A tunnel construction pre-control method and system based on risk sign evolution

Abstract

This invention relates to the field of tunnel construction safety control technology, and particularly to a tunnel construction pre-control method and system based on risk symptom evolution. The method includes the following steps: acquiring multi-source data to characterize the current construction section's state; identifying risk symptom events based on the multi-source data to characterize abnormal signs of construction risk in the current construction section; constructing a risk symptom evolution chain based on the risk symptom events, the chain being constructed according to a combination of target risk symptom events satisfying a preset evolutionary relationship; determining the target risk evolution branch corresponding to the risk symptom evolution chain based on a preset risk symptom evolution path model, and identifying the target stage node of the current construction section within that branch; and determining the corresponding construction pre-control measures for the current construction section based on the current target stage node within the target risk evolution branch.

Description

Technical Field

[0001] This invention relates to the field of tunnel construction safety control technology, and in particular to a tunnel construction pre-control method and system based on the evolution of risk signs. Background Technology

[0002] During tunnel construction in areas with adverse geological conditions, risks such as rock instability, sudden water inrush, collapse, and abnormal support deformation are not typically characterized by a single abnormal indicator. Instead, they evolve gradually as construction progresses, stemming from multiple risk indicators, including geological forecast anomalies, abnormal rock response, hydrological anomalies, and abnormal support stress. Because different risk sources exhibit variations in the order, combination, and evolutionary path of their corresponding risk indicators during formation, even if local anomalies are similar, their subsequent risk development direction and degree of danger may differ.

[0003] However, existing technologies usually rely on the over-limit of a single monitoring indicator, the occurrence of a single type of abnormal phenomenon, or fixed rules to trigger construction measures. They lack a comprehensive analysis of the temporal relationship, combination relationship, and evolution stage among multiple risk signs. Therefore, it is difficult to accurately identify the current risk status and to make targeted triggering and dynamic adjustment of construction measures according to different evolution processes. This can easily lead to problems such as delayed risk identification, inaccurate triggering of measures, or mismatch in the intensity of treatment. Summary of the Invention

[0004] On the one hand, the present invention provides a tunnel construction pre-control method based on the evolution of risk signs, comprising the following steps: Acquire multi-source data, which is used to characterize the current state of the construction section; Based on the multi-source data, risk warning events are identified, which are used to characterize abnormal signs of construction risks in the current construction section. Based on the aforementioned risk warning events, a risk warning evolution chain is constructed, which is built according to a combination of target risk warning events that satisfy a preset evolutionary relationship; Based on the preset risk symptom evolution path model, the target risk evolution branch corresponding to the risk symptom evolution chain is determined, and the target stage node of the current construction section in the target risk evolution branch is determined. Based on the current target stage node in the target risk evolution branch, determine the construction pre-control measures corresponding to the current construction section. The construction pre-control measures include measures to adjust at least one of the construction parameters, construction procedures, or on-site control conditions.

[0005] In some embodiments, the tunnel construction pre-control method based on risk symptom evolution further includes the following steps: After the implementation of construction pre-control measures, the risk symptom evolution chain is updated based on newly identified risk symptom events, and the construction pre-control measures are redefined based on the updated risk symptom evolution chain.

[0006] In some embodiments, identifying risk warning events based on the multi-source data includes the following steps: Based on the multi-source data, feature quantities of different data are extracted respectively; Candidate abnormal signs are determined based on at least one of the deviation magnitude, duration, and trend of the feature quantity relative to the corresponding baseline state; Within a preset time window, based on the timing and consistency of the occurrence of at least two types of candidate abnormal signs, candidate abnormal signs that meet the preset association triggering relationship are identified as risk sign events.

[0007] In some embodiments, the preset association triggering relationship includes: Within a preset time window, the at least two types of candidate abnormal signs appear in a preset order, and the candidate abnormal signs that appear later satisfy a preset response enhancement condition or a preset continuous correlation condition relative to the candidate abnormal signs that appear earlier.

[0008] In some embodiments, constructing a risk symptom evolution chain based on the risk symptom events includes the following steps: The risk warning events are sorted according to their occurrence time. Based on the preset event type transition rules, determine whether the preceding and subsequent risk symptom events meet the evolutionary transition conditions: When the preceding and following risk symptom events satisfy the evolutionary transition conditions, the evolutionary connection relationship between the preceding and following risk symptom events is determined based on the time interval between the preceding and following risk symptom events and according to at least one of the relationship between event intensity changes and the relationship between continuous states. Based on the evolutionary connection relationship, the corresponding risk symptom events are sequentially connected to construct the risk symptom evolution chain.

[0009] In some embodiments, the risk symptom evolution path model is generated by the following method: Obtain the historical risk warning event sequences corresponding to multiple historical construction sections and the actual risk results corresponding to each historical construction section; In each of the historical risk symptom event sequences, combinations of risk symptom events located within a preset time period before the actual risk outcome occur are extracted as candidate precursor evolution fragments. Based on the sequential connection of events and shared risk symptom events among the candidate precursor evolutionary segments, an initial evolutionary path network is constructed; Based on the actual risk results, candidate precursor evolution segments corresponding to the same risk results are screened, and the target risk evolution branch is determined. Based on at least one of the following: the intensity change boundary of risk symptom events within the target risk evolution branch, the event density change boundary, and the outcome proximity, the target risk evolution branch is divided into stages, and the stage nodes are determined. Based on the transfer characteristics of risk symptom events between adjacent stage nodes, determine the transfer conditions between adjacent stage nodes; Based on the target risk evolution branches, stage nodes, and transition conditions, a corresponding risk symptom evolution path model is generated.

[0010] In some embodiments, determining the target risk evolution branch corresponding to the risk symptom evolution chain based on a preset risk symptom evolution path model, and determining the target stage node of the current construction section in the target risk evolution branch, includes the following steps: The risk symptom evolution chain is matched with each target risk evolution branch in the risk symptom evolution path model, and the target risk evolution branch corresponding to the risk symptom evolution chain is determined based on the branch matching results. Based on the matching position of the terminal risk symptom event in the risk symptom evolution chain in the target risk evolution branch, and the transition conditions between adjacent stage nodes in the target risk evolution branch, the target stage node of the current construction section in the target risk evolution branch is determined.

