Tunnel fire water level on-line patrol system

Abstract

The application discloses a tunnel fire-fighting water level on-line patrol measurement system, the system comprises a patrol center and a pressure sensor arranged at each detection node of a fire-fighting pipe network, and the patrol center is used for: determining a water level difference sequence of a pipeline according to a historical water pressure sequence of each detection node; determining an abnormal index of the pipeline at a target moment according to a target moment water level difference and the historical moment water level difference in the water level difference sequence; determining a non-uniformity coefficient of an adjacent pipeline group at the target moment according to the abnormal index and the target moment water level difference of each pipeline in the adjacent pipeline group; and determining a failure rate of the fire-fighting pipe network at the target moment according to the non-uniformity coefficient of each adjacent pipeline group of the fire-fighting pipe network at the target moment and the abnormal index of each pipeline. Through the method of the embodiment of the application, the failure rate of the fire-fighting pipe network can be accurately determined, and the efficiency and safety of fire-fighting pipe network fault detection are improved.

Description

Technical Field

[0001] This invention relates to the field of fault detection, and in particular to an online monitoring system for fire-fighting water levels in tunnels. Background Technology

[0002] The reliable operation of fire protection piping systems is crucial for fire prevention and control. Real-time monitoring of the network status and timely detection and troubleshooting are key to ensuring firefighting capabilities. Currently, monitoring of fire protection piping systems typically involves installing pressure and water level sensors at critical nodes or ends of the network. By collecting real-time water pressure and level data within the network, the system's operational readiness is determined. When the monitored water pressure falls below a preset threshold or the water level drops abnormally, the system issues an alarm signal, prompting management personnel to conduct maintenance. This pressure and level-based detection method is widely used due to its simple structure and low cost.

[0003] However, because the pipes within the network are interconnected, an anomaly in a single pipe can cause fluctuations in the overall system water pressure, making it impossible to pinpoint the exact location of the anomaly. Furthermore, changes in flow rate caused by daily water usage in residential areas can also lead to water pressure fluctuations, which are easily misinterpreted as network faults, resulting in numerous false alarms. Summary of the Invention

[0004] This invention provides an online monitoring system for fire-fighting water levels in tunnels. By calculating the anomaly index of a single pipe and the inconsistency coefficient between pipe groups, the system achieves comprehensive fault assessment from multiple dimensions, including individual pipes and adjacent pipe groups, thereby accurately determining the failure rate of the fire-fighting pipe network and improving the safety of the fire-fighting pipe network.

[0005] In a first aspect, the present invention provides an online monitoring system for fire-fighting water levels in tunnels. The system includes an inspection center and pressure sensors configured at each detection node of the fire-fighting pipeline network. The fire-fighting pipeline network between adjacent detection nodes forms a pipeline. The pressure sensor is used to collect the water pressure at the detection node corresponding to the pressure sensor at multiple detection times to form a historical water pressure sequence. The inspection center is used for: Based on the historical water pressure sequence of each detection node, determine the water level difference sequence of each pipeline; Based on the target time water level difference of the pipeline in the water level difference sequence and the historical time water level difference, the anomaly index of the pipeline at the target time is determined, wherein the historical time is the detection time before the target time, and the anomaly index represents the extent to which the pipeline water level difference deviates from the normal water level difference. Based on the anomaly index of each pipe in the adjacent pipe group and the water level difference at the target time, the non-consistency coefficient of the adjacent pipe group at the target time is determined. The adjacent pipe group consists of two adjacent pipes in the fire protection pipe network along the water flow direction. The non-consistency coefficient characterizes the degree of synchronization of the abnormal trend changes between the adjacent pipes. The failure rate of the fire protection network at the target time is determined based on the inconsistency coefficient of each adjacent pipe group and the anomaly index of each pipe.

[0006] Specifically, the inspection center is used to determine the water level difference sequence of the pipeline based on the historical water pressure sequence of each detection node, including: Based on the historical water pressure sequence of each detection node, determine the historical water level sequence of each detection node; For any pipe in the fire protection network at any detection time, determine the head detection node and tail detection node corresponding to the pipe, and determine the water level difference of the pipe at the detection time based on the water level of the head detection node and the water level of the tail detection node in the historical water level sequence at the detection time. A water level difference sequence is generated based on the water level difference of each pipeline at each detection time.

[0007] Specifically, the inspection center is used to determine the historical water level sequence of each of the detection nodes based on the historical water pressure sequence of each node, including: Obtain the water flow density and gravitational acceleration of the fire protection pipeline network; Based on the water pressure, water flow density, and gravitational acceleration at each detection time in the historical water pressure sequence, the water level corresponding to each detection time is determined, and the water level corresponding to each detection time constitutes the historical water level sequence of the detection node.

[0008] Specifically, the inspection center is used to determine the anomaly index of the pipeline at the target time based on the target time water level difference and the historical time water level difference in the pipeline water level difference sequence, including: Based on the target water level difference in the pipeline and a preset number of historical water level differences, determine the standard deviation and mean of the water level difference. The continuous offset factor of the pipeline is determined based on the water level difference and the average water level difference at each detection time. The continuous offset factor characterizes the direction of water level difference change in the pipeline at continuous detection times. The anomaly index of the pipeline is determined based on the pipeline's continuous offset factor, the target time water level difference, the mean water level difference, and the standard deviation of the water level difference.

[0009] Specifically, the inspection center is used to determine the continuous deviation factor of the pipeline based on the water level difference and the average water level difference at each inspection time, including: Based on the water level difference and the average water level difference at each detection time, the offset direction at each detection time is determined. Different offset directions represent the relationship between the water level difference and the average water level difference at each detection time. The offset direction at the target time is determined as the reference direction; Starting from the target time, backtracking from the nearest to the farthest in time, determine the number of detection times between the target time and the target historical time, wherein the target historical time is the first historical detection time whose offset direction is different from the reference direction; The persistence offset factor is determined based on the stated quantity.

