City pipe network critical pressure prediction method based on data modeling

Abstract

The invention discloses an urban pipe network critical pressure prediction method based on data modeling, and belongs to the technical field of urban management, and the method specifically comprises the steps: building a mapping model of a pipe network water flow state and a pressure conduction time sequence, and predicting the expected time when pressure fluctuation is conducted to a designated pipe network node; before the pressure fluctuation does not reach the target node, the standard data acquisition frequency of the target node is maintained; when the predicted conduction remaining time is smaller than a set threshold value, activating an enhanced acquisition mode of the target node; according to the weight grade of the pipe section where the target node is located and the amplitude grade of the predicted pressure fluctuation, the frequency lifting multiple of the enhanced acquisition mode is adjusted in a self-adaptive mode; after the pressure fluctuation reaches a target node, correcting a conduction rate parameter in the pressure conduction time sequence mapping model based on high-density pressure data obtained in an enhancement mode; according to the invention, dynamic adjustment of pipe network pressure data monitoring is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of urban management, and in particular to a method for predicting critical pressure of an urban pipe network based on data modeling. Background Art

[0002] As an important part of urban infrastructure, urban pipe networks undertake the key task of delivering water resources to urban residents and various enterprises. With the acceleration of urbanization, the scale of cities continues to expand, and the coverage and complexity of urban pipe networks are also increasing. The stable operation of the pipe network system is crucial to ensuring the normal production and living order of the city and promoting the sustainable development of the economy and society. Once the pressure in the pipe network is abnormal, it may cause a series of problems such as insufficient water supply, pipe rupture, and water pollution, which will not only bring great inconvenience to the daily life of residents, but also cause huge economic losses and social impacts. Therefore, it is of great practical significance to effectively monitor and predict the pressure of urban pipe networks.

[0003] At present, there are some related technologies and methods in the field of urban pipe network pressure prediction. Some technologies describe the relationship between water flow and pressure in the pipe network by establishing simple mathematical models, and make preliminary predictions on future pressure changes based on historical data. Some technologies use sensors to monitor the pressure in the pipe network in real time, and issue alarms in time when the pressure fluctuates. In addition, some researchers have tried to predict the conduction of pressure fluctuations in the pipe network by analyzing the pipe network topology and water flow status.

[0004] In terms of data collection, the current mathematical model of the pipeline network mostly uses a fixed data collection frequency to periodically monitor the pressure in the pipeline network. Some other technologies use sensors to obtain real-time pressure information in the pipeline network, but lack flexibility in adjusting the data collection frequency. Summary of the invention

[0005] The purpose of the present invention is to provide a method for predicting critical pressure of urban pipe network based on data modeling to solve the following technical problems: The current mathematical model of the pipeline network lacks flexibility in adjusting data collection.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A method for predicting critical pressure of urban pipe network based on data modeling comprises the following steps: Establish a mapping model between the water flow state of the pipe network and the timing of pressure transmission, and predict the expected time for the pressure fluctuation to be transmitted to the specified pipe network node based on the real-time water flow direction, flow rate and valve start and stop status; Before the pressure fluctuation reaches the target node, the standard data acquisition frequency of the target node is maintained; when the predicted remaining conduction time is less than the set threshold, the enhanced acquisition mode of the target node is activated to increase the pressure sensor acquisition frequency to a preset multiple of the standard data acquisition frequency; Adaptively adjust the frequency boost multiple of the enhanced acquisition mode according to the weight level of the pipe section where the target node is located and the amplitude level of the predicted pressure fluctuation; When the pressure fluctuation reaches the target node, the conduction rate parameters in the pressure conduction timing mapping model are corrected based on the high-density pressure data obtained in the enhanced mode.

[0007] As a further solution of the present invention: the establishment of the mapping model includes: Mark all valve positions and water flow detection points in the three-dimensional model of the pipeline network, obtain the water flow direction scalar and flow velocity vector of each pipe section in real time through the water flow sensor array, and convert the water flow direction scalar into a pressure conduction path direction indication signal in the pipeline network topology; construct a pressure wave velocity calculation network with valves as control nodes and pipe sections as conduction paths, associate the valve opening and closing actions with water flow mutation events as the trigger source of pressure conduction, and establish a pressure wave velocity reference parameter library for each pipe section in the pressure wave velocity calculation network, including pipe type, pipe diameter size and service life; according to the dynamic connection relationship of the pipeline network topology, update the correlation between the nodes of the pressure conduction path in real time to ensure that when the valve state changes or the water flow direction changes, the pressure wave velocity calculation network automatically remaps the logical link of the conduction path.

