Dimension reduction PSO-ITA long-distance water delivery pipeline burst positioning method

Abstract

The invention discloses a dimensionality reduction PSO-ITA long-distance water pipe burst positioning method which comprises the following steps: taking a pipe burst point as a node to obtain N pipeline nodes; according to the liquid control equation of the long-distance water conveying pipeline system, calculating the pressure and the flow of a pipeline node adjacent to the node to be measured by using a characteristic line method to obtain the target pressure of the node to be measured; if any to-be-measured node of the long-distance water conveying pipeline has pipe explosion relative to the pipeline, determining target pressure of the to-be-measured node according to a pre-determined liquid control equation of the long-distance water conveying pipeline system, a pipe explosion orifice outflow equation, a fitness function, a pipe explosion position and occurrence moment coupling equation and a particle swarm optimization algorithm; and according to a particle swarm optimization algorithm and a fitness function, solving a final pipe explosion position, size and occurrence time. According to the method, for the pipe explosion problem occurring in a long-distance water conveying pipeline system, the dimensionality reduction PSO-ITA pipe explosion positioning model is used for accurate positioning, and the pipe explosion transient flow numerical simulation precision can be improved.

Description

Technical Field

[0001] This invention relates to the field of hydraulic numerical simulation calculation, specifically to a method for locating burst pipes in long-distance water conveyance pipelines using a dimension-reduced PSO-ITA method. Background Technology

[0002] In response to the water hammer phenomenon that frequently occurs in water pipeline systems, which can lead to sudden pipe bursts, existing pipe burst detection and location methods suffer from low positioning accuracy.

[0003] Traditional inverse transient burst pipe location methods assume that the burst time is known, use numerical simulation data as measured values, and employ optimization algorithms to simultaneously solve for the burst pipe location and flow rate, achieving a location accuracy of 90%. However, in reality, the burst time is unknown and must be considered in the objective function.

[0004] Traditional inverse transient burst pipe location methods cannot determine the location, size, and timing of burst pipes in a timely manner, and do not perform burst pipe location under actual conditions. The robustness of the burst pipe location model needs to be verified.

[0005] Pipe burst location based on inverse transient analysis requires an accurate transient flow model of the burst. However, previous transient hydraulic models of burst pipes often underestimated the dynamic frictional effect of water hammer, and there is still a large error between the simulation results of these transient hydraulic models and the actual situation.

[0006] In long-distance water conveyance systems, a sound optimization objective and algorithm are prerequisites for using inverse transient analysis (IRA) to locate pipe bursts. An existing ITA-based pipe burst location method assumes the burst time is known and, based on simulated data, uses an optimization algorithm to simultaneously solve for the burst location and flow rate. However, the location results deviate somewhat from the actual results. In reality, the burst time is unknown and must be considered in the objective function. Furthermore, to apply the pipe burst location method in practical engineering, measured burst data should be used to verify its accuracy. Summary of the Invention

[0007] To address the problem of locating burst pipes in long-distance water transmission systems and the limitations of existing inverse transient analysis methods, this invention provides a dimension-reduced PSO-ITA method for locating burst pipes in long-distance water transmission pipelines.

[0008] A method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA method, with the following specific steps: Based on the burst point as the node, the long-distance water transmission pipeline system is divided into left and right pipelines. The two pipelines are then divided into a grid to obtain N pipeline nodes. Considering dynamic friction, the pressure and flow rate of the pipe nodes adjacent to the node to be measured are calculated using the method of characteristics based on the liquid control equation of the long-distance water transmission pipeline system, so as to obtain the target pressure of the node to be measured. If a pipe bursts at any node to be tested in a long-distance water pipeline, the target pressure of the node to be tested is determined based on the pre-determined liquid control equations of the long-distance water pipeline system, the outflow equation of the burst orifice, the fitness function, the coupled equations of the burst location and the time of occurrence, and the particle swarm optimization algorithm. Then, the final burst location, size, and time of occurrence are solved using the particle swarm optimization algorithm and the fitness function.

