AOA-based fault positioning method for power distribution network containing distributed power supply

Abstract

The invention discloses an AOA-based fault positioning method for a power distribution network containing a distributed power supply, and belongs to the technical field of power system fault diagnosis. The method comprises the following steps: firstly, discretizing a power distribution network topology into feeder line sections, generating an actual fault current array by using fault current information collected by a feeder line terminal unit, and establishing a switch expectation function to associate a feeder line section fault state with an expectation fault current array; secondly, constructing an evaluation function based on the difference degree between the actual fault current array and the expected fault current array; and finally, performing optimization solution by adopting an improved Archimedes optimization algorithm. The method provided by the invention effectively solves the problems that the tidal current direction is changeable due to the access of the distributed power supply to the power distribution network and the adaptability of the traditional method is insufficient, and realizes rapid and accurate fault section positioning in single-point and multi-point fault scenes.

Description

Technical Field

[0001] This invention belongs to the field of power system fault diagnosis technology, and in particular relates to a fault location method for distribution networks with distributed power sources based on AOA. Background Technology

[0002] Accurate fault location in distribution networks is crucial for ensuring power supply reliability and rapid recovery. In the context of energy transition, renewable energy sources such as solar power, with their flexibility and economic advantages, are being integrated into distribution networks on a large scale as distributed generation, leading to increasingly complex system topologies. The randomness and intermittency of distributed generation output significantly reduce the accuracy of traditional fault location methods. Therefore, fault location in distribution networks containing distributed generation has become a research hotspot in the field of distribution networks.

[0003] When a phase-to-phase short-circuit fault occurs in a distribution network, the fault characteristics are obvious. Accurate fault information can be collected and extracted using feeder terminal units (FTUs), and the fault section can be located using a fast and fault-tolerant localization algorithm, which is currently the main research approach. The main process involves uploading the fault current information detected by the FTU to the supervisory control and data acquisition (SCADA) system, where intelligent algorithms calculate the fault section. Currently, distribution network fault location methods are becoming increasingly diversified, but the large-scale integration of distributed generation (DG) leads to system topology complexity and variable power flow directions, posing a significant challenge to traditional location technologies. Mainstream intelligent optimization algorithms achieve location by constructing fault state evaluation functions, but differences in algorithm architecture result in inconsistent convergence efficiency and fault tolerance performance, necessitating improvements in core algorithms to enhance location accuracy and speed. Existing optimization algorithms (such as genetic algorithms and particle swarm optimization) possess a certain degree of fault tolerance, but they often face problems such as requiring large populations, multiple iterations, or being prone to getting trapped in local optima.

[0004] Therefore, there is an urgent need for a new technical solution to address this problem. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a fault location method for distribution networks with distributed power sources based on AOA, so as to solve the problem of poor location accuracy and speed in the existing fault location methods for distribution networks with distributed power sources.

[0006] The technical solution adopted in this invention is to provide a fault location method for a distribution network with distributed power sources based on AOA, comprising the following steps:

[0007] Step 1: Fault Information Acquisition and Discretization Encoding: The distribution network is abstracted into multiple feeder segments divided by switch nodes. A binary-coded individual position vector is generated using the fault status of each feeder segment as an element. Based on the actual fault current information reported by the feeder terminal units (FTUs) installed on the distribution network, an actual fault current array is generated. Based on the individual position vectors and the preset switch expectation function, an expected fault current array is generated.

[0008] Step 2: Evaluation function design: Construct a fitness function, which is used to calculate the degree of difference between the actual fault current array and the expected fault current array;

[0009] Step 3: Fault location using Archimedes optimization algorithm: The fitness function is optimized using an improved Archimedes optimization algorithm, including initializing the population, iteratively updating the density, volume, acceleration and position vector of individuals, and converting the continuous position vector into binary code through discrete mapping; when the iteration termination condition is met, the globally optimal individual position vector is output, and its binary code corresponds to the located fault feeder segment.

