Water supply and drainage pipe network fault positioning method and system
By collecting pipeline parameters and case library clustering in fault location of water supply and drainage pipeline networks, determining sensors for detection points, calculating time delay and signal attenuation coefficients, and optimizing detection point information using Bayesian optimization algorithm, the problems of low fault positioning efficiency and insufficient accuracy in the existing technology are solved, and efficient and accurate fault positioning and self-learning of detection solutions are achieved.
Patent Information
- Application Number
- CN202510360245.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-25
- Publication Date
- 2025-06-20
AI Technical Summary
The existing ultrasonic detection methods have problems such as low detection efficiency, insufficient accuracy, high equipment costs, low resource utilization and lack of effective case library construction and update in the fault location of water supply and drainage pipelines.
By collecting the parameter information of the water supply and drainage pipe to be tested and performing cluster analysis with the case library, generating a detection plan; determine the detection point in the pipeline to arrange sensors, send ultrasonic pulses and collect the received signal, and obtain the target received signal through matching filtering; calculate the time delay and signal attenuation coefficient, establish the target optimization function, and determine the fault location through optimization solution; use Bayesian optimization algorithm to optimize the detection point information and update the case library.
It improves the accuracy and reliability of fault location, reduces the cost of detection equipment, improves resource utilization, and realizes self-learning and iteration of detection solutions by dynamically updating the case library.
Smart Images
Figure CN120176030A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of drainage pipe detection, and particularly to a method and system for fault location of water supply and drainage pipe networks. Background Art
[0002] In the maintenance and management of water supply and drainage pipe networks, fault location is a crucial task. Traditional methods usually rely on manual inspections, experience-based judgments, and simple instrument detections. With the development of technology, ultrasonic detection technology has gradually been applied to the fault diagnosis of water supply and drainage pipes. However, existing ultrasonic detection methods still have some limitations: 1) Using a unified standard or fixed template to formulate detection plans, without fully considering the influence of different pipe materials, resulting in low detection efficiency and insufficient accuracy; 2) Lacking scientific guidance for the sampling of detection points and the layout method of detection equipment, leading to high costs of detection equipment and low resource utilization, which is not conducive to accurately locating the fault position; 3) Lacking an effective case library construction and update mechanism, historical fault data not being fully utilized, making the design of new detection schemes lack sufficient reference basis. Summary of the Invention
[0003] To solve at least one of the above-mentioned technical problems, the present invention provides a method and system for fault location of water supply and drainage pipe networks.
[0004] In a first aspect, the present invention provides a method for fault location of water supply and drainage pipe networks, the method comprising:
[0005] Collecting parameter information of the water supply and drainage pipe to be measured, performing cluster analysis on the parameter information and the case library to obtain a first detection scheme for the water supply and drainage pipe to be measured; the parameter information includes the length, diameter, and material characteristics of the water supply and drainage pipe to be measured; the case library is constructed based on historical fault location data of different water supply and drainage pipes;
[0006] Determining detection points in the pipe according to the first detection scheme, arranging sensors at each detection point as signal receiving points, sending ultrasonic pulses with a fixed power multiple times at a preset interval as target emission signals, collecting multiple groups of received signals of each sensor, and obtaining target received signals of each detection point by performing matched filtering on the multiple groups of received signals;
[0007] Determining the time delay and signal attenuation coefficient of each detection point according to the target emission signal and the target received signal, establishing a target optimization function based on the time delay and signal attenuation coefficient of each detection point, and obtaining the fault location by optimizing and solving the target optimization function;
[0008] Calculating the deviation distance between the fault location and the detection points, optimizing the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance to obtain a second detection scheme; updating the case library according to the second detection scheme.
[0009] Preferably, determining the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal includes:
[0010] Δt i = t i - t0;
[0011] In the formula, Δt i is the time delay of the i-th detection point, t i is the time when the sensor at the i-th detection point receives the signal, and t0 is the signal transmission time;
[0012]
[0013] In the formula, α i is the signal attenuation coefficient of the i-th detection point, I i is the target reception signal intensity at the i-th detection point, I0 is the target transmission signal intensity, and d i is the distance from the emission point to the i-th detection point.
[0014] Preferably, establishing the target optimization function according to the time delay and signal attenuation coefficient of each detection point, and obtaining the fault location by optimizing and solving the target optimization function includes:
[0015] Construct a three-dimensional rectangular coordinate system with the pipe starting point as the origin. Let (x i , y i , z i ) be the position coordinates of the i-th detection point, and c be the propagation speed of ultrasonic waves in the medium. Then the target optimization function is:
[0016]
[0017] In the formula, F(x, y, z) is the target optimization function, w1 and w2 are the weight coefficients of the time delay and attenuation rate respectively, α0 is the theoretical attenuation coefficient of the water supply and drainage pipe to be measured, and N represents the number of detection points;
[0018] Use the gradient descent algorithm to solve the coordinates (x r , y r , z r ) that minimize F(x, y, z) to obtain the fault location.
