JP2007264849A - Distribution system extension plan selection method and distribution system extension plan selection program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a distribution system extension plan selection method with which an optimal wiring system extension plan can be selected, and the reduction of facility costs can be expected. <P>SOLUTION: In two layer tab search processing, an arithmetic unit 2 performs initialization, and adds the value of j(new substation) by one, and sets the value of i(new feeder) to an initial value 0. Then, the arithmetic unit 2 prepares a neighborhood solution relating to the capacity and installation of the substation, and adds the value of i by one, and prepares the neighborhood solution relating to feeder layout for each of the solution candidates of the substation. Then, the arithmetic unit 2 calculates evaluation functions f<SB>1</SB>and f<SB>2</SB>for each of the prepared neighborhood solutions, and shifts the solution to the most satisfactory solution. Then, the arithmetic unit 2 repeatedly performs processing while adding the values of j and i, and when the end conditions of the second layer are satisfied, confirms whether or not the end conditions of the first layer are satisfied, and when the end conditions of the first layer are satisfied, ends the processing. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a distribution system expansion plan selection method for selecting a distribution system expansion plan in a power system, and a distribution system expansion plan selection program.

  With the recent expansion of electricity liberalization in Japan, competition in the electric power business has intensified, and the load on the entire distribution system is expected to change. With this change, efficient operation may become difficult in the existing distribution network. For this reason, it is necessary to maintain an efficient system by extending the power distribution system.

  The problem of formulating a distribution system expansion plan for minimizing the cost of the distribution system in response to future changes in the distribution system load is called a distribution system expansion plan problem. The system costs that are a problem in the distribution system expansion plan mainly include a new feeder installation cost, a new substation installation cost, and a distribution loss.

  Conventionally, in an expansion plan of a distribution system in a power system, a combination of an increase in distribution lines and substations has been optimized by assuming a deterministic load (Patent Document 1).

JP 2005-57821 A

  As described above, the distribution system expansion planning problem solves the optimization problem considering the arrangement of feeders (distribution lines) and substations that constitute an efficient distribution system. However, there are a large number of candidates for new installations of feeders and substations, and it would be very expensive to install all of them. In addition, this combination optimization problem needs to be optimized under constraint conditions such as bus voltage upper and lower limit constraints, substation and feeder power capacity constraints, and distribution system radial conditions, which is a very complicated problem. Therefore, it is very difficult for the planner to obtain the optimal expansion plan.

  In recent years, with the deregulation of electric power, independent electric power companies (IPP) and specific-scale electric power companies (PPS) have entered the power distribution network, and the introduction of distributed power sources is being realized. As a result, there will be uncertainty in the future load, so it is necessary to introduce a probabilistic method even in the distribution system expansion plan. That is, it is necessary to select a system plan that is most likely to be optimized among a plurality of solutions by probabilistic evaluation.

  The present invention has been made in view of the above points. A distribution system expansion plan selection method and a distribution system expansion plan selection program that can select an optimal distribution system expansion plan and can be expected to reduce equipment costs. The purpose is to provide.

  A distribution system expansion plan selection method according to the present invention is a distribution system expansion plan selection method for selecting a distribution system expansion plan in a power system using a computer including an input device, a storage device, a calculation device, and an output device, The arithmetic unit uses a tab search to calculate a first evaluation value related to installation locations and capacities of a plurality of expansion candidate substations that are input by the input device and registered in the storage device and that are candidates for addition to the distribution system. The first evaluation value relating to the position of the distribution line to be added in the distribution system based on the first evaluation value for the plurality of additional candidate substations derived in the first layer search process A second layer search process in which the arithmetic device derives by tab search, a first evaluation value derived in the first layer search process, and the second layer search process On the basis of the second evaluation value derived in step (b), the arithmetic unit selects a substation to be added to the distribution system from the plurality of expansion candidate substations, and a selection result by the selection processing And an output process for outputting via the output device.

  A distribution system expansion plan selection program according to the present invention is a distribution system expansion plan selection program for selecting a distribution system expansion plan in a power system using a computer including an input device, a storage device, a computing device, and an output device. A first evaluation value relating to installation locations and capacities of a plurality of candidate addition substations that are input to the computer by the input device and registered in the storage device as candidates for addition to the distribution system A first layer search process in which the arithmetic device derives by tab search, and a distribution system to be added in the distribution system based on a first evaluation value for the plurality of candidate addition substations derived in the first layer search process. A second layer search process in which the arithmetic device derives a second evaluation value related to the position of the electric wire by a tab search; and the first layer search process. Based on the first evaluation value derived in step 2 and the second evaluation value derived in the second layer search process, the arithmetic unit is added to the distribution system from among the plurality of expansion candidate substations. Selection processing for selecting a substation to be executed, and output processing for outputting a selection result by the selection processing via the output device.

  ADVANTAGE OF THE INVENTION According to this invention, the optimal distribution system expansion plan can be selected, and the distribution system expansion plan selection method and distribution system expansion plan selection program which can anticipate reduction of installation cost can be provided.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a diagram showing a computer system for realizing the distribution system expansion plan selection method according to the present embodiment. This system includes an input device (or receiving device) 1, an arithmetic device 2, a storage device 3, and an output device (or transmitting device) 4. This system includes a control program for causing the input device 1, the arithmetic device 2, the storage device 3, and the output device 4 to execute various processes, and executes various processes described later according to the control program.

  The input device 1 is used when inputting various instructions and various information to the computer system. The computing device 2 performs a two-layer tab search, which will be described later, based on information such as the current distribution system, a plurality of additional candidate substations, and distribution lines that can be added, which are input from the input device 1 and registered in the storage device 3. Execute processing to select the optimal distribution system expansion plan. The output device 4 is used when displaying or transmitting the selection result of the distribution system expansion plan by the arithmetic device 2.

  In the present embodiment, an optimal distribution network can be configured by applying a two-layer tab search obtained by improving TS (tab search), which will be described later, to a distribution system expansion plan.

  Here, TS will be briefly described.

  TS is a kind of metaheuristics and is known as a technique capable of searching for a high-precision approximate solution of a global optimal solution for a combinatorial optimization problem. TS is a neighborhood search type algorithm, and it is possible to avoid falling into a local solution by using an adaptive memory called a tab list. Further, since only the tab length indicating the length of the tab list and the maximum number of iterations are parameters, it is easy to find an optimum parameter.

  Thus, TS is a robust technique with little dependency on parameters. TS is a technique that can solve a large-scale problem at a relatively high speed.

  In TS, a search is executed according to the following algorithm (not shown). First, an initial solution is generated (step S1), and a neighborhood of the current solution is created (step S2). Next, the solution is transitioned by selecting the best solution from the non-prohibited neighboring solutions for the tab list (step S3), and the tab list is updated by storing the attributes relating to the transition in the tab list (step S4). When the convergence determination is satisfied, the search is terminated, and when the convergence determination is not satisfied, the process returns to step S2 (step S5).

FIG. 2 is a conceptual diagram of TS search. In TS, it starts a search by generating an initial solution x _1. Next, a neighborhood of the initial solution x ₁ is created, and the solution is transitioned to the best solution x ₂ in the neighborhood. Create a near new from its solution x _2, a transition to the best solution x ₃ in its neighborhood. Thereafter, the same search is repeated until the convergence determination is satisfied. At the time of transition to the best solution, attributes related to the transition are stored in the tab list and used as a prohibited item for subsequent search for a certain period. As a result, solution search loops and convergence to local solutions can be avoided.

  Next, the ε constraint method will be briefly described.

  The ε constraint method is known as a kind of scalar method for obtaining a Pareto optimal solution of a multi-objective plan. For optimization of a general distribution system, a weighting coefficient method for optimizing a weighted single function of a plurality of objective functions has been applied. In contrast, in this embodiment, the ε constraint method is applied to multi-objective optimization. The ε constraint method is a method of converting an objective function into a constraint condition by designating an upper (lower) limit range for an objective function other than the r-th objective function. Therefore, this is a method for obtaining a Pareto optimal solution as a constraint problem in which only the r-th objective function f (x) is an objective function and the remaining objective function is converted into an inequality constraint called an ε constraint. The ε constraint method is formulated for the minimization problem as follows:

  A conceptual diagram of the ε constraint method for multi-objective optimization at n = 2 is shown in FIG. According to the ε constraint method, by changing ε, it is possible to obtain an optimal solution that matches the planner's purpose, and to realize a more practical plan.

  Next, the formulation of the distribution system expansion plan will be described.

  The distribution system expansion planning problem is a problem of selecting a feeder and substation layout that minimizes the cost of the distribution system when adding a substation to the distribution system. There are a large number of new facility arrangement candidates, which corresponds to a combination optimization problem for searching for an optimum combination of power distribution networks. The formulation for the distribution system expansion plan is shown below.

The evaluation functions (f ₁ , f ₂ ) are represented by the following formulas (1) and (2), respectively.

  Moreover, a constraint condition is represented by the following numerical formula (3)-numerical formula (5).

  Expression (1) is an expression indicating the equipment cost for the new configuration network. The first term of the formula (1) represents a new feeder installation cost, and the second term represents a new substation installation cost and a substation capacity cost. The capacity cost is assumed to be proportional to the square of the capacity of the substation.

  Equation (2) is an equation representing distribution loss. Formula (3) is a power flow equation, and in this example, the Very Fast Decoupled method is used. Equation (4) represents the thermal limit constraint of the line power flow, and Equation (5) represents the upper and lower limits of the voltage for each bus.

  In the distribution system expansion plan, a new network that minimizes the equations (1) and (2) is configured under the constraints of the equations (4) and (5). For this reason, the distribution system expansion planning problem can be solved by applying a two-layer tab search using the ε constraint method.

  In this embodiment, a two-layer tab search is applied to solve the distribution system expansion planning problem. In order to solve this distribution system expansion planning problem, the cost of installing new feeders and substations and the effective power loss in the configured network should be minimized as the power load increases in the future. The purpose is to construct a distribution network. However, in this embodiment, without using the original TS as it is, it is more efficient by using a two-layer tab search that divides the search into two layers, a first layer search process and a second layer search process. To do a simple search.

  In the present embodiment, in the first layer search process (first layer), the installation location of the substation and the capacity of the substation are determined from a plurality of additional candidate substations by the TS. In the second layer search process (second layer), the optimum feeder arrangement is determined by the TS.

  FIG. 4 shows a conceptual diagram of the two-layer tab search. In the distribution system expansion planning problem, substation installation, capacity variables and feeder installation variables are both integers, but they have different properties. Therefore, the vicinity of the variable related to the installation and capacity of the substation is created by “Hamming distance = 1”, and the variable related to the installation of the feeder is created by “Hamming distance = 2”.

  This is because it is known that the vicinity of the variable related to the installation of the feeder is ineffective at the Hamming distance 1 because the radial constraint is a strong constraint. As described above, it is very difficult to optimize these variables simultaneously from the viewpoint of coding. Therefore, as shown in FIG. 4, the search is divided into a two-layer structure of one layer and two layers, and variables related to installation and capacity of substations are handled in the first layer, and variables related to feeder installation in the second layer. Handle it and search more efficiently.

  Next, a two-layer tab search algorithm in this embodiment will be described. FIG. 5 is a flowchart showing an example of a two-layer tab search process that is a distribution system expansion plan selection process in the present embodiment. In the two-layer tab search process, the computing device 2 first performs initialization (step S21). Next, the computing device 2 adds 1 to the value of j (new substation) and sets the value of i (new feeder) to the initial value 0 (step S22).

  Next, the computing device 2 creates a neighborhood solution regarding the capacity and installation of the substation (step S23). Further, the arithmetic unit 2 adds 1 to the value of i (step S24), and then creates a neighborhood solution relating to feeder arrangement for each solution candidate at the substation (step S25).

Next, the arithmetic unit 2 calculates the evaluation functions f ₁ and f ₂ for each of the created neighborhoods, and transitions to the best solution (step S26).

  If the second layer end condition is not satisfied, the arithmetic unit 2 returns to step S24 (N in step S27).

  On the other hand, if the end condition for the second layer is satisfied, the arithmetic unit 2 checks whether the end condition for the first layer is satisfied (step S28). Then, the arithmetic unit 2 ends the process if the first layer end condition is satisfied (Y in step S28), and returns to step S22 if the first layer end condition is not satisfied (N in step S28). .

  And the solution obtained most frequently among the obtained several solutions is selected as a selection result of a distribution system expansion plan.

  By executing the two-layer tab search process as described above, variables relating to the substation and the feeder can be considered separately. The first layer search process (first layer) corresponds to steps S22 to S28 of the algorithm, and the second layer search process (second layer) corresponds to steps S24 to S27. Thus, the two-layer tab search has a structure in which a TS processing loop exists inside the processing loop in TS.

  In the first layer search process, the first regarding the installation location and capacity for a plurality of candidate expansion substations that are candidates for addition to the distribution system that is input by the input device 1 and registered in the storage device 3. The evaluation value is derived by tab search. In the second layer search process, the second evaluation value related to the position of the distribution line to be added in the distribution system is calculated based on the first evaluation value for the plurality of extension candidate substations derived in the first layer search process. Is derived by tab search.

  Below, the simulation result in the case of selecting a distribution system expansion plan by the two-layer tab search according to the present embodiment will be described.

Simulation conditions A simulation was performed under the following conditions (1) to (6).

  (1) Here, as an example, a system obtained by extending the IEEE32 bus distribution system shown in FIG. 6 is used. FIG. 6 shows an existing distribution system. The substation expansion candidates (100, 200,..., 600) and the feeder expansion candidates are determined in advance and are registered in the storage device 3 by the input device 1. Similarly, it is assumed that data relating to an existing distribution system is registered in the storage device 3 in advance by the input device 1.

(2) For the new installation cost of a substation, a value proportional to the square of the capacity of the installed substation is set. Similarly, the cost, resistance, and reactance related to the feeder are set to values proportional to the length of the feeder. Further, the upper limit ε ₁ of f ₁ by the ε constraint method is set to 500.

  (3) As shown in FIG. 6, new substation expansion candidates are six (100, 200, 300, 400, 500, 600), and the location and power capacity are determined by the first-layer TS. The power capacity of the installed substation is selected from 1.0, 1.5, ..., 4.0 [MW].

(4) Since the number of new extension candidates for the feeder is 75 as shown in FIG. 6, the same number of bits are used to indicate whether or not the feeder is installed. If it is installed, the bit value is “1”, and if it is not installed, the bit value is “0”. In the TS of the second layer, by exchanging bits with a Hamming distance of 2, the feeder is exchanged and the network is configured. In addition, the total number of feeder selection combinations for the distribution system expansion planning problem here is 3.77 × 10 ²² .

  (5) In this embodiment, in order to consider future changes in system load, a Monte Carlo simulation is performed in which a load change of 80% to 150% is randomly applied to each bus of the existing system. Assume that the load changes.

  (6) In order to show the effectiveness of the present embodiment, comparison with SA (Simulated Annealing) and GA (Genetic Algorithm) is performed. The parameters of each method are shown in FIG. These values are determined by prior simulation. In addition, in order to compare the robustness of each method, the same load change is given to the load of the existing system. In the present embodiment, simulation is performed for 100 different load changes.

Simulation Results The simulation results will be described below by comparing the planning method according to the present embodiment based on the two-layer tab search, the SA planning method, and the GA planning method.

FIG. 8 shows the average value of the evaluation function f _{1 in} the SA, GA and the planning method according to this embodiment for 100 load changes. Since the upper limit by the ε constraint method is set, as shown in FIG. 8, the average value of the evaluation function f ₁ is almost the same in all methods.

Next, the average value of the evaluation function f ₂ in FIG. As shown in FIG. 9, the planning method according to the present embodiment has found solutions that are 1.03 times and 1.04 times more accurate than SA and GA, respectively. This indicates that SA and GA are not effective for the planning problem as compared with the planning method according to the present embodiment. In the case of this problem, which is a radial system configuration problem, since the number of feeders is limited, an efficient search cannot be performed with GA, which is a method of combining solutions probabilistically. In addition, since the SA performs a solution search related to feeder arrangement by exchanging the Hamming distance 2, a search that is more efficient than the GA can be expected. However, it is not always effective because the feeder to be replaced is determined stochastically. On the other hand, since the planning method according to the present embodiment creates and evaluates all neighborhoods, an effective search is possible.

  Next, the result of each method relating to the average calculation time is shown in FIG. As shown in FIG. 10, with respect to the average calculation time, according to the planning method according to the present embodiment, the solution is found 1.5 times and 1.3 times faster than SA and GA. Thus, it can be seen that the planning method according to the present embodiment performs a search that is more efficient than other methods.

  Next, the optimized system is considered. FIG. 11 is an explanatory diagram showing the frequency of solutions that have become the same network with respect to 100 load changes of the planning method according to the present embodiment. The most frequent optimum system is shown in FIG. As shown in FIG. 11, the optimization system has constructed a similar network for 68 different load changes out of 100 various load changes. This optimum system is considered to be the most robust optimum system against future uncertain load changes.

  As described above, in the present embodiment, a plan for introducing a new substation and 13 new feeders as new equipment on the bus 400 is selected as the optimized system for the existing system. By installing a new substation according to this selection, it is possible to share power generation with the existing substation, thereby reducing the power generation cost and to construct two optimal distribution networks with the new feeder. Cost can be reduced.

  In this embodiment, for the distribution system expansion plan, the introduction plan is selected by a two-layer tab search that is a multi-purpose metaheuristic using the ε constraint method, so that the power generation cost adopts a new system that suppresses the distribution cost. It becomes possible. Moreover, the plan according to the planner's purpose can be made possible by using the ε constraint method for the multi-objective optimization plan.

  In this embodiment, the determination of the uncertain load change in the distribution system expansion plan is realized by performing a Monte Carlo simulation. It is possible to find the most robust and optimal network for future load changes by adopting the most frequent network among the discovered networks for various load changes by Monte Carlo simulation. It became.

  In addition, since the plan is selected by the two-layer tab search as described above, the distribution loss is more accurate (1.03 times, 1.04 times higher accuracy) than SA and GA. And the calculation time can be calculated at high speed (1.5 times, 1.3 times).

  In addition, as described above, Monte Carlo simulation takes into account multiple random future system states, and for each system state, multi-purpose using high-speed and high-precision two-layer tab search in meta-heuristics Since the power distribution expansion plan is evaluated, an optimal power distribution system expansion plan can be selected, and the equipment cost can be expected to be reduced.

It is a block diagram of a distribution system expansion plan selection system concerning one embodiment of the present invention. It is a search conceptual diagram of TS. It is a conceptual diagram of the epsilon constraint method with respect to multi-objective optimization in n = 2. It is a conceptual diagram of 2 layer tab search. It is a flowchart which shows the example of a 2 layer tab search process. It is explanatory drawing which shows the IEEE32 bus-line power distribution system. It is explanatory drawing which shows the parameter of each method. Is an explanatory diagram showing an average value of the evaluation function f ₁ for each method. Is an explanatory diagram showing an average value of the evaluation function f ₂ in each method. It is explanatory drawing which shows the average calculation time in each method. It is explanatory drawing which shows the frequency of the solution used as the same network with respect to 100 types of load changes. It is explanatory drawing which shows the optimal system of the selection result of a distribution system expansion plan.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Input device 2 ... Arithmetic device 3 ... Memory | storage device 4 ... Output device

Claims (2)

A distribution system expansion plan selection method for selecting a distribution system expansion plan in a power system using a computer including an input device, a storage device, an arithmetic device, and an output device,
The arithmetic unit uses a tab search to calculate a first evaluation value related to installation locations and capacities of a plurality of expansion candidate substations that are input by the input device and registered in the storage device and that are candidates for addition to the distribution system. A first layer search process to be derived;
Based on the first evaluation value for the plurality of expansion candidate substations derived in the first layer search process, the arithmetic device derives a second evaluation value regarding the position of the distribution line to be added in the distribution system by tab search. Second layer search processing to
Based on the first evaluation value derived in the first layer search process and the second evaluation value derived in the second layer search process, the arithmetic unit is configured to add the plurality of additional candidate substations. Selection process for selecting a substation to be added to the distribution system from
A distribution system expansion plan selection method comprising: an output process for outputting a selection result of the selection process via the output device. A distribution system expansion plan selection program for selecting a distribution system expansion plan in a power system using a computer including an input device, a storage device, an arithmetic device, and an output device,
In the computer,
The arithmetic unit uses a tab search to calculate a first evaluation value related to installation locations and capacities of a plurality of expansion candidate substations that are input by the input device and registered in the storage device and that are candidates for addition to the distribution system. A first layer search process to be derived;
Based on the first evaluation value for the plurality of expansion candidate substations derived in the first layer search process, the arithmetic device derives a second evaluation value regarding the position of the distribution line to be added in the distribution system by tab search. Second layer search processing to
Based on the first evaluation value derived in the first layer search process and the second evaluation value derived in the second layer search process, the arithmetic unit is configured to add the plurality of additional candidate substations. Selection process for selecting a substation to be added to the distribution system from
A distribution system expansion plan selection program for executing an output process for outputting a selection result of the selection process via the output device.