JP4424494B2 - Distribution system configuration optimization device and distribution system configuration optimization method - Google Patents

Description

  The present invention relates to a distribution system configuration optimizing device and a distribution system configuration optimizing method capable of representing a distribution system with a graphics model and determining an interconnection switch that minimizes distribution loss.

  In general, as shown in FIG. 1, a distribution network formed by a plurality of distribution lines is divided into a plurality of load sections by a switch in order to ensure supply reliability at the time of an accident.

  Each load section receives supply of electric power from a feeder. And the distribution line of a certain feeder is connected to the distribution lines of other feeders by a normally open interconnection switch. Each feeder of such a power distribution system employs a multi-division multi-connection system (for example, a 3-part 3-connection system or a 6-part 3-connection system) that operates in a radial configuration.

Therefore, the system configuration changes depending on which switch is used as the interconnection switch, and accordingly, the power distribution loss varies greatly. That is, since the loss is proportional to the square of the current, for example, the power loss L _A in FIG. 2 (A), the
L _A = (25 ² × 0.03) + (15 ² × 0.02) + (5 ² × 0.01) = 23.5 [W]
The power loss L _B in FIG.
^{_{L B = (10 2 × 0.03}} ) + (15 2 × 0.02) + (20 2 × 0.01) = 11.5 [W]
It becomes. Therefore, the power loss when the distribution system of FIG. 2 (B) is adopted is smaller than the power loss when the distribution system of FIG. 2 (A) is adopted.

  In the distribution system, it is necessary to determine the combination of opening and closing of a large number of switches so that the distribution loss is minimized. However, in an actual power distribution system, consumers are spread out in a mesh structure with a surface configuration, and the number of switches is usually about 1000 even at the sales office level. For this reason, the total number of system configuration candidates (that is, the total number of combinations of opening and closing switches) is enormous.

For example, if the switch is of 1000, the system configuration candidate is ^{2 1000,} also one of the configurations as could be evaluated in ^{10 -30} seconds, it takes a total of about 10293 years. When the distribution loss is calculated and compared for all combinations of opening and closing of these switches, the calculation time is enormous, which is not realistic.

  Generally, the distribution loss in the actual system is 3 to 4%, but cost reduction of several hundred million yen / year is expected by improving the distribution loss by 1%, for example. However, in the construction of a distribution system, an optimization process that minimizes power loss has not been achieved.

  By the way, in order to reduce power loss in an electric power system, an optimization technique (for example, approximate optimization technique using tabu search, genetic algorithm, simulated annealing, etc.) applying a metaheuristic method has been developed. . The metaheuristics method gives up searching for all combinations of solutions, and based on a certain rule or concept, a solution with an accuracy close to that (approximate optimal solution) can be obtained in a short time.

  However, in this metaheuristics method, the acquisition of the global optimal solution itself is not guaranteed, so it is unclear whether the obtained solution is the optimal solution, and the number of candidate solutions is large. Therefore, there is a problem that it is impossible to verify whether or not the obtained solution is an optimal solution.

  The present invention uses a strict solution method such as ROBDD (Reduced Ordered Binary Decision Diagram) in a distribution network adopting multi-division multi interconnection such as a 3-division tri-connection power distribution system, etc., and selects candidates that efficiently satisfy operational constraints. An object of the present invention is to provide a distribution system configuration optimizing device and a distribution system configuration optimizing method capable of determining a strict distribution loss minimum configuration by squeezing. The above can be applied to a power distribution network composed of any multi-divided multi-connected system. However, for the sake of convenience of explanation, the description will be made taking a 3-divided 3-connected system as an example.

