A system for deep mining and visual analysis of urban physical examination data

Abstract

The present invention belongs to the technical field of urban physical examination data analysis, and specifically discloses a system for deep mining and visual analysis of urban physical examination data, which includes a drainage data acquisition module, a drainage overload analysis module, an intersection hazard analysis module, and a hazard report generation module. The present invention effectively solves the core problems of staticity, one-sidedness, and hysteresis in the existing technology by setting a technical system of dynamic threshold correction + multi-dimensional risk modeling + drainage intersection focus analysis, and realizes a leap-forward upgrade from extensive pipe network design to refined node prevention and control, providing a quantifiable, traceable, and scalable risk management tool for urban drainage systems. At the same time, by integrating dynamic parameters such as the real-time drainage volume of drainage branches, overload index, and centralized drainage time, a drainage intersection hazard index is innovatively constructed to achieve deep risk mining focusing on the key node of the drainage intersection.

Description

Technical Field

[0001] The present invention belongs to the technical field of urban physical examination data analysis, and in particular relates to a system for deep mining and visual analysis of urban physical examination data. Background Art

[0002] With the acceleration of urbanization, the urban drainage system, as an important part of infrastructure, its operating efficiency and safety are directly related to urban flood control and drainage, water environment quality and residents' life safety, which highlights the importance of urban drainage physical examination.

[0003] Existing technologies, such as the Chinese invention patent application with application number 202411759930.5, disclose an urban drainage network optimization design system. This system integrates multiple parameters, including rainfall intensity, ground permeability, sewage concentration, and pipe cleanliness, to construct a dynamic parameter correction model. This solves the problem of traditional optimization design, which relies on single rainfall data and results in incomplete parameters that fail to reflect the actual operating status of the pipe network. It also improves drainage efficiency and the accuracy of cost-benefit analysis.

[0004] Obviously, existing technologies in drainage optimization have three core problems: staticness, one-sidedness, and hysteresis, which make it difficult to cope with complex and changeable urban drainage scenarios. At the same time, existing technologies mainly focus on the relevant influencing indicators of the pipeline and do not focus on the drainage intersection, resulting in insufficient excavation depth, and thus unable to further improve the targeted nature of subsequent drainage optimization. Summary of the Invention

[0005] In view of this, based on the above problems, a deep mining and visualization analysis system for urban physical examination data is proposed.

[0006] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides a deep mining and visual analysis system for urban physical examination data, which includes: a drainage data acquisition module, which collects the designed drainage load of the drainage intersection point in the drainage network of the area to be examined and the location of the drainage intersection point, and at the same time collects the real-time drainage volume of the drainage intersection point and each drainage branch corresponding to the drainage intersection point.

[0007] The drainage overload analysis module generates a drainage overload index for each drainage branch based on the real-time drainage volume of each drainage branch and a preset load threshold, and determines the concentrated drainage time of each drainage branch through time series analysis.

[0008] The intersection hazard analysis module generates the drainage hazard index of the drainage intersection based on the drainage overload index, current drainage volume and concentrated drainage time of each drainage branch. Based on the historical drainage monitoring data of the drainage intersection, the drainage hazard index of each historical drainage monitoring is similarly calculated to set the dynamic drainage hazard index threshold.

[0009] The danger report generation module, if the current drainage danger index exceeds the dynamic danger index threshold, matches the preset drainage danger mapping table to obtain the current drainage danger level, and generates a drainage intersection point physical examination report based on the drainage intersection point location and drainage danger level.

[0010] Compared with the existing technology, the beneficial effects of the present invention are as follows: (1) The present invention effectively solves the core problems of staticity, one-sidedness and hysteresis in the existing technology by setting a multi-dimensional evaluation system of dynamic threshold correction, multi-dimensional risk modeling and drainage intersection focus analysis, and realizes a leap-forward upgrade from extensive pipe network design to refined node prevention and control, providing a quantifiable, traceable and scalable risk management tool for urban drainage systems, which can effectively reduce the economic losses caused by urban waterlogging and extend the service life of the pipe network.

[0011] (2) This invention integrates multi-dimensional parameters such as drainage overload index, concentrated drainage time and drainage hazard index to comprehensively depict the risk transmission path in complex drainage scenarios, solving the one-sided problem caused by single parameter analysis, providing a more scientific decision-making basis for system scheduling, and facilitating the response to complex and changing urban drainage scenarios.

