CN110687397A - Active power distribution network fault positioning method based on improved artificial fish swarm algorithm - Google Patents

Active power distribution network fault positioning method based on improved artificial fish swarm algorithm Download PDF

Info

Publication number
CN110687397A
CN110687397A CN201910942736.3A CN201910942736A CN110687397A CN 110687397 A CN110687397 A CN 110687397A CN 201910942736 A CN201910942736 A CN 201910942736A CN 110687397 A CN110687397 A CN 110687397A
Authority
CN
China
Prior art keywords
distribution network
power distribution
fault
state
food
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201910942736.3A
Other languages
Chinese (zh)
Other versions
CN110687397B (en
Inventor
胡珏
韦钢
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Electric Power University
Original Assignee
Shanghai Electric Power University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Electric Power University filed Critical Shanghai Electric Power University
Priority to CN201910942736.3A priority Critical patent/CN110687397B/en
Publication of CN110687397A publication Critical patent/CN110687397A/en
Application granted granted Critical
Publication of CN110687397B publication Critical patent/CN110687397B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
    • G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/088—Aspects of digital computing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y04S10/52—Outage or fault management, e.g. fault detection or location

Abstract

The invention relates to an active power distribution network fault positioning method based on an improved artificial fish swarm algorithm, which comprises the following steps of: 1) according to the improved integer programming model of the power distribution network after the fault, the power distribution network is equivalent to a fault vector consisting of three integers; 2) based on an improved artificial fish swarm algorithm, generating an initial fish swarm by using the fault vector obtained in the step 1), wherein each fish represents an expected state of each section of the power distribution network, and starting iterative optimization; 3) and evaluating the quality of the position of the artificial fish through the food concentration, updating iteration until a finishing condition is met, and acquiring the positioned fault position. Compared with the prior art, the method has the advantages of being suitable for the problem of fault location of the active power distribution networks of different scales, having higher fault tolerance capability on distortion information and the like.