[0008] As a further solution of the present invention: the prediction process of the pressure fluctuation transmission time is: The main path of pressure conduction is determined according to the current water flow direction, and the pipe network nodes that may be affected are marked in sequence downstream along the main path; based on the pipe type, pipe diameter and service life data stored in the pressure wave velocity benchmark model, the theoretical conduction velocity of the pressure wave under interference-free conditions is calculated, and the theoretical conduction time baseline value of each pipe section is generated; combined with the branch diversion ratio data obtained in real time, the main path pressure wave velocity attenuation caused by branch diversion is calculated and superimposed on the theoretical conduction time baseline value; at the same time, the pipe wall state parameters uploaded by the pipe section corrosion detection device are called, and the theoretical conduction time baseline value is dynamically corrected according to the influence of the degree of corrosion on the pressure wave velocity, and finally a conduction time prediction result including environmental interference factors is generated.

[0009] As a further solution of the present invention: the activation process of the enhanced acquisition mode includes: Multiple warning stages are set in the pressure conduction timing prediction model, and each warning stage corresponds to a different remaining conduction time interval and acquisition frequency adjustment strategy; when the predicted remaining conduction time enters the first warning stage, the pressure sensor of the target node is triggered to start the medium-frequency acquisition mode. In this mode, the sensor acquisition frequency is increased to the set first multiple of the conventional mode, and the pressure fluctuation feature pre-analysis function is turned on; when the remaining conduction time enters the second warning stage, it switches to the high-frequency acquisition mode, and the sensor acquisition frequency is increased to the set second multiple of the conventional mode. At the same time, the multi-sensor synchronous calibration mechanism is started to eliminate single-point data errors; after the pressure fluctuation actually reaches the target node, the high-frequency acquisition mode is maintained until the pressure data fluctuation amplitude drops below the preset steady-state threshold, and then the conventional acquisition frequency is automatically restored.

[0010] As a further solution of the present invention: the process of setting the pipe segment weight level is as follows: The basic weight coefficient is divided according to the level of the pipe section in the pipe network topology structure, where the main pipeline connects the core water supply facilities and the regional allocation node, and is assigned the highest basic weight; the secondary pipeline connects the regional allocation node and the secondary user cluster, and is assigned a medium basic weight; the branch pipeline connects the terminal user node, and is assigned the lowest basic weight; the weight correction value generated by the user density data in the service area is superimposed on the basic weight coefficient, and the user density is comprehensively calculated by the number of terminal water metering devices connected to the pipe section and the historical peak water consumption; the risk weighting factor in the historical leakage event database of the pipe section is further loaded, and the risk weighting factor is determined according to the frequency and severity of the event based on the pipe section with leakage or maintenance records within a set period of time.

[0011] As a further solution of the present invention: the classification process of the pressure fluctuation amplitude is: The pressure design value parameter library of the target pipe section is obtained, and the ratio of the current predicted pressure fluctuation peak to the pressure design value is divided into multiple pressure level intervals, and each interval corresponds to a differentiated enhanced acquisition strategy; for low-level pressure fluctuations, only the linear frequency increase strategy of the target node is triggered, that is, the acquisition frequency increases at a fixed ratio and the data storage adopts a conventional compression algorithm; for medium-level pressure fluctuations, the exponential frequency increase strategy and the auxiliary monitoring mode of the adjacent nodes are triggered synchronously, and while the acquisition frequency of the target node is doubled, the pressure sensors of the adjacent nodes are awakened for auxiliary data verification; for high-level pressure fluctuations, on the basis of exponentially increasing the acquisition frequency of the target node, the monitoring equipment of all related branch nodes in the pipeline system is forcibly activated to form a radial collaborative monitoring network centered on the target node, and the multi-node pressure waveforms are integrated in real time through the data bus for cross-validation.

