Active power distribution network fault positioning method based on improved crown porcupine optimization algorithm

Abstract

The invention relates to the field of power system power distribution network fault positioning, in particular to an active power distribution network fault positioning method based on an improved crown porcupine optimization algorithm. The method comprises the following steps of: according to a topological structure of an active power distribution network, carrying out discretization coding on fault current on section switch nodes, and reading fault current state information of section switches detected by a feeder terminal unit and an actual fault current matrix of an ith switch node; establishing an expected fault current state function of a logic relation between a feeder line section and a section switch state, and establishing an objective function by using a difference value between the state function and an actual state code; and optimizing the target function based on an improved crown porcupine optimization algorithm to obtain a global optimal solution, namely positioning a fault point. Under the condition that different numbers of distributed power supplies at different positions are connected to the power distribution network, the algorithm provided by the invention improves the positioning speed and the positioning precision of the fault section, and provides a powerful guarantee for the reliability of a power system.

Description

Technical Field

[0001] The invention relates to the field of fault location of an active distribution network, and in particular to a fault location method of an active distribution network based on an improved crested porcupine optimization algorithm. Background Art

[0002] Fault location in active distribution networks is one of the key technologies to ensure reliable operation of power grids. At present, most methods for fault location in distribution networks use information collection and feedback based on fault indicators (FID) or feeder terminal units (FTU). However, FTU operates outdoors for a long time and is susceptible to harsh environments and electromagnetic and lightning interference, which often leads to information loss and distortion during information collection and upload, thus affecting the accuracy and reliability of fault location.

[0003] Existing fault location methods often lack sufficient fault tolerance when facing missing or distorted FTU information, resulting in inaccurate positioning results or even failure to complete fault location. Therefore, developing a fault location method that can maintain high accuracy when FTU information is incomplete has important practical application value. Summary of the invention

[0004] In view of the problem of inaccurate positioning caused by missing or distorted FTU information, the present invention provides an active distribution network fault positioning method based on an improved Crested Porcupine Optimizer (CPO). By introducing and improving the Crested Porcupine Optimizer (CPO), rapid and accurate positioning of distribution network faults is achieved.

[0005] The technical solution of the present invention is: a method for locating faults in an active distribution network based on an improved crown porcupine optimization algorithm, comprising: discretizing and encoding the fault current on the segmented switch node according to the topological structure of the active distribution network, reading the fault current state information of the segmented switch element detected by the feeder terminal unit (FTU), and calculating the actual fault current matrix S of the i-th switch node. i The positive direction of the node fault current is specified as follows: when there is no distributed power source in the feeder segment, the direction from the system power source to the load is the positive direction; when there is a distributed power source in the feeder segment, the direction from the system power source to the distributed power source is the positive direction. The fault current encoding of each node is:

[0006]

[0007] Among them, I i Encodes the fault current at node i.

[0008] The expected fault current state function of the logical relationship between the feeder section and the section switch state is established, and the objective function is established using the difference between the state function and the actual state code; for a switch in the feeder section, the feeder section from the system power supply to the switch is defined as the upstream section, and the feeder section from the switch to the distributed power supply or load is defined as the downstream section. The expected fault current state function of each switch node can be expressed as

[0009]

[0010] in, is the expected fault current state function at the Lth section at node i; n1 is the total number of feeder sections upstream of the i-th switch; n2 is the total number of feeder sections downstream of the i-th switch; n DG is the total number of downstream distributed power sources; K j is the jth distributed power switch in the downstream; L j is the fault state of the jth feeder section in the upstream or downstream section. The expected fault current array is generated and the objective function W(S) is obtained:

[0011]

[0012] in, is the total number of switches in the active distribution network; n is the total number of feeder segments in the active distribution network; δ is the weight coefficient, is the sum of the number of all faulty feeder sections.