[0012] (3) The present invention innovatively constructs a drainage intersection risk index by integrating dynamic parameters such as the real-time drainage volume of drainage branches, overload index, and concentrated drainage time, thereby achieving in-depth risk mining focusing on the key node of drainage intersection, ensuring the dynamic, comprehensive and real-time risk assessment of the drainage system, providing technical support for precise optimization in complex urban drainage scenarios, and filling the core shortcoming of the current inability to adapt to the dynamic carrying capacity of the pipeline network.

[0013] (4) The present invention can effectively avoid the assessment deviation caused by parameter solidification by setting a dynamic drainage risk index threshold, enabling the system to respond in real time to dynamic factors such as changes in pipe network load, seasonal characteristics or equipment aging, and significantly enhance the environmental adaptability of drainage risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0015] Figure 1 This is a schematic diagram of the system module structure connection of the present invention.

[0016] Figure 2 It is a flow chart of the overall implementation steps of the present invention.

[0017] Figure 3This is a flow chart of the centralized drainage time confirmation steps of the present invention. DETAILED DESCRIPTION

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0019] See also Figure 1 and Figure 2 As shown, the present invention provides a system for deep mining and visual analysis of urban physical examination data, which includes: a drainage data acquisition module, a drainage overload analysis module, an intersection risk analysis module and a risk report generation module.

[0020] In the above, the drainage overload analysis module is connected to the drainage data acquisition module and the intersection hazard analysis module respectively, and the intersection hazard analysis module is connected to the hazard report generation module.

[0021] The drainage data collection module collects the designed drainage load of the drainage intersection and the location of the drainage intersection in the drainage network of the area to be examined, and at the same time collects the real-time drainage volume of the drainage intersection and each drainage branch corresponding to the drainage intersection.

[0022] It should be noted that drainage data collection requires a sensor array consisting of electromagnetic flowmeters and liquid level gauges, which collect real-time flow and water levels for pipes ranging from DN50 to DN2000 via RS485 / 4G communication protocols. An integrated GIS positioning unit captures the longitude and latitude coordinates of drainage intersections and spatially correlated data for network nodes.

[0023] For example, the installation location of the sensor can be referenced as follows: set a monitoring section at the straight pipe section of each branch 1-2 times the pipe diameter upstream of the intersection, avoid flow disturbance areas such as elbows and valves, uniformly install high-precision ultrasonic flow meters, and install high-precision ultrasonic flow meters at the water inlet and outlet positions of the drainage intersection respectively.

[0024] The drainage overload analysis module generates a drainage overload index for each drainage branch based on the real-time drainage volume of each drainage branch and a preset load threshold, and determines the concentrated drainage time of each drainage branch through time series analysis.

[0025] Specifically, the specific analysis process of the drainage overload index is as follows: A1. Calculate the difference between the real-time drainage and the design drainage load to obtain the real-time drainage difference, compare the drainage difference with the design drainage load to obtain the relative proportion of the real-time drainage exceeding the design load, recorded as the drainage full load ratio, calculate the average drainage difference through mean calculation, and extract the maximum drainage full load ratio.

[0026] A2. The real-time drainage volume is organized into a drainage difference time series. The continuous duration of the drainage difference exceeding the preset threshold is tracked by setting a time window, and recorded as the duration of full drainage load.

[0027] A3. Calculate the standard deviation of the drainage difference time series to obtain the volatility of the drainage difference, and calculate the average volatility by means of mean.

[0028] A4. Overload compensation triggering is determined based on the average drainage difference, duration of full drainage, and average fluctuation rate. If overload compensation is triggered, an overload compensation assessment factor is set.

[0029] A5. After normalizing the average drainage difference, maximum drainage full load ratio, drainage full load duration and average fluctuation rate, a preliminary drainage overload index is obtained through linear weighted calculation, and the preliminary drainage overload index is compensated by the overload compensation evaluation factor to obtain the final drainage overload index of each drainage branch.

[0030] It should be added that the normalization described in the present invention can adopt the minimum-maximum normalization method, which will not be explained later.

[0031] Furthermore, the overload compensation triggering judgment condition in step A4 is: if any two of the average drainage difference, the duration of full drainage load and the average fluctuation rate exceed the corresponding set thresholds, it is determined that overload compensation is triggered.

[0032] Furthermore, the specific setting of the overload compensation evaluation factor in step A4 is as follows: B1. Parameter items that exceed the corresponding set thresholds are selected from the average drainage difference, the duration of full drainage load, and the average fluctuation rate, recorded as deviation parameter items, and form a deviation parameter item combination.

[0033] B2. Match the deviation parameter items to a preset parameter deviation basic compensation mapping table to obtain a reference overload compensation factor.

[0034] B3. Normalize the deviation parameter item, and input the weighted sum of the processing results as a variable into the Sigmoid function to output the target benchmark overload compensation factor.