Description

Active power distribution network fault positioning method based on improved artificial fish swarm algorithm
Technical Field
The invention relates to a fault positioning method for an active power distribution network of distributed energy, in particular to a fault positioning method for an active power distribution network based on an improved artificial fish swarm algorithm.
Background
With the rapid development of distributed power generation technology, Distributed Generators (DG) are enabled to supply power to loads or grids. It can be seen that the traditional power distribution network is changed into an active power distribution network with a more complex structure and a non-single trend direction and containing distributed power supplies. Due to the access of the DGs, the fault characteristics are greatly different from those of the traditional power distribution network after the fault occurs, so that the original fault positioning method of the power distribution network is not applicable any more. Therefore, it is of practical significance to find a quick and accurate fault positioning method suitable for the active power distribution network. The active power distribution network fault location based on the artificial intelligence algorithm is characterized in that firstly, fault vectors are constructed according to an integer programming model of the power distribution network faults, then the artificial intelligence algorithm is used for solving, and finally fault sections are found.
In the current stage, the problem of power distribution network fault location after the DG is connected mainly includes: 1. the access position of the DG influences the power flow direction of the power distribution network after the fault, and the access capacity of the DG directly influences the contribution capacity of the DG to the fault current. 2. The DG's contribution has uncertainty and the fault signature varies widely due to the different types of DG. 3. The net rack of the low-voltage distribution network is complex, the number of branches is large, and the distribution of line parameters is uneven.
At present, the fault location research on an active power distribution network containing distributed energy is less, and the difficulty mainly lies in that: 1) the fault is positioned based on the fault characteristics and the fault quantity by using the traditional method, the fault positioning accuracy is reduced when the distribution network has more branches, different fault characteristics of DGs cannot be fully considered, and the influence of transition resistance is large. 2) The fault location based on the non-sound information mainly utilizes an artificial intelligence algorithm to solve, the existing research is difficult to have high fault tolerance rate on the distorted fault information, and the algorithm with strong randomness is easy to be trapped into local optimization, so that the location failure is caused.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide the active power distribution network fault positioning method which is high in fault tolerance rate, suitable for active power distribution networks of different scales and based on the improved artificial fish swarm algorithm.
The purpose of the invention can be realized by the following technical scheme:
an active power distribution network fault positioning method based on an improved artificial fish swarm algorithm comprises the following steps:
step one, according to an improved integer programming model of the power distribution network after the fault, the power distribution network is equivalent to a fault vector formed by three integers.
The equivalent of the power distribution network is that the main content of a fault vector formed by three integers is as follows:
an integer programming model based on power distribution network fault location takes a circuit breaker, a section switch and a tie switch as nodes, takes feeder line state information as an optimization variable, takes the direction of a system power supply flowing to a user and a DG as a positive direction, and when the fault current direction uploaded by an FTU is the same as the positive direction, a switch state value S is obtainedjIf the fault current direction is opposite to the positive direction, S is 1j-1, if no fault current is detected, S j0; when a fault occurs, according to the current out-of-limit information uploaded by the FTU, namely the actual state value of each switch, the power distribution network architecture can be equivalent to an integer expressed fault vector, and the dimension of the vector is the total number of power distribution network sections.
Step two, based on an improved artificial fish swarm algorithm, generating an initial fish swarm by using the fault vector obtained in the step one, wherein each fish represents an expected state of each section of the power distribution network, and starting iterative optimization; the method specifically comprises the following steps:
21) initializing fish school, and representing the state of AF individual as vector Xi＝{a1,a2,……,anIn which a isiFor the optimization variables, the vector X represents the expected state of each feeder line interval in the power distribution networkiThe method is characterized by comprising 0 or 1, selecting proper fish school scale M according to the scale of the problem, and randomly generating M AF individuals.
22) And respectively carrying out net-collecting mechanism judgment, foraging behavior and rear-end collision behavior on the current AF, and judging the advancing direction of the AF by using the food concentration.
The specific content judged by the network receiving mechanism is as follows:
food concentration A if AF is presentαIs approximately zero or lower than the set value ALOWIf AF is towards the center position XcenterAdvancing;
the specific contents of the foraging behavior are as follows:
let a certain AF state be XaAt a current food concentration of AaSelecting a random state X within its field of viewbIf X isbConcentration of food at the location AbGreater than the current food concentration AaIt proceeds in this direction, otherwise another random state X is selected againb(ii) a If the advance condition which is met is not found after the trial for a certain number of times, executing a random behavior; regarding the default behavior of foraging behavior, AF is determined to be in a free swimming state, and when the foraging behavior does not meet the advancing condition after trying for a certain number of times, a state is randomly selected in the visual field range, and the AF is made to move towards the direction.
The specific contents of the rear-end collision behavior are as follows:
let a certain AF state be XaThe total number of partners in the visual field is m, and the partner state of the partner with the highest food concentration in the assumed position is XbIf A is satisfiedb/m＞δAaThen, it indicates the partner XbThe food concentration at the position is high, and the symbiosis is realized, and AF is towards XbIf not, the AF is foraging.
23) The AF clustering is determined and whether or not the progression is continued is determined. Specifically, the method comprises the following steps:
let a certain AF state be XaAt a current food concentration of AaThe total number of buddies in the visual field is m and the central position X of the buddiescenterIf the concentration at the center position satisfies Acenter/m＞δAaIf the concentration of food at the position of the partner center is higher and the partner center can live in symbiosis, the AF advances towards the direction; if not, the AF is subjected to foraging behavior.
24) And recording the current optimal AF state in real time by using a bulletin board method, comparing the AF state with the optimal state, updating the finally searched optimal state, and indicating that the optimal state is searched if the bulletin board is not updated for three times.