[0011] In some embodiments, determining the target risk evolution branch corresponding to the risk symptom evolution chain based on the branch matching results includes the following steps: Determine the matching degree of event type sequence, event intensity change, and event connection relationship between the risk symptom evolution chain and each target risk evolution branch; Based on the event type sequence matching degree, event intensity change matching degree, and event connection relationship matching degree, the comprehensive matching degree corresponding to each target risk evolution branch is determined; The target risk evolution branch with the highest overall matching degree and that meets the preset matching conditions is determined as the target risk evolution branch corresponding to the risk symptom evolution chain.

[0012] In some embodiments, determining the construction pre-control measures corresponding to the current construction section based on the current target stage node in the target risk evolution branch includes the following steps: Invoke the set of measures corresponding to the target risk evolution branch; Select at least one construction control measure corresponding to the target stage node from the set of measures; At least one construction control measure selected from the screening will be combined as the construction pre-control measure for the current construction section.

[0013] On the other hand, the present invention also provides a tunnel construction pre-control system based on risk symptom evolution, including a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the tunnel construction pre-control method based on risk symptom evolution proposed in any of the above embodiments.

[0014] The tunnel construction pre-control method and system based on risk symptom evolution provided by this invention includes at least the following gains: This invention does not rely solely on a single abnormal indicator or fixed handling rules to judge construction risks. Instead, it acquires multi-source data, identifies risk symptom events, and constructs a risk symptom evolution chain. This transforms scattered and heterogeneous abnormal information into risk characterization results with temporal and evolutionary relationships, thereby more comprehensively reflecting the dynamic evolution process of construction risks in the current construction section and improving the accuracy of identifying the direction and stage of risk evolution.

[0015] Furthermore, this invention introduces a risk symptom evolution path model. By matching the risk symptom evolution chain corresponding to the current construction section with the target risk evolution branch in the model, the corresponding target risk evolution branch and target stage node are determined. This allows the risk status of the current construction section to be mapped to a specific risk evolution path and development stage, thereby improving the pertinence and rationality of the timing, triggering level, and matching of construction pre-control measures. Furthermore, after the implementation of construction pre-control measures, the present invention can update the risk symptom evolution chain based on newly identified risk symptom events and redetermine the construction pre-control measures accordingly. This enables dynamic correction of continuously changing risk symptoms during construction, avoids the gradual mismatch between the originally determined construction pre-control measures and the actual risk evolution state of the current construction section, and further improves the continuity, adaptability, and reliability of tunnel construction risk pre-control. Attached Figure Description

[0016] From the following description of embodiments in conjunction with the accompanying drawings, aspects, features, and advantages of the present invention will become clearer and more readily understood, in which: Figure 1 This is a schematic diagram of the tunnel construction pre-control method based on risk symptom evolution provided by the present invention. Figure 2 This is a schematic diagram of the risk warning event identification process provided by the present invention; Figure 3 A schematic diagram illustrating the construction process of the risk symptom evolution chain provided by this invention; Figure 4A schematic diagram illustrating the risk symptom evolution path model generation process provided by this invention; Figure 5 This is a reference diagram of an initial evolution path network provided by the present invention; Figure 6 This is a schematic diagram of the system modules of the tunnel construction pre-control method based on risk symptom evolution provided by the present invention. Detailed Implementation

[0017] In the following description, specific details such as systems, structures, and techniques are set forth for illustrative purposes and not limiting, in order to provide a thorough understanding of the embodiments of this application. Those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details.

[0018] In the description of this application, detailed descriptions of well-known systems, devices, circuits and methods have been omitted so as not to obscure the description of this application with unnecessary details; in addition, the terms "first", "second", "third", "fourth", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0019] In existing tunnel construction pre-control, due to the differences in the timing, association, and evolution path of abnormal signs corresponding to different risk sources, it is difficult to accurately identify the risk evolution status of the current construction section by relying solely on a single abnormal indicator or fixed handling rules. This can easily lead to inaccurate risk stage judgment, delayed triggering of pre-control measures, or insufficient adaptability of measures.

[0020] Please see Figure 1 In one embodiment of the present invention, a tunnel construction pre-control method based on risk symptom evolution is proposed, which specifically includes the following steps: S01. Obtain multi-source data, which is used to characterize the current construction section.

[0021] In this embodiment, the multi-source data is a data set from different collection objects, different collection methods or different monitoring dimensions, used to comprehensively characterize the geological changes, construction effects and abnormal response of the current construction section during the construction process.

[0022] For example, the multi-source data may include at least one of geological exploration data, construction parameter data, surrounding rock response monitoring data, and field auxiliary data. The geological exploration data may include geological forecast data, drilling data, geological sketch data, or geological logging data. The construction parameter data may include drilling speed, propulsion pressure, torque, excavation footage, blasting parameters, muck removal information, or support construction parameters. The surrounding rock response monitoring data may include surrounding rock displacement, crown settlement, perimeter convergence, support structure stress, crack propagation, seepage, water pressure, or microseismic response data. The field auxiliary data may include construction logs, video images, ambient temperature, ambient humidity, dust concentration, or equipment operating status data.

[0023] It is understood that the multi-source data can be selected and combined according to the monitoring conditions, construction methods and risk types of the current construction section, and form a data set corresponding to the current construction section through time alignment, section correspondence or spatial registration, so as to characterize the comprehensive status of the construction section at the current moment or within the target time window.

[0024] S02. Based on the multi-source data, identify risk warning events, which are used to characterize abnormal signs of construction risks in the current construction section.

[0025] In this embodiment, risk warning events are the event results obtained by event-based characterization of abnormal changes in construction risks in the current construction section, such as abnormally enhanced surrounding rock deformation events, abnormally increased seepage events, abnormal changes in support stress events, or abnormal fluctuations in construction disturbance events.

[0026] Because multi-source data is heterogeneous and continuously changing, it is difficult to conduct unified comparison, correlation analysis and evolutionary characterization of different abnormal changes based on the original multi-source data. Therefore, characterizing abnormal changes as risk symptom events that can be identified individually is beneficial for unified expression, correlation analysis and sequential connection of abnormal changes from different sources.