[0013] Based on the improved crown porcupine optimization algorithm, the objective function is optimized to obtain the global optimal solution, that is, to locate the fault point. This includes: parameter and population initialization. Including the population number N, the maximum number of iterations T_max and the individual position. According to the cyclic population reduction technology, the optimal population size is updated, and the convergence factor, individual quality, average output and random number are set to update the optimal position and fitness of the crown porcupine. The mathematical model of population diversity is as follows:

[0014]

[0015] Where: N' represents the initial population, T is the variable of the number of cycles, and t is the current function evaluation; T max is the maximum number of times the function is evaluated, % represents the remainder or modulus operator; N min is the minimum number of individuals in a newly generated population.

[0016] The exploration phase mainly determines whether to perform global search or local development by comparing the size of two generated random numbers (represented by τ8 and τ9). When τ8<τ9, global exploration behavior is implemented, and the more commonly used normal distribution is used to generate random values ​​(represented by τ6 and τ7). When the random value τ6<τ7, the first defense strategy is implemented, and the individual uses visual intimidation to intimidate the predator. Its mathematical model is:

[0017]

[0018]

[0019] Where: is the new position of the ith individual, is the location information of the i-th individual at the current moment, is the current position of the random individual, is the vector generated between the current individual and the random individual, is the optimal solution for function evaluation t, is the position vector generated between the current individual and a randomly selected individual in the population to represent the position of the predator at iteration t; τ1 is a random number based on a normal distribution, τ2 is a random number between [0,1]; r is a random number between [1,N].

[0020] When τ6≥τ7, the second defense strategy is implemented, and the individual uses noise to drive away the predator. Its mathematical model is:

[0021]

[0022] Where: is a randomly generated binary vector between [0,1], is the location of the predator, is the current position of random individual r1, is the current position of the random individual r2, r1 and r2 are random integers between [1, N], and τ3 is a random value between [0, 1]. Therefore, the second defense strategy uses the approach or distance of the predator to avoid the population from falling into the local optimal problem.

[0023] When τ8 ≥ τ9, the individual implements local development behavior by comparing the random number τ 10 and T f The size of τ determines 10 <T f The third defense strategy is to avoid predators by creating odors. The mathematical model is:

[0024]

[0025] Where: r3 is a random number in [1,N], δ is a parameter that controls the search direction; is the position of the i-th individual at iteration t, γ t is the defense factor, τ3 is a random value between [0,1]; It is the odor diffusion factor.

[0026] When τ 10 ≥T f When , the fourth defense strategy is adopted, and the individual defends by physical attack means. Its mathematical model is:

[0027]

[0028] Where: α is the convergence speed factor; τ4 is a random value in [0,1], is the individual average force affecting the ith predator, given by the law of inelastic collisions.

[0029]

[0030] Where: m i is the mass of the i-th predator at the iteration; is the final velocity of the ith individual at the next iteration t+1, is the initial velocity of the i-th individual at iteration t; τ6 is a random value vector generated between [0,1].

[0031] Substitute the inverse Cauchy distribution into the first defense strategy, and use the tangent flight operator to change the weight distribution of the fourth defense strategy; execute different defense strategies according to the size of the random number comparison, so as to update the population position and fitness, and input the optimal weight into the objective function. Considering the iteration order in the optimization algorithm, the inverse Cauchy cumulative distribution function is first added to the first defense strategy, and the iteration step size is controlled to accelerate the optimization of the fitness value. The improved mathematical model of the first defense strategy is as follows:

[0032]

[0033] Where p is a random number uniformly distributed in [0,1];

[0034] Then, the tangent flight operator is substituted into the fourth defense strategy to adjust the final weight distribution and balance the global search capability and convergence speed, which is expressed as a tangent function with the following formula:

[0035]

[0036] Where v is a random number uniformly distributed in (1, d), and d is the dimension.

[0037] Determine whether the maximum number of iterations has been reached. If the conditions are met, the global best population individual is output, that is, the fault section is located, and accurate positioning is achieved. Otherwise, the position and fitness of the crested porcupine are updated. The minimum value of the objective function value corresponds to the global optimal solution. According to the cycle corresponding to the minimum objective function value, the fault section of the active distribution network is obtained, thereby realizing the fault location of the active distribution network.

