JPH11289664A - Power distribution system control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent a state from getting worse due to a calculation error, by judging the state of a current power distribution system before calculating power flow, and by preventing execution of calculation processing when the state is appropriate. SOLUTION: Before calculating power flow, system information and the detection data of a target system are inputted and stored (S1), and the presence or absence of a place that deviates from the adjustment tolerance and prescribed voltage width of a local device is confirmed according to the detection data (S2). Furthermore, it is judged whether there is no tolerance or a deviation place exists according to the detection data (S3), reactive power in each detection place is increased in the case of NO or it is judged whether the minimum voltage value is not reached or not (S4), and no adjustment is required for preventing calculation processing from being executed in the case of NO. On the other hand, in the case of YES, the degree of appropriateness of a current state for a target voltage and the reactive power is judged (S5), at the same time, it is judged whether the current state is appropriate or not (S6), and no adjustment is required for preventing the calculation processing from being executed in the appropriate case. In the inappropriate case, a target local device is determined (S7) and the calculation processing is started.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distribution system control system for controlling voltage and reactive power of each section of a distribution system.

[0002]

2. Description of the Related Art In general, a distribution system includes a bus connected to a distribution transformer installed in a substation, and a plurality of high-voltage distribution lines connected to the bus. Is connected as a local device.

In a conventional power distribution system, all the voltage information and current information detected at the detection points set in each section of the system are collected by a monitoring device, and information of each section of the system is displayed. Whether or not to adjust the voltage in a specific manner and the value of the adjustment value are determined, and the tap of the distribution transformer is selected based on the result. The voltage regulators connected to each high-voltage distribution line are manually set to a predetermined set value based on the configuration of the distribution system, and the voltage of the high-voltage distribution line detected at the installation location of each voltage regulator is set. The voltage regulator was controlled so as to maintain the set value.

[0004] With the recent increase in demand for interconnection of distributed power sources and the like, power quality management in a distribution system has been performed by a conventional control system that assumes only a fixed one-way power flow and its slow fluctuation. It is expected that they will not be able to respond sufficiently. Further, from the viewpoint of global warming, it is required to minimize the power loss of the system from the viewpoint of energy saving more than ever. That is, the current voltage regulation of the distribution system is mainly performed by a voltage regulation transformer under load (hereinafter, LR) in a distribution substation.
It is called T (Load Ratio Transformer). ), And in the case of long-distance distribution lines, a step voltage regulator (hereinafter, referred to as SV
It is called R (Step Voltage Regulator). ) Is added.

Although it is necessary to perform a power flow calculation for this distribution system control system, the power flow simulator of the related art performs the power flow calculation using the Newton-Raphson method based on an admittance matrix which is a known method of analyzing an electric circuit. Thus, the voltage value and the reactive power value at each detection position are calculated. However, in prior art simulators,
There is a problem that the calculation accuracy is relatively low and the calculation speed is relatively slow. In order to solve this problem, the present inventor has proposed a conventional technology document "Hiroyuki Fudo et al.," Development of distribution system voltage / reactive power control method using GA / NN ", IEICE Technical Meeting, Power Technology and Power System Technology Joint Study Group, PE-97-75, PSE-97-75, 1997
Oct. 7, 2008 "proposes a method of controlling a power distribution system using a genetic algorithm and a neural network.

[0006]

However, in this method, when the voltage and the reactive power at each detection position are calculated using a neural network, there is not a small error, and even if the distribution system is in an appropriate state. However, there is a problem that the state may be degraded.

An object of the present invention is to solve the above problems,
An object of the present invention is to provide a power distribution system control system capable of avoiding a situation in which a state deteriorates due to an error when a voltage and a reactive power at each detection position are calculated using a neural network.

[0008]

According to a first aspect of the present invention, there is provided a distribution system control system having a bus connected to a distribution transformer and a plurality of high-voltage distribution lines connected to the bus. In a distribution system control system that controls the fluctuation range of the voltage of each high-voltage distribution line to be within a set range, a central device that controls the operation of the distribution system and controls the operation of the high-voltage distribution line at least. A local device that adjusts the voltage value and the reactive power value of each of the high-voltage distribution lines by a predetermined adjustment amount based on a control signal from the central device; A voltage sensor and a current sensor that respectively detect a voltage and a current of the high-voltage distribution line at at least one detection position distant from the installation location; The a detection device to be transmitted to the central unit, the central unit comprises an input layer, at least one
A layer having an intermediate layer and an output layer, which has been learned based on predetermined teacher data in advance, and has a voltage adjustment amount and a phase adjustment amount of each of the local devices, a predetermined load active power in a section between the detection devices, and Load power reactive power, input data including the transmission voltage and the transmission phase from the distribution transformer is input, and the power flow calculation is performed by outputting output data including the voltage value and the reactive power value at each of the detection positions. Using a neural network to be performed, based on a predetermined initial adjustment amount, predetermined load active power and load reactive power in a section between the respective detection devices, and a transmission voltage and a transmission phase from the distribution transformer. Calculating means for calculating a voltage value and a reactive power value at each of the detection positions; and, prior to the processing by the calculating means, the voltage value and the current value detected by each of the detecting devices. Based on the output data, it is determined whether or not the current power distribution system is in an appropriate state using predetermined appropriate conditions, and when it is in an appropriate state, control is performed so as not to execute the processing by the calculation means. Means.

The power distribution system control system according to claim 2 is the power distribution system control system according to claim 1,
The appropriate condition is a voltage value or a reactive power value at each of the current detection positions detected by each of the detection devices,
It is characterized in that the sum of the differences from the predetermined target value is equal to or less than a predetermined threshold value.

[0010]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a configuration of a distribution system control system according to an embodiment of the present invention. In FIG. 1, SS is a distribution transformer installed in a distribution substation, B is a bus connected to the secondary side of the transformer SS, and F1, F2,... Are connected to the bus B. High voltage distribution line. Although not shown, the high voltage distribution lines F1, F2, ...
Are connected to low-voltage distribution lines (not shown) via respective transformers.

Reference numeral 1 denotes a central device for controlling a distribution system. This central device 1 is provided, for example, in a distribution substation. In the present embodiment, at least one local device 2 is provided for each high-voltage distribution line in order to perform voltage adjustment and reactive power adjustment of each of the high-voltage distribution lines F1, F2,. Each of the high-voltage distribution lines F1, F2,... Is provided with a detection device 3 that detects three-phase voltages and currents of the high-voltage distribution line at detection points set at at least one location away from the local device 2. I have. A data communication line 4 is provided for performing communication between the central device 1 and each of the local devices 2 and the detection device 3. On the side of the central unit 1, the local unit 2 and the detecting unit 3
A data transmission / reception device 5 for receiving data transmitted to the local device 1 and transmitting data from the central device 1 to each local device 2 is provided.

Each of the local devices 2 includes a local device 2A having a voltage regulator and a reactive power regulator, and a voltage of a corresponding one of the high-voltage distribution lines F1 and F2 at a detection point set near a location where the local device 2A is installed. Voltage sensor 2B and internal current sensor 2C for detecting current and current respectively, a device controller 2D for controlling a voltage regulator and a reactive power regulator constituting the local device 2A, and control of the voltage regulator and the reactive power regulator A communication control and control target setting device 2 </ b> E that performs target value setting processing and communication with the central device 1 is provided. The internal voltage sensor 2B includes, for example, an instrument transformer, and the internal current sensor 2C includes, for example, a current transformer.