The distribution system configuration optimizing device according to the present invention includes a distribution system information storage device that stores information on a distribution system and the distribution system represented by a graphics model, which is closed by a rising portion that is less than X-connected (such as triple-connected) and node model information storage device for storing information into a plurality of partial feeders which are, the multiple parts feeders stored in the node model information storage device to all cases search, Tashi satisfy the radial configuration constraints Strictly determine all the combinations of the partial feeder creation means for obtaining the division mode satisfying the Y division (three divisions, etc.) constraint and the division mode satisfying the feeder capacity restriction among the division modes obtained by the partial feeder creation unit. solution and detail calculation means for calculating the (ROBDD etc.), the combination of the divided manner calculated by the precise calculation means, said portion in the feeder The distribution losses calculated for, characterized by comprising a means for identifying the one candidate 1 is distribution losses from the candidate that satisfies the voltage drop constraint becomes minimum as distribution loss minimum configuration.

The power distribution system configuration optimization method of the present invention includes a step of dividing a power distribution system represented by a graphics model into a plurality of closed partial feeders that are less than X-connected (such as triple-connected) by a rising portion, the feeder meets the radial configuration constraints, Y split the steps of a dividing manner that satisfies (3 splits) constraints exhaustive search, the dividing manner to meet the line capacity constraint of the feeder roots of the dividing manner determined by the exhaustive search calculating the exact algorithm all combinations of the combinations of the divided manner calculated by exact algorithm, the power distribution loss in the portion feeder calculates, distribution losses from the candidate that satisfies the voltage drop constraints Identifying one candidate with the smallest distribution loss configuration as a minimum distribution loss configuration.

  The exact solution described above is a method in which an optimal solution is searched strictly, and the obtained solution is always optimal. Conventionally, due to the size of the system, it was considered difficult to apply the exact solution within a practical calculation time. However, in the present invention, the steps are executed as described above, and thus the distribution system The exact solution can be applied to the determination of

  According to the present invention, in a power distribution network that employs a multi-partition multi-linkage power system such as a three-part / three-system power distribution system, a strict solution such as ROBDD is used to efficiently squeeze candidates that satisfy operational constraints. Therefore, it is possible to provide a distribution system configuration optimization device and a distribution system configuration optimization method that can determine the minimum distribution loss configuration.

  An embodiment of a power distribution system configuration optimizing device of the present invention will be described. As shown in FIG. 3, the distribution system configuration optimizing device 1 includes a distribution system information storage device 11, a node model information storage device 12, a partial feeder creation unit 13, a precision calculation unit 14, and a loss calculation unit 15. The distribution loss minimum configuration specifying means 16 and the operation display means 17 are provided.

  The distribution system information storage device 11 stores information on the distribution system. The node model information storage device 12 stores information obtained by dividing a power distribution system represented by a graphics model into a plurality of closed partial feeders that are less than three connected by a rising portion.

The partial feeder creation unit 13, each part feeder, satisfies the radial configuration constraints, the dividing manner satisfying about 3 split system can be exhaustive search.

The precise calculation means 14 can calculate all candidates satisfying the feeder line capacity limitation at the feeder root by an exact solution . Moreover, the loss calculation means 15 can calculate the power distribution loss in the partial feeder.

  The distribution loss minimum configuration specifying unit 16 can specify one candidate having the minimum distribution loss as a distribution loss minimum configuration among candidates satisfying the voltage drop constraint. Note that the function of the loss calculation means 14 may be executed by the distribution loss minimum configuration specifying means 16. The operation display means 17 is a user interface.

  One embodiment of the power distribution system configuration optimizing method of the present invention will be described with reference to FIG. 4, first, distribution system information is extracted from the distribution system information storage device (distribution system information DB) 11 (S110). As will be described later, this information includes system configuration information, line type information, load information, distribution line information, switch information, and the like. Next, the extracted information is converted into a graphics model (S120). Thereby, nodes and sides to be described later are defined.

  Next, the graphics model is divided into partial feeders (S130). That is, the partial feeder creating means 13 creates a partial feeder from the graphics model. As will be described later, one unit is a node (feeder root branch point) and a line connecting the nodes (feeder root branch point).

