WO2021153901A1 - Loop-based distribution network reconfiguration method - Google Patents

Abstract

The present invention relates to a distribution network reconfiguration method, and more specifically, to a method for reconfiguring a distribution network on the basis of a loop. The purpose of the present invention is to reconfigure a distribution network while minimizing losses in lines in consideration of the operation efficiency of the network. The present invention is characterized by comprising: a network simplification step for omitting a portion of a distribution network and contracting the network; a loop searching step for defining a loop by using the network simplified in the network simplification step, and searching for a switch included in the defined loop; a distribution network analysis step for defining a switch combination that can be isolated in the network by using the switch found in the loop searching step, and analyzing detailed switch combinations that satisfy the radial structure of the network; and a network reconfiguration step for using the detailed switch combinations to calculate the line losses, and deriving a global optimal solution on the basis of the calculated line losses.

Description

Loop-based distribution system reconfiguration method

The present invention relates to a method for reconfiguring a distribution system, and more particularly, to a method for reconfiguring a distribution system based on a loop.

Distribution network reconfiguration (DNR) is an operation technique of a distribution system that relocates the load borne by each feeder by appropriately controlling the closing/opening state of switchgear existing on the line. In general, a distribution system may include numerous power facilities and lines in order to stably supply high-quality power and to minimize breakdowns and construction sections. Therefore, the structure of the distribution system may have a complex radial, loop, and mesh structure, and the operator of the system operates some switchgear existing in the loop and mesh structure for the convenience of protecting the protection devices and controlling the voltage at the end of the line. It is opened and operated after converting the system to a radial structure.

However, the operating efficiency of the distribution system may deteriorate or an isolated section may occur depending on the positions of the switchgear in an open state. Therefore, it is necessary to appropriately control the switchgear in the open state and to minimize the power loss occurring in the line by changing the system configuration, to alleviate the voltage reinforcement by the load, or to improve the uneven distribution of the load between the feeders. To this end, the system analysis for the switchgear combination should be preceded, and a system configuration that satisfies the radial structure but has good operating efficiency should be searched for.

The power distribution system can be connected with various types of consumers such as residential, general, industrial, educational, agricultural, etc., and these various consumers have different load patterns and different fluctuation characteristics by time, day of week, month, and season. In addition, distributed generation (DG) based on renewable energy, which is connected to the grid in large quantities in modern times, is affected by natural phenomena such as solar radiation, cloudiness, wind volume, and temperature, and the output changes rapidly. Therefore, in addition to these load fluctuation characteristics, there is a need for a method for optimizing and analyzing the system in response to the topology change due to construction and failure occurring in the system and the system connection and separation due to the microgrid capable of independent operation.

The best way to find the global optimal solution in optimization analysis is to simulate all possible cases. However, the number of cases to be considered in the systemic reconstruction process is

Therefore, a lot of time is required to solve the systematic reconstruction. Therefore, there is a need for a method for deriving the global optimal solution of the system and reducing the time required for solving the system reconstruction.

In addition, since the optimal system configuration changes depending on the uncontrollable switchgear at the time when the system reconfiguration is performed, the output amount of distributed power and independent operation, a method for searching for the optimal system configuration according to the current system state is required.

An object of the present invention is to perform reconfiguration of the distribution system, but to minimize the loss of the line in consideration of the operating efficiency of the system.

In addition, an object of the present invention is to derive a global optimal solution in consideration of the current state of the system, but to reduce the time required for solving the system reconstruction.

A loop-based distribution system reconfiguration method according to a preferred embodiment of the present invention defines a loop using a simplified system by omitting a part of the distribution system, abbreviated system simplification step, and the system simplification step, and defining The loop search step of searching the switchgear included in the loop, and the switchgear combination in which isolation occurs in the system is defined using the switchgear found by the loop search step, and a detailed switchgear combination that satisfies the radial structure of the distribution system It characterized in that it comprises a system reconfiguration step of calculating a line loss using the distribution system analysis step to analyze and the detailed switchgear combination, and deriving a global optimal solution based on the calculated line loss.

In addition, the system simplification step is characterized by determining a situation in which an isolated section can be created, and simplifying the system in response to the determined situation, but abbreviated lines with the SBN as the center.

In addition, the distribution system analysis step is characterized in that when a plurality of power sources exist in the distribution system, a virtual branch (VB) that cannot be opened between each power source is formed.

In addition, the system reconfiguration step is characterized in that the center of the BSS is identified and the number of detailed switchgear that can be selected in one BSS is limited.

In addition, the system reconfiguration step is characterized in that the search for a BSS that maintains the always-in state, and limiting the searched BSS.

In addition, the present invention may further provide a loop-based distribution system reconfiguration program stored in a medium for performing the loop-based distribution system reconfiguration method in combination with a computer.

In addition, the present invention may further provide a server system in which the loop-based distribution system reconfiguration program is stored and capable of transmitting the loop-based distribution system reconfiguration program through a communication network.

In addition, the present invention stores the loop-based distribution system reconfiguration program and reconfigures the distribution system by the loop-based distribution system reconfiguration program (SDMS: Smart Distribution Management System), a distribution operating system (DMS) : Distribution Management System) and distribution automation system (DAS: Distribution Automation System) can be provided.

The present invention defines a switchgear combination in which isolation occurs in the system and analyzes the detailed switchgear combination that satisfies the radial structure of the distribution system, thereby minimizing line loss.

In addition, the present invention determines a situation in which an isolated section can be generated, and simplifies the system in response to the determined situation, but reduces the time consumed in solving the system reconfiguration by reducing the lines around the SBN.

1 is a view for explaining a loop structure of a power distribution system.

2 is a view for explaining the radial structure of the distribution system.

3 is a diagram for explaining an example of an incorrect system configuration.

4 is a view for explaining the relationship between the radial structure and the loop structure.

5 is a view for explaining the conversion of the loop structure to the radial structure.

6 is a view for explaining before system simplification is performed.

7 is a view for explaining a condition for generating an isolated section.

8 is a diagram for explaining a condition for generating an isolated section.

9 is a diagram for explaining system simplification.

10 is a flowchart of a systematic simplification algorithm.

11 is a view for explaining the intersecting line S14 in the system.

12 is a diagram for explaining a limit of a loop search using a line connection relationship.

13 is a diagram for explaining system simplification.

14 is a diagram for explaining a loop search process.

15 is a diagram for explaining an initial radial system.

16 is a diagram for explaining an affiliated BSS search for each loop.

17 is a diagram for explaining a system drawn with data.

FIG. 18 is a diagram for explaining the loop present in FIG. 16 .

19 is a diagram for explaining optimization of loop 2;

20 is a flowchart of a loop search algorithm.

21 is a view for explaining the opening of a line included in one loop.

22 is a view for explaining the opening of a line included in two loops.

23 is a view for explaining a process of reflecting the switch in the open state.

24 is a diagram for explaining a system in which two or more power sources exist.

25 is a flowchart of a current state reflection algorithm.

26 is a diagram for explaining an isolation condition of a node.

27 is a diagram for explaining the IEEE 33 BUS DISTRIBUTION NETWORK.

28 is a diagram for explaining a BSS combination in which isolation occurs.

29 is a view for explaining selections 1 and 2 of Table 13;

30 is a view for explaining selection 3 of Table 13 and a detailed switchgear.

31 is a diagram for explaining a single line diagram of a distribution system.

32 is a flow chart for simulating all possible combinations.