[0027] Furthermore, in some embodiments, step S02, which involves identifying risk warning events based on the multi-source data, includes, for example... Figure 2 The steps shown are as follows: S021. Based on the multi-source data, extract the feature quantities of different data respectively.

[0028] In this implementation, the feature quantity corresponding to any data source refers to the parameter extracted from the data of that data source and used to characterize the state or change characteristics of the data.

[0029] Since data from different sources differ in data format, sampling method, and variation characteristics, it is necessary to extract corresponding feature quantities for different data types in the specific implementation, so as to serve as input for subsequent anomaly detection.

[0030] In one specific implementation, for geological exploration data, at least one of the following features can be extracted: degree of reflection anomaly, wave velocity deviation characteristics, lithological variation characteristics, or adverse geological indication characteristics. For example, the degree of reflection anomaly can be determined as the feature based on the difference in reflection amplitude between the target section and the reference section in geological prediction data; alternatively, adverse geological indication characteristics can be determined as the feature based on changes in drilling resistance, rebound, and / or core integrity in drilling data.

[0031] In another specific embodiment, for the construction parameter data, at least one of the following features can be extracted: drilling speed change rate, propulsion pressure fluctuation amplitude, torque fluctuation degree, excavation footage deviation degree, or blasting parameter change degree. For example, the drilling speed change rate can be determined as the feature based on the rate of change of drilling speed within a continuous sampling period; the propulsion pressure fluctuation amplitude can be determined as the feature based on the deviation of propulsion pressure from the average pressure within a preset time window; and the torque fluctuation degree can be determined as the feature based on the torque fluctuation range within a preset period.

[0032] In another specific embodiment, for the surrounding rock response monitoring data, at least one characteristic quantity can be extracted from the following: surrounding rock displacement change rate, crown settlement increment, perimeter convergence rate, support structure stress change amplitude, seepage rate increase, water pressure fluctuation amplitude, or microseismic activity degree. For example, the surrounding rock displacement change rate can be determined as the characteristic quantity based on the displacement difference between adjacent monitoring times; the seepage rate increase can be determined as the characteristic quantity based on the seepage increment per unit time.

[0033] In another specific embodiment, for on-site auxiliary data, at least one feature quantity can be extracted from image change features, log record features, or equipment operating status change features. For example, log record features can be determined as the feature quantity based on abnormal keyword records in the construction log; image change features can be determined as the feature quantity based on information such as water accumulation, rockfall, and crack expansion in video images.

[0034] S022. Based on at least one of the deviation magnitude, duration, and trend of the characteristic quantity relative to the corresponding baseline state, determine the candidate abnormal signs.

[0035] In this part of the implementation, the reference state refers to a reference state used for comparison with the corresponding characteristic quantity, and is used to characterize the reference level of the current construction section under normal construction conditions, stable surrounding rock conditions, or preset safety conditions.

[0036] Furthermore, the candidate anomaly refers to an anomaly that may be related to construction risks, characterized by an abnormal deviation of the feature quantity from the corresponding baseline state.

[0037] Specifically, step S022 can compare each of the aforementioned feature quantities with the corresponding baseline state, and determine whether it meets the preset abnormal conditions based on at least one of the deviation magnitude, duration, and trend of change; when the preset abnormal conditions are met, the corresponding abnormal change can be identified as a candidate abnormal sign.

[0038] For example, when the surrounding rock convergence rate is continuously higher than the reference value corresponding to the stable construction stage, the corresponding abnormal change is identified as a candidate abnormal sign of "abnormal surrounding rock deformation"; when the seepage volume is continuously higher than the normal seepage level within a preset time window, the corresponding abnormal change is identified as a candidate abnormal sign of "abnormally enhanced seepage"; when the torque fluctuation amplitude exceeds the set threshold, the corresponding abnormal change is identified as a candidate abnormal sign of "abnormal fluctuation of construction disturbance".

[0039] S023. Within a preset time window, based on the occurrence sequence and consistency of changes of at least two types of candidate abnormal signs, candidate abnormal signs that meet the preset association triggering relationship are identified as risk sign events.

[0040] In this part of the implementation, the preset association triggering relationship refers to the relationship in which at least two types of candidate abnormal signs meet a preset order within a preset time window, and the candidate abnormal signs that appear later meet preset response enhancement conditions or preset continuous association conditions relative to the candidate abnormal signs that appear earlier. This is used to characterize the triggering association between different candidate abnormal signs that is related to the evolution of construction risks.

[0041] Furthermore, the occurrence sequence is used to characterize the sequential occurrence relationship between at least two types of candidate abnormal signs; the change consistency is used to characterize the consistent, synergistic, or mutually reinforcing relationship between the at least two types of candidate abnormal signs in terms of change direction, change intensity, or duration.

[0042] The preset response enhancement condition can be used to characterize the enhanced trend of subsequent candidate abnormal signs in terms of abnormal amplitude, abnormal frequency, or abnormal growth rate compared to the earlier candidate abnormal signs; the preset continuous association condition can be used to characterize the continuous coexistence or continuous coordinated change of the at least two types of candidate abnormal signs within a preset time window.

[0043] Specifically, step S023 can analyze the occurrence sequence and consistency of changes of at least two types of candidate abnormal signs within a preset time window; when it is determined that they meet the preset association triggering relationship, the corresponding candidate abnormal sign is confirmed as a risk sign event.

[0044] For example, if the candidate abnormal sign of "abnormal fluctuation of construction disturbance" is detected first, and then the candidate abnormal sign of "abnormal enhancement of seepage" is detected within a preset time window, and the seepage growth rate is significantly increased compared to the previous stage, then the combination of candidate abnormal signs can be identified as a "water-rich disturbance risk sign event"; if the candidate abnormal sign of "abnormal deformation of surrounding rock" is detected first, and then the candidate abnormal sign of "abnormal support stress" is detected, and the two continue to enhance synergistically within a preset time window, then the combination of candidate abnormal signs can be identified as a "surrounding rock instability risk sign event".

[0045] S03. Based on the risk symptom events, construct a risk symptom evolution chain, which is constructed according to a combination of target risk symptom events that satisfy a preset evolutionary relationship.