[0009] Specific steps: S1. Based on the burst point, the long-distance water transmission pipeline system is divided into left and right pipelines. The pipelines on both sides are then divided into grids to obtain multiple nodes to be measured in the pipeline, and the liquid control equations of the long-distance water transmission pipeline system are constructed. S2. Based on the liquid control equation, construct a transient flow model of the bursting tube coupled with the method of characteristics and dynamic friction. S3. Construct the fitness function, initialize the population size and iteration count Nmax, and construct the inverse transient burst pipe location model; S4. Determine the optimization parameters, taking into account the weights of the burst pipe location, size, and timing. S5. Based on time monitoring technology, the time of pipe burst and the location of pipe burst are coupled and solved to realize the transformation of three-variable optimization into two-variable optimization; S6. Based on the transient flow model of the bursting tube, the particle swarm optimization algorithm is used to solve the inverse transient bursting tube localization model in S3.

[0010] In S1, the liquid control equation for the long-distance water pipeline system is: In the formula: H The piezometric head is measured in meters (m). V The average flow velocity across the pipe cross-section is given in m / s. g The acceleration due to gravity is m / s². 2 ; a Wave speed, m / s; x The length along the pipe axis is in meters (m). t The propagation time is s; J = J s + J u The head loss per unit fluid caused by friction, where constant friction... J s= fV|V| / 2gD , D The diameter of the pipe is in meters (m). f The friction coefficient of the Darcy-Wiesbach pipe; dynamic friction resistance.J u Its composition is: In the formula: ν m is the kinematic viscosity of the fluid. 2 / s; y i It is a weighted function; To calculate the time step, s; m i and n i These are the coefficients in the weighting function that depend on the fluid flow regime; According to the method of characteristics, the hyperbolic partial differential equation system is transformed into an ordinary differential equation system, which gives: Integrating the system of ordinary differential equations along the positive and negative characteristic lines respectively, we get: In the formula, H P , H B , H M The pressure head at points P, B, and M are respectively; Q P , Q B , Q M The cross-sectional flow rates at points P, B, and M are respectively; A The pipe area; η These are integral approximate control coefficients, typically η = 0.5 ~ 1, when η When = 0.5, it is a semi-explicit solution; when η When = 1.0, it is an implicit solution; make B = a / ( gA ), The equation of the characteristic line can be further expressed as: In the formula, The solution to the equation of the characteristic line is: Combining the method of characteristics described above, a transient flow model for bursting pipes is further constructed: In the formula: C d A g This can represent the outflow coefficient of a burst pipe. C d The flow coefficient at the pipe rupture orifice is... A g The area of ​​the burst pipe orifice; The pressure at the pipe burst point during the transient process. ,in H B For the piezometer head at the point of pipe rupture in the system, Z B Elevation of the burst pipe point. D B The diameter of the burst pipe orifice. Q 1. Q 2 represents the flow rates corresponding to the positive and negative characteristic lines at the pipe burst node, respectively; The aforementioned outflow equation from the burst tube orifice refers to the equation in the transient flow model of the burst tube: The transient flow model of a bursting pipe is an extended application based on the outflow equation at the orifice of a bursting pipe.

[0011] Solving the characteristic line equations simultaneously, the transient head at the pipe rupture point. H B and traffic Q B for: in: .

[0012] In S3, an inverse transient burst pipe location model is constructed, with the fitness function being: In the formula: n is the number of data points collected by the high-frequency pressure sensor; H i This is the pressure sensor reading, in meters (m). H i * This is the calculated value, m, from the transient flow model of the bursting tube. X B The location of the burst pipe is in meters (m). A g The burst pipe area is m 2 ; T B The time of the pipe burst is s; LThe total length of the pipeline is in meters (m). T The total time for calculating the model is in seconds.

[0013] In S4, the weighting coefficients for the optimization parameters are determined as follows: In optimization algorithms, different weights for the three variables—burst location, size, and occurrence time—result in different convergence speeds and algorithm stability. Appropriate variable weights can, to some extent, prevent the algorithm from getting trapped in local optima. Parameter sensitivity analysis is a method to evaluate the impact of the uncertainty of parameters (burst location, size, and occurrence time) on the output results. Therefore, this invention uses relative sensitivity analysis to determine the weights of these three parameters: In the formula: F The fitness function is defined as MSE. x n To affect the fitness function F Variables.