[0010] The encoding rule for the actual fault current array in step 1 is as follows: for any switching node:

[0011] If no fault current is detected or cannot be detected, the code is 0;

[0012] If a fault current is detected and its direction is the same as the specified positive direction, it is coded as 1;

[0013] If a fault current is detected and its direction is opposite to the specified positive direction, it is coded as -1;

[0014] The specified positive direction is as follows: when there is no distributed power source in the feeder segment, the direction is from the system power source to the electrical load; when there is a distributed power source in the feeder segment, the direction is from the system power source to the distributed power source.

[0015] The specific expression for the switch expectation function in step 1 is as follows:

[0016] ;

[0017] In the formula: n1 is the total number of feeder segments upstream of the i-th switch; n2 is the total number of feeder segments downstream of the i-th switch; n DG K represents the total number of downstream distributed power sources. j For the j-th distributed power switch downstream; L j This represents the fault status of the j-th feeder segment in the upstream or downstream section.

[0018] The specific expression for the fitness function in step 2 is as follows:

[0019] ;

[0020] In the formula: W is a maximum number; N is the total number of switches in the distribution network containing distributed generation; S i This refers to the actual fault current array information of the i-th switching node detected by the FTU device. This represents the expected fault current state array information for each switching node; n is the total number of feeder segments in the distribution network containing distributed generation. These are preset weighting coefficients; This is the sum of the number of all faulty feeder segments.

[0021] The discretization process of the improved Archimedes optimization algorithm in step 3 is to use the sigmoid function to map the continuous position vector to discrete values ​​of 0 or 1.

[0022] In step 3, the optimization solution balances the global exploration and local exploitation capabilities of the algorithm by setting the transfer operator TF and the density factor d. The calculation formulas for the transfer operator TF and the density factor d are as follows:

[0023] ;

[0024] Where t represents the current iteration number, T max This represents the maximum number of iterations.

[0025] The process of iteratively updating the individual position in step 3 includes:

[0026] If TF ≤ 0.5, perform global exploration. The update formulas for individual acceleration acc and position x are:

[0027] ;

[0028] ;

[0029] ;

[0030] If TF > 0.5, perform local development, and the update formulas for individual acceleration acc and position x are:

[0031] ;

[0032] ;

[0033] in, In generation t+1, i thThe acceleration of an individual is represented by den, vol, and acc, which represent the density, volume, and acceleration of the individual, respectively. The subscript mr represents a random individual, the subscript best represents the current best individual, C1 and C2 are constants, rand is a random number in [0,1], F is the direction factor, and T is a constant.

[0034] The direction factor F is calculated as follows:

[0035] ;

[0036] in, , It is a constant.

[0037] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a fault location method for a distribution network with distributed power sources based on AOA.

[0038] Through the above design scheme, the present invention can bring the following beneficial effects:

[0039] This invention improves the standard continuous domain Archimedes Optimization Algorithm (AOA) through discretization (e.g., using sigmoid mapping), adapting it to the binary encoding characteristics of distribution network fault states. This enables fault location in both single-point and multi-point fault scenarios. Compared to traditional algorithms (such as genetic algorithms and particle swarm optimization), the AOA algorithm converges to the optimal solution faster, effectively improving the speed of fault location and shortening the overall fault handling time, thus increasing the efficiency of power restoration. Furthermore, it fully considers the variability of distribution network topology and power flow direction after distributed generation (DG) integration. By explicitly defining the positive direction of the fault current and constructing the corresponding switch expectation function, the method can flexibly adapt to complex scenarios with different numbers and locations of DG integration. This method demonstrates good location performance in distribution networks with no DG, partial DG integration, or full DG integration. This invention successfully transfers the advantages of the AOA algorithm in engineering optimization problems to the field of power system fault diagnosis, providing a new solution to the problem of fault location in active distribution networks. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the overall process of the fault location method for distribution networks with distributed power sources based on AOA according to the present invention.

[0041] Figure 2 This is an IEEE-33 node power system diagram for the fault location method of a distribution network with distributed power sources based on AOA according to the present invention. Detailed Implementation

[0042] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0043] The fault location method for distribution networks with distributed power sources based on AOA of the present invention includes the following steps:

[0044] Step 1: Fault information acquisition and discretization coding.

