Power distribution network traveling wave fault location method and device based on improved barrel theory optimization algorithm, computer equipment and storage medium

Abstract

The embodiment of the application relates to the field of power systems, and provides a power distribution network traveling wave fault positioning method and device based on an improved barrel theory optimization algorithm, computer equipment and a storage medium, the method comprising: determining a traveling wave energy attenuation matching item and a topological connectivity consistency item based on a basic physical relationship of traveling wave propagation of a target power distribution network to be positioned; constructing a double-dimension fused power distribution network optimization model conforming to traveling wave propagation characteristics and power distribution network topological constraints according to the traveling wave energy attenuation matching item and the topological connectivity consistency item; and solving the power distribution network optimization model based on the improved barrel theory optimization algorithm to obtain a power distribution network traveling wave fault positioning result associated with the target power distribution network. The implementation of the method improves the efficiency and accuracy of traveling wave fault positioning.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a method, apparatus, computer equipment, and storage medium for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm. Background Technology

[0002] In the field of power system technology, distribution network topologies are complex and operating environments are highly variable. After a fault occurs, quickly and accurately locating the fault point is crucial for shortening power outage time and improving power supply reliability.

[0003] However, in traditional techniques, single-end methods are prone to misjudgment or omission, making wavefront identification difficult, reducing positioning reliability, and easily causing positioning confusion; the shortcomings of dual-end methods are mainly reflected in time synchronization, parameter dependence, and application scenario limitations; the computational cost of multi-target particle swarm optimization methods is high, the uncertainty of fault handling is large, and it is easy to cause the search to get stuck in local optima or oscillation, making it difficult to guarantee the reliability of positioning results and resulting in low search efficiency.

[0004] In summary, traditional traveling wave fault location technology for distribution networks suffers from low efficiency and accuracy in the fault location process. Summary of the Invention

[0005] This application provides a method, apparatus, computer equipment, and storage medium for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm. More specifically, this application provides a method, apparatus, computer equipment, computer storage medium, and computer program product for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm, thereby improving the efficiency and accuracy of traveling wave fault location.

[0006] In a first aspect, embodiments of this application provide a method for locating traveling wave faults in a distribution network based on an improved barrel theory optimization algorithm, comprising:

[0007] Based on the fundamental physical relationship of traveling wave propagation in the target distribution network to be fault located, the traveling wave energy attenuation matching term and topology connectivity consistency term are determined.

[0008] Based on the traveling wave energy attenuation matching term and the topology connectivity consistency term, a two-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints is constructed.

[0009] Based on the improved barrel theory optimization algorithm, the distribution network optimization model is solved to obtain the distribution network traveling wave fault location results associated with the target distribution network.

[0010] Optionally, in some embodiments of this application, determining the traveling wave energy attenuation matching term and topology connectivity consistency term based on the fundamental physical relationship of the traveling wave propagation of the target distribution network to be fault located includes:

[0011] The degree of matching is quantified by summing relative errors to determine the traveling wave energy attenuation matching term;

[0012] By using the ratio of the actual line length to the straight-line distance, combined with topology validity verification, the degree of fit between the virtual fault point and the actual structure of the distribution network is quantified, and the topology connectivity consistency item is determined.

[0013] Optionally, in some embodiments of this application, the step of quantifying the matching degree by summing relative errors to determine the traveling wave energy attenuation matching term includes:

[0014] Obtain the measured peak energy of the traveling wave at the monitoring point and the theoretical traveling wave energy corresponding to the virtual fault point;

[0015] The relative energy error is determined based on the measured peak energy of the traveling wave at the monitoring point and the theoretical traveling wave energy corresponding to the virtual fault point.

[0016] Based on the total number of monitoring points, the relative energy error is summed to obtain the traveling wave energy attenuation matching term.

[0017] Optionally, in some embodiments of this application, the step of quantifying the fit between the virtual fault point and the actual structure of the distribution network by combining the ratio of the actual line length to the straight-line distance with topology validity verification, and determining the topology connectivity consistency item, includes:

[0018] Obtain the actual line length from the fault point to the monitoring point, and the straight-line distance from the fault point to the monitoring point;

[0019] Based on the total number of monitoring points, the proportional terms of the actual line length from the fault point to the monitoring point and the straight-line distance from the fault point to the monitoring point are summed to determine the proportional summation term;

[0020] The topology connectivity consistency term is determined based on the proportional summation term and the topology validity indicator function.

[0021] Optionally, in some embodiments of this application, the step of constructing a dual-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints based on the traveling wave energy attenuation matching term and the topology connectivity consistency term includes:

[0022] The objective function of the distribution network optimization model is determined based on the traveling wave energy attenuation matching term and the topology connectivity consistency term.

[0023] Based on the traveling wave energy attenuation matching constraint, the constraint conditions of the power distribution network optimization model are determined;

[0024] The power distribution network optimization model is determined based on the objective function and the constraints.

[0025] Optionally, in some embodiments of this application, determining the objective function of the distribution network optimization model based on the traveling wave energy attenuation matching term and the topology connectivity consistency term includes:

[0026] Obtain dynamic weight coefficients that conform to preset summation rules;

[0027] Based on the dynamic weighting coefficients, the traveling wave energy attenuation matching term and the topology connectivity consistency term are weighted and summed to obtain the objective function of the distribution network optimization model.

[0028] Optionally, in some embodiments of this application, the step of solving the distribution network optimization model based on the improved barrel theory optimization algorithm to obtain the distribution network traveling wave fault location results associated with the target distribution network includes:

[0029] Initial data is determined based on data preprocessing and feature extraction techniques;

[0030] Pseudo-random initialization is performed based on the initial data to obtain the initialization result. The initialization result is then validated for topological validity to obtain a valid initialization result.

[0031] Based on the effective initialization results, fitness evaluation and weakness identification are performed, global balance update and repair are carried out, and the algorithm iteration process begins.

[0032] If the algorithm iteration process stalls, differential chaotic perturbation processing is performed based on the perturbation factor;

[0033] If the iteration termination condition using dual-judgment is met, the algorithm iteration process is terminated and the optimal solution is output to determine the location result of the traveling wave fault in the distribution network.

[0034] Secondly, embodiments of this application provide a distribution network traveling wave fault location device based on an improved barrel theory optimization algorithm, which has the function of implementing the distribution network traveling wave fault location method based on the improved barrel theory optimization algorithm provided in the first aspect above. The function can be implemented in hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above function, and the modules can be software and / or hardware.

[0035] In one possible design, the device includes:

[0036] The data item determination module is used to determine the traveling wave energy attenuation matching item and the topology connectivity consistency item based on the basic physical relationship of the traveling wave propagation of the target distribution network to be located for fault location.

[0037] The model building module is used to construct a two-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints based on the traveling wave energy attenuation matching term and the topology connectivity consistency term.

[0038] The model solving module is used to solve the distribution network optimization model based on the improved barrel theory optimization algorithm to obtain the distribution network traveling wave fault location results associated with the target distribution network.

[0039] In another aspect, this application provides a computer device including at least one connected processor and a memory, wherein the memory is used to store program code, and the processor is used to call the program code in the memory to execute the methods described in the above aspects.

[0040] In another aspect, this application provides a computer storage medium including instructions that, when executed on a computer, cause the computer to perform the methods described in the above aspects.

[0041] In another aspect, this application provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the methods described in the above aspects.

[0042] Compared to traditional technologies, the technical solution of this application, at the theoretical framework level, constructs a dual-objective optimization model. It quantifies physical correlations through energy attenuation laws and locks reasonable regions using topological constraints. This avoids the defects of insufficient time synchronization accuracy and eliminates the need for preset fixed wave speeds. It can adapt to changes in line materials and environment, fundamentally solving the problem of dependence on external equipment and fixed parameters in existing technologies. In terms of structural adaptability, through line classification coding and connectivity matrix construction, the topological characteristics of the distribution network are quantified and embedded into the entire algorithm process. The structured initial population allocates the number of "barrels" according to the fault occurrence rate and energy sensitivity, prioritizing the coverage of complex areas. Global balance updates only interact information within the connected domain, avoiding cross-regional interference. Differential disturbances strictly follow the line direction and boundary constraints, ensuring that all iteration steps conform to the physical layout of the distribution network. Thus, the overall efficiency and accuracy of traveling wave fault location are improved. Attached Figure Description

[0043] Figure 1 This is an application environment diagram from one embodiment;

[0044] Figure 2 This is a flowchart of one embodiment;

[0045] Figure 3 This is a flowchart illustrating the specific implementation process of fault location in one embodiment;

[0046] Figure 4 This is a simplified topology diagram in one embodiment;

[0047] Figure 5 This is a schematic diagram illustrating the fault identification accuracy under different initial phase angles in one embodiment.