Here, a known load voltage regulator (LRT) can be used as the voltage regulator of the local device 2A. In addition, as the reactive power regulator, a device in which a shunt reactor or a power capacitor is selectively connected to a line via a switch, a synchronous phase adjuster, or the like can be used. Communication control and control target setting device 2E
A transmitter that transmits voltage information and current information detected by the internal voltage sensor 2B and the internal current sensor 2C, respectively, and status information indicating the status of the local device 2 to the central device 1 via the data communication line 4; A communication device including a receiver for receiving a signal transmitted from the central device 1, and a voltage control target value and a reactive power control target value, which are control target values of a voltage and a reactive power of a corresponding high-voltage distribution line, respectively. And a control target setting device for setting the target. The device control device 2D also controls the voltage regulator and the reactive power so as to maintain the voltage and the reactive power of each part of the corresponding high-voltage distribution line at the voltage control target value and the reactive power control target value set by the control target setting device, respectively. Control the vessel.

The detecting device 3 includes a voltage sensor 3 for detecting the voltage and current of the high-voltage distribution line at the installation location.
A and current sensors 3B and these sensors 3A and 3B
And a communication slave station 3C for transmitting the detected voltage information and current information to the data transmitting / receiving device 5 of the central device 1 via the data communication line 4. The voltage sensor 3A is composed of, for example, an instrument transformer, and the current sensor 3B is composed of a current transformer.

The central unit 1 executes the following processing. (A) Ascertain system information (address and connection of a detection device, switching on / off of a VS, type of a local device) of a power distribution system. (B) Based on the detection data from the local device 2 and the detection device 3, a load current or the like at each detection position is predicted, and the voltage at each detection position is calculated from the impedance map and the detection information. (C) collecting detection data from the local devices 2 and 3 by polling the devices 2 and 3;
Check for voltage deviation, increase in reactive power, etc. (D) Using the genetic algorithm and the neural network, which will be described in detail later, determine the local device 2 to be in the optimum state and calculate the adjustment amount. (E) Create an impedance map based on the detection data and system information. (F) The power flow is calculated using the impedance map and the detection information, and the respective target values of the voltage and the reactive power at the position of each local device 2 are calculated. (G) The power flow is calculated using the impedance map, and after confirming the calculated adjustment amount, a macro command is issued to the local device 2.

The communication control and control target setting device 2E of the local device 2 executes the following processing. (1) A line voltage and a phase current are input from the own voltage sensor 3A and current sensor 3B, and the input detection information is averaged to calculate a temporal average value for a predetermined time. (2) In response to the polling from the central unit 1, the latest information (voltage, current, reactive power) averaged at the present time and the current adjustment amount are returned as a response. (3) The current adjustment amount is constantly monitored, it is determined whether or not there is a margin for the adjustment width, and a reply is sent to the central device 1 as a polling response. (4) The amount of adjustment is calculated with respect to the macro command (target value) from the central device 1, and it is determined whether or not the adjustment is possible. (5) The set target value of the local device 2A is determined and passed to the device control device 2D together with the detection information.

Further, the communication slave station 3C of the detection device 3
Perform the following processing. (1) A line voltage and a phase current are input from the own voltage sensor 3A and current sensor 3B, and the input detection information is averaged to calculate a temporal average value for a predetermined time. (2) In response to the polling from the central unit 1, the latest information (voltage, current, reactive power) averaged at the present time is returned as a response.

Next, a specific example of the distribution system control system will be described. In this system, the state of each part of the system is grasped by an automatic power distribution system, and the line adjustment devices distributed and arranged in each part of the system are controlled based on the state. FIG. 2 shows the concept of the system including the central unit to be considered. The system includes a central device 1 for overall control, an LRT target correction unit 90 for a substation, and an inverter-controlled regulator (hereinafter referred to as an ICR (Inverter Controlled Regulato) as a line adjustment device installed in a distribution line.
r). ) 91, a thyristor-controlled reactor (hereinafter referred to as TCR (Thyristor Controlled Reactor)) 92, and a sensor 93 that is installed at a key point in a distribution line and sequentially takes in voltage, current, and power factor information.
These are linked by a distribution automation transmission line 94. In this configuration, the LRT target correction unit 90 is connected to the voltage adjustment relay 9.
5 has a function of collectively controlling the voltage of the banks for appropriately increasing or decreasing the output voltage of the LRT 96 controlled by the LRT 96. Also, I
The CR91 has a voltage adjustment function at the installation point and a reactive power compensation function on the power supply side line, and the TCR92 mainly has
It has a delayed reactive power injection function to improve overcompensation by C. Note that these devices 91, 92, and 96 receive control target values from the central device 1 (the LRT target correction unit 90 is a voltage correction target value, the ICR 91 is a voltage and reactive power control target value,
The TCR 92 receives the reactive power control target value) and performs feedback control by itself to follow this target value. Therefore, it is important for the central unit 1 to generate a target value for realizing optimal voltage and reactive power control as a power distribution system and to instruct individual local devices 2 such as the ICR 91 and the TCR 92. Becomes

Next, a voltage and reactive power control method will be described. Optimization of voltage and reactive power control is a combinatorial optimization problem. In the present embodiment, an optimal optimal solution for voltage and reactive power control in the power distribution system is obtained at high speed using a genetic algorithm (hereinafter, referred to as GA) and a neural network (hereinafter, referred to as NN).

FIGS. 3 and 4 are flowcharts showing the central processing performed by the central equipment 1 of FIG. In FIG. 3, first, in step S <b> 1, input and storage of system information and detection data of a target system from the local device 2 and the detection device 3 are performed. Next, based on the detection data input in step S2, the local device 2
Check if there is any deviation from the adjustment margin and the specified voltage range. Further, in step S3, it is determined whether there is no margin or there is a departure point in the data. If YES, the process proceeds to step S7, whereas if NO, the process proceeds to step S4. In step S4, it is determined whether the reactive power at each detection position has increased or the voltage is not at the lower limit value. If NO, there is no need to adjust, so the process returns to step S1, while
If YES, the process proceeds to step S5. In step S5, the appropriateness of the current state with respect to the target voltage and the reactive power is determined. In step S6, it is determined whether or not the current state is inappropriate. On the other hand, if it is inappropriate (YES), the target local device 2 is specified in step S7. That is, the local device 2 to be controlled is determined based on the system configuration (the number of strings in the GA is determined), and the process proceeds to step S11 in FIG.

In the central processing of FIG.
Steps S11 to S15 are control processes for determining an adjustment amount using GA and NN.
In step 2, NN is used. In step S11 of FIG. 4, a process of generating an initial string (adjustment amount) group is performed to generate a data group for power flow calculation, that is, each adjustment amount of the target local device 2 is randomly generated within the adjustment width, and adjustment is performed. Generate a specified number of quantity groups. Next, in step S12, a fitness calculation process is executed, and the voltage V and the reactive power Q at each detection position when the local device 2 of the target system operates with the given adjustment amount are calculated by the power flow calculation (that is, the regulation). Then, a function value of a comprehensive objective function (or evaluation function) f (x) described later is calculated based on the calculated voltage V and reactive power Q. Here, the relationship between the adjustment amount and the power flow distribution is learned by the NN in advance, and the adjustment amount is obtained at a high speed by an inference method using the NN. Then, in step S13, it is determined whether or not the end condition is satisfied. When satisfied, the process proceeds to step S16, while when not satisfied (NO), a selection process is performed in step S14 to obtain the end condition. Then, the reciprocal of each objective function f (x) is calculated, and is selected in descending order of the reciprocal value indicating the fitness, and a plurality of predetermined adjustment amounts at that time are selected as strings having high fitness.
Next, in step S15, crossover and mutation processes are performed, and a crossover point is randomly determined from the selected adjustment amounts, and a predetermined number of adjustment amount groups (corrected strings) are newly generated. Then, bit inversion is performed in a string among a plurality of predetermined low-fitness strings in the generated group, mutation processing is performed, and the process returns to step S12 to execute the fitness calculation processing. .