[0035] B4. The product of the baseline overload compensation factor and the baseline overload compensation factor is used as the final overload compensation assessment factor.

[0036] For example, the threshold corresponding to the drainage difference is set to a range of ±5% to ±20% of the set drainage load, and the specific value can be set comprehensively in combination with the pipeline health assessment report.

[0037] Understandably, overload risk in drainage system operation is influenced by the coupling of multiple parameters, including drainage differentials, duration, and volatility. A single parameter or simple weighting cannot accurately reflect the risk differences between complex deviation scenarios, such as short-term, highly volatile overloads and long-term, mild overloads. The three-tiered architecture of "deviation screening - benchmark mapping - dynamic correction" combines qualitative and quantitative analysis, quantifying risks across multiple dimensions, and significantly improving system robustness.

[0038] It needs to be explained that the main purpose of setting the overload compensation assessment factor is to accurately identify complex overload risk scenarios in the operation of the drainage system through the systematic integration and dynamic quantification of multi-dimensional parameters, effectively solve the one-sidedness of single parameter evaluation, and provide accurate quantitative basis for drainage scheduling strategy optimization and equipment operation and maintenance plan formulation, significantly improving the system's adaptive ability in the face of complex load fluctuations.

[0039] Exemplarily, different basic compensation coefficients are assigned according to the number and type of deviation parameters, such as the deviation combination of two parameters vs. the deviation of the full parameter combination, to reflect the complexity and impact of the deviation. For example, when there are three deviation parameter items, the benchmark overload compensation factor is set to 1; when there are two deviation parameter items and the average drainage difference and the duration of drainage full load exceed the corresponding set thresholds, the benchmark overload compensation factor is set to 0.8; when there are two deviation parameter items and the average drainage difference and the average volatility exceed the corresponding set thresholds, the benchmark overload compensation factor is set to 0.6; when there are two deviation parameter items and the duration of drainage full load and the average volatility exceed the corresponding set thresholds, the benchmark compensation factor is set to 0.5.

[0040] Understandably, the differences in the baseline overload compensation factor values ​​for different deviation parameter combinations are primarily based on the contribution of each parameter to overload risk and the complexity of the combined risk. For example, average drainage difference, as a direct overload criterion, is the core parameter determining the compensation factor level. Combinations in which these factors are present, such as "drainage difference and duration" and "drainage difference and volatility," directly correlate with the intensity of immediate overloads, have significantly higher compensation factors than combinations without drainage difference, such as "duration and volatility." The three-parameter deviations simultaneously encompass overload intensity, persistence, and volatility, creating a multi-dimensional risk stack, and are therefore assigned the highest compensation factor. The "drainage difference and duration" combination, reflecting the cumulative risk of persistent overloads, has a higher compensation factor than the "drainage difference and volatility" combination, which only reflects intermittent shocks. Combinations without drainage difference, however, have the lowest compensation factors because they do not exceed immediate drainage capacity and are primarily latent risks. This design achieves a differentiated response to complex overload scenarios by highlighting core risk indicators and quantifying the effects of risk stacking.

[0041] It should be added that the benchmark compensation factor is based on a preset parameter deviation basic compensation mapping table, which converts industry standards, historical failure cases and engineering practice experience into quantifiable basic coefficients to ensure standardized responses to known typical deviation scenarios such as slight over-limit of a single parameter and risk of multi-parameter combination, and avoids assessment confusion caused by over-reliance on real-time data, such as when the three full deviations trigger the highest compensation. The target compensation factor captures the subtle fluctuation characteristics of the current operating data in real time through the normalization of the deviation parameters and the Sigmoid function mapping, to compensate for the dynamic change scenarios that the preset rules cannot cover. Combining the two, efficient response to simple scenarios and fine distinction between complex scenarios can be achieved.

[0042] More specifically, see Figure 3 As shown, the specific confirmation time of the centralized drainage time includes: D1, importing the historical average daily drainage of each drainage branch, dividing the real-time drainage according to the preset time period, and constructing a drainage time series data set including the average flow, peak flow and flow change rate of each time period.

[0043] D2. Use the local maximum search algorithm to detect the drainage time series data set of each drainage branch, and count the candidate peak periods that meet the peak flow threshold and whose continuous duration exceeds the minimum duration threshold.

[0044] D3. Cluster the candidate peak periods using the OPTICS density clustering algorithm, merge adjacent periods with overlapping time or intervals less than a first preset time length, and generate concentrated drainage period clusters.

[0045] The OPTICS density clustering algorithm is an existing algorithm, and the specific clustering process will not be described in detail.