[0048] Figure 6 This is a schematic diagram illustrating the fault location accuracy under different transition resistances in one embodiment.

[0049] Figure 7 This is a structural block diagram of the device in one embodiment;

[0050] Figure 8 This is an internal structural diagram of a computer device in one embodiment;

[0051] Figure 9 This is a diagram of the internal structure of a computer device in another embodiment. Detailed Implementation

[0052] The terms "first," "second," etc., used in the embodiments of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or modules is not necessarily limited to those steps or modules explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. The division of modules appearing in the embodiments of this application is only a logical division. In actual applications, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not performed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interface, and the indirect coupling or communication connection between modules may be electrical or other similar forms. None of these are limited in the embodiments of this application. Furthermore, the modules or sub-modules described as separate components may or may not be physically separate, may or may not be physical modules, or may be distributed among multiple circuit modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiments of this application.

[0053] Figure 1As shown in the application environment diagram of one embodiment, this application provides a method for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm, which can be applied to, for example... Figure 1 In the application scenario shown, terminal 102 communicates with server 104 via a network.

[0054] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0055] It should be noted that the terminal 102 involved in the embodiments of this application can be a wired terminal or a wireless terminal, and can be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connectivity, or other processing devices connected to a wireless modem. The wireless terminal can communicate with one or more core networks via a wireless access network, and the wireless terminal can be a mobile terminal, such as a mobile phone or a computer with a mobile terminal.

[0056] Figure 2 This is a flowchart illustrating one embodiment, such as... Figure 2 As shown in the embodiments of this application, the method for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm includes:

[0057] S2100, based on the fundamental physical relationship of traveling wave propagation in the target distribution network to be fault located, determines the traveling wave energy attenuation matching term and topology connectivity consistency term.

[0058] Fault location refers to the process of determining the specific location of a fault in the distribution network; the target distribution network refers to the actual distribution network for which fault location analysis is to be carried out.

[0059] Among them, the fundamental physical relationship of traveling wave propagation refers to the inherent physical laws that traveling waves follow in the propagation of the power distribution network.

[0060] Among them, the traveling wave energy attenuation matching term is used to characterize the degree of agreement between the measured and theoretical traveling wave energies.

[0061] Among them, the topology connectivity consistency term is used to reflect the matching status between the fault point and the line topology.

[0062] S2200, based on the traveling wave energy attenuation matching term and the topology connectivity consistency term, constructs a two-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints.

[0063] Among them, dual-dimensional fusion refers to the organic combination of energy matching and topological connectivity; the distribution network optimization model refers to a dual-objective mathematical model used for fault location and solution.

[0064] S2300, based on an improved barrel theory optimization algorithm, solves the distribution network optimization model and obtains the distribution network traveling wave fault location results associated with the target distribution network.

[0065] Among them, the improved barrel theory optimization algorithm refers to the optimization solution method that introduces a short-board repair mechanism.

[0066] Among them, the distribution network traveling wave fault location result refers to the final fault location information output by the algorithm.

[0067] Compared to traditional traveling wave fault location methods, this embodiment first determines the traveling wave energy attenuation matching term and topology connectivity consistency term based on the fundamental physical relationships of traveling wave propagation in the target distribution network to be fault located. Then, based on the traveling wave energy attenuation matching term and topology connectivity consistency term, a dual-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints is constructed. Finally, based on an improved barrel theory optimization algorithm, the distribution network optimization model is solved to obtain the traveling wave fault location result associated with the target distribution network. The technical solution of this embodiment, at the theoretical framework level, constructs a dual-objective optimization model, quantifies physical correlation through energy attenuation laws, and uses topology constraints... By locking onto a reasonable area, the algorithm avoids the shortcomings of insufficient time synchronization accuracy and eliminates the need for preset fixed wave velocities. It can adapt to changes in line materials and environment, fundamentally solving the problem of dependence on external equipment and fixed parameters in existing technologies. In terms of structural adaptability, the algorithm quantifies and embeds the topological characteristics of the distribution network into the entire process through line classification coding and connectivity matrix construction. The structured initial population allocates the number of "barrels" according to the fault occurrence rate and energy sensitivity, prioritizing the coverage of complex areas. Global balance updates only interact information within the connected domain, avoiding cross-regional interference. Differentiated disturbances strictly follow the line direction and boundary constraints, ensuring that all iteration steps conform to the physical layout of the distribution network. As a result, the overall efficiency and accuracy of traveling wave fault location are improved.

[0068] In a specific embodiment, the technical principles of the basic physical relationship of traveling wave propagation in the target distribution network are introduced.

[0069] This application constructs the fault location problem as a global optimization problem strongly constrained by physical laws, and uses an improved barrel theory optimization algorithm for efficient and robust solution, ultimately directly outputting an accurate fault location coordinate.

[0070] For traditional Time-Difference-of-Origin (TDOA) fault location theory, an optimization framework based on the dual-dimensional fusion of "traveling wave propagation characteristics – distribution network topology constraints" is proposed. This framework uses the energy attenuation law of traveling waves in the distribution network and the line topology connectivity as its core basis to construct an optimization model with clear physical meaning and deep adaptation to the distribution network structure, thereby achieving accurate fault location inversion. After the traveling wave originates at the fault point, it propagates along the distribution network lines to each monitoring point. This process follows both the physical law of energy attenuation and is strictly constrained by the line topology. This dual characteristic has a strong correlation with the fault location, forming the theoretical basis of the location method.

[0071] Therefore, it is necessary to first clarify the fundamental physical relationships of traveling wave propagation. Specifically, the technical principles of the fundamental physical relationships of traveling wave propagation are as follows.

[0072] The theoretical time for the traveling wave to reach the i-th monitoring point can be expressed as:

[0073] ;

[0074] Where t0 is the time when the fault occurs (unknown), and D... i (x) represents the known propagation path length from fault point x to the i-th monitoring point, and v is the propagation speed of the traveling wave. This formula establishes a direct physical connection between the arrival time of the traveling wave and the fault location. Based on the above-mentioned traveling wave propagation characteristics and distribution network structure features, a dual-core optimization objective function of "traveling wave energy attenuation matching - topological connectivity consistency" is constructed. This function forms a single-objective optimization problem through weighted fusion, which not only ensures the physical rationality of the location but also forces the fault point to conform to the actual structure of the distribution network.

[0075] Optionally, in some embodiments of this application, based on the fundamental physical relationship of traveling wave propagation in the target distribution network to be fault located, the traveling wave energy attenuation matching term and topology connectivity consistency term are determined, including: quantifying the degree of matching by summing relative errors to determine the traveling wave energy attenuation matching term; and quantifying the fit between the virtual fault point and the actual structure of the distribution network by combining the ratio of the actual line length to the straight-line distance with topology validity verification to determine the topology connectivity consistency term.

[0076] Among them, relative error summation refers to quantifying the overall matching degree by accumulating the energy errors of each measuring point; actual line length refers to the real transmission distance from the fault point to the monitoring point along the line; straight distance refers to the geometric straight length from the fault point to the monitoring point; ratio is used to reflect the degree of difference between the line direction and the straight path; topology validity verification refers to judging whether the virtual fault point conforms to the power grid topology rules.

[0077] In this embodiment, two indicators are determined by summing relative errors and topology verification, which improves the stability of model construction, increases the accuracy of fault location, improves the optimization solution effect, and enhances the technical capability of distribution network fault location.