[0045] Step 2: Design the evaluation function.

[0046] Step 3: The Archimedes optimization algorithm is used to locate the fault.

[0047] The method for fault information acquisition and discretization coding in step 1 is as follows:

[0048] The distribution network is divided into several feeder segments by multiple nodes, using components such as sectionalizing switches, tie switches, and circuit breakers as nodes. The fault operation state of each feeder segment (0 for normal, 1 for fault) serves as the position vector of an individual in the Archimedes optimization algorithm. Based on this position vector, the expected fault current array for each node is generated. When a fault occurs in a feeder segment of the distribution network, the FTU (Fault Transfer Unit) detects the actual fault current state of each node and reports it to generate the actual fault current array. A fitness function is constructed based on the difference between the expected fault current array corresponding to each individual and the fault current array actually reported by the FTU, serving as the iterative basis for the Archimedes optimization algorithm. During the algorithm iteration, the individual position (i.e., the feeder segment state) is updated by simulating the interaction of density, volume, and acceleration of an object in a fluid. After operations such as density adaptive update, position vector discrete mapping, and acceleration guidance, if the termination condition is met, the globally optimal individual is output, and its position vector represents the fault state of each feeder segment in the current distribution network, thereby achieving the location of the faulty feeder segment.

[0049] The standard Archimedes optimization algorithm is only applicable to continuous domain problems, while the distribution network fault location problem is based on the discrete domain. Therefore, the Archimedes optimization algorithm is discretized, and its individuals are encoded in binary. Each dimension of the position component of each individual corresponds to the fault state of a feeder segment: the value is 1 when the feeder segment is faulty, and 0 when there is no fault. Therefore, an individual's position vector can be represented as L1, L2, ..., Ln, where Ln takes the value 0 or 1.

[0050] When distributed generation (DG) sources are connected to the distribution network, the network topology, power flow distribution, and power flow direction of the entire distribution network will change. To adapt to this change, the positive direction of the node fault current detected by the FTU device is specified as follows: when there is no DG source in the feeder segment, the direction from the system power source to the load is the positive direction; when there is a DG source in the feeder segment, the direction from the system power source to the DG source is the positive direction. According to this specification, there are three cases for encoding the node fault current, as shown in Table 1.

[0051] Table 1. Coding of node fault currents

[0052]

[0053] The expected fault state of a switching node is closely related to the fault states of its upstream and downstream feeder segments, and the state of the distributed power supply switches in the downstream segment. A switching expectation function is established. For a given switch in a feeder segment, the feeder segment from the system power source to that switch is defined as the upstream segment, and the feeder segment from that switch to the distributed power supply or load is defined as the downstream segment. Then, the expected fault current state function for each switching node is... It can be represented as follows:

[0054] ;

[0055] In the formula: n1 is the total number of feeder segments upstream of the i-th switch; n2 is the total number of feeder segments downstream of the i-th switch; n DG K represents the total number of downstream distributed power sources. j For the j-th distributed power switch downstream; L j This represents the fault state of the j-th feeder segment in the upstream or downstream section. The desired fault current array is then generated from this.

[0056] The method for designing the evaluation function in step 2 is as follows:

[0057] When a fault occurs in a feeder section of a distribution network containing distributed generation, its fitness is represented by the difference between the actual fault current array detected by the FTU equipment and the expected function array of the switch. Only when the actual fault current information and the expected function value are most similar can the difference between the two be minimized to obtain the optimal solution. The expansion rate function can be expressed as follows:

[0058] ;

[0059] In the formula: W is a maximum number (taken as 10). 2 To ensure the evaluation function remains positive, the problem is transformed into a maximum value optimization problem; N represents the total number of switches in a distribution network containing distributed generation sources; S i This refers to the actual fault current array information of the i-th switching node detected by the FTU device. This represents the expected fault current state array information for each switching node; n is the total number of feeder segments in the distribution network containing distributed generation. The weighting coefficient is set to 0.5 in this paper to prevent misjudgment. This is the sum of the number of all faulty feeder segments.