If the predetermined termination condition is satisfied in step S13, the process proceeds to step S16 where the target local device 2
Are determined, and in step S17, the target voltage and the target reactive power calculated by the conventional power flow calculation based on the adjustment amounts for each device obtained by GA and NN are transmitted as target values of the local device. After the final confirmation using a conventional power flow calculation method (method for obtaining an exact solution), the validity of the adjustment amount is confirmed by a macro simulation of the distribution system, and then, in step S18, the determination is made in step S16. The macro command including the adjustment amount is transmitted via the data communication line 4 to the communication control and control target setting device 2E of the local device 2.
After that, the process returns to step S1 in FIG.

FIG. 7 is a flowchart showing the central device interrupt processing executed by the central device 1 of FIG. In FIG. 7, first, it is determined in step S41 whether there is a system change in the power distribution system, and in step S42, it is determined whether there is new detection data from the local device 2 and the detection device 3 or not. . When YES is determined in the step S41, the processing from the step S51 to the step S58 is executed, while the processing in the step S42 is performed.
If YES in step S61 to step S61
The processing up to 66 is executed.

In step S51, system information is input, in step S52, switch on / off information is input, and in step S53, detection data is input. Next, in step S54, an impedance map is created based on the above information. Then, step S55
In step S, a tidal current is calculated based on the impedance map and the detected data, and learning data is created.
After learning and updating the NN based on the learning data in 56, the process returns to step S41.

In step S61, system information is input, in step S62, switch on / off information is input, and in step S63, detection data is input. Next, in step S64, an impedance map is created based on the above information, and in step S65, power flow calculation is performed based on the impedance map and the detected data to create learning data. Further, based on the learning data created in step S66, NN
After the learning of 500 is performed and the NN 500 is updated, the process returns to step S41. The processing from steps S61 to S66 may be configured to be executed, for example, every time new detection data is input periodically.

Next, the process of creating an impedance map in steps S54 and S64 will be described.
In this process, the impedance between the detection positions is calculated based on the input system information and the detection data (voltage, current, reactive power) from each detection position, and is used for power flow calculation. Here, the creation of the impedance map is calculated when the system change information is input and when the detection data is input. In addition, the impedance map recognizes the positional relationship between the detection devices from the system information, and calculates the impedance between the detection devices based on the input detection data.

Further, a description will be given of a process of creating learning data of the neural network by calculating the power flow in steps S57 and S65. In this process, an impedance map is created at the time of inputting detection data, and a power flow calculation is performed using the impedance map and the detection data to calculate a voltage and a reactive power at each detection position.

Next, additional learning of the network is performed based on the adjustment amount (input) of each local device 2 input as the detection data, the voltage calculated by the power flow calculation, and the reactive power (teacher data at the output).

Next, formulation of the optimization problem of voltage and reactive power will be described. The problem of optimizing the voltage and the reactive power in the distribution substation on a bank-by-bank basis can be considered as a problem of optimizing a combination of adjustment amounts, and can be formulated as follows. That is, for the purpose of optimizing the system voltage and minimizing the reactive power, a comprehensive objective function f defined by the following equation consisting of the objective function of the voltage and the objective function of the reactive power
(X) is used.

[0031]

(Equation 1)

Here, m is the number of distribution lines per bank, a
And b are weighting factors, V (x) is an objective function of voltage, Q
(X) is the objective function of the reactive power. Of these, the weighting factors a and b are characteristics that should be determined for each distribution line in consideration of the situation of the distribution line. In the present embodiment, the weighting factors a and b are determined based on the state of the distribution line based on the detection information. The values of the coefficients a and b are automatically selected as shown in FIG.

The selection of the weight coefficient in FIG. 8 is performed as follows. (I) If there is any voltage deviation at any one of the detection points, the weight coefficient at the time of voltage deviation is used. Even when all the detection points are in the “normal state”, if the states are in the states ST1 and ST3, it is considered that the voltage and the reactive power should be adjusted.
And b are both 0.5. Also, the voltage objective function V
(X) was defined as follows.

[0034]

(Equation 2)

Here, Vi is each detected voltage before the operation of the local device 2, and Vc is a predicted voltage value at the detection point when the local device 2 is operated calculated by NN of the central device 1.
Vo is a target value (fixed value) to be determined for each distribution line, but in the present embodiment, each detection voltage is 6300 to 6900.
If it is out of the range of V, the voltage is 6600V, and if it is not out of the range, it is 6400V. Further, the objective function Q (x) of the reactive power is defined as the following equation.

[0036]

(Equation 3)

Here, i is the number of detections per distribution line, Qi is each reactive power detected before the operation of the local device 2, and Qc is the case where the local device 2 calculated by the NN of the central unit 1 operates. Is the predicted value of the reactive power at the detection point.

Next, the constraints will be described. First,
Under the constraint conditions of the voltage upper and lower limits, the transmission voltage of the distribution line and the voltage at the detection point must satisfy the following equations. Here, the sending voltage must be within the LRT tap adjustment range, and the voltage at the detection point must be a voltage that sets the terminal voltage of the customer to 101 ± 6V.

[0039]

Vfmin ≦ Vf ≦ Vfmax

Vsmin ≦ Vs ≦ Vsmax

Here, Vf is the voltage of the high voltage distribution line (feeder), Vfmax is the maximum value of the voltage, and Vfmin is the minimum value of the voltage. Vs is the detection voltage, Vsm
ax is the maximum value of the detection voltage, and Vsmin is the minimum value of the detection voltage.

Next, under the constraint condition of the adjustment amount of the ICR 91, the voltage adjustment amount and the reactive power adjustment amount must not exceed the range of the adjustment capability determined by the specification of the ICR 91 as shown in the following equations.

[0042]

[ _Expression 6] V _ICR min ≦ V _ICR ≦ V _ICR max

[ _Equation 7] Q _ICR min ≦ Q _ICR ≦ Q _ICR max

Here, V _ICR is the voltage adjustment value of _ICR 91, V _ICR max is its upper limit, and V _ICR min is its lower limit. Q _ICR is the reactive power (phase) adjustment value of the _ICR 91, Q _ICR max is its upper limit, and Q _ICR min is its lower limit.

Further, under the constraint of the amount of adjustment of the TCR 92, the amount of reactive power adjustment must not exceed the range of the adjustment capability determined by the specifications of the TCR 92 as in the following equation.

[0045]

[ _Equation 8] Q _TCR ≦ Q _TCR max

Here, Q _TCR is the reactive power adjustment value of the _TCR 92, and Q _TCR max is its upper limit.

Next, the formulation by the combination of GA and NN500 will be described. Genetic algorithm (G
In A), a method is required in which the string length is short and it is difficult to generate a dead string by string manipulation. On the other hand, the ICR 91 performs voltage adjustment and phase adjustment, and
Since the reactive power adjustment of No. 2 actually adjusts the power factor angle of the installation point of the TCR 91, the phase adjustment is performed. Therefore, in the present embodiment, a string expression satisfying the above condition is shown below. A string expression of the voltage adjustment amount and a string expression of the phase adjustment amount are used.

In the string representation of the voltage adjustment amount, I
Since the controllable voltage adjustment amount of CR91 is ± 330 V, the minimum controllable voltage width is 5 V / step, and
If V is expressed as 1 bit, 330 V is 1 in binary.
000010 and represented by 7 bits. Here, assuming that the sign of ± is 1 bit, the change of the adjustment amount can be expressed in a string as shown in FIG. In this case, since the voltage adjustment range is 330 V at the maximum, a string exceeding this range is a lethal gene.