[0078] Optionally, in some embodiments of this application, the degree of matching is quantified by summing relative errors to determine the traveling wave energy attenuation matching term, including: obtaining the measured peak traveling wave energy at the monitoring point and the theoretical traveling wave energy corresponding to the virtual fault point; determining the energy relative error based on the measured peak traveling wave energy at the monitoring point and the theoretical traveling wave energy corresponding to the virtual fault point; and summing the energy relative errors based on the total number of monitoring points to obtain the traveling wave energy attenuation matching term.

[0079] For example, the traveling wave energy attenuation matching term E(x) is the core physical indicator of the optimization objective. Essentially, it infers the fault distance by comparing the theoretical energy attenuation curve corresponding to the virtual fault point with the measured energy at the monitoring point. When a traveling wave propagates in a distribution network, its energy continuously attenuates due to line losses, reflections from branch nodes, and refractions. The degree of attenuation is directly related to the distance from the fault point to the monitoring point, the line material, and the node type; the farther the fault point, the more severe the energy attenuation. Different line materials have different attenuation coefficients per unit length, with copper core cables exhibiting higher energy losses than overhead lines. Based on these characteristics, E(x) quantifies the matching degree using the form of relative error summation.

[0080] The specific formula for the traveling wave energy attenuation matching term E(x) is as follows:

[0081] ;

[0082] Where n is the total number of monitoring points. The measured peak energy of the traveling wave at the monitoring point is taken as the square of the effective value of the voltage transient component. Let be the theoretical traveling wave energy from the virtual fault point x to the i-th monitoring point at time t.

[0083] The calculation needs to take into account both line attenuation and node transmission characteristics:

[0084] ;

[0085] in, The initial traveling wave energy at the fault point is estimated from the short-circuit capacity of the distribution network. The node transmission coefficient; Indicates the sampling time sequence number; Energy attenuation coefficient per unit length of the line; The actual propagation path length from the virtual fault point x to the i-th monitoring point; Belongs to set Node / line segment number; At the virtual fault point x, the set of all nodes / line segments that have a topological connection with the i-th monitoring point.

[0086] In this embodiment, the measured peak energy of the traveling wave at the monitoring point refers to the maximum energy value of the traveling wave actually collected by the monitoring point, which can be denoted as... The theoretical traveling wave energy corresponding to a virtual fault point refers to the theoretical traveling wave energy value generated by the virtual fault point. Specifically, the theoretical traveling wave energy corresponding to virtual fault point x can be denoted as... The relative energy error is used to characterize the degree of deviation between the measured and theoretical traveling wave energy, and can be denoted as... The total number of monitoring points, n, refers to the total number of all monitoring points involved in fault location. Summation refers to the operation of accumulating the relative energy errors of all monitoring points.

[0087] In this embodiment, by obtaining measured and theoretical energy, calculating relative errors, and summing them to determine energy matching terms, the accuracy of energy matching is improved, the precision of fault location is enhanced, and the reliability of distribution network fault location is strengthened.

[0088] Optionally, in some embodiments of this application, the degree of fit between the virtual fault point and the actual structure of the distribution network is quantified by combining the ratio of the actual line length to the straight-line distance with topology validity verification, and the topology connectivity consistency term is determined. This includes: obtaining the actual line length from the fault point to the monitoring point and the straight-line distance from the fault point to the monitoring point; summing the ratio terms of the actual line length from the fault point to the monitoring point and the straight-line distance from the fault point to the monitoring point based on the total number of monitoring points to determine the ratio summation term; and determining the topology connectivity consistency term based on the ratio summation term and the topology validity indicator function.

[0089] For example, the topology connectivity consistency term T(x) is designed to address the pain point of traditional algorithms where "fault points are detached from actual lines". Since a fault point in a distribution network must fall on a certain actual line segment, T(x) quantifies the fit between the virtual fault point and the actual structure of the distribution network by combining the ratio of the actual line length to the straight-line distance with topology validity verification.

[0090] The expression for the topological connectivity consistency term T(x) is:

[0091] ;

[0092] in This is the actual line length from the fault point to the monitoring point. The ratio of the two distances is closer to 1, indicating that the propagation path closely follows the route of the line. This is a topology validity indicator function. If the path falls on the line and is reasonable (I(x)=1), then the penalty is doubled (I(x)=2), forcing the algorithm to abandon unreasonable paths.

[0093] The specific meaning of the topology validity indicator function is as follows.

[0094] I(x)=1 means that the fault point x is located on a real line segment in the distribution network, and the propagation path from x to all monitoring points strictly follows the topological connection relationship of the distribution network.

[0095] I(x)=2 indicates that the fault point x is not on any real line segment, or although it is on the line, the propagation path to the monitoring point does not match the actual topology. In this case, the penalty is doubled, forcing the algorithm to abandon or correct the solution.

[0096] The entire optimization model has a clear logical closed loop: by minimizing E(x), the approximate distance of the fault is determined by utilizing the energy decay law; by minimizing T(x), invalid searches in non-line areas are eliminated; the two work together to achieve precise location.

[0097] In this embodiment, the actual line length This refers to the actual path length from the fault point to the monitoring point along the power distribution line; straight-line distance. This refers to the geometric straight-line length between the fault point and the monitoring point; proportional term Used to characterize ratio relationships; proportional summation term This refers to the value obtained by summing the proportional terms corresponding to each monitoring point; Topology validity indicator function. Used to characterize whether a fault point conforms to the power grid topology.

[0098] In this embodiment, the topology terms are determined by combining the ratio of actual line length to straight-line distance with topology validity verification, which improves the rationality of topology constraints, increases the accuracy of fault location, improves the stability of optimization solutions, and enhances the reliability of distribution network fault location.

[0099] Optionally, in some embodiments of this application, a dual-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints is constructed based on the traveling wave energy attenuation matching term and the topology connectivity consistency term. This includes: determining the objective function of the distribution network optimization model based on the traveling wave energy attenuation matching term and the topology connectivity consistency term; determining the constraint conditions of the distribution network optimization model based on the traveling wave energy attenuation matching constraint; and determining the distribution network optimization model based on the objective function and the constraint conditions.

[0100] Here, the objective function refers to the mathematical expression that characterizes the solution objective of the distribution network optimization model; the traveling wave energy attenuation matching constraint refers to the condition that limits the accuracy of traveling wave energy matching; the constraint condition refers to the restriction condition that must be followed in the solution process of the distribution network optimization model; and the distribution network optimization model refers to the fault location mathematical model that integrates the objective function and the constraint condition.

[0101] In this embodiment, by determining the objective function and constraints, a distribution network optimization model is constructed, which improves the rationality and applicability of the model, enhances the optimization effect of fault location, and strengthens the accuracy and stability of distribution network fault location.

[0102] Optionally, in some embodiments of this application, the objective function of the distribution network optimization model is determined based on the traveling wave energy attenuation matching term and the topology connectivity consistency term, including: obtaining dynamic weight coefficients that conform to a preset summation rule; and performing a weighted summation of the traveling wave energy attenuation matching term and the topology connectivity consistency term based on the dynamic weight coefficients to obtain the objective function of the distribution network optimization model.

[0103] Among them, the preset summation rule refers to the pre-set weight coefficient superposition calculation rule; the dynamic weight coefficient refers to the weight calculation parameter that can be adaptively adjusted; and the weighted summation refers to the operation of superimposing two indicators according to their weights.

[0104] In this embodiment, the objective function is constructed by weighting and summing the two indicators using dynamic weighting coefficients, which improves the adaptability of the objective function, enhances the stability of the optimization solution, and strengthens the accuracy and robustness of distribution network fault location.

[0105] For example, in conjunction with the above embodiments, the objective function of the distribution network optimization model is described below.

[0106] This model is a bi-objective optimization problem, the core of which is to minimize the traveling wave energy attenuation matching error and the topological connectivity consistency penalty, which can be expressed as: Therefore, the objective function of the distribution network optimization model is specifically expressed as: ;

[0107] in, Traveling wave energy attenuation matching error; Topological connectivity consistency penalty.

[0108] in and These are dynamic weighting coefficients. In the formula, ω1 and ω2 are complementary, and their weights sum to 1, satisfying the condition... + =1, the value of which is flexibly adjusted according to the actual structure of the distribution network. The preset summation rule refers to the dynamic weighting coefficient satisfying... + =1 rule.