33 is a diagram for explaining a BSS for calculating a loss.

34 is a diagram for explaining various convergence characteristic curves of a genetic algorithm.

35 is a flow chart of a genetic algorithm.

36 is a diagram for explaining IEEE 118 BUS DISTRIBUTION NETWORK zone division.

37 is a diagram for explaining a BSS participating in system reconfiguration.

38 is a diagram for explaining IEEE 118 BUS DISTRIBUTION NETWORK load classification.

39 is a diagram for explaining a Monday load pattern in summer.

In the event of a breakdown in the distribution system or a construction plan is established, rapid section isolation and power recovery can be performed so that only a minimal number of areas experience a power outage. This is possible because there are numerous interconnected lines and backup lines in the system, and due to these lines, the loop structure of the distribution system can be formed as shown in FIG. 1 . Referring to FIG. 1 , the loop structure of the distribution system may be defined as shown in Table 1.

Loop in feeder	Loop formed inside one feeder
Loop between substations	The transformer/substation is the same but Loop formed between different feeders
Loop between feeders	Loops formed in different transformers/substations

When the distribution system is operated in a loop or mesh structure, the protection coordination between the protection devices is complicated because the flow of the current is bidirectional. In addition, it becomes difficult to control the voltage of the power facilities to maintain an appropriate value of the voltage compared to the radial structure. Therefore, the operator of the system changes some of the switchgear existing in the distribution system to an open state and operates the system by converting the system structure into a radial structure as shown in FIG. 2 . Depending on the position of the open switch, unequal distribution of the load may occur as shown in (a) of FIG. 3, thereby deteriorating operating efficiency, and an isolated section in which power is not supplied may appear as shown in (b) of FIG. 3 . Therefore, while converting the loop structure to the radial structure, a combination of switchgear that sufficiently satisfies the operating efficiency should be searched. This will be described in detail in the loop-based distribution system analysis method to be described later.

Referring to FIG. 4 , the radial structure of the power distribution system may have the following characteristics.

- Every node in a system with a radial structure can have a unique line to the power source.

- If one line is added in a system with a radial structure, one loop can be formed.

All load nodes in the distribution system can have a unique line to the power node. In addition, each time a line is added between existing nodes without adding a node, one loop may be added. Through Equation 1, it is possible to determine the number of loops existing in the distribution system of the loop structure.

Referring to FIG. 5 , the loop structure can be converted into a radial structure by reversely utilizing a condition for forming a loop in the radial structure. If one line is opened in each loop, a radial structure as shown in Fig. 5 (a) can be obtained, and an isolated section as shown in Fig. 5 (b) may occur depending on the opened line. As a result, if the switchgear included in the loop existing in the system can be identified and the conditions under which the isolation occurs can be specified, it is easy to find a switchgear combination that is converted from the loop structure to the radial structure. Based on this relationship, the following facts can be specified.

- If one line is opened in each loop, the loop structure can be converted to a radial structure. However, a switch combination in which an isolated section occurs in the system cannot be selected.

3. Data processing plan for reconfiguration of distribution system

In the present invention, a part of the system may be omitted and abbreviated prior to analyzing the structure of the distribution system. Through this, the number of switch combinations to be considered can be minimized by removing unnecessary components for structural analysis of the distribution system. In addition, by simplifying the system, it is possible to simply search for the rope existing in the system and the switches belonging to it. Accordingly, it is possible to improve the performance of the future distribution system reconfiguration.

go. Loop-based system simplification

1) Super Branch Node (SBN)

Nodes existing in the distribution system can be classified as shown in Table 2 according to the number of connected lines. By utilizing the structural relationship between the node and the line, it is possible to grasp the condition in which isolation occurs inevitably, and to remove components unnecessary for analyzing the system structure. A line having such a node may be represented as shown in FIG. 6 .

Kinds	connected to each other	note
End Node	One	located at the beginning or end of a lineage
Connected Nodes	2	connected with two lines
Super Branch Nodes (SBNs)	3 or more	connection with three or more lines

Referring to FIG. 7 (a), when an end node is present and a switchgear is opened in a line set whose end is an SBN, an isolated section may inevitably be generated. That is, the section composed of the SBN and the end node becomes an unnecessary component for solving the system reconfiguration in which the switch is opened, so it can be omitted as shown in FIG. 7(b).

Referring to FIG. 8 ( a ) , in a section in which both ends are composed of SBNs, when two or more switchgear are opened, an isolated section may inevitably be created. That is, since only one switch should be opened in the corresponding set, it can be abbreviated to one line as shown in FIG. 8(b).

Based on this fact, if the lines are arranged around the SBN, it is possible to prevent the selection of a switchgear that inevitably causes isolation.

2) Branch Switch Set (BSS)

In order to minimize the situation where isolation occurs, the system can be simplified by omitting all connected nodes in the system. Thereafter, a set of lines may be defined as shown in Table 3 according to the type of node existing at the end of the line centered on the SBN, and the defined set of lines may be expressed as shown in FIG. 9 .

Kinds	Linked node type	Conditions for creating an isolated section
BSS #1	SBN - SBN	Two or more switchgear open
BSS #2	SBN - end node	One or more switchgear open

As described above, in the case of BSS #1, an isolation section does not occur even when one switch is opened, but in the case of BSS #2, an isolation section occurs even when only one switch is opened. Therefore, by omitting BSS #2 from the said system|strain, the system|strain of FIG.8(b) can be acquired.

By performing the system simplification work through such abbreviations and omissions, the system can be expressed with a minimum of SBN and BSS, and unnecessary isolation can be minimized in the structural analysis of the system.

10 is a flowchart of a systematic simplification algorithm. Referring to FIG. 10 , the systematic simplification flow described above may become more clear.

me. loop search method

It is a simple task for a person to intuitively look at the schematic or floor plan and grasp each loop and the switches included in the loop. However, systematic logic is required to define the loop existing in the system using only line data and node data and to accurately identify the switchgear included in the loop.

1) Necessity of loop navigation

To convert a system with a loop structure to a radial structure, one switchgear must be opened in each loop. However, depending on the selected combination of switchgear, an isolated section may occur in the corresponding system. Therefore, a systematic and logical procedure is required to select a switchgear while avoiding the condition in which isolation occurs. For this, first, accurate information about the switch included in each loop should be acquired.

2) Loop search process

The simplest way to explore a loop using data is to track the connection relationship of the line. You just need to select one node and search the path back to the same node, but in a complex system, you can only get results by going through a lot of trial and error. Also, as shown in FIG. 11 , when a line S14 intersecting another line exists in the middle of the lines included in the two loops, accurate loop information cannot be obtained.

Due to the existence of the crossing line S14 of FIG. 11 , as shown in FIG. 12 , a maximum of six loops may be searched in an area where only three loops should be searched. Therefore, a method other than a method that utilizes the connection relationship of the lines is required.

The present invention utilizes the property that 'one loop can be added every time a line is added' and 'a loop can disappear one by one whenever a switch constituting the loop is opened' described above. You can browse the included switchgear.