[0012] As a further solution of the present invention: the process of dynamically correcting the conduction rate parameter in the pressure conduction timing prediction model is: In the enhanced acquisition mode, the waveform characteristics of the pressure rising edge recorded by the pressure sensor are extracted, including the peak arrival time, slope change trend and oscillation attenuation period; the measured waveform is aligned and compared with the waveform propagation template preset in the prediction model in the time domain, and the deviation indicators of the two in three dimensions, namely, the peak arrival time difference, waveform morphology similarity and energy decay rate, are calculated; when the deviation in any dimension exceeds the corresponding fault tolerance threshold, the conduction rate parameter optimization program is automatically triggered, and the conduction efficiency parameters of the corresponding pipe section in the pressure wave velocity benchmark model are reversely corrected according to the deviation direction; after the parameter update is completed, the remaining conduction time of all downstream nodes is recalculated, and the prediction result change information is prompted through the visual interface, and the corrected parameters are synchronized to the historical database for subsequent event analysis.

[0013] As a further solution of the present invention: if the actual pressure fluctuation arrival time of the target node in the enhanced acquisition mode is earlier than the predicted time, the reverse conduction path tracing program is started to reversely search for possible abnormal conduction paths along the pressure wave arrival direction in the three-dimensional model of the pipeline network; during the search process, the pipe segment connection relationship database is called to identify all hidden conduction channels that are physically connected but not included in the current main path prediction; for the discovered hidden channels, their pipe segment attribute data are extracted and loaded into the pressure wave velocity calculation network, and the corresponding conduction parameter calculation branch is added; after completing the path update, the pressure conduction timing prediction model is re-run, and the path characteristics and parameter correction records of the hidden channels are added to the abnormal event knowledge base.

[0014] Beneficial effects of the present invention: The present invention can accurately predict the expected time of pressure fluctuation transmission to the designated pipe network node by establishing a mapping model of pipe network water flow state and pressure transmission time series, combined with real-time water flow direction, flow rate and valve start-stop state, and solves the problem of large prediction deviation caused by insufficient consideration of dynamic factors in the prior art. In terms of data acquisition, the standard acquisition frequency is maintained before the pressure fluctuation reaches the target node, and the enhanced acquisition mode is activated when the predicted transmission remaining time is less than the set threshold, and the frequency enhancement multiple is adaptively adjusted according to the pipe section weight level and the pressure fluctuation amplitude level, avoiding the data redundancy or insufficient collection caused by the fixed acquisition frequency, and timely obtaining detailed pressure data for monitoring the key pipe network conditions. At the same time, after the pressure fluctuation reaches the target node, the conduction rate parameters in the pressure transmission time series mapping model are corrected based on the high-density pressure data in the enhanced mode, and the prediction model can be optimized by using real-time data, so as to improve the adaptability and accuracy of the model and effectively cope with the complex and changeable pipe network operation environment. In addition, when the actual pressure fluctuation arrives earlier than the predicted time, the reverse conduction path tracing program is started, which can find the hidden conduction channel and update the prediction model, further improving the accuracy and reliability of the pressure conduction prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the accompanying drawings.

[0016] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0018] See also Figure 1 As shown, the present invention is a method for predicting critical pressure of urban pipe network based on data modeling, comprising the following steps: First, a mapping model between the water flow state of the pipe network and the pressure conduction timing is established. In the urban pipe network system, the water flow state is in dynamic change, which is affected by a variety of factors such as real-time water flow direction, flow velocity, and valve start and stop status. In order to accurately predict the expected time for pressure fluctuations to be transmitted to the specified pipe network node, these factors need to be fully considered. By constructing a mapping model, the water flow state is associated with the pressure conduction timing, and the intrinsic relationship between them is described in the form of a mathematical model. In the process of model establishment, relevant knowledge such as pipe network topology and water flow dynamics principles will be used, combined with real-time collected data such as water flow direction, flow velocity, and valve status, to achieve accurate prediction of the pressure fluctuation conduction time.