[0019] Preferably, calculating the deviation distance between the fault location and the detection point, and optimizing the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance to obtain the second detection scheme includes:
[0020] Calculate the deviation distance d bias between the fault location and the detection point:
[0021]
[0022] Define the objective function f(X), where X represents the position vector of the detection point, to minimize the deviation distance d bias Construct the objective function:
[0023] f(X) = -d bias ;
[0024] Construct a Gaussian process model, initialize the kernel function and hyperparameters of the model, select a preset number of detection point samples as the data set, and use the data set to train the Gaussian process model to construct a prior distribution;
[0025] Determine the acquisition function, and find a new detection point X according to the acquisition function new , and add X new to the data set to update the posterior distribution of the Gaussian process model; Iteratively loop through the steps of updating the posterior distribution until the model converges to obtain the optimal solution and the optimized deviation distance d bias-new ;
[0026] Adjust the position of the detection point in the first detection plan according to the optimal solution, and update the number of detection points according to the optimized deviation distance d bias-new to generate a second detection plan;
[0027] Among them, updating the number of detection points according to the optimized deviation distance d bias-new includes:
[0028]
[0029] In the formula, N represents the number of detection points before updating, N new represents the number of detection points after updating, and k represents the adjustment coefficient, with a value range of 0.1 to 0.5.
[0030] In a second aspect, the present invention also provides a fault location system for a water supply and drainage pipe network, and the system includes:
[0031] A detection plan clustering unit, configured to collect parameter information of the water supply and drainage pipe to be measured, perform clustering analysis on the parameter information and the case base, and obtain a first detection plan for the water supply and drainage pipe to be measured; the parameter information includes the length, diameter, and material characteristics of the water supply and drainage pipe to be measured; the case base is constructed based on historical fault location data of different water supply and drainage pipes;
[0032] The detection point layout unit is used to determine the detection points in the pipeline according to the first detection scheme, lay out sensors at each detection point as signal receiving points, send ultrasonic pulses with a fixed power multiple times at a preset interval as the target transmission signal, collect multiple groups of received signals of each sensor, and obtain the target received signal of each detection point by performing matched filtering on the multiple groups of received signals;
[0033] The fault location positioning unit is used to determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target received signal, establish a target optimization function based on the time delay and signal attenuation coefficient of each detection point, and obtain the fault location by optimizing and solving the target optimization function;
[0034] The detection scheme update unit is used to calculate the deviation distance between the fault location and the detection points, optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance to obtain the second detection scheme; update the case base according to the second detection scheme.
[0035] Preferably, the fault location positioning unit, used to determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target received signal, includes:
[0036] Δt i =t i -t0;
[0037] In the formula, Δt i is the time delay of the i-th detection point, t i is the time when the sensor at the i-th detection point receives the signal, and t0 is the signal transmission time;
[0038]
[0039] In the formula, α i is the signal attenuation coefficient of the i-th detection point, I i is the intensity of the target received signal at the i-th detection point, I0 is the intensity of the target transmission signal, and d i is the distance from the emission point to the i-th detection point.
[0040] Preferably, the fault location positioning unit, used to establish a target optimization function based on the time delay and signal attenuation coefficient of each detection point and obtain the fault location by optimizing and solving the target optimization function, includes:
[0041] Construct a three-dimensional rectangular coordinate system with the starting point of the pipeline as the origin. Let (x i , y i , z i ) be the position coordinates of the i-th detection point, and c be the propagation speed of ultrasonic waves in the medium. Then the target optimization function is:
[0042]
[0043] Wherein, F(x, y, z) is the target optimization function, w1 and w2 are the weight coefficients of the time delay and the attenuation rate respectively, α0 is the theoretical attenuation coefficient of the water supply and drainage pipe to be measured, and N represents the number of detection points;
[0044] Use the gradient descent algorithm to solve the coordinates (x r , y r , z r ) that minimize F(x, y, z) to obtain the fault location.
[0045] Preferably, the detection scheme updating unit is configured to calculate the deviation distance between the fault location and the detection points, and optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance to obtain a second detection scheme, including:
[0046] Calculate the deviation distance d between the fault location and the detection points bias :
[0047]
[0048] Define the objective function f(X), where X represents the position vector of the detection points, to minimize the deviation distance d bias Construct the objective function:
[0049] f(X) = -d bias ;
[0050] Construct a Gaussian process model, initialize the kernel function and hyperparameters of the model, select a preset number of detection point samples as the data set, and use the data set to train the Gaussian process model to construct a prior distribution;
[0051] Determine the acquisition function, and find a new detection point X new , add X new to the data set to update the posterior distribution of the Gaussian process model; iterate and loop the steps of updating the posterior distribution until the model converges to obtain the optimal solution and the optimized deviation distance d bias-new ;
[0052] Adjust the detection point positions in the first detection scheme according to the optimal solution, and update the number of detection points according to the optimized deviation distance d bias-new to generate a second detection scheme;
[0053] Wherein, updating the number of detection points according to the optimized deviation distance d bias-new includes:
[0054]
[0055] Wherein, N represents the number of detection points before update, and N new represents the number of detection points after update, and k represents an adjustment coefficient, whose value range is 0.1 to 0.5.