[0060] The method for fault location using the Archimedes optimization algorithm in step 3 is as follows:

[0061] The Archimedes Optimization Algorithm, inspired by Archimedes' principle, simulates the process by which an object immersed in a liquid continuously adjusts its properties through collisions or buoyancy to eventually reach an equilibrium state. It is a population-based optimization algorithm. In the initialization phase, the algorithm randomly generates an initial population of a certain size, evaluates it, and selects the optimal individual to guide the population. Next, it sets transition operators and a density factor to control and balance the global exploration and local development phases. The algorithm continuously updates the individual positions and the optimal solution until the termination condition is met.

[0062] During the initialization phase, the AOA initialization population location is shown in the following formula:

[0063] ;

[0064] In the formula [lb i ,ub i [] indicates the search space.

[0065] The density (den), volume (vol), and acceleration (acc) of each individual are initialized as shown in the following equation:

[0066] ;

[0067] At this stage, AOA will evaluate the population and select the current best individual (x). best ), the density of the optimal individual (den) best ), volume (vol) best ) and the acceleration of the optimal individual (acc) best And using the density and volume of the optimal individual, update the volume and density of other individuals:

[0068] ;

[0069] Next, we set the transfer operator TF and the density factor d:

[0070] ;

[0071] Where t represents the current iteration number, T maxThis represents the maximum number of iterations.

[0072] Global exploration phase (collisions occur between objects):

[0073] If TF≤0.5, AOA performs a global exploration and updates the individual's acceleration as shown in the following formula:

[0074] ;

[0075] in, In generation t+1, i th Individual acceleration, and Let represent the density and volume of random individuals in the current iteration.

[0076] AOA normalizes individual acceleration according to the following formula and is used for updating individual position:

[0077] ;

[0078] in, In generation t+1, i th Acceleration after individual normalization; u and l are used to adjust the range of normalization.

[0079] At this stage, the individual position is updated as follows:

[0080] ;

[0081] Where C1 is a fixed constant, Random numbers.

[0082] Partial development phase (no collisions between objects):

[0083] If TF > 0.5, AOA performs localized development, updating the individual's acceleration according to the following formula and normalizing the acceleration:

[0084] ;

[0085] At this stage, the individual position is updated using the following formula:

[0086] ;

[0087] Where C2 is a fixed constant; ,and F is the direction factor, used to determine the direction of position update during iteration, and is defined as:

[0088] ;

[0089] in, , It is a fixed constant.

[0090] The flowchart of this invention is as follows Figure 1 As shown, this simulation experiment is based on the Matlab platform. A mathematical model of the IEEE-33 node distribution network structure with DG is built to verify the effectiveness of the method of the present invention. Its system structure is as follows. Figure 2 As shown, the AOA algorithm was used to conduct multiple single-point and multi-point fault comparison simulation experiments when distributed power sources were connected to the distribution network.

[0091] (1) Single point of failure

[0092] For a multi-distributed power source distribution network model, single-point fault location is simulated. [K1,K2,K3]=[1,1,1], that is, distributed power sources are connected to the distribution network. If a fault occurs in section L28, the information reported by the FTU device is [1,1,1,1,1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,1,1,1,1,1,1,1,1,1,-1,0,0,0,0], and the output location result is [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0], and the location result is L28.

[0093] (2) Multiple points of failure

[0094] Simulations were performed on multi-point fault location. [K1,K2,K3]=[1,1,1], meaning that all distributed power sources are connected to the distribution network. If faults occur in sections L10, L28, and L32, the information reported by the FTU device is [1,1,1,1,1,1,1,1,1,1,1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,1,1,1,1,1,1,1,-1,1,1,1,0], and the output location result is [0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,1,0], and the location result is L10, L28, and L32.

[0095] The implementation of the present invention is not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and are included within the protection scope of the present invention.