In order to define the number of loops that exist in the distribution system and the BSS included in each loop, the feature that 'a loop can disappear one by one whenever a switch composing a loop is opened' can be utilized. 13 (a) is a diagram illustrating a system before simplification, and FIG. 13 (b) is a diagram illustrating a system after simplification. In the simplified system as shown in FIG. 13(b), if one arbitrary line is deleted as shown in FIG. 14, one loop is removed and a line included in the corresponding loop can be defined. That is, loop 1 can be defined while S1 is deleted, and loop 2 can be defined while S2 is deleted. At this time, an end node may appear as an arbitrary line is deleted, and the line S3 connected thereto is removed to leave only the loop component. Finally, while deleting S6, loop 3 can be defined, and S4 and S5 connected to the end node can be removed. As a result, the system of FIG. 13(b) can be defined as initial loop data as shown in Table 4.

loop	Included BSS
One	S1
2	S2
3	S6

In order to understand all the BSSs included in each loop, it is possible to utilize the characteristic of the radial system that 'one loop can be added every time one line is added'.

In the system of FIG. 13 (b), if the BSSs S1, S2, and S6 identified as included in each loop of Table 4 are opened, the system of the radial structure as shown in FIG. 15 can be secured.

Here, as shown in FIG. 16 , data of all the lines included in one loop can be searched by inputting the lines included in each loop of Table 4 one by one and removing the lines connected to the end node and the connected node.

Accordingly, finally, the line of FIG. 13(b) may be defined as the loop data of Table 5. As shown in FIG.

loop	BSS in loop
One	S1, S3, S5
2	S2, S3, S4, S5
3	S6, S4, S5

3) Considerations when exploring loops

In searching for loops existing in the system, an essential element is that all BSSs must belong to at least one loop. If the loop is searched through the process described above, all BSSs can be searched without exception. However, there may be some considerations when exploring loops using only data. Since this increases the number of meaningless switch combinations in the future loop analysis process, it may be desirable to resolve them in advance.

A) When there is no loop between feeders and only loops within the feeder exist

The section corresponding to 'Loop in feeder' of FIG. 1 is treated as 'BSS #2' from the viewpoint of determining the loop structure, and thus may be omitted. The optimal switchgear for this loop structure can be searched after extracting only the relevant part.

B) When loop data of a different combination is searched for in the same data

Even when the same data is given as shown in Table 6, different loop data can be defined as shown in Table 7 by expressing different lines (a) and (b) as shown in FIG. 17 . This results in a difference in the switchgear that can be selected in each loop in the process of using loop data, but does not affect the loop-based analysis itself.

BSS	From	To		BSS	From	To
S1	One	2		S4	2	3
S2	One	3		S5	2	4
S3	One	4		S6	3	4

loop number	BSS in loop
	(a)	(b)
One	S1, S2, S4	S1, S2, S4
2	S4, S5, S6	S1, S3, S5
3	S2, S3, S6	S2, S3, S6

C) When Optimal Loop is not found

Since an arbitrary switch is opened and a switch included in one loop is searched based on this, another loop may exist inside the loop as shown in FIG. 18( b ). In this case, since the number of switchgear that can be selected in each loop does not increase, it may adversely affect the solution of the system reconfiguration, so it should be addressed in advance.

Looking at the loops of FIG. 18 , it appears that loop 1 exists inside loop 2. If loop 1 included in loop 2 is removed, loop 2 can be an optimal loop with a minimum number of switchgear. In order to solve this problem only with data, a loop pair in which two or more BSSs are shared may be searched, and the number of shared BSSs may be compared with the number of non-shared BSSs. When the number of shared BSSs is greater than the number of non-shared BSSs, the two loops may have a containment relationship, and finally the loops may be optimized through data substitution.

loop	comparison loop	shared track	non-shared track	inclusive relationship
2	One	2	One	O
	3	2	One	O

Referring to Table 8, as a result of comparison, it can be determined that loop 2 and loop 1 and loop 2 and loop 3 all have an inclusive relationship. When a person intuitively views the plan view, it can be seen that only loop 1 is included in loop 2, but it can be determined that loop 3 is also included in loop 2 from the data point of view. That is, it should be noted that different optimal loop data may be obtained from the same data as shown in FIGS. 19 and 9 .

loop number	BSS in loop
	(a)	(b)
One	S1, S3, S5	S1, S3, S5
2	S1, S2, S4	S2, S3, S6
3	S6, S4, S5	S6, S4, S5

20 is a flowchart of a loop search algorithm. Referring to FIG. 20 , the flow of the loop search described above may become more clear.

4. Loop-based distribution system analysis method

The present invention can actively utilize the loop concept, which is the basic framework of the distribution system, for solving the distribution system reconfiguration. It is possible to reflect the situation in which the line is temporarily uncontrollable due to the occurrence of a failure in the distribution system and the establishment of a construction plan, and the situation in which multiple power sources exist in one system and must be divided into multiple radial structures in the loop data. Through this, it is possible to respond more flexibly to the physical fluctuations of the system. In addition, since the condition under which an isolated section occurs can be specified using the loop-based data obtained by the loop search described above, the logic of searching the switchgear that converts the loop structure to the radial structure without tracking the connection relationship of the line can be designed.

go. How to reflect the current state of the distribution system

1) Open and uncontrollable switchgear

In a loop system defined only by SBN and BSS through data processing, when one switch is opened, a line previously defined as 'BSS #1' can be converted into two 'BSS #2'. At this time, if the corresponding BSS is included in only one loop as shown in FIG. 21 , the loop is deleted, and if the BSS is included in two or more loops, loop data may be updated by integrating a plurality of loops as shown in FIG. 22 .

Referring to FIG. 23 , when two switchgear, S10 and S37, exist in an open state in the same system, it may be the same as BSS S8 and S2 are opened. In this case, since S2 belongs to one loop and S8 belongs to two loops, S2 may be removed together with loop 1, and loop 3 and loop 4 may be integrated into one while S8 is removed. That is, only three loops may remain in a system in which five loops existed by reflecting the current state of the system.

2) Multiple power sources exist

When a distribution system, a substation, a transformer, and a large number of distributed power sources capable of independent operation are connected to one distribution system, a plurality of radial structures connected to independent power sources should be formed as a result of the distribution system reconfiguration.

Referring to FIG. 24 , when a plurality of power sources exist, a virtual branch (VB) that cannot be opened may be formed between the respective power sources, and thus a connection relationship between the power sources may be formed. there is. When treated as an independent power source, it can be treated as a line that is connected but cannot be opened, and when treated as a load power source, it can be treated as an open line. Accordingly, the added loop can be reflected in the loop data, and at the same time, it can be applied to the loop-based analysis without any specific restrictions while securing the independence of the power source. In this case, the current state may be reflected so that the previously processed data is not changed because the SBN existing in the system is increased due to the virtual line.

25 is a flowchart of a current state reflection algorithm. Referring to FIG. 25 , the flow of the method for reflecting the current state of the power distribution system described above may become more clear.

me. Define the condition for occurrence of an isolated section

When one BSS existing in each loop is opened to solve the systematic reconstruction, an isolation period may occur according to the selected combination. The condition in which the isolation occurs can be defined in advance using the processed loop data and line data described above.

Isolation may mean that all lines connected to one node are switched to an open state. However, in the loop-based analysis method, only one line is selected. Therefore, as shown in FIG. 26 , the number of loops including one node must be greater than or equal to the number of BSSs connected to the node to have a possibility that the node is isolated. When a node with the possibility of isolation is identified, a switchgear combination in which a single node is isolated can be defined through the connection relationship of the line. In addition, when a plurality of nodes are treated as one lump as shown in FIGS. 26 (c) to (e), a condition in which multiple nodes are isolated may be generated if the above-described relationship is satisfied.