[0056] In a third aspect, the present invention further provides an electronic device, including a processor and a memory. The memory is used to store computer program code, and the computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes the method according to the first aspect and any possible implementation manner thereof as described above.
[0057] In a fourth aspect, the present invention further provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor is caused to execute the method according to the first aspect and any possible implementation manner thereof as described above.
[0058] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0059] 1) By collecting detailed parameter information of the water supply and drainage pipes to be measured, including length, diameter, and material properties, and performing clustering analysis with the case library, a targeted first detection scheme is generated. Compared with the uncertainty of formulating a scheme based on manual experience, this method can provide a scientific basis on the basis of historical cases, making the first detection scheme more reasonable and effective.
[0060] 2) By sending ultrasonic pulses with a fixed power multiple times at preset intervals at each detection point, performing matched filtering processing on multiple groups of received signals, and extracting a more accurate target received signal. In this way, the signal-to-noise ratio can be enhanced, the influence of random noise is reduced, and the reliability of fault location is improved. Calculate the time delay and signal attenuation coefficient according to the target transmitted signal and the received signal, establish a target optimization function, and determine the fault location through optimization solution, thereby realizing the accurate positioning of the fault location.
[0061] 3) Use the Bayesian optimization algorithm to optimize the detection point information in combination with the deviation distance to form a second detection scheme, and feedback the second detection scheme after each optimization to the case library to keep the case library dynamically updated. Through the role of the Bayesian optimization algorithm in optimizing the detection point layout in combination with the actual deviation distance, not only the accuracy of fault location is improved, but also the flexibility and robustness of the system are enhanced, enabling it to better cope with complex and changeable actual environments. Through continuous optimization of the second detection scheme, self-learning iteration of the detection scheme is realized, ensuring long-term stable high-performance detection.
[0062] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and do not limit the present disclosure. Brief Description of the Drawings
[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the background art, the drawings required for use in the embodiments of the present invention or the background art will be described below.
[0064] The drawings herein are incorporated into the specification and form a part of this specification. These drawings show embodiments in accordance with the present disclosure and, together with the specification, are used to illustrate the technical solutions of the present disclosure.
[0065] Figure 1 It is a schematic flowchart of a method for fault location of a water supply and drainage pipe network provided by an embodiment of the present invention;
[0066] Figure 2 It is a schematic structural diagram of a water supply and drainage pipe network fault location system provided by an embodiment of the present invention. Detailed Description of the Embodiments
[0067] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0068] The mention of "embodiment" in this article means that the specific features, structures, or characteristics described in connection with the embodiment may be included in at least one embodiment of the present invention. The phrase appears in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.
[0069] Please refer to Figure 1 , Figure 1 It is a schematic flowchart of a method for fault location of a water supply and drainage pipe network provided by an embodiment of the present invention. As Figure 1 shown, the method includes:
[0070] S10. Collect parameter information of the water supply and drainage pipe to be measured, perform cluster analysis on the parameter information and the case library, and obtain a first detection plan for the water supply and drainage pipe to be measured; the parameter information includes the length, diameter, and material properties of the water supply and drainage pipe to be measured; the case library is constructed based on historical fault location data of different water supply and drainage pipes.
[0071] First, construct a case library based on the historical fault location data of different water supply and drainage pipes. Extract the historical fault location data of different types of water supply and drainage pipes from past maintenance records, including but not limited to detection methods, fault types, locations, discovery times, and repair measures. Clean, organize, and convert this data into a unified format to ensure that each record contains complete parameter information and corresponding fault characteristics. As new fault cases accumulate, update the case library regularly to maintain its timeliness and representativeness. Among them, typical cases or cases with high detection efficiency and accuracy should be specially marked.
[0072] Traditional ultrasonic detection methods usually determine the detection plan based on manual experience. For example, for the layout of signal generators and receivers, usually only historical data is used to screen out the fault points that frequently appear in the water supply and drainage pipes at high frequencies first, and then receivers are arranged at the positions of these points and their peripheries. Such a method has strong randomness and is very likely to cause the failure to identify the real fault points. Another method is to have no prior evaluation and only arrange them evenly at a certain interval according to the length of the water pipe, as much as possible to ensure that every position can be detected. However, this relatively conventional method requires a large amount of cost, including the time and labor costs of laying detection equipment in the early stage, equipment costs, and later maintenance costs, and is not conducive to popularizing to daily detection work.
[0073] Therefore, in order to provide a relatively scientific and reasonable detection plan in this embodiment, the parameter information of the water supply and drainage pipe to be measured is considered, and the parameter information is subjected to cluster analysis with the case library to obtain the first detection plan for the water supply and drainage pipe to be measured. The parameter information includes the length, diameter, and material characteristics of the water supply and drainage pipe to be measured.