[0014] In S5, considering that simultaneously optimizing the three variables of pipe burst location, size, and timing is a three-dimensional search problem, the particle swarm optimization algorithm requires a large search space and dimension, and is prone to getting trapped in local optima. To address this issue, this invention considers that there is an inherent relationship between the timing and location of pipe bursts, that is, given the wave velocity in the water pipeline system, the timing of pipe bursts can be represented by the pipe burst location: In the formula: T B The moment when the pipe burst occurred. T PT1 The descent time point at the location of the PT1 sensor. X pt1 For the position of the sensor, X B This indicates the location of the burst pipe. The burst pressure data collected at the location of the PT1 sensor was detected using time inflection point detection technology to pinpoint the burst point of the burst tube. T PT1 Based on inflection point detection technology, the time of tube bursting can be directly calculated from the tube bursting position during each optimization iteration, realizing the transformation from three-variable optimization to two-variable optimization. This significantly reduces the search space and dimensionality required by the particle swarm optimization algorithm, and to a certain extent avoids the optimization algorithm from getting trapped in local optima, thereby improving its computational efficiency and accuracy.

[0015] In S6, the particle swarm optimization algorithm: Particle Swarm Optimization (PSO) is a swarm intelligence optimization algorithm. Its core principle is to update the position and velocity of particles by simulating the social behavior of flocks of birds or schools of fish. In each iteration, the particles update their positions and velocities based on their individual optimal behavior. pbest ) and group optimal ( gbest Update its speed: In the formula: v i ( t ) is a particle i In time t speed; x i ( t ) is a particle i In time t Location, W Inertial weights; c 1. c 2 is the acceleration constant; r 1. r 2 is a random number in the range [0,1]. pbest i For particles i The historical best position; gbest This represents the historical best position for the entire group. The particles move according to the updated velocity:

[0016] To improve the convergence of particle swarm optimization, its inertia weights... w It can decrease linearly over time: In the formula: w max , w min For initial and final weights; N max This represents the maximum number of iterations.

[0017] In S1, the mesh is generated as follows: In the dimension-reduced PSO-ITA model for locating burst pipes in long-distance water pipelines, the computational grid length is... =1.2m, if the assumed distance between the burst point and the upstream pressure tank is less than 1.2m, then the length of the burst point from the upstream pressure tank ( X B ) as the computational grid length This model calculates the grid length by inputting the grid length. It can automatically generate a grid for calculation: In the formula: To calculate the grid time interval, To calculate the grid length, Let be the wave velocity of the liquid inside the pipe, and CR be the Coulomb number. NP represents the total calculated length of the L-shaped pipe, NINT is the floor function, and NP is the number of grids required for the calculation.

[0018] Beneficial effects: Compared with existing technologies, this invention provides a dimensionality-reduced PSO-ITA method for locating burst pipes in long-distance water pipelines. By fully considering the attenuation effect of dynamic friction, a transient flow model coupling the characteristic line method and dynamic friction is constructed. The model's calculation results are compared and analyzed with experimental data to verify its accuracy. Then, the burst occurrence time is considered in the objective function, and the relationship between the burst occurrence time and the burst location is fully analyzed. A dimensionality-reduced PSO-ITA burst pipe location model is proposed, transforming the optimization from three variables to two variables, significantly reducing the search space and dimensionality required by the particle swarm optimization algorithm. Compared with ordinary burst pipe location models, the dimensionality-reduced PSO-ITA burst pipe location model achieves a significant improvement in positioning accuracy and computational efficiency. Attached Figure Description

[0019] Figure 1 This is a flowchart of a method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA method according to the present invention; Figure 2 This is a schematic diagram of the tube bursting test apparatus; Figure 3 This is a comparison of the transient flow model of the burst tube and the experimental pressure fluctuation curve; Figure 4 This is a comparison chart of the calculation accuracy and calculation time of the dimension-reduced PSO-ITA burst pipe model (burst pipe location error). Figure 5 This is a comparison chart of the calculation accuracy and calculation time of the dimension-reduced PSO-ITA burst pipe model (burst pipe size positioning error). Figure 6 This is a comparison chart of the calculation accuracy and calculation time of the dimension-reduced PSO-ITA burst pipe model (error in the time of burst pipe occurrence). Figure 7 This is a comparison chart of the calculation accuracy and calculation time of the dimension-reduced PSO-ITA burst pipe model (comparison of burst pipe location calculation time). Detailed Implementation

[0020] The principles and technical advantages of the present invention will be further explained below with reference to specific implementation examples and corresponding drawings. It should be noted that these examples are only used to illustrate the function of the present invention and are not intended to limit the scope of application of the present invention. After reading the relevant principles and functions of the present invention, any modifications to the present invention in various equivalent forms should be within the scope limited by the claims appended to this application.