[0046] D4. Extract the start time, end time, and peak drainage ratio of each time period cluster. The peak drainage ratio is the ratio of the total drainage in the cluster to the historical daily average drainage.

[0047] D5. Assign priority weights to each time period cluster based on the peak discharge volume ratio, select the time period clusters with weights greater than the preset weight threshold, and calculate the arithmetic mean of their start and end times as the time plot point.

[0048] For example, if a cluster contains three candidate time periods with start times of 7:15, 7:30, and 7:45, the average start time is 7:30, and the average end time is obtained similarly, which reflects the typical start and end patterns of centralized drainage.

[0049] D6. Calculate the standard deviation of the start and end times of each candidate peak time period within each focus time period cluster. If the standard deviation of the start and end times within the focus time period cluster is less than the second preset threshold, then expand the core time period with the time plot point as the center and the standard deviation as the boundary. If the standard deviation exceeds the third preset threshold, then take the minimum start time and the maximum end time within the cluster as the core time period boundary.

[0050] D7. The core time periods whose intervals are less than the second preset time length and whose total peak drainage volume ratio after merging is greater than the third reference threshold are merged twice to obtain the final concentrated drainage time.

[0051] In a specific embodiment, the first preset duration may be set to 30 minutes, and the second preset threshold and the third preset threshold corresponding to the standard deviation may be set to 30% and 60% respectively.

[0052] For example, if the core time period for Cluster 1 is 8:00-9:00 and the core time period for Cluster 2 is 9:15-10:00, and the interval is 15 minutes, the time interval can be merged into 8:00-10:00, as it is less than the preset threshold of 30 minutes. This ensures that the main drainage load is covered within the minimum time range, providing a precise time benchmark for pipe network scheduling.

[0053] It should be explained that the operation and management of the drainage system is similar to the flow control of the traffic system. It is necessary to accurately grasp the time and intensity of the "flow peak" to avoid system "overload" and paralysis. In traffic management, if the traffic density and duration of the morning and evening peaks are not taken into account, it is very easy to cause road congestion and even traffic accidents. Similarly, if the drainage system ignores the time characteristics of the centralized drainage of each branch, it may cause sewage overflow, waterlogging and other risks due to the instantaneous flow exceeding the carrying capacity of the pipeline network. Different drainage branches are affected by the function of the service area, drainage type and weather conditions, and the time and intensity of their drainage peaks vary significantly.

[0054] Apparently, accurately identifying the concentrated drainage time periods for each drainage branch can significantly improve the operational efficiency and safety of the drainage system. Furthermore, it can assist in pipeline network planning and expansion decisions. By analyzing long-term concentrated drainage data, weak links in the pipeline network can be identified, providing data support for new construction or renovation projects, and shifting the drainage system from reactive response to proactive management.

[0055] Furthermore, the process of determining the candidate peak period in step D2 is as follows: setting a double threshold for peak determination, wherein the double threshold is the time when the flow value exceeds the daily average flow. times and the duration exceeds the preset second duration.

[0056] The real-time discharge volume is organized into a discharge volume sequence, and the discharge volume sequence is scanned point by point. When it is detected that the flow exceeds the double threshold for the first time at a certain moment and is in a non-peak state, it is marked as the peak starting point.

[0057] Starting from the peak starting point, continuously verify whether the subsequent traffic continues to meet the dual threshold conditions. If so, mark the time from the peak starting point to the time when the traffic first fails to meet the threshold conditions as a candidate peak period.

[0058] It should be added that the preset second duration can be set to 30 minutes. The specific value can be 1.5.

[0059] The intersection hazard analysis module generates a drainage hazard index for a drainage intersection based on the drainage overload index, current drainage volume, and concentrated drainage time of each drainage branch, and based on the historical drainage monitoring data of the drainage intersection, similarly calculates the drainage hazard index at each historical drainage monitoring time and sets a dynamic drainage hazard index threshold.

[0060] Specifically, the specific generation process of the drainage hazard index of the drainage intersection includes: E1. Based on the concentrated drainage time period of each drainage branch, the drainage concentration of the drainage intersection is calculated. If the drainage concentration is less than the set threshold, the ratio of the current drainage volume of each drainage branch to the design drainage load of the drainage intersection is recorded as the current drainage load ratio, and used as the drainage hazard index.

[0061] E2. Otherwise, extract the measured water volume of the water inlet and outlet from the real-time drainage volume of the drainage intersection and calculate the drainage anomaly coefficient.

[0062] E3. Drainage branches whose drainage overload index is greater than the set warning overload index are recorded as focus branches, and the ratio of the focus branches to the total number of drainage branches is calculated and recorded as the warning coverage ratio.