[0109] In addition, in multi-branch areas =0.6, main line area =0.6. Specifically, the value of the dynamic weighting coefficient is flexibly adjusted according to the actual structure of the distribution network: in areas with multiple branches and dense nodes, the line topology imposes a stronger constraint on the traveling wave propagation path, and in this case, the value is set to ( =0.6), strengthening the influence of topological constraints; in long-distance, branchless trunk lines, the traveling wave energy attenuation pattern is more stable, setting ( =0.6), highlighting the role of energy matching.

[0110] Specifically, in multi-branch, densely nodeed regions: topological constraints have a stronger impact on the propagation path of traveling waves. The weight coefficient of the topological connectivity consistency term is set to 0.6 to strengthen the effect of topological constraints, and the remaining 0.4 is allocated to the traveling wave energy attenuation matching term.

[0111] Specifically, in long-distance, branchless trunk lines: the energy attenuation pattern of traveling waves is more stable. The weight coefficient of the traveling wave energy attenuation matching term is set to 0.6 to highlight the core role of energy matching, and the remaining 0.4 is allocated to the topological connectivity consistency term.

[0112] The weight value of 0.6 in this application was determined through repeated iterative testing and verification using multiple sets of samples (1000 training samples and 400 test samples) of the 10kV distribution network simulation model on the PSCAD platform. This value represents the optimal weight for the core constraint, which can highlight the dominant constraint while retaining the auxiliary verification role of the secondary items, thus achieving dual-objective collaborative optimization.

[0113] For example, in conjunction with the above embodiments, the constraints of the distribution network optimization model are described below.

[0114] The constraints of the distribution network optimization model include a traveling wave energy attenuation matching constraint. Specifically, when a fault traveling wave propagates in a cable, its energy attenuates exponentially, and transmission and reflection occur when it passes through nodes. This constraint requires that the theoretical traveling wave energy calculated from the virtual fault point x... It must closely match the traveling wave energy measured at the monitoring point.

[0115] The mathematical expression for the traveling wave energy attenuation matching constraint is:

[0116] ;

[0117] Where t: sampling time sequence number; T: total number of sampling times; i: monitoring point sequence number; n: total number of monitoring points; : The measured traveling wave energy at the i-th monitoring point at time t.

[0118] Optionally, in some embodiments of this application, the distribution network optimization model is solved based on the improved barrel theory optimization algorithm to obtain the distribution network traveling wave fault location result associated with the target distribution network. This includes: determining initial data based on data preprocessing and feature extraction techniques; performing pseudo-random initialization based on the initial data to obtain initialization results; performing topology validity verification on the initialization results to obtain valid initialization results; performing fitness evaluation and bottleneck identification based on the valid initialization results; performing global balance update and repair; and entering the algorithm iteration process; performing differentiated chaotic perturbation processing based on perturbation factors when the algorithm iteration process stalls; and terminating the algorithm iteration process and outputting the optimal solution when the iteration termination condition using dual judgment is met, thereby determining the distribution network traveling wave fault location result.

[0119] This embodiment specifically includes the following steps:

[0120] S3100 determines the initial data based on data preprocessing and feature extraction techniques.

[0121] Among them, data preprocessing refers to the filtering and noise reduction of the original fault data; feature extraction technology refers to the technology of extracting key positioning features from the traveling wave signal; and initial data refers to the data used for algorithm calculation after preprocessing and extraction.

[0122] For example, in order to solve the above-mentioned nonlinear, multimodal optimization problem, a new optimization model solution algorithm adapted to the distribution network structure is proposed. The improved BTO algorithm takes "the capacity of the barrel is determined by the shortest plank" as its core logic, and reconstructs the definition of "plank" and the iteration mechanism according to the characteristics of the distribution network to ensure that each step of the process is deeply bound to the distribution network structure.

[0123] The data preprocessing and feature extraction process is as follows.

[0124] The required data include: fault traveling wave transient data (time-domain waveforms of voltage transient components at each monitoring point), distribution network topology and line parameter data (line classification, energy attenuation coefficient α per unit length, node transmission coefficient, line length range, fault occurrence rate, node / line segment coordinates), and basic distribution network operation data (short-circuit capacity, line material properties).

[0125] The data preprocessing process includes:

[0126] For fault traveling wave transient data: noise is removed, the square of the effective value of the voltage transient component is calculated to obtain the measured peak energy of the traveling wave, and the data is normalized.

[0127] For distribution network topology and line parameter data: classify and encode them according to main line / branch line / terminal line, construct a binary connectivity matrix, and verify and correct invalid topology data;

[0128] Based on the basic data of distribution network operation: estimate the initial traveling wave energy at the fault point using short-circuit capacity, and calibrate the attenuation coefficient and node transmission coefficient of lines with different materials.

[0129] The feature extraction process includes:

[0130] The measured peak energy of the traveling wave is extracted from the preprocessed transient data of the fault traveling wave; the line attenuation coefficient, node transmission coefficient, ratio of actual line length to straight distance, and topology validity indicator function parameters are extracted from the preprocessed distribution network topology and line parameter data.

[0131] Finally, all the data obtained after data preprocessing and feature extraction of fault traveling wave transient data, distribution network topology and line parameter data, and distribution network operation basic data are summarized to form the initial data.

[0132] S3200 performs pseudo-random initialization based on initial data to obtain initialization results, and performs topology validity verification on the initialization results to obtain valid initialization results.

[0133] Among them, pseudo-random initialization refers to the operation of generating the initial population in a pseudo-random manner; initialization result refers to the set of virtual fault points after the initial population is generated; topology validity verification refers to verifying whether the fault points conform to the power grid topology rules; and valid initialization result refers to a valid initial population that passes the topology verification.

[0134] For example, the quality of the initial population directly affects the iteration efficiency and positioning accuracy. If pseudo-random initialization is used, it is easy to cause virtual fault points to be concentrated on the main line, ignoring high-fault areas such as branch lines and terminal lines.

[0135] Therefore, the algorithm first classifies and encodes all line segments of the distribution network according to "main line (L) - branch line (B) - terminal line (E)", and simultaneously labels the energy attenuation coefficient α and node transmission coefficient of each line. Length range [L] min, L max And failure rate.

[0136] Subsequently, the initial number of "barrels" was allocated based on the energy sensitivity and failure rate of the lines: the terminal lines had significant energy attenuation and frequent failures, so the sampling density was the highest; the branch lines had dense nodes and complex propagation paths, so the sampling density was the second highest; the trunk lines had simple structures and low failure rates, so the sampling density was relatively low, ensuring that the initial population prioritized coverage of high-risk and complex structural areas.

[0137] Among them, the energy sensitivity is first classified and coded by the distribution network line segments and the parameters of this type are labeled simultaneously. Then, the initial number of "barrels" (i.e. the initial population sampling density) is allocated according to the energy sensitivity and fault occurrence rate of the line. The above parameters are labeling items rather than the calculation basis for allocating the initial population.

[0138] The data sources for each labeled parameter are as follows:

[0139] Energy attenuation coefficient α and node transmission coefficient: These are calibrated based on the inherent physical properties of the distribution network lines (overhead / copper core cable), laying method, and node type, combined with the energy attenuation law of traveling wave propagation and node transmission characteristics. Length range [Lmin, Lmax]: Derived from distribution network topology and line parameter data, i.e., the actual physical dimensions of the lines exported from the Distribution Management System (DMS) / Geographic Information System (GIS). Fault occurrence rate: Obtained based on historical fault operation statistics of the distribution network, combined with statistical analysis of the operating characteristics of main lines, branch lines, and terminal lines.

[0140] Finally, all initial virtual fault points must pass topology validity verification, strictly falling on actual line segments and not crossing nodes or falling in non-line areas, thus laying a structurally sound foundation for subsequent iterative optimization. The rules for topology validity verification are as follows:

[0141] Crossing Nodes: B1 is the end point of the M1-B1 line and the starting point of the B1-B2 line. If the initial virtual fault point is generated at "4.2km on the M1-B1 line", this location has crossed node B1 and entered the B1-B2 line section, making it an invalid point that crosses a node. Falling into Non-Line Areas: If the initial virtual fault point is generated in the blank area between the M1-B1 line and the B2-B3 line, this location has no actual line segment coverage, making it an invalid point falling into non-line areas. Virtual fault points in both of these cases will be directly removed by topology validity verification and will not be included in the initial population.