25) And introducing a life mechanism to enable the AF to increase the information of the packaging life cycle.
Survival index E of AF after introduction of life mechanismsaThe expression of (a) is:
Figure BDA0002223361430000031
in the formula, AaFood concentration for AF Current location, η is a consumption factor, init denotes that AF is performedInitializing, wherein T is a dynamic life cycle and has an expression as follows:
Figure BDA0002223361430000032
in the formula, Amaxλ is a proportionality coefficient, which is the maximum value of the currently known food concentration, i.e. the food concentration corresponding to the optimal AF state.
And step three, evaluating the advantages and disadvantages of the positions of the artificial fishes through the food concentration, updating iteration until the ending condition is met, and acquiring the positioned fault positions. Specifically, the method comprises the following steps:
obtaining a food evaluation value corresponding to each fish in the AF colony as a target function of a feasible solution, obtaining the food concentration of the AF position according to the food evaluation value, evaluating the quality of the artificial fish position based on the food concentration of the AF position, and indicating that the optimal state is found when the updating iteration is carried out until an end condition is met or the bulletin board is not updated for three times, wherein the AF recorded in the bulletin board is the optimal state and corresponds to a vector XiThe section number corresponding to the middle 1 is the positioned fault section.
Food evaluation value F corresponding to each fish in AF colonyiThe expression of (S) is:
Figure BDA0002223361430000041
wherein S is an AF individual; t issThe total number of switches of the power distribution network is obtained, L is the total number of feeders of the power distribution network, SjThe switch state values actually uploaded by the FTU,
Figure BDA0002223361430000042
alpha is a dimensional weight coefficient and takes a value of 0-1 for the state function of the constructed j-th switch]X oflIs a single feeder field status value.
Food concentration A at the location of AFfoodThe expression of (a) is:
Afood＝SNet-Fi(S)
in the formula, SNetIs a constantTerm, the value of which is determined by the scale of the solution to the distribution network, AfoodThe larger the value of (a), the greater the concentration of the food, the better the feasible solution.
Compared with the prior art, the invention has the following advantages:
(1) the method is used for positioning the fault position according to the direction of the fault current in the power distribution network without considering the uncertainty of DG output, the established integer programming fault model is suitable for the active power distribution network containing the distributed power supply, and the constructed fitness function is suitable for the fault positioning problem of the active power distribution network with different scales;
(2) the bionic algorithm has higher calculation efficiency, higher fault-tolerant capability on distortion information and more advantages in the application of active power distribution network fault location.
Drawings
FIG. 1 is a schematic flow chart of an active power distribution network fault location method based on an artificial fish swarm algorithm according to the invention;
FIG. 2 is a simplified diagram of an IEEE33 node active power distribution network;
fig. 3 is a comparison diagram of the AFSA, PSO, and SGA algorithm optimization curves under different fault conditions in the embodiment of the present invention, where fig. 3(a) is an AFSA, PSO, and SGA algorithm optimization curve under a single distortion condition, fig. 3(b) is an AFSA, PSO, and SGA algorithm optimization curve under a single distortion condition, fig. 3(c) is an AFSA, PSO, and SGA algorithm optimization curve under a multiple distortion condition, and fig. 3(d) is an AFSA, PSO, and SGA algorithm optimization curve under a multiple distortion condition.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of protection of the present invention.
Examples
The invention relates to an active power distribution network fault positioning method based on an artificial fish swarm algorithm, which is based on the following theoretical basis:
1. improved integer programming model
An integer planning model for power distribution network fault location is to equate a complex power distribution network architecture to a fault vector expressed by an integer according to current out-of-limit information uploaded by a Feeder Terminal Unit (FTU) after a fault occurs. The circuit breaker, the section switch and the interconnection switch are used as nodes, and the feeder line state information is used as an optimization variable.
The switching of DGs influences the power flow direction of the power distribution network after a fault occurs, and the upstream and downstream relation between the feeders can be changed in real time along with the switching of the DGs. Therefore, it is necessary to construct a model containing three parameters of '-1', '0', '1'. The invention assumes that the direction of the system power flow to the user and the DG is taken as the positive direction so as to avoid being influenced by the change of the DG access position. When the fault current direction uploaded by the FTU is the same as the positive direction, the switch state value Sj1. If the fault current direction is opposite to the positive direction, SjIs-1. If no fault current is detected, Sj＝0。
When a fault occurs, the power distribution network architecture can be equivalent to a fault vector expressed by an integer according to the current out-of-limit information uploaded by the FTU.
2. Construction of State function
When a fault occurs, the information uploaded by the FTU is the actual state value of each switch, and the fault location is realized by judging the actual state of each feeder line section, so that it is necessary to construct a state function to express the conversion from the state value of the feeder line section to the state value of the switch.
In an active power distribution network, switching of a DG needs to be considered, for a certain switch, not only a downstream feeder line but also an upstream feeder line exist, and a defined state function is as follows:
Figure BDA0002223361430000052
in the formula:
Figure BDA0002223361430000054
namely the state function of the j switch; sjdIndicating the state of the feeder downstream of the switch, SjuIs the state of the feeder line upstream of the switch No. j, namely the value of the optimizing variable;
Figure BDA0002223361430000053
is a logical or function of the state of all the feeders upstream of switch j,is a logical or function of the state of all the feeders downstream of switch j.
Construction of TDGiTo indicate the condition of the grid-connected operation of the DG, when the i-th DG is put into operation, the state is indicated as '1', otherwise, the state is '0'. It can be seen that when there is no DG grid connected, the state function is nowThe method is equivalent to the state function of a single-power radial distribution network.
3. Description of the Artificial Fish
Each Artificial Fish (AF) encapsulates its own data information (such as field of view, moving step length, etc.) and a series of daily interactions (foraging, herding, etc.), and receives environmental information to make corresponding activities. Namely, AFSA (Artificial Fish swarm Algorithm) simulates a series of real behaviors of foraging, herding, rear-end collision and the like of Fish in reality, and seeks an optimal solution in a given range.