[0038] The present invention improves the first and fourth defense strategies in the crown porcupine optimization algorithm, thereby improving the adaptability and fault tolerance of the algorithm in complex environments and achieving rapid and accurate positioning of distribution network faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flow chart of the method of the present invention;

[0040] Figure 2 This is a flow chart of the improved crested porcupine optimization algorithm in the present invention;

[0041] Figure 3 A comparison convergence curve diagram of the multi-point fault location method according to an embodiment of the present invention;

[0042] Figure 4 This is a schematic diagram of the network topology structure of the active power distribution network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] like Figure 1-4 As shown, the present invention proposes a method for locating faults in an active distribution network based on an improved crown porcupine optimization algorithm. The method comprises:

[0045] According to the topological structure of the active distribution network, the fault current on the section switch node is discretized and encoded, and the fault current state information of the section switch detected by the feeder terminal unit (FTU) is read. The actual fault current matrix S of the i-th switch node i ;

[0046] Establishing the expected fault current state function of the logical relationship between the feeder section and the section switch state, and establishing the objective function using the difference between the state function and the actual state code;

[0047] The objective function is optimized based on the improved crested porcupine optimization algorithm to obtain the global optimal solution, that is, to locate the fault point.

[0048] The crowned porcupine algorithm is a novel swarm intelligence optimization algorithm. The present invention applies it to fault location in active distribution networks. While accelerating the convergence speed, it also improves the diversity of the group, which helps to avoid falling into the local minimum. The present invention improves the defect that traditional intelligent algorithms are prone to converge to the local optimum, and improves the efficiency and accuracy of fault location, which is helpful for locating and troubleshooting faults in actual projects.

[0049] In order to improve the applicability of the crested porcupine algorithm in fault location of different active distribution networks, the defense mechanism of the crested porcupine algorithm is improved by using the Cauchy inverse cumulative distribution function and the tangent operator, which improves the optimization performance of the algorithm. In the application, it is proved to have better convergence results through the benchmark test function.

[0050] Fault current discretization encoding methods include:

[0051] Switch node actual fault current array S i The coding scheme of the node fault current stipulates that the positive direction of the node fault current is: when there is no distributed power supply in the feeder segment, the direction from the system power supply to the load is the positive direction; when there is a distributed power supply in the feeder segment, the direction from the system power supply to the distributed power supply is the positive direction. The fault current coding of each node is:

[0052]

[0053] Among them, I i Encodes the fault current at node i.

[0054] For a switch in a feeder section, the feeder section from the system power supply to the switch is defined as the upstream section, and the feeder section from the switch to the distributed power supply or load is defined as the downstream section;

[0055] Then the expected fault current state function of each switch node is for:

[0056]

[0057] in, is the expected fault current state function at the Lth section at node i; n1 is the total number of feeder sections upstream of the i-th switch; n2 is the total number of feeder sections downstream of the i-th switch; n DG is the total number of downstream distributed power sources; K j is the jth distributed power switch in the downstream; L j is the fault status of the jth feeder segment in the upstream or downstream section.

[0058] The objective function W(S) is:

[0059]

[0060] Where n is the total number of feeder segments in the active distribution network; δ is the weight coefficient, is the sum of the number of all faulty feeder sections.

[0061] Specifically, the effectiveness of the improved crested porcupine optimization algorithm proposed in the present invention is first verified. Figure 2 The flowchart of the improved crown porcupine optimization algorithm described in the embodiment of the present invention; the specific steps are as follows:

[0062] Step 1: Parameter and population initialization. The parameters are uniformly set: the population size is 50, the dimension is 30, the maximum number of iterations is 500, and the algorithm test is repeated 20 times.

[0063] Step 2: Calculate the individual fitness of the crested porcupine. Update the optimal population size based on the cyclic population reduction technique, set the convergence factor, individual quality, average output and random number, and update the optimal position and fitness of the crested porcupine. The test function uses a multi-peak function The dimension is 30, the search range is [-100, 100], where F3 is the actual output result, x i is the i-th sample, n1 is the total number of samples, which is 500 in total.