[0010] Specifically, the inspection center is used to determine the inconsistency coefficient of the adjacent pipeline group at the target time based on the anomaly index of each pipeline in the adjacent pipeline group and the water level difference at the target time, including: Determine the preceding and following pipes in the adjacent pipe group, based on the direction of water flow; Based on the anomaly index of the preceding pipeline at the target time and the detection time before the target time, determine the change in the anomaly index of the preceding pipeline at the target time. Based on the anomaly index of the subsequent pipeline at the target time and the detection time before the target time, determine the change in the anomaly index of the subsequent pipeline at the target time. The first factor is determined based on the abnormal index change of the preceding pipeline and the abnormal index change of the following pipeline. The second factor is determined based on the target time water level difference between the preceding pipeline and the target time water level difference between the following pipeline; The inconsistency coefficient of the adjacent pipe group is determined based on the first factor and the second factor.

[0011] Specifically, the inspection center is used to determine a second factor based on the target time water level difference between the preceding pipeline and the target time water level difference between the following pipeline, including: The rate of change of water level difference is determined based on the target time water level difference of the preceding pipeline and the target time water level difference of the following pipeline. The water level difference change rate is mapped to a preset numerical range to obtain the second factor.

[0012] Specifically, the inspection center is used to determine the failure rate of the fire protection network at the target time based on the inconsistency coefficient of each adjacent pipe group and the anomaly index of each pipe, including: Identify target adjacent pipeline groups whose inconsistency coefficient is less than a preset threshold at the target time. Connect the pipes of each adjacent pipe group of each target according to the water flow sequence to obtain multiple continuous pipe segments. The pipe segment containing the most pipes is determined as the target pipe segment. The non-consistency ratio factor is determined based on the number of pipes in the target pipe section and the total number of pipes in the fire protection pipe network. The directional coordination factor is determined based on the number of pipes with positive anomaly indices, the number of pipes with negative anomaly indices, and the total number of pipes in the target pipeline segment. The boundary isolation coefficient is determined based on the inconsistency coefficient of each adjacent edge pipe group consisting of edge pipes in the target pipe segment and their adjacent non-target pipe segments, and the mean of the inconsistency coefficient of each adjacent internal pipe group in the target pipe segment that does not contain edge pipes. The failure rate of the fire protection network is determined based on the non-consistency ratio factor, the directional coordination factor, the boundary isolation coefficient, and the anomaly index of each pipe in the fire protection network.

[0013] Specifically, the inspection center is used to determine the failure rate of the fire protection network based on the inconsistency ratio factor, the directional coordination factor, the boundary isolation coefficient, and the anomaly index of each pipeline in the fire protection network, including: Based on the abnormal index and standard abnormal index of each pipe in the fire protection pipeline network, determine the pipe deviation of each pipe; The average pipe deviation is determined based on the pipe deviation of each pipe and the total number of pipes in the fire protection network; The failure rate of the fire protection pipeline network is determined based on the non-uniformity scaling factor, the directional coordination factor, the boundary isolation coefficient, and the average pipeline deviation.

[0014] Specifically, the inspection center, after determining the failure rate of the fire protection pipe network, also includes: If the failure rate of the fire protection pipeline network is greater than the failure threshold, the target abnormal pipeline is determined according to the magnitude of the abnormal index of each pipeline in the target pipeline segment.

[0015] Secondly, the present invention provides a method for online monitoring of tunnel fire water levels, specifically configured in an inspection center of an online monitoring system for tunnel fire water levels, the method comprising: Based on the historical water pressure sequence of each detection node, determine the water level difference sequence of each pipeline; Based on the target time water level difference of the pipeline in the water level difference sequence and the historical time water level difference, the anomaly index of the pipeline at the target time is determined, wherein the historical time is the detection time before the target time, and the anomaly index represents the extent to which the pipeline water level difference deviates from the normal water level difference. Based on the anomaly index of each pipe in the adjacent pipe group and the water level difference at the target time, the non-consistency coefficient of the adjacent pipe group at the target time is determined. The adjacent pipe group consists of two adjacent pipes in the fire protection pipe network along the water flow direction. The non-consistency coefficient characterizes the degree of synchronization of the abnormal trend changes between the adjacent pipes. The failure rate of the fire protection network at the target time is determined based on the inconsistency coefficient of each adjacent pipe group and the anomaly index of each pipe.

[0016] Beneficial Effects: This invention provides an online monitoring system for fire-fighting water levels in tunnels. By calculating anomaly indices, it quantifies the deviation of pipeline water level differences from normal levels, enabling accurate identification of abnormal conditions in individual pipelines. Simultaneously, by utilizing the anomaly indices of each pipeline in adjacent pipeline groups and the target time water level difference, a non-consistency coefficient is determined. Based on the synchronicity of abnormal trend changes between adjacent pipelines along the water flow direction, it distinguishes between genuine faults and disturbances caused by normal residential water use, effectively reducing false alarm rates. Finally, by combining the non-consistency coefficients of each adjacent pipeline group and the anomaly indices of each pipeline, the failure rate of the fire-fighting pipeline network is determined. This achieves multi-scale fault assessment from individual pipelines to pipeline groups and then to the entire network, improving the accuracy of fault detection and the safety of the fire-fighting pipeline network. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0018] Figure 1 This is a schematic diagram of the structure of an online tunnel fire water level monitoring system provided in an embodiment of the present invention; Figure 2 A flowchart of an online monitoring method for tunnel fire water level provided in an embodiment of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0020] Example 1 Figure 1 This is a schematic diagram of the structure of an online monitoring system for fire water levels in tunnels provided by an embodiment of the present invention. It is specifically applicable to the fault detection and fault location of fire protection pipe networks, and is particularly suitable for fault detection of tunnel fire protection pipe networks. The structure of tunnel fire protection pipe networks is generally a serial structure, that is, the pipes of each network are connected in a serial sequence.