[0046] In this embodiment, the risk symptom evolution chain is an event association sequence formed by connecting multiple risk symptom events in a preset evolutionary order, used to characterize the evolution process of abnormal construction risk symptoms in the current construction section.

[0047] Since a single risk symptom event can only reflect a local anomaly, this embodiment constructs a risk symptom evolution chain, which is beneficial to further characterize the dynamic development state of construction risks and provides a basis for subsequent risk evolution branch determination and stage node determination.

[0048] Furthermore, in some embodiments, step S03, which involves constructing a risk symptom evolution chain based on the risk symptom events, includes, for example... Figure 3 The steps shown are as follows: S031. Sort the risk warning events according to their occurrence time.

[0049] Specifically, step S031 sorts each risk symptom event by time according to the occurrence time, the start time of the duration interval, or the first identification time of each risk symptom event.

[0050] For example, if the risk signs of “abnormal fluctuation of construction disturbance”, “abnormal increase of water seepage” and “abnormal increase of surrounding rock deformation” are identified successively in a certain construction section, they are sorted into a candidate event sequence according to their corresponding occurrence time.

[0051] S032. Based on the preset event type transfer rules, determine whether the preceding risk symptom event and the following risk symptom event meet the evolution transfer conditions.

[0052] In this part of the implementation, the event type transfer rule refers to the preset rule that allows the establishment of an evolutionary transfer relationship between different types of risk symptom events, and is used to limit whether there is a type constraint for establishing an evolutionary transfer relationship between different types of risk symptom events.

[0053] The evolutionary transition condition refers to the determination condition used to determine whether there is a condition for establishing an evolutionary transition relationship between a previous risk symptom event and a subsequent risk symptom event. It includes at least one of the preset time adjacency condition, type association condition, or response connection condition.

[0054] For example, when the risk symptom event of "abnormal fluctuation of construction disturbance" appears first, and the risk symptom event of "abnormal increase in seepage" subsequently appears within a preset time interval, it can be determined that the two meet the evolution and transfer conditions; when the risk symptom event of "abnormal increase in seepage" appears first, and the risk symptom event of "abnormal increase in surrounding rock deformation" subsequently appears within a preset time interval, and the two meet the preset type association conditions or response connection conditions, it can also be determined that the two meet the evolution and transfer conditions.

[0055] Conversely, if the subsequent risk symptom event does not occur within the preset time interval, or if the preset type association conditions or response connection conditions are not met between the preceding and following risk symptom events, it can be determined that the preceding and following risk symptom events do not meet the evolution and transfer conditions.

[0056] For example, if no related risk event, such as "abnormal increase in seepage", occurs for a long period after the occurrence of the "abnormal fluctuation in construction disturbance" risk event, or if a risk event, such as "abnormal change in support stress", occurs subsequently, but the two do not meet the preset type association conditions, it can be determined that the evolution and transfer conditions are not met.

[0057] S033. When the preceding and following risk symptom events satisfy the evolutionary transition conditions, the evolutionary connection relationship between the preceding and following risk symptom events is determined based on the time interval between the preceding and following risk symptom events and according to at least one of the event intensity change relationship and the continuous state connection relationship.

[0058] In this implementation, the evolutionary connection relationship refers to the connection relationship used to characterize the evolutionary connection between successive risk symptom events.

[0059] Wherein, the time interval is used to characterize the time difference between the occurrence of successive risk warning events; the event intensity change relationship is used to characterize the change relationship of the subsequent risk warning event relative to the previous risk warning event in terms of abnormal intensity, abnormal frequency, or abnormal growth rate; the continuous state connection relationship is used to characterize whether the subsequent risk warning event has occurred and formed a relationship of continuous overlap, continuous synergy, or continuous development with the previous risk warning event before the previous risk warning event has ended.

[0060] Specifically, step S033 can determine the evolutionary connection relationship between the preceding and following risk symptom events based on the determination that the preceding and following risk symptom events meet the above evolutionary transition conditions, combined with the time interval, and based on at least one of the event intensity change relationship and the continuous state connection relationship.

[0061] For example, if the risk symptom event of "abnormally enhanced seepage" occurs shortly after the risk symptom event of "abnormal fluctuation of construction disturbance" and the seepage rate continues to increase, it can be determined that there is an enhanced evolutionary connection between the two; if the risk symptom event of "abnormally enhanced surrounding rock deformation" occurs during the period when the risk symptom event of "abnormally enhanced seepage" continues to exist, it can be determined that there is a continuous evolutionary connection between the two.

[0062] S034. Based on the evolutionary connection relationship, the corresponding risk symptom events are sequentially connected to construct the risk symptom evolution chain.

[0063] Specifically, step S034 connects the risk symptom events that satisfy the evolutionary connection relationship in chronological order to form a risk symptom evolution chain describing the evolution process of risk symptoms in the current construction section.

[0064] For example, when the risk signs of “abnormal fluctuations in construction disturbance”, “abnormal enhancement of seepage”, and “abnormal enhancement of surrounding rock deformation” are identified in sequence, and the adjacent risk signs meet the preset evolutionary relationship, a risk sign evolution chain of “abnormal fluctuations in construction disturbance → abnormal enhancement of seepage → abnormal enhancement of surrounding rock deformation” can be constructed. If the risk sign of “abnormal change in support stress” is further identified, and it meets the continuous connection relationship with the previous risk sign event, the risk sign evolution chain can be extended to “abnormal fluctuations in construction disturbance → abnormal enhancement of seepage → abnormal enhancement of surrounding rock deformation → abnormal change in support stress”.

[0065] In this embodiment, based on the constructed risk symptom evolution chain, the risk evolution status of the current construction section can be more accurately characterized by combining the temporal relationship and evolutionary association between multiple risk symptom events, in addition to timely identification of individual risk symptom events. This improves the accuracy of the timing, triggering level, and matching of construction pre-control measures.

[0066] S04. Based on the preset risk symptom evolution path model, determine the target risk evolution branch corresponding to the risk symptom evolution chain, and determine the target stage node of the current construction section in the target risk evolution branch.