[0019] Secondly, when the pressure fluctuation has not yet reached the target node, maintain the standard data collection frequency of the target node. The standard data collection frequency is pre-set according to the normal operation of the pipeline network and the data collection requirements, and can meet the basic monitoring requirements of the pipeline network pressure data under normal circumstances. When the remaining time for the pressure fluctuation transmission predicted by the mapping model is less than the set threshold, it means that the pressure fluctuation is about to reach the target node. At this time, activate the enhanced collection mode of the target node and increase the collection frequency of the pressure sensor to a preset multiple of the standard data collection frequency. The purpose of the enhanced collection mode is to obtain more intensive and detailed pressure data during the critical period of pressure fluctuations, so as to more accurately analyze the pressure changes.

[0020] Then, according to the weight level of the pipe section where the target node is located and the amplitude level of the predicted pressure fluctuation, the frequency boost multiple of the enhanced acquisition mode is adaptively adjusted. The weight level of the pipe section is determined by comprehensively considering multiple factors such as the level of the pipe section in the pipe network topology, the user density in the service area, and the historical leakage events of the pipe section, reflecting the importance of the pipe section in the entire pipe network system. The amplitude level of pressure fluctuation is divided according to the ratio of the pressure design value of the target pipe section to the current predicted pressure fluctuation peak value, reflecting the severity of the pressure fluctuation. According to these two levels, the frequency boost multiple in the enhanced acquisition mode is dynamically adjusted, so that the data acquisition frequency can be flexibly adjusted according to the actual situation, which not only ensures that sufficient data can be obtained in key pipe sections and high-amplitude pressure fluctuations, but also avoids unnecessary data collection, improving the efficiency and pertinence of data collection.

[0021] Finally, when the pressure fluctuation reaches the target node, the conduction rate parameters in the pressure conduction time series mapping model are corrected based on the high-density pressure data obtained in the enhanced mode. Due to the complex and changeable operating environment of the actual pipeline network and the existence of many uncertain factors, the conduction rate parameters in the mapping model may deviate from the actual situation. By analyzing the high-density pressure data obtained in the enhanced acquisition mode, the actual situation of pressure conduction can be understood more accurately. These actual data are compared and analyzed with the model prediction results to find out the deviation of the conduction rate parameters and correct them. The corrected model can more accurately reflect the actual situation of pressure conduction in the pipeline network, thereby improving the accuracy and reliability of subsequent pressure fluctuation predictions.

[0022] In a preferred embodiment of the present invention, regarding the establishment of the mapping model, the specific operations are as follows: In the constructed three-dimensional model of the pipe network, all valve positions and water flow detection points are accurately marked. The water flow direction scalar and flow velocity vector of each pipe section are obtained in real time and accurately using the water flow sensor array. The obtained water flow direction scalar is converted into a pressure conduction path direction indication signal suitable for the pipe network topology structure according to a specific conversion algorithm. The signal can intuitively and clearly indicate the direction of pressure conduction. Then, a pressure wave velocity calculation network with valves as control nodes and pipe sections as conduction paths is constructed. In this network system, the opening and closing actions of valves and water flow mutation events are set as trigger sources for pressure conduction through logical association. In addition, for each pipe section in the pressure wave velocity calculation network, a pressure wave velocity benchmark parameter library is specially established, which stores key information such as pipe type, pipe diameter and service life in detail. As the pipe network is running, the dynamic connection relationship of the pipe network topology structure is constantly changing. Based on the information monitored in real time, the correlation between nodes in the pressure conduction path is updated in a timely manner. Ensure that when the valve state changes or the water flow direction changes, the pressure wave velocity calculation network can automatically and quickly remap the logical links of the conduction path, thereby ensuring that the mapping model is always closely aligned with the actual operation of the pipeline network and achieving accurate simulation of the pressure conduction path.