[0021] like Figure 1As shown, the tunnel fire water level online monitoring system includes: a monitoring center 110 and pressure sensors 120 configured at each detection node of the fire pipeline network. The fire pipeline network between adjacent detection nodes forms a pipeline 130.

[0022] Specifically, the pressure sensor is used to collect the water pressure at multiple detection times of the detection node corresponding to the pressure sensor, forming a historical water pressure sequence.

[0023] Specifically, the detection nodes are the locations where pressure sensors are installed on the fire protection pipeline network. These can be connection points between various pipes in the network, including valve wells, fire hydrants, and sluice gates. The pressure sensors are pressure measuring devices installed at the detection nodes to collect water pressure data at multiple detection times. Each pressure sensor synchronously detects the water pressure at its corresponding location based on a predetermined detection time. For example, the pressure sensor's range can be 0~1.6 MPa, and its accuracy can be 0.5%FS. The pipelines are the sections of the fire protection pipeline network between two adjacent detection nodes, used to transmit fire-fighting water. Different pipelines are used for water source transmission to different physical entities, and their water pressures vary due to their different functional levels. The historical water pressure sequence is a data sequence formed by arranging the water pressure data collected by the pressure sensors at multiple detection times in chronological order, reflecting the temporal changes in water pressure at the detection nodes. For example, if the detection time interval is 1 minute, the length of the historical water pressure sequence can be 60, meaning the historical water pressure sequence includes water pressure data from the most recent hour.

[0024] The inspection center is used for: Based on the historical water pressure sequence of each detection node, the water level difference sequence of each pipeline is determined.

[0025] The water level difference sequence is a data sequence formed by arranging the water level differences of each pipeline in the order of detection time; the water level difference is calculated by the water pressure at the detection nodes at both ends of the pipeline.

[0026] Specifically, the method for determining the water level difference sequence is as follows: Based on the historical water pressure sequence of each detection node, determine the historical water level sequence of each detection node; For any pipe in the fire protection network at any detection time, determine the head detection node and tail detection node corresponding to the pipe. Based on the water level of the head detection node and the water level of the tail detection node at the detection time in the historical water level sequence, determine the water level difference of the pipe at the detection time. Generate a water level difference sequence based on the water level difference of each pipe at each detection time.

[0027] Specifically, since different detection nodes correspond to different fire protection physical entities, the historical water level sequence of the detection nodes includes the water level value of the detection node at each detection time corresponding to the water pressure. The specific calculation method is as follows: obtain the water flow density and gravitational acceleration of the fire protection pipe network; determine the water level corresponding to each detection time based on the water pressure, the water flow density and the gravitational acceleration at each detection time in the historical water pressure sequence, and the water level corresponding to each detection time constitutes the historical water level sequence of the detection node.

[0028] For example, water level value The calculation formula is: ,in, For water pressure, For water flow density, This is the acceleration due to gravity.

[0029] Specifically, for the pipe section between adjacent detection nodes, the detection node corresponding to the water flowing into the pipe (i.e., the detection node closer to the water supply end) is the pipe head detection node, and the detection node corresponding to the water flowing out of the pipe (the detection node farther from the water supply end) is the pipe tail detection node. The water level difference in the pipe can then be obtained by subtracting the water level at the tail detection node from the water level at the head detection node. Furthermore, after calculating the water level difference for each pipe at each detection time, the water level differences can be grouped based on the detection time or the pipe section to obtain a water level difference sequence. For example, the water level difference sequence is represented as follows: the historical water level difference corresponding to each historical detection time of the pipe and the current water level difference at the current detection time constitute the water level difference sequence.

[0030] Based on the target time water level difference and the historical time water level difference in the pipeline in the water level difference sequence, the anomaly index of the pipeline at the target time is determined, wherein the historical time is the detection time before the target time, and the anomaly index represents the extent to which the pipeline water level difference deviates from the normal water level difference.

[0031] Among them, the target time water level difference is the water level difference corresponding to the target time, which can be the current detection time, and the historical time is the detection time before the target time. The historical time water level difference is the water level difference corresponding to the detection time before the target time.

[0032] Specifically, the most direct manifestation of pipeline anomalies is abnormal water level differences, such as leaks caused by pipeline rupture. Therefore, the pipeline anomaly index is used to characterize the extent to which the pipeline's water level difference deviates from the normal water level difference. The normal water level difference of a pipeline can be determined by averaging multiple historical water level differences under normal safe operating conditions.

[0033] Optional methods for determining anomaly indices include: Based on the target water level difference in the pipeline and a preset number of historical water level differences, the standard deviation and mean of the water level difference are determined; based on the water level difference and mean of the water level difference at each detection time, the continuous offset factor of the pipeline is determined, wherein the continuous offset factor characterizes the direction of water level difference change in the pipeline at continuous detection times; based on the continuous offset factor, the target water level difference, the mean of the water level difference, and the standard deviation of the water level difference, the anomaly index of the pipeline is determined.

[0034] The preset quantity represents the historical data observation range for this anomaly detection. For example, the preset quantity can be 59, meaning the observation range is the water level difference corresponding to the 59 detection times before the target time. The mean water level difference is the average of the historical water level differences before the target time used in the calculation. Specifically, due to water flow and natural water level fluctuations, the water level difference in the pipeline is in a dynamic and stable process under normal conditions. If the water level difference continues to increase, it indicates that a leak may occur near the tail detection node; if the water level difference decreases, it indicates that a blockage may occur near the tail detection node. Therefore, the continuous offset factor represents the direction of water level difference change in the pipeline during continuous detection times, i.e., the duration of continuous increase or decrease in water level difference. The standard deviation of the water level difference is used to characterize the degree of deviation of the water level difference from the mean water level difference, and the calculation formula is: Where D is the standard deviation of the water level difference, and N is the total number of historical water level differences included in the calculation. The average water level difference. Let be the i-th historical water level difference of the pipeline.