[0142] Based on the effective initialization results, S3300 performs fitness evaluation and bottleneck identification, performs global balance update and repair, and enters the algorithm iteration process.

[0143] Among them, fitness evaluation refers to the operation of calculating the objective function to judge the quality of individuals; bottleneck identification refers to the process of identifying the term with the largest error in energy or topology; global balance update refers to the operation of co-optimizing individuals and the global optimal solution; repair refers to the process of directional correction of bottleneck errors; and algorithm iteration process refers to the complete execution flow of the algorithm iteratively seeking optimization.

[0144] For example, during the fitness evaluation process, for each "barrel" representing a virtual fault point in the algorithm, its two core indicators, traveling wave energy attenuation matching error E(x) and topology connectivity consistency error T(x), are first calculated and normalized; then, based on the dynamic weight coefficients... The objective function value is obtained by weighted calculation. This value is used as the fitness value of the "barrel". The smaller the value, the higher the matching degree of the location of the corresponding virtual fault point and the rationality of the topology, which provides a quantitative basis for subsequent short-bone identification.

[0145] The "bottleneck identification" step in the iteration process requires combining the characteristics of the power distribution network structure and energy propagation characteristics to identify the core bottlenecks currently affecting positioning accuracy.

[0146] First, the two core indicators of each "barrel" are normalized. If the topology connectivity consistency error T(x) > 0.5, it indicates that the virtual fault point may deviate from the actual line or the propagation path is unreasonable. In this case, the topology constraint is treated as the "shortest plank" and repaired first.

[0147] During the repair process, the adjustment direction strictly follows the distribution network line direction. For example, if the B2-B3 branch runs north-south, the virtual fault point will only be adjusted along the north-south direction to avoid drifting to the adjacent B2-B4 branch. The adjustment step size is dynamically set according to the line type. The main line is longer, so the step size is set to 0.1km. The branch line step size is set to 0.05km. The terminal line is shorter and has higher accuracy requirements, so the step size is set to 0.03km to ensure that the adjustment is both efficient and accurate.

[0148] If the energy attenuation matching error E(x) > 0.3, it indicates that the energy propagation pattern corresponding to the virtual fault point deviates significantly from the measured data. In this case, energy matching is used as the "shortest plank" and adjusted directionally in conjunction with the line attenuation coefficient and the node transmission coefficient.

[0149] If the measured energy is greater than the theoretical energy, it means that the virtual fault point is too far from the monitoring point and needs to be adjusted towards the monitoring point. The adjustment range is negatively correlated with the attenuation coefficient to avoid over-adjustment due to high energy sensitivity.

[0150] If the measured energy is less than the theoretical energy, adjust the direction away from the monitoring point to gradually reduce the energy matching error. The adjustment magnitude is also negatively correlated with the attenuation coefficient.

[0151] This negative correlation adjustment principle is a unified guideline for energy bottleneck repair. Regardless of whether the virtual fault point is adjusted closer to or further away from the monitoring point, the adjustment range will be set according to this rule: the larger the line attenuation coefficient, the more sensitive the energy changes with distance, and the smaller the adjustment range, to avoid over-adjustment and deviation from the actual fault point due to high energy sensitivity; the smaller the attenuation coefficient, the more gradual the energy change, and the larger the adjustment range can be appropriately increased to improve the iterative convergence efficiency.

[0152] S3400 performs differentiated chaotic perturbation processing based on perturbation factors when the algorithm iteration process stalls.

[0153] Among them, stagnation refers to the state in which the optimal solution is not updated for a long time during algorithm iteration; perturbation factor refers to the calculation parameter used to generate the perturbation amplitude; differentiated chaotic perturbation processing refers to the operation of applying chaotic perturbations of different intensities to different regions.

[0154] For example, the multi-branch, topologically dispersed structure of distribution networks leads to significant differences in the traveling wave propagation patterns and fault distribution characteristics of lines in different areas. If indiscriminate global information exchange is used, it can easily lead to invalid information interference in disconnected areas, causing virtual fault points to drift to physically infeasible areas, reducing iteration efficiency and location accuracy. Therefore, a global balance update mechanism based on line connectivity constraints is designed to ensure that information exchange strictly follows the distribution network topology, achieving a precise balance between individual optimization and group collaboration. First, all line segments in the distribution network are uniformly encoded, and a binary connectivity matrix M is constructed based on the actual connection relationships of the lines, where N is the total number of distribution network line segments. The matrix elements are defined as follows: ;

[0155] For example, if branch B2-B3 shares node B2 with branch B1-B2 and also shares node B2 with branch B2-B4, then M B3,B2 =1、M B3,B4 =1, while the terminal lines E7-E8, which have no connection relationship, satisfy M. B3,E7 =0.

[0156] To avoid invalid information transmission across regions, each "barrel," or virtual fault point, only extracts globally optimal information within the connected domain of its local line segment. Specifically, for the m-th "barrel" located on line segment p, its globally optimal reference object must satisfy M... pq The selection process involves filtering from the "barrels" corresponding to all line segments q with a value of 1, i.e., only accepting the optimal solution feedback from connected branches and rejecting information interference from disconnected regions. Based on the connectivity matrix and information interaction rules, the global balance adjustment formula is designed as follows:

[0157] ;

[0158] This is the updated value of the nth index of the mth "barrel" in the (k+1)th iteration; This is the current value of the index in the k-th iteration; This represents the current value of the nth metric of the globally optimal "barrel" within the connected component; M is the global learning factor, ranging from [0.3, 0.7], used to balance the weights of individual autonomous optimization and group information sharing, avoiding the trap of local optima caused by over-reliance on group information; pq M is used as an element of the connectivity matrix to implement the quantized embedding of topological constraints. When line segments p and q are not connected, M... pq =0, global information interaction items fail, the "barrel" only updates itself iteratively, effectively preventing virtual fault points from drifting to disconnected areas; when the line segment is connected, M pq =1, guides the individual optimization direction through the group's optimal information, improves convergence efficiency, and ensures the structural rationality and physical feasibility of the iterative process.

[0159] If the objective function value of the globally optimal "barrel" does not improve significantly for K consecutive iterations during the iteration process, it indicates that the algorithm may be trapped in a local optimum.

[0160] The specific data manifestation of "no significant improvement" refers to the fact that the objective function value F(x) corresponding to the global optimal "barrel" remains within a very small fluctuation range in K consecutive iterations, without a significant downward trend. In other words, during the iteration process, this core value does not effectively converge towards a better direction (smaller value) and always stays in a certain numerical range, indicating that the algorithm's search for the global optimal solution has stalled and is likely to fall into a local optimum trap.

[0161] The quantitative criterion for "significant improvement" refers to the following: when the decrease in the objective function value exceeds the preset threshold ε (usually 0.1~0.01) in a simulation verification scenario combining the engineering accuracy requirements for fault location in the distribution network, the result is considered "significant improvement"; conversely, if the decrease is ≤ε for K consecutive generations, or if the value does not decrease, or even fluctuates slightly upward, it is considered "no significant improvement".

[0162] The value of the threshold ε is set based on the meter-level accuracy requirement for fault location in the distribution network. It has been verified by PSCAD platform 10kV distribution network simulation. This range can accurately distinguish the effective convergence and stagnation states of the algorithm and is suitable for the actual needs of distribution network projects.

[0163] Considering the differences in traveling wave propagation characteristics and local optimal risks in different structural areas of the distribution network, such as main lines, branch lines, and terminal lines, a differentiated stagnation breakthrough mechanism is designed to achieve effective escape from the global optimal solution while ensuring structural constraints.