[0021] This invention provides a method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA method, such as... Figure 1 The flowchart shown indicates that the method may include: Step S1: Using the burst point as a node, divide the long-distance water pipeline system into left and right side pipes. Then, mesh the two side pipes to obtain multiple nodes to be tested, specifically including: Using the burst point as a node, the long-distance water transmission pipeline system is divided into left and right pipelines, spaced according to a predetermined calculated length. The invention performs mesh generation for long-distance water pipelines. The computational mesh length of the transient flow model for pipe rupture in this invention is [value missing]. =1.2m, if the assumed distance between the burst point and the upstream pressure tank is less than 1.2m, then the length of the burst point from the upstream pressure tank ( X B ) as the computational grid length .

[0022] Step S2 involves constructing a transient flow model of a bursting tube coupled with the method of characteristics and dynamic friction, specifically including: The transient flow model for pipe bursts considers the governing equations of long-distance water pipeline systems, the outflow equation from the burst orifice, and the continuity equation. Furthermore, it incorporates dynamic friction terms and uses the method of characteristics to solve these equations, obtaining the target pressure at the pipe node under test. The transient flow model for pipe bursts, coupled with the method of characteristics and dynamic friction, predicts experimental values ​​in the time domain. The experimental values ​​are then compared with measured values ​​to verify the accuracy of the transient flow model, providing a theoretical basis for pipe burst location using inverse transient analysis.

[0023] Transient head at pipe rupture H B for: Step S3 involves constructing the fitness function, initializing the population size and iteration count Nmax, and building an inverse transient burst pipe localization model, specifically including: To quantify the average deviation between simulated and measured values, a corresponding fitness function is proposed. An objective function is constructed that simultaneously includes the location, size, and time of pipe burst. The location, size, and time of pipe burst are set to be within a reasonable range. The population size is initialized to 14, and the maximum number of iterations of the particle swarm optimization algorithm is set to 20.

[0024] The fitness function is:

[0025] Step S4, determine the weights of the optimization parameters (burst location, size, and occurrence time), specifically including: Parameter sensitivity analysis is a method to evaluate the impact of uncertainties in parameters (burst location, size, and timing) on ​​the output results. Therefore, this invention uses relative sensitivity analysis to determine the weights of these three parameters and uses relative sensitivity analysis to calculate the burst location (…). X B ),size( A g ) and the time of occurrence ( T B The weights for ) are 1.237, 2.317 and 0.816, respectively.

[0026] Step S5, based on time monitoring technology, couples the time of pipe rupture with the location of pipe rupture, specifically including: Considering the inherent relationship between the timing and location of a pipe burst—that is, given the known wave velocity in a water pipeline system—the timing of the burst can be represented by its location. In the formula: T B The moment when the pipe burst occurred. T PT1 The descent time point at the location of the PT1 sensor. X pt1 For the position of the sensor, X B This indicates the location of the burst pipe.

[0027] To detect the burst pressure data collected at the PT1 sensor location, the burst point of the burst tube is determined. T PT1The method employs time inflection point detection technology. The detection process is as follows: First, wavelet filtering is applied to the high-frequency pressure data collected by the PT1 pressure sensor to remove noise. Then, pressure data drop points are detected by sequentially traversing the pressure data list in chronological order, searching for points where the pressure value decreases among multiple consecutive data points, and recording the index, number of drops, and magnitude of these points in a list. Finally, based on the point with the largest drop magnitude in the list, the index and number of drops of that point are found to determine the location of the pressure drop point.

[0028] Based on inflection point detection technology, the time of tube bursting can be directly calculated from the tube bursting position during each optimization iteration, realizing the transformation from three-variable optimization to two-variable optimization. This significantly reduces the search space and dimensionality required by the particle swarm optimization algorithm, and to a certain extent avoids the optimization algorithm from getting trapped in local optima, thereby improving its computational efficiency and accuracy.

[0029] Step S6: Solve the inverse transient burst pipe location model using the particle swarm optimization algorithm, specifically including: Particle Swarm Optimization (PSO) is a swarm intelligence optimization algorithm that updates particle positions and velocities by simulating the social behavior of flocks of birds or schools of fish. It achieves this through individual optimality (…). pbest ) and group optimal ( gbest With the dual guidance of particle swarm optimization (PSO) and the ability to rapidly converge towards high-quality solutions, the particles exhibit efficient convergence, especially in the early stages of optimization. The fitness function is solved using PSO, and after the algorithm iterates to its maximum number of iterations (Nmax), the location, size, and time of the burst pipe are obtained. The optimized burst pipe location, size, and time are compared with the actual burst pipe location, size, and time to calculate the error of the PSO-ITA burst pipe localization model.