[0067] In this embodiment, the risk symptom evolution path model includes multiple target risk evolution branches, each of which is used to characterize the evolution direction, stage division, and stage transition conditions of different construction risks. Specifically, step S04 matches the risk symptom evolution chain corresponding to the current construction section with each target risk evolution branch in the risk symptom evolution path model to determine the corresponding target risk evolution branch and target stage node, thereby providing a basis for matching subsequent construction pre-control measures.

[0068] Furthermore, the risk symptom evolution path model preset in this embodiment, through, as... Figure 4 The following steps are shown to generate: S0401. Obtain the historical risk symptom event sequence corresponding to multiple historical construction sections and the actual risk results corresponding to each historical construction section.

[0069] The historical risk warning event sequence refers to the sequence of risk warning events identified during the construction of historical construction sections, arranged in chronological order; the actual risk outcome refers to the risk outcome that ultimately occurred in the corresponding historical construction section.

[0070] S0402. From each of the historical risk symptom event sequences, extract the combination of risk symptom events located within a preset period before the actual risk outcome occurs, as candidate precursor evolution fragments.

[0071] The candidate precursor evolution fragment refers to a combination of risk symptom events that precede the actual risk outcome and have a precursory indicative role for that actual risk outcome.

[0072] It is understood that by extracting the candidate precursor evolution fragments, the precursor evolution process related to the actual risk outcome can be screened out from the historical risk symptom event sequence.

[0073] S0403. Construct an initial evolutionary path network based on the sequential connection of events and shared risk symptom events among the candidate precursor evolutionary segments.

[0074] The initial evolutionary path network refers to a network structure built based on the sequential connection and event sharing relationships among multiple candidate precursor evolutionary segments, used to characterize the potential evolutionary associations between different candidate precursor evolutionary segments.

[0075] For example, for the same actual risk result "surrounding rock instability risk", the following candidate precursor evolution segments were extracted from multiple historical construction sections: Candidate precursor evolution segment A ("abnormal fluctuation of construction disturbance → abnormal enhancement of seepage → abnormal enhancement of surrounding rock deformation"); Candidate precursor evolution segment B ("abnormal enhancement of seepage → abnormal enhancement of surrounding rock deformation → abnormal change of support stress"); Candidate precursor evolution segment C ("abnormal fluctuation of construction disturbance → abnormal enhancement of seepage → abnormal change of support stress").

[0076] Furthermore, based on the event sequence connection relationship and shared risk symptom events among the above candidate precursor evolution fragments, “abnormal fluctuation of construction disturbance” is set as node N1, “abnormal enhancement of seepage” is set as node N2, “abnormal enhancement of surrounding rock deformation” is set as node N3, “abnormal change of support stress” is set as node N4, and “surrounding rock instability risk result” is set as result node R.

[0077] Furthermore, based on the event sequence connections and shared risk symptom events among the aforementioned candidate precursor evolutionary segments, a system is constructed as follows: Figure 5 The directed connections shown are as follows: node N1 points to node N2, node N2 points to node N3, node N2 can also directly point to node N4, node N3 points to node N4, and node N4 points to the result node R. Thus, an initial evolutionary path network containing branching and convergence relationships can be formed based on the sequential connection relationships and shared event relationships among multiple candidate precursor evolutionary segments.

[0078] S0404. Based on the actual risk results, candidate precursor evolution segments corresponding to the same risk results are selected, and the target risk evolution branch is determined.

[0079] It is understood that any outcome node in the initial evolutionary path network is used to characterize an actual risk outcome, and each outcome node corresponds to at least one candidate precursor evolutionary segment to characterize that different candidate precursor evolutionary segments can be attributed to the same actual risk outcome.

[0080] In this embodiment, step S0404 further filters candidate precursor evolution fragments that point to the same actual risk outcome, and merges the filtered candidate precursor evolution fragments to determine the target risk evolution branch corresponding to the actual risk outcome.

[0081] S0405. Based on at least one of the following: the intensity change boundary of risk symptom events within the target risk evolution branch, the event density change boundary, and the result proximity, divide the target risk evolution branch into stages and determine the stage nodes.

[0082] In this embodiment, the stage node refers to a node used to characterize different stages of risk development in the evolution branch of the target risk. It can be understood that the stage node may correspond to an inflection point in the development of risk from low to high, from weak to strong, or from dispersed to concentrated.

[0083] The intensity change boundary refers to the boundary where the risk symptom events within the target risk evolution branch undergo significant changes in abnormal amplitude, abnormal frequency, or abnormal growth rate; the event density change boundary refers to the boundary where the frequency of occurrence of risk symptom events or the degree of event clustering changes significantly within a unit time range; and the result proximity refers to the degree of proximity of the corresponding risk symptom event in the target risk evolution branch to the actual risk outcome at that time.

[0084] Specifically, step S0405 can analyze at least one of the following along the evolution direction of the target risk evolution branch: the intensity change of the risk symptom event, the event density change, and the result proximity. The target risk evolution branch can be divided into stages at the locations where the corresponding characteristics change significantly, thereby determining the corresponding stage nodes.

[0085] For example, if in a certain branch of the risk evolution, the early stage mainly presents risk warning events such as "abnormal fluctuations in construction disturbance" and "abnormally enhanced seepage," and the event intensity is low and the event distribution is relatively scattered, then this part can be classified as the early warning stage. If the subsequent risk warning event of "abnormally enhanced surrounding rock deformation" appears, and the intensity of the risk warning event increases significantly and the frequency of the event increases significantly per unit time, then the next stage node can be determined at the corresponding position, and the subsequent branch segment can be classified as the risk enhancement stage. If the risk warning event of "abnormal changes in support stress" appears even later, and the corresponding risk warning event is closer to the time when the actual risk result occurs, then the subsequent stage node can be further determined, and the subsequent branch segment can be classified as the high-risk approaching stage.

[0086] S0406. Determine the transfer conditions between adjacent stage nodes based on the transfer characteristics of risk symptom events between adjacent stage nodes.

[0087] In this embodiment, the transfer feature refers to the feature used to characterize the transfer pattern of risk symptom events between adjacent stage nodes; further, the transfer feature may include the dominant transfer direction and / or transfer frequency of risk symptom events between adjacent stage nodes.