[0063] E4. Comprehensively analyze the drainage anomaly coefficient and early warning coverage ratio to obtain the drainage hazard compensation factor. Use the drainage hazard compensation factor to correct the current drainage load ratio to obtain the final drainage hazard index.

[0064] The embodiment of the present invention innovatively constructs a drainage intersection hazard index by integrating dynamic parameters such as the real-time drainage volume of drainage branches, overload index, and concentrated drainage time, thereby achieving in-depth risk mining focusing on the key node of the drainage intersection, ensuring the dynamic, comprehensive and real-time risk assessment of the drainage system, providing technical support for precise optimization in complex urban drainage scenarios, and filling the core shortcoming of the current inability to adapt to the dynamic carrying capacity of the pipeline network.

[0065] Understandably, when drainage concentration is low, drainage from each branch is dispersed, and the load at the intersection is mainly determined by the total flow. Therefore, the current drainage load ratio is directly used as the hazard index to ensure a simple and efficient assessment. When drainage concentration is high, concentrated drainage from multiple branches can easily lead to coupling risks such as flow superposition and pressure imbalance. Therefore, the drainage anomaly coefficient and early warning coverage ratio are introduced, and the load ratio is corrected by the hazard compensation factor to fully capture systemic risks. Its significance lies in achieving accurate identification of the dangerous state of drainage intersections through differentiated assessments based on different scenarios, taking into account the computational efficiency of simple scenarios and the comprehensiveness of risks in complex scenarios. Among them, the drainage anomaly coefficient reflects the dynamic balance of water inlet and outlet, and the early warning coverage ratio reflects the prevalence of overloaded branches.

[0066] It should be added that, by combining the multi-level calculation of drainage anomaly coefficient, early warning coverage ratio and concentration, a complete chain of "real-time data collection - multi-dimensional risk assessment - dynamic threshold correction - precise early warning response" is formed, which can effectively respond to real-time changes in complex drainage scenarios.

[0067] Furthermore, the statistical process of drainage concentration includes: mapping the concentrated drainage periods of all drainage branches to a unified time axis to generate a binary matrix .

[0068] Based on the number of drainage branches connected to the drainage intersection, the current cumulative drainage time period is divided into monitoring time windows.

[0069] For each monitoring time window, the number of drainage branches within the monitoring time window is counted and recorded as the number of concentrated branches. , the number of concentrated branches is the matrix The sum of the elements in each row, Indicates the monitoring time window number, .

[0070] Based on the centralized drainage time period of each drainage branch, the sum of the drainage time of all drainage branches is calculated and recorded as .

[0071] Calculating drainage concentration , , is the total number of drainage branches.

[0072] For example, the intersection connects three branches, the current cumulative drainage time period is 8:00-11:00, and the time windows are divided into 8:00-9:00, 9:00-10:00, and 10:00-11:00.

[0073] Count the number of branches that are simultaneously in the centralized drainage state. For example, if the centralized drainage period for branch A is 08:00-09:00 and for branch B is 08:30-09:30, then the number of overlaps within the 08:30-09:00 window is 2.

[0074] Understandably, It represents the maximum proportion of centralized drainage branches within a single window, reflecting the intensity of centralized drainage. To calculate the ratio of the total concentrated drainage time of all branches to the total number of minutes in a day. For example, if the total concentrated drainage time of all branches is 800 minutes, then , indicating that the total duration of centralized drainage accounts for 56% of the day, reflecting the continuity of centralized drainage.

[0075] Multiplying the two results in drainage concentration , which combines the intensity and continuity of centralized drainage. When the intensity is high and the continuity is strong, This standardized calculation enables drainage concentration to comprehensively and objectively reflect the concentration characteristics of the drainage system in the time dimension, providing a quantitative basis for subsequent drainage risk assessment.

[0076] It is also understandable that by calculating the drainage concentration in two dimensions, namely the proportion of overlapping branches in the time window and the proportion of total concentrated time, the spatiotemporal coupling effect of concentrated drainage of multiple branches can be accurately captured, which significantly improves the accuracy of overload analysis.

[0077] Furthermore, the specific statistical process of calculating the drainage anomaly coefficient includes: subtracting the measured water volume at the water inlet and the water outlet to obtain the inlet and outlet water volume difference.

[0078] The real-time inlet and outlet water volume difference is calculated by standard deviation to obtain the inlet and outlet water volume difference volatility , and calculate the rate of change of the difference between inlet and outlet water volume .