[0035] Specifically, the calculation method for the pipeline's persistent offset factor is as follows: Based on the water level difference and the average water level difference at each detection time, the offset direction at each detection time is determined, wherein different offset directions characterize the relationship between the water level difference and the average water level difference at each detection time; the offset direction at the target time is determined as the reference direction; starting from the target time, backtracking from the nearest to the farthest in time, the number of detection times between the target time and the target historical time is determined, wherein the target historical time is the first historical detection time whose offset direction is different from the reference direction; the continuous offset factor is determined based on the number.

[0036] The offset direction at a given detection time represents the relationship between the water level difference at that time and the average water level difference, specifically including positive offset, negative offset, and no offset. If the water level difference at a given detection time is greater than the average water level difference, the offset direction at that time is positive; if the water level difference at a given detection time is less than the average water level difference, the offset direction at that time is negative; if they are equal, there is no offset. A continuous positive offset indicates a potential leakage risk near the detection node at the end of the pipeline; a continuous negative offset indicates a potential blockage risk near the detection node at the end of the pipeline. The reference direction is defined as the offset direction at the target time, used as a benchmark for comparing the continuity of offset directions when tracing back from the target time in chronological order from the nearest to the furthest point in time. The target historical time is the first historical detection time where the offset direction differs from the reference direction when tracing back from the target time in chronological order from the nearest to the furthest point in time.

[0037] For example, if the offset direction sequence is represented as [+,+,-,-,+,-,-], where different elements are arranged in chronological order of detection times, and the last element (-) represents the offset direction of the target time (negative offset), then the target historical time (+) is the third-to-last detection time, and the number of detection times between the target time and the target historical time is 1. In a special case, if there is no target historical time in the offset direction sequence, then L is (the length of the offset direction sequence - 1). The persistence offset factor can then be calculated using the following formula: in, Let be the continuous offset factor of pipe i at detection time t. This refers to the number of historical detection times between the target time and the target historical time. Specifically, compared to a single positive (negative) shift, a sustained positive (negative) shift, i.e., a long-term large (small) water level difference, indicates a higher degree of anomaly risk. Therefore, the persistence shift factor reflects the enhancing effect of this persistence on anomaly risk. Thus, if... The larger the value, the more significant the impact of persistent offset on anomalous results; however, to prevent... Too long With the continuous expansion, the embodiments of the present invention adopt the writing strategy of the above formula form, when When it is too big, If it can be close to 1, then its maximum influence weight is 1, and there will be no numerical explosion.

[0038] Furthermore, based on the pipeline's persistent deviation factor, the target time water level difference, the mean water level difference, and the standard deviation of the water level difference, the formula for determining the pipeline's anomaly index is as follows: in, Let be the anomaly index of pipeline i at detection time t. Let be the water level difference in pipe i at detection time t. The average water level difference. The standard deviation of the water level difference. As a persistent offset factor, To prevent extremely small positive values ​​from being divided by zero, it can be 0.001.

[0039] Specifically, when the hydraulics of the pipeline network are stable, Approaching 0, anomaly index A value close to 0 indicates no abnormality or a very small degree of abnormality. When the current water level difference is too large, it indicates that there may be a leak at the end of the pipe; when This indicates a possible blockage at the end of the pipe. Furthermore, when... Unchanged A continuous increase indicates that the water level difference is consistently shifting in the same direction. If the water level difference continues to increase or decrease, the likelihood of an anomaly is greater. Even with a continuous shift factor... Zero, factor The value remains at 1, thus preserving the contribution of the deviation from the target time itself to the anomaly index. Furthermore, different pipe sections in the network have varying lengths, diameters, and flow resistances, resulting in different fluctuations in water level differences (i.e., standard deviations) under normal operating conditions. The differences are inherent. For example, a 0.1-meter absolute deviation is considered severely abnormal (5 times the standard deviation) for a pipe section with normally minimal fluctuations (D=0.02), but within the normal noise range for a pipe section near the pump with normally larger fluctuations (D=0.2). Therefore, dividing by the standard deviation in the formula... The purpose of this is to standardize the degree of variation in water level differences between different pipelines, so as to eliminate the influence caused by the differences in the characteristics of the pipelines themselves, and make the anomaly index comparable between different pipelines.

[0040] Based on the anomaly index of each pipe in the adjacent pipe group and the water level difference at the target time, the non-consistency coefficient of the adjacent pipe group at the target time is determined. The adjacent pipe group consists of two adjacent pipes in the fire protection pipe network along the water flow direction. The non-consistency coefficient characterizes the degree of synchronization of the abnormal trend changes between the adjacent pipes.

[0041] In this context, an adjacent pipe group is a combination of two pipes that are adjacent to each other in the direction of water flow within the fire protection pipe network. Specifically, it can include the preceding and following pipes corresponding to the flow sequence of the water. The preceding pipe is the upstream pipe in the adjacent pipe group, defined by the direction of water flow, and the following pipe is the downstream pipe in the adjacent pipe group, defined by the direction of water flow.

[0042] Specifically, due to the hydraulic continuity between pipelines, anomalies at a specific location can affect other pipelines. Therefore, analyzing the degree of coordination among the anomaly trends of adjacent pipelines in terms of occurrence time, direction of change, and magnitude helps in inferring the type and location of the anomaly source. In other words, calculating the inconsistency coefficient allows for fault analysis and location.

[0043] Optional methods for determining the inconsistency coefficient include: The preceding and following pipelines in the adjacent pipeline group are determined by the direction of water flow. Based on the anomaly index of the preceding pipeline at the target time and the detection time preceding the target time, the change in the anomaly index of the preceding pipeline at the target time is determined. Based on the anomaly index of the following pipeline at the target time and the detection time preceding the target time, the change in the anomaly index of the following pipeline at the target time is determined. Based on the changes in the anomaly index of the preceding and following pipelines, a first factor is determined. Based on the water level difference between the preceding and following pipelines at the target time, a second factor is determined. Based on the first and second factors, the inconsistency coefficient of the adjacent pipeline group is determined.