[0074] Specifically, clustering algorithms such as K-means, DBSCAN, and hierarchical clustering can be used to classify the data in the case library according to the pipe parameter information. For the selection of features, key features that can significantly distinguish different types of pipes and their fault modes are considered, such as length, diameter, and material characteristics. Before clustering, define a reasonable distance metric method, such as Euclidean distance, cosine similarity, etc., to measure the similarity between the pipe to be measured and each cluster in the case library. Based on the clustering results, find one or more clusters that are closest to the characteristics of the pipe to be measured as reference templates. Finally, referring to the successful detection experience within the selected cluster, design a preliminary first detection plan for the pipe to be measured, covering aspects such as the selection of detection points, sensor layout, and transmission signal setting. Finally, manual review can also be carried out to ensure the rationality and feasibility of the first detection plan.
[0075] Therefore, by utilizing existing historical fault data and advanced clustering analysis techniques, customized detection strategies can be developed according to the characteristics of different pipelines, enabling the rapid generation of personalized detection schemes for specific pipelines, reducing the time and resource investment required for preliminary planning. Based on learning from a large number of actual cases, the generated detection schemes are more in line with the actual situation, improving the accuracy and reliability of fault location.
[0076] S20. Determine the detection points inside the pipeline according to the first detection scheme. Install sensors at each detection point as signal receiving points. Send ultrasonic pulses with a fixed power multiple times at a preset interval as the target transmission signal. Collect multiple groups of received signals from each sensor, and obtain the target received signal at each detection point through matched filtering of the multiple groups of received signals.
[0077] In this step, according to the first detection scheme generated in step S10, clarify the specific positions where detection points need to be set. Based on the recommended first detection scheme, key features of the pipeline can also be considered, such as elbows, tees, and diameter changes, as well as areas with a high incidence of historical faults. Ensure that the detection points are evenly distributed inside the pipeline, covering all parts where problems may exist, and consider potential obstacles in the signal propagation path.
[0078] When installing sensors, ultrasonic sensors suitable for the internal environment of the pipeline can be selected to ensure that they have sufficient sensitivity and stability and can work properly under harsh conditions. Firmly install the sensors at the predetermined detection points, and special brackets or adhesives can be used to ensure that the sensors are in close contact with the pipeline wall, reducing signal loss. Equip each sensor with a data transmission cable or wireless module for real-time acquisition and transmission of the received ultrasonic signals. Further, determine the ultrasonic pulse transmission signal. According to the pipeline material and the estimated signal attenuation, adjust parameters such as the frequency and power of the ultrasonic transmitter to ensure that the best penetration depth and reflection intensity can be obtained for each transmission. Trigger the ultrasonic pulse transmission at a preset time interval through the control system, such as every few seconds or minutes, to ensure that the signal covers the entire monitoring period. Keep the power of each transmission constant to avoid measurement errors caused by power fluctuations and ensure the comparability of data at different time points. When the transmitter sends ultrasonic pulses, all installed sensors start receiving echo signals simultaneously and digitize and store them in the local memory or upload them to the central database. To improve the signal-to-noise ratio, multiple (such as more than 10 times) signal acquisitions should be performed at each detection point to form multiple groups of received data sets.
[0079] To improve the quality of the received signal, in this embodiment, matched filtering processing is preferentially adopted. First, based on the known characteristics of the transmitted signal, an ideal received signal template is pre-designed as the basis for subsequent matched filtering. Then, the matched filtering algorithm is used to process each group of received signals, and the true reflected signal closest to the template is extracted by maximizing the correlation coefficient, effectively removing background noise and other interference factors. Finally, the target received signal at each detection point is obtained, which reflects the internal structure of the pipeline and possible fault information. In a preferred embodiment, the signals collected multiple times at a certain detection point can also be averaged as the target received signal at this detection point for subsequent fault location analysis process.
[0080] Therefore, in this embodiment, by performing matched filtering on multiple groups of received signals, the signal-to-noise ratio is significantly enhanced, enabling weak but important reflected signals to be clearly presented, thereby improving the accuracy of fault location. The strategy of multiple transmissions with a fixed power ensures the consistency of each measurement condition, reduces the influence of random factors, and makes the results more stable and reliable. By reasonably arranging multiple detection points and performing high-frequency signal acquisition, a full-range scan of the pipeline interior is achieved without missing any potential problem areas. The entire process is highly automated, and from sensor deployment to signal acquisition and then to data analysis, it is all automatically completed by the system, greatly reducing the manual operation burden and improving work efficiency.
[0081] S30. Determine the time delay and signal attenuation coefficient at each detection point according to the target transmitted signal and the target received signal, establish a target optimization function based on the time delay and signal attenuation coefficient at each detection point, and obtain the fault location by optimizing and solving the target optimization function.
[0082] Specifically, in this step, determining the time delay and signal attenuation coefficient at each detection point according to the target transmitted signal and the target received signal includes:
[0083] Δt i =t i -t0;
[0084] In the formula, Δt i is the time delay at the i-th detection point, t i is the time when the sensor at the i-th detection point receives the signal, and t0 is the signal transmission time;
[0085]
[0086] In the formula, α i is the signal attenuation coefficient at the i-th detection point, I i is the intensity of the target received signal at the i-th detection point, I0 is the intensity of the target transmitted signal, and d i is the distance from the transmission point to the i-th detection point.
