied to the soft soil foundation pit risk assessment method according to any one of claims 1 to 7, characterized in that: It includes a foundation pit data acquisition and processing module, a foundation pit feature generation module, a foundation pit risk analysis module, and a foundation pit status triggering module; The foundation pit data acquisition and processing module acquires the construction log sequence Log and the monitoring data sequence Mon of the soft soil foundation pit area; and performs time synchronization, missing measurement compensation and anomaly correction to form an aligned data framework Frm. The foundation pit feature generation module extracts rhythm change features based on the aligned data framework Frm and constructs a rhythm feature vector Ryt; within a preset sliding time window, it generates a rhythm disturbance index Cpi based on the statistical mean and variance of the historical interval of soft soil foundation pits as a dynamic benchmark. The foundation pit risk analysis module establishes a sensitivity mapping library Map from the rhythm disturbance index Cpi to the multi-source response increment Res through correlation and lag analysis; The foundation pit state triggering module calculates the comprehensive instability index based on the sensitivity mapping library Map, generates the instability risk index Rin, inputs it into the preset graded control standard for processing, generates the trigger signal Trig, and outputs the control command Cmd to issue a notification.

Images