[0044] Among them, the change in the abnormal index represents the magnitude of the change in the abnormal index of the pipeline at adjacent detection times. If the change in the abnormal index suddenly increases, it indicates that there is a greater possibility of an abnormality.

[0045] Specifically, the inconsistency coefficient Y of adjacent pipe groups can be calculated using the following formula: ; in, The first factor characterizes the anomalous exponential deviation between the preceding and following pipelines. Specifically, the first factor is... Where i is the preceding pipe in the adjacent pipe group, i+1 is the following pipe in the adjacent pipe group, and t is the target time. The change in the anomaly index of pipeline i at the target time t and the anomaly index at the previous detection time (t-1) is specifically... That is, the anomaly index at the target time minus the anomaly index at the previous detection time; The change in the anomaly index of pipeline i+1 at the target time t and the anomaly index at the previous detection time (t-1) is specifically... Furthermore, the closer the first factor is to 1, the greater the relative deviation between the abnormal index changes of the subsequent pipeline and the preceding pipeline, indicating high inconsistency; the closer the first factor is to 0, the smaller the relative deviation between the abnormal index changes of the subsequent pipeline and the preceding pipeline, indicating relatively consistent changes. It should be noted that if... and This indicates that there is normal noise in the preceding pipeline, while there is a significant change in the following pipeline. The inconsistency between the two suggests... =1, if and This indicates that both the preceding and following pipelines exhibit minimal noise, and their consistency is consistent. , This is a preset noise factor, used to prevent the first factor value from becoming abnormal due to noise.

[0046] The second factor characterizes the deviation in water level difference between preceding and subsequent pipelines. ,like If the value is very small, it indicates a high degree of consistency between the preceding and following pipelines, then the second factor ≈ 0. A large value indicates low consistency between the preceding and following pipelines, therefore the second factor ≈ 1. Further, special case analysis: if... This indicates that the water level difference in the preceding pipeline is equal in magnitude and opposite in direction to the water level difference in the following pipeline. Therefore, it means that the water level difference in the preceding pipeline increases while the water level difference in the following pipeline decreases. This is a serious inconsistency. .when At this point, it means that the water level difference between the front and rear pipes is zero, indicating that the front and rear pipes are completely identical. .

[0047] For example, consider a leak: suppose the preceding pipeline... When a leak occurs, the water level difference increases sharply, and the amplitude of the abnormal index changes accordingly. Meanwhile, subsequent pipelines Due to upstream leakage, the inflow rate decreased, the water level difference dropped, and the situation also deviated from normal, resulting in an increase in the magnitude of abnormal index changes. At this point, the abnormal changes in the two pipelines are in the same direction and the magnitude of the changes is relatively close, therefore Smaller, approaching zero, final The values ​​are also relatively small. This indicates that there is a synchronous abnormal change trend between adjacent pipes, which is consistent with the physical law that water leakage will simultaneously disturb the upstream and downstream pipe sections.

[0048] The failure rate of the fire protection network at the target time is determined based on the inconsistency coefficient of each adjacent pipe group and the anomaly index of each pipe.

[0049] Specifically, the method for determining the failure rate may include the following sub-steps: Sub-step 1: Determine the non-consistent scaling factor Since abnormal water flow changes involve multiple continuous pipes, the key to failure rate determination is to identify the target pipe segment affected by the abnormality, and then quantify the target pipe segment into a factor that can be used to calculate the failure rate based on the relevant data of the target pipe segment, that is, to determine the non-consistency ratio factor.

[0050] Specifically, identify target adjacent pipe groups whose inconsistency coefficient is less than a preset threshold at the target time; connect the pipes of each target adjacent pipe group according to the water flow sequence to obtain multiple continuous pipe segments, and determine the pipe segment with the most pipes as the target pipe segment; determine the inconsistency ratio factor based on the number of pipes in the target pipe segment and the total number of pipes in the fire protection pipe network.

[0051] The inconsistency coefficient is set to a value of 1. The preset threshold can be set to 0.5.

[0052] Specifically, the impact of anomalies on pipelines is continuous. For example, if pipelines are connected in the order of water flow as ABC, and an anomaly is located at pipeline A, its impact will not skip pipeline B and directly affect pipeline C. Therefore, the possible range of impact of an anomaly can only be A, AB, or ABC, and there will be no jumps like AC. Thus, a pipeline segment refers to a continuous set of pipelines affected by the same anomaly; the target pipeline segment is the pipeline segment that needs to be focused on during this anomaly detection, or it can be understood as the pipeline segment that meets preset detection conditions. These preset conditions can be based on the number of pipelines, such as the pipeline segment containing the most pipelines.

[0053] It should be added that, since the inconsistency coefficient corresponds to adjacent pipe groups, a pipe in a fire protection network may correspond to one or two inconsistency coefficients. In actual determination: if both inconsistency coefficients corresponding to a pipe are less than a preset threshold, then the pipe exhibits consistency with its preceding and following adjacent pipes, and the pipe can be included as an interior point of a continuous pipe segment; if only one of the two inconsistency coefficients corresponding to a pipe is less than the preset threshold, it is an edge pipe (endpoint) of a continuous pipe segment; if both inconsistency coefficients corresponding to a pipe are not less than the preset threshold, then the pipe does not belong to any continuous pipe segment. If a pipe in a fire protection network corresponds to one inconsistency coefficient, and that inconsistency coefficient is less than the preset threshold, then it is determined as an endpoint in a continuous pipe segment.

[0054] For example, the formula for calculating the non-consistency scaling factor R is as follows: Where L represents the number of pipes in the target pipe segment, and m represents the total number of pipes in the fire protection network. In this embodiment of the invention, the target pipe segment is the pipe segment containing the most pipes. The longer the target pipe segment, the wider the scope of the anomaly, and the richer the anomaly information contained therein. Therefore, using data from the target pipe segment for failure rate analysis yields relatively higher accuracy. In summary, R actually represents the proportion of the longest continuous pipe segment with a non-consistency coefficient less than a preset threshold among all pipes. Furthermore, since a smaller non-consistency coefficient indicates more synchronized anomaly changes between adjacent pipes, R reflects the proportion of pipes with highly synchronized anomaly trends among all pipes.