[0079] Statistical drainage anomaly coefficient , , is a natural constant, and are the pre-set allowable values ​​of the fluctuation rate and change rate of the difference between the inlet and outlet water volume under normal conditions, and are weights for setting the fluctuation rate of the difference between inlet and outlet water volume and the rate of change of the difference between inlet and outlet water volume respectively.

[0080] Understandably, the exponential function used in the formula more accurately reflects the nonlinear relationship between the degree of drainage anomaly and its fluctuation and rate of change. In actual drainage systems, slight deviations from the normal range have little impact on the anomaly coefficient, while as the degree of deviation increases, the anomaly increases at a faster rate. The characteristics of the exponential function make the formula change smoothly when the deviation is small and respond sensitively when the deviation is large, which conforms to this nonlinear law. This avoids the problem that a linear weighted sum function alone cannot reflect this nonlinear characteristic: "the larger the deviation, the more significant the anomaly."

[0081] In the formula, Operation ensures that only or The corresponding factors are correct This generates a positive contribution. It highlights abnormalities that exceed normal reference values, avoiding interference with the abnormal coefficient caused by minor fluctuations, such as those that do not exceed the reference value. A simple linear summation would include all fluctuations, including minor changes within the normal range, in the calculation, reducing the accuracy of the abnormality assessment.

[0082] Understandably, the pre-set allowable values ​​for the fluctuation and rate of change of the inflow and outflow flow difference under normal conditions are primarily derived from statistical analysis of long-term historical drainage system operating data, from which typical parameter ranges under normal operating conditions are extracted. Reference is made to the definition of system operating parameters in relevant drainage industry standards and specifications.

[0083] In another specific embodiment, the specific setting process of the dynamic drainage risk index threshold includes: counting the maximum number of consecutive monitoring times in which the drainage risk index exceeds the preset baseline threshold in historical monitoring, and dividing it by the total number of drainage monitoring times to obtain the continuous drainage monitoring times ratio.

[0084] Arrange the drainage hazard index in ascending order according to the monitoring time, use the median order as the dividing point, and calculate the proportion of over-limit monitoring times and the average drainage hazard index after the dividing point.

[0085] The correction factor is obtained by comprehensively analyzing the ratio of continuous drainage monitoring times, the ratio of over-limit monitoring times and the average drainage hazard index.

[0086] The pre-set baseline drainage hazard index threshold is corrected by the correction factor to obtain the corrected drainage hazard index threshold.

[0087] The embodiment of the present invention can effectively avoid assessment deviations caused by parameter solidification by setting a dynamic drainage risk index threshold, enabling the system to respond in real time to dynamic factors such as changes in pipe network load, seasonal characteristics, or equipment aging, thereby significantly enhancing the environmental adaptability of drainage risk assessment.

[0088] It can be understood that the proportion of over-limit monitoring times after the split point is the ratio of the total number of drainage monitoring times after the split order where the drainage risk index is greater than the corresponding set threshold to the total number of drainage monitoring times after the split order.

[0089] It's important to note that the purpose of setting split points is to divide historical data into a baseline segment and a recent focus segment. By focusing on data after the split point, the latest evolutionary trends in the pipeline network's status can be effectively captured. Choosing a median order rather than a fixed time window as the split point avoids the subjectivity of manual divisions while dynamically adapting to changes in data volume, ensuring a balanced data volume between the baseline and focus segments. This reduces interference from fixed issues or outdated operating conditions in early historical data, enhances sensitivity to recent risk signals, such as sudden overloads, and enables dynamic threshold adjustments to better align with the current system's actual carrying capacity.

[0090] It should also be noted that the ratio of continuous drainage monitoring times quantifies the cumulative effect of long-term overloads, revealing the degradation trend of the pipeline network under sustained high loads. The proportion of overload monitoring times focuses on the frequency of recent risk evolution, capturing the risks of sudden environmental changes or accelerated pipeline aging. The average drainage hazard index after the split point measures the hazard intensity of recent overload events, avoiding misjudgment of high-frequency low-risk or low-frequency high-risk events. This achieves full-scenario coverage and precise classification of risk perception.

[0091] This embodiment of the present invention integrates multiple parameters, including the drainage overload index, concentrated drainage time, and drainage risk index, to comprehensively depict the risk transmission path in complex drainage scenarios. This addresses the one-sidedness of single-parameter analysis, provides a more scientific decision-making basis for system scheduling, and facilitates responses to complex and ever-changing urban drainage scenarios.

[0092] The danger report generation module matches the preset drainage danger mapping table to obtain the current drainage danger level if the current drainage danger index exceeds the dynamic danger index threshold, and generates a drainage intersection point physical examination report based on the drainage intersection point location and the drainage danger level.