[0023] In another preferred embodiment of the present invention, the specific steps of the pressure fluctuation transmission time prediction process are as follows: First, according to the current real-time monitored water flow direction, the main path of pressure conduction is accurately determined by analyzing the pipe network topology and the principle of hydrodynamics. Along the determined main path, according to certain rules, the pipe network nodes that may be affected by pressure fluctuations are marked in the downstream direction. Then, based on the data such as pipe type, pipe diameter and service life pre-stored in the pressure wave velocity benchmark model, the corresponding calculation formula is used to calculate the theoretical conduction velocity of the pressure wave under ideal conditions without interference. Based on this theoretical conduction velocity, the theoretical conduction time baseline value of each pipe section is further generated. At the same time, the branch diversion ratio data in the pipe network is obtained in real time, and according to the principle of hydrodynamics, the main path pressure wave velocity attenuation caused by branch diversion is calculated, and the attenuation is superimposed on the previously generated theoretical conduction time baseline value. In addition, the pipe wall state parameters uploaded in real time by the pipe section corrosion detection device are called, and the theoretical conduction time baseline value is dynamically corrected in combination with the existing research results on the influence of corrosion degree on pressure wave velocity. Taking all the above factors into consideration, a comprehensive and accurate conduction time prediction result including various environmental interference factors is finally generated, providing reliable data support for subsequent pipeline pressure monitoring and analysis.

[0024] In another preferred embodiment of the present invention, the activation process of the enhanced acquisition mode is as follows: In the pressure conduction timing prediction model, multiple early warning stages are set scientifically and reasonably according to the characteristics of the pressure fluctuation conduction time and the needs of data collection. Each early warning stage corresponds to a different remaining conduction time interval, and a special acquisition frequency adjustment strategy is formulated for each interval. When the remaining conduction time calculated by the prediction model enters the first early warning stage, the system will automatically trigger the pressure sensor of the target node to start the medium-frequency acquisition mode. In this medium-frequency acquisition mode, the acquisition frequency of the sensor is increased according to the pre-set first multiple, which is significantly improved compared to the conventional mode, and more pressure data can be obtained. At the same time, the pressure fluctuation feature pre-analysis function is turned on to perform a preliminary feature analysis of the upcoming pressure fluctuations to provide basic data for subsequent processing.

[0025] When the remaining conduction time advances further and enters the second warning stage, the system will automatically switch the pressure sensor's acquisition mode to the high-frequency acquisition mode. At this time, the sensor's acquisition frequency is increased according to the set second multiplier, further significantly increasing the frequency of data acquisition to meet the requirements for data accuracy when pressure fluctuations are approaching. In addition, when switching to the high-frequency acquisition mode, the multi-sensor synchronous calibration mechanism is started. Through this mechanism, the data of multiple sensors are synchronously calibrated, effectively eliminating the possible errors in single-point data and ensuring that the collected data is accurate and reliable.

[0026] After the pressure fluctuation actually reaches the target node, in order to fully and accurately monitor the entire process of pressure fluctuation, the system will continue to maintain the high-frequency acquisition mode until the fluctuation amplitude of the pressure data drops below the preset steady-state threshold, indicating that the pressure fluctuation has stabilized. At this time, the system will automatically return to the normal acquisition frequency, thereby ensuring data quality while reasonably controlling the resource consumption of data acquisition.

[0027] In another preferred embodiment of the present invention, the specific steps of setting the pipe segment weight level are as follows: First, the basic weight coefficients of the pipe sections are divided according to the hierarchical relationship of the pipe sections in the pipe network topology. Among them, the main pipeline, as the key part connecting the core water supply facilities and the regional distribution nodes, undertakes the main water delivery task in the entire pipe network system. Its importance is self-evident, so it is given the highest basic weight. The secondary pipeline connects the regional distribution nodes and the secondary user clusters, and plays a connecting role in the pipe network system, and is given a medium basic weight. The branch pipeline mainly connects the end user nodes, which are relatively less important in the pipe network and are given the lowest basic weight.

[0028] Next, on the basis of the basic weight coefficient, the weight correction value generated by the user density data in the service area is superimposed. The user density is comprehensively calculated by the number of terminal water metering devices connected to the pipe section and the historical peak water consumption. The higher the user density, the more users the pipe section serves, the greater the impact on users, and therefore the higher the corresponding weight correction value.

[0029] Finally, the risk weighting factor in the historical leakage event database of the pipe section is further loaded. This risk weighting factor is determined based on the pipe sections with leakage or maintenance records within the set time, according to the frequency and severity of the incident. For pipe sections with a high frequency and severity of leakage or maintenance events within the set time, a higher risk weighting factor is assigned to increase its weight level; conversely, a lower risk weighting factor is assigned. Through the above steps, a pipe section weight level setting system is finally formed that comprehensively considers multi-dimensional factors such as the pipeline network topology, user impact range, and risk history.