[0164] The main line area, due to its simple line structure, lack of branch interference, and stable traveling wave energy attenuation, exhibits shallow local optimum trap depths. Therefore, a small-amplitude disturbance is adopted, with a base disturbance amplitude δ1 = 0.02 km, and the disturbance direction is along the extension direction of the main line. This avoids disrupting the established stable search trend due to excessive disturbance, ensuring positioning stability. In contrast, the branch line area, with its dense nodes and intersecting branches, experiences multiple reflections and refractions during traveling wave propagation, resulting in a large number of local optimum traps. A medium-amplitude disturbance is adopted, with a base disturbance amplitude δ1 = 0.05 km, balancing escape capability and structural constraints. The terminal line area, due to its short line length, significant signal attenuation, and weak fault characteristics, exhibits prominent local optimum problems. A large-amplitude disturbance is adopted, with a base disturbance amplitude δ1 = 0.08 km, ensuring that the virtual fault point after disturbance does not exceed the line boundary, maintaining structural rationality. To ensure that the disturbance behavior conforms to the traveling wave propagation law, the disturbance factor comprehensively considers the line energy attenuation characteristics and topology length characteristics, and the calculation formula is as follows:

[0165] ;

[0166] in, , These are the virtual fault point locations for the (k+1)th and kth iterations, respectively; The amplitude of the basic disturbance; This represents the energy attenuation coefficient per unit length of the current line segment; This represents the actual line length from the current virtual fault point to the monitoring point. This represents the total length of the current line segment, reflecting the constraints of the topology location on disturbances and preventing the fault point from shifting excessively to both ends of the line. The disturbance direction vector is 1 or -1 along the current line segment extension direction; γ1 and γ2 are weighting coefficients. This approach avoids local optima while adhering to physical laws and structural constraints.

[0167] S3500, when the iteration termination condition using dual-judgment is met, terminates the algorithm iteration process and outputs the optimal solution to determine the location result of the traveling wave fault in the distribution network.

[0168] The iteration termination condition employs a dual-judgment approach: when the objective function value of the global optimal solution F(x) < ϵ and T(x) < 0.05, it indicates that the virtual fault point satisfies both physical energy matching and conforms to the distribution network structure, and the iteration terminates. If the iteration count reaches a preset maximum value and the conditions are still not met, the iteration also automatically terminates and outputs the optimal solution, ensuring that the engineering requirements of the distribution network are met. Here, ϵ refers to the convergence threshold of the objective function, used to determine whether the positioning accuracy meets the standard; this threshold can be determined comprehensively based on the allowable engineering error, acquisition accuracy, and solution accuracy.

[0169] In this embodiment, by combining a complete iterative process with topology verification and chaotic perturbation, the stability of the algorithm solution is improved, the accuracy of fault location is increased, the global optimization effect is improved, and the technical capability of distribution network fault location is enhanced.

[0170] in addition, Figure 3 The following is a flowchart of the specific implementation of fault location in one embodiment. For specific limitations, please refer to the above description of the specific limitations of a distribution network traveling wave fault location method based on an improved barrel theory optimization algorithm, which will not be repeated here.

[0171] The technical research process and other technical details of this application are described below with reference to a specific embodiment.

[0172] Distribution networks have complex topologies and operate in variable environments. Quickly and accurately locating the fault point after a failure is crucial for minimizing power outage time and improving power supply reliability. Existing fault location methods are mainly divided into methods based on power frequency electrical quantities and methods based on transient traveling waves.

[0173] In traditional techniques, the impedance method calculates the fault distance by measuring the voltage and current amplitudes and phases after a fault. Although it is low-cost, its accuracy is easily affected by inaccurate line parameters, transition resistance, system operating mode, and the current injected by distributed power sources, which limits its application in complex distribution networks.

[0174] The traveling wave method utilizes the high-frequency transient traveling wave generated instantaneously and propagating to both ends of the line for fault location. It boasts advantages such as high theoretical accuracy and insensitivity to transition resistance. The double-ended traveling wave location method is a typical example. Its basic principle is to calculate the fault distance x by accurately measuring the time difference Δt between the arrival times of the traveling wave front at both ends of the line, combined with the known traveling wave velocity V and line length L. However, the practical application of this method heavily relies on the accuracy of two core components: 1) strict synchronization of clocks at both ends of the line; and 2) accurate calibration of the arrival time of the traveling wave front. Wave front calibration, in particular, is highly susceptible to errors due to noise interference, waveform distortion, and complex reflections and refractions. Even a small time calibration deviation can lead to a location error of tens to hundreds of meters.

[0175] To reduce the direct dependence on wavefront calibration accuracy, intelligent optimization algorithms have been introduced into the research field in recent years, transforming the fault location problem into an optimization problem of searching for the optimal fault location in the solution space. The core idea is to construct an objective function describing the difference between the "theoretical traveling wave waveform corresponding to the assumed fault location" and the "actual recorded waveform," and then minimize this function through an optimization algorithm to deduce the most probable fault point.

[0176] In a collusion technique called "A Traveling Wave Fault Location Method for Distribution Networks Based on Multi-Objective Particle Swarm Optimization," the scheme recognizes the limitations of a single objective function in fault location and attempts to simultaneously optimize multiple potentially conflicting indices to improve the robustness and rationality of the location results. However, multi-objective particle swarm optimization outputs a set of non-dominated solutions, rather than a single deterministic solution. In fault location scenarios requiring clear and rapid response, an additional complex decision-making mechanism is needed to select the final result from the frontier. This not only introduces computational overhead and subjectivity but may also lead to the selection of a non-optimal solution due to inappropriate decision criteria, causing delays and uncertainties in the location decision. Furthermore, its performance is highly dependent on parameters such as inertia weight ω, learning factor c1, and external archive size. In the sharp, multi-peaked objective space of distribution network fault location, improper parameter settings can easily cause the particle swarm to converge prematurely to a local Pareto front or exhibit oscillations that prevent convergence, thus reducing the reliability of the location results.

[0177] Therefore, the shortcomings of the existing technology are as follows:

[0178] The core drawback of the single-ended method lies in the identification of the reflected wavefront at the fault point and its adaptability to complex operating conditions. Its wavefront identification is easily affected by line signal attenuation, environmental noise, and various interference signals, which can easily lead to misjudgment or missed judgment. For scenarios such as high-resistance faults and arc grounding faults, the reflected wave signal at the fault point is relatively weak, which will further increase the difficulty of wavefront identification and reduce the reliability of positioning. At the same time, this method is highly dependent on the accuracy of the wave velocity value, which will dynamically change with the line material, temperature, humidity and other operating conditions, directly introducing positioning errors. In scenarios with branch lines or multiple fault points, the source of the reflected wave becomes more complex, making it difficult to distinguish the reflected wave from the fault point and the reflected wave from the branch node, thus causing positioning confusion.

[0179] The drawbacks of the two-end method mainly lie in time synchronization, parameter dependence, and application scenario limitations. It requires extremely high time synchronization accuracy at both ends of the line. Even a synchronization error at the microsecond level can significantly amplify the positioning error. Moreover, the positioning accuracy also depends on the accuracy of the wave velocity value. The deviation between the wave velocity and the actual line conditions will be directly transmitted to the positioning result. In addition, the two-end method requires the deployment of data transmission and synchronization equipment at both ends, which not only increases the deployment cost, but also causes positioning failure due to equipment failure or communication interruption. The initial traveling wave front may be blocked by interference signals such as lightning strikes or operational overvoltages, causing wave front identification errors. At the same time, this method is more suitable for purely straight lines. When there are T-junctions or parallel equipment on the line, the propagation path of the initial wave front becomes more complex, which in turn introduces additional positioning errors.

[0180] In the engineering scenario of distribution network fault location, which requires rapid and definitive response, existing methods based on multi-objective particle swarm optimization (MOPSO) have several inherent limitations. First, the method ultimately outputs a Pareto optimal set of solutions, rather than a single deterministic solution. This forces maintenance personnel to expend additional time and computational resources to select the final location point from the set based on potentially subjective criteria. This decision-making process not only delays fault handling but also introduces uncertainty into the results. Second, the algorithm's convergence performance is extremely sensitive to parameters such as inertia weights and learning factors. In the complex, multi-peak objective space unique to the traveling wave model of the distribution network, slight errors in parameter settings can easily lead to the search getting trapped in local optima or oscillations, making it difficult to guarantee the reliability of the location results. A deeper problem lies in the fact that the update mechanism of standard MOPSO is a purely mathematical optimization, failing to effectively incorporate the deterministic physical laws that traveling wave propagation must follow. This results in a significant waste of computation on blindly exploring physically infeasible regions, leading to low search efficiency.