SBN No.	Number of BSSs connected to SBNs	Number of Loops with SBN
3	3	2
6	3	3
8	3	3
9	3	3
12	3	2
15	3	2
21	3	2
29	3	2

As shown in Table 10, by comparing the number of BSSs connected to each SBN and the number of loops to which the SBNs belong in the system of FIG. Based on the corresponding SBN, a BSS combination in which a single node is isolated and a condition in which multiple nodes are isolated may be defined as shown in Table 11 and FIG. 28 .

SBN No.	BSS combinations where isolation occurs
6	S1, S4, S5
8	S4, S6, S7
9	S6, S8, S9
6, 8	S1, S5, S6, S7
8, 9	S4, S7, S8, S9
6, 8, 9	S1, S5, S7, S8, S9

all. How to determine the radial structure of the distribution system

In order for the loop structure to be converted into a radial structure, one line must be opened in each loop. In this case, if the same line is selected in different loops, an isolated section may occur for the same reason as in FIG. 7 . Also, when a combination of lines in which isolation occurs is selected, an isolated section may occur for the same reason as in FIG. 26 . That is, if the two cases in which the isolated section occurs do not correspond, the corresponding combination may satisfy the radial structure.

loop
NO	Path
One	S4, S6, S9, S12, S5
2	S6, S8, S11, S7
3	S8, S9, S10
4	S1, S3, S4, S7
5	S1, S2, S5

Table 12 is loop data corresponding to the system in FIG. 27 . Using the loop data in Table 12, BSS combinations such as selections 1 to 3 in Table 13 can be selected in each loop. Here, selection 1 may have an isolated structure as shown in FIG. 29 (a) because the same line is selected in different loops, and selection 2 is similar to FIG. 29 (b) because a combination in which SBNs 8 and 9 are isolated is selected. They can have the same isolated structure. Since option 3 does not correspond to the two cases described above, the radial structure shown in FIG. 30 (a) may be satisfied. In addition, since each BSS has a detailed switchgear set as shown in FIG. 30(b), the radial structure can be satisfied no matter which switchgear combination is selected among them.

loop	Path		Choice 1	choice 2	choice 3	detail opener	Select
One	S4, S6, S9, S12, S5	S4	S9	S12	S15, S16, S17, S36, S32, S31, S30, S29		S36
2	S6, S8, S11, S7	S11	S7	S7	S33		S33
3	S8, S9, S10	S8	S8	S8	S9, S10, S11		S9
4	S1, S3, S4, S7	S4	S4	S4	S6, S7		S7
5	S1, S2, S5	S2	S2	S2	S22, S23, S24, S37	S37

If the loop-based system analysis method is used, it is possible to easily identify the switch combinations that cause isolation in the system as shown in Table 14, and as shown in Table 15, the detailed switch combinations that satisfy the radial structure of the actual distribution system can be obtained. Here, Table 15 shows the results of determining the radial structure using the data derived through the data processing method described above for the IEEE 33/69/118 BUS DISTRIBUTION, which is a representative benchmark model of the distribution system.

Case		Number	Reminder
Selectable full BSS combinations		720	720
Combinations with the same BSS		308	412
Single SBN Isolation	6	22	350
	8	18
	9	22
Multiple SBN Isolation	6, 8	6	336
	8, 9	6
	6, 8, 9	2

IEEE Test System	IEEE33 BUS	IEEE69 BUS	IEEE118 BUS
Number of BSS combinations that require radial judgment	720	720	10,450,944,000
The number of BSS combinations that satisfy the radial structure	336	336	678,187,952
Total number of possible actuator combinations	50,751	407,924	3* More than

5. Globally optimal system reconstruction for the purpose of minimizing loss

The present invention can utilize a genetic algorithm that is one of artificial intelligence techniques and a case in which all possible combinations are considered to perform distribution system reconfiguration for the purpose of minimizing line loss. In this case, in order to minimize the calculation time of line loss, a 'distflow branch equation' technique specialized for calculating the current in the distribution system may be used. Thereafter, it is possible to improve the performance of the grid reconfiguration through the constraint that utilizes the characteristics of the distribution system while performing the global optimal distribution system reconfiguration using the possible combination of all switchgear and the genetic algorithm.

go. Algae Calculation for Distribution Grid

Calculation of current flow in the power system is performed at each point through line data and node data.

It is a systematic analysis method that calculates and estimates the current flowing in the track. The analysis techniques of the 'Newton-Raphson' and 'Gauss-Seidel' series, which are often used for tidal current calculation, are specialized in the transmission system among the power systems. . However, since the distribution system is a radial structure with very low interconnectivity, analysis techniques for the transmission system generate unnecessary computational processes. Therefore, the tidal current calculation technique that improves the operation time and is optimized for the distribution system has been proposed as follows.

The 'Distflow branch equation' proposed by Baran can be mainly used as a tidal current analysis technique for the distribution system. This is the line impedance of the system

and load

When the value of is known, the voltage of all nodes

is 1[pu], and through Equation 5.1 'Backward update',

to calculate After that, using Equation 5.2 'Forward update',

, and finally, when the measured power value at the end of the line is the same as the calculated value, the calculation is terminated. If they do not match, the error may be improved by repeatedly performing Equations 2 and 3 with appropriate compensation.

here,

is the active and reactive power flowing in line i,

is the impedance component of line i. Also, referring to FIG. 31 , power distribution at a branching point k in the 'Forward Update' process may be distributed based on the power added in the 'Backward Update' process. If the line loss value and voltage drop are ignored in Equations 2 and 3, 'Simplified Distflow', which is a simpler tidal current calculation expression, can be obtained as Equations 4 and 5.

Table 16 compares the execution times of 'Newton-Raphson method' and 'Simplified Dist-flow' in the IEEE 33/69/118 BUS DISTRIBUTION SYSTEM under 17 CPU 3.60 GHz, RAM 16GB, and Matlab environment.

IEEE Test System	IEEE33 BUS	IEEE69 BUS	IEEE118 BUS
Gauss-Seidel	0.085 seconds	14 seconds	1.75 seconds
Newton-Raphson	0.005 seconds	0.01 seconds	0.02 seconds
Simplified Distflow	8.27* candle	1.22* candle	2.5* candle

As shown in Table 17, when arranging switchgear combinations according to line loss values in IEEE 118 BUS DISTRIBUTION NETWORK, there is no significant difference in the order of switchgear combinations except for Rank 5 and 6. Therefore, when solving the distribution system reconfiguration problem, it is more important to find a switch combination that minimizes loss than to obtain an accurate loss value.

BSS Combination	Simplified Distflow power loss [kW]	Rank		Newton-Raphson power loss [kW]	Rank
One	0.779561	One		0.853923	One
2	0.779939	2		0.854376	2
3	0.780121	3		0.854709	3
4	0.780181	4		0.854743	4
5	0.780249	5		0.855031	6
6	0.780500	6		0.854993	5

me. Simulation of all possible combinations

The simplest and surest way to find the global optimal solution is to perform systematic analysis for all possible combinations by controlling the closing/opening of switches belonging to all BSS combinations determined as radial structures through structural analysis of the system. Thereafter, the switch combination in which the line loss is the least can be treated as the global optimal solution of the corresponding system.

1) Complete Investigation Process

By the method described above, it is possible to search for the BSS included in each loop through data processing of the system, and to derive a BSS combination that satisfies the radial structure. In addition, since detailed switchgear constituting each BSS is known, a detailed switchgear combination for which actual closing/opening control is to be performed can be grasped, and accordingly, line loss can be calculated. 32 is a flow chart for simulating all possible combinations. Referring to FIG. 32 , the flow of simulation for all possible combinations can be made clearer.