In the string expression of the phase adjustment amount, the controllable phase adjustment amounts of the ICR 91 and the TCR 92 are ±
Assuming that the angle is 5 degrees and the minimum controllable phase width is 0.1 degree / step, when 0.1 degrees is expressed as 1 bit, 5 degrees is 110010 in a binary number and is represented by 6 bits. Here, assuming that the sign of ± is 1 bit, the change in the adjustment amount can be expressed in a string as shown in FIG. In this case, since the phase adjustment width is 5 degrees at the maximum, a string exceeding this is regarded as a lethal gene.

Further, the string length will be described.
Strings are generated for each distribution line. in this case,
The string length varies depending on the number of local devices 2 of the distribution line. For this reason, a method is adopted in which the maximum number of local devices 2 is set, and the string length is changed according to the type and number of target devices. FIG. 11 shows an example of the string length for three ICRs and three TCRs.

In the method of generating the initial string, a binary number 0 or 1 is randomly allocated to each bit of the adjustment amount of each local device 2. When a lethal gene that exceeds the adjustment width is generated, the allocation is performed again to generate N strings. Although the currently used adjustment amount is not an optimal solution, it is considered to be a sub-optimal solution, so that it is included in one of the initial string groups.

Next, the configuration of the neural network (NN) 500 to be applied will be described. The string (adjustment amount) information and the information not applicable in the GA (active power and reactive power at each detection point, transmission voltage, transmission phase) are represented by N
Sent to the processing unit of the N500 and have a network in which the environment of the target distribution line has been learned (remembered) in advance. Method), and outputs the calculation result (voltage and reactive power at each detection point) to the GA processing unit. A back propagation model having a learning function is used for NN500, and the number of hidden layers is one.
One network is configured for each distribution line as a layer, and the number of intermediate units is determined based on verification results described later. FIG. 15 shows a configuration example of the NN 500 in an eight-section distribution line in which three ICRs and three TCRs are installed. In the present embodiment, the intermediate layer 200 is one layer, but the present invention is not limited to this, and a plurality of layers may be provided.

In the NN 500 shown in FIG.
Is configured to include an input layer 100, an intermediate layer 200, and an output layer 300. The input layer 100 includes 27 input layer units 101-1 to 101-27, and includes an intermediate layer 200.
Represents a plurality of M intermediate layer units 201-1 to 201-
M, and the output layer 300 includes 28 output layer units 3
01-1 to 301-28. Here, the following data is input to each of the input layer units 101-1 to 101-27. (A) ICR _{V1 to} ICR _V3 : voltage adjustment amounts of three ICRs 91; (b) ICR _{Q1 to} ICR _Q3 : reactive power (phase) adjustment amounts of three ICRs 91; (c) TCR _{Q1 to} TCR _Q3 : 3 (D) P _{1 to} P ₈ : Active load power (fixed value) in the section between the detectors 3, (e) Q _{1 to} Q ₈ : Each detector 3 (F) Vs: transmission voltage from the distribution substation (detected by the central unit), and (g) θs: transmission phase from the distribution substation (fixed value). Detected by the central unit.).

Next, the input layer units 101-1 through 101-1
01-17 distributes the input data to M and outputs the data to each of the intermediate layer units 201-1 to 201-M.
Each of the intermediate layer units 201-1 to 201-M distributes 28 pieces of data weighted by a linear combination with a predetermined weighting coefficient to the 27 pieces of input data, and outputs the output layer units 301-1 to 301. Output to -28. Further, the output layer units 301-1 to 301-2
Numeral 8 adds the input M data and outputs the following data. (A) V _{C1 to} V _C14 : voltage values at each detection position, and (b) Q _{C1 to} Q _C14 : reactive power values at each detection position.

Here, the output y of each of the intermediate layer units 201-1 to 201-M can be represented by the following equation, where a (x) is a response function and w _i is a link load.

[0056]

Y = a (x) = w ₁ · ICR _V1 + w ₂ · ICR _V2 + w ₃ · ICR _V3 + w ₄
・ ICR _Q1 + ... + w ₂₅・ Q ₈ + w ₂₆・ Vs + w ₂₇・ θs

Next, the fitness function will be described. NN
Based on the output data from 500 (the voltage value and the reactive power value at each detection position), the objective function f
(X) is obtained, and the evaluation by GA is performed by a fitness function g by the following equation.
This is performed using (x).

[0058]

G (x) = 1 / f (x)

Further, the GA has a characteristic that the rise of the evaluation value is saturated as generations are repeated, so that the process is terminated when any of the following conditions is satisfied. (Condition 1) When the set calculation time has been reached. (Condition 2) When the maximum evaluation value is not updated over the specified number of generations.

In the GA selection and selection process, a specified number of highly adaptable strings are left in the next generation in the previous generation group, and some of them are flagged as mutations for mutation prohibition. In this case, a string having a high degree of fitness remains unconditionally in the next generation, but an elite string may spread rapidly in a group and fall into a local solution. I have. That is, an elite preservation strategy is adopted, and a method is adopted in which individuals with the highest fitness in the group are left as they are in the next generation. While individuals with high fitness remain unconditionally in the next generation, there is a risk that the elite individuals will spread rapidly in the population and fall into local solutions. Although the actual elite conservation strategy prohibits mutations in elite individuals. In the present embodiment, mutation is possible for individuals other than the arbitrarily set individuals.

In the GA crossover process, two strings are selected at random from the selected strings, the crossover points are randomly determined for each voltage and phase adjustment amount, and each crossover is performed by one point. The specified number of strings is generated by excluding the lethal gene.
An example of crossover is shown in FIG. FIG. 14 is a diagram showing an example of multipoint crossover of strings in GA.
As shown in (1), one point crossover is performed for each adjustment amount of each local device 2.

Further, in the GA mutation processing, the mutation is a conversion of a numerical value at each string position. For this reason, a method is adopted in which an arbitrary local device 2 causing a mutation for each voltage and phase adjustment amount is randomly selected, and an arbitrary bit of the adjustment amount is inverted. FIG. 13 shows an example of mutation.

Next, a method of applying the genetic algorithm (GA) will be described. For the following input data, an adjustment amount of the local device 2 is calculated by the GA so that the objective function f (x) is minimized (the NN 500 uses the GA
Used in the fitness calculation process inside the process. ) Output the optimal adjustment amount. (A) GA input data (a) Target local device 2: Target local device 2 to be optimized. However, in the case of the ICR 91, information on whether only the voltage adjustment amount is optimized or whether the phase adjustment is also performed is included. (B) Weight coefficient of objective function: weight coefficients a and b of objective function f (x). (C) Target voltage: a target voltage for obtaining the objective function f (x). Here, the following input information must be data at the same time. (A) Initial value of the adjustment amount of each local device 2: the current adjustment amount of each local device 2. (B) Impedance map: a current impedance in a section between the detection devices 3. (C) V, I, Q of the first detection device 3: Current measured values of current, voltage, and reactive power in the first detection device 3. (2) Output data (a) Adjustment amount of local device 2 that has been optimized: The optimum adjustment amount for each local device 2 in input data.

In the GA of this embodiment, prior information is used as follows. If the voltage does not deviate from the specified voltage, the adjustment amount of each currently operated local device 2 is not an optimal solution, but it can be said that it is a somewhat good solution.
The optimal solution should also be near that adjustment. By utilizing this, a method is used in which the search space is narrowed down and the calculation time is shortened by reflecting the adjustment amount of each local device 2 currently operated in the initially generated individual. In the case where the voltage deviates from the specified voltage, the following method is not used.