[0087] It should be noted that according to the propagation characteristics of ultrasonic waves in a medium, the time delay from the signal emission point to the reception point is proportional to the actual distance between the two points. In an ideal situation, if there are no obstacles or fault points, the time delay at each detection point should conform to the above relationship. However, when there is a fault inside the pipeline, the fault point will change the signal path, resulting in a discrepancy between the actually received time delay and the theoretical value. Therefore, by measuring the time difference of the signal reaching each detection point, the position of the obstacle or defect on the signal propagation path can be inferred. Since the propagation speed of ultrasonic waves in different media is known, the spatial position of the fault point can be deduced based on the time delay.
[0088] Ultrasonic waves will attenuate during propagation, and the degree of attenuation depends on the inherent properties of the material, such as density, elastic modulus, and physical obstacles on the path. For an intact pipeline, the signal attenuation coefficient is known. However, when cracks, corrosion, or other abnormal conditions occur inside the pipeline, the material properties in the local area change, resulting in additional attenuation, making the actually measured attenuation coefficient different from the theoretical value. Therefore, by using the variation law of signal intensity with distance and combining the inherent attenuation characteristics of the material, the fault point can be further verified and accurately located. When there is a significant difference between the actual attenuation coefficient and the theoretical value, it often means that there may be an abnormal situation at that place. This embodiment aims to combine the two key indicators of time delay and signal attenuation to construct a comprehensive objective optimization function, ensuring that both the time characteristics and energy loss characteristics of signal propagation are considered simultaneously to improve the positioning accuracy.
[0089] Furthermore, establishing the objective optimization function based on the time delay and signal attenuation coefficient of each detection point and obtaining the fault position by optimizing and solving the objective optimization function includes:
[0090] Taking the starting point of the pipeline as the origin to construct a three-dimensional rectangular coordinate system, and setting (x i , y i , z i ) as the position coordinates of the i-th detection point, and c as the propagation speed of ultrasonic waves in the medium, then the objective optimization function is:
[0091]
[0092] In the formula, F(x, y, z) is the objective optimization function, w1 and w2 are the weight coefficients of time delay and attenuation rate respectively, α0 is the theoretical attenuation coefficient of the water supply and drainage pipe to be measured, and N represents the number of detection points;
[0093] Using the gradient descent algorithm to solve the coordinates (x r , y r , z r), the fault location is obtained.
[0094] In the above formula, the first term represents the error between the actual time delay and the theoretical time delay. When the fault point is located at (x, y, z), this error should be as small as possible because the calculated distance is closest to the actual situation at this time. The second term (α i -α0) 2 represents the difference between the actual attenuation coefficient and the theoretical attenuation coefficient. The presence of a fault point will cause an increase in local attenuation, so this term also reflects the possibility of a fault. w1 and w2 are used to adjust the influence ratios of the time delay and the attenuation rate on the total error respectively, ensuring that the two are balanced within a reasonable range.
[0095] Furthermore, it is necessary to solve the coordinates (x r , y r , z r ) that minimize F(x, y, z) to find the most likely fault location. Because when F(x, y, z) reaches the minimum value, it means that the time delays and attenuation coefficients of all detection points are as close as possible to the theoretical expected values, indicating that the selected location can best explain the observed data. The minimized F(x, y, z) corresponds to a location that can simultaneously satisfy the time and energy propagation laws, that is, the location where the fault point is located. Because near the fault point, the propagation path and attenuation characteristics of the signal will change significantly, deviating from the normal situation. Finally, with the help of numerical optimization techniques such as gradient descent, search for the point that minimizes the objective function in three-dimensional space, which is the most likely fault location. These algorithms have good convergence and robustness, can effectively handle complex non-linear problems, find the global optimal solution from a complex environment, and have high computational efficiency.
[0096] Therefore, through the above objective function and solution process, combined with the two key indicators of time delay and signal attenuation, comprehensively evaluate the possibility of each potential location, and determine the most likely fault point location through mathematical optimization means. This method not only improves the positioning accuracy but also enhances the adaptability and reliability of the system.
[0097] S40. Calculate the deviation distance between the fault location and the detection points, optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance, and obtain the second detection scheme; update the case library according to the second detection scheme.