[0030] In another preferred embodiment of the present invention, regarding the classification process of the pressure fluctuation amplitude, the specific operations are as follows: First, obtain the pressure design value parameter library of the target pipe section, which stores key parameters such as the pressure design value of the target pipe section. Then, compare the current predicted pressure fluctuation peak with the pressure design value and calculate their ratio. According to the size of this ratio, it is divided into multiple different pressure level intervals. Each pressure level interval corresponds to a differentiated enhanced acquisition strategy to meet the data acquisition needs under different degrees of pressure fluctuations.

[0031] For low-level pressure fluctuations, that is, when the ratio is in a lower range, only the linear frequency increase strategy of the target node is triggered. Under this strategy, the acquisition frequency increases at a fixed ratio, and the data storage uses a conventional compression algorithm to minimize the pressure on data storage while ensuring that certain pressure data can be obtained.

[0032] For medium-level pressure fluctuations, when the ratio is in the medium range, the exponential frequency increase strategy and the auxiliary monitoring mode of the adjacent nodes are triggered synchronously. While the acquisition frequency of the target node is exponentially multiplied, the pressure sensor of the adjacent node is awakened and put into the auxiliary monitoring state, and the pressure data of the target node is auxiliary verified, thereby improving the accuracy and reliability of the data.

[0033] For high-level pressure fluctuations, when the ratio is in a higher range, the monitoring equipment of all related branch nodes in the pipe network system is activated on the basis of exponentially increasing the acquisition frequency of the target node. These devices together form a radial collaborative monitoring network centered on the target node, integrating the pressure waveform data of multiple nodes in real time through the data bus, and cross-validating to comprehensively and accurately monitor the high-level pressure fluctuations, providing strong data support for the safe operation of the pipe network.

[0034] In another preferred embodiment of the present invention, the process of dynamically correcting the conduction rate parameter of the pressure conduction timing prediction model is as follows: After the enhanced acquisition mode is turned on, the characteristic information of the pressure rising edge waveform is extracted from the data recorded by the pressure sensor. These features include the peak arrival time, that is, the moment when the pressure rises to the peak; the slope change trend, which reflects the speed of the pressure rise; and the oscillation decay period, that is, the time period during which the pressure fluctuation gradually weakens. Next, the measured pressure waveform is aligned and compared with the pre-set waveform propagation template in the prediction model in the time domain. During the comparison process, the deviation indicators of the two in three key dimensions are calculated: the peak arrival time difference, which reflects the difference between the measured and predicted peak arrival times; the waveform morphology similarity, which measures the similarity in shape between the measured waveform and the template waveform; the energy decay rate deviation, which reflects the difference in the pressure energy decay rate.

[0035] Once the deviation value of any dimension exceeds the corresponding fault tolerance threshold, the system will automatically trigger the conduction rate parameter optimization program. This program will reversely correct the conduction efficiency parameters of the corresponding pipe section in the pressure wave velocity benchmark model according to the direction of the deviation. For example, if the measured wave peak arrives earlier than the predicted time, it means that the conduction rate may be faster than the model estimate. At this time, the conduction efficiency parameter will be appropriately increased; otherwise, it will be reduced.

[0036] After completing the parameter update operation, the remaining conduction time of all downstream nodes is recalculated. At the same time, the change information of the prediction results is prompted to the relevant personnel through the visual interface, so that they can grasp the situation in time. In addition, the corrected parameters are synchronized to the historical database for subsequent analysis and reference for similar events.

[0037] In a preferred case of this embodiment, if the actual pressure fluctuation arrival time of the target node is earlier than the predicted time in the enhanced acquisition mode, the reverse conduction path tracing program will be started. The program will perform a reverse search along the direction of the pressure wave arrival in the three-dimensional model of the pipe network to find possible abnormal conduction paths. During the search, the pipe segment connection relationship database is called to identify all hidden conduction channels that are physically interconnected but not included in the current main path prediction range.