[0064] According to the cyclic population reduction technique, the mathematical model of population diversity is as follows:

[0065]

[0066] Where: N' represents the initial population, T is the variable of the number of cycles, and t is the current function evaluation; T max is the maximum number of times the function is evaluated, % represents the remainder or modulus operator; N min is the minimum number of individuals in a newly generated population.

[0067] By comparing the size of the two generated random numbers (represented by τ8 and τ9), the two behaviors of global search or local development are decided. When τ8<τ9, global exploration behavior is implemented, and the more commonly used normal distribution is used to generate random values ​​(represented by τ6 and τ7). When the random value τ6<τ7, the first defense strategy is implemented, and the individual uses visual intimidation to intimidate the predator. Its mathematical model is:

[0068]

[0069]

[0070] Where: is the new position of the ith individual, is the location information of the i-th individual at the current moment, is the current position of the random individual, is the vector generated between the current individual and the random individual, is the optimal solution for function evaluation t, is the position vector generated between the current individual and a randomly selected individual in the population to represent the position of the predator at iteration t; τ1 is a random number based on a normal distribution, τ2 is a random number between [0,1]; r is a random number between [1,N].

[0071] When τ6≥τ7, the second defense strategy is implemented, and the individual uses noise to drive away the predator. Its mathematical model is:

[0072]

[0073] Where: is a randomly generated binary vector between [0,1], is the location of the predator, is the current position of the random individual r1, r1 and r2 are random integers between [1, N], and τ3 is a random value between [0, 1]. Therefore, the second defense strategy uses the approach or distance of the predator to avoid the population from falling into the local optimal problem.

[0074] When τ8 ≥ τ9, the individual implements local development behavior by comparing the random number τ 10 and T f The size of τ determines 10 <T f The third defense strategy is to avoid predators by creating odors. The mathematical model is:

[0075]

[0076] Where: r3 is a random number in [1,N], δ is a parameter that controls the search direction; is the position of the i-th individual at iteration t, γ t is the defense factor, τ3 is a random value between [0,1]; It is the odor diffusion factor.

[0077] When τ 10 ≥T f When , the fourth defense strategy is adopted, and the individual defends by physical attack means. Its mathematical model is:

[0078]

[0079] Where: α is the convergence speed factor; τ4 is a random value in [0,1], τ5 is a random value in [0,1], is the individual average force affecting the ith predator, given by the law of inelastic collisions.

[0080]

[0081] Where: m i is the mass of the i-th predator at the iteration; is the final velocity of the ith individual at the next iteration t+1, is the initial velocity of the i-th individual at iteration t; τ6 is a random value vector generated between [0,1].

[0082] Step 3: Substitute the inverse Cauchy distribution into the first defense strategy, and use the tangent flight operator to change the weight distribution of the fourth defense strategy; execute different defense strategies according to the size of the random number comparison, and update the population position and fitness. Considering the iteration order in the optimization algorithm, the inverse Cauchy cumulative distribution function is first added to the first defense strategy, and the iteration step size is controlled to accelerate the optimization of the fitness value. The improved mathematical model of the first defense strategy is as follows:

[0083]

[0084] Where p is a random number uniformly distributed in [0,1];

[0085] Then, the tangent flight operator is substituted into the fourth defense strategy to adjust the final weight distribution and balance the global search capability and convergence speed, which is expressed as a tangent function with the following formula:

[0086]

[0087] Where v is a random number uniformly distributed in (1, d), and d is the dimension.

[0088] Step 4: Determine whether the current solution is better than the optimal solution. If so, replace the optimal solution with the current solution; otherwise, continue to keep the previous optimal solution.

[0089] Step 5: Determine whether the maximum number of iterations has been reached. If the condition is met, output the optimal solution, otherwise jump to step 3 to continue updating.

[0090] In order to verify and intuitively express the convergence performance of the improved crested porcupine algorithm, two classic optimization algorithms are selected to compare with the algorithm before improvement, namely, the whale optimization algorithm (WOA), the Harris Hawk optimization algorithm (HHO) and the crested porcupine algorithm (CPO). The improved CPO algorithm is named CTCPO. The test results of the multimodal function F3 are shown in Figure 2. Figure 3 As shown in the figure, it can be seen from the convergence curve of the function that the CTCPO algorithm shows a faster convergence speed in the multi-peak test function F3. Compared with the algorithm before improvement, its parameter optimization process is more efficient and stable, and the effectiveness of its algorithm advantage is proved.