Claims

1. A fault location method for distribution networks with distributed generation based on AOA, characterized in that: Includes the following steps: Step 1: Fault Information Acquisition and Discretization Encoding: The distribution network is abstracted into multiple feeder segments divided by switch nodes. A binary-coded individual position vector is generated using the fault status of each feeder segment as an element. Based on the actual fault current information reported by the feeder terminal units (FTUs) installed on the distribution network, an actual fault current array is generated. Based on the individual position vectors and the preset switch expectation function, an expected fault current array is generated. Step 2: Evaluation function design: Construct a fitness function, which is used to calculate the degree of difference between the actual fault current array and the expected fault current array; Step 3: Fault location using Archimedes optimization algorithm: The fitness function is optimized using an improved Archimedes optimization algorithm, including initializing the population, iteratively updating the density, volume, acceleration and position vector of individuals, and converting the continuous position vector into binary code through discrete mapping; when the iteration termination condition is met, the globally optimal individual position vector is output, and its binary code corresponds to the located fault feeder segment.

2. The fault location method for a distribution network with distributed power sources based on AOA according to claim 1, characterized in that: The encoding rule for the actual fault current array in step 1 is as follows: for any switching node: If no fault current is detected or cannot be detected, the code is 0; If a fault current is detected and its direction is the same as the specified positive direction, it is coded as 1; If a fault current is detected and its direction is opposite to the specified positive direction, it is coded as -1; The specified positive direction is as follows: when there is no distributed power source in the feeder segment, the direction is from the system power source to the electrical load; when there is a distributed power source in the feeder segment, the direction is from the system power source to the distributed power source.

3. The fault location method for a distribution network with distributed power sources based on AOA according to claim 1, characterized in that: The specific expression for the switch expectation function in step 1 is as follows: ； In the formula: n1 is the total number of feeder segments upstream of the i-th switch; n2 is the total number of feeder segments downstream of the i-th switch; n DG K represents the total number of downstream distributed power sources. j For the j-th distributed power switch downstream; L j This represents the fault status of the j-th feeder segment in the upstream or downstream section.

4. The fault location method for a distribution network with distributed power sources based on AOA according to claim 1, characterized in that: The specific expression for the fitness function in step 2 is as follows: ； In the formula: W is a maximum number; N is the total number of switches in the distribution network containing distributed generation; S i This refers to the actual fault current array information of the i-th switching node detected by the FTU device. This represents the expected fault current state array information for each switching node; n is the total number of feeder segments in the distribution network containing distributed generation. These are preset weighting coefficients; This is the sum of the number of all faulty feeder segments.

5. The fault location method for a distribution network with distributed power sources based on AOA according to claim 1, characterized in that: The discretization process of the improved Archimedes optimization algorithm in step 3 is to use the sigmoid function to map the continuous position vector to discrete values ​​of 0 or 1.

6. The fault location method for a distribution network with distributed power sources based on AOA according to claim 1, characterized in that: In step 3, the optimization solution balances the global exploration and local exploitation capabilities of the algorithm by setting the transfer operator TF and the density factor d. The calculation formulas for the transfer operator TF and the density factor d are as follows: ； Where t represents the current iteration number, T max This represents the maximum number of iterations.

7. The fault location method for a distribution network with distributed power sources based on AOA according to claim 6, characterized in that: The process of iteratively updating the individual position in step 3 includes: If TF ≤ 0.5, perform global exploration. The update formulas for individual acceleration acc and position x are: ； ； ； If TF > 0.5, perform local development, and the update formulas for individual acceleration acc and position x are: ； ； in, In generation t+1, i th The acceleration of an individual is represented by den, vol, and acc, which represent the density, volume, and acceleration of the individual, respectively. The subscript mr represents a random individual, the subscript best represents the current best individual, C1 and C2 are constants, rand is a random number in [0,1], F is the direction factor, and T is a constant.

8. The fault location method for a distribution network with distributed power sources based on AOA according to claim 7, characterized in that: The direction factor F is calculated as follows: ； in, , It is a constant.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1 to 8.

Images