[0055] Sub-step 2: Determine the directional coordination factor Specifically, the directional coordination factor is determined based on the number of pipes with positive anomaly indices, the number of pipes with negative anomaly indices, and the total number of pipes in the target pipeline segment.

[0056] Specifically, directional synergy factor ,in, and These represent the number of pipes with positive and negative anomaly indices, respectively. L is the number of pipes in the target pipe segment. When L=0, it indicates that the target pipe segment does not exist, the fire protection network is normal, and the failure rate can be directly output as 0. When Q is large, it indicates... and The significant differences in quantity and the predominance of the abnormal direction (either the entire section shows a high water level difference or the entire section shows a low water level difference) indicate that the pipe section has been subjected to a uniform hydraulic disturbance, and this disturbance has had a widespread impact on the continuous pipeline. and When the quantities are close and the directions cancel each other out, this dimension does not show obvious consistency, the overall noise is large, and it does not help much in fault rate detection. Therefore, Q should be a smaller value.

[0057] Sub-step 3: Determine the boundary isolation coefficient Specifically, the boundary isolation coefficient is determined based on the inconsistency coefficient of each adjacent edge pipe group formed by the edge pipe in the target pipe segment and its adjacent non-target pipe segment, and the mean of the inconsistency coefficient of each adjacent internal pipe group in the target pipe segment that does not contain edge pipes.

[0058] In this context, the target pipe segment is considered a continuous set of pipes. Edge pipes refer to those within the target pipe segment that are located at the beginning or end of the flow path and connect to pipes outside the target pipe segment. The number of edge pipes (0, 1, or 2) depends on the relative position of the target pipe segment within the fire protection network. Specifically: 0 indicates the target pipe segment covers the entire fire protection network; 1 indicates the beginning or end of the target pipe segment coincides with the overall beginning or end of the fire protection network; 2 indicates neither the beginning nor the end of the target pipe segment overlaps with the overall beginning or end of the fire protection network. In an adjacent edge pipe group, one pipe is an edge pipe and the other is a non-edge pipe; in an internal adjacent pipe group, both pipes are non-edge pipes of the target pipe segment.

[0059] For example, the fire protection pipe network is ABCDEFGH.

[0060] If the target pipe segment is ABCDEFGH, then the edge pipe is 0, the adjacent edge pipe group does not exist, and the internal adjacent pipe group is AB, BC, CD, DE, EF, FG, GH.

[0061] If the target pipe segment is ABCD, then the edge pipe is D, the adjacent edge pipe group is DE, and the adjacent internal pipe groups are AB and BC.

[0062] If the target pipe segment example is BCDEF, then the edge pipes are B and F, the adjacent edge pipe groups are AB and FG, and the adjacent internal pipe groups are CD and DE.

[0063] If the target pipe segment example is CDE, then the edge pipes are C and E, the adjacent edge pipe groups are BC and EF, and the internal adjacent pipe groups do not exist.

[0064] For example, the formula for calculating the boundary isolation coefficient I of a target pipe segment containing two edge pipes is as follows: ; in, and These are the inconsistency coefficients of the adjacent edge pipe groups corresponding to the first and last edge pipes of the target pipe segment. This represents the mean of the inconsistency coefficients for each internal group of adjacent pipes. This is to find the minimum value to prevent the denominator from being zero. It should be noted that, generally speaking, It will be greater than or equal to 1, except in very rare special cases, due to and Extremely close, thus making Slightly larger This results in I being less than 1. Therefore, I < 1 is truncated to make I = 1. Furthermore, for a target pipe segment containing only one edge pipe, Similarly, if It is also truncated to 1. For target pipe segments without adjacent edge pipe groups or internal adjacent pipe groups, in order to effectively transmit local feature data downstream, I is preset to 2, indicating that the boundary conditions or internal conditions are unknown. In summary, the value of I is... .

[0065] Specifically, I represents the degree of isolation between the target pipeline segment and external pipeline segments, reflecting whether the abnormal change patterns within the target pipeline segment are effectively "blocked" by its edge pipelines, thus exhibiting a significant statistical difference from the external pipeline segments. As can be seen from the above conditions for generating the target pipeline group, the inconsistency coefficient of the edge-adjacent pipeline group of the target pipeline segment is greater than a threshold. Therefore, the inconsistency coefficient of the edge-adjacent pipeline group must be greater than the inconsistency coefficient of the internal adjacent pipeline group, i.e., I is not less than 1. The larger the I value, the more significant the boundary isolation effect, and the more effectively the anomaly is confined within the target pipeline segment.

[0066] Sub-step 4: Determine the failure rate of the fire protection network based on the non-consistency ratio factor, the directional coordination factor, the boundary isolation coefficient, and the anomaly index of each pipe in the fire protection network.

[0067] Specifically, the pipe deviation of each pipe is determined based on the abnormal index and standard abnormal index of each pipe in the fire protection pipe network; the average pipe deviation is determined based on the pipe deviation of each pipe and the total number of pipes in the fire protection pipe network; and the failure rate of the fire protection pipe network is determined based on the non-consistency ratio factor, the directional coordination factor, the boundary isolation coefficient, and the average pipe deviation.

[0068] The standard anomaly index is the average of the anomaly indices corresponding to each pipe in the fire protection network under normal operating conditions. The average pipe deviation characterizes the average net anomaly intensity of the pipes in the fire protection network.

[0069] Specifically, the formula for calculating the failure rate is: in, Let R be the failure rate of the fire protection network at time t, R be the non-uniformity scaling factor, Q be the directional coordination factor, and A be the degree of boundary isolation. A = I is the boundary isolation coefficient, and α is the scaling factor, used to adjust the influence of the boundary isolation coefficient on the overall failure rate, preset to α=0.5. B represents the average net anomaly intensity of the pipelines in the fire protection network, B= ], Let be the anomaly index of pipeline i at detection time t. This is the standard anomaly index for pipeline i under normal operating conditions. Under ideal conditions, In engineering practice, the value is 0. It can take values ​​in the range [-0.2, 0.2]. m is the total number of pipes in the fire protection network.