[0091] Build a mathematical model of the IEEE 33-node active distribution network structure, and its structure diagram is as follows Figure 4 As shown, L1—L 33 There are 33 feeder segments, S1-S 33 There are 33 switch nodes, and K1-K3 are the access switches of 3 distributed power sources. Due to the access of distributed power sources, the topology of the distribution network and the complexity of fault location are increased. The expected fault current state function of the logical relationship between the feeder section and the segmented switch state is established, and the objective function W(S) is established using the difference between the state function and the actual state code; for a switch in the feeder section, the feeder section between the system power supply and the switch is defined as the upstream section, and the feeder section between the switch and the distributed power source or load is defined as the downstream section.

[0092] For multi-point fault location simulation, assume [K1, K2, K3] = [0, 1, 0], segment L 14 ,L 20 ,L 31 When a fault occurs, the information reported by the FTU is [1,1,1,1,1,1,1,1,1,1,1,1,-1,-1,-1,1,1,1,0,0,0,0,0,0,0,0,0,0,1,1,0,0] and the output positioning result is [0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0], which enables accurate identification of the faulty section.

[0093] In summary, in the fault location of active distribution networks, the proposed active distribution network fault location method based on the improved crown porcupine optimization algorithm can achieve accurate location of multi-point faults, and has a faster solution speed and better optimization performance than other traditional optimization algorithms.

[0094] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for locating faults in active distribution networks based on an improved crown porcupine optimization algorithm, characterized in that: The method includes: According to the topological structure of the active distribution network, the fault current on the section switch node is discretized and encoded, and the fault current state information of the section switch detected by the feeder terminal unit is read, and the actual fault current matrix of the i-th switch node; Establishing the expected fault current state function of the logical relationship between the feeder section and the section switch state, and establishing the objective function using the difference between the state function and the actual state code; The objective function is optimized based on the improved crested porcupine optimization algorithm to obtain the global optimal solution, that is, to locate the fault point.

2. The method for locating faults in an active distribution network based on an improved crown porcupine optimization algorithm according to claim 1, characterized in that: Fault current discretization encoding methods include: Switch node actual fault current array S i The coding scheme of the node fault current stipulates that the positive direction of the node fault current is: when there is no distributed power supply in the feeder segment, the direction from the system power supply to the load is the positive direction; when there is a distributed power supply in the feeder segment, the direction from the system power supply to the distributed power supply is the positive direction. The fault current coding of each node is: Among them, I i Encodes the fault current at node i.

3. The method for locating faults in an active distribution network based on an improved crown porcupine optimization algorithm according to claim 2 is characterized in that: For a switch in a feeder section, the feeder section from the system power supply to the switch is defined as the upstream section, and the feeder section from the switch to the distributed power supply or load is defined as the downstream section; Then the expected fault current state function of each switch node is: in, is the expected fault current state function at the Lth section at node i; n1 is the total number of feeder sections upstream of the i-th switch; n2 is the total number of feeder sections downstream of the i-th switch; n DG is the total number of downstream distributed power sources; K j is the jth distributed power switch in the downstream; L j is the fault status of the jth feeder segment in the upstream or downstream section.

4. The method for locating faults in an active distribution network based on an improved crown porcupine optimization algorithm according to claim 3 is characterized in that: The objective function W(S) is: in, is the total number of switches in the active distribution network; n is the total number of feeder segments in the active distribution network; δ is the weight coefficient, is the sum of the number of all faulty feeder sections.