[0038] For the hidden channel found by the search, the attribute data of its pipe section, such as pipe type, pipe diameter, service life, etc., are extracted and loaded into the pressure wave velocity calculation network. At the same time, a conduction parameter calculation branch corresponding to the hidden channel is added to the network to accurately calculate the conduction of the pressure wave in the channel.

[0039] After completing the path update, rerun the pressure conduction timing prediction model so that the model can take into account the newly discovered hidden channel, thereby obtaining a more accurate pressure conduction prediction result. Finally, the path characteristics of the hidden channel and the relevant records of parameter correction are added to the abnormal event knowledge base to provide experience and data support for subsequent processing of similar abnormal situations.

[0040] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for predicting critical pressure of urban pipe network based on data modeling, characterized in that: The following steps are involved: Establish a mapping model between the water flow state of the pipe network and the timing of pressure transmission, and predict the expected time for the pressure fluctuation to be transmitted to the specified pipe network node based on the real-time water flow direction, flow rate and valve start and stop status; Before the pressure fluctuation reaches the target node, the standard data acquisition frequency of the target node is maintained; when the predicted remaining conduction time is less than the set threshold, the enhanced acquisition mode of the target node is activated to increase the pressure sensor acquisition frequency to a preset multiple of the standard data acquisition frequency; Adaptively adjust the frequency boost multiple of the enhanced acquisition mode according to the weight level of the pipe section where the target node is located and the amplitude level of the predicted pressure fluctuation; When the pressure fluctuation reaches the target node, the conduction rate parameters in the pressure conduction timing mapping model are corrected based on the high-density pressure data obtained in the enhanced mode.

2. According to the method for predicting critical pressure of urban pipe network based on data modeling according to claim 1, it is characterized in that: The establishment of the mapping model includes: Mark all valve positions and water flow detection points in the three-dimensional model of the pipeline network, obtain the water flow direction scalar and flow velocity vector of each pipe section in real time through the water flow sensor array, and convert the water flow direction scalar into a pressure conduction path direction indication signal in the pipeline network topology; construct a pressure wave velocity calculation network with valves as control nodes and pipe sections as conduction paths, associate the valve opening and closing actions with water flow mutation events as the trigger source of pressure conduction, and establish a pressure wave velocity reference parameter library for each pipe section in the pressure wave velocity calculation network, including pipe type, pipe diameter size and service life; according to the dynamic connection relationship of the pipeline network topology, update the correlation between the nodes of the pressure conduction path in real time to ensure that when the valve state changes or the water flow direction changes, the pressure wave velocity calculation network automatically remaps the logical link of the conduction path.

3. The method for predicting critical pressure of urban pipe network based on data modeling according to claim 1 is characterized in that: The prediction process of pressure fluctuation transmission time is: The main path of pressure conduction is determined according to the current water flow direction, and the pipe network nodes that may be affected are marked in sequence downstream along the main path; based on the pipe type, pipe diameter and service life data stored in the pressure wave velocity benchmark model, the theoretical conduction velocity of the pressure wave under interference-free conditions is calculated, and the theoretical conduction time baseline value of each pipe section is generated; combined with the branch diversion ratio data obtained in real time, the main path pressure wave velocity attenuation caused by branch diversion is calculated and superimposed on the theoretical conduction time baseline value; at the same time, the pipe wall state parameters uploaded by the pipe section corrosion detection device are called, and the theoretical conduction time baseline value is dynamically corrected according to the influence of the degree of corrosion on the pressure wave velocity, and finally a conduction time prediction result including environmental interference factors is generated.

4. The method for predicting critical pressure of urban pipe network based on data modeling according to claim 1 is characterized in that: The activation process of the enhanced acquisition mode includes: Multiple warning stages are set in the pressure conduction timing prediction model, and each warning stage corresponds to a different remaining conduction time interval and acquisition frequency adjustment strategy; when the predicted remaining conduction time enters the first warning stage, the pressure sensor of the target node is triggered to start the medium-frequency acquisition mode. In this mode, the sensor acquisition frequency is increased to the set first multiple of the conventional mode, and the pressure fluctuation feature pre-analysis function is turned on; when the remaining conduction time enters the second warning stage, it switches to the high-frequency acquisition mode, and the sensor acquisition frequency is increased to the set second multiple of the conventional mode. At the same time, the multi-sensor synchronous calibration mechanism is started to eliminate single-point data errors; after the pressure fluctuation actually reaches the target node, the high-frequency acquisition mode is maintained until the pressure data fluctuation amplitude drops below the preset steady-state threshold, and then the conventional acquisition frequency is automatically restored.