[0093] The embodiment of the present invention effectively solves the core problems of staticity, one-sidedness and hysteresis in the existing technology by setting up a technical system of dynamic threshold correction + multi-dimensional risk modeling + drainage intersection focus analysis, and realizes a leap-forward upgrade from extensive pipe network design to refined node prevention and control, providing a quantifiable, traceable and scalable risk management tool for urban drainage systems, which can effectively reduce the economic losses of urban waterlogging and extend the service life of the pipe network.

[0094] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the scope of protection of the present invention.

Claims

1. A city physical examination data deep mining and visual analysis system, characterized by: The system includes: The drainage data collection module collects the designed drainage capacity and location of the drainage intersection in the drainage network of the area to be examined, and also collects the real-time drainage volume of the drainage intersection and the drainage branches corresponding to the drainage intersection; a drainage overload analysis module, which generates a drainage overload index for each drainage branch based on the real-time drainage volume of each drainage branch and a preset load threshold, and determines the concentrated drainage time of each drainage branch through time series analysis; The intersection hazard analysis module generates a drainage hazard index for the drainage intersection based on the drainage overload index, current drainage volume, and concentrated drainage time of each drainage branch. Based on the historical drainage monitoring data of the drainage intersection, the drainage hazard index at each historical drainage monitoring time is calculated and a dynamic drainage hazard index threshold is set. The hazard report generation module, if the current drainage hazard index exceeds the dynamic hazard index threshold, matches the preset drainage hazard mapping table to obtain the current drainage hazard level, and generates a drainage intersection point physical examination report based on the drainage intersection point location and drainage hazard level; The specific analysis process of the drainage overload index is as follows: The difference between the real-time drainage and the design drainage load is calculated to obtain the real-time drainage difference. The drainage difference is compared with the design drainage load to obtain the relative proportion of the real-time drainage exceeding the design load, which is recorded as the drainage full load ratio. The average drainage difference is calculated by averaging, and the maximum drainage full load ratio is extracted. The real-time drainage volume is combined into a drainage difference time series. The continuous duration of the drainage difference exceeding the preset threshold is tracked by setting a time window, which is recorded as the duration of full drainage load. The standard deviation of the drainage difference time series is calculated to obtain the volatility of the drainage difference, and the average volatility is obtained by mean calculation; Overload compensation triggering is determined based on the average drainage difference, duration of full-load drainage, and average fluctuation rate. If overload compensation is triggered, an overload compensation assessment factor is set. The average drainage difference, maximum drainage full load ratio, drainage full load duration and average fluctuation rate are normalized and then a preliminary drainage overload index is obtained by linear weighted calculation. The preliminary drainage overload index is compensated by the overload compensation evaluation factor to obtain the final drainage overload index of each drainage branch.

2. The city health examination data deep mining and visual analysis system according to claim 1 is characterized by: The overload compensation triggering judgment condition is: if any two of the average drainage difference, the duration of full drainage and the average fluctuation rate exceed the corresponding set thresholds, it is determined that overload compensation is triggered.

3. The city health examination data deep mining and visual analysis system according to claim 1 is characterized by: The specific setting of the overload compensation assessment factor is as follows: Parameter items exceeding the corresponding set thresholds are selected from the average drainage difference, the duration of full drainage load, and the average fluctuation rate, recorded as deviation parameter items, and formed into a deviation parameter item combination; Matching the deviation parameter item combination to a preset parameter deviation basic compensation mapping table to obtain a reference overload compensation factor; Normalizing the deviation parameter item, and inputting the weighted sum of the processing results as a variable into a Sigmoid function to output a target benchmark overload compensation factor; The product of the baseline overload compensation factor and the baseline overload compensation factor is taken as the final overload compensation assessment factor.

4. The city health examination data deep mining and visual analysis system according to claim 1 is characterized by: The specific confirmation time of the centralized drainage time includes: Import the historical average daily discharge of each drainage branch, divide the real-time discharge into preset time periods, and construct a time series data set of discharge that includes the average flow, peak flow, and flow change rate for each period; The drainage time series data set of each drainage branch is detected by the local maximum search algorithm, and the candidate peak periods that meet the peak flow threshold and whose continuous duration exceeds the minimum duration threshold are counted; Clustering candidate peak periods using the OPTICS density clustering algorithm, merging adjacent periods with overlapping time or intervals less than a first preset time length, and generating concentrated drainage period clusters; Extract the start time, end time and peak drainage ratio of each time period cluster. The peak drainage ratio is the ratio of the total drainage in the cluster to the historical daily average drainage; Assign priority weights to each time period cluster based on the peak discharge ratio, select the time period clusters with weights greater than the preset weight threshold, and calculate the arithmetic mean of their start and end times as the time plot point; Calculate the standard deviation of the start and end times of each candidate peak period within each focus period cluster. If the standard deviation of the start and end times within the focus period cluster is less than the second preset threshold, expand the core period with the time plot point as the center and the standard deviation as the boundary. If the standard deviation exceeds the third preset threshold, take the minimum start time and the maximum end time within the cluster as the core period boundary. The core time periods whose intervals are less than the second preset time length and whose total peak drainage volume ratio after merging is greater than the third reference threshold are merged twice to obtain the final concentrated drainage time.