[0181] Based on this, this application provides a method for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm, also known as a method for locating traveling wave faults in distribution networks based on an improved BTO algorithm, or a novel fault location method based on chaotic particle swarm optimization. Details are as follows.

[0182] This application provides a method for fault location in distribution networks based on an improved barrel theory optimization algorithm. First, the fault location problem is constructed as a multimodal optimization model based on the arrival time difference of the traveling waves. To address the inherent limitation of the standard particle swarm optimization (PSO) algorithm in solving such problems—its susceptibility to local optima due to insufficient population diversity—this method introduces chaos theory for dual improvements:

[0183] (1) Utilize the ergodicity of chaotic mapping to generate an initial population, ensuring a more comprehensive initial exploration of the entire solution space;

[0184] (2) A chaotic perturbation mechanism is introduced during the iteration process. When the algorithm stagnates, a perturbation is applied to the current optimal solution to help it escape local traps, thereby enhancing the global search capability. Simulation results show that this method effectively overcomes the "premature convergence" problem of the standard algorithm, significantly improves the accuracy, convergence speed and robustness of fault location, and provides strong technical support for rapid and accurate fault diagnosis in complex power grid environments.

[0185] To address the inherent limitations of existing single-ended methods, double-ended traveling wave methods, and multi-objective optimization algorithms in distribution network fault location, this application aims to construct a more robust, efficient, and direct fault location solution. Specifically, the objectives are as follows:

[0186] (1) Fundamentally avoid the dependence on precise calibration of a single wavefront and high-precision time synchronization: By transforming the problem into an optimization model based on key feature matching, this method does not rely on the precise capture of the absolute time of a specific reflection wavefront or initial wavefront, thereby effectively overcoming the core problems of easy misjudgment of wavefront identification in the single-end method and the stringent requirements of synchronization accuracy in the double-end method, and improving the fault tolerance capability under complex working conditions such as noise interference and high impedance faults.

[0187] (2) Significantly reduce sensitivity to fixed traveling wave velocity values: By using wave velocity as a correctable or co-optimized parameter in the optimization model, the algorithm can adapt to the dynamic changes in wave velocity caused by line material and environmental factors, thereby reducing the systematic error introduced by inaccurate wave velocity.

[0188] (3) Provide a single, clear location decision output to avoid complex post-event decision-making: Unlike the decision delay and uncertainty brought about by the Pareto solution set output by the MOPSO method, the improved Barrel Theory Optimization (BTO) algorithm automatically converges to a single overall optimal solution during the optimization process through its inherent "short-board repair" and "global balance" mechanism, directly outputting the definite fault location, which meets the requirements of engineering applications for fast and clear response.

[0189] (4) Improve the convergence stability and efficiency of the optimization process: By improving the BTO algorithm, an adaptive search strategy that is insensitive to parameter settings and can effectively cope with the multi-peak characteristics of the distribution network location problem is designed. At the same time, the physical constraints of traveling wave propagation are deeply embedded in the search logic of the algorithm to guide the algorithm to explore efficiently in the physically feasible solution space, avoid blind search, thereby speeding up the convergence speed and effectively preventing getting trapped in local optima or oscillations.

[0190] The key technical points of the distribution network traveling wave fault location method based on the improved barrel theory optimization algorithm provided in this application are concentrated on the construction of a dual-objective optimization model and the design of an improved BTO algorithm, as follows: This application completely abandons the traditional TDOA time difference theory. One of the core protection points is to construct a dual-objective optimization model of "traveling wave energy attenuation - topology connectivity consistency". The traveling wave energy attenuation matching error E(x) is quantitatively modeled by combining the theoretical energy calculation formula of the attenuation coefficient per unit length of the line and the energy transmission coefficient of the node. The topology connectivity consistency error T(x) is quantified by the ratio of the actual line length to the straight distance and constrained by the dual penalty mechanism of the topology effectiveness indicator function I(x). At the same time, a dynamic weight adjustment strategy is adopted to allocate values ​​according to the structural characteristics of the distribution network. This whole set of dual-objective modeling logic is the core protection content. Another core protection point is the improved BTO algorithm adapted to the multi-branch, densely node structure of the distribution network. Its key mechanisms include a structured initial population generation method, which allocates the number of initial "barrels" according to the "line fault occurrence rate and energy sensitivity", giving priority to covering high-risk areas such as end lines and branch lines. All initial virtual fault points need to pass the topology validity verification. The energy bottleneck is combined with the measured and theoretical energy deviation direction and the line attenuation coefficient to achieve adaptive adjustment of the step size. This strategy effectively improves the iteration efficiency and positioning accuracy.

[0191] The advantages of the distribution network traveling wave fault location method based on the improved barrel theory optimization algorithm provided in this application are as follows:

[0192] Compared with existing techniques based on multi-objective particle swarm optimization (MOPSO), as well as comparative techniques such as traditional particle swarm optimization (PSO) and genetic algorithms (GA), this application has significant advantages in theoretical framework, structural adaptability, convergence performance, robustness, and engineering practicality, as detailed below:

[0193] At the theoretical framework level, existing technologies generally rely on TDOA time difference theory or modeling based on single physical characteristics. MOPSO, traditional PSO, and other technologies have not broken free from their dependence on high-precision time synchronization equipment or fixed traveling wave speeds, making them susceptible to positioning deviations caused by dynamic changes in wave speed and node reflection interference. In contrast, this application constructs a dual-objective optimization model of "traveling wave energy attenuation - topology connectivity consistency". It quantifies physical correlations through energy attenuation laws and uses topological constraints to lock in reasonable areas. This avoids the defects of insufficient time synchronization accuracy and eliminates the need to preset a fixed wave speed. It can adapt to changes in line material and environment, fundamentally solving the problem of dependence on external equipment and fixed parameters in existing technologies.

[0194] Regarding structural adaptability, existing optimization processes are disconnected from the multi-branch, densely noded structural characteristics of distribution networks: the initial populations of MOPSO and traditional GA rely on pseudo-random generation, which easily leads to particle concentration on the main line, ignoring high-fault-risk areas such as branch lines and terminal lines; global information interaction lacks topological constraints, easily resulting in invalid interference across non-connected regions, and even causing virtual fault points to deviate from actual lines. This application, through line classification coding and connectivity matrix construction, quantifies and embeds the topological characteristics of the distribution network into the entire algorithm process. The structured initial population allocates the number of "barrels" according to the fault occurrence rate and energy sensitivity, prioritizing the coverage of complex areas; global balance updates only interact information within connected domains, avoiding cross-regional interference; differentiated disturbances strictly follow the line direction and boundary constraints, ensuring that all iteration steps conform to the physical layout of the distribution network, perfectly solving the core pain point of "algorithm and structure disconnect" in existing technologies.

[0195] The following specific embodiment illustrates the experimental effect of the distribution network traveling wave fault location method based on the improved barrel theory optimization algorithm provided in this application.

[0196] Figure 4 This is a simplified topology diagram in one embodiment; Figure 5 This is a schematic diagram illustrating the fault identification accuracy under different initial phase angles in one embodiment. Figure 6 This is a schematic diagram illustrating the fault location accuracy under different transition resistances in one embodiment.

[0197] refer to Figures 4 to 6 To quantitatively evaluate the effectiveness and superiority of the proposed distribution network fault accurate location method based on the improved barrel theory optimization algorithm (BTO), a system was built on the PSCAD platform based on a simplified topology. Figure 4 The 10kV distribution network model shown has line lengths as shown in Table 1.

[0198] Table 1:

[0199]

[0200] The circuit model adopts a frequency-dependent phase domain model with a sampling frequency of 1MHz.

[0201] A sample database was constructed based on a power distribution network simulation model. Considering different fault locations, a virtual fault point was set every 50m, resulting in a total of 1872 samples, which were used as the training sample set. Then, a single-phase ground fault was set at a random location as the test sample set.