[0098] Specifically, the calculating the deviation distance between the fault location and the detection points, optimizing the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance, and obtaining the second detection scheme includes:
[0099] Calculate the deviation distance d bias :
[0100]
[0101] Define the objective function f(X), where X represents the position vector of the detection point, to minimize the deviation distance d bias Construct the objective function:
[0102] f(X) = -d bias ;
[0103] Construct a Gaussian process model, initialize the kernel function and hyperparameters of the model, select a preset number of detection point samples as the data set, and use the data set to train the Gaussian process model to construct a prior distribution;
[0104] Determine the acquisition function and find a new detection point X according to the acquisition function new , and add X new to the data set to update the posterior distribution of the Gaussian process model; Iteratively loop through the steps of updating the posterior distribution until the model converges to obtain the optimal solution and the optimized deviation distance d bias-new ;
[0105] Adjust the position of the detection point in the first detection scheme according to the optimal solution, and update the number of detection points according to the optimized deviation distance d bias-new to generate a second detection scheme;
[0106] Among them, updating the number of detection points according to the optimized deviation distance d bias-new includes:
[0107]
[0108] In the formula, N represents the number of detection points before updating, and N new represents the number of detection points after updating, and k represents the adjustment coefficient, with a value range of 0.1 to 0.5.
[0109] In this embodiment, first calculate the deviation distance according to the determined fault position (x r , y r , z r ) and the position of each detection point (x i , y i , z i ). With the goal of minimizing the deviation distance, define the objective function f(X) = -d bias, the negative sign here is to convert the maximization problem into a minimization problem because Bayesian optimization usually seeks the maximum value. Further, a Gaussian process model is constructed, an appropriate kernel function such as the squared exponential kernel and hyperparameters are selected, and the Gaussian process model is initialized. A preset number of detection point samples are selected as the initial dataset for training the Gaussian process model to establish a prior distribution. The acquisition function and acquisition strategy are determined. Commonly used acquisition functions include expected improvement, upper confidence bound, etc. These functions help determine the next most promising detection point. Then the newly selected detection point is added to the dataset, the Gaussian process model is retrained, and its posterior distribution is updated. Repeat the above acquisition and update steps until the model converges or reaches the predetermined maximum number of iterations, and finally obtain the optimal solution and the optimized deviation distance. Adjust the specific positions of the detection points in the first detection plan according to the optimization results, and update the number of detection points based on the optimized deviation distance using the formula of N new . Finally, the optimized second detection plan and its related data are fed back to the case library to keep the case library dynamically updated for future reference.
[0110] Therefore, by accurately calculating the deviation distance and combining with the Bayesian optimization algorithm, the fault point can be more accurately located, reducing the possibility of misjudgment. Dynamically adjusting the positions and numbers of the detection points avoids unnecessary over-detection, effectively saving labor and material costs and improving resource utilization. Introducing the Bayesian optimization algorithm enables the system to have the ability of self-learning and evolution, and can continuously improve the detection strategy in practice to adapt to the demand changes in different environments. Through continuous optimization of the second detection plan, the self-learning iteration of the detection plan is realized, and the case library data is enriched to ensure long-term stable high-performance detection.
[0111] See Figure 2 , in one embodiment, the present invention also provides a fault location system for a water supply and drainage pipe network, the system includes:
[0112] A detection plan clustering unit 100, configured to collect parameter information of a water supply and drainage pipe to be measured, perform clustering analysis on the parameter information and a case library to obtain a first detection plan for the water supply and drainage pipe to be measured; the parameter information includes the length, diameter and material properties of the water supply and drainage pipe to be measured; the case library is constructed according to historical fault location data of different water supply and drainage pipes;
[0113] A detection point layout unit 200, configured to determine detection points in the pipeline according to the first detection plan, arrange sensors at each detection point as signal receiving points, send ultrasonic pulses with a fixed power multiple times at a preset interval as target emission signals, collect multiple groups of received signals of each sensor, and obtain the target received signal of each detection point by performing matched filtering on the multiple groups of received signals;
[0114] The fault location unit 300 is configured to determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal, establish a target optimization function based on the time delay and signal attenuation coefficient of each detection point, and obtain the fault location by optimizing and solving the target optimization function;
[0115] The detection scheme update unit 400 is configured to calculate the deviation distance between the fault location and the detection point, optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance to obtain a second detection scheme; update the case base according to the second detection scheme.
[0116] In one embodiment, the fault location unit 300 is configured to determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal, including:
[0117] Δt i = t i - t0;
[0118] In the formula, Δt i is the time delay of the i-th detection point, t i is the time when the sensor at the i-th detection point receives the signal, and t0 is the signal transmission time;
[0119]
[0120] In the formula, α i is the signal attenuation coefficient of the i-th detection point, I i is the target reception signal strength at the i-th detection point, I0 is the target transmission signal strength, and d i is the distance from the emission point to the i-th detection point.
[0121] In one embodiment, the fault location unit 300 is configured to establish a target optimization function based on the time delay and signal attenuation coefficient of each detection point, and obtain the fault location by optimizing and solving the target optimization function, including:
[0122] Construct a three-dimensional rectangular coordinate system with the starting point of the pipeline as the origin. Let (x i , y i , z i ) be the position coordinates of the i-th detection point, and c be the propagation speed of ultrasonic waves in the medium. Then the target optimization function is:
[0123]
[0124] In the formula, F(x, y, z) is the target optimization function, w1 and w2 are the weight coefficients of the time delay and the attenuation rate respectively, α0 is the theoretical attenuation coefficient of the water supply and drainage pipe to be measured, and N represents the number of detection points;
[0125] Use the gradient descent algorithm to solve for the coordinates (x r , y r , z r ) that minimize F(x, y, z) to obtain the fault location.