[0030] Example 1 Reference Figures 2-3 This embodiment provides a numerical simulation example of the transient flow model of a bursting pipe. A simplified diagram of the experimental setup is shown below. Figure 2 As shown, this test system mainly consists of a water supply pipeline system and a measurement system.

[0031] The water supply pipeline system mainly includes: a pressure-stabilized water supply tank, pipeline valves, steel pipes, and an underground reservoir. In this test system, the pressure tank was selected as the upstream boundary. The pressure tank has a design pressure of 1.6 MPa and a volume of 2.2 m³. A stop valve was used as the downstream blind end boundary, and a side valve (DN15 manual ball valve) was used to simulate a pipe burst with high-frequency transient pressure drop. The simulated burst points were 84.49 m and 152.98 m from the upstream pressure tank, respectively. The length of the steel pipe used in the test was... L =241.52 m, pipe diameter D=0.05 m, pipe wall thickness is 3.5 mm, average pipe slope is 0.4°, and steady-state friction coefficient is 0.022. The measurement system mainly includes one electromagnetic flowmeter and two pressure sensors. The experiment uses the electromagnetic flowmeter to collect the steady-state flow rate of the pipe and the pressure sensors to collect the transient pressure. Specific locations are as follows: Figure 5 As shown. The electromagnetic flowmeter located near the pressure tank has an accuracy of 0.2% and a measuring range of 0.707~10m. 3 The main technical parameters of the high-precision pressure sensor are: accuracy 0.2%, measurement range 0~1.0MPa, and sampling frequency 2000Hz.

[0032] The pipe burst experiment was first conducted by pressurizing the pressure tank with a submersible pump to change the water head in the upstream manometer to reach the target pressure. Then, the target flow rate was achieved by adjusting the opening of the side valve, and the steady-state flow rate of the pipeline measured by the electromagnetic flowmeter was recorded. Next, the side valve was randomly and rapidly opened within a 5-second timeframe, with the opening time strictly controlled between 0.05 and 0.08 seconds, to simulate the transient wave generated by the pipe burst. The pressure signal was continuously collected by a pressure sensor for 10 seconds until the wave completely dissipated.

[0033] The calculation results of the transient flow model of the bursting tube are as follows: Figure 4 , 5 As shown in Figures 6 and 7, the experimental and model-calculated values ​​fit well, and the R² of the constructed burst pipe transient flow model is close to 1, indicating that the error between the calculated and experimental values ​​is small, thus verifying the accuracy of the burst pipe model. Therefore, the burst pipe transient flow model coupled with the method of characteristics and dynamic friction can predict experimental values ​​well in the time domain. Furthermore, inflection point detection can accurately locate the pressure drop inflection point.

[0034] Example 2 This embodiment provides a case study of the positioning effect of a dimensionally reduced PSO-ITA burst pipe positioning model.

[0035] The localization effect of the dimension-reduced PSO-ITA burst pipe localization model is as follows: Figure 4 , 5 As shown in Figures 6 and 7, based on the time detection technology, the time of tube bursting can be directly calculated from the tube bursting position during each optimization iteration, realizing the transformation from three-variable optimization to two-variable optimization. This significantly reduces the search space and dimension required by the particle swarm optimization algorithm, thereby avoiding the optimization algorithm from getting trapped in local optima and improving its computational efficiency.

[0036] The dimensionality-reduced PSO-ITA burst pipe location model has high computational accuracy and efficiency. The error range of the burst pipe location is 0.01%~1.87%, the error range of the burst pipe size is 0.06%~2.94%, the error range of the burst pipe occurrence time is 0.01%~0.16%, and the model calculation time ranges from 32.2s to 44.4s.

[0037] Compared to the standard PSO-ITA model, the dimensionality-reduced PSO-ITA burst pipe location model improves the accuracy of burst pipe location by 1.3%–13.3%, burst pipe size by 1.8%–26.0%, and burst occurrence time by 0.2%–0.6%, while also increasing computational efficiency by 198.3%–279.3%. These results demonstrate that the dimensionality-reduced PSO-ITA burst pipe location model can quickly locate bursts in long-distance water pipeline systems, even achieving high accuracy.