5. The method for predicting critical pressure of urban pipe network based on data modeling according to claim 1 is characterized in that: The process of setting the pipe segment weight level is as follows: The basic weight coefficient is divided according to the level of the pipe section in the pipe network topology structure, where the main pipeline connects the core water supply facilities and the regional allocation node, and is assigned the highest basic weight; the secondary pipeline connects the regional allocation node and the secondary user cluster, and is assigned a medium basic weight; the branch pipeline connects the terminal user node, and is assigned the lowest basic weight; the weight correction value generated by the user density data in the service area is superimposed on the basic weight coefficient, and the user density is comprehensively calculated by the number of terminal water metering devices connected to the pipe section and the historical peak water consumption; the risk weighting factor in the historical leakage event database of the pipe section is further loaded, and the risk weighting factor is determined according to the frequency and severity of the event based on the pipe section with leakage or maintenance records within a set period of time.

6. The method for predicting critical pressure of urban pipe network based on data modeling according to claim 1 is characterized in that: The classification process of pressure fluctuation amplitude is: Obtain the pressure design value parameter library of the target pipe section, divide the ratio of the current predicted pressure fluctuation peak to the pressure design value into multiple pressure level intervals, and each interval corresponds to a differentiated enhanced acquisition strategy; for low-level pressure fluctuations, only the linear frequency increase strategy of the target node is triggered, that is, the acquisition frequency increases at a fixed ratio and the data storage adopts a conventional compression algorithm; For medium-level pressure fluctuations, the exponential frequency increase strategy and the auxiliary monitoring mode of adjacent nodes are triggered synchronously. While the acquisition frequency of the target node is doubled, the pressure sensors of the adjacent nodes are awakened for auxiliary data verification. For high-level pressure fluctuations, on the basis of exponentially increasing the acquisition frequency of the target node, the monitoring equipment of all related branch nodes in the pipeline system is forcibly activated to form a radial collaborative monitoring network centered on the target node, and the multi-node pressure waveforms are integrated in real time through the data bus for cross-verification.

7. The method for predicting critical pressure of urban pipe network based on data modeling according to claim 1 is characterized in that: The process of dynamically correcting the conduction velocity parameters in the pressure conduction timing prediction model is as follows: In the enhanced acquisition mode, the pressure rising edge waveform characteristics recorded by the pressure sensor are extracted, including the peak arrival time, slope change trend and oscillation attenuation period; The measured waveform is aligned and compared with the waveform propagation template preset in the prediction model in the time domain, and the deviation indicators of the two in three dimensions, namely, the peak arrival time difference, waveform morphology similarity and energy decay rate, are calculated; when the deviation in any dimension exceeds the corresponding fault tolerance threshold, the conduction rate parameter optimization program is automatically triggered, and the conduction efficiency parameters of the corresponding pipe section in the pressure wave velocity benchmark model are reversely corrected according to the deviation direction; after the parameter update is completed, the remaining conduction time of all downstream nodes is recalculated, and the prediction result change information is prompted through a visual interface, and the corrected parameters are synchronized to the historical database for subsequent event analysis.

8. The method for predicting critical pressure of urban pipe network based on data modeling according to claim 7 is characterized in that: If the actual pressure fluctuation arrival time of the target node in the enhanced acquisition mode is earlier than the predicted time, the reverse conduction path tracing program is started to search for possible abnormal conduction paths in the three-dimensional model of the pipe network in the reverse direction of the pressure wave arrival direction; during the search process, the pipe segment connection relationship database is called to identify all hidden conduction channels that are physically connected but not included in the current main path prediction; for the discovered hidden channels, their pipe segment attribute data is extracted and loaded into the pressure wave velocity calculation network, and the corresponding conduction parameter calculation branch is added; After completing the path update, the pressure conduction timing prediction model is rerun, and the path characteristics and parameter correction records of the hidden channel are added to the abnormal event knowledge base.

Images