[0126] In one embodiment, the detection scheme update unit 400 is configured to calculate the deviation distance between the fault location and the detection points, and optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance to obtain a second detection scheme, including:
[0127] Calculate the deviation distance d between the fault location and the detection points bias :
[0128]
[0129] Define the objective function f(X), where X represents the position vector of the detection points, to minimize the deviation distance d bias Construct the objective function:
[0130] f(X) = -d bias ;
[0131] Construct a Gaussian process model, initialize the kernel function and hyperparameters of the model, select a preset number of detection point samples as the data set, and use the data set to train the Gaussian process model to construct a prior distribution;
[0132] Determine the acquisition function, and find a new detection point X new , and add X new to the data set to update the posterior distribution of the Gaussian process model; Iteratively loop through the steps of updating the posterior distribution until the model converges to obtain the optimal solution and the optimized deviation distance d bias-new ;
[0133] Adjust the detection point positions in the first detection scheme according to the optimal solution, and update the number of detection points according to the optimized deviation distance d bias-new to generate a second detection scheme;
[0134] Wherein, updating the number of detection points according to the optimized deviation distance d bias-new includes:
[0135]
[0136] In the formula, N represents the number of detection points before update, N new represents the number of detection points after update, and k represents the adjustment coefficient, and its value range is 0.1 to 0.5.
[0137] It can be understood that the functions or modules included in the system provided in this embodiment can be used to execute the methods described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments. For the sake of brevity, it will not be repeated here.
[0138] The present invention also provides an electronic device, including a processor and a memory. The memory is used to store computer program code, and the computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes the method in any of the above possible implementation manners.
[0139] The present invention also provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by the processor of an electronic device, the processor is caused to execute the method in any of the above possible implementation manners.
[0140] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.
[0141] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, device, and unit can refer to the corresponding processes in the foregoing method embodiments and will not be repeated here. Those skilled in the art can also clearly understand that each embodiment of the present invention has different emphases in description. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, the parts not described or not described in detail in a certain embodiment can be referred to the records of other embodiments.
Claims
1. A method for locating faults in a water supply and drainage network, characterized in that: The method comprises: Collect parameter information of the water supply and drainage pipe to be tested, perform cluster analysis on the parameter information and the case library, and obtain a first detection scheme for the water supply and drainage pipe to be tested; the parameter information includes the length, diameter and material properties of the water supply and drainage pipe to be tested; the case library is constructed based on historical fault location data of different water supply and drainage pipes; Determine the detection points in the pipeline according to the first detection scheme, arrange sensors at each detection point as signal receiving points, send ultrasonic pulses with fixed power as target transmission signals multiple times at preset intervals, collect multiple groups of receiving signals from each sensor, and obtain the target receiving signal of each detection point by matching filtering the multiple groups of receiving signals; Determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal, establish a target optimization function according to the time delay and signal attenuation coefficient of each detection point, and obtain the fault location by optimizing and solving the target optimization function; The deviation distance between the fault location and the detection point is calculated, and the detection point information in the first detection scheme is optimized according to the Bayesian optimization algorithm and the deviation distance to obtain the second detection scheme; and the case library is updated according to the second detection scheme.
2. The method for locating faults in a water supply and drainage network according to claim 1, characterized in that: The step of determining the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal comprises: Δt i =t i -t0; In the formula, Δt i is the time delay of the ith detection point, t i is the time when the sensor at the i-th detection point receives the signal, and t0 is the signal transmission time; In the formula, α i is the signal attenuation coefficient of the i-th detection point, I i is the target receiving signal strength at the i-th detection point, I0 is the target transmitting signal strength, d i is the distance from the emission point to the i-th detection point.
3. The method for locating faults in a water supply and drainage network according to claim 2, characterized in that: The target optimization function is established according to the time delay and signal attenuation coefficient of each detection point, and the fault location is obtained by optimizing and solving the target optimization function, including: Construct a three-dimensional rectangular coordinate system with the starting point of the pipeline as the origin. Let (x i ,y i ,z i ) is the position coordinate of the i-th detection point, c is the propagation speed of ultrasonic wave in the medium, then the objective optimization function is: Where F(x,y,z) is the target optimization function, w1 and w2 are the weight coefficients of time delay and attenuation rate, α0 is the theoretical attenuation coefficient of the water supply and drainage pipe to be tested, and N represents the number of detection points; Use the gradient descent algorithm to find the coordinates (x) that minimize F(x,y,z) r ,y r ,z r ) to get the fault location.