3) Limitations of full investigation

When performing a full investigation of the IEEE 33/69 BUS DISTRIBUTION SYSTEM, the system reconfiguration solving time takes 4.2 seconds and 50 seconds, respectively, as shown in Table 18. However, in the case of IEEE 118 BUS DISTRIBUTION SYSTEM, which has a more complex structure, there are more than 10,450,944,000 BSS combinations that need to determine the radial structure, and the number of detailed switchgear combinations far exceeds 3,000 trillion. Therefore, as the system becomes larger and more complex, searching for the optimal system configuration by simulating all possible combinations requires a very long execution time and memory, and is impossible to perform in reality.

IEEE Test System		IEEE 33 BUS		IEEE 69 BUS		IEEE 118 BUS
All possible switchgear combinations	50,751	407,924	3* More than
Distflow branch equation	4.2 seconds	50 seconds	-

4) Measures to improve the performance of the full investigation

Even if the system is abbreviated, the time required for a full investigation is prolonged because the number of detailed switchgear combinations that need to be analyzed is large. In order to effectively shorten this, the fact that the nodes of the distribution system are divided to have a constant load can be used. By identifying the center of the BSS and limiting the number of detailed switchgear that can be selected within one BSS, it is possible to conduct a full investigation more quickly.

Using Equations 6 and 7, it is possible to calculate a line loss value according to the direction of current when a specific switchgear is opened in the corresponding BSS. And, it is possible to treat the switchgear that minimizes the loss occurring in the entire line of the BSS as a center line.

Referring to FIG. 33 , it can be divided into a left section and a right section according to the opening position of the switchgear, and the line loss according to each section is shown in Table 19. As a result, S3 having the minimum total loss may be treated as the center line of FIG. 33 , and the switchgear utilized in the corresponding BSS may be defined as S1, S3, and S6.

open switch	Left Segment Loss	right leg loss	total loss
S1	0	0.9968	0.9968
S2	0.0246	0.8993	0.9239
S3	0.3827	0.1595	0.5422
S4	0.4048	0.1390	0.5438
S5	0.5701	0.549	1.1191
S6	0.9878	0	0.9878

In the case of 'IEEE 69 BUS DISTRIBUTION NETWORK', as shown in Table 20, the number of systematic analysis times and execution time can be significantly improved by identifying the center of the line set and restricting the openable detailed switchgear as shown in Table 21.

Branch Switch Set data
BSS	Before included Switch	After included switch
One	s4, s5, s6, s7, s8	s4, s6, s8
2	s46, s47, s48, s49, s73	s3, s38, s39, s42
3	s3, s35, s36, s37, s38, s39, s40, s41, s42	s46, s48, s73
4	s9, s10	s9, s10
5	s52, s53, s54, s55, s56, s57, s58	s52, s57, s58
6	s11, s12	s11, s12
7	s69		s69
8	s13, s14	s13, s14
9	s70		s70
10	s15, s16, s17, s18, s19, s20	s15, s17, s20
11	s43, s44, s45, s71	s43, s45, s71
12	s21, s22, s23, s24, s25, s26, s72, s64, s63, s62, s61, s60, s59	s21, s72, s59

IEEE Test System		IEEE 33 BUS		IEEE 69 BUS
All possible switchgear combinations	6,768	23,949
Distflow branch equation	0.5 seconds	3 seconds

all. Simulation using genetic algorithm

It is a difficult task in reality to analyze the combination of all the detailed switchgear that have a radial structure in a given distribution system as a loop system. As shown in Table 18, since the number of switches satisfying the radial structure increases exponentially as the system becomes larger and more complex, a method for efficiently searching for the optimal switch combination is required.

For this combinatorial optimization problem, a wide variety of meta-heuristic methods and artificial intelligence methods such as TS, SA, GA, ACO, and PSO are being studied. It is the best method for the problem. The most important thing to utilize this is to express the genetic trait that is the basis of the algorithm in the form of an array. In the phylogenetic reconstruction technique using the existing genetic algorithm, a specific logic was not explained for expressing the switchgear components in an array form, but the genetic trait can be expressed more systematically and logically by using the loop-based analysis technique proposed in the present invention. .

2) Loop-based lineage reconstruction through genetic algorithm

In order to apply the loop-based data to the genetic algorithm, genes constituting the genome can be defined as BSSs randomly selected in each loop as shown in ' Selection 1, 2, 3' in Table 13. After that, each BSS can also randomly select a detailed switchgear to form a detailed switchgear combination subordinate to the BSS. As shown in Table 22, when the first generation is formed as much as the input population, the radial structure can be determined based on the loop data. In this case, traits that do not satisfy the radial structure may be given a lower weight in the future by giving a high level of penalty to the fitness part.

It is possible to preserve the dielectric with optimal line loss as it is to the next generation and perform crossing over the entire dielectric. In addition, by utilizing 'Roulette wheel selection', it is possible to differentiate the possibility that each trait can be selected according to the line loss value of each trait. Since the lower the line loss, the better the value. Accordingly, the fitness of each trait can be calculated by using Equations 8 and 9 to give the fitness. The x and 100 traits in Table 22 were penalized because an isolated section occurred.

Generation											fitness
No	BSS					Switch
One	S12	S7	S8	S4	S2	S36	S33	S9	S7	S37	2.18
:	:	:	:	:	:	:	:	:	:	:	:
x	S4	S11	S8	S4	S5	S7	S35	S9	S6	S27	0.68
:	:	:	:	:	:	:	:	:	:	:	:
99	S12	S8	S10	S1	S2	S32	S11	S14	S4	S37	1.16
100	S9	S7	S8	S4	S2	S34	S33	S10	S7	S24	0.68

Since the BSSs selected for each loop do not affect each other, the diversity of genetic traits can be secured by setting a high level of mutation probability and performing crossover using the 'Uniform crossover method'. Two parental traits can cross over according to their cross probability. At this time, the crossing proceeds with respect to the BSS, and the detailed switchgear may be performed in the same manner as the BSS. If crossover is not performed, the parental trait can be passed on to the offspring.

parental trait
No	BSS					Switch
One	S12	S7	S8	S4	S2	S36	S33	S10	S7	S37
cross			↕	↕
99	S12	S8	S10	S1	S2	S32	S11	S14	S4	S37

offspring trait
No	BSS					Switch
One	S12	S7	S10	S1	S2	S36	S33	S14	S4		S37
2	S12	S8	S8	S4	S2	S32	S11	S10	S7	S37

Thereafter, as shown in Table 25, the next generation can be constructed by applying the mutation probability to each gene. Through mutation, the BSS selected in each loop can be changed randomly, and the detailed switchgear can also be randomly reselected accordingly.

offspring trait
No	BSS					Switch
One	S12	S7	S10	S1	S2	S36	S33	S14	S4	S37
Mut	O	X	X	O		X
2	S12	S8	S8	S4	S2	S32	S11	S10	S7	S37
Mut	X	O	X	X	O

offspring trait
No	BSS					Switch
One	S5	S7	S10	S3	S2	S26	S33	S14	S2		S37
2	S12	S11	S8	S4	S1	S32	S21	S10	S7	S3

The genetic algorithm has various convergence characteristics as shown in FIG. 34 according to the genetic trait formed in the first generation even when the same parameters are input. In addition, IEEE 118 BUS DISTRIBUTION NETWORK' has more than 12,000 local optimal solutions. The present invention can perform migration between islands by forming a plurality of islands independently performing parallel operations. As a result, by sharing the dielectric being formed on each island, the possibility of deriving a global optimal solution can be improved. And, it is possible to operate the Elite island by extracting only the optimal genetic traits from each island. Tables 27 to 29 show the optimal lineage reconstruction using the genetic algorithm for various parameters in the CPU 3.60GHz, RAM 16GB, and Matlab environment. is the result of Among the input parameters, the elite ratio to the total population was fixed at 1%, the crossover probability was 85%, and the migration probability was fixed to 1% of the total population. In addition, if the radial structure is not satisfied or the fitness value has an insignificant numerical value even if it is satisfied, a penalty corresponding to 120% of the power loss value in the state in which the entire switchgear is turned on may be given. Here, each fixed probability may be experimentally varied.

The IEEE 33/69 BUS DISTRIBUTION NETWORK of Tables 27 and 28 is the average of 1000 repetitions for all cases, and the IEEE 118 BUS DISTRIBUTION NETWORK of Table 29 is the average of 100 repetitions. In addition, the optimal solution calculation time in the simulation can be obtained because the information of the global optimal solution for a static system is known, and through this, the performance of the genetic algorithm can be verified. Since the genetic algorithm is an algorithm based on randomness, it cannot guarantee a global optimal solution unless a long execution time is taken into account. However, even if the global optimal solution is not searched, the obtained solution has an excellent result value, so that each parameter can be appropriately selected according to the situation.

Since the currently implemented genetic algorithm program operates the independent operation of the island with one process core, it is not performed as a practical parallel operation. can be improved effectively.

No	population	Max generation	Island number	Muta tion rate	Execution time	Optimal solution
						rate [%]	avg time[s]/generation	min time[s]/ generation	max time[s]/ generation
One	100	10	5	40	0.36	88.6	0.226/6.1	0.045/1	0.433/10
2	100	20	5	40	0.73	99.7	0.260/6.9	0.045/1	0.711/20
3	100	40	5	40	1.44	100	0.268/6.9	0.044/1	6.629/20
4	100	40	One	40	0.84	89.8	0.356/17	0.015/1	1.167/40
5	100	40	3	40	0.95	99.5	0.222/8.1	0.027/1	0.980/20
6	100	20	One	20	0.26	55.4	0.138/10.8	0.009/1	0.311/20
7	100	20	3	20	0.53	95.3	0.207/7.8	0.029/1	0.541/20
8	100	20	5	20	0.80	99.6	0.258/6.3	0.046/1	0.803/20
9	100	20	10	20	1.49	100	0.363/4.7	0.095/1	0.829/20
10	50	20	10	20	0.86	99.4	0.293/6.6	0.053/1	0.921/20
11	50	10	10	10	0.42	91.4	0.255/5.2	0.049/1	0.523/10
12	100	10	10	10	0.80	99.9	0.370/4.5	0.096/1	0.816/10
13	50	20	5	40	0.37	86.4	0.182/9.7	0.024/1	0.44/20
14	50	30	5	40	0.59	96.4	0.219/11.1	0.025/1	0.575/40
15	50	40	5	40	1.29	98.4	0.386/8.8	0.041/1	1.742/40
16	30	40	10	40	0.85	98.7	0.253/11.5	0.030/1	0.867/40
17	50	40	10	40	1.37	100	0.277/7.6	0.046/1	1.141/40
18	50	20	10	40	0.72	98.9	0.279/7.4	0.049/1	0.717/40

No	population	Max generation	Island number	Muta tion rate	Execution time	Optimal solution
						rate [%]	avg time[s]/generation	min time[s]/ generation	max time[s]/ generation
One	100	20	5	20	1.06	99	0.255/4.3	0.067/1	1.052/20
2	100	20	5	40	0.99	99.1	0.242/4.3	0.063/1	0.957/40
3	100	10	5	40	0.57	97.0	0.268/4.8	0.074/1	0.552/40
4	100	40	5	40	1.90	100	0.258/4.6	0.065/1	1.372/40
5	100	20	5	60	0.98	99.4	0.283/5.2	0.066/1	0.928/69
6	100	20	10	20	2.02	100	0.374/3.1	0.138/1	3.079/20
7	50	20	10	20	1.03	98.8	0.277/4.7	0.069/1	1.059/20
8	50	10	20	40	0.99	99.9	0.416/3.8	0.146/1	0.988/40

No	population	Max generation	Island number	Muta tion rate	Execution time	Optimal solution
						rate [%]	avg time[s]/generation	min time[s]/ generation	max time[s]/ generation
One	1000	100	10	0	189.06	64	60.0/44.3	22.60/26	126.54/73
2	1000	100	10	20	68.27	67	28.7/84.0	11.91/51	65.17/117
3	1000	100	10	40	45.37	41	21.5/47.0	10.43/28	37.08/78
4	1000	100	10	60	34.61	23	25.3/73.6	17.33/51	33.16/96
5	1000	100	20	40	88.13	81	34.6/43.3	17.14/24	87.55/100
6	1000	200	10	40	96.89	62	31.6/67.9	10.74/27	92.35/174
7	1000	300	10	40	138.92	66	35.5/80.5	12.01/31	131.30/282
8	500	100	10	40	24.68	44	14.2/59.9	8.15/36	24.66/99
9	500	200	10	40	49.81	52	19.6/81.0	7.90/36	53.52/196
10	500	300	10	40	74.68	59	24.4/101.4	6.70/31	76.16/299
11	1000	200	5	40	51.21	46	19.5/79.5	8.49/38	47.10/180
12	1000	200	15	40	134.33	75	41.4/65.5	18.19/32	131.21/200
13	500	100	20	20	70.73	70	28.0/43.2	14.35/24	64.61/92
14	500	100	20	40	48.09	59	23.4/51.4	51.86/100	13.49/30
15	100	100	20	20	11.74	17	8.5/73.9	5.77/52	11.54/97
16	100	200	20	20	24.71	31	12.0/100.1	7.08/62	20.22/166
17	100	300	20	20	36.76	23	15.9/134.6	4.26/44	34.04/284
18	100	100	50	20	42.03	55	21.6/54.4	13.38/36	40.85/97
19	100	100	100	20	82.93	90	35.2/45.9	20.94/29	77.06/92
20	100	100	100	40	56.68	72	29.6/54.8	14.91/30	55.03/96
21	200	100	100	20	117.95	91	53.2/49.7	29.22/32	105.21/93

35 is a flowchart of a genetic algorithm. Referring to FIG. 35 , it may be clearer for the loop-based systemic reconstruction flow through the genetic algorithm.

La. How to improve the performance of the distribution grid reconfiguration

In a fixed system in which physical fluctuations do not occur due to load drop, line opening, independent operation, etc., the system configuration that satisfies the optimal operating efficiency may vary depending on the current electrical state of the system. However, due to the characteristics of the distribution system, the BSS connected to the start point of the feeder or a large load can be fixed in the always-on state without participating in system reconfiguration. That is, when the load change rate from the worst condition to the best condition is applied in one system, if the BSS in which the input state is always maintained can be specified, the BSS can be treated as an unnecessary component in the optimal system configuration search process. .