[0070] Specifically, a larger R indicates a wider range of abnormal impact and a higher probability of systemic failure. Q=1 indicates that all pipelines deviate in the same direction, with obvious fault characteristics; Q close to 0 indicates a mixture of positive and negative values, mostly random noise interference. I is determined by the exponential term ( Feature enhancement is performed when I=1 (neutral baseline): the exponent term is 0, meaning that the abnormal target pipeline segment is not significantly different from other pipelines, therefore, it may be a false alarm. Even in this case... The strength parameters are all high, and the failure rate is set to 0. When I>1, it indicates that the inconsistency coefficient of adjacent pipe groups at the edge is greater than that of adjacent pipe groups within the interior. That is, the consistency of the internal pipes is extremely high, the fault is confined to the target pipe segment, the confidence of fault location is high, and the failure rate increases. Furthermore, The average pipeline deviation characterizes the average net anomaly intensity of pipelines in the fire protection network. It is then standardized using the tanh function to obtain the relative confidence level of the anomaly index. =0 indicates that the average net anomaly intensity is 0, meaning there are no faults. A value close to 1 indicates a very high average net anomaly intensity and an extremely high failure rate in the fire protection pipeline network.

[0071] In summary, the failure rate calculation formula comprehensively calculates the failure rate from four aspects: R (spatial impact range), Q (consistency of anomaly direction), A (spatial location of anomaly), and B (average net anomaly intensity). Multiplying these four factors means that all conditions must be met simultaneously for the failure rate to increase significantly. This conservative strategy aligns with the safety requirement of "extremely high cost of false alarms" in tunnel fire protection, avoiding unnecessary alarms triggered by anomalies in a single indicator.

[0072] Furthermore, the inspection center, after determining the failure rate of the fire protection pipe network, also includes: If the failure rate of the fire protection pipeline network is greater than the failure threshold, the target abnormal pipeline is determined according to the magnitude of the abnormal index of each pipeline in the target pipeline segment.

[0073] Specifically, when the current failure rate is greater than the failure threshold (preset to 0.4), it indicates that there is a high probability of a failure in the fire protection pipeline network. Therefore, the target abnormal pipeline can be determined based on the magnitude of the abnormal index of each pipeline in the target pipeline segment. For example, the pipeline with the highest absolute value of the abnormal index in the target pipeline segment can be determined as the target abnormal pipeline. The target abnormal pipeline is the pipeline that may be blocked or leaking. In this way, the fault can be accurately located and the maintenance efficiency can be improved.

[0074] This invention provides an online monitoring system for fire-fighting water levels in tunnels. By calculating an anomaly index, it quantifies the deviation of pipeline water level differences from normal levels, enabling accurate identification of abnormal conditions in individual pipelines. Simultaneously, it determines a non-consistency coefficient using the anomaly index of each pipeline in adjacent pipeline groups and the target time water level difference. Based on the synchronicity of abnormal trend changes between adjacent pipelines along the water flow direction, it distinguishes between genuine faults and disturbances caused by normal residential water use, effectively reducing false alarm rates. Finally, it determines the fault rate of the fire-fighting pipeline network by comprehensively considering the non-consistency coefficients of each adjacent pipeline group and the anomaly index of each pipeline. This achieves comprehensive fault assessment from four aspects: spatial impact range, consistency of abnormal direction, spatial location of abnormality, and average net anomaly intensity, improving the accuracy of fault detection and the safety of the fire-fighting pipeline network.

[0075] Example 2 Figure 2 A flowchart of an online monitoring method for tunnel fire water level provided by an embodiment of the present invention. This method is specifically configured in the inspection center of an online monitoring system for tunnel fire water level, and includes: Step 210: Determine the water level difference sequence of each pipeline based on the historical water pressure sequence of each detection node.

[0076] Step 220: Determine the anomaly index of the pipeline at the target time based on the target time water level difference and the historical time water level difference in the pipeline in the water level difference sequence. The historical time is the detection time before the target time. The anomaly index represents the extent to which the pipeline water level difference deviates from the normal water level difference.

[0077] Step 230: Determine the non-consistency coefficient of the adjacent pipe group at the target time based on the abnormal index of each pipe in the adjacent pipe group and the water level difference at the target time. The adjacent pipe group consists of two adjacent pipes in the fire protection pipe network along the water flow direction. The non-consistency coefficient represents the degree of synchronization of abnormal trend changes between adjacent pipes.

[0078] Step 240: Determine the failure rate of the fire protection network at the target time based on the inconsistency coefficient of each adjacent pipe group and the abnormality index of each pipe.

[0079] The inspection center in the tunnel fire water level online inspection system provided in this embodiment of the invention can execute the tunnel fire water level online inspection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

Claims

1. A tunnel fire-fighting water level online monitoring system, characterized in that, The system includes an inspection center and pressure sensors configured at each detection node of the fire protection pipeline network. The fire protection pipeline network between adjacent detection nodes forms a pipeline. The pressure sensor is used to collect the water pressure at the corresponding detection node at multiple detection times to form a historical water pressure sequence. The inspection center is used for: Based on the historical water pressure sequence of each detection node, determine the water level difference sequence of each pipeline; Based on the target time water level difference of the pipeline in the water level difference sequence and the historical time water level difference, the anomaly index of the pipeline at the target time is determined, wherein the historical time is the detection time before the target time, and the anomaly index represents the extent to which the pipeline water level difference deviates from the normal water level difference. Based on the anomaly index of each pipe in the adjacent pipe group and the water level difference at the target time, the non-consistency coefficient of the adjacent pipe group at the target time is determined. The adjacent pipe group consists of two adjacent pipes in the fire protection pipe network along the water flow direction. The non-consistency coefficient characterizes the degree of synchronization of the abnormal trend changes between the adjacent pipes. The failure rate of the fire protection network at the target time is determined based on the inconsistency coefficient of each adjacent pipe group and the anomaly index of each pipe.