[0202] This project randomly sets the fault location, transition resistance, and initial phase angle to generate a new sample set. By comparing the localization results under different initial phase angles, the robustness of the algorithm under strong and weak fault signals is verified, thus validating the generalizability of the proposed method for fault localization. The parameter setting methods for different operating conditions are shown in Table 2, generating a total of 400 fault samples.

[0203] Table 2:

[0204]

[0205] The proposed method was compared with common methods on 400 samples, namely the PSO fault location method and the GA fault location method, with the initial phase angle and transition resistance selected as key variables for analysis. The accuracy of the proposed method was compared with that of three common distribution network fault identification methods. Figure 5 The diagram shows a comparison of the recognition accuracy of the proposed method with other commonly used methods across different initial phase angle ranges. It can be seen that the recognition accuracy of the proposed method consistently remains above 96% across the entire initial phase angle range, with minimal fluctuation in the curve. Figure 6 The accuracy of the proposed method compared with that of the comparative method is shown across different transition resistance ranges. The accuracy of the proposed method remains consistently above 97%.

[0206] Any technical feature in the embodiment corresponding to any of the above embodiments is also applicable to the subsequent embodiments in this application, and similar aspects will not be repeated hereafter.

[0207] The above describes a method for locating traveling wave faults in a distribution network based on an improved barrel theory optimization algorithm in the embodiments of this application. The following describes the apparatus, system, or device that performs the above method.

[0208] Figure 7 Here is a structural block diagram of the device in one embodiment, with reference to Figure 7 The distribution network traveling wave fault location device based on the improved barrel theory optimization algorithm includes:

[0209] The data item determination module 701 is used to determine the traveling wave energy attenuation matching item and the topology connectivity consistency item based on the basic physical relationship of the traveling wave propagation of the target distribution network to be located for fault location.

[0210] The model building module 702 is used to construct a two-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints based on the traveling wave energy attenuation matching term and the topology connectivity consistency term.

[0211] The model solving module 703 is used to solve the distribution network optimization model based on the improved barrel theory optimization algorithm to obtain the distribution network traveling wave fault location results associated with the target distribution network.

[0212] In this embodiment of the application, based on, as follows Figure 7 The connection relationships between the various modules or units shown in the diagram improve the efficiency and accuracy of traveling wave fault location through the cooperation between these modules or units.

[0213] In another embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, it includes a processor, memory, input / output interfaces, and a communication interface. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface is connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores relevant data. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. The computer program can be executed by the processor to implement the various methods described in the above embodiments.

[0214] In yet another embodiment, a computer device is provided, such as a terminal, whose internal structure diagram may be as follows: Figure 9 As shown, it includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. The computer program can be executed by the processor to implement the various methods described in the above embodiments.

[0215] Those skilled in the art will understand that Figure 8 and Figure 9The structure shown is only a block diagram of a part of the structure related to the present application and does not constitute a limitation on the computer device on which the present application is applied. It may also include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements, in order to realize the function of the computer device.

[0216] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0217] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the systems, devices, equipment, modules or units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0218] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, devices, or methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0219] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0220] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0221] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product.

[0222] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium, or a semiconductor medium (e.g., a solid-state drive), etc.

[0223] The technical solutions provided by the embodiments of this application have been described in detail above. Specific examples have been used in the embodiments of this application to illustrate the principles and implementation methods of the embodiments of this application. The description of the above embodiments is only for the purpose of helping to understand the methods and core ideas of the embodiments of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the embodiments of this application. Therefore, the content of this specification should not be construed as a limitation on the embodiments of this application.

Claims

1. A method for locating traveling wave faults in distribution networks based on an improved barrel theory optimization algorithm, characterized in that, The method includes: Based on the fundamental physical relationship of traveling wave propagation in the target distribution network to be fault located, the traveling wave energy attenuation matching term and topology connectivity consistency term are determined. Based on the traveling wave energy attenuation matching term and the topology connectivity consistency term, a two-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints is constructed. Based on the improved barrel theory optimization algorithm, the distribution network optimization model is solved to obtain the distribution network traveling wave fault location results associated with the target distribution network.

2. The method according to claim 1, characterized in that, The determination of the traveling wave energy attenuation matching term and topology connectivity consistency term based on the fundamental physical relationship of the traveling wave propagation in the target distribution network to be fault located includes: The degree of matching is quantified by summing relative errors to determine the traveling wave energy attenuation matching term; By using the ratio of the actual line length to the straight-line distance, combined with topology validity verification, the degree of fit between the virtual fault point and the actual structure of the distribution network is quantified, and the topology connectivity consistency item is determined.

3. The method according to claim 2, characterized in that, The method of quantifying the degree of matching by summing relative errors to determine the traveling wave energy attenuation matching term includes: Obtain the measured peak energy of the traveling wave at the monitoring point and the theoretical traveling wave energy corresponding to the virtual fault point; The relative energy error is determined based on the measured peak energy of the traveling wave at the monitoring point and the theoretical traveling wave energy corresponding to the virtual fault point. Based on the total number of monitoring points, the relative energy error is summed to obtain the traveling wave energy attenuation matching term.

4. The method according to claim 2, characterized in that, The ratio of actual line length to straight-line distance, combined with topology validity verification, quantifies the fit between virtual fault points and the actual distribution network structure, and determines the topology connectivity consistency item, including: Obtain the actual line length from the fault point to the monitoring point, and the straight-line distance from the fault point to the monitoring point; Based on the total number of monitoring points, the proportional terms of the actual line length from the fault point to the monitoring point and the straight-line distance from the fault point to the monitoring point are summed to determine the proportional summation term; The topology connectivity consistency term is determined based on the proportional summation term and the topology validity indicator function.

5. The method according to claim 1, characterized in that, The process of constructing a two-dimensional fusion distribution network optimization model that conforms to both traveling wave propagation characteristics and distribution network topology constraints, based on the traveling wave energy attenuation matching term and the topology connectivity consistency term, includes: The objective function of the distribution network optimization model is determined based on the traveling wave energy attenuation matching term and the topology connectivity consistency term. Based on the traveling wave energy attenuation matching constraint, the constraint conditions of the power distribution network optimization model are determined; The power distribution network optimization model is determined based on the objective function and the constraints.

6. The method according to claim 5, characterized in that, The step of determining the objective function of the distribution network optimization model based on the traveling wave energy attenuation matching term and the topology connectivity consistency term includes: Obtain dynamic weight coefficients that conform to preset summation rules; Based on the dynamic weighting coefficients, the traveling wave energy attenuation matching term and the topology connectivity consistency term are weighted and summed to obtain the objective function of the distribution network optimization model.

7. The method according to claim 1, characterized in that, The improved barrel theory-based optimization algorithm solves the distribution network optimization model to obtain the traveling wave fault location results associated with the target distribution network, including: Initial data is determined based on data preprocessing and feature extraction techniques; Pseudo-random initialization is performed based on the initial data to obtain the initialization result. The initialization result is then validated for topological validity to obtain a valid initialization result. Based on the effective initialization results, fitness evaluation and weakness identification are performed, global balance update and repair are carried out, and the algorithm iteration process begins. If the algorithm iteration process stalls, differential chaotic perturbation processing is performed based on the perturbation factor; If the iteration termination condition using dual-judgment is met, the algorithm iteration process is terminated and the optimal solution is output to determine the location result of the traveling wave fault in the distribution network.

8. A distribution network traveling wave fault location device based on an improved barrel theory optimization algorithm, characterized in that, The device includes: The data item determination module is used to determine the traveling wave energy attenuation matching item and the topology connectivity consistency item based on the basic physical relationship of the traveling wave propagation of the target distribution network to be located for fault location. The model building module is used to construct a two-dimensional fusion distribution network optimization model that conforms to the traveling wave propagation characteristics and distribution network topology constraints based on the traveling wave energy attenuation matching term and the topology connectivity consistency term. The model solving module is used to solve the distribution network optimization model based on the improved barrel theory optimization algorithm to obtain the distribution network traveling wave fault location results associated with the target distribution network.

9. A computer device, characterized in that, The computer device includes: At least one processor and memory; The memory is used to store program code, and the processor is used to call the program code stored in the memory to execute the method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, It includes instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

Images