5. The method for locating faults in an active distribution network based on an improved crested porcupine optimization algorithm according to claim 1 is characterized in that: The improved crown porcupine optimization algorithm includes: a method for locating faults in active distribution networks based on the improved crown porcupine optimization algorithm Parameters and population initialization: including population number N, maximum number of iterations T_max and individual positions; Calculate the individual fitness of crested porcupines: Update the optimal population size based on the cyclic population reduction technique, set the convergence factor, individual quality, average output and random number, and update the optimal position and fitness of crested porcupines; Substitute the inverse Cauchy distribution into the first defense strategy, and use the tangent flight operator to change the weight distribution of the fourth defense strategy; execute different defense strategies according to the size of random number comparison, so as to update the population position and fitness; Determine whether the current solution is better than the optimal solution. If so, replace the optimal solution with the current solution; otherwise, continue to keep the previous optimal solution. Determine whether the maximum number of iterations has been reached; if the condition is met, output the global best population individual, that is, locate the fault section, otherwise continue to update the crested porcupine position and fitness.

6. The method for locating faults in an active distribution network based on an improved crested porcupine optimization algorithm according to claim 1, characterized in that: According to the cyclic population reduction technique, the mathematical model of population diversity is as follows: Where: N' represents the initial population, T is the variable of the number of cycles, and t is the current function evaluation; T max is the maximum number of times the function is evaluated, % represents the remainder or modulus operator; N min is the minimum number of individuals in a newly generated population.

7. The method for locating faults in an active distribution network based on an improved crested porcupine optimization algorithm according to claim 1 is characterized in that: In calculating the individual fitness of crested porcupines, By comparing the sizes of the two generated random numbers τ8 and τ9, the two behaviors of global search or local development are decided; when τ8<τ9, global exploration behavior is implemented, and the random values ​​τ6 and τ7 are generated using normal distribution; when the random value τ6<τ7, the first defense strategy is implemented, and the individual uses visual intimidation to the predator. The mathematical model is: Where: is the new position of the ith individual, is the location information of the i-th individual at the current moment, is the current position of the random individual, is the vector generated between the current individual and the random individual, is the optimal solution for function evaluation t, is the position vector generated between the current individual and a randomly selected individual in the population to represent the position of the predator at iteration t; τ1 is a random number based on a normal distribution, and τ2 is a random number between [0,1]; r is a random number between [1,N]; When τ6≥τ7, the second defense strategy is implemented, and the individual uses noise to drive away the predator. Its mathematical model is: Where: is a randomly generated binary vector between [0,1], is the location of the predator, is the current position of random individual r1, is the current position of the random individual r2, r1 and r2 are random integers between [1, N], and τ3 is a random value between [0, 1]; When τ8 ≥ τ9, the individual implements local development behavior by comparing the random number τ 10 and T f The size of τ determines 10 <T f The third defense strategy is to avoid predators by creating odors. The mathematical model is: Where: r3 is a random number in [1,N], δ is a parameter that controls the search direction; is the position of the i-th individual at iteration t, γ t is the defense factor, τ3 is a random value between [0,1]; is the odor diffusion factor; When τ 10 ≥T f When , the fourth defense strategy is adopted, and the individual defends by physical attack means. Its mathematical model is: Where: α is the convergence speed factor; τ4 is a random value in [0,1], τ5 is a random value in [0,1], is the average force of the individuals affecting the ith predator; where Where: m i is the mass of the i-th predator at the iteration; is the final velocity of the ith individual at the next iteration t+1, is the initial velocity of the i-th individual at iteration t; τ6 is a random value vector generated between [0,1].

8. The method for locating faults in an active distribution network based on an improved crested porcupine optimization algorithm according to claim 7 is characterized in that: The Cauchy inverse cumulative distribution function is added to the first defense strategy, and the iteration step size is controlled to accelerate the optimization of the fitness value. The improved mathematical model of the first defense strategy is as follows: Where p is a random number uniformly distributed in [0,1]; Then, the tangent flight operator is substituted into the fourth defense strategy to adjust the final weight distribution to balance the global search capability and convergence speed, which is expressed as a tangent function with the following formula: Where v is a random number uniformly distributed in (1, d), and d is the dimension.

9. The method for locating faults in an active distribution network based on an improved crested porcupine optimization algorithm according to claim 1, characterized in that: The minimum value of the objective function corresponds to the global optimal solution. According to the cycle corresponding to the minimum objective function value, the fault section of the active distribution network is obtained, thereby realizing the fault location of the active distribution network.

Images