[0088] The dominant shift direction refers to the main shift trend exhibited when risk symptom events shift from one stage to the next between adjacent stage nodes; the shift frequency refers to the frequency or proportion of occurrence of the corresponding risk symptom events shifting according to the dominant shift direction in historical samples.

[0089] For example, if, between the early warning stage node and the risk enhancement stage node, the risk symptom events mainly manifest as a shift from "abnormal fluctuations in construction disturbance" to "abnormally enhanced seepage" or "abnormally enhanced surrounding rock deformation," and such a shift occurs frequently in historical samples, then the corresponding event shift relationship can be determined as the shift condition from the early warning stage node to the risk enhancement stage node. If, between the risk enhancement stage node and the high-risk adjacent stage node, the risk symptom events mainly manifest as a shift from "abnormally enhanced surrounding rock deformation" to "abnormal changes in support stress," and the frequency of such a shift reaches a preset condition, then the corresponding event shift relationship can be determined as the shift condition from the risk enhancement stage node to the high-risk adjacent stage node.

[0090] S0407. Based on the target risk evolution branches, stage nodes, and transition conditions, generate the corresponding risk symptom evolution path model.

[0091] In this embodiment, the risk symptom evolution path model is a model structure formed by associating and integrating the target risk evolution branches, stage nodes, and transition conditions between adjacent stage nodes. It is used to characterize the evolution path and stage evolution law of risk symptom events under the corresponding actual risk results.

[0092] Specifically, in step S0407, the target risk evolution branch can be used as the main path, the stage nodes can be set at the corresponding positions of the target risk evolution branch, and the transfer conditions between adjacent stage nodes can be set as the transfer rules between the corresponding stage nodes, thereby generating the corresponding risk symptom evolution path model.

[0093] Furthermore, in this embodiment, step S04, which involves determining the target risk evolution branch corresponding to the risk symptom evolution chain based on a preset risk symptom evolution path model, and identifying the target stage node of the current construction section within that target risk evolution branch, specifically includes the following steps: S041. Perform branch matching between the risk symptom evolution chain and each target risk evolution branch in the risk symptom evolution path model, and determine the target risk evolution branch corresponding to the risk symptom evolution chain based on the branch matching results.

[0094] In this embodiment, the branch matching result refers to the matching situation between the risk symptom evolution chain and the corresponding target risk evolution branch, which is used to characterize the degree of consistency between the two.

[0095] Specifically, the risk symptom evolution chain can be denoted as chain A, and any branch of the target risk evolution can be denoted as branch B. Let chain A contain m risk symptom events arranged in chronological order, and branch B contain n risk symptom events arranged in branch order.

[0096] Furthermore, step S041 can compare chain A and branch B from three dimensions: the order of risk symptom event types, the characteristics of event intensity changes, and the event connection relationship, and determine the branch matching result between chain A and branch B based on the comparison results.

[0097] The event type sequence matching degree can be determined based on the length of the longest event sequence between chain A and branch B that maintains a sequential match according to the risk symptom event types.

[0098] For example, if the longest event sequence length is L, the event type order matching degree can be expressed as 2L / (m+n).

[0099] The matching degree of event intensity change can be determined based on the intensity difference between the matched risk symptom events in chain A and branch B.

[0100] For example, the intensity of risk symptom events can be pre-divided into multiple levels, and the intensity matching value of a single event can be determined based on the intensity level difference between corresponding matching events. Then, the intensity matching values ​​of each single event can be averaged to obtain the event intensity change matching degree.

[0101] The matching degree of event connectivity can be determined based on the consistency of the evolutionary connectivity between adjacent matching risk symptom events in chain A and branch B.

[0102] For example, if the evolutionary connectivity between adjacent matching risk symptom events is the same, the corresponding connectivity matching value can be recorded as 1; if the evolutionary connectivity between adjacent matching risk symptom events belongs to the same type of connectivity, the corresponding connectivity matching value can be recorded as 0.5; if the evolutionary connectivity between adjacent matching risk symptom events is inconsistent, the corresponding connectivity matching value can be recorded as 0; then, the average of each connectivity matching value is calculated to obtain the event connectivity matching degree.

[0103] In some specific implementations, step S041 further determines the comprehensive matching degree between chain A and branch B based on the event type sequence matching degree, event intensity change matching degree, and event connection relationship matching degree. For example, the comprehensive matching degree can be expressed as: M = α × M1 + β × M2 + γ × M3, where M1 represents the event type sequence matching degree, M2 represents the event intensity change matching degree, M3 represents the event connection relationship matching degree, and α, β, and γ are preset weights, with α + β + γ = 1.

[0104] In these specific implementations, the comprehensive matching degree between the risk symptom evolution chain and each target risk evolution branch is calculated, and the target risk evolution branch with the highest comprehensive matching degree and that meets the preset matching conditions is determined as the target risk evolution branch corresponding to the risk symptom evolution chain.

[0105] S042. Based on the matching position of the terminal risk symptom event in the risk symptom evolution chain in the target risk evolution branch, and the transfer conditions between adjacent stage nodes in the target risk evolution branch, determine the target stage node of the current construction section in the target risk evolution branch.

[0106] In this embodiment, the target stage node refers to the stage node corresponding to the current construction section in the target risk evolution branch, which is used to characterize the development stage of the current construction section in the corresponding risk evolution process.

[0107] Specifically, after determining the target risk evolution branch corresponding to the risk symptom evolution chain in step S041, the matching position of the terminal risk symptom event in the risk symptom evolution chain in the target risk evolution branch can be further determined, and the risk development stage of the current construction section can be determined by combining the transfer conditions between adjacent stage nodes in the target risk evolution branch.

[0108] The matching position refers to the corresponding position of the terminal risk symptom event in the target risk evolution branch, which is used to characterize the degree of advancement of the risk symptom evolution state corresponding to the current construction section in the target risk evolution branch; the transfer condition refers to the condition that should be met when a stage transfer occurs between adjacent stage nodes, which is used to characterize the evolutionary connection law between different risk development stages.