Next, a node model is created (S140) and registered in the node model storage unit 12 (S150). And the mode which passes restrictions with respect to this node model is extracted (S160). S160 may include a step of searching all partial feeders for a division mode that satisfies the radial configuration constraint and satisfies the three-partition constraint.

Next, an exact solution is applied by the precision calculation means 14 (S170). For example, ROBDD is applied to this exact solution . Finally, the distribution loss is obtained by the loss calculation means 15 and the minimum distribution loss configuration is specified (S180).

《Graphics modeling of distribution system and creation of partial feeder》
FIG. 5 is a graphics model showing a typical example of a power distribution system. In FIG. 5, feeders F1 and F2 are drawn from the distribution substation G1, and feeders F3 and F4 are drawn from the distribution substation G2. A plurality of switches are provided on a distribution line laid out based on rising points U1, U2, U3, U4 (indicated by black circles) of feeders F1, F2, F3, and F4. These switches include normally open switches (indicated by white squares) and normally closed switches (indicated by black squares), and normally open switches function as interconnected switches. . A section load exists in the distribution line between the switches, and in FIG. 5, the section load is indicated by ampere [A].

  When the selection of the interconnection switch is inappropriate, the supply voltage of the distribution system decreases. FIGS. 6A and 6B show the relationship between the supply voltage and feeder capacity (current capacity) of the distribution system of the three-divided three-link system. In FIG. 6A, the electric power from the distribution substation G is supplied to the rising point U (indicated by a black circle) through the feeder F. The distribution line drawn from the rising point U is closed by the interconnection switches A3, A4, A5 during normal operation, the load section closed by the rising point U, the switches A1, A5, the switches A1, A2, 150 [A] is supplied to each of the load section closed by A3 and the load section closed by the switches A2 and A4. In FIG. 6A, these load sections are indicated by dotted circles. The interconnection switch is selected so that the supply voltage at the rising point U (6900 [V] in this embodiment) does not fall below the voltage allowable limit (6600 [V] in this embodiment). It is necessary (see FIG. 6B).

《Graphics modeling of system and creation of partial feeder》
As illustrated in FIG. 5, the actual distribution line of the distribution system has a radial configuration with the rising points U1, U2, U3, U4 of the feeders F1, F2, F3, F4 as base points. These radial distribution lines are divided into small areas by normally open switches, and these small areas are interconnected by the switches.

In the present invention, the graphics model of the distribution system as shown in FIG. 5 is created as follows. First, information on the distribution system is extracted from the distribution system information DB. This distribution system information includes
・ System configuration information (system configuration and information on equipment connection status)
・ Line type information (type of electric wire)
・ Load information (load amount for each section (kW))
・ Distribution line information (rising point, maximum transmission current (annual maximum value of distribution line current), substation feeder number, distribution line rated capacity (allowable current value of distribution line))
-Includes on / off information for switches.

Next, the information on the power distribution system is converted into a graphics model (node model) using nodes. For this conversion, processes (1) to (4) are performed.
(1) The switch and load section are sides with both ends as nodes.
(2) A branch point (a portion where distribution lines intersect) is defined as a node.
(3) Distribution substations and underground lines leading to the rise point are excluded from conversion.
(4) Assign numbers to edges and nodes.

  FIG. 7 shows the result of conversion to the node model. This result is stored in the node model database. In FIG. 7, nodes at connection points in the track section are indicated by white circles, nodes at rising points are indicated by black circles, and nodes at feeder root branch points are indicated by double circles. In addition, the switch is indicated by a thick line with hatching, the load section is indicated by a solid line, and the feeder root load section is indicated by a thick black line. In FIG. 7, the node is indicated by a 2-digit number and the side is indicated by a 3-digit number.

  Next, as shown in FIG. 8, the distribution line connecting the rise and the rise is divided into units (partial feeders PF1 to PF4) so as to satisfy the triple connection.

