A comprehensive pipe gallery water level monitoring processing method and system

Abstract

The application provides a comprehensive pipe gallery water level monitoring processing method and system, relates to the technical field of water level monitoring, and the method comprises the following steps: collecting water level data of each monitoring point of the comprehensive pipe gallery in real time by using a water level sensor in a distributed sensor network, combining a pre-constructed three-dimensional model, constructing a water level spatial distribution map by using a Kriging interpolation algorithm, dividing risk levels according to water level data thresholds and dynamically dividing drainage partitions according to the risk levels; if the risk levels do not meet expectations, starting water pump drainage of the corresponding drainage partition, and pushing alarm information to an operation and maintenance terminal through an Internet of Things platform. The application can realize more refined monitoring on the water level condition in the comprehensive pipe gallery.

Description

Technical Field

[0001] This application relates to the technical field of water level monitoring, and in particular to a method and system for monitoring and processing water levels in integrated utility tunnels. Background Technology

[0002] With the acceleration of urbanization, urban underground utility tunnels, as a crucial component of urban infrastructure, undertake the laying of various pipelines such as electricity, communications, gas, and water supply and drainage. They effectively solve problems such as repeated road excavation and pipeline safety, improving the efficiency and safety of urban operations. However, utility tunnels are typically located in enclosed underground environments, making them susceptible to water accumulation due to rainfall and groundwater infiltration. This water accumulation not only damages the various pipelines within the tunnel, affecting their normal operation, but can also lead to safety accidents such as electric shock and gas leaks, posing a serious threat to the safe and stable operation of the city. Therefore, real-time and accurate monitoring of water levels in utility tunnels, timely detection of anomalies, and effective measures to address them are of significant practical importance.

[0003] Chinese invention patent application CN120410188A discloses an intelligent risk monitoring system and method for underground utility tunnels based on multi-source heterogeneous data. This patent collects multi-source heterogeneous data to conduct an overall risk assessment of underground utility tunnels, generating risk multivariate coefficients and early warning signals. This method generates early warning signals based on a global risk coefficient threshold; however, local water level anomalies may not reach the global threshold standard (e.g., a short-term rise in water level in a certain area without triggering an overall risk escalation), thus posing a risk of missed warnings. Summary of the Invention

[0004] In order to achieve more precise monitoring of water levels within integrated utility tunnels, this application provides a method and system for monitoring and processing water levels in integrated utility tunnels.

[0005] Firstly, this application provides a method for monitoring and processing water levels in integrated utility tunnels, employing the following technical solution:

[0006] A method for monitoring and managing water levels in integrated utility tunnels includes the following steps:

[0007] Real-time water level data of each monitoring point in the integrated utility tunnel is collected by water level sensors in the distributed sensor network. Based on the real-time water level data and the pre-built three-dimensional model of the integrated utility tunnel, a water level spatial distribution map is constructed using the Kriging interpolation algorithm. The water level spatial distribution map is divided into different risk levels using water level data thresholds, and drainage zones are dynamically divided according to the risk levels.

[0008] When the risk level does not meet expectations, the water pumps in the corresponding drainage zone are activated to drain the water, and alarm information is pushed to the operation and maintenance terminal through the Internet of Things platform.

[0009] This application utilizes water level sensors in a distributed sensor network to collect real-time water level data from various monitoring points within the integrated utility tunnel. This coverage allows for comprehensive monitoring of water levels at different locations within the tunnel, minimizing blind spots. Subsequently, based on the real-time water level data and a pre-built 3D model of the integrated utility tunnel, this application employs a Kriging interpolation algorithm to construct a spatial distribution map of the water level. This transforms abstract water level data into intuitive graphics, enabling maintenance personnel to clearly understand the spatial distribution of water levels in different areas of the integrated utility tunnel and quickly locate areas with high water levels.

[0010] This application divides the spatial distribution map of water levels into different risk levels by setting water level data thresholds. A quantitative risk assessment method provides a clear basis for operation and maintenance decisions, enabling maintenance personnel to take corresponding measures based on different risk levels and improve response efficiency. Subsequently, this application dynamically divides drainage zones according to risk levels. This intelligent zoning method can flexibly adjust drainage strategies based on the division of drainage zones. When the water level risk in certain areas is high, the pumps in the corresponding drainage zones can be activated specifically for drainage, reducing unnecessary energy waste and improving the operational efficiency of the drainage system. Through comprehensive monitoring, accurate assessment, and timely handling of abnormal water levels, this application effectively reduces the risk of safety accidents caused by water accumulation in the integrated utility tunnel, ensuring the safety of facilities and personnel within the tunnel. The intelligent monitoring, control, and alarm system reduces the workload of manual inspections and operations, improving the efficiency and accuracy of operation and maintenance management.

[0011] Optionally, each monitoring point is equipped with a first sensor and a second sensor. The range of the first sensor is smaller than that of the second sensor. When the water level data monitored by the first sensor reaches the warning threshold, the second sensor is triggered to monitor the water level data synchronously. When the difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor at the same monitoring point exceeds the preset difference threshold, an alarm signal is issued.

[0012] This application incorporates a first sensor and a second sensor at each monitoring point. The first sensor has a smaller measurement range, making it more sensitive to low water level changes and enabling more accurate detection of minute fluctuations. The second sensor has a larger measurement range, allowing it to monitor higher water levels. The combination of these two sensors enables effective monitoring of different water level stages, minimizing monitoring blind spots caused by range limitations. Because the first sensor has a smaller measurement range, it primarily performs monitoring tasks under normal water level conditions, reducing the number of times the second sensor operates at low water levels. This also reduces wear and tear on the second sensor, extending the overall lifespan of the monitoring equipment and lowering maintenance and replacement costs.

[0013] When the water level data monitored by the first sensor reaches the warning threshold, the second sensor is triggered to simultaneously monitor the water level data. This application uses a highly sensitive first sensor for initial warning assessment, followed by further confirmation by a second sensor with a larger measurement range, reducing the possibility of false alarms and improving alarm accuracy. In the event of a sudden and significant rise in the water level in the utility tunnel, the first sensor issues a warning first, and the second sensor responds quickly and begins synchronous monitoring. Even if the water level rises beyond the measurement range of the first sensor, the second sensor can continue monitoring, ensuring that no critical water level change information is missed, providing more reliable data support for responding to emergencies.

[0014] Subsequently, this application compares the water level data monitored by the first and second sensors at the same monitoring point. When the difference between the two exceeds a preset difference threshold, an alarm signal is issued. Based on the difference in the monitoring data of the two sensors, maintenance personnel can make a preliminary judgment on whether it is a sensor malfunction or an actual water level anomaly. This helps maintenance personnel to quickly locate the problem, reduce the time and cost of troubleshooting, and improve maintenance efficiency.

[0015] Optionally, when the difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor at the same monitoring point exceeds a preset difference threshold, the method further includes:

[0016] Monitoring points where the difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor exceeds a preset difference threshold are recorded as abnormal locations. Water level data at adjacent monitoring points of the abnormal location are obtained. Based on the water level data at adjacent monitoring points, an inverse distance weighted interpolation algorithm is used to calculate the theoretical water level data at the abnormal location and update the water level distribution map.

[0017] This application identifies anomaly locations by marking monitoring points with differences exceeding a preset threshold as abnormal locations, thereby quickly and accurately pinpointing areas within the integrated utility tunnel where water level data is abnormal, thus improving the efficiency and accuracy of anomaly location. Subsequently, this application obtains water level data from monitoring points adjacent to the anomaly location. By analyzing data from normal monitoring points surrounding the anomaly location, a reliable basis for calculating the theoretical water level at the anomaly location is provided. An inverse distance weighted interpolation algorithm is used to calculate the theoretical water level data at the anomaly location, fully considering the spatial distance relationship between adjacent monitoring points and the anomaly location, and rationally allocating the contribution weight of each adjacent point's data to the theoretical water level. This makes the calculated theoretical water level data more scientific, reasonable, and closer to the actual situation.

[0018] When abnormal data is detected at monitoring points, this application can compensate for the impact of abnormal data to a certain extent by calculating the theoretical water level and updating the distribution map, thereby improving the continuity and stability of the water level monitoring process. Even if some monitoring points have data problems, this application can still provide relatively accurate water level distribution information, ensuring the normal operation of the integrated utility tunnel.

[0019] Optionally, when the risk level meets expectations, the method further includes:

[0020] Historical water level data and historical water level change rate at each monitoring point are obtained. All historical water level data are integrated into a historical water level data sequence according to the order of collection time, and historical water level change rate is integrated into a historical water level change rate sequence. A neural network model is constructed, and the neural network model is trained using the historical water level data sequence and the historical water level change rate sequence to obtain the trained neural network model.

[0021] Real-time water level data is input into the trained neural network model, which outputs the predicted rate of change. When the predicted rate of change exceeds a preset rate threshold, the pump drainage rate is increased.

[0022] This application acquires historical water level data and historical water level change rates at various monitoring points and integrates them into sequences to fully leverage the value of the large amount of historical data accumulated during the monitoring of water levels in integrated utility tunnels. This historical data includes water level changes under different times and operating conditions. Subsequently, this application trains a neural network model using the integrated historical water level data sequences and historical water level change rate sequences. The neural network model can automatically learn the complex relationship between water level changes and water level change rates from historical data, thereby constructing a model that closely reflects the actual water level change patterns of integrated utility tunnels. Compared to traditional linear prediction methods, it can more accurately reflect the dynamic changes in water levels.

[0023] Optionally, the method further includes:

[0024] Various structural parameters at each monitoring point are obtained. Based on the open channel flow theory and Bernoulli equation in fluid mechanics, the influence factor of the structural parameters on the water level change rate is calculated. The product of the predicted change rate and the influence factor is used as the new predicted change rate.

[0025] The structure of integrated utility tunnels is complex and diverse. Structural parameters at different monitoring points (such as the slope, cross-sectional shape, and roughness of the tunnel) can significantly affect the flow state of water, thereby influencing the rate of water level change. This application obtains these structural parameters and calculates their influence factors on the rate of water level change based on open channel flow theory and Bernoulli's equation in fluid mechanics. This allows for a more comprehensive and accurate consideration of the impact of actual engineering factors on water level changes, overcoming the shortcomings of predictions based solely on historical water level data and making the prediction results closer to reality.