5. The city health examination data deep mining and visual analysis system according to claim 4 is characterized by: The process of determining the candidate peak period is as follows: Set a double threshold for peak determination, the double threshold is the flow value exceeds the daily average flow times and the duration exceeds the preset second duration; The real-time discharge volume is organized into a discharge volume sequence, and the discharge volume sequence is scanned point by point. When it is detected that the flow rate at a certain moment exceeds the double threshold for the first time and is in a non-peak state, it is marked as a peak starting point; Starting from the peak starting point, continuously verify whether the subsequent traffic continues to meet the dual threshold conditions. If so, mark the time from the peak starting point to the time when the traffic first fails to meet the threshold conditions as a candidate peak period.

6. The city physical examination data deep mining and visual analysis system according to claim 1 is characterized by: The specific process of generating the drainage hazard index of the drainage intersection point includes: Based on the concentrated drainage time period of each drainage branch, the drainage concentration of the drainage intersection is calculated. If the drainage concentration is less than the set threshold, the ratio of the current drainage volume of each drainage branch to the design drainage load of the drainage intersection is recorded as the current drainage load ratio and used as the drainage hazard index. Otherwise, the measured water volume of the water inlet and outlet is extracted from the real-time drainage volume of the drainage intersection and the drainage anomaly coefficient is calculated; Drainage branches whose drainage overload index is greater than the set warning overload index are recorded as concerned branches, and the ratio of concerned branches to the total number of drainage branches is calculated and recorded as the warning coverage ratio; The drainage hazard compensation factor is obtained by comprehensively analyzing the drainage anomaly coefficient and the early warning coverage ratio. The final drainage hazard index is obtained by correcting the current drainage load ratio using the drainage hazard compensation factor.

7. The city health examination data deep mining and visual analysis system according to claim 6 is characterized by: The statistical process of drainage concentration includes: Map the centralized drainage periods of all drainage branches to a unified time axis to generate a binary matrix ; Based on the number of drainage branches connected to the drainage intersection, the current cumulative drainage time period is divided into monitoring time windows; For each monitoring time window, the number of drainage branches within the monitoring time window is counted and recorded as the number of concentrated branches. , the number of concentrated branches is the matrix The sum of the elements in each row, Indicates the monitoring time window number, ; Based on the centralized drainage time period of each drainage branch, the sum of the drainage time of all drainage branches is calculated and recorded as ; Calculating drainage concentration , , is the total number of drainage branches.

8. The city health examination data deep mining and visual analysis system according to claim 6 is characterized by: The specific statistical process of statistical drainage anomaly coefficient includes: The difference between the measured water volume at the water inlet and the water outlet is obtained by subtracting the measured water volume at the water inlet and the water outlet; The real-time inlet and outlet water volume difference is calculated by standard deviation to obtain the inlet and outlet water volume difference volatility , and calculate the rate of change of the difference between inlet and outlet water volume ; Statistical drainage anomaly coefficient , , is a natural constant, and are the pre-set allowable values ​​of the fluctuation rate and change rate of the difference between the inlet and outlet water volume under normal conditions, and are weights for setting the fluctuation rate of the difference between inlet and outlet water volume and the rate of change of the difference between inlet and outlet water volume respectively.

9. The city physical examination data deep mining and visual analysis system according to claim 1 is characterized by: The specific setting process of the dynamic drainage hazard index threshold includes: The maximum number of consecutive monitoring times in which the drainage hazard index exceeded the preset baseline threshold in historical monitoring is calculated, and the ratio of the consecutive drainage monitoring times is obtained by adding the total number of drainage monitoring times. Arrange the drainage hazard index in ascending order according to the monitoring time, use the median order as the dividing point, and calculate the proportion of over-limit monitoring times and the average drainage hazard index after the dividing point; The correction factor is obtained by comprehensively analyzing the ratio of continuous drainage monitoring times, the ratio of over-limit monitoring times and the average drainage hazard index; The pre-set baseline drainage hazard index threshold is corrected by the correction factor to obtain the corrected drainage hazard index threshold.

Images