4. The method for locating faults in a water supply and drainage network according to claim 3, characterized in that: The calculating of the deviation distance between the fault position and the detection point, optimizing the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance, and obtaining the second detection scheme includes: Calculate the deviation distance d between the fault location and the detection point bias : Define the objective function f(X), where X represents the position vector of the detection point to minimize the deviation distance d bias Construct the objective function: f(X)=-d bias ; Construct a Gaussian process model, initialize the kernel function and hyperparameters of the model, select a preset number of detection point samples as a data set, and use the data set to train the Gaussian process model to construct a prior distribution; Determine the acquisition function and find a new detection point X based on the acquisition function new , X new Add to the data set to update the posterior distribution of the Gaussian process model; iterate the steps of updating the posterior distribution until the model converges to obtain the optimal solution and the optimized deviation distance d bias-new ; According to the optimal solution, the detection point position in the first detection scheme is adjusted, and the deviation distance d after optimization is calculated. bias-new Update the number of detection points and generate a second detection plan; Among them, according to the optimized deviation distance d bias-new Update the number of detection points including: In the formula, N represents the number of detection points before updating, N new It indicates the number of updated detection points, k indicates the adjustment coefficient, and its value range is 0.1~0.
5.
5. A water supply and drainage network fault location system, characterized in that: The system comprises: A detection scheme clustering unit is used to collect parameter information of the water supply and drainage pipe to be tested, and cluster the parameter information with the case library to obtain a first detection scheme for the water supply and drainage pipe to be tested; the parameter information includes the length, diameter and material properties of the water supply and drainage pipe to be tested; the case library is constructed based on historical fault location data of different water supply and drainage pipes; A detection point arrangement unit is used to determine the detection points in the pipeline according to the first detection scheme, arrange a sensor at each detection point as a signal receiving point, send ultrasonic pulses with fixed power as target transmission signals for multiple times at preset intervals, collect multiple groups of receiving signals from each sensor, and obtain the target receiving signal of each detection point by matching and filtering the multiple groups of receiving signals; A fault location positioning unit is used to determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal, establish a target optimization function according to the time delay and signal attenuation coefficient of each detection point, and obtain the fault location by optimizing and solving the target optimization function; The detection scheme updating unit is used to calculate the deviation distance between the fault location and the detection point, optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance, and obtain the second detection scheme; and update the case library according to the second detection scheme.
6. The water supply and drainage network fault location system according to claim 5, characterized in that: The fault location positioning unit is used to determine the time delay and signal attenuation coefficient of each detection point according to the target transmission signal and the target reception signal, including: Δt i =t i -t0; In the formula, Δt i is the time delay of the ith detection point, t i is the time when the sensor at the i-th detection point receives the signal, and t0 is the signal transmission time; In the formula, α i is the signal attenuation coefficient of the i-th detection point, I i is the target receiving signal strength at the i-th detection point, I0 is the target transmitting signal strength, d i is the distance from the emission point to the i-th detection point.
7. The water supply and drainage network fault location system according to claim 6, characterized in that: The fault location positioning unit is used to establish a target optimization function according to the time delay and signal attenuation coefficient of each detection point, and obtain the fault location by optimizing and solving the target optimization function, including: Construct a three-dimensional rectangular coordinate system with the starting point of the pipeline as the origin. Let (x i ,y i ,z i ) is the position coordinate of the i-th detection point, c is the propagation speed of ultrasonic wave in the medium, then the objective optimization function is: Where F(x,y,z) is the target optimization function, w1 and w2 are the weight coefficients of time delay and attenuation rate, α0 is the theoretical attenuation coefficient of the water supply and drainage pipe to be tested, and N represents the number of detection points; Use the gradient descent algorithm to find the coordinates (x) that minimize F(x,y,z) r ,y r ,z r ) to get the fault location.
8. The water supply and drainage network fault location system according to claim 7, characterized in that: The detection scheme updating unit is used to calculate the deviation distance between the fault position and the detection point, optimize the detection point information in the first detection scheme according to the Bayesian optimization algorithm and the deviation distance, and obtain the second detection scheme, including: Calculate the deviation distance d between the fault location and the detection point bias : Define the objective function f(X), where X represents the position vector of the detection point to minimize the deviation distance d bias Construct the objective function: f(X)=-d bias ; Construct a Gaussian process model, initialize the kernel function and hyperparameters of the model, select a preset number of detection point samples as a data set, and use the data set to train the Gaussian process model to construct a prior distribution; Determine the acquisition function and find a new detection point X based on the acquisition function new , X new Add to the data set to update the posterior distribution of the Gaussian process model; iterate the steps of updating the posterior distribution until the model converges to obtain the optimal solution and the optimized deviation distance d bias-new ; According to the optimal solution, the detection point position in the first detection scheme is adjusted, and the deviation distance d after optimization is calculated. bias-new Update the number of detection points and generate a second detection plan; Among them, according to the optimized deviation distance d bias-new Update the number of detection points including: In the formula, N represents the number of detection points before updating, N new It indicates the number of updated detection points, k indicates the adjustment coefficient, and its value range is 0.1~0.
5.
9. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store computer program codes, wherein the computer program codes include computer instructions, and when the processor executes the computer instructions, the electronic device executes the water supply and drainage network fault locating method according to any one of claims 1 to 4.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the water supply and drainage network fault locating method according to any one of claims 1 to 4.