Here, the initial generation method (reflection of the initial state)
In, an individual that reflects the initial value of the adjustment amount (the adjustment amount currently being operated) as it is is put into one of the initially generated individuals. Although the currently used adjustment amount is not the optimal solution, it can be said that it is a good solution to some extent (when the voltage does not deviate from the specified voltage range), so that an efficient search can be performed by reflecting the adjustment amount in the initial generation. It has the advantage that.

FIG. 5 is a flowchart showing the fitness calculation process (step S12), which is a subroutine of FIG. As shown in FIG. 5, first, 1 is substituted for a parameter n in step S21, and n is set in step S22.
The adjustment amount of each local device 2 is obtained from the gene content of the th individual. Next, in step S23, the voltage value V and the reactive power value Q at each detection position in the adjustment amount are obtained using the neural network 500. And
In step S24, an objective function f (x) is calculated, and a fitness function is calculated. Further, step S2
In step 5, it is determined whether or not the parameter n has reached the number of individuals. If it has not reached (NO), the parameter n is incremented by 1 in step S26, and the process returns to step S22 to repeat the above processing. On the other hand, if n has reached the number of individuals in step S25 (YES), the process returns to the original main routine.

FIG. 6 is a flowchart showing the selection and selection process (step S14), which is a subroutine of FIG.
As shown in FIG. 6, in the selection process, in step S31, a specified individual having a high fitness is selected from the population of the previous generation. Next, in step S32, a mutation prohibition flag is set for several individuals having high fitness as elite individuals. Then, the process returns to the original main routine.

In determining the end of GA, the reciprocal g (x) = 1 / f (x) of the objective function f (x) is used as the objective function indicating the fitness. End the calculation. If the following condition is satisfied before the end time, the process ends at that time. When calculating with GA, as the generations increase, the degree of increase in the evaluation value of the objective function decreases so as to be inversely proportional to the exponential function, and eventually the evaluation value does not change regardless of how many generations are calculated . By utilizing this, if "the maximum evaluation value is not updated by the specified generation", the calculation is terminated assuming that the optimum solution has been reached. Even if the optimum solution exceeds the specified voltage range, it is not regarded as a lethal gene. As the calculation result, the adjustment amount of the target local device 2 is output.

Next, an initial learning method of the NN 500 will be described. First, an impedance map is created. FIG. 16 is a block diagram showing a process of generating an impedance map in a certain load state used in the distribution system control system of FIG. As shown in FIG. 16, in the generation processing, an impedance map in the current load state is created from the detection data and the adjustment amount of each local device 2. Here, detection is performed several times a day, and an impedance map of several patterns is created. Next, learning data in each impedance map is created. FIG.
FIG. 2 is a block diagram showing a calculation process of a voltage value V and a reactive power value Q at a sensor for each adjustment amount used in the distribution system control system of FIG. 1. As shown in FIG. 17, for each impedance map, the voltage value V and the reactive power value Q of each detection device 3 when each local device 2 is changed in a uniform pattern are configured by NN500. Calculated by the conventional tidal current simulator. The voltage value V and the reactive power value Q are used as learning data. As described above, a pattern is set so that the adjustment amount is uniform, and the output voltage value V and reactive power value Q are used as learning data.

The tidal current simulator of the prior art is NN50
Based on input data similar to 0, a power flow calculation is performed using the Newton-Raphson method based on an admittance matrix, which is a known method of analyzing an electric circuit, to calculate output data similar to NN500. This tidal current simulator is used not only for learning of the NN 500 but also for the process of step S17 in FIG.

FIG. 18 is a block diagram showing a learning method of the neural network (NN) 500 of FIG.
Learning is performed based on the learning data created above. In addition,
The learning of the NN 500 is performed separately for each high-voltage distribution line. In addition, in the additional learning of the NN 500, it is necessary to reflect the latest learning data on the NN 500 because the NN 500 learned in the initial learning alone cannot cope with a change in load or the like. Therefore, additional learning is performed (see steps S58 and S66 in FIG. 7).

FIG. 19 is a block diagram showing a method of incorporating learning data into the neural network (NN) 500 of FIG. As shown in FIG.
Using the shift register 401 connected to each of the input layer units 101-1 to 101-27 and the shift register 402 connected to each of the output layer units 301-1 to 301-28 of the output layer 300. A method of taking in data (the creation method is the same as in the initial learning) and learning a sequential network is used. At this time, since the learning data increases in a snowball manner with time, the learning data is discarded in order from the oldest data. Here, the number of learning data is kept constant. The new learning data is a set of 5 to 10 data obtained by changing the adjustment amount of the local device 2. Also, each input data and teacher data are paired.

Next, a process of confirming that the specified voltage width deviates within a short period will be described. In this process, a voltage deviation is confirmed as follows. (1) Check the impedance map (line impedance between the detection devices 3) for each high-voltage distribution line. (2) For each high-voltage distribution line, the voltage at each detection position is calculated as shown in FIG. 20 from the transmission voltage (predicted value), the line impedance, and the load current (predicted value).

(Calculation example) (a) In the case of only line impedance

V ₂ = V ₁ : Predicted voltage value at the detection position of address 1 (b) With load (line impedance: Z1)

V ₂ = V ₁ −I ₁ Z ₁

Equation 13] _{Z 1 = (r 12 + X} 12) 1/2: Voltage predictive value at the detection position of the address 1

(3) From the voltage values at the respective detection positions calculated by the above calculation, it is confirmed whether or not any voltage value deviates from the specified voltage range. (4) After confirming all addresses (all detection positions), if there is a high-voltage distribution line that is recognized as having a deviation, the target high-voltage distribution line and address, and predicted values (voltage, current, impedance: load Z is the reactive power It is calculated together with the predicted value.)
Then, the process proceeds to the macro simulation process (step S17 in FIG. 4).

Next, the process of selecting the target local device 2 or LRT 96 in step S7 in FIG. 3 will be described.
This processing includes (a) calculating the adjustment amount of the local device 2 as a result of checking the voltage and the reactive power value from the detection data and the data calculated from the prediction, and (b) calculating the adjustment amount of the local device 2. There are two cases in which the selection of the local device 2 needs to be reviewed. (A) Initial setting: In the initial setting, two cases are conceivable: a case of voltage deviation correction and a case of other (no tolerance, increase of reactive power, or lower voltage limit). (I) In the case of voltage deviation correction, start work in the morning, lunch break,
Focusing on the suspension of employment in the evening, etc., the local device 2 to be targeted is limited in advance in order to correct such a load fluctuation for a relatively short time as soon as possible. First, a voltage deviation point is recognized from an address, and an ICR 91 having a voltage adjustment function higher than the deviation point is limited as a target local device 2 for each high-voltage distribution line. When the ICR 91 is higher than the voltage deviation point, the process proceeds to the next process as selecting the LRT 96 for the high voltage distribution line. (II) In cases other than the voltage deviation correction, all the local devices 2 are recognized for each high-voltage distribution line.

(B) Cases other than the Initial Setting (a) In this process, in order to correct the voltage deviation, the local device 2 is once limited to obtain the adjustment amount and the voltage deviation is corrected. Since the correction cannot be made only by the local device 2, it is necessary to reselect the local device 2, and it is limited to the case of correcting the voltage deviation. (B) To correct the voltage deviation, an ICR 91 higher than the limited ICR 91 is added to the high-voltage distribution line as the target local device 2. Currently limited ICR91
If there is no higher ICR91, the high voltage distribution line is L
Proceed to the next process as RT96 selection.