The fault section positioning of the power distribution network is realized by judging the actual state of each feeder line section, and the process of artificial fish school iteration optimization is realized by solving the problem by using AFSA. The artificial fish population can be described as the solution space Z ═ X1,X2,……,Xi,……,XMWhere M is the fish size, i.e. total AFAnd (4) counting. The status of an AF individual can be represented as a vector Xi＝{a1,a2,……,anIn which a isiThe discrete optimization variables represent expected states of feeder intervals in the power distribution network, and n is the total number of sections of the power distribution network. The iterative optimization process is the combined optimization problem of discrete variables. In the optimizing process, the advantages and disadvantages of the positions of the artificial fishes are evaluated through the food concentration, iteration is updated until the ending condition is met, namely the maximum iteration times are reached or the bulletin board is not updated for 3 times, and the optimal position is found. At this time, AF recorded in the bulletin board is in an optimum state corresponding to the vector XiThe segment number corresponding to the '1' in the sequence is the fault segment.
Based on the theory, the active power distribution network fault positioning method based on the artificial fish swarm algorithm comprises the following steps:
step 1, according to an integer programming model of the power distribution network after the fault, the power distribution network is equivalent to a fault vector consisting of three integers.
And 2, generating an initial fish swarm, wherein each fish represents an expected state of each section of the power distribution network, and starting iterative optimization.
And 3, evaluating the advantages and disadvantages of the positions of the artificial fishes through the food concentration, and updating the iteration until the ending condition is met, namely the maximum iteration times are reached or the bulletin board is not updated for 3 times, which indicates that the optimum is found. At this time, AF recorded in the bulletin board is in an optimum state corresponding to the vector XiThe segment number corresponding to the '1' in the sequence is the fault segment. .
The specific implementation process comprises the following steps:
1) according to the proposed integer programming model of the active power distribution network, the complex active power distribution network after the fault is equivalent to a fault vector consisting of three integers of 0, 1 and-1, and the dimension of the vector is the total number of the sections of the power distribution network.
2) Initializing a fish school: the status of an AF individual can be represented as a vector Xi＝{a1,a2,……,anIs a decision vector, where aiFor the optimization variable, vector XiAnd the system consists of 0 or 1 and represents the expected state of each feeder line interval in the power distribution network. In the form of expression X1＝[0,0,1,0,1,…,0,0,0,0,1,0]. And selecting a proper fish school scale M according to the scale of the problem, and randomly generating M AF individuals.
3) A net collecting mechanism: when the AFSA is applied to a power distribution network with more nodes, the fish school scale to be set is large, and AF with an undesirable initial state may appear at the initial stage of the iterative process. Aiming at the problem, the invention provides a net collecting mechanism, if the food concentration A of the AF position isαIs approximately zero or lower than the set value ALOWThen to the central position XcenterAdvancing, represented by the formula:
Figure BDA0002223361430000071
wherein, Xa|nextThe next state of AF is represented, and the Most function represents taking a value that is Most common or closest to the corresponding position.
4) Foraging behavior: some AF state is XaThe current food concentration is AaSelecting a random state X within its field of viewbIf X isbConcentration of food at the location AbGreater than the current food concentration AaThen proceed in that direction, otherwise another random state X is selected againb(ii) a If no satisfied advance condition is found after Try Try times, then a random action is executed, which is expressed as follows:
Figure BDA0002223361430000072
5) and (3) rear-end collision behavior: some AF state is XaThe total number of partners in its field of view is m. The partner state of the position with the highest food concentration in the search partner is XbIf A isb/m＞δAaThen, it indicates the partner XbThe food concentration at the position is high, the symbiosis can be realized, and the direction of X is changed when the conditions are metbIs advanced in the direction of (1); if not, the AF performs foraging behavior, which is expressed by the following formula:
Figure BDA0002223361430000073
6) clustering behavior: some AF state is XaThe total number of buddies in the visual field is m and the central position X of the buddiescenterIf the concentration at the center position satisfies Acenter/m＞δAaIf the food concentration is higher and the symbiosis is possible, the food concentration is higher at the center of the partner, and if the conditions are met, the food concentration goes forward; if not, the AF performs foraging behavior, which is expressed by the following formula:
Figure BDA0002223361430000074
7) random behavior: i.e., the default behavior of foraging behavior, can be considered free-swimming of AF. Some AF state is XaIf the foraging behavior does not satisfy the advance condition after Try attempts, a state is randomly selected within the visual field range and moved in the direction, which is expressed by the following formula:
Xa|next＝Xa·rand(Step)
8) bulletin board: the effect is to record the current optimum AF state in real time. In the optimizing process, after the action of each AF is finished, the state of each AF is compared with the record in the bulletin board, and if the updated state of the AF is superior to the bulletin board, the bulletin board is updated to be in the new state; otherwise, the bulletin board record is unchanged. The arrangement of the bulletin board enables the optimal state to be recorded in real time, and the finally found optimal solution can be conveniently extracted.
9) The life mechanism is as follows: in order to reduce the influence of local extreme values on global optimization and improve efficiency, a life mechanism is introduced, namely, AF is enabled to increase the information of the packaging life cycle. The AF has the strongest vitality near the global extreme value, and the AF trapped in the local extreme value disappears along with the increase of the iteration number, so that the storage space is saved while the optimizing capability is improved, and the AF is initialized if food at the position of the AF is not enough to maintain the life of the AF, as shown in the following formula:
Figure BDA0002223361430000081
wherein E is a survival index; a represents the food concentration at the AF current position; t is the dynamic life cycle, and eta is the consumption factor. init indicates that the AF is initialized.
T is a dynamic life cycle, so that the life cycle of AF is improved along with the progress of the optimization process, so as to achieve the purpose that AF with poor position and local extremum does not generate excessive redundancy, and the expression is:
Figure BDA0002223361430000082
in the formula, AmaxIs the maximum value of the currently known food concentration, i.e. the optimal record in the bulletin board. λ is a proportionality coefficient. The specific implementation flow is shown in fig. 1.
10) Evaluation of food concentration: the key to the evaluation of the food concentration at the location of the artificial fish is the construction of an objective function. The principle of the structure is as follows: and the fault vector determined by the AF individual vector after the operation of the state function is the smallest difference with the equivalent fault vector of the actual power distribution network. The invention adopts the target function containing the dimensional weight coefficient, accords with the concept of the minimum set in the fault diagnosis theory, and reduces the possibility of avoiding the missed judgment and the erroneous judgment. As shown in the following formula:
Figure BDA0002223361430000083
in the formula: fi(S) is a food evaluation value corresponding to each fish in the AF group, namely an objective function value which can be solved. S represents an AF individual; t issThe total number of the switches of the power distribution network is obtained; l is the total number of the feeder lines of the power distribution network; sjThe switch state value is actually uploaded by the FTU;
Figure BDA0002223361430000092
a state function for the constructed switch # j; alpha is a weight-preserving coefficient and takes the value of 0-1]X oflFor a single feeder section state value, the visible dimension term is determined by the number of faulty feeder sections in the optimization process.
The food concentration at the location of AF, i.e. the fitness function of the optimization process, can be constructed as follows:
Afood＝SNet-Fi(S)
wherein: fi(S) is the above defined objective function; sNetThe constant term is determined by the numerical value of the constant term according to the scale of solving the power distribution network; a. thefoodIndicating the concentration of the food and judging the quality of the feasible solution. Can see AfoodThe larger the value of (a), the greater the concentration of the food, the better the feasible solution.
The present embodiment takes an active distribution network of IEEE33 node as an example for simulation, as shown in fig. 2. In the example distribution network, S represents a system power supply, DG1, DG2 and DG3 represent distributed power supplies, and K1, K2 and K3 are switching switches of the distributed power supplies respectively. The active power distribution network fault positioning method based on the artificial fish swarm algorithm does not need to consider the fault characteristics of different DGs and the uncertainty of output, so that the types of the DGs do not influence the simulation result.
The difference of the fault type and the number respectively; and fault positioning is carried out under the conditions of missing report or distortion of FTU uploading information, so that the effectiveness, the accuracy and the high fault tolerance of the proposed algorithm are verified. And the algorithm is compared and analyzed with other intelligent algorithms, so that the superiority of the algorithm in fault positioning application is demonstrated.
(1) Analysis of positioning results
The results are shown in Table 1. It can be seen that under the condition of a plurality of DG accesses, the algorithm can accurately output the faulty feeder line. The algorithm can correctly position the single fault and the multiple faults. Meanwhile, under the condition that a plurality of pieces of information of fault current information uploaded by the FTU are missed or distorted, the FTU can also be accurately positioned. The simulation result can prove that the algorithm can dynamically adapt to DG switching, has good effectiveness and high fault tolerance on information of missing report or distortion.
TABLE 1 positioning results of different fault situations during DG access
Figure BDA0002223361430000091
(2) Fault tolerance, rapidity analysis
In order to verify the superiority of the method in the active power distribution network fault location application, an artificial fish swarm algorithm AFSA is compared with a genetic algorithm SGA and a basic particle swarm PSO algorithm which are widely applied.
Firstly, aiming at different fault conditions, namely single/multiple faults without information distortion and single/multiple faults with information missing report and distortion, the method carries out comparative analysis on algorithm accuracy, rapidity and fault tolerance. The optimization curve pairs of the AFSA, SGA and PSO algorithms are shown in FIG. 3.
a. The single fault of the preset fault feeder line (27) has no information distortion. As can be seen from comparison of the optimization curves in FIG. 3(a), the three algorithms of AFSA, PSO and SGA can reach ideal adaptive values in the optimization process, and the position of a fault point can be accurately found. In comparison, the AFSA algorithm can complete the iteration the fastest, PSO times, while the average number of iterations of the SGA exceeds 20.
b. When the single fault of the fault feeder (27) is preset, distortion information is generated at switches S4 and S14. As can be seen from the analysis of the optimization curve in fig. 3(b), in the presence of distortion information, the SGA may generate a locally optimal solution too early, and even if a fault location is finally found, the number of iterations is significantly greater than that of the AFSA and PSO algorithms. Compared with other two algorithms, a small amount of distortion information hardly influences the AFSA algorithm, and the algorithm can still quickly and accurately locate the fault position.
c. Presetting multiple faults of fault feeder lines (9) (30) without information distortion. As can be seen from the comparison of the optimization curves in fig. 3(c), the AFSA algorithm has a small average number of iterations, and the situation of falling into local optimum is not substantially generated. The PSO algorithm has more average iteration times, the premature problem appears in the early stage, but the optimal solution can be found after 50 iterations. Even if the iteration times exceed 100 times, the SGA still does not tend to the highest adaptive value, does not find the optimal solution, and falls into local optimization for many times in the iteration process, and can hardly complete fault positioning.
d. Presetting multiple faults of the fault feeder (9) (30), and switching information distortion or failure report at 5 points of S13, S14, S18, S25 and S27. As can be seen from the optimization curve in fig. 3(d), after the SGA algorithm iterates 300 times, an optimal solution is not found yet and the premature problem is obvious. The PSO algorithm is also difficult to find the optimum and is prone to premature when dealing with a large amount of information distortion. The AFSA algorithm can be used for accurately positioning the fault, and although the iteration times are increased along with the increase of the distortion information, the optimization process can be completed quickly.
In conclusion, the SGA algorithm has serious premature problems, is easy to generate local optimal solutions too early, and has too many iteration times; the PSO algorithm has poor fault tolerance and more iteration times, and the two algorithms are not suitable for fault location of the active power distribution network with information distortion. The AFSA method provided by the invention has the advantages of good stability, high convergence rate and high fault tolerance, and is suitable for a multi-branch large-scale active power distribution network.
(3) Analysis of accuracy
The following different fault types and positions were selected and the three algorithms were subjected to 20 fault simulations for further comparison, respectively, with the results shown in table 2.
TABLE 2 comparison of fault location accuracy for three algorithms
Figure BDA0002223361430000111
The comparison of table 2 shows that the fault location method of the present invention has high fault tolerance to information distortion, and also highlights the strong accuracy in locating multiple faults. Compared with the method disclosed by the invention, the PSO algorithm has the advantages of general accuracy in experiments and poor fault tolerance on distortion information. The SGA algorithm can hardly realize multiple fault location with more distortion information, and is not suitable for a large-scale active power distribution network because the fault location accuracy of a simple power distribution network is not high.
While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and those skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (10)

1. An active power distribution network fault positioning method based on an improved artificial fish swarm algorithm is characterized by comprising the following steps:
1) according to the improved integer programming model of the power distribution network after the fault, the power distribution network is equivalent to a fault vector consisting of three integers;
2) based on an improved artificial fish swarm algorithm, generating an initial fish swarm by using the fault vector obtained in the step 1), wherein each fish represents an expected state of each section of the power distribution network, and starting iterative optimization;
3) and evaluating the quality of the position of the artificial fish through the food concentration, updating iteration until a finishing condition is met, and acquiring the positioned fault position.
2. The active power distribution network fault location method based on the improved artificial fish swarm algorithm according to claim 1, wherein in the step 1), the main content of the power distribution network equivalent fault vector formed by three integers is as follows:
an integer programming model based on power distribution network fault location takes a circuit breaker, a section switch and a tie switch as nodes, takes feeder line state information as an optimization variable, takes the direction of a system power supply flowing to a user and a DG as a positive direction, and when the fault current direction uploaded by an FTU is the same as the positive direction, a switch state value S is obtainedjIf the fault current direction is opposite to the positive direction, S is 1j-1, if no fault current is detected, Sj0; when a fault occurs, according to the current out-of-limit information uploaded by the FTU, namely the actual state value of each switch, the power distribution network architecture can be equivalent to an integer expressed fault vector, and the dimension of the vector is the total number of power distribution network sections.
3. The active power distribution network fault location method based on the improved artificial fish swarm algorithm according to claim 1, wherein the step 2) specifically comprises the following steps:
21) initializing fish school, and representing the state of AF individual as vector Xi＝{a1,a2,……,anIn which a isiFor the optimization variables, the vector X represents the expected state of each feeder line interval in the power distribution networkiThe method comprises the following steps of (1) selecting a proper fish school scale M according to the scale of problems, and randomly generating M AF individuals;
22) respectively carrying out net-collecting mechanism judgment, foraging behavior and rear-end collision behavior on the current AF, and judging the advancing direction of the AF by using the food concentration;
23) judging the AF cluster, and judging whether the AF cluster continues to advance;
24) recording the current optimal AF state in real time by using a bulletin board method, comparing the AF state with the optimal state, updating the optimal state searched finally, and if the bulletin board is not updated for three times, indicating that the optimal state is searched;
25) and introducing a life mechanism to enable the AF to increase the information of the packaging life cycle.
4. The active power distribution network fault location method based on the improved artificial fish swarm algorithm according to claim 3, wherein the specific content of the step 3) is as follows:
obtaining a food evaluation value corresponding to each fish in the AF colony as a target function of a feasible solution, obtaining the food concentration of the AF position according to the food evaluation value, evaluating the quality of the artificial fish position based on the food concentration of the AF position, and indicating that the optimal state is found when the updating iteration is carried out until an end condition is met or the bulletin board is not updated for three times, wherein the AF recorded in the bulletin board is the optimal state and corresponds to a vector XiThe section number corresponding to the middle 1 is the positioned fault section.
5. The active power distribution network fault location method based on the improved artificial fish swarm algorithm as claimed in claim 4, wherein the food evaluation value F corresponding to each fish in the AF swarmiThe expression of (S) is:
Figure FDA0002223361420000021
wherein S is an AF individual; t issThe total number of switches of the power distribution network is obtained, L is the total number of feeders of the power distribution network, SjThe switch state values actually uploaded by the FTU,
Figure FDA0002223361420000022
alpha is a dimensional weight coefficient and takes a value of 0-1 for the state function of the constructed j-th switch]X oflIs a single feeder field status value.
6. The active distribution network fault location method based on the improved artificial fish swarm algorithm as claimed in claim 5, wherein the food concentration A of AF isfoodThe expression of (a) is:
Afood＝SNet-Fi(S)
in the formula, SNetIs a constant term, the value of which is determined according to the scale of solving the distribution network, AfoodThe larger the value of (a), the greater the concentration of the food, the better the feasible solution.
7. The active power distribution network fault location method based on the improved artificial fish swarm algorithm according to claim 3, wherein in step 22), the specific content judged by the network receiving mechanism is as follows:
food concentration A if AF is presentαIs approximately zero or lower than the set value ALOWIf AF is towards the center position XcenterAdvancing;
the specific contents of the foraging behavior are as follows:
let a certain AF state be XaAt a current food concentration of AaSelecting a random state X within its field of viewbIf X isbConcentration of food at the location AbGreater than the current food concentration AaIt proceeds in this direction, otherwise another random state X is selected againb(ii) a If the advance condition which is met is not found after the trial for a certain number of times, executing a random behavior;
the specific contents of the rear-end collision behavior are as follows:
let a certain AF state be XaThe total number of partners in the visual field is m, and the partner state of the partner with the highest food concentration in the assumed position is XbIf A is satisfiedb/m＞δAaThen, it indicates the partner XbThe food concentration at the position is high, and the symbiosis is realized, and AF is towards XbIf not, the AF is foraging.
8. The active distribution network fault location method based on the improved artificial fish swarm algorithm according to claim 7, in step 22), regarding the default behavior of foraging behavior, the AF is determined as a free-swimming state, and when the foraging behavior does not satisfy the forward condition after trying for a certain number of times, a state is randomly selected within the visual field range, so that the AF moves towards the direction.
9. The active power distribution network fault location method based on the improved artificial fish swarm algorithm according to claim 3, wherein the specific content of the step 23) is as follows:
let a certain AF state be XaAt a current food concentration of AaThe total number of buddies in the visual field is m and the central position X of the buddiescenterIf the concentration at the center position satisfies Acenter/m＞δAaIf the concentration of food at the position of the partner center is higher and the partner center can live in symbiosis, the AF advances towards the direction; if not, the AF is subjected to foraging behavior.
10. The active power distribution network fault location method based on the improved artificial fish swarm algorithm as claimed in claim 3, wherein in step 25), after the life mechanism is introduced, the survival index E of AF is obtainedaThe expression of (a) is:
in the formula, AaFood concentration at AF current position, eta isAnd an init represents that the AF is initialized, T is a dynamic life cycle, and the expression is as follows:
Figure FDA0002223361420000032
in the formula, Amaxλ is a proportionality coefficient, which is the maximum value of the currently known food concentration, i.e. the food concentration corresponding to the optimal AF state.
CN201910942736.3A 2019-09-30 2019-09-30 Active power distribution network fault positioning method based on improved artificial fish swarm algorithm Active CN110687397B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910942736.3A CN110687397B (en) 2019-09-30 2019-09-30 Active power distribution network fault positioning method based on improved artificial fish swarm algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910942736.3A CN110687397B (en) 2019-09-30 2019-09-30 Active power distribution network fault positioning method based on improved artificial fish swarm algorithm

Publications (2)

Publication Number Publication Date
CN110687397A true CN110687397A (en) 2020-01-14
CN110687397B CN110687397B (en) 2022-08-23

Family

ID=69111098

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910942736.3A Active CN110687397B (en) 2019-09-30 2019-09-30 Active power distribution network fault positioning method based on improved artificial fish swarm algorithm

Country Status (1)

Country Link
CN (1) CN110687397B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111638421A (en) * 2020-05-20 2020-09-08 国网河北省电力有限公司电力科学研究院 Method for positioning fault section of power distribution network containing distributed power generation grid connection
CN113239623A (en) * 2021-05-17 2021-08-10 上海交通大学 Fault positioning method suitable for electric power equipment
CN113420912A (en) * 2021-06-04 2021-09-21 国网江西省电力有限公司电力科学研究院 Method for identifying users with low-voltage abnormality of power distribution network

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102184338A (en) * 2011-06-08 2011-09-14 南京航空航天大学 Artificial fish swarm algorithm based direction optimal configuration method of piezoelectric actuator
US20130057296A1 (en) * 2011-09-06 2013-03-07 General Electric Company Systems, Methods, and Apparatus for Locating Possible Fault Locations in an Electrical Power Network
CN103927459A (en) * 2014-05-04 2014-07-16 华北电力大学（保定） Method for locating faults of power distribution network with distributed power supplies
CN106249109A (en) * 2016-09-20 2016-12-21 广西电网有限责任公司钦州供电局 Electrical power distribution network fault location method and system
CN108693771A (en) * 2017-04-10 2018-10-23 南京理工大学 A kind of distribution network failure section location algorithm based on Multiple-population Genetic Algorithm