2. The system according to claim 1, characterized in that, The inspection center is used to determine the water level difference sequence of each pipeline based on the historical water pressure sequence of each detection node, including: Based on the historical water pressure sequence of each detection node, determine the historical water level sequence of each detection node; For any pipe in the fire protection network at any detection time, determine the head detection node and tail detection node corresponding to the pipe, and determine the water level difference of the pipe at the detection time based on the water level of the head detection node and the water level of the tail detection node in the historical water level sequence at the detection time. A water level difference sequence is generated based on the water level difference of each pipeline at each detection time.

3. The system according to claim 2, characterized in that, The inspection center is used to determine the historical water level sequence of each of the aforementioned detection nodes based on the historical water pressure sequence, including: Obtain the water flow density and gravitational acceleration of the fire protection pipeline network; Based on the water pressure, water flow density, and gravitational acceleration at each detection time in the historical water pressure sequence, the water level corresponding to each detection time is determined, and the water level corresponding to each detection time constitutes the historical water level sequence of the detection node.

4. The system according to claim 1, characterized in that, The inspection center is used to determine the anomaly index of the pipeline at the target time based on the target time water level difference and the historical time water level difference in the pipeline water level difference sequence, including: Based on the target water level difference in the pipeline and a preset number of historical water level differences, determine the standard deviation and mean value of the water level difference. The continuous offset factor of the pipeline is determined based on the water level difference and the average water level difference at each detection time. The continuous offset factor characterizes the direction of water level difference change in the pipeline at continuous detection times. The anomaly index of the pipeline is determined based on the pipeline's continuous offset factor, the target time water level difference, the mean water level difference, and the standard deviation of the water level difference.

5. The system according to claim 4, characterized in that, The inspection center is used to determine the continuous deviation factor of the pipeline based on the water level difference and the average water level difference at each inspection time, including: Based on the water level difference and the average water level difference at each detection time, the offset direction at each detection time is determined. Different offset directions represent the relationship between the water level difference and the average water level difference at each detection time. The offset direction at the target time is determined as the reference direction; Starting from the target time, backtracking from the nearest to the farthest in time, determine the number of detection times between the target time and the target historical time, wherein the target historical time is the first historical detection time whose offset direction is different from the reference direction; The persistence offset factor is determined based on the stated quantity.

6. The system according to claim 1, characterized in that, The inspection center is used to determine the inconsistency coefficient of the adjacent pipeline group at a target time based on the anomaly index of each pipeline in the adjacent pipeline group and the water level difference at the target time, including: Determine the preceding and following pipes in the adjacent pipe group, based on the direction of water flow; Based on the anomaly index of the preceding pipeline at the target time and the detection time before the target time, determine the change in the anomaly index of the preceding pipeline at the target time. Based on the anomaly index of the subsequent pipeline at the target time and the detection time before the target time, determine the change in the anomaly index of the subsequent pipeline at the target time. The first factor is determined based on the abnormal index change of the preceding pipeline and the abnormal index change of the following pipeline. The second factor is determined based on the target time water level difference between the preceding pipeline and the target time water level difference between the following pipeline; The inconsistency coefficient of the adjacent pipe group is determined based on the first factor and the second factor.

7. The system according to claim 6, characterized in that, The inspection center is used to determine a second factor based on the target time water level difference between the preceding pipeline and the following pipeline, including: The rate of change of water level difference is determined based on the target time water level difference of the preceding pipeline and the target time water level difference of the following pipeline. The water level difference change rate is mapped to a preset numerical range to obtain the second factor.

8. The system according to claim 1, characterized in that, The inspection center is used to determine the failure rate of the fire protection network at a target time based on the inconsistency coefficient of each adjacent pipe group and the anomaly index of each pipe, including: Identify target adjacent pipeline groups whose inconsistency coefficient is less than a preset threshold at the target time. Connect the pipes of each adjacent pipe group of each target according to the water flow sequence to obtain multiple continuous pipe segments. The pipe segment containing the most pipes is determined as the target pipe segment. The non-consistency ratio factor is determined based on the number of pipes in the target pipe section and the total number of pipes in the fire protection pipe network. The directional coordination factor is determined based on the number of pipes with positive anomaly indices, the number of pipes with negative anomaly indices, and the total number of pipes in the target pipeline segment. The boundary isolation coefficient is determined based on the inconsistency coefficient of each adjacent edge pipe group consisting of edge pipes in the target pipe segment and their adjacent non-target pipe segments, and the mean of the inconsistency coefficient of each adjacent internal pipe group in the target pipe segment that does not contain edge pipes. The failure rate of the fire protection network is determined based on the non-consistency ratio factor, the directional coordination factor, the boundary isolation coefficient, and the anomaly index of each pipe in the fire protection network.

9. The system according to claim 8, characterized in that, The inspection center is used to determine the failure rate of the fire protection network based on the inconsistency ratio factor, the directional coordination factor, the boundary isolation coefficient, and the anomaly index of each pipeline in the fire protection network, including: Based on the abnormal index and standard abnormal index of each pipe in the fire protection pipeline network, determine the pipe deviation of each pipe; The average pipe deviation is determined based on the pipe deviation of each pipe and the total number of pipes in the fire protection network; The failure rate of the fire protection pipeline network is determined based on the non-uniformity scaling factor, the directional coordination factor, the boundary isolation coefficient, and the average pipeline deviation.

10. The system according to claim 8, characterized in that, After determining the failure rate of the fire protection pipeline network, the inspection center is also used for: If the failure rate of the fire protection pipeline network is greater than the failure threshold, the target abnormal pipeline is determined according to the magnitude of the abnormal index of each pipeline in the target pipeline segment.

Images