And as shown in FIG.
(1) Radial configuration constraint (2) Feeder capacity constraint (3) Three-division constraint (4) Among the voltage drop constraints, the search for the interconnection switch is performed so as to satisfy the conditions (1) and (3). In FIG. 9, the search results of the partial feeders PF1, PF2, PF3, and PF4 are indicated by symbols K1, K2, K3, and K4, and the interconnection switches are indicated by blanks.

In search result code K3, 3 aspect which does not satisfy the divided constraint (dividing manner), that is, divided four or more in divided distribution line (switches are included three or more) but dividing manner occurs The switch in this division mode is excluded from the candidates for the interconnected switch.

Next, as shown in FIG. 10 and FIG. 11, a search is made for a division mode by the interconnection switch that satisfies (2). FIGS. 10 and 11 show a case where (2) is satisfied when the interconnection switch is selected in a divided manner surrounded by an ellipse. FIGS. 12A and 12B show calculation results (first and second constraint satisfaction solutions) of the system based on FIGS. 10 and 11 .

Next, processing by an exact solution is performed. That is, one solution is selected from each partial feeder, and all combinations of solutions satisfying the feeder capacity constraint of (2) and the three-partition constraint of (3) are used using ROBDD (Reduced Ordered Binary Decision Diagram). To calculate. ROBDD is a graph representing a binary decision graph (BDD) representing a logical function in a more compact and efficient manner as shown in FIGS. FIGS. 13A and 13B show an example in which the BDD for obtaining the path from x1 to x3 (logical function f (x1, x2, x3)) is changed to ROBDD.

In the arithmetic constraint equation, the feeder capacity constraint is
S (x ₁₁ , x ₁₂ ,..., X _ij ) ≦ 450
Represented by
Configuration selection constraint (restriction that only one constraint satisfaction configuration can be selected from each partial feeder)
R (x ₁₁ , x ₁₂ ,..., X _ij ) = 1
Represented by
The 3 division restriction (restriction that each feeder has less than 3 closed switches) is
F (x ₁₁ , x ₁₂ ,..., X _ij ) = 2
(However, x _ij : 0-1 variable (all the solutions satisfying 1 if the jth constraint satisfaction solution of the partial feeder i is selected is 0) is extracted.) This is equivalent to finding all the routes to node “1” shown in FIG. In FIG. 14, the logical expression W by SOB, R, F ROBDD is
W (x ₁₁ , x ₁₂ ,..., X _ij ) = 1
W (x ₁ , x ₂ , x ₃ ) = x ₁ + <x ₂ > x ₃
(<X ₂ > is a conjugate of x ₂ )
It is represented by

Next, processing for specifying the minimum distribution loss configuration is performed. That is, (4) Among the division modes that are candidates satisfying the voltage drop constraint, the configuration that minimizes the distribution loss is determined as the optimum system configuration. Here, the distribution loss Loss is
Loss = (loss calculated at each partial feeder) + (loss at each feeder root section)
It is represented by

The distribution loss (including voltage drop) is calculated from the power flow of each feeder, and the configuration that minimizes the distribution loss is determined from the configurations that satisfy the voltage drop constraint of (4). FIG. 15A shows a case where the distribution loss satisfies the voltage drop constraint, and FIG. 15B shows a case where the distribution loss does not satisfy the voltage drop constraint.

A graphics model of a three-division three-way distribution system as shown in FIG.
Number of switches: 140
Number of feeders: 40
When the configuration of the distribution loss is determined by applying the method of the present invention to 2140 candidates = about 1.39 × 1042, the number of candidates is about 1 when the first problem is passed. The number of candidates is 4.44 × 1013, and the number of candidates becomes 32 when the processing by the exact solution is passed, and the number of candidates becomes 2 when the processing for specifying the minimum distribution loss configuration is passed.

  In this case, the minimum loss was 816.632 [kW], the total system demand was 49.731 MW, and the loss was 1.642%. In other words, since the loss in the actual system is 3-4%, the proposed method improves the loss by about 1.5%, and a cost reduction of 400 million yen can be expected in one year. The calculation time was about 30 minutes using a CPU with an internal clock of 1.7 GHz.

It is a figure which shows a power distribution network. It is a figure which shows the example from which a distribution loss changes with the selection modes of a connection switch, (A) is a figure which shows the case where a power loss is large, (B) shows the case where a power loss is small. It is a block diagram which shows one Embodiment of the power distribution system structure optimization apparatus of this invention. It is a flowchart which shows one Embodiment of the power distribution system structure optimization method of this invention. It is a graphics model which shows the typical example of a power distribution system. (A), (B) is a figure which shows the relationship between the supply voltage and feeder capacity | capacitance (current capacity) of the distribution system of a 3 division | segmentation 3 interconnection system. It is a figure which shows the conversion result to a node model. It is a figure which shows a mode that the distribution line which connects a rise and a rise was divided | segmented so that the triple connection might be satisfy | filled per unit. It is a figure which shows the search result when searching for the interconnection switch to satisfy the radial configuration constraint, the three division constraint among the radial configuration constraint, the feeder capacity constraint, the three division constraint, and the voltage drop constraint as operation constraints. is there. It is a figure which shows the result of having searched the division | segmentation aspect by the interconnection switch which satisfy | fills feeder capacity | capacitance restrictions. It is a figure which shows the other result which searched the division | segmentation aspect by the interconnection switch which satisfy | fills feeder capacity restrictions. (A), (B) is a figure which shows the 1st, 2nd constraint satisfaction solution. (A) is a figure which shows the calculation method of BDD, (B) is a figure which shows the calculation method of ROBDD. It is explanatory drawing which shows the calculation method by ROBDD concretely. (A) shows the case where the distribution loss satisfies the voltage drop constraint, and (B) is an explanatory diagram showing the case where the distribution loss does not satisfy the voltage drop constraint. It is a figure which shows the graphics model of a 3 division | segmentation 3 interconnection distribution system.

Explanation of symbols

11 Distribution System Information Storage Device 12 Node Model Information Storage Device 13 Partial Feeder Creation Unit 14 Precision Calculation Unit 15 Loss Calculation Unit 16 Distribution Loss Minimum Configuration Identification Unit 17 Operation Display Unit

Claims (2)

A power distribution system information storage device storing power distribution system information;
A node model information storage device for storing information obtained by dividing the power distribution system represented by a graphics model into a plurality of closed partial feeders that are less than X-linked by a rising portion;
The multiple parts feeders stored in the node model information storage device to all cases search, meets radial configuration constraints, the portion feeder creation means for obtaining a divided manner which satisfies the Y split constraints,
And precision calculating means for calculating the exact algorithm all combinations of the divided manner to meet the line capacity constraint of the feeder roots of the dividing manner determined by that part feeder creation means,
The combination of the divided manner calculated by the precise calculation means, said portion of the power distribution loss in the feeder is calculated and electric distribution losses minimized constituting one candidate 1 distribution loss is minimized from the candidate that satisfies the hypotensive constraints Means to identify as ,
A distribution system configuration optimizing device characterized by comprising: Dividing the distribution system represented by the graphics model into a plurality of closed partial feeders that are less than X-linked by the rising portion;
A step wherein each piece feeder meets the radial configuration constraints and exhaustive search for dividing manner which satisfies the Y split constraints,
Calculating all the combinations of the division modes satisfying the line capacity constraint of the feeder root among the division modes obtained by the exhaustive search by an exact solution ;
The combination of the divided manner calculated by the exact algorithm, the power distribution loss in the portion feeder is calculated, the one candidate 1 distribution loss is minimized from the candidate that satisfies the voltage drop constraints as distribution losses minimum configuration Identifying steps;
A distribution system configuration optimizing method characterized by comprising:

Images