[0038] This embodiment also provides an electronic device suitable for a dimensionality-reduced PSO-ITA burst pipe location model, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the simulation location of the dimensionality-reduced PSO-ITA burst pipe location model as proposed in the above embodiment.

[0039] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the simulation and positioning method of the dimensionality-reduced PSO-ITA burst pipe positioning model proposed in the above embodiments.

[0040] The storage medium proposed in this embodiment and the dimensionality reduction PSO-ITA burst pipe location model simulation location method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0041] As can be seen from the results of the embodiments, in the field of pipe burst location of long-distance water transmission pipelines, compared with the current pipe burst location technology, the present invention, by fully considering the attenuation effect of dynamic friction, has a determination coefficient (R2) close to 1 for the transient flow model of pipe burst coupled with the characteristic line method and dynamic friction, indicating the accuracy of the transient flow model of pipe burst.

[0042] The PSO algorithm has high accuracy in solving the burst pipe location model. Based on time detection technology, a dimensionality-reduced PSO-ITA burst pipe location model is proposed, which realizes the transformation from three-variable optimization to two-variable optimization, significantly reducing the search space and dimensionality required by the PSO algorithm, and has high computational accuracy and efficiency.

[0043] Based on this, this invention specifically studies a numerical simulation method for transient flow in pipe bursts. For pipe bursts occurring in long-distance water transmission pipeline systems, a dimension-reduced PSO-ITA pipe burst location model is used for accurate location, which helps improve the accuracy of numerical simulation of transient flow in pipe bursts and improves the accuracy of pipe burst location in water transmission pipeline systems. Timely and accurate pipe burst location is of great significance for ensuring safe water supply and conserving water resources.

Claims

1. A method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA method, characterized in that... The specific steps are as follows: Based on the burst point as the node, the long-distance water transmission pipeline system is divided into left and right pipelines. The two pipelines are then divided into a grid to obtain N pipeline nodes. Considering dynamic friction, the pressure and flow rate of the pipe nodes adjacent to the node to be measured are calculated using the method of characteristics based on the liquid control equation of the long-distance water transmission pipeline system, so as to obtain the target pressure of the node to be measured. If a pipe bursts at any node to be tested in a long-distance water pipeline, the target pressure of the node to be tested is determined based on the pre-determined liquid control equations of the long-distance water pipeline system, the outflow equation of the burst orifice, the fitness function, the coupled equations of the burst location and the time of occurrence, and the particle swarm optimization algorithm. Then, the final burst location, size, and time of occurrence are solved using the particle swarm optimization algorithm and the fitness function.

2. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA approach as described in claim 1, characterized in that... S1. Based on the burst point, the long-distance water transmission pipeline system is divided into left and right pipelines. The pipelines on both sides are then divided into grids to obtain multiple nodes to be measured in the pipeline, and the liquid control equations of the long-distance water transmission pipeline system are constructed. S2. Based on the liquid control equation, construct a transient flow model of the bursting tube coupled with the method of characteristics and dynamic friction. S3. Construct the fitness function, initialize the population size and iteration count Nmax, and construct the inverse transient burst pipe location model; S4. Determine the optimization parameters, taking into account the weights of the burst pipe location, size, and timing. S5. Based on time monitoring technology, the time of pipe burst and the location of pipe burst are coupled and solved to realize the transformation of three-variable optimization into two-variable optimization; S6. Based on the transient flow model of the bursting tube, the particle swarm optimization algorithm is used to solve the inverse transient bursting tube localization model in S3.

3. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA approach as described in claim 2, characterized in that... In S1, the liquid control equation for the long-distance water pipeline system is: In the formula: H The piezometric head is measured in meters (m). V The average flow velocity across the pipe cross-section is given in m / s. g The acceleration due to gravity is m / s². 2 ; a Wave speed, m / s; x The length along the pipe axis is in meters (m). t The propagation time is s; J = J s + J u The head loss per unit fluid caused by friction, where constant friction... J s= fV|V| / 2gD , D The diameter of the pipe is in meters (m). f The friction coefficient of the Darcy-Wiesbach pipe; dynamic friction resistance. J u Its composition is: In the formula: ν m is the kinematic viscosity of the fluid. 2 / s; y i It is a weighted function; To calculate the time step, s; m i and n i These are the coefficients in the weighting function that depend on the fluid flow regime; According to the method of characteristics, the hyperbolic partial differential equation system is transformed into an ordinary differential equation system, which gives: Integrating the system of ordinary differential equations along the positive and negative characteristic lines respectively, we get: In the formula, H P , H B , H M The pressure head at points P, B, and M are respectively; Q P , Q B , Q M The cross-sectional flow rates at points P, B, and M are respectively; A The pipe area; η These are integral approximate control coefficients, typically η = 0.5 ~ 1, when η When = 0.5, it is a semi-explicit solution; when η When = 1.0, it is an implicit solution; make B = a / ( gA ), The equation of the characteristic line can be further expressed as: In the formula, The solution to the equation of the characteristic line is: Combining the method of characteristics described above, a transient flow model for bursting pipes is further constructed: In the formula: C d A g This can represent the outflow coefficient of a burst pipe. C d The flow coefficient at the pipe rupture orifice is... A g The area of ​​the burst pipe orifice; The pressure at the pipe burst point during the transient process. ,in H B For the piezometer head at the point of pipe rupture in the system, Z B The elevation of the burst pipe point. D B The diameter of the burst pipe orifice. Q 1. Q 2 represents the flow rates corresponding to the positive and negative characteristic lines at the pipe burst node, respectively; Solving the characteristic line equations simultaneously, the transient head at the pipe rupture point. H B and traffic Q B for: in:

4. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA approach as described in claim 1, characterized in that... In S3, an inverse transient burst pipe location model is constructed, with the fitness function being: In the formula: n is the number of data points collected by the high-frequency pressure sensor; H i This is the pressure sensor reading, in meters (m). H i * This is the calculated value, m, from the transient flow model of the bursting tube. X B The location of the burst pipe is in meters (m). A g The burst pipe area is m 2 ; T B The time of the pipe burst is s; L The total length of the pipeline is in meters (m). T The total time, in seconds, is calculated for the model.

5. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA as described in claim 1, characterized in that... In S4, the weighting coefficients for the optimization parameters are determined as follows: In the optimization algorithm, the weights of the three variables—burst location, size, and time of occurrence—are different, resulting in different convergence speeds and varying algorithm stability. Relative sensitivity analysis is used to determine the weights of these three parameters: In the formula: F The fitness function is defined as MSE. x n To affect the fitness function F Variables.

6. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA approach as described in claim 1, characterized in that... In S5, given the wave velocity in the water pipeline system, the moment of pipe bursting can be represented by the location of the burst: In the formula: T B The moment when the pipe burst occurred. T PT1 The descent time point at the location of the PT1 sensor. X pt1 For the position of the sensor, X B This indicates the location of the burst pipe. The burst pressure data collected at the location of the PT1 sensor was detected using time inflection point detection technology to pinpoint the burst point of the burst tube. T PT1 Based on inflection point detection technology, the time of pipe bursting is directly calculated from the location of pipe bursting during each optimization iteration.

7. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA approach as described in claim 1, characterized in that... In S6, the particle swarm optimization algorithm: In each iteration, the particle updates its velocity based on the individual optimality and the swarm optimality: In the formula: v i ( t ) is a particle i In time t speed; x i ( t ) is a particle i In time t Location, W Inertial weight; c 1. c 2 is the acceleration constant; r 1. r 2 is a random number in the range [0,1]. pbest i For particles i The historical best position; gbest This represents the historical best position for the entire group. The particles move according to the updated velocity: To improve the convergence of particle swarm optimization, its inertia weights... w It can decrease linearly over time: In the formula: w max , w min For initial and final weights; N max This represents the maximum number of iterations.

8. The method for locating burst pipes in long-distance water transmission pipelines using a dimension-reduced PSO-ITA approach as described in claim 2, characterized in that... In S1, the mesh is generated as follows: In the dimension-reduced PSO-ITA model for locating burst pipes in long-distance water pipelines, the computational grid length is... =1.2m, if the assumed distance between the burst point and the upstream pressure tank is less than 1.2m, then the length of the burst point from the upstream pressure tank ( X B ) as the computational grid length Calculate the grid length by inputting the grid length. Automatically generate a grid for calculation: In the formula: To calculate the grid time interval, To calculate the grid length, Let be the wave velocity of the liquid inside the pipe, and CR be the Coulomb number. NP represents the total calculated length of the L-shaped pipe, NINT is the floor function, and NP is the number of grids required for the calculation.

Images