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102184338A (en) * 2011-06-08 2011-09-14 南京航空航天大学 Artificial fish swarm algorithm based direction optimal configuration method of piezoelectric actuator
US20130057296A1 (en) * 2011-09-06 2013-03-07 General Electric Company Systems, Methods, and Apparatus for Locating Possible Fault Locations in an Electrical Power Network
CN103927459A (en) * 2014-05-04 2014-07-16 华北电力大学（保定） Method for locating faults of power distribution network with distributed power supplies
CN106249109A (en) * 2016-09-20 2016-12-21 广西电网有限责任公司钦州供电局 Electrical power distribution network fault location method and system
CN108693771A (en) * 2017-04-10 2018-10-23 南京理工大学 A kind of distribution network failure section location algorithm based on Multiple-population Genetic Algorithm

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
李晓磊: "一种新型的智能优化方法-人工鱼群算法", 《中国优秀博硕士学位论文全文数据库（博士）信息科技辑》 *
段大伟 等: "基于混合算法的配电网故障定位", 《黑龙江科学》 *
陈磊 等: "基于人工鱼群算法的配电网故障定位", 《电子测量技术》 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111638421A (en) * 2020-05-20 2020-09-08 国网河北省电力有限公司电力科学研究院 Method for positioning fault section of power distribution network containing distributed power generation grid connection
CN111638421B (en) * 2020-05-20 2022-05-31 国网河北省电力有限公司电力科学研究院 Method for positioning fault section of power distribution network containing distributed power generation grid connection
CN113239623A (en) * 2021-05-17 2021-08-10 上海交通大学 Fault positioning method suitable for electric power equipment
CN113239623B (en) * 2021-05-17 2022-11-18 上海交通大学 Fault positioning method suitable for electric power equipment
CN113420912A (en) * 2021-06-04 2021-09-21 国网江西省电力有限公司电力科学研究院 Method for identifying users with low-voltage abnormality of power distribution network
CN113420912B (en) * 2021-06-04 2022-04-26 国网江西省电力有限公司电力科学研究院 Method for identifying users with low-voltage abnormality of power distribution network

Also Published As

Publication number Publication date
CN110687397B (en) 2022-08-23

Similar Documents

Publication Publication Date Title
CN110687397B (en) Active power distribution network fault positioning method based on improved artificial fish swarm algorithm
Truong et al. A quasi-oppositional-chaotic symbiotic organisms search algorithm for optimal allocation of DG in radial distribution networks
CN106532778B (en) Method for calculating maximum access capacity of distributed photovoltaic grid connection
CN107994582B (en) Method and system for reconstructing power distribution network containing distributed power supply
Yong An improved artificial fish swarm algorithm for optimal operation of cascade reservoirs
CN113178864A (en) Power distribution network power supply fault recovery method and device and terminal equipment
CN106777449A (en) Distribution Network Reconfiguration based on binary particle swarm algorithm
CN104867062A (en) Low-loss power distribution network optimization and reconfiguration method based on genetic algorithm
CN112149264A (en) Active power distribution network planning method based on improved Harris eagle optimization algorithm
Azad-Farsani et al. A new hybrid CPSO-TLBO optimization algorithm for distribution network reconfiguration
Zhang et al. Evolutionary based Pareto optimization algorithms for bi-objective PV array reconfiguration under partial shading conditions
CN113673065B (en) Loss reduction method for automatic reconstruction of power distribution network
Chicco et al. An overview of the probability-based methods for optimal electrical distribution system reconfiguration
CN111881626B (en) Distribution network planning method for promoting DG (distributed generation) digestion
CN107276093B (en) The Probabilistic Load calculation method cut down based on scene
CN108710742A (en) A kind of fault section location method based on PGSA-GA hybrid algorithms
CN114167216A (en) Complex power distribution network fault positioning method based on improved mantray foraging optimization algorithm
Kanoongo et al. Blended biogeography based optimization for different economic load dispatch problem
Qi Operation ControlMethod of Relay Protection in Flexible DC Distribution Network Compatible with Distributed Power Supply.
Sulaima et al. A 16kV distribution network reconfiguration by using evolutionaring programming for loss minimizing
CN115986744B (en) Power distribution network power flow optimization method containing shared energy storage
Qi Research on prediction model of improved BP neural network optimized by genetic algorithm
Li et al. PSO-AFSA global maximum power point tracking algorithm with adaptive evolutionary strategy for PV system
Zhou et al. A power supply restoration method of distribution network based on NSGA-II algorithm
Wang et al. Distribution network loss calculation and reactive power optimization loss reduction analysis

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant