JP2009065788A - Optimal voltage controller for distribution system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an optimal voltage control device for optimally controlling the voltage of a distribution system by performing centralized cooperative control on controllers such as LRT, SVR, SC, ShR, and SVC.
Using a genetic algorithm (GA), voltage control is performed on a distribution system 1 in which distribution lines 12 and 13 including voltage regulators 15, 16, 17 and 18 are connected to a distribution transformer 11. In the optimum voltage control device 2, the binary display obtained by connecting the control signals of each voltage regulator in binary numbers is used as a gene, and the estimated voltage value and the voltage target value at each node of the distribution system when the gene is applied. Search for the best gene that minimizes the objective function according to GA until the end condition is satisfied, using the sum of the differences for the nodes as the objective function, and operate the controller based on the control signal indicated by the obtained best gene To decide.
[Selection] Figure 1

Description

  The present invention relates to an optimum voltage control device that optimizes the voltage of a power distribution system including various controllers by centralized cooperative control.

Usually, a distribution system includes a busbar connected to a distribution transformer installed at a substation and a plurality of high-voltage distribution lines connected to the busbar. Each high-voltage distribution line is connected as a local device with a voltage regulator for compensating for a voltage drop.
In the conventional power system, the power flow is unidirectional from the upper system to the lower system, so the voltage of the power system is controlled on the premise of this.

  However, in recent years, clean natural energy power generation and highly efficient cogeneration systems have attracted attention due to problems such as global warming and fossil fuel depletion, and power sources that can reduce environmental impact such as wind power generation and solar power generation. Has increased. Many of these power sources are installed as distributed power sources near customers because of the advantages of reducing losses in transmission lines, short-term construction, and low investment risk.

By the way, the output of natural energy power generation such as wind power generation and solar power generation is greatly affected by climate weather, so in power distribution systems in which these power sources are distributed and connected to lower systems, reverse power flow that does not exist in conventional systems occurs. Sometimes.
Therefore, it is difficult to suppress voltage fluctuations at each node with the conventional control method based on the premise that when the distributed power supply increases, power flows from the upper system to the lower system. However, since voltage fluctuations adversely affect consumer equipment, it is necessary to keep consumer voltage within a certain range.

  For voltage control of a conventional distribution system, on-load tap switching transformers (LRT) in distribution substations, automatic voltage regulators (SVR) on distribution lines, etc. are based on voltage / current measurement values, which are local information. Many methods are used to calculate and determine the tap position. In particular, in the SVR, a target point called a load center point is defined on the terminal side of the SVR, the voltage value at the load center point is estimated using the voltage / current measurement values at the own end, and the voltage value falls within an appropriate range. So that the position of the transformer tap is controlled.

However, as the number of distributed power sources increases, the following problems arise.
(1) There are problems such as reverse power flow from distributed power sources, unstable output of natural energy power generation, and uneven voltage distribution of each feeder due to uneven load distribution. Even if the value falls within the proper range, it may deviate from the proper range at other points such as the end portion.
(2) LRT and SVR can be coordinated by time limit and sensitivity, but they do not always operate optimally as a whole system.
Therefore, there is a need for a voltage controller that achieves voltage optimization that is consistent across the entire system.

Non-Patent Document 1 discloses an autonomous distributed voltage control method for stabilizing a consumer voltage in a distribution system. According to the disclosed method, parallel capacitors (SC) are distributed in a distribution system, and a voltage control device provided in parallel thereto is used to maintain a constant voltage by using a moving average value of voltage fluctuation, which is self-end information. Controls turning on and off, without the need to exchange control information using an expensive central computer system or a communication medium that covers the entire system. Can be kept in.
However, in the disclosed method, since the on / off control is performed based only on the local information, there remains a problem from the viewpoint of optimally controlling the voltage of the entire system.

  Patent Document 1 discloses a distribution line voltage control method for transmitting a voltage corresponding to a load current of a transmission line in a distribution system including a distributed power source. In the disclosed method, the voltage drop value is estimated from the current of the busbar and the current of the interconnection line with the distributed power source, and the tap of the transformer is switched by switching the transformer tap according to the difference between the voltage drop value and the voltage of the busbar. It is intended to stabilize. According to the disclosed centralized control method, distribution lines other than the interconnection distribution line can output an appropriate voltage corresponding to the load from the distribution transformer without being affected by unstable output of the distributed power supply.

  In addition, when adjusting the voltage of the distribution system using a load-tap switching transformer (LRT) or an automatic voltage regulator (SVR) for compensating the voltage drop provided on the line in the distribution substation, the LRT is a bank unit. Therefore, the voltage of each part of the system is not always the optimum value due to the influence of the load characteristics for each distribution line and the roughness of the tap of the SVR. In addition, the static capacitor (SC) used for reactive power adjustment causes a voltage increase due to the Ferrant effect at a light load and also causes an increase in power loss.

Patent Document 2 discloses a distribution system control system that uses a genetic algorithm (GA) and a neural network (NN) to control the voltage distribution and reactive power distribution of a distribution system so that the power loss is minimized. Yes.
The disclosed method is a method of controlling the voltage of a distribution system having a bus connected to a distribution transformer and a high-voltage distribution line connected to the bus, and is provided with a high voltage based on a control signal from a central device provided in the high-voltage distribution line. The local device that adjusts the voltage and reactive power of the distribution line, and the tide calculation using NN based on the voltage and current of the high-voltage distribution line, and using GA to find the optimal combination of adjustment amounts of the local device A central device.

  An LRT is used to control the voltage regulator, which is a local device, and a reactive power regulator is a device that selectively connects a shunt reactor or power capacitor to the line via a switch, a synchronous phase adjuster, or the like. . In the description of the embodiment, an LRT is used for a substation, and an inverter-controlled regulator (ICR) and a thyristor-controlled reactor (TCR) are used for distribution lines.

In GA, the string expression of the voltage adjustment amount and phase adjustment amount of these local devices is used as a gene, the sum of the voltage deviation and reactive power deviation in the distribution line is the objective function, and the inverse is the evaluation function, and the genetic algorithm is applied. To search for a gene that minimizes the evaluation function.
The NN is a simulation device for accurately estimating a voltage value and a power value at each point when applying an adjustment amount represented by a generated string. The NN always maintains an optimal structure by learning.

However, every controller is installed in an actual power distribution system, and these existing controllers, that is, LRT, SVR, SC, ShR (Shunt reactor or shunt reactor), SVC (Static reactive power compensation) The optimal control of the distribution system voltage for the devices) has not been sufficiently studied, and there was no knowledge about the cooperative control considering all the controllers.
JP 2007-143313 A JP-A-11-289663 Katsuki Kabemura “Mock-up test for application of self-sustained distributed voltage regulator to high-voltage distribution lines” Denki Theory B, Vol. 121, No. 12, 2001

  Accordingly, the problem to be solved by the present invention is to provide an optimum voltage control device that optimally controls the voltage of the distribution system by performing centralized cooperative control on controllers such as LRT, SVR, SC, ShR, and SVC. That is.

  In order to solve the above problems, the optimum voltage control device for a distribution system according to the present invention controls the voltage of a distribution system in which a distribution line having a voltage regulator is connected to a distribution transformer using a genetic algorithm. Then, the binary display that connects the control signal of each voltage regulator in binary number is used as a gene, and the difference between the voltage estimated value and the voltage target value at each node of the distribution system when the gene is applied Using the total value as the objective function, search for the best gene that minimizes the objective function according to the genetic algorithm (GA) until the termination condition is satisfied, and operate the controller based on the control signal indicated by the obtained best gene It is characterized by determining.

The gene is a value obtained by displaying the LRT and SVR tap positions actually arranged in the distribution system in binary numbers, a binary number indicating ON / OFF operation of ShR and SC, and a binary number indicating the control amount of SVC, respectively. It is represented by the corresponding binary number.
The objective function may include a transmission loss.
The GA termination condition may be that the number of generations reaches a predetermined number such as 1000s.
The GA operation preferably includes selection, crossover, and mutation.
In addition, the optimal voltage control apparatus of the power distribution system of this invention can be provided with the communication path which connects between a central control apparatus and a central control apparatus, and each voltage regulator, and can perform centralized control.

According to the optimum voltage control device for a power distribution system of the present invention, by appropriately selecting a gene and using GA, cooperative control considering all voltage regulating devices in the power distribution system is achieved, and in an extremely short time. The control method can be determined. Further, by appropriately selecting the objective function, each node voltage can be controlled so as to approach the target value.
Furthermore, when deciding on the control of each voltage regulator, GA can be used as an approximate solution at high speed even when the solution has multimodality due to a large-scale optimization problem or the feasible region is a discrete set. Can be reached.
In the apparatus of the present invention, even when a distributed power source such as wind power generation or solar power generation exists in the distribution system, it is sufficient to include the output of the distributed power source when estimating the voltage value at each node. The procedure for determining the optimum operation of the voltage regulator does not require major changes and can be easily applied.

Hereinafter, the present invention will be described in detail based on embodiments with reference to the drawings.
FIG. 1 is a block diagram showing a voltage control device for a distribution system according to one embodiment of the present invention, FIG. 2 is a flowchart showing an example of a genetic algorithm procedure used in this embodiment, and FIG. 4 is a configuration diagram and an equivalent circuit diagram of the target SVC, FIG. 4 is an equivalent model diagram and an equivalent four-terminal circuit of the target LRT in the present embodiment, and FIG. 5 is a power distribution targeted in a simulation for confirming the effect of the present embodiment. Fig. 6 is a load pattern in the distribution system targeted by the simulation, Fig. 7 is a graph showing the daily change in the maximum output of photovoltaic power generation, and Fig. 8-10 is the voltage of this embodiment by simulation. Drawing explaining the effect in case 1 of a control apparatus, FIGS. 11-13 are drawings explaining the effect in case 2 of the voltage control apparatus of a present Example by simulation, FIGS. 14-16 are simulations By tio emission is a diagram illustrating the effect of the case 3 of the voltage control apparatus of the present embodiment.

FIG. 1 shows a voltage control device for a distribution system according to the present embodiment.
The distribution system 1 targeted by the voltage control apparatus of the present embodiment is a distribution system composed of high-voltage wiring connected to a distribution substation, and includes one bank or sub-series 12 and 13 of a plurality of banks. In addition, for example, a distributed power source DG14 such as a solar power generator is often connected to a residential area distribution system.
As a voltage regulator for controlling the voltage of the distribution system 1, there are the LRT 11 installed in the distribution substation, the SVR 15 provided in the distribution line, or SC 17 and ShR 18. Further, the distribution line may be provided with the SVC 16.

The voltage control device of the power distribution system according to the present embodiment is a device that performs centralized and coordinated control of the voltage of the power distribution system 1 and is composed of a central control device 2 installed in a central management room or the like, and a control signal is transmitted via the communication network 3. Is transmitted to each adjustment device in the distribution system.
Note that measurement signals at various points of the distribution system 1 required by the central control device 2 are also collected via the communication network 3.

  The central controller 2 calculates the optimum control method and control value of each voltage regulator using a genetic algorithm (GA) and transmits it to the voltage regulator in the distribution system. Perform cooperative control. GA is an excellent fast approximate solution that can be used when a large-scale optimization problem has multimodality, or when the feasible region is a discrete set.

  In the GA of the present embodiment, a binary string formed by continuously displaying binary operation command signals of each voltage regulator is used as a gene. For LRT and SVR, a string indicating the tap position in binary notation, for ShR and SC, a string indicating on / off as 0, 1, and for SVC, a string indicating the control amount in binary notation is generated and joined in series Becomes a gene.

Further, the objective function is a value obtained by adding all the absolute values of the difference between the voltage value manifested at each node and the target voltage value, and the evaluation criterion is that this objective function is minimized. By solving the objective function minimization problem, the amount of operation of each voltage regulator is determined, and the voltage of the entire distribution system is optimized.
Transmission loss can also be taken into account. In order to evaluate appropriately considering the balance between maximizing voltage margin and reducing power loss, it is necessary to select a weighting factor appropriately.
The voltage at each node is estimated by simulation within a central control unit that executes GA.

As an example of the objective function, equation (1) can be given. This objective function is set so that a sufficient voltage margin can be secured so as to be robust against steep voltage fluctuations, and the transmission loss can be evaluated in the same manner.
F = ω ₁ Σ | V _nref −V _n | + ω ₂ Loss (1)
here,
F: objective function _Vnref : voltage target value of _nth node _Vn : voltage of nth node (estimated value)
Loss: transmission loss ω ₁ ω _2: weighting factor

In addition, in the sending transformer in the distribution substation, it is necessary to satisfy the constraints of the equations (2) and (3).
_{V 0min ≦ V 0 ≦ V 0max} (2)
t _min ≦ t ≦ t _max (3)
here,
V _0: feed voltage V _0min: minimum V _0max of delivery Voltage: Maximum value of the feed voltage t: tap ratio of the transformer t _min: minimum ratio of the transformer tap t _max: maximum ratio of the transformer taps

FIG. 2 is a flowchart showing the flow of the basic operation of the GA used in this embodiment.
First, N, for example, 50 strings having random binary numbers are generated, and an initial generation in which a population having a gene represented by those strings enters is formed as the current generation (S1). . Here, the number of individuals in one generation was N.
Next, the fitness for the objective function is calculated for all individuals in the current generation (S2). At this time, the control state of each voltage regulator represented by the gene is applied to a distribution system model in accordance with the actual situation, the voltage at each node is estimated, and the transmission loss is estimated and used.

The solid of the current generation is performed in the order of regeneration (S3), crossover (S4), and mutation (S5), and the result is stored in the next generation (S6).
In the reproduction, the individual is reproduced according to a certain rule according to the fitness of the selected individual (S3). There are various selection methods such as roulette selection, ranking selection, and tournament selection. Individuals that were not selected will be rejected and regeneration will be modeled after natural selection.
Crossover is a model of how an organism leaves offspring, and is an operation of generating a new individual by exchanging part of the gene of an individual selected according to the fitness (S4). Select two individuals, randomly select two crossing points in the gene string, and replace the part between the crossing points. Crossovers are often used.
Mutation is an operation of changing a part of an individual gene to generate a new individual (S5).

Until the number of next-generation individuals reaches N, the operations of regeneration, crossover, and mutation are repeated (S), and when the number reaches N, the contents of the next-generation are all transferred to the current generation (S8).
The procedure from S2 is repeated until the number of generations reaches G, for example, 1000 (S9), and when it reaches G, finally the individual with the highest fitness in the current generation is output as a solution. (S10).
The solution of GA is converted into an electrical signal suitable for the voltage regulator in the central controller 2 and transmitted to the voltage regulator in the distribution system 1 via the communication network 3, and the voltage adjustment of the distribution system is performed accordingly. Is called.

Since the voltage state at each node in the objective function of GA is configured to be equally involved, according to the solution obtained by GA, the voltage at the node in the distribution system is close to the target value on average. . That is, by using the GA, the nodes of the power distribution system are voltage-controlled cooperatively.
GA is an excellent high-speed approximate solution that can obtain an optimal approximate solution at high speed even when the solution has multimodality or the feasible region is a discrete set.

  In the voltage control apparatus of the present embodiment, the fitness of the objective function must be evaluated for N individuals, and therefore it is necessary to accurately and quickly estimate the voltage state at each node when the gene string of the individual is applied. There is. For this reason, the central controller 2 is provided with a model obtained by copying the actual power distribution system, and this model is required to be accurate and simple in terms of voltage calculation.

  FIG. 3 is a diagram showing an SVC analysis model. The SVC is a device that generates reactive power at high speed by controlling the current of a reactor or a capacitor with a thyristor. FIG. 3A shows a basic structure, and FIG. 3B shows an equivalent circuit. For the purpose of the GA operation, it is sufficient to simplify and treat it as the reactive power supply source shown in FIG.

The LRT is a load tap change transformer that can change the tap position by centralized control. 4A shows an equivalent circuit of the LRT, and FIG. 4B shows an equivalent four-terminal circuit.
Now, when the turns ratio of the windings of the transformer is switched from the standard transformation ratio to n times, the primary voltage and current expressed in the unit method between the secondary voltage V ₂ and the current I ₂ are as follows. There is a relationship represented by Equation (4).

一方、等価四端子回路の入出力電圧電流関係式は（５）式となる。

On the other hand, the input / output voltage / current relational expression of the equivalent four-terminal circuit is the expression (5).

式（４）と式（５）の係数行列の各要素は互いに等しいので、
ｚ＝Ｚ_T／ｎ， ｙ₁＝ｎ(ｎ−１)／Ｚ_T，ｙ₂＝(１−ｎ)／Ｚ_T （６）
という関係が成立する。
したがって、ＬＲＴをインピーダンス線図上に表示するときは、各要素の値を（６）式で求めて、等価四端子回路で表すことができる。
なお、ＳＶＲは、集中制御によってタップ位置を変更することができるＬＲＴ変圧器と同様に扱うものとする。

Since the elements of the coefficient matrix of Equation (4) and Equation (5) are equal to each other,
_{z = Z T / n, y} 1 = n (n-1) / Z T, y 2 = (1-n) / Z T (6)
The relationship is established.
Therefore, when the LRT is displayed on the impedance diagram, the value of each element can be obtained by equation (6) and represented by an equivalent four-terminal circuit.
Note that the SVR is handled in the same manner as an LRT transformer that can change the tap position by centralized control.

FIG. 5 is a diagram showing an example of a simulation model that is used to verify the performance of the power distribution system voltage control apparatus according to the present embodiment and is formed assuming an actual power distribution system.
The model assumes a distribution system in a residential area, and a photovoltaic power generation facility, which is a distributed power source DG, is installed at the node N33. The line capacity of this system model was 2,500 kVA.
The SVR was installed between the node N6 and the node N7 suitable for adjusting the voltage of the entire distribution system. In addition, SC and ShR are installed at the positions of nodes N2, N4, N6, N8, N10, and N12 so that they are equally arranged in an area where the voltage drop is large.

FIG. 6 is a drawing showing a load fluctuation pattern in one day of the distribution system. In residential areas, unlike commercial areas and factory areas, there is a feature that peak load comes around 20:00.
Moreover, FIG. 7 is drawing which represents the change of 1 day of photovoltaic power generation output. In this verification simulation, for the sake of simplicity, it is assumed that solar power generation always operates at a power factor of 1.
Note that the weighting factor in the objective function used in this simulation incorporates both the voltage margin and the power loss suppression into the evaluation in a well-balanced manner. Therefore, by trial and error, the voltage margin weighting factor ω1 = 1.0 and the transmission loss weighting factor. ω2 = 10.0.
Moreover, the transformer tap of LRT and SVR can use 11 taps in the range of 0.9 to 1.0 in increments of 0.02.
GA parameters were set to 1000 generations, 50 individuals within a generation, 1.0 reproduction probability, 0.8 crossover probability, and 0.5 mutation probability.

8 to 16 are drawings showing the results of a performance verification simulation according to the voltage control apparatus of this embodiment. The simulation was performed for four cases: no control, transformer tap control LRT + SVR only, transformer tap control + SC + ShR control, transformer tap + SC + ShR + SVC control.
The figure displays the voltage at each node under control for each sub-series.
The simulation was performed by selecting three characteristic times in the load variation pattern.

Case 1 shown in FIGS. 8 to 10 is at 6 o'clock when the customer load is minimum, and the photovoltaic power generation shows 18% output. 8 shows the voltage distribution in the system of nodes N1 to N13, FIG. 9 shows the system of nodes N111 to N114, and FIG. 10 shows the voltage distribution in the system of nodes N31 to N33.
In this case, since the conditions are not so strict as compared with the other cases, the margin was within 5% in any control method. In addition, each node voltage can be set to a target value of approximately 1.0 pu by cooperatively controlling all the voltage regulators. Even when SVC was not used, the same control result as that obtained when all voltage regulators were cooperatively controlled could be obtained.

Case 2 shown in FIGS. 11 to 13 is at 12 o'clock when the solar power generation reaches the maximum output, and the consumer load is an intermediate value. 11 shows the voltage distribution in the system of nodes N1 to N13, FIG. 12 shows the system of nodes N111 to N114, and FIG. 13 shows the voltage distribution in the system of nodes N31 to N33.
In FIG. 11, it was found that the voltage lower limit value was deviated when there was no control, and there was a sufficient voltage margin when the voltage regulator was controlled cooperatively. Thereby, it can be confirmed that the coordinated control of the voltage regulator has an effect. Moreover, in FIG. 13, it can confirm that the voltage of a terminal side is high and the reverse power flow from a distributed power supply exists.

Case 3 shown in FIGS. 14 to 16 is at 20:00 when the customer load is maximum, and the output of the photovoltaic power generation is zero. 14 shows a system of nodes N1 to N13, FIG. 15 shows a system of nodes N111 to N114, and FIG. 16 shows a voltage distribution in a system of nodes N31 to N33.
It can be seen that in the case of no control, the voltage lowers below the lower limit value, and in the case of the feed voltage control + SVR tap control, the voltage upper limit value is exceeded. It is considered that the request for minimum transmission loss is strong and the voltage upper limit value is deviated at the node N1 close to the sending transformer.
In case 3 as well, when all the voltage regulators are cooperatively controlled, it can be confirmed that the voltages at all the nodes are within the allowable range and that there is a large voltage margin.

  From this simulation result, it can be seen that when all the devices are controlled in a coordinated manner, the deviation of each node voltage can be reduced, and a large voltage margin can be obtained. Further, by performing cooperative control, it is possible to reduce the capacity of each control device such as SC, ShR, and SVC.

It is a block diagram which shows the voltage control apparatus of the power distribution system which concerns on one Example of this invention. It is a flowchart which shows the example of a procedure of the genetic algorithm utilized for a present Example. It is the block diagram and equivalent circuit diagram of SVC made into object in a present Example. It is the equivalent model figure and equivalent four terminal circuit of LRT made into object in a present Example. It is a systematic diagram of the power distribution system made into the object by the simulation which confirms the effect of a present Example. It is a load pattern in the power distribution system targeted in the simulation. It is a graph showing the change of the solar power generation maximum output on the 1st. It is drawing explaining the effect about the nodes N1-13 in case 1 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about the nodes N111-114 in case 1 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about the nodes N31-33 in case 1 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about the nodes N1-13 in case 2 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about the nodes N111-114 in case 2 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about node N31-33 in case 2 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about the nodes N1-13 in case 3 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about the nodes N111-114 in case 3 of the voltage control apparatus of a present Example by simulation. It is drawing explaining the effect about node N31-33 in case 3 of the voltage control apparatus of a present Example by simulation.

Explanation of symbols

1 Power Distribution System 2 Central Controller 3 Communication Network 11 LRT
12, 13 Subseries 14 Distributed power supply 15 SVR
16 SVC
17 SC
18 ShR

Claims (4)

  2 is an optimum voltage control device for controlling the voltage of a distribution system in which a distribution line having a voltage regulator is connected to a distribution transformer, and connecting the control signals of the voltage regulators in binary numbers. The objective expression is a value obtained by adding the difference between the voltage estimated value and the voltage target value at each node of the distribution system when the gene is applied as a target, and using the objective function as the objective function, according to the genetic algorithm (GA). Searching for the best gene that minimizes the function until the end condition is satisfied, and determining the operation of the controller based on the control signal indicated by the obtained best gene, optimal voltage control of the distribution system apparatus.   The voltage regulator includes a load tap change transformer (LRT) in a distribution substation, an automatic voltage regulator (SVR) provided on the distribution line, a shunt capacitor (SC), a shunt reactor (ShR), and static reactive power. 2. The optimum voltage of the distribution system according to claim 1, comprising at least one compensator (SVC), wherein the gene is represented by a binary number concatenated with a binary number representing a control signal of the voltage regulator. Control device.   The optimal voltage control device for a distribution system according to claim 1 or 2, wherein the objective function further takes into account power transmission loss.   4. The distribution system optimum voltage control device according to claim 1, wherein the estimated voltage value is calculated using a model of the distribution system stored in the optimum voltage control device.

Images