Next, a description will be given of a process for obtaining the adjustment amount of the local device 2 that minimizes the objective function. In this process, the objective function and the target local device 2 are determined for each high-voltage distribution line from the pre-process, and under that condition, the adjustment amount of the local device 2 that minimizes the objective function is obtained. (1) The adjustment amount of the local device 2 that minimizes the objective function is obtained for each target high-voltage distribution line. (2) When the LRT 96 is not selected as the target local device 2, the adjustment amount of the target local device 2 is set at random, and the voltage and reactive power at each detection position at the adjustment amount are obtained using the NN500. From the above values, the objective function f
(X) is calculated, the adjustment amount of the local device 2 is updated until the end condition is satisfied, and the objective function f (x) is calculated. When the termination condition is satisfied, the local device adjustment amount at that time is set as a solution that minimizes f (x) by GA. (3) When the adjustment amount of the local device 2 that minimizes the objective function f (x) is obtained, the voltage at each detection position at that time is confirmed, and there is no portion that deviates from the specified voltage width. I do. If there is no voltage deviation, the adjustment amount of the target local device 2 in the target high-voltage distribution line is determined. If there is a voltage deviation, the process proceeds to the local device 2 selection process to review the target local device 2. (4) When the LRT 96 is selected as the target local device 2, a correction voltage is calculated, a transmission voltage corrected by the correction voltage is obtained, and how much the transmission voltage changes from the current state is recognized. Then, it calculates how each detected voltage of the high voltage distribution line connected to the LRT 96 changes. Therefore, it is confirmed whether a voltage deviation point newly occurs. If it does not occur, the constraint condition (sending voltage) is changed and a macro simulation process is performed. If it occurs, the high-voltage distribution line having no problem at present is deteriorated. Therefore, the LRT 96 is not corrected, and is recognized as an uncontrollable portion (a section where voltage deviation cannot be avoided), and the above (1) and (2) ) Is performed. Then, if the voltage deviation section is an uncontrollable part, the adjustment amount of the target local device 2 is determined as it is.

Further, the power flow is calculated by a simplified system, and the voltage and reactive power at each detection position when the local device 2 operates with the obtained adjustment amount are obtained, and it is confirmed whether or not the data is improved. (Step S17 in FIG. 4) will be described. (1) In this process, the power flow calculation is performed based on the adjustment amount of the local device 2 for each target high-voltage distribution line determined in the preprocessing and the system diagram (impedance map) of the target high-voltage distribution line created earlier. Then, the voltage and the reactive power at each detection position are calculated. (2) From the result and the detected data, the following objective function (average of the improvement rate) is calculated, and the improvement rate of the voltage and the reactive power is 5
If it is 0% or less, it is determined that the data has been improved. If one of the improvement rates exceeds 50%, the one with the worse improvement rate is recognized, and the adjustment amount is calculated again. (I) Voltage objective function fv (x):

[0080]

[Equation 14]

Here, Vi: detection data (voltage) at each detection position, Vf: voltage at each detection position calculated by power flow calculation, Vo: target voltage, n: number of detections. (II) Reactive power objective function fq (x):

[0082]

(Equation 15)

Here, Qi: detection data (reactive power value) at each detection position, Qf: reactive power value at each detection position calculated by power flow calculation, Qo: target reactive power, n: number of detections.

FIG. 21 is a graph showing a specific data processing method in the averaging process used in the distribution system control system of FIG. In the distribution system control system of the present embodiment, the central device 1 takes in data from sensors (hereinafter, referred to as sensors) of the local devices 2 and the detecting devices 3 and performs system optimization control while grasping the contents. Since it is impossible to simultaneously input data from the local devices 2 and the detection devices 3 scattered in the high-voltage distribution lines F1 and F2, a time difference occurs at the time of input. How to deal with this time difference will be described. Specifically, when collecting data of each sensor by polling from the central unit 1, since there is a time difference between the first collected detection data and the last collected detection data, they are simply treated as the same. Can not. Therefore, in this embodiment, a temporal averaging process is performed so that the detected data can be regarded as being collected at the same time. For example, if it is assumed that the collection time error is 1% or less in order to regard the collected detection data as simultaneous data, the detection data is averaged by each detection device in 100 times the polling cycle, and the obtained data is averaged. By collecting the data of the average value, it is treated as detection data at the same time (within a time error of 1%). However, if the detected data exceeds a certain width from the previously detected data, it is regarded as unique data and the data is discarded.

Further, the processing of steps S5 and S6 in FIG. 3 will be described. Even when an optimization processing request is issued from the remote monitoring function, there is a possibility that the current state may be worse than the current state due to the influence of the error of the NN500. In this case, it is determined that the current situation is almost optimal, and the optimization process, that is,
It is confirmed that the processing after step S7 in FIG. 3 is not executed (NO in step S6).

[0086]

(Equation 16)

[Equation 17]

Here, 60: error of voltage at each detection point of NN500 (V), 100: error of reactive power at each detection point of NN500 (kVar), Vi: current error at each detection position Voltages Vo, Vo: target voltage, Qi: reactive power at each current detection position, n: number of detections. That is, the appropriate condition represented by Expressions 16 and 17 is that the sum of the difference between the voltage value or the reactive power value at each of the current detection positions detected by each detection device 3 and the predetermined target value is a predetermined value. That is, it is not less than the threshold value.

[0088]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In order to verify the operation of the distribution system control system of FIG. 1 configured as described above, the present inventor performed the following simulation.

First, a simple verification of the power flow calculation function by the NN 500 will be described. In the NN 500, the calculation accuracy of the objective function greatly depends on the structure of the network. Therefore, how to determine the structure of the network becomes a problem. For this reason, the structure was examined and verified using the number of intermediate units as a parameter.

FIG. 22 is a block diagram showing a configuration of a verification power distribution system for simulating the power distribution system control system of FIG. The distribution system for verification is as shown in FIG. 22, and the number of units is the input layer 35 and the output layer 24. The number of intermediate units was 30, 40, and 50. Also,
For the learning data and the evaluation data, the transmission voltage is 690.
0V, each section load is 1A step in the range of 25 to 50A, load power factor is 0.01 step in the range of 0.4 to 0.9, voltage adjustment amount of ICR91 is 5V step in the range of ± 330V, The phase adjustment amounts of the ICR 91 and the TCR 92 were determined by random numbers in 0.1 degree steps within a range of ± 5 degrees. Conventional data flow calculation is executed using these data as input values, and the obtained voltage and reactive power value at each detection position are used as learning data, and the number of times of learning is 100, 200, and 40.
0 was set to 3 patterns. Evaluation data (data obtained by conventional power flow calculation) was generated in the same manner as the learning data.

In the evaluation of the NN500, the evaluation data based on the conventional power flow calculation result is regarded as a true value, and the error between the voltage and the reactive power at each detection position calculated by the NN500 and the evaluation data based on the conventional power flow calculation result. According to Voltage tolerance is 6
It is 60 V, which is equivalent to 1% of 600 V, and the allowable error of the reactive power is 300 kvar, which is equivalent to 10% of the maximum reactive power generated under the conditions for creating the evaluation data.

In determining the learning cutoff error of the NN 500, the learning cutoff error is determined in advance prior to the simple verification of the power flow calculation function by the NN 500.
The mean square error after the start of learning by the sequential learning method with 30 intermediate units and 100 learning data was determined. In this case, the mean square error was defined as shown in the following equation.

[0093]

(Equation 18)

FIG. 23 is a graph showing a learning curve of the neural network (NN) 500 in the simulation result of the distribution system control system of FIG. The mean square error after the start of learning is as shown in FIG. 23, and the learning cutoff error is equivalent to 15 minutes when the progress is slow (the number of learning times is 6).
000 times) was set to 0.0002.