[0109] Furthermore, if the matching location of the terminal risk symptom event is located within the stage segment corresponding to a certain stage node, the current construction segment can be determined as being at that stage node; if the matching location of the terminal risk symptom event is close to the transition boundary between adjacent stage nodes and satisfies the transition condition from the previous stage node to the next stage node, the current construction segment can be determined as being at the next stage node, or as being in a transitional stage state from the previous stage node to the next stage node.

[0110] For example, if the target risk evolution branch is "abnormal fluctuation of construction disturbance → abnormal enhancement of seepage → abnormal enhancement of surrounding rock deformation → abnormal change of support stress", and the terminal risk symptom event in the risk symptom evolution chain is "abnormal enhancement of surrounding rock deformation", and the matching position of this terminal risk symptom event in the target risk evolution branch is located within the stage segment corresponding to the risk enhancement stage, then the current construction segment can be determined as being at the risk enhancement stage node; if the risk symptom event "abnormal change of support stress" is subsequently identified, and the transfer condition of transferring from the risk enhancement stage node to the high-risk adjacent stage node is met, then the current construction segment can be determined as being at the high-risk adjacent stage node.

[0111] By using the above methods, based on the determination of the target risk evolution branch, it is possible to further determine the stage of the current construction section in the corresponding risk evolution process, thereby providing a basis for the graded triggering and targeted matching of subsequent construction pre-control measures.

[0112] S05. Based on the current target stage node in the target risk evolution branch, determine the construction pre-control measures corresponding to the current construction section.

[0113] The construction pre-control measures include measures to adjust at least one of the construction parameters, construction procedures, or on-site control conditions.

[0114] It is understood that this embodiment, by combining the target risk evolution branches and target stage nodes to determine the construction pre-control measures, can provide the current construction section with construction control measures that are adapted to its risk evolution direction and stage.

[0115] Specifically, step S05 can first call the set of measures corresponding to the target risk evolution branch, then select at least one construction control measure corresponding to the current target stage node from the set of measures, and determine the selected construction control measure as the construction pre-control measure corresponding to the current construction section.

[0116] The set of measures refers to a set of construction control measures pre-established for a specific target risk evolution branch; the construction control measures corresponding to the target stage node refer to measures that are adapted to the current risk development stage.

[0117] For example, when the target risk evolution branch corresponding to the current construction section is determined to be the "water-rich disturbance - surrounding rock instability" risk evolution branch, and the target stage node is an early warning stage node, the construction pre-control measures corresponding to the current construction section can be determined to be increasing the monitoring frequency, adjusting drilling parameters, or strengthening advanced detection; when the target stage node is a risk enhancement stage node, the construction pre-control measures corresponding to the current construction section can be determined to be shortening the excavation footage, strengthening support, or implementing local grouting in advance; when the target stage node is a high-risk approach stage node, the construction pre-control measures corresponding to the current construction section can be determined to be suspending high-disturbance operations, switching construction procedures, strengthening drainage and depressurization, or implementing emergency reinforcement.

[0118] By using the above methods, the evolution branches of the target risk and the target stage nodes can be linked with the corresponding construction control measures, thereby improving the timing, trigger level, and relevance of the construction pre-control measures.

[0119] In some embodiments, to address the problem that risk symptom events may continue to change after the implementation of construction pre-control measures, leading to a gradual mismatch between the originally determined construction pre-control measures and the actual risk evolution state of the current construction section, the tunnel construction pre-control method based on risk symptom evolution provided by the present invention further includes the following steps: S06. After the implementation of construction pre-control measures, based on the newly identified risk symptom events, update the risk symptom evolution chain, and based on the updated risk symptom evolution chain, redetermine the construction pre-control measures.

[0120] Specifically, step S06 is used to make rolling corrections to the risk evolution status of the current construction section after the implementation of construction pre-control measures.

[0121] Furthermore, if a newly identified risk symptom event after the implementation of construction pre-control measures satisfies a preset evolutionary relationship with an existing risk symptom event, the newly identified risk symptom event is connected to the original risk symptom evolution chain to update the risk symptom evolution chain; and based on the updated risk symptom evolution chain, the corresponding target risk evolution branch and target stage node are re-determined, thereby redetermining the construction pre-control measures corresponding to the current construction section.

[0122] Through the above methods, after the implementation of construction pre-control measures, the original construction pre-control measures can be dynamically modified based on continuously changing risk symptom events, thereby reducing the risk of mismatch between the originally determined measures and the actual risk evolution state of the current construction section.

[0123] In some embodiments, the present invention also provides a tunnel construction pre-control system based on risk symptom evolution; see [link to relevant documentation]. Figure 6 , Figure 6This is a schematic diagram of the tunnel construction pre-control system based on risk symptom evolution provided in an embodiment of the present invention.

[0124] In these embodiments, the tunnel construction pre-control system based on risk symptom evolution includes an input device, a processor, a memory, and an output device, wherein the input device, processor, memory, and output device are interconnected.

[0125] Further, the input device is used to acquire multi-source data corresponding to the current construction section. The multi-source data may include geological exploration data, construction parameter data, surrounding rock response monitoring data, and / or on-site auxiliary data. The memory is used to store computer programs, preset parameters, risk symptom evolution path models, and intermediate processing results. The processor is used to call and execute the computer program to implement the tunnel construction pre-control method based on risk symptom evolution described in any of the above embodiments. Specifically, it may include: identifying risk symptom events based on the multi-source data acquired by the input device, constructing a risk symptom evolution chain, determining the corresponding target risk evolution branch and target stage node based on the preset risk symptom evolution path model, and determining the construction pre-control measures corresponding to the current construction section based on the target risk evolution branch and target stage node. The output device is used to output risk symptom events, risk symptom evolution chains, target risk evolution branches, target stage nodes, and / or construction pre-control measures.

[0126] Furthermore, the output device may be at least one of a display, an alarm device, a printing device, or a communication interface, for use in outputting corresponding early warning information and construction pre-control measures to construction site management personnel.

[0127] Furthermore, the memory may be a read-only memory (ROM), random access memory (RAM), flash memory, hard disk, solid-state drive, optical disk, or other non-transitory computer-readable storage medium.