For this purpose, the initial system of 'IEEE 118 BUS DISTRIBUTION NETWORK' was divided into four zones as shown in FIG. 36 . In addition, the BSS that maintains the constant input state was searched by assigning the load change rate as shown in Table 30 to each zone as the scenario in Table 31. In this case, the load variation rates corresponding to types A to D can be randomly assigned to each node, and fixed values are assigned to types E and F to simulate extreme situations. Cases 1 to 5 used only types E and F, and cases 6 to 13 were repeated 5 times to search for a global optimal solution. BSS finally participating in the lineage reconstruction can be abbreviated to FIG. 37 and Table 32.

type	load variation		type	load variation
A	+30 to +70 %		D	+50 to -50%
B	-30 to +30 %		E	+70%
C	-70 to -30%		F	-70%

CASE		Zone 1	Zone 2	Zone 3	Zone 4
One	E	E	E	E
2	F	F	F	F
3	E	F	F	E
4	F	E	E	F
5	E	F	E	F
6	B	B	B	B
7	A	A	A	A
8	C	C	C	C
9	A	C	C	A
10	A	B	C	C
11	A	A	C	A
12	C	A	C	C
13	D	D	D	D

Loop
No	Before Path fixed	After Path fixed
One	S39, S40, S41	S40, S41
2	S2, S3, S28, S35, S40	S35, S40
3	S37, S36, S35, S41	S37, S36
4	S38, S31, S33, S32		S33
5	S27, S28, S29, S31, S34, S36	S29, S31
6	S1, S2, S4, S7, S18, S26, S27, S30, S38	S26, S38
7	S29, S30, S32	S29, S32
8	S16, S18, S20, S25		S25
9	S20, S21, S24, S19	S21, S24
10	S14, S15, S21, S23, S13	S15, S21, S23
11	S8, S15, S22, S9, S12, S13		S22
12	S6, S7, S9, S12, S14, S16, S19	S14, S19
13	S6, S7, S8, S17		S17
14	S10, S12, S13, S11	S10, S13
15	S4, S5, S6, S9, S11	S11

By utilizing the realistic characteristics of the distribution system, it is possible to limit BSSs that do not participate in the distribution system reconstruction solution. Thereafter, by applying the reduced loop-based data to the genetic algorithm, distribution system reconstruction can be performed, and simulation results for various parameters may be shown in Table 33.

No	population	Max generation	Island number	Mutation rate	Execution on time	Optimal solution
						rate[%]	avg time[s]/generation	min time[s]/gneration	max time[s]/generation
One	100	100	10	40	13.83	76.5	8.58/60	2.886/20	14.334/99
2	100	100	20	40	27.60	94.5	14.26/50	5.842/20	27.507/99
3	100	100	20	20	31.52	99.6	11.02/34	4.721/14	28.752/92
4	200	100	10	40	24.46	48.6	18.02/73	5.419/21	24.719/99
5	200	100	10	20	27.73	84.9	16.20/57	6.487/23	28.087/98

Through this, the search probability and speed of the global optimal system configuration of 'IEEE 118 BUS DISTRIBUTION NETWORK' could be improved from 90% and 35.2 seconds to 99.6% and 11.02 seconds. The method of restricting the BSS not participating in such phylogenetic reconstruction is not limited to the genetic algorithm, but can be a method to improve the performance of the phylogenetic reconstruction itself.

mind. case study

1) IEEE 33/69/118 BUS DISTRIBUTION NETWORK

Table 34 shows the results of performing global optimal system reconfiguration targeting IEEE 33/69/118 BUS DISTRIBUTION NETWORK.

Case	division	BSS Combination	Detailed switchgear combination	Loss[kW]
33	Initial	S2, S7, S9, S11, S12	s37, s33, s34, s35, s36	192
	Optimal	S2, S4, S8, S10, S12	s37, s7, s9, s14, s32	134
69	Initial	S2, S7, S9, S11, S12	s73, s69, s70, s71, s72	224
	Optimal	S5, S7, S8, S9, S12	s56, s69, s14, s70, s61	110
118	Initial	S8, S9, S10, S14, S15, S20, S23, S24, S26. S32, S33, S34, S35, S37, S40	s124, s120, s119, s132, s118, s123, s121, s122, s125, s126, s128, s127, s129, s130, s131	1298
	Optimal	S6, S11, S13, S19, S21, S22, S23, S25, S29, S31, S33, S35, S37, S38, S40	s39, s23, s25, s34, s50, s42, s121, s58, s71, s74, s97, s129, s130, s95, s109	853

2) Grid reconfiguration considering load variability

The system of 'IEEE 118 BUS DISTRIBUTION NETWORK' can be divided into six zones as shown in FIG. 38 . And, among the loads of the six contract types of Korea Electric Power Corporation, 'general type', 'residential type', 'industrial type', and 'education type' can be used to allocate loads.

In order to give differential loads to each zone, loads in the range shown in Table 35 can be randomly assigned according to the gross domestic product and power consumption by contract type. Using the load pattern for each time period corresponding to Monday in summer in Table 36 and FIG. 39 where the deviation between the minimum and maximum values of the load during the year is the largest, a real-time distribution system reconstruction solution can be performed after establishing a system environment similar to the actual one. there is.

load type	quantity	Load range (MW)
		minimum	maximum
general type	26	0.0748	0.1496
residential	45	0.1496	0.2244
industrial	21	0.2992	0.3741
General + Residential	15	0.0748	0.2244
educational	11	0.1496	0.2244

case	Normal	dwelling	industry	education
One	42.12	73.89	68.11	32.23
2	39.00	63.88	67.85	30.98
3	36.87	57.44	67.58	29.96
4	35.52	53.86	67.49	29.44
5	35.22	52.60	67.58	29.07
6	36.76	52.63	68.17	29.00
7	41.77	59.85	70.36	30.91
8	51.50	70.11	76.24	39.65
9	68.50	73.68	89.33	59.91
10	84.66	74.41	98.36	82.45
11	93.75	74.06	99.62	91.85
12	97.05	74.27	100.00	95.89
13	97.23	75.19	90.85	95.37
14	99.29	75.90	96.78	99.49
15	100.00	75.49	99.09	100.00
16	99.82	75.65	98.45	96.40
17	99.47	76.93	98.48	88.99
18	97.23	80.25	94.80	75.18
19	89.20	85.49	90.94	62.78
20	83.78	93.54	88.22	57.42
21	77.40	99.14	86.12	53.82
22	69.14	100.00	83.66	48.09
23	59.94	96.40	83.69	42.36
24	52.15	88.35	82.90	38.11

Referring to FIG. 39 , as a result of simulation for calculating the global systematic optimal solution for 24 hours a day corresponding to Monday in summer, the results shown in Table 37 may be derived. At this time, if the optimal system reconfiguration is performed for each time period, the line loss of 6.784MWh compared to the initial system can be improved by changing the system configuration 6 times. By changing Equation 8 to Equation 10, 24 load patterns in one switchgear combination can be used to obtain one optimal system composition for 24 hours. At this time, when the system is operated without switchgear control with one optimal system configuration, the line loss of 6.684MWh can be improved compared to the initial system.

here,

is the power loss value at time t for the i-th system configuration. The difference in loss between performing optimal system reconfiguration for each time period and operating as one system configuration is 0.1 MWh. That is, unless a physical change occurs in a general system, the performance of system reconfiguration for each time period may appear very insignificant.