Next, the results of the unit verification will be described.
Ninety-six different data were input to the NN500 and the conventional power flow calculation, and the difference between the calculated voltage and reactive power at each detection position was evaluated using the above evaluation criteria. In the case of learning times 100 and 200, the number of intermediate units is 30, 4
It was found that the evaluation standard value was not reached in all of 0 and 50, and that both of the functions were not satisfactory. However, when the number of times of learning was 200, both the calculation accuracy of the voltage and the reactive power was improved with respect to the case of 100. When the number of times of learning is 400, the evaluation reference value is almost achieved in all of the number of intermediate units 30, 40, and 50.

FIG. 24 shows a simulation result of the distribution system control system of FIG.
FIG. 25 is a graph showing the number of data with respect to the error range of the voltage, which is the result of the unit verification when the number of intermediate units is 30, and FIG. 25 is a simulation result of the distribution system control system of FIG. It is a graph which shows the number of data with respect to the error range of the reactive power which is a verification result. FIG. 26 is a table showing input conditions of an experiment in the simulation of the distribution system control system of FIG.

That is, FIGS. 24 and 25 show examples of the results of the unit verification with 30 intermediate units. When the input conditions are the same as in FIG. 26, a comparison between the output value of the conventional power flow calculation and the output value of the NN power flow calculation is shown in FIG.
As shown in FIG. 27, it can be seen that the power flow calculation function by the NN 500 is at a practically usable level. Furthermore, if the function of ICR91 is added in the conventional power flow calculation method, the processing speed per distribution line is about 30 s, whereas NN
In the case of 500, it was also confirmed that calculation could be performed at a high speed of 10 ms or less. Therefore, a substantially real-time operation can be performed.

Next, verification of the GA function will be described. The number of strings per distribution line is 100 and 200 as GA parameters, and the selection rate is 5
Assuming 0%, an appropriate number of strings is selected from the value of the objective function and the degree of convergence depending on the generation. FIG. 28 and FIG.
Reference numeral 9 denotes the distribution system FF of the verification wiring system in the simulation of the distribution system control system of FIG.
1 is a block diagram showing FF2. In other words, the distribution system for verification is as shown in FIGS. 28 and 29. The number of ICRs installed per distribution line is two for distribution system 1 and three for distribution system 2, and the load at each point is as shown in FIG. 30. In performing this verification, the fixed load value in FIG. 30 is also used in the fitness calculation process by the NN 500 in step S12 in FIG.
Only the amount of adjustment was made with random numbers, and pre-learning was performed with 400 learning data.

FIG. 31 shows the change of the objective function implemented for the distribution systems FF1 and FF2 shown in FIGS. 28 and 29 with 100 and 200 strings and up to 100 generations. As is clear from FIG. 31, the final convergence value of the objective function is the same for both, but the number of strings converged is smaller in the case of 200 strings than in the case of 100. It was decided to. In addition, both the voltage and the reactive power at all the detection positions in the distribution systems FF1 and FF2 have been improved from the current state. When the number of strings is 200, the average value of the objective function of the voltage of the distribution system FF1 is 0.47, the average value of the objective function of the reactive power is 0.92, and the average value of the objective function of the voltage of the distribution system FF2 is 0. .49, and the average value of the objective function of the reactive power was 0.48.

Further, the general verification of the GA and NN functions will be described. Comprehensive verification of GA and NN functions consists of a central unit 1 equipped with GA and NN functions and a simulated local device 2.
Are connected via the data communication line 4, and the measurement frame from the simulated local device 2 is returned to the polling frame and the target value command frame from the central unit 1. The GA and NN functions were confirmed by transmitting and receiving these data.

As the verification conditions at this time, the distribution system corresponds to the verification distribution system FF2 in FIG. 29, and the load pattern used is that of the verification distribution system FF2 in FIG. The initial condition in this case was “voltage deviates from the lower limit at the detection position”. The target voltage was 6650 V, the allowable voltage lower limit was 6300 V, and the GA termination condition was 300 non-update generations.

FIG. 32 is a graph showing a voltage improvement effect before and after the verification in the simulation of the distribution system control system of FIG. 1. FIG. 33 is a graph showing the reactive power before and after the verification in the simulation of the distribution system control system of FIG. It is a graph which shows an improvement effect. As is clear from FIGS. 32 and 33, the voltage and the reactive power at each detection position before and after the verification are shown. The optimization function of GA and NN worked effectively for both voltage and reactive power, confirming the validity of the system.

As described above, according to the present embodiment, since the distribution system control system is configured using the NN 500 and the GA, the voltage and reactive power of the entire distribution system can be reduced based on the detected data at several locations. Compared to the example, it is possible to adjust efficiently and adaptively, with higher accuracy and at a higher speed in real time, thereby minimizing the power loss of the whole system. Also, based on the result of the adjustment amount obtained using the NN500 and the GA, after confirming the reliability of the control by the conventional deterministic power flow calculation method,
Since each local device 2 is controlled, it is possible to control the power distribution system with higher accuracy and reliability. Further, the detection data from the sensors of the local devices 2 and the detection devices 3 uses average data over a certain period of time, and
Since the power distribution system is controlled by the N500 and the GA, the time error between each sensor and between the central device 1 and each sensor can be reduced, and the control of the power distribution system can be performed with higher accuracy and reliability.

Further, according to the present embodiment, since the power flow calculation of the power distribution system is executed by using the NN 500, the power flow simulation of the mutual influence of the complicated power distribution system can be realized. As a result, it can be executed in real time with higher accuracy and at higher speed, thereby minimizing the power loss of the entire system. Furthermore, according to the present embodiment, before the calculation of the adjustment amount by the NN 500 and the GA, the calculation of the adjustment amount by the NN 500 and the GA is performed when the current distribution system satisfies a predetermined appropriate condition. , The current state is maintained, so that the influence of the error caused by the NN 500 is minimized, and the process of deteriorating the state of the current distribution system is prevented, thereby preventing erroneous control and improving control efficiency. Can be.

The present apparatus can be applied not only to the power distribution system but also to power flow calculation or state calculation of any network having various control devices.

[0106]

As described above in detail, according to the distribution system control system according to the first aspect of the present invention, the bus connected to the distribution transformer and the plurality of high-voltage distribution lines connected to the bus are connected to each other. In a distribution system control system that controls the fluctuation range of the voltage of each high-voltage distribution line of the distribution system having a set range, a central device that controls the operation of the distribution system and controls the operation of each of the high-voltage distribution lines At least 1 for
Provided based on a control signal from the central device,
A local device that adjusts the voltage value and the reactive power value of each of the high-voltage distribution lines by a predetermined adjustment amount; and a local device that adjusts the high-voltage distribution line at at least one detection position apart from the installation location of the local device of each of the high-voltage distribution lines. A detection device for transmitting a detection signal of the detected voltage value and current value to the central device, the central device comprising: an input layer; A layer having an intermediate layer and an output layer, which is learned based on predetermined teacher data in advance, the voltage adjustment amount and the phase adjustment amount of each of the local devices, the predetermined load active power and the predetermined load active power in a section between the detection devices. Inputting the input data including the load reactive power and the transmission voltage and the transmission phase from the distribution transformer, and outputting the voltage and the reactive power value at each of the detection positions. Using a neural network that performs power flow calculation by outputting data, a predetermined initial adjustment amount, a predetermined load active power and a load reactive power in a section between the respective detection devices, and a signal from the distribution transformer. Calculating means for calculating a voltage value and a reactive power value at each of the detection positions based on the sending voltage and the sending phase of the detecting means; and a voltage value and a current detected by each of the detecting devices before the processing by the calculating means. Based on the value detection data, it is determined whether or not the current distribution system is in an appropriate state using predetermined appropriate conditions, and when it is in an appropriate state, control is performed so as not to execute the processing by the calculation means. Control means for performing the operation. Therefore, by minimizing the effect of the calculation error by the neural network and avoiding the process of deteriorating the state of the current distribution system, erroneous control is prevented,
Control efficiency can be improved.