[0128] The above system structure enables the identification and evolution analysis of risk signs in the current construction section, as well as the output of construction pre-control measures, thereby providing system support for risk pre-control during tunnel construction.

[0129] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0130] It should be noted that the above embodiments can be freely combined as needed. The above are merely preferred embodiments of the present invention; it should be observed that those skilled in the art can make several improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A tunnel construction pre-control method based on risk symptom evolution, characterized in that, The steps include the following: Acquire multi-source data, which is used to characterize the current state of the construction section; Based on the multi-source data, risk warning events are identified, which are used to characterize abnormal signs of construction risks in the current construction section. Based on the aforementioned risk warning events, a risk warning evolution chain is constructed, which is built according to a combination of target risk warning events that satisfy a preset evolutionary relationship. Based on the preset risk symptom evolution path model, the target risk evolution branch corresponding to the risk symptom evolution chain is determined, and the target stage node of the current construction section in the target risk evolution branch is determined. Based on the current target stage node in the target risk evolution branch, determine the construction pre-control measures corresponding to the current construction section. The construction pre-control measures include measures to adjust at least one of the construction parameters, construction procedures, or on-site control conditions.

2. The tunnel construction pre-control method based on risk symptom evolution according to claim 1, characterized in that, It also includes the following steps: After the implementation of construction pre-control measures, the risk symptom evolution chain is updated based on newly identified risk symptom events, and the construction pre-control measures are redefined based on the updated risk symptom evolution chain.

3. The tunnel construction pre-control method based on risk symptom evolution according to claim 1 or 2, characterized in that, The step of identifying risk warning events based on the multi-source data includes the following steps: Based on the multi-source data, feature quantities of different data are extracted respectively; Candidate abnormal signs are determined based on at least one of the deviation magnitude, duration, and trend of the feature quantity relative to the corresponding baseline state; Within a preset time window, based on the timing and consistency of the appearance of at least two types of candidate abnormal signs, candidate abnormal signs that meet the preset association triggering relationship are identified as risk sign events.

4. The tunnel construction pre-control method based on risk symptom evolution according to claim 3, characterized in that, The preset association triggering relationship includes: Within a preset time window, the at least two types of candidate abnormal signs appear in a preset order, and the candidate abnormal signs that appear later satisfy a preset response enhancement condition or a preset continuous correlation condition relative to the candidate abnormal signs that appear earlier.

5. The tunnel construction pre-control method based on risk symptom evolution according to claim 1 or 2, characterized in that, The process of constructing a risk symptom evolution chain based on the aforementioned risk symptom events includes the following steps: The risk warning events are sorted according to their occurrence time. Based on the preset event type transition rules, determine whether the preceding and subsequent risk symptom events meet the evolutionary transition conditions: When the preceding and following risk symptom events satisfy the evolutionary transition conditions, the evolutionary connection relationship between the preceding and following risk symptom events is determined based on the time interval between the preceding and following risk symptom events and according to at least one of the relationship between event intensity changes and the relationship between continuous states. Based on the evolutionary connection relationship, the corresponding risk symptom events are sequentially connected to construct the risk symptom evolution chain.

6. The tunnel construction pre-control method based on risk symptom evolution according to claim 1 or 2, characterized in that, The risk symptom evolution path model is generated using the following method: Obtain the historical risk warning event sequences corresponding to multiple historical construction sections and the actual risk results corresponding to each historical construction section; In each of the historical risk symptom event sequences, combinations of risk symptom events located within a preset time period before the actual risk outcome occur are extracted as candidate precursor evolution fragments. Based on the sequential connection of events and shared risk symptom events among the candidate precursor evolutionary segments, an initial evolutionary path network is constructed; Based on the actual risk results, candidate precursor evolution segments corresponding to the same risk results are screened, and the target risk evolution branch is determined. Based on at least one of the following: the intensity change boundary of risk symptom events within the target risk evolution branch, the event density change boundary, and the outcome proximity, the target risk evolution branch is divided into stages, and the stage nodes are determined. Based on the transfer characteristics of risk symptom events between adjacent stage nodes, determine the transfer conditions between adjacent stage nodes; Based on the target risk evolution branches, stage nodes, and transition conditions, a corresponding risk symptom evolution path model is generated.

7. The tunnel construction pre-control method based on risk symptom evolution according to claim 6, characterized in that, The process of determining the target risk evolution branch corresponding to the risk symptom evolution chain based on the preset risk symptom evolution path model, and identifying the target stage node of the current construction section within the target risk evolution branch, includes the following steps: The risk symptom evolution chain is matched with each target risk evolution branch in the risk symptom evolution path model, and the target risk evolution branch corresponding to the risk symptom evolution chain is determined based on the branch matching results. Based on the matching position of the terminal risk symptom event in the risk symptom evolution chain in the target risk evolution branch, and the transition conditions between adjacent stage nodes in the target risk evolution branch, the target stage node of the current construction section in the target risk evolution branch is determined.

8. The tunnel construction pre-control method based on risk symptom evolution according to claim 7, characterized in that, The step of determining the target risk evolution branch corresponding to the risk symptom evolution chain based on the matching results of each branch includes the following steps: Determine the matching degree of event type sequence, event intensity change, and event connection relationship between the risk symptom evolution chain and each target risk evolution branch; Based on the event type sequence matching degree, event intensity change matching degree, and event connection relationship matching degree, the comprehensive matching degree corresponding to each target risk evolution branch is determined; The target risk evolution branch with the highest overall matching degree and that meets the preset matching conditions is determined as the target risk evolution branch corresponding to the risk symptom evolution chain.

9. The tunnel construction pre-control method based on risk symptom evolution according to claim 1 or 2, characterized in that, The step of determining the construction pre-control measures corresponding to the current construction section based on the current target stage node in the target risk evolution branch includes the following steps: Invoke the set of measures corresponding to the target risk evolution branch; Select at least one construction control measure corresponding to the target stage node from the set of measures; At least one construction control measure selected from the screening will be combined as the construction pre-control measure for the current construction section.

10. A tunnel construction pre-control system based on risk symptom evolution, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the tunnel construction pre-control method based on risk symptom evolution as described in any one of claims 1 to 9.

Images