time	full load [MW]	Line loss [MW]
		early lineage	Optimal system for each time zone	one optimal strain
One	16.609	0.678	0.478	0.485
2	15.173	0.577	0.406	0.41
3	14.233	0.517	0.363	0.367
4	13.709	0.487	0.34	0.344
5	13.537	0.478	0.333	0.337
6	13.643	0.485	0.338	0.342
7	14.953	0.568	0.399	0.403
8	17.241	0.739	0.523	0.527
9	19.726	0.974	0.692	0.694
10	21.570	1.169	0.835	0.835
11	22.168	1.222	0.877	0.877
12	22.435	1.247	0.896	0.896
13	21.886	1.136	0.829	0.829
14	22.570	1.234	0.895	0.895
15	22.724	1.263	0.913	0.913
16	22.617	1.250	0.902	0.902
17	22.618	1.252	0.90	0.900
18	22.403	1.213	0.871	0.871
19	22.217	1.189	0.851	0.852
20	22.716	1.239	0.883	0.889
21	22.947	1.271	0.900	0.911
22	22.434	1.226	0.864	0.877
23	21.499	1.146	0.804	0.815
24	20.027	1.011	0.706	0.716
total	471.655	23.571	16.798	16.887
Improvement compared to the initial system [MWh]			6.784	6.684

3) Systemic reorganization considering malignant load

It is assumed that a malicious load having output fluctuation characteristics as shown in Table 38 exists in the nodes 49, 60, 84, 97, and 106 of the line of FIG. 36, and the global optimal system reconstruction simulation for 24 hours a day was performed.

time	Relative load		time		Relative load
7	0.063		14	0.805
8	0.268		15	0.675
9	0.383		16	0.423
10	0.363		17	0.380
11	0.779		18	0.208
12	0.990		19	0.048
13	1.000		20	0.005

As a result of the simulation, the results shown in Table 39 were derived. At this time, if the optimal system reconfiguration is performed for each time period, the line loss of 5.005MWh compared to the initial system can be improved by 12 system configuration changes. Since the difference in loss improvement between performing optimal system reconfiguration for each time period and operating a system with one system configuration is 1.068MWh, the performance of system reconfiguration by time period in a system with rapid load fluctuations can be improved. can be checked In other words, it is possible to perform an economic evaluation based on the loss improvement amount to perform appropriate control according to the judgment of the operator.

time	full load [MW]	Line loss [MW]
		early lineage	Optimal system for each time zone	one optimal strain
One	16.609	0.678	0.478	0.481
2	15.173	0.577	0.406	0.406
3	14.233	0.517	0.363	0.363
4	13.709	0.487	0.340	0.34
5	13.537	0.478	0.333	0.333
6	13.628	0.484	0.338	0.338
7	14.008	0.515	0.365	0.365
8	13.221	0.526	0.384	0.401
9	13.981	0.641	0.481	0.514
10	16.125	0.796	0.601	0.636
11	10.483	0.692	0.545	0.683
12	7.585	0.737	0.584	0.794
13	6.886	0.701	0.546	0.767
14	10.495	0.703	0.555	0.706
15	12.599	0.729	0.577	0.685
16	16.272	0.822	0.629	0.683
17	16.918	0.853	0.645	0.689
18	19.283	0.969	0.713	0.731
19	21.497	1.126	0.878	0.884
20	22.641	1.233	0.878	0.884
21	22.947	1.271	0.900	0.908
22	22.434	1.226	0.864	0.872
23	21.499	1.146	0.804	0.809
24	20.027	1.011	0.706	0.709
total	375.79	18.918	13.913	14.981
Improvement compared to the initial system [MWh]			5.005	3.937

6. Conclusion

As a result of performing global optimal system reconstruction simulation using a genetic algorithm, if the load fluctuation is not large, the performance may be very small even if different load patterns are applied for each time period. However, in the presence of a malignant load with large load fluctuations, the performance of the optimal system reconfiguration for the purpose of minimizing the line loss of the system may be visibly revealed. As a result, the current distribution system reconfiguration may seem unnecessary, but if various distributed power sources and a microgrid capable of irregular independent operation are connected to the grid in the future, the meaning of the global optimal grid reconfiguration may increase.

In addition, by performing the optimal system reconfiguration based on the current state of the system, it is possible to perform more optimal system operation than before. Through this, it will be possible to suppress the expansion of power facilities by improving the utilization rate of existing facilities, and to increase the grid connection rate of new and renewable energy sources, and furthermore, it can become the basis for system modernization, automation, and intelligence.

In addition, the loop-based distribution system reconfiguration method according to an embodiment of the present invention is substantially performed by a computer system in which a loop-based distribution system reconfiguration program is installed.

That is, the present invention may be provided in the form of a computer in which the loop-based distribution system reconfiguration program is stored, a smart distribution operating system, a distribution operating system, or a distribution automation system.

In addition, the loop-based distribution system reconfiguration program is stored in a server system, and the computer system may download and install the loop-based distribution system reconfiguration program from the server system, and then perform distribution system reconfiguration.

In addition, the loop-based distribution system reconfiguration program may be separately stored and provided in a recording medium, and the recording medium is specially designed and configured for the present invention or is known and used by a person skilled in the art of computer software. It may be possible, for example, hard disks, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD and DVD, magneto-optical recording media capable of both magnetic and optical recording, ROM, RAM, flash It may be a hardware device specially configured to store and execute program instructions, alone or in combination, such as a memory.

In addition, the loop-based distribution system reconfiguration program may be a program composed of program instructions, local data files, local data structures, etc. alone or in combination. It may be a program written in high-level language code that can be executed by a computer.

Claims (10)

1. A system simplification step of omitting a part of the distribution system and abbreviated;

   a loop search step of defining a loop using the simplified system by the system simplification step and searching for a switch included in the defined loop;

   a distribution system analysis step of defining a switchgear combination in which isolation occurs in the system using the switchgear found by the loop search step, and analyzing the detailed switchgear combination satisfying the radial structure of the distribution system; and

   and calculating a line loss using the detailed switchgear combination, and deriving a global optimal solution based on the calculated line loss.
2. The method of claim 1,

   The system simplification step determines a situation in which an isolated section can be generated, and simplifies the system in response to the determined situation.
3. The method of claim 1,

   The distribution system analysis step is a loop-based distribution system reconfiguration method, characterized in that when a plurality of power sources exist in the distribution system, a virtual branch (VB) that cannot be opened between the respective power sources is formed.
4. The method of claim 1,

   In the grid reconfiguration step, the loop-based distribution system reconfiguration method, characterized in that the number of detailed switchgear that can be selected in one BSS is limited by identifying the center of the BSS.
5. The method of claim 1,

   The grid reconfiguration step is a loop-based distribution system reconfiguration method, characterized in that the search for a BSS that maintains the always-in state, and limiting the searched BSS.
6. A loop-based distribution system reconfiguration program stored in a medium for performing the distribution system reconfiguration method of any one of claims 1 to 5 in combination with a computer.
7. A server system that stores the distribution system reconfiguration program of claim 6 and transmits the distribution system reconfiguration program through a communication network.
8. A smart distribution management system that stores the distribution system reconfiguration program of claim 6 and reconfigures the distribution system by the distribution system reconfiguration program (SDMS: Smart Distribution Management System).
9. A distribution management system (DMS) that stores the distribution system reconfiguration program of claim 6 and reconfigures the distribution system by the distribution system reconfiguration program.
10. A distribution automation system that stores the distribution system reconfiguration program of claim 6 and reconfigures the distribution system by the distribution system reconfiguration program (DAS: Distribution Automation System).

Images