[0026] Integrated utility tunnels may operate under different conditions at different times. For example, during maintenance, some areas within the tunnel may be closed or the water flow path may be altered, causing changes in the impact of structural parameters on water level changes. This application can dynamically calculate the influencing factors based on actual structural parameters and correct the predicted rate of change in real time, thereby adapting to various complex operating conditions and maintaining high prediction accuracy. Because the new predicted rate of change more accurately reflects the actual situation of water level changes, this application can respond more quickly when abnormal water level trends occur, adjust drainage strategies in advance, shorten the time interval from the discovery of water level anomalies to the implementation of effective control measures, effectively reduce the risk of damage to the structure and facilities of the integrated utility tunnel caused by rising water levels, and improve emergency response capabilities.

[0027] This application comprehensively considers the influence of various factors such as historical water level data and structural parameters on water level changes, forming a more comprehensive and reliable prediction model. Compared with single-factor prediction methods, multi-factor comprehensive prediction can better cope with various uncertainties and improve anti-interference ability and stability.

[0028] Optionally, the structural parameters include the cross-sectional dimensions of the pipe gallery, the slope, and the roughness coefficient, and the influencing factors are calculated through the following steps:

[0029] Calculation of hydraulic radius and Chezy coefficient of water flow in integrated utility tunnel based on open channel flow theory;

[0030] Based on the hydraulic radius and Chezy coefficient, and combined with Bernoulli's equation, the correlation function between the water level change rate and various structural parameters is derived.

[0031] The weight coefficients corresponding to each structural parameter are calculated based on the correlation function, and the weight coefficients are standardized and used as the influence factors.

[0032] This application calculates the hydraulic radius of water flow within a utility tunnel based on open channel flow theory. The hydraulic radius comprehensively considers the influence of the tunnel's cross-sectional dimensions on the water flow. Utility tunnels with different cross-sectional dimensions have different water passage sections and hydraulic perimeters. The hydraulic radius accurately characterizes the relationship between the effective range of water flow and boundary resistance. This application also calculates the Chezy coefficient in conjunction with the hydraulic radius. The Chezy coefficient reflects the influence of pipe roughness on water flow resistance. As a structural parameter, the roughness coefficient is linked to water flow resistance through the Chezy coefficient, making the calculation process more consistent with the actual water flow within the utility tunnel and reducing calculation deviations caused by neglecting roughness.

[0033] Based on the hydraulic radius and Chezy coefficient, this application derives the correlation function between the water level change rate and various structural parameters by combining the Bernoulli equation. The Bernoulli equation describes the principle of energy conservation in the flow process of water. By introducing the hydraulic radius and Chezy coefficient, the complex relationship between structural parameters and the water level change rate is clearly expressed in the form of a mathematical function, so that the prediction of the water level change rate has a more scientific and accurate basis.

[0034] This application calculates the weighting coefficients corresponding to each structural parameter based on the correlation function to quantify the influence of each structural parameter on the rate of water level change. Then, this application standardizes the weighting coefficients to obtain the influence factor. Since the influence factor accurately reflects the impact of each structural parameter on the rate of water level change, multiplying the predicted rate of change by the influence factor to obtain the new predicted rate of change allows for a more accurate prediction of future water level trends. Adjusting the pump drainage rate based on this more accurate prediction can reduce the problems of insufficient or excessive drainage caused by inaccurate predictions.

[0035] Optionally, the structural parameters further include pipe corrosion rate, historical leakage count, and concrete carbonation depth; the method further includes:

[0036] The aging index of the integrated utility tunnel is calculated based on the pipeline corrosion rate, historical leakage frequency, and carbonation depth of the concrete structure. The attenuation factor of the water level data threshold is set based on the aging index, and the product of the water level data threshold and the attenuation factor is updated to the water level data threshold.

[0037] Over time, the corrosion rate of pipelines, the number of historical leaks, and the carbonation depth of concrete structures will continuously change, and the aging index will also be dynamically adjusted accordingly. This application calculates the aging index of the integrated utility tunnel based on these three aging indicators, and adjusts the water level data threshold used to classify risk levels based on the aging index, so that risk assessment can keep up with the aging process of the integrated utility tunnel structure. Setting different water level data thresholds at different aging stages can better balance the needs of safe operation and normal use of the integrated utility tunnel.

[0038] Optionally, the method further includes:

[0039] The sampling frequency adjustment coefficient k is calculated based on the predicted rate of change and leakage frequency at each monitoring point in the drainage zone. The calculation model is as follows:

[0040] ;

[0041] in, and These are the weighting coefficients; Let be the predicted rate of change at the i-th monitoring point; This represents the maximum permissible water flow rate at the i-th monitoring point; Let i be the number of leaks at the i-th monitoring point; This represents the total number of monitoring times at the i-th monitoring point;

[0042] The sampling frequency of the water level sensor at each monitoring point is adjusted using an adjustment coefficient k.

[0043] This application increases the data collection frequency at monitoring points with large predicted change rates, enabling more timely acquisition of water level data. When abnormal situations such as rapid rises or falls in water level occur, this application can react quickly and issue early warning signals in a timely manner, giving operation and maintenance personnel more time to take countermeasures, thereby effectively reducing the risk of damage to the integrated utility tunnel caused by abnormal water levels.

[0044] Different monitoring points exhibit varying water level change characteristics and risk levels. By adjusting the sampling frequency of the water level sensors at each monitoring point using a sampling frequency adjustment coefficient, sensor resources can be allocated more rationally. This concentrates more resources on key monitoring points, improving data acquisition quality and real-time performance, and enhancing resource utilization efficiency. The calculation model for the sampling frequency adjustment coefficient considers dynamic factors such as the predicted rate of change and the number of leaks, enabling the water level monitoring system to automatically adjust the sampling frequency according to the actual conditions of the integrated utility tunnel, achieving adaptive monitoring and better coping with various complex and changing operating conditions.

[0045] Optionally, the method further includes:

[0046] Based on the risk level and spatial location of each monitoring point, the integrated utility tunnel is divided into several initial partitions using the K-means clustering algorithm;

[0047] Obtain the average water level change rate and connectivity parameters of each initial zone, calculate the hydraulic conduction coefficient of adjacent initial zones, and merge adjacent initial zones to form a new drainage zone when the hydraulic conduction coefficient of adjacent initial zones exceeds the preset conduction threshold. Configure the pump operating parameters according to the new drainage zone.

[0048] This application combines the risk level and spatial location of each monitoring point, employing the K-means clustering algorithm to divide the initial zones, thus fully considering the actual conditions of the integrated utility tunnel. Areas with high risk levels may have structural defects or be prone to leakage; rationally dividing them from surrounding areas facilitates differentiated management of areas with different risk levels, improving the targeting of management. This application does not simply divide areas by geographical region, but comprehensively considers risk and spatial factors, ensuring relatively consistent hydraulic characteristics within each initial zone.

[0049] This application obtains the average water level change rate and connectivity parameters of each initial zone and calculates the hydraulic conduction coefficient of adjacent initial zones. When the hydraulic conduction coefficient exceeds a preset conduction threshold, it indicates that the water flow between adjacent initial zones is closely connected, and they are merged into a new drainage zone. This dynamic adjustment mechanism can optimize the zones in real time according to the actual hydraulic conditions within the integrated utility tunnel, making the drainage zones more consistent with the actual water flow. During operation, the integrated utility tunnel may be affected by various factors, such as changes in rainfall and alterations in leakage due to structural aging. These factors can affect water level changes and hydraulic conduction characteristics within the integrated utility tunnel. By dynamically merging initial zones, this application can automatically adapt to these changes, maintaining the rationality and effectiveness of the zones at all times, and improving the adaptability and stability of the entire water level monitoring and control system.

[0050] Secondly, this application provides an integrated utility tunnel water level monitoring and processing system, which adopts the following technical solution:

[0051] A comprehensive utility tunnel water level monitoring and processing system includes: a memory and a processor.

[0052] The memory contains a computer-readable storage medium;

[0053] When the processor processes a computer program stored on the computer-readable storage medium, it implements the method as described in the first aspect.

[0054] In summary, this application includes at least one of the following beneficial technical effects:

[0055] 1. This application utilizes water level sensors in a distributed sensor network to collect real-time water level data from various monitoring points within the integrated utility tunnel. This coverage allows for comprehensive monitoring of water levels at different locations within the tunnel, minimizing blind spots. Subsequently, based on the real-time water level data and a pre-built 3D model of the integrated utility tunnel, this application employs a Kriging interpolation algorithm to construct a spatial distribution map of the water level. This transforms abstract water level data into intuitive graphics, enabling maintenance personnel to clearly understand the spatial distribution of water levels in different areas of the integrated utility tunnel and quickly locate areas with high water levels.

[0056] 2. This application effectively reduces the risk of safety accidents caused by water accumulation in the integrated utility tunnel by comprehensively monitoring, accurately assessing and promptly handling abnormal water levels, ensuring the safety of facilities and personnel within the integrated utility tunnel. The intelligent monitoring, control and alarm system reduces the workload of manual inspection and operation, and improves the efficiency and accuracy of operation and maintenance management. Attached Figure Description

[0057] Figure 1 This is a flowchart of Embodiment 1 of this application;

[0058] Figure 2 This is a flowchart of Embodiment 3 of this application. Detailed Implementation

[0059] The following combination Figure 1 and Figure 2 This application will be described in further detail.

[0060] Example 1: This example discloses a method for monitoring and processing water levels in integrated utility tunnels, referring to... Figure 1 The method includes: S11 area division and S12 risk assessment. First, water level data from monitoring points in the integrated utility tunnel are collected in real time using water level sensors in a distributed sensor network. Combined with a pre-built 3D model, a spatial distribution map of water level is constructed using the Kriging interpolation algorithm. Risk levels are divided according to water level data thresholds, and drainage zones are dynamically divided accordingly. If the risk level does not meet expectations, the water pumps in the corresponding drainage zone are activated to drain water. At the same time, alarm information is pushed to the operation and maintenance terminal through the Internet of Things platform. The execution process of each step in this embodiment is as follows:

[0061] S11 is a designated area. In the complex underground space environment of the integrated utility tunnel, a distributed sensor network is constructed. This network consists of numerous water level sensors, which are deployed at various monitoring points within the integrated utility tunnel to collect real-time water level data.

[0062] A pre-built 3D model of a utility tunnel is a digital representation of its physical structure. This model includes detailed information about the tunnel's geometry, dimensions, layout, and internal facilities, such as pipeline routes and support locations. 3D modeling technology allows for the intuitive graphical representation of the complex structure of the utility tunnel, providing a spatial reference framework for simulating and analyzing water level distribution.

[0063] Kriging interpolation is a spatial interpolation method based on geostatistical principles. It can reasonably estimate and interpolate the water level in unknown areas using water level data from known monitoring points. In the water level analysis of integrated utility tunnels, the real-time water level data of each monitoring point where the water level sensor is located is known, while the water level conditions at other locations within the integrated utility tunnel are unknown.

[0064] Kriging interpolation takes into account the autocorrelation of spatial data, meaning that there is a certain correlation between water level data at adjacent locations. By analyzing the spatial relationships and water level change patterns between known monitoring points, Kriging interpolation can generate a smooth spatial distribution surface of water levels, thereby constructing a spatial distribution map of water levels within the integrated utility tunnel. This spatial distribution map graphically displays the water level levels at different locations within the integrated utility tunnel, allowing maintenance personnel to intuitively understand the overall water level situation within the tunnel.

[0065] To quantify the water level risk within the integrated utility tunnel, it is necessary to set water level data thresholds. These thresholds are determined comprehensively based on factors such as the design standards of the integrated utility tunnel, historical water level data, and safe operation requirements. The water level data thresholds include: low water level threshold, medium water level threshold, and high water level threshold.

[0066] Based on the real-time constructed spatial distribution map of water levels, the real-time water level data at each location is compared with a set water level threshold. If the real-time water level data is below the low water level threshold, the area is considered to be at a low risk level; if the real-time water level data is between the low and medium water level thresholds, it is at a medium risk level; and if the real-time water level data is above the high water level threshold, it is at a high risk level. In this way, the spatial distribution map of water levels is divided into different risk levels.

[0067] Based on the risk level classification results of the water level spatial distribution map, drainage zones are dynamically divided. High-risk areas are designated as independent drainage zones for priority drainage treatment. Adjacent areas with similar risk levels are merged into one drainage zone for unified management of drainage equipment. The dynamically divided drainage zones can be adjusted in real time according to water level changes, ensuring the rational allocation and effective utilization of drainage resources.

[0068] In other embodiments, the interconnected areas of integrated utility tunnels corresponding to the same risk level can also be divided into a drainage zone.

[0069] S12 Risk Assessment: After the drainage zoning is completed, this embodiment will assess the risk level of each drainage zoning zone in real time to determine whether it meets expectations. A drainage zoning zone with a low risk level is considered to meet expectations, while a drainage zoning zone with a medium or high risk level is considered to not meet expectations.

[0070] If the risk level of a drainage zone does not meet expectations, for example, if the risk level of a drainage zone suddenly rises to a high risk level, it indicates that there is a risk of water accumulation in the area. Drainage measures need to be taken in a timely manner, the water pumps of the corresponding drainage zone will be automatically started to drain water, and alarm information will be pushed to the operation and maintenance terminal through the Internet of Things platform.

[0071] If the risk level of the drainage zone meets expectations, no action will be taken.

[0072] This embodiment enables real-time monitoring, risk assessment, and timely handling of water levels in integrated utility tunnels, effectively reducing the risk of water accumulation within the tunnels and ensuring the normal operation of facilities and the safety of personnel.

[0073] Example 2: This example differs from Example 1 in that the method further includes:

[0074] At each monitoring point, a first sensor and a second sensor are installed. The first sensor has a smaller measurement range, meaning it has higher accuracy and resolution, enabling more precise measurement of changes in water level from low to medium thresholds. The second sensor has a larger measurement range, capable of measuring a wider range of water levels, suitable for monitoring environments where extreme water levels may occur. When real-time water level data exceeds the range of the first sensor, the second sensor can continue to collect real-time water level data, ensuring no data loss due to excessively high water levels.

[0075] Because some water level sensors, such as piezoresistive water level sensors, ultrasonic water level sensors, or some low-precision water level sensors, may have an increased measurement error range near the end of their range, leading to unclear measurement results, this embodiment triggers a second sensor when the water level data monitored by the first sensor reaches a warning threshold. This second sensor then monitors the water level data at the same frequency and time as the first sensor, with the warning threshold being equal to 80% of the range of the first sensor. In other embodiments, other values ​​can be set as needed.

[0076] When the difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor at the same monitoring point exceeds a preset difference threshold, an alarm signal is issued. At the same time, the monitoring point where the difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor exceeds the preset difference threshold is recorded as an abnormal location.

[0077] Water level data from adjacent monitoring points at the anomaly location are obtained. Based on the water level data from the adjacent monitoring points, an inverse distance weighted interpolation algorithm is used to calculate the theoretical water level data at the anomaly location, and the water level distribution map is updated. The calculation process for the theoretical water level data is as follows:

[0078] Let the anomaly location be P0, and its adjacent monitoring points be P1, P2, ..., P1. n The real-time water level data of adjacent monitoring points are Z1, Z2, ..., Z n Adjacent monitoring point P i The distance to the abnormal location P0 is d i The formula for calculating the theoretical water level Z0 at the abnormal location P0 is:

[0079] ;

[0080] Where m is a positive real number, and in this embodiment m=2, it is used to control the degree of influence of distance on weight. The larger the value of m, the greater the influence of points that are closer to each other on the interpolation result.

[0081] The first sensor, with a smaller range, has higher resolution and accuracy, enabling precise measurement of minute changes within a low to medium water level range. The second sensor, with a larger range, can withstand measurements over a higher water level range and is suitable for monitoring environments where extreme water levels may occur. In this embodiment, two water level sensors with different ranges are set up at the monitoring point, which can meet both the need to capture minute changes in water level data and the need to monitor high water levels.

[0082] Example 3: Reference Figure 2 The difference between this embodiment and Embodiment 1 is that, when the risk level meets expectations, the method further includes:

[0083] S31 modeling and training acquires historical water level data and historical water level change rates at each monitoring point. In this embodiment, the historical water level change rate is obtained by calculating the water level change between two adjacent collection time points and then dividing it by the time interval.

[0084] All historical water level data were integrated into a historical water level data sequence according to the chronological order of collection time, and the historical water level change rate was integrated into a historical water level change rate sequence according to the chronological order of time.

[0085] A neural network model is constructed, which can use a recurrent neural network (RNN) or its variants, such as a long short-term memory network (LSTM) or a gated recurrent unit (GRU). The neural network model includes an input layer, hidden layers, and an output layer. The number of neurons in the input layer is the same as the dimension of the historical water level data sequence. The hidden layer consists of two LSTM units and one Dropout layer. The LSTM units address the gradient vanishing and gradient exploding problems in traditional RNNs by introducing gating mechanisms (input gate, forget gate, and output gate), enabling them to better capture long-term dependencies in time-series data. The first LSTM unit has 64 neurons, and the second LSTM unit has 32 neurons. The Dropout layer follows the two LSTM units. The output layer has one neuron and uses a linear activation function.

[0086] The neural network model was trained using historical water level data sequences and historical water level change rate sequences to obtain the trained neural network model.

[0087] S32 predicts water velocity by inputting real-time water level data into a trained neural network model and outputting the predicted rate of change. When the predicted rate of change exceeds a preset rate threshold, the pump discharge rate is increased.

[0088] S33 calculates the influence factor by obtaining various structural parameters at each monitoring point, including the cross-sectional dimensions of the pipe gallery, its slope, and its roughness coefficient. Based on the open channel flow theory and Bernoulli's equation in fluid mechanics, the influence factor of the structural parameters on the water level change rate is calculated. The influence factor is calculated through the following steps:

[0089] The hydraulic radius and Chezy coefficient of the water flow in the pipe gallery are calculated based on the open channel flow theory. The calculation model for the hydraulic radius is as follows:

[0090] ;

[0091] Where R is the hydraulic radius; a is the width of the water surface; and h is the water depth.

[0092] The calculation model for the Chezy coefficient C is as follows:

[0093] ;

[0094] Where e is the roughness coefficient, which is related to the roughness of the inner wall of the integrated utility tunnel and is determined through experiments or experience.

[0095] Based on the hydraulic radius and Chezy coefficient, and combined with Bernoulli's equation, the correlation function between the water level change rate and various structural parameters is derived.

[0096] Bernoulli's equation describes the principle of conservation of total mechanical energy in the steady flow of an ideal fluid. In open channel flow, Bernoulli's equation can be expressed as:

[0097] ;

[0098] in, , These are the water level elevations at the upstream and downstream sections of the integrated utility tunnel, respectively. , These are the pressures at the upstream and downstream sections of the integrated utility tunnel, respectively. The density of water; It is the acceleration due to gravity; , , respectively, represent the average flow velocities at the upstream and downstream sections of the integrated utility tunnel; A represents the head loss.

[0099] In open channel flow within a utility tunnel, assuming the flow is uniform or gradually varied, and the upstream and downstream pressures are equal, Bernoulli's equation can be simplified to:

[0100] ;

[0101] Head loss is calculated using the Darcy-Weisbach formula or the Manning formula. Combined with the Chezy coefficient, head loss can be expressed as:

[0102] ;

[0103] Where Q is the flow rate; e is the roughness coefficient; L is the flow path length; S is the cross-sectional area of ​​the water passage; and R is the hydraulic radius.

[0104] The relationship between flow rate Q, average flow velocity u, and cross-sectional area S is as follows:

[0105] ;

[0106] The average flow velocity u is calculated using the Chezy coefficient, and the calculation model is as follows:

[0107] ;

[0108] Where f is the slope within the integrated utility tunnel.

[0109] The rate of water level change is derived by differentiating and transforming Bernoulli's equation, yielding a correlation function between the rate of water level change and the interface dimensions, slope, and roughness coefficient of the pipe gallery. The calculation process of this correlation function is as follows:

[0110] ;

[0111] Substituting the above formula and taking partial derivatives with respect to the cross-sectional area S, hydraulic radius R, slope f, and roughness coefficient e, we obtain:

[0112] ;

[0113] in, This is the correlation function, and its specific form can be approximated by Taylor expansion or sensitivity analysis. This embodiment takes Taylor expansion as an example to illustrate its form.

[0114] ;

[0115] ;

[0116] Since the Chezy coefficient C is independent of the roughness coefficient a, and the cross-sectional area S of the water passage is positively correlated with the width a of the water surface, therefore,

[0117] ;

[0118] When the width 'a' of the water surface is much greater than the depth 'h' ,therefore,

[0119] ;

[0120] Subsequent calculations are similar and will not be repeated in this embodiment.

[0121] ;

[0122] ;

[0123] ;

[0124] in, , , and As a weighted system, it is determined by a normalized sensitivity index to reflect the contribution of various structural parameters to water level changes.

[0125] The weight coefficients corresponding to each structural parameter are calculated based on the correlation function. After standardization, these weight coefficients are used as the influence factors. The calculation process is as follows:

[0126] Within the range of variation of each structural parameter, select a set of benchmark values, and then change the value of one structural parameter at a time while keeping the other parameters as benchmark values, calculate the rate of water level change. The relative rate of change is standardized and used as the weighting coefficient for various structural parameters.

[0127] The product of the predicted rate of change and the influencing factor is used as the new predicted rate of change.

[0128] The S34 optimization threshold, the structural parameters also include pipeline corrosion rate, historical leakage frequency, and concrete carbonation depth. The aging index of the integrated utility tunnel is calculated based on the pipeline corrosion rate, historical leakage frequency, and concrete carbonation depth. The calculation model for the aging index is as follows:

[0129] ;

[0130] ;

[0131] ;

[0132] ;

[0133] in, The normalized pipeline corrosion rate; This represents the current pipeline corrosion rate. This represents the maximum permissible corrosion rate for the pipeline. The normalized historical leakage count; is the weight of the number of leaks in the j-th instance; M is the total number of leaks. The carbonation depth of the concrete structure after normalization. This represents the current carbonation depth of the concrete structure. This is the critical value for the carbonation depth of concrete structures.

[0134] The attenuation factor for the water level data threshold is set based on the aging index. The product of the water level data threshold and the attenuation factor is then used to update the water level data threshold. The calculation model for the attenuation factor is as follows:

[0135] ;

[0136] Where DF is the attenuation factor; This is an adjustment coefficient, with a value ranging from 0.2 to 0.5.

[0137] S35 optimizes the sampling frequency by calculating the sampling frequency adjustment coefficient k based on the predicted rate of change and leakage frequency at each monitoring point in the drainage zone. The calculation model is as follows:

[0138] ;

[0139] in, and These are the weighting coefficients; Let be the predicted rate of change at the i-th monitoring point; This represents the maximum permissible water flow rate at the i-th monitoring point; Let i be the number of leaks at the i-th monitoring point; This represents the total number of monitoring times at the i-th monitoring point;

[0140] The acquisition frequency of the water level sensor at each monitoring point is adjusted by an adjustment coefficient. That is, the product of the current acquisition frequency of the water level sensor and the adjustment coefficient k is used as the new acquisition frequency. The water level sensor acquires the real-time water level data at the corresponding monitoring point at the new acquisition frequency.

[0141] The S36 optimized zoning method divides the integrated utility tunnel into several initial zones based on the risk level and spatial location of each monitoring point using the K-means clustering algorithm. The process is as follows:

[0142] The risk level of each monitoring point is converted into a numerical index: high risk = 3, medium risk = 2, and low risk = 1. A three-dimensional feature vector [x, y, risk] is constructed for each monitoring point, where x and y are the normalized planar coordinates, and risk is the normalized numerical index.

[0143] The optimal number of clusters is determined by combining the Elbow Method with the Silhouette Score. The point with the highest risk level is selected as the first centroid. The K-means++ algorithm is used to optimize the initial centroid distribution until the centroid position no longer changes significantly or the maximum number of iterations is reached. Each monitoring point is then assigned a cluster label, forming K initial partitions.

[0144] Obtain the average water level change rate and connectivity parameter for each initial partition. If two initial partitions are geographically adjacent, their connectivity parameter is 1; otherwise, it is 0.

[0145] The hydraulic conductivity coefficients of adjacent initial partitions are calculated using the following model:

[0146] ;

[0147] in, Let r be the hydraulic conductivity coefficient between the initial partition r and the initial partition s; This is the connectivity parameter between the initial partition r and the initial partition s; This is an empirical coefficient, with a value ranging from 0.1 to 0.3; The average water level change rate of the initial partition r and the average water level change rate of the initial partition s The absolute difference.

[0148] When the hydraulic conductivity of adjacent initial zones exceeds a preset conductivity threshold, the adjacent initial zones are merged to form a new drainage zone. The pump operating parameters are then configured according to the new drainage zone. In this embodiment, the process of configuring the pump operating parameters according to the new drainage zone is as follows:

[0149] Based on the geographical location of the drainage zones and the layout of the water pumps, one or more water pumps are designated as the main pumps responsible for drainage for each final drainage zone. When the real-time water level data of a drainage zone exceeds the middle water level threshold, the operating power of the main pumps responsible for drainage is increased; when the real-time water level data of a drainage zone exceeds the high water level threshold, all pumps are turned on for drainage.

[0150] In other embodiments, the operating parameters of the water pump can be set according to specific needs.

[0151] Example 4: This example discloses a comprehensive utility tunnel water level monitoring and processing system, the system including: a memory and a processor.

[0152] The memory contains a computer-readable storage medium;

[0153] When the processor processes the computer program stored on the computer-readable storage medium, it implements the integrated utility tunnel water level monitoring and processing method.

[0154] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A utility tunnel water level monitoring processing method, characterized in that, The method comprises the steps of: Collecting real-time water level data of each monitoring point in the comprehensive pipe gallery through water level sensors in the distributed sensor network, constructing a water level spatial distribution map based on the real-time water level data and a pre-constructed three-dimensional model of the comprehensive pipe gallery, dividing the water level spatial distribution map into different risk levels using a water level data threshold, and dynamically dividing drainage zones according to the risk levels; When the risk level does not meet the expectation, the water pump corresponding to the drainage zone is started to drain water, and an alarm information is pushed to the operation and maintenance terminal through the Internet of Things platform; Each monitoring point is provided with a first sensor and a second sensor, the range of the first sensor is smaller than that of the second sensor, when the water level data monitored by the first sensor reaches a warning threshold, the second sensor is triggered to synchronously monitor the water level data, when the difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor at the same monitoring point exceeds a preset difference threshold, an alarm signal is sent, the monitoring point whose difference between the water level data monitored by the first sensor and the water level data monitored by the second sensor exceeds the preset difference threshold is recorded as an abnormal position, the water level data at the adjacent monitoring points of the abnormal position is obtained, the theoretical water level data at the abnormal position is calculated according to the water level data at the adjacent monitoring points by using an inverse distance weighted interpolation algorithm, and the water level distribution map is updated; The method further comprises the steps of: According to the risk level and the spatial position of each monitoring point, the comprehensive pipe gallery is divided into a plurality of initial zones by using a K-means clustering algorithm; The average water level change rate and the connectivity parameter of each initial zone are obtained, if two initial zones are adjacent in geography, the connectivity parameter of the two initial zones is 1, otherwise, the connectivity parameter is 0; The hydraulic conduction coefficient of adjacent initial zones is calculated, and the calculation model of the hydraulic conduction coefficient is as follows: ； wherein, is the hydraulic conductivity between the initial zone r and the initial zone s; is the connectivity parameter between the initial zone r and the initial zone s; is an empirical coefficient, having a value comprised between 0.1 and 0.3; is the average rate of variation of the water level of the initial zone r and the average rate of variation of the water level of the initial zone s of the absolute difference. When the hydraulic conduction coefficient of adjacent initial zones exceeds a preset conduction threshold, the adjacent initial zones are merged to form a new drainage zone, and the water pump operation parameters are configured according to the new drainage zone.

2. The utility tunnel water level monitoring process of claim 1, wherein, When the risk level meets the expectation, the method further comprises the steps of: The historical water level data and the historical water level change rate at each monitoring point are obtained, all the historical water level data are integrated into a historical water level data sequence according to the collection time sequence, the historical water level change rate is integrated into a historical water level change rate sequence, a neural network model is constructed, the historical water level data sequence and the historical water level change rate sequence are used to train the neural network model, and a trained neural network model is obtained; The real-time water level data is input into the trained neural network model, and a predicted change rate is output, when the predicted change rate exceeds a preset rate threshold, the water pump drainage rate is increased.

3. The utility tunnel water level monitoring process of claim 2, wherein, The method further comprises the steps of: Various structural parameters at each monitoring point are obtained, an influence factor of the structural parameters on the water level change rate is calculated based on the open channel flow theory and Bernoulli equation in fluid mechanics, and the product of the predicted change rate and the influence factor is taken as a new predicted change rate.

4. The utility tunnel water level monitoring process of claim 3, wherein, The structural parameters include the pipe gallery cross-sectional size, the slope and the roughness coefficient, and the influence factor is calculated through the following steps: The hydraulic radius and the Chezy coefficient of water flow in the comprehensive pipe gallery are calculated based on the open channel flow theory. Based on the hydraulic radius and Chezy coefficient, the correlation function of water level change rate and each structural parameter is derived by Bernoulli equation; According to the correlation function, the weight coefficient corresponding to each structural parameter is calculated, and the weight coefficient is standardized and taken as an influence factor after processing.

5. The utility tunnel water level monitoring process of claim 4, wherein, The structural parameters also include pipeline corrosion rate, historical leakage times and concrete structure carbonation depth, and the method further comprises: According to the pipeline corrosion rate, the historical leakage times and the concrete structure carbonation depth, the aging index of the comprehensive pipe gallery is calculated, the decay factor of the water level data threshold is set according to the aging index, and the product of the water level data threshold and the decay factor is updated as the water level data threshold.

6. The utility tunnel water level monitoring process of claim 3, wherein, The method further comprises: According to the predicted change rate and the leakage times of each monitoring point in the drainage partition, the sampling frequency adjustment coefficient k is calculated, and the calculation model is as follows: ； wherein, and are weight coefficients; is a predicted rate of change at the i-th monitoring point; is a maximum allowed water flow rate at the i-th monitoring point; is a number of leaks at the i-th monitoring point; is a total number of monitoring at the i-th monitoring point; The collection frequency of the water level sensor at each monitoring point is adjusted by using the adjustment coefficient k.

7. A utility tunnel water level monitoring processing system characterized by, It comprises: A memory and a processor, The memory has a computer readable storage medium stored therein; When the processor processes the computer program stored on the computer readable storage medium, the method as claimed in any one of claims 1-6 is realized.

Images