According to the distribution system control system of the second aspect, in the distribution system control system of the first aspect, the appropriate condition is a voltage at each of the current detection positions detected by each of the detection devices. The sum of the difference between the value or the reactive power value and the predetermined target value is equal to or less than a predetermined threshold value. Therefore, erroneous control can be prevented and control efficiency can be improved by minimizing the effect of the calculation error by the neural network and avoiding the process of deteriorating the current state of the power distribution system.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a distribution system control system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing an example of a specific example of the distribution system control system of FIG.

FIG. 3 is a flowchart showing a first part of the central device processing executed by the central device 1 of FIG. 1;

FIG. 4 is a flowchart showing a second part of the central device processing executed by the central device 1 of FIG. 1;

FIG. 5 is a flowchart showing a fitness calculation process which is a subroutine of FIG. 4;

FIG. 6 is a flowchart illustrating a selection culling process that is a subroutine of FIG. 4;

FIG. 7 is a flowchart showing central device interrupt processing executed by the central device 1 of FIG. 1;

8 is a diagram showing a setting table of a weight coefficient in an objective function used in the distribution system control system of FIG. 1;

FIG. 9 is a diagram illustrating an example of a string expression of a voltage adjustment amount used in the distribution system control system of FIG. 1;

FIG. 10 is a diagram illustrating an example of a string expression of a phase adjustment amount used in the distribution system control system of FIG. 1;

11 is a diagram illustrating an example of an individual string used in the distribution system control system of FIG. 1;

FIG. 12 is a diagram illustrating an example of crossover of strings used in the distribution system control system of FIG. 1;

FIG. 13 is a diagram illustrating an example of mutation of a string used in the distribution system control system of FIG. 1;

FIG. 14 is a diagram showing an example of multiple crossover of strings used in the distribution system control system of FIG. 1;

15 is a block diagram showing a configuration of a neural network (NN) 500 used in the distribution system control system of FIG.

FIG. 16 is a block diagram showing a process of generating an impedance map in a certain load state used in the distribution system control system of FIG. 1;

17 is a diagram illustrating a voltage value V and a reactive power value Q of a sensor at each adjustment amount used in the distribution system control system of FIG. 1;
It is a block diagram which shows the calculation process of.

FIG. 18 shows the neural network (NN) shown in FIG.
It is a block diagram which shows the learning method of 500.

FIG. 19 shows the neural network (NN) shown in FIG.
FIG. 5 is a block diagram illustrating a method of incorporating learning data into a learning program;

20 is a diagram showing an example of voltage calculation from load prediction in the distribution system control system of FIG.

21 is a graph showing a method of processing singular data in the averaging process used in the distribution system control system of FIG.

FIG. 22 is a block diagram illustrating a configuration of a verification power distribution system that simulates the power distribution system control system of FIG. 1;

23 is a neural network (NN) 5 in a simulation result of the distribution system control system of FIG.
12 is a graph showing a learning curve of 00.

24 is a graph showing a simulation result of the distribution system control system of FIG. 1 and showing the number of data with respect to a voltage error range as a result of unit verification when the number of intermediate units is 30. FIG.

25 is a graph showing simulation results of the distribution system control system of FIG. 1 and showing the number of data with respect to the error range of the reactive power, which is the result of unit verification when the number of intermediate units is 30. FIG.

26 is a table showing input conditions of an experiment in the simulation of the distribution system control system of FIG. 1;

FIG. 27 is a graph showing simulation results of the distribution system control system of FIG. 1 and showing voltage and reactive power with respect to distribution system span length.

FIG. 28 is a block diagram showing a distribution system FF1 of a verification wiring system in the simulation of the distribution system control system of FIG. 1;

FIG. 29 is a block diagram showing a distribution system FF2 of the verification wiring system in the simulation of the distribution system control system of FIG. 1;

30 is a table showing load values at respective load points in the distribution systems FF1 and FF2 of the verification wiring system in the simulation of the distribution system control system of FIG.

FIG. 31 is a graph showing an evaluation curve of the number of strings in a simulation of the distribution system control system of FIG. 1;

FIG. 32 is a graph showing the effect of improving voltage before and after verification in a simulation of the distribution system control system of FIG. 1;

FIG. 33 is a graph showing the effect of improving the reactive power before and after verification in the simulation of the distribution system control system of FIG. 1;

[Explanation of symbols]

 Reference Signs List 1 central device, 2 local device, 2A local device, 2B voltage sensor, 2C current sensor, 2D device control device, 2E communication control and control target setting device, 3 detection device, 3A voltage sensor , 3B: current sensor, 3C: communication slave station, 4: data communication line, 5: data transmission / reception device, 90: LRT target correction unit, 91: ICR, 92: TCR, 93: sensor, 94: transmission line for power distribution automation Reference numeral 95: voltage adjustment relay; 96: LRT; 100: input layer; 101-1 to 101-27: input layer unit; 200: intermediate layer; 201-1 to 201-M: intermediate layer unit; 300: output layer; 301-1 to 301-28: output layer unit, 401, 402: shift register, SS: distribution transformer, B: bus, F1, F2: high-voltage distribution line.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Masaru Yukawa 2-1-1, Tagawa, Yodogawa-ku, Osaka-shi, Osaka Dainai Co., Ltd. (72) Haruya Abe 2-1-1, Tagawa, Yodogawa-ku, Osaka-shi, Osaka No.11 Daihen Co., Ltd. (72) Inventor Masao Shimamoto 2-1-1, Tagawa, Yodogawa-ku, Osaka-shi, Osaka Prefecture (72) Inventor Takashi Hashimoto 2-1-1, Tagawa, Yodogawa-ku, Osaka-shi, Osaka Daihen Co., Ltd.

Claims (2)

[Claims] An electric power distribution system having a bus connected to a distribution transformer and a plurality of high-voltage distribution lines connected to the bus is controlled so that the fluctuation range of the voltage of each high-voltage distribution line falls within a set range. In the distribution system control system, a central device that controls the operation of the distribution system and controls at least one of the high-voltage distribution lines based on a control signal from the central device. A local device that adjusts a voltage value and a reactive power value of the distribution line by a predetermined adjustment amount, and a voltage and a current of the high-voltage distribution line at at least one detection position apart from a location where the local device of each of the high-voltage distribution lines is installed. A voltage sensor and a current sensor for respectively detecting the detected voltage value and the current value, and a detection device for transmitting detection data of the detected voltage value and current value to the central device. Comprising a layer, at least one intermediate layer, an output layer,
It is learned based on predetermined teacher data in advance, and the voltage adjustment amount and the phase adjustment amount of each of the local devices, the predetermined load active power and the load reactive power in the section between the detection devices, and the distribution transformer. The input data including the transmission voltage and the transmission phase is input, and the output data including the voltage value and the reactive power value at each of the detection positions is output. The amount of adjustment, the predetermined load active power and the load reactive power in the section between the detection devices, and based on the transmission voltage and the transmission phase from the distribution transformer,
Calculating means for calculating a voltage value and a reactive power value at each of the detection positions; and, prior to the processing of the calculating means, the current power distribution based on the detection data of the voltage value and the current value detected by each of the detection devices. Control means for determining whether or not the system is in an appropriate state using predetermined appropriate conditions, and controlling the system so as not to execute the processing by the calculating means when the system is in an appropriate state. Distribution system control system. 2. The appropriate condition is that a sum of a difference between a current voltage value or a reactive power value detected by each of the detection devices at each of the detection positions and a predetermined target value is equal to or less than a predetermined threshold value. The distribution system control system according to claim 1, wherein: