WO2014087539A1

WO2014087539A1 - Power grid control system and power grid control method

Info

Publication number: WO2014087539A1
Application number: PCT/JP2012/081797
Authority: WO
Inventors: 山根　憲一郎; 翔太 ▲逢▼見; 正俊熊谷
Original assignee: 株式会社日立製作所
Priority date: 2012-12-07
Filing date: 2012-12-07
Publication date: 2014-06-12
Also published as: JPWO2014087539A1; JP5962770B2

Abstract

The problem addressed by the present invention is to perform highly reliable power control. A high reliability region (1801) between a prescribed upper limit and lower limit and a low reliability region (1802) excluding the high reliability region (1801) are preliminarily set as reliability region data. When performing power control in a distributed control mode where a control amount is calculated using computation model because of a deterioration in the state of communication, a power grid control system determines whether or not there is measured data in the low reliability region, calculates a provisional amount of control based on the amount of control in the high reliability region when the measured data for low reliability exists, and performs control using the provisional amount of control.

Description

Power system control system and power system control method

The present invention relates to a technology of a power system control system and a power system control method for controlling a power system.

Control by the control device such as SVR (Step Voltage Regulator), SVC (Static Voltage Compensator), etc., so that the voltage at its own end (position where the control device is installed) becomes the target voltage. There is technology to do.

Furthermore, so-called centralized control is disclosed in which the monitoring and control server collectively grasps the state of the entire power system and gives an optimal control command to each control device (see, for example, Patent Document 1). In addition, a technique is also disclosed in which when reactive power is monitored on the SVR side and reactive power is measured for a predetermined time or longer, the SVC is estimated to be operating and tap switching control of the SVR is performed (for example, Patent Document 2). This technology is a so-called autonomous distributed control technology that does not assume communication.

JP 2007-288877 A JP 2011-217581 A

In recent years, so-called all-electric houses that use cooking utensils, water heaters, air conditioners, lighting devices, and the like in houses have become popular. In addition, large-capacity loads such as electric water heaters and electric vehicles that produce hot water using cheap electricity at night are becoming widespread, and power demand is diversifying.

On the other hand, auxiliary power generation using distributed power sources such as solar power generation, fuel cells, and home storage batteries is being performed in various places. Therefore, due to such a situation, fluctuations in the state (voltage) of the power system, in particular the distribution system, become large, and maintaining this within an appropriate range (101 ± 6V or 202 ± 20V) defined by the Electricity Business Law, It is expected to become increasingly difficult in the future.

In the conventional technology, the voltage fluctuation at its own end is suppressed to an appropriate range by SVR or SVC. However, SVR and SVC basically operate independently and do not cooperate with other control devices in the vicinity. Therefore, if a large-capacity load is provided in the power system or a large number of distributed power sources are connected to the power system, it may be difficult to suppress voltage fluctuations in the power system within an appropriate range.

The technique described in Patent Document 1 is effective when the communication environment between the monitoring control server and each control device is stable and high-speed and high-quality communication can be performed. However, for example, there are many cases where a high-speed and high-quality communication environment is not in place such as in remote areas in mountainous areas. In such a case where a high-speed and high-quality communication environment cannot be used, it is conceivable that the technique described in Patent Document 1 may delay the transmission of a control command from the monitoring control server to the control device or may not reach the control command. . Therefore, when a high-speed and high-quality communication environment cannot be used, there is a possibility that appropriate control cannot be sufficiently performed with the technique described in Patent Document 1.

The technique described in Patent Document 2 is considered to be effective when the SVR is on the upstream side (substation side) of the SVC and is installed at a position where the distance is short. However, in cases other than such a configuration, or when the positional relationship is not such that reactive power output by the SVC can be correctly detected on the SVR side (SVR is upstream of the SVC and the distance is short, etc.), Patent Literature In the technique described in No. 2, it is considered difficult to perform appropriate control. Similarly, when two or more SVCs are connected to the same power system, similarly, it is considered difficult to accurately estimate the reactive power output by each SVC using the SVR.

In the future, the possibility of expansion of voltage fluctuations is expected to increase with the spread of large-capacity loads and distributed power sources. As a countermeasure to such a problem, it is expected that a large number of control devices including SVR and SVC are installed in the power system. In such a case, it is conceivable that there are many power systems that do not satisfy the specific installation conditions (high-speed and high-quality communication environment). In such a case, it is considered difficult to appropriately control the voltage of the power system with the techniques described in Patent Document 1 and Patent Document 2.

Furthermore, if a large disturbance in the output of the load or the distributed power supply occurs and the reliability of the sensor output or the like decreases, it may be difficult to perform appropriate control with the conventional control technology.

The present invention has been made in view of such a background, and an object of the present invention is to perform power control with high reliability.

In order to solve the above problems, the present invention is characterized in that, in the distributed control mode, when the reliability of the measurement data is low, the electric power system is controlled based on the temporary control amount.

According to the present invention, highly reliable power control can be performed.

It is a figure showing an example of composition of an electric power system control system concerning a 1st embodiment. It is a figure which shows the hardware structural example of the central apparatus and control apparatus which concern on this embodiment. The structural example of the electric power grid | system which concerns on this embodiment is shown. It is a figure which shows the structural example of the node management data which concerns on 1st Embodiment. It is a figure which shows the structural example of the branch management data which concerns on 1st Embodiment. It is a figure which shows the structural example of the control apparatus management data which concerns on 1st Embodiment. It is a figure which shows the structural example of the controlled variable data which concerns on 1st Embodiment. It is a figure which shows the structural example of the measurement data which concerns on 1st Embodiment. It is a figure for demonstrating transfer of the control mode which concerns on 1st Embodiment. It is a flowchart which shows the procedure of the control amount calculation process in the central apparatus which concerns on 1st Embodiment. It is a flowchart which shows the procedure of the calculation model calculation process in the central apparatus which concerns on 1st Embodiment. It is a figure for demonstrating the specific meaning of the calculation model which concerns on 1st Embodiment. It is a figure which shows a specific example of the model data which concern on 1st Embodiment. It is a flowchart which shows the process sequence of the control apparatus which concerns on 1st Embodiment. It is a figure for demonstrating switching a calculation model according to a time slot | zone. It is a figure which shows the structural example of the central apparatus which concerns on 2nd Embodiment. It is a figure which shows the structural example of the control apparatus which concerns on 2nd Embodiment. It is a figure which shows the example of the confidence interval which concerns on 2nd Embodiment. It is a figure which shows the example of the confidence area data which concern on 2nd Embodiment. It is a flowchart which shows the procedure of the control processing which concerns on 2nd Embodiment. It is a figure for demonstrating the calculation method of the temporary control amount which concerns on 2nd Embodiment (the 1). It is a figure for demonstrating the calculation method of the temporary control amount which concerns on 2nd Embodiment (the 2). It is a figure which shows the structural example of the control apparatus which concerns on 3rd Embodiment. It is a flowchart which shows the procedure of the control processing which concerns on 3rd Embodiment.

Next, modes for carrying out the present invention (referred to as “embodiments”) will be described in detail with reference to the drawings as appropriate. In addition, in each drawing, about the same component, the same code | symbol is attached | subjected and description is abbreviate | omitted.

[First Embodiment]
The power system control system 5 according to the first embodiment will be described with reference to FIGS. In this embodiment, the power system which monitors the communication state of the communication network 4 for controlling the power system comprised by a transformer etc., switches a control mode according to a communication state, and controls the control apparatus 2 appropriately. An example of the control system 5 will be described.

<System configuration>
FIG. 1 is a diagram illustrating a configuration example of a power system control system according to the first embodiment.
In the power system control system 5, a central device (management device) 1, a plurality of control devices 2 (2a to 2c), and a plurality of sensors 3 (3a to 3c: measuring devices) can communicate with each other via a communication network 4. It is connected.

(Central device)
The central device 1 is configured as a computer system, for example.
The central apparatus 1 includes a communication state monitoring unit 101, a power state estimation unit 102, a control amount calculation processing unit 103, a model generation unit 104, a transmission / reception processing unit (transmission unit) 105, a measurement data storage unit 111, a system data storage unit 112, and A model data storage unit 113 is included.
The communication status monitoring unit 101 monitors the communication status of the communication network 4. The communication state monitoring unit 101 collects time-stamped measurement data from each sensor 3 at a predetermined constant monitoring cycle Tm, and the collected time-stamped measurement data is stored in the measurement data storage unit 111 as measurement data. Store. Hereinafter, the measurement data is assumed to be measurement data with a time stamp.

The power state estimation unit 102 estimates the power state of the power system control system 5. Specifically, the power state estimation unit 102 determines observability by determining whether or not a sufficient number of measurement data for estimating the power state has been obtained. Then, the power state estimation unit 102 uses the measurement data and system data determined to be normal by a method according to whether it is observable or unobservable, and uses the power state (active power, reactive power, voltage, etc.) of the entire power system. ).
The control amount calculation processing unit 103 calculates the control amount of each control device 2 and transmits each control amount to the control device 2. When the control device 2 is an LRT (Load Ratio control Transformer) or SVR, the control amount is a tap number (transformation ratio). In this case, the control device 2 refers to a tap map (a list in which tap numbers corresponding to the transformation ratio are included) included in the controlled variable, and switches the tap to the corresponding tap number. Note that the switching operation is not performed when the tap number has already been reached.
The model generation unit 104 calculates a power state and a control amount calculation model. The calculation model will be described later.
The transmission / reception processor 105 transmits / receives various data to / from each control device 2 and each sensor 3. That is, the transmission / reception processing unit 105 transmits various data such as control amount data, model data, and system data described later to the control device 2. The data transmitted to these control devices 2 may be transmitted at different periods for each data type.

Measured data transmission cycle is predetermined in consideration of each power system, communication line specifications, number of communication devices, target performance, and the like. The transmission period is set to a value such as 1 minute, 3 minutes, 10 minutes, 30 minutes, 60 minutes, or the like. These specific numerical values are examples, and are not limited to these values.

The measurement data storage unit 111 stores measurement data that is data related to the measurement values measured by the sensor 3. The measurement data will be described later.
The system data storage unit 112 stores system data that is data relating to the configuration of the power system, the specifications of the devices that configure the power system, and the like. System data will be described later.
The model data storage unit 113 stores model data that is data related to the power state and the control amount calculation model generated by the model generation unit 104.

(Control device)
Next, the configuration of the control device 2 will be described.
The control device 2 is a device for mainly controlling the voltage in the state of the power system such as LRT, SVR, SVC, switched capacitor, battery-equipped PCS (Power Conditioning System), etc., and includes a controller 200 and a control processing unit. 204.
The controller 200 determines a control amount, and includes a transmission / reception processing unit 201, a control mode determination processing unit 202, a control amount calculation processing unit 203, a system data storage unit 211, a model data storage unit 212, and a measurement data storage unit 213. Have.
The transmission / reception processing unit 201 transmits / receives data to / from the central device 1 and each sensor 3.
The control mode determination processing unit 202 determines any one of a plurality of control modes based on the communication state. The control mode includes a central control mode, a distributed control mode, and an autonomous control mode. Each control mode will be described later.
When the control mode determined by the control mode determination processing unit 202 is the distributed control mode, the control amount calculation processing unit 203 calculates a control amount output by itself using an arithmetic model.

In the system data storage unit 211, system data sent from the central apparatus 1 is stored.
The model data storage unit 212 stores model data sent from the central apparatus 1.
The measurement data storage unit 213 stores measurement data that is data related to the measurement values measured by the sensor 3.
That is, the system data storage unit 211, the measurement data storage unit 212, and the model data storage unit 213 of the control device 2 are similar to the system data storage unit 112, the model data storage unit 113, and the measurement data storage unit 111 in the central device 1. Data is stored. However, when the control device 2 performs control in the autonomous control mode to be described later, only measurement data from the sensor 3 provided in the self-control device 2 of the control device 2 may be accumulated as measurement data. .

The control processing unit 204 performs control output according to the control amount sent from the central apparatus 1 or the control amount calculated by the control amount calculation processing unit 203. The control processing unit 204 refers to a tap map (a list in which tap numbers corresponding to the transformation ratio are included) included in the controlled variable, and switches the taps to the corresponding tap numbers.

When the control device 2 is an SVC, a switched capacitor, or a PCS with a battery, there are two types of control outputs: reactive power output or target voltage. When the control output is a reactive power output, it is output as “advance 50 kvar” or “delay 30 kvar”.

When the control output is the target voltage, the control processing unit 204 first monitors the difference between the voltage at the installation point of the control device 2 and the target voltage. Then, the control processing unit 204 uses the difference and the reactance between the installation point of the control device 2 and the inverter (capacitor) to set the target voltage, for example, by PI (Proportional Integral) control (proportional control, integral control). The reactive power to be output is determined so as to match, and the output operation is performed according to the reactive power.

(Sensor)
Next, the sensor 3 will be described.
The sensor 3 is a device that measures a power state quantity at an installation position on the power system or the like of the sensor 3 itself. The power state quantity measured by the sensor 3 is transmitted to the central device 1 and the control device 2 via the communication network 4. Data of the power state quantity measured by the sensor 3 is measurement data.

Each sensor 3 measures active power P, reactive power Q, voltage V, etc. every predetermined time, such as every 15 minutes, according to the type and role in a predetermined time period, and outputs it as measurement data. be able to. Further, the sensor 3 can also measure and output a current (which may include a flow direction), a voltage, and a power factor. Note that the active power P and the reactive power Q can also be calculated by a calculation formula using current, voltage, and power factor. In the case of a three-phase three-wire AC circuit, active power P, reactive power Q, current, and power factor are measured in phase units, and voltage is measured in line units. In order to simplify the measurement, only the representative phase (for example, the U phase) or between the representative lines (for example, the UV phase) may be measured.

The transmission / reception processing unit 105 of the central apparatus 1 stores the measurement data received from the sensor 3 in the measurement data storage unit 111. Similarly, the transmission / reception processing unit 201 of the control device 2 stores the measurement data received from the sensor 3 in the measurement data storage unit 213. The measurement data will be described later.

(Communication network)
Next, the communication network 4 will be described.
The communication network 4 is a communication network that connects the central device 1, the control device 2, and the sensor 3 to each other. Each of the central device 1, the control device 2, and the sensor 3 transmits and receives various data such as control commands and measurement data to each other using the communication network 4. The communication network 4 may be a wired network such as a public line such as a telephone line, Ethernet (registered trademark), a dedicated communication line, and a power line carrier communication line. Alternatively, the communication network 4 may be a wireless network such as a mobile phone communication network, a PHS (Personal Handy-phone System), a business radio, a satellite line, a wireless LAN (Local Area Network), or ZigBee (registered trademark).

<Hardware configuration>
FIG. 2 is a diagram illustrating a hardware configuration example of the central device and the control device according to the present embodiment.
FIG. 2A is a diagram illustrating a hardware configuration example of the central device.
The central apparatus 1 has a CPU (Central Processing Unit) 301, a RAM (Random Access Memory) 302, a ROM (Read Only Memory) 303, an HD (Hard Disk) 304, a communication interface 305 such as a LAN (Local Area Network) card, etc. as a bus. 306 is connected.
In the central apparatus 1 of FIG. 1, the communication state monitoring unit 101, the power state estimation unit 102, the control amount calculation processing unit 103, the model generation unit 104, and the transmission / reception processing unit 105 are expanded in the RAM 302 by a program stored in the HD 304. This is realized by being executed by the CPU 301.
Further, the measurement data storage unit 111, the system data storage unit 112, and the model data storage unit 113 are realized by the HD 304.

FIG. 2B is a diagram illustrating a hardware configuration example of the control device.
In the control device 2, a CPU 401, a ROM 402, and a communication interface 403 are connected via a bus 404.
The control processing unit 204, the controller 200, the transmission / reception processing unit 201 constituting the controller 200, the control mode determination processing unit 202, and the control amount calculation processing unit 203 in the control device 2 of FIG. Realize by executing.
Further, the system data storage unit 211, the model data storage unit 212, and the measurement data storage unit 213 are formed in the ROM 402.

<Example of power system configuration>
FIG. 3 shows a configuration example of the power system according to the present embodiment.
The power system is roughly divided into nodes 31 (31a to 31g) and branches 32 (32a to 32f), and each node 31 and branch 32 has attribute data.

The node 31 is a place where a predetermined device is installed.
For example, the node 31a is a transformer installed in a substation, to which a sensor 3a (3) and a control device 2a (2) are connected. The node 31b is a pole transformer installed on a utility pole, to which a sensor 3b (3) is connected. The node 31c is a pole transformer installed on the utility pole, but the control device 2 and the sensor 3 are not installed there. The node 31d indicates a load or power source such as a customer or a distributed power source, to which a sensor 3c (3) and a control device 2b (2) are connected. The node 31e is a utility pole in which a pole transformer is installed, and the sensor 3 is not installed there. The node 31f is a utility pole in which a pole transformer is installed, and a sensor 3d (3) is installed there. The last node 31g indicates a load or power source such as a customer or a distributed power source, to which a sensor 3e (3) and a control device 2c (2) are connected.

Next, the branch 32 (32a to 32f) will be described. The branch 32 is a path section between the node 31 and the node 31, and specifically, a section of a power transmission line or a distribution line. In other words, nodes other than the branch 32 are nodes 31. Each branch 32 has a resistance R and a reactance X as its impedance. Strictly speaking, the branch 32 also has a capacitance, but in this example, it is considered to be sufficiently smaller than the others and is ignored.

The branch 32a in FIG. 3 is a section in which the start point node is 31a and the end point node is 31b. The start point of the branch 32b is a node 31b and the end point is a node 31c. Furthermore, the start point of the branch 32c is a node 31c, and the end point is a node 31d. The start point of the branch 32d is a node 31d, and the end point is a node 31e. The start point of the branch 32e is a node 31e, and the end point is a node 31f. The start point of the branch 32f is a node 31f and the end point is a node 31g.

<System data>
Next, various data constituting the system data will be described with reference to FIGS.
(Node management data)
FIG. 4 is a diagram illustrating a configuration example of node management data according to the first embodiment.
The node management data is data that forms part of the system data 10.
The node management data has fields of a node ID (Identification), a substation flag, a pole transformer flag, a sensor ID, a control device ID, a measured value, and a power state estimated value.

The node ID is information for identifying the node.
The substation flag is information indicating the presence or absence of a substation.
The pole transformer flag is information indicating the presence or absence of a pole transformer.
The sensor ID is information for identifying the sensor 3 and information indicating the presence or absence of the sensor 3.
The control device ID is information for identifying the control device 2 and information for identifying the presence or absence of the control device 2.
The measured value is a measured value by the sensor 3. The measured value is included in the measured data.

The power state estimated value is a value of a power state estimated using a calculation model described later using all or a part of measurement data sent from each sensor 3 at each node. Specifically, the power state estimated value is an estimated value of active power P (PA to PG), reactive power Q (QA to QG), estimated voltage V (VA to VG), and the like.
The timing at which the power state estimated value is calculated is preferably synchronized with the sensor 3 (for example, every 15 minutes). In the

nodes

31c and 31e where the sensor 3 is not installed, the power state at the nodes is estimated from the power flow calculation and stored in the node management data. In many cases, the power state estimated value and the measured value are different from each other.
In addition, when a centralized control mode or an autonomous control mode, which will be described later, is being performed, a power state estimated value calculated based on power flow calculation or the like is stored in the power state estimated value.

As shown in FIG. 4, it is desirable that the power state estimation value is calculated for all nodes. Alternatively, the user may specify a node for calculating the power state estimated value.

In addition to the items shown in FIG. 4, other items may be added to the node management data. The shown items are divided into a plurality of node management data, and the divided node management data are linked to each other. Or the structure matched with a pointer etc. may be sufficient. The same can be said for not only the node management data but also other data described later.

(Branch management data)
FIG. 5 is a diagram illustrating a configuration example of branch management data according to the first embodiment.
The branch management data is data that forms part of the system data.
The branch management data includes fields of branch ID, start node ID, end node ID, resistance value R (Ω), and reactance X (Ω).
The branch ID is information for identifying a branch.
The start point node ID is the node ID of the node that is the start point of the target branch.
The end node ID is the node ID of the node that is the end point of the target branch.
The resistance value is the resistance value of the target branch.
The reactance is the reactance of the target branch.
The resistance value and reactance are calculated from the voltage at the node, the material of the branch, and the like.

(Control device management data)
FIG. 6 is a diagram illustrating a configuration example of control device management data according to the first embodiment.
The control device management data is data that forms part of the system data.
The control device management data has fields of control device ID, reference voltage (V), LDC (Line Drop Compensator) parameters R, X (Ω), operation time limit (sec), and rated capacity (kVA / kvar). Yes.
The control device ID is information for identifying the control device 2.
The reference voltage, the LDC parameter, and the rated capacity are values determined by the specifications of each control device 2.
The LDC parameter is used for a device in which the control device 2 adjusts a voltage according to a load such as LRT or SVR. Specifically, the impedance (resistance, reactance) and dead zone from the control device 2 to a reference point are used. Is numerical data composed of Therefore, when the control device 2 is other than a device that adjusts the voltage according to the load such as LRT or SVR, the LDC parameter is invalid.
The operation time period is the time from when the control device 2 operates until the next operation becomes possible.

(Control amount data)
FIG. 7 is a diagram illustrating a configuration example of control amount data according to the first embodiment.
The control amount data has a control device ID, a control amount, and a time stamp.
The control device ID is an ID for identifying the control device 2 and is also a destination of control amount data.
The control amount is a control amount in the control device 2 corresponding to the control device ID.
The time stamp is the date and time when the control amount data is transmitted.
The control amount data in FIG. 7 is data transmitted from the central device 1 to the control device 2, and when the control device 2 performs centralized control, the control device 2 identifies itself with the control amount included in the control amount data. Control.

(Measurement data)
FIG. 8 is a diagram illustrating a configuration example of measurement data according to the first embodiment.
One record of measurement data corresponds to the measurement data sent from the sensor 3.
The measurement data includes a time stamp, a sensor ID, and each measurement value measured by the target sensor 3. In the present embodiment, the measured values are active power P (kW), reactive power Q (kvar), and voltage V (V). In addition to this, there are current, power factor, power flow direction (current direction), and the like. May be included.
The time stamp is the date and time when the measurement value is measured by the target sensor 3.
The sensor ID is information for identifying the sensor 3.

<Control mode transition>
FIG. 9 is a diagram for explaining the transition of the control mode according to the first embodiment.
In the present embodiment, there are three power system control modes: (1) centralized control mode, (2) distributed control mode, and (3) autonomous control mode.
These three control modes will be described before describing the transition of the control mode.

(1) Centralized control mode The centralized control mode is a mode in which the central device 1 controls all the control devices 2, and corresponds to the centralized control of the technique described in Patent Literature 1. Specifically, the central device 1 calculates the control amount of each control device 2, and each control device 2 controls itself based on the control amount sent from the central device 1. The central control mode is excellent in terms of efficiency and the like because the central device 1 performs all control.

(2) Distributed control mode The distributed control mode calculates and controls its own control amount based on the power state / control amount calculation model calculated by the central device 1 and the measurement data acquired from the sensor 3. Mode. In the distributed control mode, the control amount from the central device 1 is not used, and the control amount can be calculated by the control device 2 itself.

(3) Autonomous control mode The autonomous control mode is a mode in which the control device 2 calculates and controls its own control fee regardless of the central device 1, and corresponds to the autonomous distributed control of the technology described in Patent Document 2. .
The distributed control mode that is controlled after estimating the power state of the power system is different from the autonomous control mode that performs control so as to simply cancel the deviation from the target voltage at the local node.

When the control is performed in the autonomous control mode, the control amount calculation processing unit 203 of the control device 2 performs a predetermined predetermined value based on the control parameter of each control device 2 included in the system data and the measurement data. The control amount is calculated according to the method. For example, when the control device 2 is LRT or SVR, the control amount to be output by itself is calculated based on the constant voltage control in the SVC, the switched capacitor, and the battery-equipped PCS based on the LDC method, and the control processing unit 204 Performs its own control based on this control amount.

The priority order is (1) centralized control mode, (2) distributed control mode, and (3) autonomous control mode.

These control modes are shifted depending on the communication state.
Control mode transition control will be described with reference to FIG.
First, in the centralized control mode, when the central device 1 calculates the control amount of the control device 2, it transmits the calculated control amount to the control device 2 as control amount data. At this time, the control amount data is given a time stamp that is a transmission date and time.
The control device 2 compares the difference Td between the time stamp of the control amount data previously received from the central device 1 and the current time with a preset first predetermined time T1. Here, the first predetermined time T1 is set to be equal to or longer than the transmission period Tc of control amount data by the central device 1 (T1 ≧ Tc).
Here, when Td <T1, that is, when the control amount data is transmitted from the central device 1 to the control device 2 at predetermined time intervals, the control device 2 has a communication soundness. It is determined that it is maintained, and control based on the centralized control mode is performed (S1).

In the case of Td ≧ T1, that is, when the control amount data is not transmitted from the central device 1 to the control device 2 at predetermined time intervals (communication amount is equal to or less than a predetermined value), the control device 2 Then, it is determined that the soundness of communication has been lost, and the process shifts to the distributed control mode (S2).
Under the distributed control mode, the control device 2 estimates the power state of the power system in itself based on the power state estimation model received from the central device 1 and predetermined measurement data. Furthermore, the control device 2 calculates its own control amount using the control amount calculation model, and executes a predetermined control operation (performs a predetermined control output). The power state estimation model is a calculation model for estimating the power state from the measured value, and the control amount calculation model is a calculation model for calculating the control amount in itself from the estimated power state.

Therefore, by performing the distributed control mode, the control device 2 is sent from the central device 1 in advance even when the control amount calculated by the central device 1 cannot be received in a predetermined cycle (when the communication state deteriorates). The control amount can be calculated based on the calculation model, and appropriate control can be performed.

During control in the distributed control mode, when the communication state recovers soundness and the control device 2 receives the control amount data from the central device 1 within the first predetermined time T1 (Td <T1), the control device 2 It is determined that the communication state has recovered soundness, and the control mode is shifted from the distributed control mode to the centralized control mode (S3).
Here, the control device 2 may immediately shift to the centralized control mode when the control amount data can be received within the first predetermined time T1. Alternatively, the control device 2 continues the distributed control mode until new control amount data is received, and the difference Te between the time when the new control amount is received and the time when the previous control amount is received is When Te <T1, the distributed control mode may be shifted to the centralized control mode.

When the time during which the communication state has deteriorated continues for a long time and the elapsed time Ts after the transition to the distributed control mode becomes equal to or longer than the effective period T2 of the power state estimation model (Ts ≧ T2), the control device 2 The control mode is shifted from the distributed control mode to the autonomous control mode (S4).

By using the distributed control mode, the control device 2 can relatively accurately perform control for stabilizing the voltage of the entire power system. However, since the power supply and demand state varies depending on time and season, it is not desirable to continue using the same calculation model for a long period of time. In addition, for example, the configuration of the power system may be changed when a new distributed power source is connected to the power system or the customer's equipment is discarded. The change in the configuration of the power system also affects the estimation accuracy of the power state, and the appropriate control amount changes accordingly. Continuing to use an arithmetic model that is far from the actual state of power supply and demand in this way reduces the estimation accuracy of the power state and does not contribute to maintaining the stability of the power system. Therefore, a period T2 that can be used effectively is set in advance in the calculation model of the present embodiment.
Here, the effective period T2 is set longer than the predetermined time T1 (T2> T1 ≧ Tc).
The control device 2 under the autonomous control mode performs control so as to eliminate the deviation between the target voltage of the local node and the measured voltage based on the measurement data from the sensor 3 of the local node.

When the control amount data reception cycle from the central device 1 is within the first predetermined cycle T1 during the control in the autonomous control mode (Td <T1), the control device 2 switches from the autonomous control mode to the centralized control mode. (S5). At this time, the control device 2 may immediately shift to the centralized control mode. Alternatively, the control device 2 waits until a new control amount is received, and when the difference Te between the previous control amount reception time and the current control amount reception time is Te <T1, the autonomous control mode changes to the centralized control mode. The control mode may be shifted to.

An example is given to understand the transition timing of the control mode. When the communication state is healthy, the central device 1 transmits a control amount (control amount data) calculated based on each measurement data to each control device 2 at a control cycle Tc such as several seconds to several minutes. At this time, the central apparatus 1 has previously transmitted a calculation model including a power state estimation model and a control amount calculation model to the control apparatus 2. If the communication state deteriorates and control amount data cannot be received even after waiting for the predetermined time T1 set to several minutes, the control device 2 shifts from the central control mode to the distributed control mode. That is, the control device 2 stops using the control amount data sent from the central device 1 and performs control using the arithmetic model sent together with the control amount data. Incidentally, the control amount data is continuously transmitted from the central device 1 to the control device 2 even in the distributed control mode and the autonomous control mode.

For example, when a predetermined time T2 such as one day to several days, one week to several weeks, or one month to several months has elapsed since the transition to the distributed control mode, the control device 2 changes from the distributed control mode to the autonomous control mode. Transition. The valid period T2 of the power state estimation model may be defined as an elapsed time since the creation, or may be defined by a date and time, for example, “valid until September 30, 2012”. These specific numerical values are merely examples for understanding and are naturally not limited to the above numerical values.

<Flowchart>
Next, a processing procedure in the power system control system 5 will be described with reference to the flowcharts of FIGS. 10, 11, and 14.
(Central device: Control amount calculation processing)
FIG. 10 is a flowchart illustrating a control amount calculation process in the central apparatus according to the first embodiment.
The process shown in FIG. 10 is a process for calculating the control amount of the control device 2 used in the centralized control mode.
The communication state monitoring unit 101 accumulates measurement data (with a time stamp) sent from each sensor 3 at a predetermined constant monitoring cycle Tm (S101).
Next, the communication state monitoring unit 101 determines the soundness (normal or not) of the communication state with the central apparatus 1 for each sensor 3 (S102). Measurement data used for power state estimation changes depending on the determination here. That is, the power state estimation unit 102 does not use the measurement data determined to be abnormal in step S102 for power state estimation.

For example, when the measurement data from the sensor 3 is to be collected at a predetermined time period Tm, the communication state monitoring unit 101 determines the difference between the time stamp of the measurement data received last time and the current time as the predetermined time period. Compare with Tm. When the time period Tm is exceeded ((current time−time stamp of latest measurement data)> Tm), the communication state monitoring unit 101 determines that the soundness of the communication state is deteriorated. It can be determined that the soundness of the communication state decreases as the difference between the current time and the time stamp of the latest measurement data increases.

Here, the cause of the deterioration of the soundness of the communication state may be the cause of the communication network 4 itself or the cause of the sensor 3 itself. Possible causes of the communication network 4 include, for example, communication congestion, radio wave interference due to electromagnetic waves from obstacles and electronic devices, disconnection, and the like. Possible causes of the sensor 3 itself include, for example, a failure of the sensor 3 and a temporary suspension of processing due to overload. Therefore, by determining the soundness of the communication state between the sensor 3 and the central device 1, it is possible to determine whether or not the sensor 3 is operating normally.

Then, the power state estimation unit 102 performs observability determination in the power state estimation according to the number of measurement data determined to have a normal communication state in step S102 (hereinafter, normal measurement data number). It is determined whether or not observation is possible (S103). “Observable” means that the number of measurement data that can be acquired is sufficient to estimate the power state, and is determined by the following method, for example. First, the power state estimation unit 102 sets the total number of power states (for example, active power P, reactive power Q, voltage V) of nodes and branches of the target power system as Nd. And the electric power state estimation part 102 calculates the ratio (Nn / Nd) of the normal measurement data number Nn with respect to the total number Nd. Subsequently, the power state estimation unit 102 determines that observation is possible if the ratio of Nn to Nd (Nn / Nd) is equal to or greater than a predetermined value, and otherwise determines that observation is impossible.

As a result of step S103, when it is determined that observation is possible (S103 → Yes), the power state estimation unit 102 calculates the power flow at each monitoring node using the measurement data and system data of the sensor 3 determined to be normal. Is estimated (S104), and the process proceeds to Step S106. Here, the monitoring node is a node whose voltage should be monitored, and is set by the user. If the user does not set a monitoring node, all nodes become monitoring nodes.

The power state estimation unit 102 estimates the power state in the following procedure. First, an initial value for the power state of each node is set via the input unit. And the electric power state estimation part 102 performs the tidal current calculation based on the set initial value. Subsequently, the power state estimation unit 102 calculates the sum of the squares of the estimated values related to the power states (active power P, reactive power Q, estimated voltage V) obtained by the power flow calculation and the deviations of the measured values included in the measurement data. A solution regarding the power state of each node is obtained by repeated calculation so as to be minimized. In this way, an estimated power state value at an arbitrary point in the power system is finally obtained.
The power state estimation unit 102 adds a time stamp to the obtained power state estimation value and stores it in the node management data (FIG. 4) of the system data storage unit 112 as measurement data. Such calculation of the power state estimation value is a known method.

As a result of step S103, when it is determined that it is not observable (unobservable) (S103 → No), the power state estimation unit 102 uses the measurement data and system data of the sensor 3 determined to be normal to calculate the power flow. (S105), and the process proceeds to step S106.
For each node and each branch, the power state estimation unit 102 establishes equations (power equations) relating to active power, reactive power, and estimated voltage, and solves them using measurement data, whereby the power state of each monitoring node and each branch The power flow calculation for obtaining (active power P, reactive power Q, estimated voltage V) is performed. Then, the power state estimation unit 102 uses the power state obtained by the power flow calculation as a power state estimation value, and stores this power state estimation value as measurement data in the node management data (FIG. 4) of the system data storage unit 112. .
As described above, the power state estimation unit 102 obtains the power state estimation value at an arbitrary point of the power system through the processes of steps S104 and S105.

Next, the control amount calculation processing unit 103 calculates the control amount that each control device 2 should output using the power state estimation value calculated in steps S104 and S105 (S106).
For example, the control amount calculation processing unit 103 calculates the control amount by the following procedure. Here, for example, the sum of the squares of the deviations of the estimated voltage from each target voltage at a plurality of predetermined points on the power system is used as the objective function. Here, the estimated voltage is a voltage calculated in steps S104 and S105 in FIG. 10, and the target voltage is a reference voltage (see FIG. 5) to be output in the control device 2. The control amount calculation processing unit 103 calculates the optimal control amount of each control device 2 so as to minimize this objective function. Thereby, the control amount calculation processing unit 103 calculates an optimal control amount for stabilizing the voltage of the entire power system.

Here, the control amount is a tap number (transformation ratio) of a voltage regulator such as LRT or SVR, SVC, a switched capacitor, reactive power of a PCS with a battery, or the like.
There are various methods for minimizing the objective function, such as hill climbing, quadratic programming, and tabu search. Which solution method to use is determined by the user according to the nature of the objective function and the nature of the controlled variable (continuous value, discrete value).
The calculated control amount may be stored in the system data storage unit 112 or the like as a control amount history.

Then, the transmission / reception processing unit 105 transmits the calculated control amount of each control device 2 as control amount data to each control device 2 via the communication network 4 (S107).

10, the central device 1 determines the soundness of the measurement data from each sensor 3 and performs power state estimation according to the soundness. Furthermore, the central apparatus 1 calculates the control amount of each control apparatus 2 based on the calculated power state estimation value, and transmits it.

(Central unit: calculation model calculation process)
FIG. 11 is a flowchart illustrating the procedure of the calculation model calculation process in the central apparatus according to the first embodiment.
The process shown in FIG. 10 is a process for calculating a control amount used in the centralized control mode, whereas the process shown in FIG. 11 is a process for calculating an arithmetic model used in the distributed control mode. . The central controller 2 executes the process shown in FIG. 10 and the process shown in FIG. 11 in parallel.
First, the model generation unit 104 uses the measurement values (FIG. 8) to measure the power state measured values (active power P, reactive power Q) by each sensor 3, and the power state estimation unit 102 in FIG. Various power state estimation values (active power P, reactive power Q, estimated voltage V) of each node and each branch estimated in S104 and S105, and control amounts of each control device 2 calculated in step S106 of FIG. Data is read for use as teacher data (S201).

Next, the model generation unit 104 uses the power state measurement value actually measured by the sensor 3 and the power state estimation value of each node and each branch estimated by the power state estimation unit 102 in the process illustrated in FIG. Then, a power state estimation model is generated (S202). The generation of the power state estimation model will be described later.

Subsequently, the model generation unit 104 controls the power state estimation value of each node and each branch estimated by the power state estimation unit 102 in FIG. 10 and the control of each control device 2 calculated by the control amount calculation processing unit 103 in FIG. The control amount calculation model is generated using the amount (S203). The generation of the control amount calculation model will be described later.

Then, the model generation unit 104 packages the power state estimation model and control amount calculation model data (coefficient parameters) generated for each control device 2 in a predetermined combination. For example, with respect to a certain control device 2 (node i), the power state estimation model is a combination of a measurement value of each node x and node y as an input, and a control amount calculation model is a combination of a local node (node i) Packaging is performed.

Next, the transmission / reception processing unit 105 transmits packaged data (model data) of each operation model to each control device 2 via the communication network 4 (S204).

11, the central device 1 generates a calculation model such as a power state estimation model and a control amount calculation model from the measurement data, and transmits the generated calculation model to each control device 2.

(Generation of calculation model)
Next, a method for generating a calculation model such as a power state estimation model and a control amount calculation model will be described.
The generation of the arithmetic model can use a linear model or a nonlinear model. Here, a linear model will be described as an example.
First, generation of a power state estimation model will be described.
The input (explanatory variable) of the power state estimation model is an actual measurement value (active power P, reactive power Q) of the power state measured by each sensor 3. The output from the power state estimation model is a power state estimation value of each node (or each branch) in the power system estimated by the power state estimation unit 102.

The model generation unit 104 identifies the coefficients a _nk and b _nk in the linear model represented by the following formula (1). These coefficients a _nk and b _nk become a power state estimation model.

Here, S _n represents the complex power at any node n. j represents an imaginary unit. P _k and Q _k indicate active power (real component of complex power) and reactive power (imaginary component of complex power) measured by the sensor 3 (sensor node k), respectively. a _nk and b _nk are coefficient parameters of the power state estimation model, and indicate the influence of the sensor node k at the node n.
The model generation unit 104 substitutes the input data (P _k , Q _k ) and the output data (S _n ) at each node into the equation (1), and uses the least square method or the like to calculate the coefficients (a _nk , b _nk). ) Is identified.

In the control device 2, when the basic calculation formula (formula (1)) is set in advance, the central device 1 simply transmits the coefficient parameters a _nk and b _nk to the control device 2, and the power state estimation model. Can be updated. That is, the amount of information can be reduced by using the coefficient in Equation (1) as an arithmetic model.

Next, generation of a control amount calculation model will be described.
The input (explanatory variable) of the control amount calculation model is the power state estimation value of each node estimated by the power state estimation unit 102 in steps S104 and S105 of FIG. The output (objective variable) of the control amount calculation model is the control amount of each control device 2 calculated by the control amount calculation processing unit 103 in step S106 in FIG. 10, and is used to stabilize the voltage of the entire power system. The control amount is close to the optimum (sub-optimal).

The model generation unit 104 identifies coefficients c _im and d _im in the linear model represented by the following formula (2). The coefficients c _im and d _im serve as a control amount calculation model.

Here, C _i is the control amount (complex power) at the node i calculated by the control amount calculation processing unit 103 in the processing of FIG. j represents an imaginary unit. P _m and Q _m are power state estimation values (active power P _m and reactive power Q _m ) of each node m (monitoring node m) estimated by the power state estimation unit 102 in the process of FIG. c _im and d _im are coefficient parameters of the control amount calculation model, and indicate the influence of the node m on the node i.
As for P _m and Q _m , if measured values (power state values P and Q of measurement data) measured by the sensor 3 are available, they may be used.
The model generation unit 104 substitutes the input data (power state estimation values P _m , Q _m ) and output data (C _i ) at each node into the equation (2), and uses each of the coefficients (c _im , d _im ).

As described above, the control amount Ci is a tap number (transformation ratio) of a voltage regulator such as LRT or SVR, SVC, a switched capacitor, reactive power of a battery-equipped PCS, or the like.

When the basic calculation formula (formula (2)) is set in advance in the control device 2, the central device 1 simply transmits the coefficient parameters c _im and d _im to the control device 2 to control the control amount calculation model. Can be updated. That is, the amount of information can be reduced by using the coefficient in Equation (2) as an arithmetic model.

As described above, if the power state estimation model, the control amount calculation model, and the measurement data (P _k , Q _k ) are used, each control device 2 can stabilize the voltage of the entire power system. The control amount to be output can be easily calculated.
Therefore, if each control device 2 has an operation model such as a power state estimation model and a control amount calculation model, even if the communication state of the communication network 4 between the central device 1 and the control device 2 is not good, Appropriate control can be performed using the measurement data at its own end.

(Model data)
Next, model data will be described.
The model data includes parameters a _nk and b _nk constituting the power state estimation model shown in Expression (1), and parameters c _im and d _im constituting the control amount calculation model shown in Expression (2). Expressions (1) and (2) are examples and may be changed as appropriate, and the parameters in that case are coefficient parameters according to the expressions.

It is necessary to previously set target nodes (sensor node k and monitoring node m) in each calculation model.
If all the nodes where the sensor 3 is installed are set as the sensor node k, the accuracy of the power state estimated value can be improved. Similarly, if all the nodes are set as the monitoring node m, it can be expected to improve the accuracy of the control amount (close to the optimum control amount).

However, if the communication state of the communication network 4 cannot always be expected to be healthy, there is a possibility that measurement data cannot be acquired from all the sensors 3. Therefore, as described below, the estimation may be performed using only the measurement data of some of the sensors 3.

For example, instead of using all the nodes as sensor nodes, the node i of the control device 2 may include only the local node i. That is, each control device 2 may not refer to the measurement data of the sensors 3 other than its own end node. The sensor 3 of the local node is a sensor 3 that is directly associated with the control device 2, that is, a sensor 3 that is provided at a node common to the node of the control device 2. In other words, the sensor 3 of the local node is the sensor 3 installed in the control device 2.

In the case of the power system example shown in FIG. 3, the local node of the control device 2a is the node 31a, and the sensor 3 of the local node is the sensor 3a. Similarly, the own end node of the control device 2b is the node 31d, and the sensor 3 of the own end node is the sensor 3c. Similarly, the own end node of the control device 2c is the node 31g, and the sensor 3 of the own end node is the sensor 3e.

If the measurement data referred to by the control device 2 is limited to the measurement data from the sensor 3 of the local node, the control device 2 uses the power state estimation model even if the soundness of the communication state is lost. Can be estimated.

Note that there is no particular reason for reducing the target node for the monitoring node m, unlike the sensor node. That is, all nodes may be monitoring nodes. This is because if the power state is estimated, the control amount can be calculated using the control amount calculation model. That is, even if the measurement data is missing, it is possible to estimate the power state itself, so it is not necessary to consider the missing node m in the generation of the control amount calculation model.

Next, the specific meaning of the calculation model will be described with reference to FIG.
FIG. 12 is a diagram for explaining the specific meaning of the calculation model according to the first embodiment, where (a) shows a power state estimation model and (b) shows a control amount calculation model.
In the power state estimation model indicated by the thick solid line 1201 in FIG. 12A, the explanatory variable on the horizontal axis is a measurement value included in the measurement data, and the output on the vertical axis is the power state estimation value. That is, the horizontal axis is a measurement value actually measured by the sensor 3 accumulated in step S101 in FIG. 10, and the vertical axis is a power state estimation value in each node calculated in steps S104 and S105 in FIG. In short, in the graph of FIG. 12 (a), the horizontal axis represents the _P k, _{Q k} of the formula (1), the vertical axis represents _{S n} of the formula (1).
Accordingly, the horizontal axis should be originally two-dimensional coordinates, but here, it is assumed as one-dimensional coordinates for convenience. Also, the plot point 1202 is a point obtained by plotting in association actual measured value P _k, and Q _k, the power state estimate S _n which is estimated based on the actual measurement values. A solid line 1201 is a straight line calculated by the least square method based on the plot point 1202. The coefficients a _nk and b _{nk of} Equation (1), that is, the power state estimation model corresponds to the slope of the solid line 1201.

In the control amount calculation model indicated by the thick solid line 1211 in FIG. 12B, the horizontal axis input is the power state estimation value, and the vertical axis output is the calculated control amount. That is, the horizontal axis is the estimated power state value at each node calculated in steps S104 and S105 in FIG. 10, and the vertical axis is the control amount at each node calculated in step S106 in FIG. In short, in the graph of FIG. 12B, the horizontal axis indicates P _m and Q _m in the equation (2), and the vertical axis indicates C _i in the equation (2).
Accordingly, the horizontal axis should be originally two-dimensional coordinates, but here, it is assumed as one-dimensional coordinates for convenience. The plot point 1212 is a point plotted by associating the power state estimated values P _m and Q _m with the calculated control amount C _i . A solid line 1212 is a straight line calculated by the least square method based on the plot point 1211. The coefficients c _im and d _im in Equation (2), that is, the power state estimation model corresponds to the slope of the solid line 1211.

Each of these calculation models shows a case in which the input on the horizontal axis is two variables (as described above, it is assumed to be one-dimensional for convenience in FIG. 12), but the present invention is not necessarily limited to this. There may be input. That is, in this embodiment, the active power P and the reactive power Q are used as explanatory variables for obtaining each calculation model, but other power state values (for example, voltages) may be used as explanatory variables.
However, from the viewpoint of accuracy, it is desirable that the explanatory variable has a high correlation with the objective variable (power state, control amount) to be estimated.

FIG. 13 is a diagram illustrating a specific example of model data according to the first embodiment. FIG. 13 shows an example of the power state estimation model.
In FIG. 13, a calculation model for each case is stored for each control device 2. In FIG. 13, the model data 1301 is for the control device 2a (FIG. 1), the model data 1302 is for the control device 2b (FIG. 1), and the model data 1303 is for the control device 2c (FIG. 1). Hereinafter, description will be given with reference to the power state estimation model in the model data 1301 of the control device 2a.
Here, the case indicates a missing pattern of measurement data. If the power system includes the sensors 3a to 3e, for example, “Case 1” is a case where measurement data can be normally acquired from all the sensors 3a to 3e. In other words, in “Case 1”, the power state is estimated by the calculation model using the measurement data from all the sensors 3a to 3e. In other words, “Case 1” is a power state estimation model used when the control device 2 a can acquire measurement data from all the sensors 3. Note that “a1a” indicates a power state estimation model “a _nk (formula (1))” indicating the influence from the sensor 3a in the state of “case 1”.

Also, “Case 2” is a case in which measurement data can be acquired from the sensors 3a to 3d among the sensors 3a to 3e, and only the measurement data from the sensor 3e cannot be acquired. Here, the coefficient parameter when the measurement data cannot be acquired is set to 0, and otherwise, the coefficient value is stored. In “Case 2”, the power state is estimated by the calculation model based only on the measurement data of the sensors 3a to 3d. In other words, “Case 2” is a power state estimation model used when the control device 2a cannot acquire measurement data only from the sensor 3e.

“Case 3” is a case where, among the sensors 3a to 3e, measurement data can be acquired from the sensors 3a to 3c and 3e, and only measurement data from the sensor 3d cannot be acquired. That is, in “Case 3”, the power state estimation model is generated based only on the measurement data of the sensors 3a to 3c, 3e. In short, “Case 3” is a power state estimation model used when the control device 2a cannot acquire measurement data only from the sensor 3d.

Similarly, the central apparatus 1 generates a power state estimation model for all combinations of measurement data that can be normally acquired in step S202 of FIG.
Here, the power state estimation model is also generated when the measurement data can be acquired only from any one of the sensors 3 (case n). Moreover, if the measurement data from the sensor 3 of the end node of the control device 2 can be acquired, a power state estimation model for the control device 2 can be generated. In the case where the control device 2 is configured to be able to acquire measurement data from the sensor 3 of a remote node other than the sensor 3 of the own node, the power state estimation based only on the measurement data from the sensor 3 of the remote node A model is also generated.
Each case is referred to as “model 1”, “model 2”,.
Such a power state estimation model for each case is generated for each control device 2.

Although the power state estimation model has been described with reference to FIG. 11, the central apparatus 1 also includes the control amount calculation model (c _im , d _im (Expression (2))) as well as the model data together with the power state estimation model illustrated in FIG. 11. To the control device 2. Since the control variable calculation model uses the explanatory variable as the power state estimation value, it is not necessary to obtain a control variable calculation model for each case as shown in FIG. That is, the estimation of the power state is performed for all the nodes set as the monitoring nodes, so that the power state estimation value is not lost.

Thus, a power state estimation model is prepared for each control device 2 and for each combination of measurement data missing states. Thereby, even when a part of measurement data cannot be acquired due to deterioration of the communication state or failure of the sensor 3, the power state estimation unit 102 can generate the power state estimation model corresponding to the acquired combination of measurement data. Can be used to calculate the power state estimate.

When generating the operation model as a linear model, the power state estimation model has two parameters a _nk and b _nk (hereinafter, appropriately described as a and b), and the control amount calculation model also has two parameters c _im and d _im ( Hereinafter, these are described as c and d as appropriate. Therefore, the data size of each calculation model can be reduced. Thus, according to this embodiment, since the size of each calculation model can be reduced, an increase in communication load can be suppressed even if a plurality of model data is periodically transmitted to each control device 2. Therefore, even when the communication speed of the communication network 4 is low, model data having a small data size can be transmitted normally. Moreover, since each calculation model can be used with a small number of parameters, even when the performance of the CPU 201 (FIG. 2) of the control device 2 is low, an appropriate control amount can be obtained by estimating the power state. In other words, it is not necessary to mount a high-performance CPU 201 or the like in the control device 2, and the manufacturing cost can be reduced.

(Processing in the control device)
FIG. 14 is a flowchart illustrating a processing procedure of the control device according to the first embodiment.
First, the control mode determination processing unit 202 determines whether data is received from the central device 1 or the sensor 3 (S301).
As a result of step S301, when data is not received (S301 → No), the control mode determination processing unit 202 returns the process to step S301 and waits for data reception.
When data is received as a result of step S301 (S301 → Yes), the control mode determination processing unit 202 determines whether or not the received data is control amount data (S302).
If the result of step S302 is not control amount data (S302 → No), the control mode determination processing unit 202 stores data in each data storage unit according to the data type (S303), and returns the process to step S301. For example, if the received data is model data, the control mode determination processing unit 202 stores the data in the model data storage unit 212. If the received data is system data, the control mode determination processing unit 202 stores the system data. Stored in the unit 211.

If the result of step S302 is control amount data (S302 → Yes), the time stamp attached to the control amount data is referred to, and the time Td between the time stamp of the control amount data received last time and the current time Td Is calculated, and it is determined whether or not this Td is smaller than the preset time T1 (Td <T1) (S304).

As a result of step S304, when Td <T1 (S304 → Yes), that is, when the control amount data has reached within a predetermined time, the control mode determination processing unit 202 sets the control mode to the centralized control mode ( S305). If the control mode is already the centralized control mode, the control mode determination processing unit 202 maintains the current control mode.
The control mode determination processing unit 202 temporarily stores the time stamp of the control amount data acquired this time in the memory (S306), and transmits the control amount in the control amount data received by the control processing unit 204 to the measurement device. Thus, the measurement apparatus is controlled (S313).

On the other hand, if Td ≧ T1 as a result of step S304 (S304 → No), the control mode determination processing unit 202 determines that the time Ts after shifting to the distributed control mode is smaller than the predetermined time T2 (Ts <T1). ) Is determined (S307). Note that when the current control mode is not the distributed control mode, Ts = 0.

As a result of step S307, if Ts <T2 (S307 → Yes), that is, if within a predetermined time after shifting to the distributed control mode, the control mode determination processing unit 202 sets the control mode to the distributed control mode. (S308). If the control mode is already the distributed control mode, the control mode determination processing unit 202 maintains the current control mode.
Then, the control amount calculation processing unit 203 reads measurement data from the measurement data storage unit 213 (S309).

Subsequently, the control amount calculation processing unit 203 reads the calculation model from the model data storage unit 212 (S310), performs the calculation of Expression (1) from the electric energy state estimation model and the measurement data in the read calculation model, The power state is estimated (S311). Specifically, the control amount calculation processing unit 203 estimates active power and reactive power in its own control device 2. In step S310, the control amount calculation processing unit 203 may read an operation model corresponding to the missing state from the model data illustrated in FIG. 13 according to the missing state of the measurement data.

Further, the control amount calculation processing unit 203 calculates the control amount of itself by performing the calculation of Expression (2) from the control amount calculation model of the calculation model read in step S309 and the power state estimated in step S311. (S312).
Then, the control processing unit 204 controls itself with the control amount calculated in step S312 (S313).

If Ts ≧ T2 as a result of step S307 (S307h → No), that is, if it is a predetermined time or longer after shifting to the distributed control mode, the control mode determination processing unit 202 sets the control mode to the autonomous control mode. (S314). If the control mode is already the autonomous control mode, the control mode determination processing unit 202 maintains the current control mode.
Then, the control amount calculation processing unit 203 reads the system data from the system data storage unit 211 (S315), and further reads the measurement data from the measurement data storage unit 213 (S316).
Then, the control amount calculation processing unit 203 calculates its own control amount from the read system data and measurement data (S312). Since the control amount calculation method in the autonomous control mode has been described above, detailed description thereof is omitted here.
Then, the control processing unit 204 controls itself with the control amount calculated in step S312 (S313).
The system data may or may not have a time stamp.

<Calculation model switching by time zone>
FIG. 15 is a diagram for explaining switching of the calculation model according to the time zone.
The power supply / demand condition varies depending on the weather, temperature, etc., but also varies depending on the time of day such as morning, noon, and night. For example, in the case of a private house, electric power demand increases due to preparation of meals in the morning and evening. There is a high possibility of being absent during the daytime, so power demand will decrease. On the contrary, the amount of solar power generation in the morning and evening is small, and the amount of power generation in the daytime is large. For this reason, in a house where solar power generation is installed, power consumption from the power plant is small during the daytime and large during the morning and evening. As described above, the supply and demand state of electric power varies depending on the characteristics of consumers such as private houses, factories, and commercial facilities, but also varies greatly depending on time zones.
Therefore, the power system control system 5 according to the present embodiment has a plurality of times such as a first time zone (morning / evening), a second time zone (daytime), and a third time zone (night), for example. A power state estimation model group and a control amount calculation model are generated for each segment and time zone.

As shown in FIG. 13, a plurality of power state estimation models of each control device 2 are generated according to combinations of acquired measurement data.
Here, as shown in FIG. 15, the central device 1 generates a calculation model for each control device 2 and for each time zone (that is, according to a predetermined living condition). In FIG. 15, the time zone is divided into a first time zone (for example, morning and evening), a second time zone (for example, noon), and a third time zone (for example, night), but is not limited thereto. For example, it may be divided every hour. Alternatively, it may be divided according to the weather. In this case, the control amount calculation processing unit 203 may acquire the weather via the Web and use a calculation model according to the weather.
For example, in the first time zone, a power state estimation model group 1501 and a control amount calculation model 1502a are generated as calculation models for the control device 2a (FIG. 1). Here, the power state estimation model “group” is used because different power state estimation models are generated for each case even in the same control device 2 as illustrated in FIG. 13.

Similarly, a power state estimation model group 1501b and a control amount calculation model 1502b are generated as calculation models for the control device 2b (FIG. 1). In addition, as an arithmetic model for the control device 2c (FIG. 1), a power state estimation model group 1501c and a control amount calculation model 1502c are generated.

Similarly, in the second time zone, a power state estimation model group 1511a (for the control device 2a), 1511b (for the control device 2b), 1511c (for the control device 2c), a control amount calculation model 1512a (for the control device 2a), 1512b (for the control device 2b) and 1512c (for the control device 2c) are generated.
Further, in the third time zone, power state estimation model groups 1521a (for the control device 2a), 1521b (for the control device 2b), 1521c (for the control device 2c), control amount calculation models 1522a (for the control device 2a), 1522b (For the control device 2b) and 1522c (for the control device 2c) are generated.
Incidentally, each of the power state estimation model groups 1501a to 1501c, 1511a to 1511c, and 1521a to 1521c corresponds to the power state estimation model group shown in FIG.

When the time zone is changed, the control amount calculation processing unit 203 of the control device 2 switches to the power state estimation model corresponding to the time zone, and further calculates a sub-optimal control amount based on the control amount calculation model.

The control device 2 according to the first embodiment can select any one control mode from a centralized control mode, a distributed control mode, and an autonomous control mode according to a communication state from among a plurality of control modes. Therefore, the control device 2 can appropriately operate according to the state of the communication network 4.

Furthermore, the control device 2 according to the first embodiment performs control in the centralized control mode when the communication state is healthy, and performs control in the distributed control mode when the communication state deteriorates, and the validity period of the power state estimation model has passed. If it does, control in autonomous control mode. Therefore, the power state can be controlled according to the degree of soundness of the communication state.

And the central apparatus 1 of 1st Embodiment accumulate | stores and manages the measurement data which are the actual measurement values of the sensor 3, the log | history of an estimated power state value, and the log | history of the calculated control amount as measurement data, Calculation models such as a power state estimation model and a control amount calculation model are generated based on the measurement data. Accordingly, it is possible to relatively easily obtain a relatively high accuracy calculation model.

In addition, the central device 1 of the first embodiment calculates the calculation model as a coefficient of a measurement value acquired from the sensor 3 or a power state estimation value. Therefore, the size of the model data that is the data of the operation model can be reduced, and the load on the communication network 4 can be suppressed even when the model data is distributed from the central device 1 to the plurality of control devices 2. For this reason, the power system control system 5 of the present embodiment can appropriately control the power system even in an environment where the communication speed is low and the communication quality is unstable.

In the power system control system 5 according to the first embodiment, the effective period is set in the power state estimation model used in the distributed control mode, and the power system control system 5 has the power state estimation model that has passed the effective period. Switch to autonomous control mode without using. Therefore, according to the first embodiment, it is possible to suppress the execution of the distributed control far from the change in the power supply / demand state and the change in the configuration of the power system, thereby improving the reliability of the system.

The central apparatus 1 according to the first embodiment generates a power state estimation model according to a combination of measurement data that can be normally acquired (FIG. 13). For this reason, even if the state of the communication network 4 is poor and the control device 2 cannot receive the measurement data from all the sensors 3, the control amount is estimated so as to stabilize the voltage of the entire power system as much as possible by estimating the power state. Can be calculated. Therefore, the power system control system 5 according to the first embodiment can estimate the power state according to the communication environment even in an area where the communication environment is bad, and can appropriately control the power state of the power system.

Further, the central device 1 according to the first embodiment generates a power state estimation model according to a predetermined living condition such as a time zone and weather (FIG. 15). Thereby, it can respond to the power consumption which fluctuates according to living conditions.

[Second Embodiment]
Next, a second embodiment of the present invention will be described with reference to FIGS.
The second embodiment is intended to perform appropriate control according to the reliability of the measurement data by providing a confidence interval for the measurement data.
In the second embodiment, the power state estimation model is described as an example of the calculation model. However, the present invention is not limited to this, and the control amount calculation model can be similarly handled.

<Configuration of central unit>
FIG. 16 is a diagram illustrating a configuration example of the central device according to the second embodiment.
In the central device 1A in FIG. 16, the same components as those in the central device 1 in FIG. 1 are denoted by the same reference numerals as those in FIG. In addition, since the structure of the electric power system control system 5 whole in 2nd Embodiment is the same as that of 1st Embodiment, illustration and description are abbreviate | omitted.
The central apparatus 1A in FIG. 16 is different from the central apparatus 1 according to the first embodiment in that a confidence interval processing unit 106 is added and a confidence interval data storage unit 114 is added.
The confidence interval processing unit 106 is realized by developing a program stored in the HD 304 or the like in FIG. 2A in the RAM 302 and executing it by the CPU 301.

The reliability processing unit 205 outputs data that takes into account the confidence interval for the measurement data. The operation of the confidence interval processing unit 106 will be described later.
The confidence interval data storage unit 114 stores confidence interval data that is data relating to a confidence interval described later.

<Control device configuration>
FIG. 17 is a diagram illustrating a configuration example of a control device according to the second embodiment.
In FIG. 17, the same components as those of the control device 2 of FIG. 1 are denoted by the same reference numerals as those of FIG.
The controller 200a in the control device 2A in FIG. 17 differs from the measurement device according to the first embodiment in that a reliability processing unit 205 is added and a confidence interval data storage unit 214 is added. Further, the control amount calculation processing unit 203a is different from the first embodiment in that a temporary control amount is calculated when low-reliable measurement data exists.
The control processing unit 204a is different from the first embodiment in that the control processing unit 204a performs control by the temporary control amount when there is low-reliability measurement data.
The control amount calculation processing unit 203a, the control processing unit 204a, and the reliability processing unit 205 are realized by the CPU 401 executing a program stored in the ROM 402 in FIG.
The confidence interval data storage unit 214 stores confidence interval data (reliability information: interval information) that is data relating to a confidence interval described later.

<Confidence interval>
FIG. 18 is a diagram illustrating an example of a confidence interval according to the second embodiment, in which (a) illustrates a confidence interval for active power and (b) illustrates a confidence interval for reactive power.
In an actual power system, the power state value fluctuates due to a temporal load fluctuation (disturbance) or the like. And it is thought that the fluctuation | variation of such electric power state values (active power, reactive power, voltage) has a bias in a certain range.
When such fluctuations occur, as shown in the first embodiment, the calculation model is not calculated using all the obtainable data, but the data within the defined confidence interval is used. With respect to data outside the confidence interval, improvement of the stability of the calculation model can be expected by performing a predetermined correction.

First, the confidence interval of active power will be described with reference to FIG.
Here, it is assumed that the bias of the power state value follows a normal distribution.
Assuming that the variation in active power follows a normal distribution, assuming that the average value is P _M and the standard deviation is σ _p , the normal distribution of active power is a distribution as shown in FIG.
Then, a highly reliable interval (first interval) 1801 is defined as in the following expression (3).

P _M −α · σ _p ≦ P ≦ P _M + α · σ _p (3)

The high confidence interval 1801 represented by the expression (3) corresponds to the high confidence interval in FIG. Other than the high confidence interval 1801 is a low confidence interval (second interval) 1802.
Note that α shown in Equation (3) is a coefficient for setting the high confidence interval 1801. For example, if a confidence interval is set with α = 1.64, the probability that an active power value will occur in the high confidence interval 1801 is 0.9. Further, for example, if the highly reliable interval 1801 is set with α = 1.96, the probability that the value of reactive power will occur in the highly reliable interval 1801 is 0.95.

Similarly, a high-reliability interval 1801 for reactive power is also set (see FIG. 18B).
In addition, the highly reliable section 1801 of reactive power is prescribed | regulated by the following formula | equation (4). Other than the high confidence interval 1801, the low confidence interval 1802 is provided.

Q _M −α · σ _q ≦ Q ≦ Q _M + α · σ _q (4)

Q _M and σ _q in equation (4) are the average value and standard deviation of the reactive power values, respectively, and α is the same coefficient as in equation (3).
The active power value, the average value of the reactive power value, and the standard deviation are the average value and the standard deviation of the active power value and the reactive power value within a predetermined time.

As described above, the confidence interval processing unit 106 calculates the average value and the standard deviation of each measurement data, and sets the high confidence interval 1801 defined by the equations (3) and (4) based on the average value and the standard deviation. Then, a highly reliable interval 1801 (upper limit value, lower limit value) of the measurement data is calculated.
The upper limit values are P _M + α · σ _p in equation (3) and Q _M + α · σ _q in equation (4). Similarly, the lower limit values are P _M −α · σ _p in Equation (3) and Q _M −α · σ _q in Equation (4).

For example, the confidence interval data shown in the format of FIG. 19 is stored in the confidence interval data storage unit 114.
As shown in FIG. 19, the confidence interval data stores the upper limit value and the lower limit value of the high confidence interval for each sensor 3 and each measurement data type. In the measurement data type, “P” indicates active power, and “Q” indicates reactive power. In addition, “90%”, “95%”, and “99%” are establishments in which a measurement value is generated within a defined high confidence interval.
Further, the transmission / reception processing unit 105 of the central device 1A transmits the high confidence interval data in the confidence interval data storage unit 114 to the control device 2A, and the transmission / reception processing unit 201 of the control device 2A transmits the received data related to the high reliability interval to the confidence interval. The data is stored in the data storage unit 114. This process is performed, for example, with the system data in steps S301 to S303 in FIG.

Note that the method of setting the high confidence interval and the low confidence interval as shown in FIG. 18 is an example. For example, it may be based on a beta distribution or a gamma distribution instead of the normal distribution.
Moreover, you may set different normal distribution for every time slot | zone shown in FIG.

(Control processing)
FIG. 20 is a flowchart illustrating a procedure of control processing according to the second embodiment.
Note that the processing in FIG. 20 is processing that is inserted into the locations of steps S311 to S313 of the distributed control processing in FIG.
First, the reliability processing unit 205 determines whether each of the measurement data (active power and reactive power) stored in the measurement data storage unit 213 is highly reliable (S401). In this process, the reliability processing unit 205 refers to the confidence interval data stored in the confidence interval data storage unit 214 and determines whether the reliability interval corresponds to the high reliability interval or the low reliability interval shown in FIG. Each measurement data is determined.

For example, the upper limit value and lower limit value of the active power, which are measurement data of the sensor 3 when the confidence interval is 90%, are Pau and Pal, respectively, according to the confidence interval data shown in FIG. Since the lower limit values are defined as Qau and Qal, respectively, the reliability processing unit 205 determines whether each of the active power and reactive power of the measurement data is within the range of the upper and lower limit values. The reliability processing unit 205 determines that the reliability of each measurement data is high when the value is within the range of the upper and lower limit values, and otherwise determines that the reliability is low. These measurement data are explanatory variables (input data) of the power state estimation model. However, if all the measurement data are determined to be highly reliable, the measurement data is determined to be high reliability as the explanatory variable of the calculation model. Otherwise, it is determined to be unreliable. That is, when there are a plurality of types of measurement data as explanatory variables, if at least one reliability is low, it is determined that the reliability is low as an explanatory variable of the calculation model.

As a result of step S401, when it is determined that all measurement data is highly reliable (S401 → Yes), the control amount calculation processing unit 203a performs the same processing as steps S311 to S313 in FIG. Take control.

As a result of step S401, when there is even one piece of measurement data determined to be unreliable (S401 → No), the control amount calculation processing unit 203a calculates a temporary control amount (S402). A method for calculating the temporary control amount will be described later.
Then, the control processing unit 204a controls the measuring device with the temporary control amount calculated in step S402 (S403).
Next, the control amount calculation processing unit 203a acquires the measurement data of the sensor 3 (own end node sensor) of the own end node via the transmission / reception processing unit 201 after a predetermined time from the control by the temporary control amount, and the sensor 3 point A target deviation index ΔV indicating how much the voltage of the current is deviating from the target is calculated (S404). The calculation method of the temporary control amount will be described later.

Then, the control processing unit 204a determines whether ΔV is equal to or less than a predetermined value (S405).
As a result of step S405, if ΔV is equal to or smaller than the predetermined value (S405 → Yes), the control processing unit 204a finishes the process because the control is close to the target voltage.
If ΔV is not less than or equal to the predetermined value as a result of step S405 (S405 → No), the control processing unit 204a calculates a corrected temporary control amount (S406), returns the process to step S403, and uses the corrected temporary control amount as the temporary control amount. Process. A method for calculating the corrected temporary control amount will be described later.

(Temporary control amount calculation method)
21 and 22 are diagrams for explaining a method for calculating a temporary control amount according to the second embodiment.
When there is low-reliability measurement data, the control amount calculation processing unit 203a calculates a temporary control amount as shown in step S402 in FIG. 21 and 22, the horizontal axis should originally be the two-dimensional coordinates of the active power P and the reactive power Q, but here, the one-dimensional coordinates are used for convenience.
For example, the following three methods can be considered as a method for calculating the temporary control amount.
In FIGS. 21 and 22, the high confidence interval is “xcl” to “xcu”, and the others are low confidence intervals. Here, xcl is the upper limit value of the high confidence interval, and xcu is the lower limit value of the high reliability interval. Hereinafter, the upper limit value and the lower limit value may be collectively referred to as a boundary value. “X1” is a measured value. X0 is the average value of the measured values. The slope of the straight line 2100 indicates the calculation model. In the description of FIGS. 21 and 22, f (•) represents an arithmetic expression of the control amount calculation model according to Expression (2).

In the first method, as shown in FIG. 21A, an operation model in a highly reliable interval is applied as it is.
That is, as shown in FIG. 21A, the temporary control amount at the measurement value x1 with low reliability is calculated using the calculation model in the high reliability interval. That is, the control processing unit 204a calculates the temporary control amount 2111 using the formula (2) itself.

In the second method, as shown in FIG. 21B, the control amount at the boundary values xcl and xcu in the high confidence interval closest to the measured value x1 is set as the temporary control amount 2112. That is, S = f (xcu or xcl) is set as the temporary control amount 2112.
In the example of FIG. 21B, since the boundary value of the high confidence interval closest to the measurement value x1 is “xcu”, S = f (xcu) is the temporary control amount 2112 calculated. When the boundary value of the high confidence interval closest to the measured value x1 is xcl, S = f (xcl) is a calculated temporary control amount.

In the third method, a temporary control amount is calculated based on at least one of a control amount lower limit 2201 and a control amount upper limit 2202 calculated by a predetermined method as shown in FIG.
Here, the control amount lower limit 2201 and the control amount upper limit 2202 are functions of the measurement value x, and are represented by, for example, Expression (5) and Expression (6).

S = c · (x−xd) ^k + f (xd) (x∈XL, x ≧ xd) (5)

S = c · (x−xd) ^k + f (xd) (x∈XL, x <xd) (6)

Here, Equation (5) is a case where the measured value x belongs to the low confidence interval XA shown in FIG. 22, and Equation (6) is a case where the measured value x belongs to the low confidence interval XB shown in FIG.
“Xd” is a boundary value (either xcu or xcl) close to the measurement value x1. c is a predetermined coefficient parameter. When c represents a control amount lower limit and “c” represents a control amount upper limit, “c” represents a positive real number having a relationship of “cu> cl”. It is. k is a predetermined “exponential parameter”. When the measured value x belongs to the low confidence interval and “x> xd”, “0 <k <1”, and “x <xd”. “K ≧ 1” is set. XL indicates a low confidence interval.

The graph of the control amount upper limit calculated by the equations (5) and (6) is denoted by reference numeral 2202 in FIG. 22, and the graph of the control amount lower limit is denoted by reference numeral 2201 in FIG.
The control amount calculation processing unit 203a calculates the control amount lower limit 2211 and the control amount upper limit 2212 by substituting the measured value x1 into the equations (5) and (6).
Then, the control amount calculation processing unit 203a calculates an average value 2213 of the control amount lower limit 2211 and the control amount upper limit 2212, and uses the average value 2213 as a temporary control amount. Alternatively, the control amount calculation processing unit 203a may use either the control amount upper limit 2211 or the control amount lower limit 2212 as a temporary control amount.

The calculation formulas for the control amount lower limit and the control amount upper limit are not necessarily limited to the formulas (5) and (6). For example, the measured value and the control amount history data in the low confidence interval are accumulated more than a predetermined amount. The equation calculated using the least square method may be used.

(Calculation of corrected temporary control amount)
Next, a method for calculating the corrected temporary control amount in step S406 in FIG. 20 will be described.
When the reliability of the measurement data is low, the control amount calculation processing unit 203a calculates the temporary control amount using the method described above. Then, as shown in step S403 of FIG. 20, the control processing unit 204a performs control according to the temporary control amount.
Then, the control amount calculation processing unit 203a acquires the measurement data of the sensor 3 of the local node via the transmission / reception processing unit after a predetermined time from the control by the temporary control amount, and the voltage at the point of this sensor 3 is preset. A target deviation index ΔV indicating how much deviation is from the target voltage is calculated according to the equation (7) (step S404 in FIG. 20).

ΔV = Vm−Vref (7)

Here, Vm is a measurement value (measurement voltage), and Vref is a target voltage.
Next, when the target deviation index ΔV is not less than or equal to the predetermined value (step S405 → No in FIG. 20), the control amount calculation processing unit 203a calculates the corrected temporary control amount Sa according to the equation (8) (step in FIG. 20). S406).

Sa = Kp · ΔV (8)

Here, Kp is a proportional gain. Equation (8) calculates the corrected temporary control amount Sa according to the proportional control, but the corrected control amount Sa may be calculated according to the proportional-integral control as shown in Equation (9).

Here, Ki is an integral gain, and T is a predetermined time. The proportional gain and the integral gain are determined according to the type of the control device 2A and the power system configuration (impedance, etc.), and are set in advance by sensitivity analysis or the like.

Then, the control processing unit 204a performs control again using the corrected temporary control amount Sa calculated in this way as a temporary control amount. Then, the control processing unit 204a obtains the target deviation index ΔV again after a predetermined time of control by the corrected temporary control amount Sa, and uses it to calculate the corrected control amount again.
By repeating the above processing until ΔV becomes a predetermined value or less, control close to the target voltage can be achieved.

In the second embodiment, when at least one piece of low-reliability measurement data exists, control is performed using a temporary control amount. However, the present invention is not limited to this. For example, a predetermined ratio of all measurement data When the measurement data is low-reliability, control using a temporary control amount may be performed.
Further, for example, x1-xcl, xcu-x1, etc. are converted into low reliability by using xcl, xcu, etc. in FIG. 21, and the control amount calculation processing unit 203a and the control processing unit 204a provide this low reliability. Whether or not to perform the control by the temporary control amount may be determined depending on the ratio of the numerical value that is equal to or greater than a predetermined numerical value.

Further, when the model generation unit 104 in the central apparatus 1A generates an operation model, a confidence interval may be used. That is, the model generation unit 104 may exclude the low-reliability measurement data and calculate the power state estimation model. At this time, it is desirable to use the basic calculation formula of the calculation model when there is low-reliability measurement data.

The reliability of the measurement data may be insufficient due to output disturbances in the load or the distributed power source. Even in such a case, by using the technique described in the second embodiment, control in consideration of reliability can be performed, and as a result, appropriate control can be performed.
When there is low-reliability measurement data, the control amount calculation processing unit 203a calculates a temporary control amount based on the control amount based on the high-reliability measurement data. Then, the control processing unit 204a performs control based on the temporary control amount, the control amount calculation processing unit 203a calculates the target deviation index from the output, and the control processing unit 204a is based on the corrected temporary control amount based on the target deviation index. The control is repeated until the target deviation index becomes a predetermined value or less. In this way, even when low-reliability measurement data exists, the control amount calculated based on the high-reliability measurement data can be used, so that the control reliability can be improved.

[Third Embodiment]
Next, a third embodiment of the present invention will be described with reference to FIGS.
An object of the third embodiment is to notify another control device 2B that measurement data with low reliability exists when measurement data is determined to be low reliability.

<Control device configuration>
FIG. 23 is a diagram illustrating a configuration example of a control device according to the third embodiment.
23, the same components as those of the control device 2B of FIG. 15 are denoted by the same reference numerals as those of FIG.
The control device 2B in FIG. 23 has a function of notifying the other control device 2B via the transmission / reception processing unit 201 when the reliability processing unit 205b determines that measurement data with low reliability exists. This is the difference from the control device 2A in FIG.
Since the configuration of the power system control system 5 and the configuration of the central device 1 are the same as those in FIGS. 1 and 14, illustration and description thereof are omitted.

(Control processing)
FIG. 24 is a flowchart illustrating a procedure of control processing according to the third embodiment.
Note that the processing in FIG. 24 is processing that is inserted into the locations of steps S311 to S313 of the distributed control processing in FIG. In the processing of FIG. 24, the same processing as that of FIG. 20 is denoted by the same step number as in FIG.

As a result of step S401, if all are highly reliable (S401 → Yes), has the reliability processing unit 205b received low reliability information that is information indicating that low-reliability measurement data exists from the other control device 2B? It is determined whether or not (S501).
If the low reliability information is not received as a result of step S501 (S501 → No), the control amount calculation processing unit 203a performs the same processing as steps S311 to S313 in FIG. 14 to control the measuring device.
As a result of step S401, whether there is even one piece of low-reliability measurement data (S401 → No), or if the result of step S501 is low-reliability information (S501 → Yes), the reliability processing unit 205b Reliability information, which is information indicating that low-reliability measurement data exists, is transmitted to another control device 2B (S502). Thereafter, the control amount calculation processing unit 203a proceeds to the process of step S402.

In the third embodiment, the reliability processing unit 205b notifies the other control device 2B of the fact when there is even one piece of low-reliability measurement data. If the measured data is low-reliability, this may be notified to the other control device 2B.

According to the third embodiment, when low-reliability measurement data exists, the control device 2B transmits a message to that effect to the other control device 2B, and receives the notification that low-reliability measurement data exists. Even if all of the measurement data are highly reliable, the measurement data is controlled with low reliability, so that it is possible to perform distributed control that is unified throughout the power system control system 5.
As a result, stable voltage control is possible in the entire power system control system 5.

In addition, this invention is not limited to above-described embodiment, Various modifications are included. For example, a time stamp may be attached to the system data. Moreover, although this embodiment presupposes application to an electric power grid | system system, application to a road traffic signal system etc. is also possible.
The calculation of the calculation model is not limited to the least square method, and multiple regression analysis, principal component analysis, or the like may be used.
Furthermore, in the second and third embodiments, the confidence intervals for the active power P and the reactive power Q are provided. However, the present invention is not limited thereto, and for example, a confidence interval for the voltage V may be provided.

The above-described embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to having all the configurations described. In addition, a part of the configuration of a certain embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of a certain embodiment. In addition, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

Each of the above-described configurations, functions, processing units 101 to 106, 201 to 205, 203a, 204a, 203b, and 205b, storage units 111 to 114, 211 to 214, etc. You may implement | achieve by a hardware by designing with an integrated circuit. Further, as shown in FIG. 2, the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by a processor such as the

CPU

301 or 401. As shown in FIG. 2, data such as programs, tables, and files for realizing each function are stored in the HD 304, as well as a memory, a recording device such as an SSD (Solid State Drive), or an IC (Integrated Circuit). It can be stored in a recording medium such as a card, an SD (Secure Digital) card, a DVD (Digital Versatile Disc), or a flash memory.
In each embodiment, control lines and information lines are those that are considered necessary for explanation, and not all control lines and information lines are necessarily shown on the product. In practice, it can be considered that almost all configurations are connected to each other.

1,1A Central device (management device)
2, 2a to 2c, 2A,

2B Control device

3, 3a to 3e Sensor (measuring device)
DESCRIPTION OF SYMBOLS 5 Power system control system 101 Communication state monitoring part 102 Power state estimation part 103 Control amount calculation process part 104 Model generation part 105 Transmission / reception part (transmission part)
106 confidence interval processing unit 111 measurement data storage unit 112 system data storage unit 113 model data storage unit (including calculation model)
114 confidence interval data storage (including reliability information)
200 controller 201 transmission / reception processing unit 202 control mode

determination processing unit

203, 203a control amount

calculation processing unit

204, 204a, 204b

control processing unit

205, 205b reliability processing unit 211 system data storage unit 212 model data storage unit (including calculation model) )
213 Measurement data storage unit 214 Confidence interval data storage unit (including reliability information)
1801 High confidence interval (first interval)
1802 Low confidence interval (second interval)

Claims

A model generation unit that generates an arithmetic model for estimating a control amount in the power system based on measurement values input from a plurality of measurement devices;
A transmission unit for transmitting the generated calculation model to the control device;
A management device having
A storage unit that stores information on a first section between a predetermined upper limit value and a lower limit value, and information on a second section other than the first section, as section information;
The measurement value occurring in the first interval is more reliable than the measurement value occurring in the second interval,
A reliability processing unit that determines whether a measurement value transmitted from the measurement device belongs to the first section or the second section based on the section information;
The amount of communication with the management device is monitored, and when the amount of communication is equal to or less than a predetermined value and the measured value is the second interval, the calculation is performed using the calculation model in the first interval. Calculate the temporary control amount based on the control amount, perform own control based on the temporary control amount,
When the communication amount is equal to or less than a predetermined value and the measured value is the first section, the control amount is calculated based on the calculation model, and the control is performed based on the control amount. A control processing unit to perform,
A power system control system comprising: a control device having:
The control processing unit
Control the own device based on the temporary control amount,
Calculating a target deviation index that is a difference between an output value as a result of the control and a target output value in the device itself;
Based on the target deviation index, the temporary control amount is newly calculated,
The power system control system according to claim 1, wherein the control of the device based on the temporary control amount is repeated until the target deviation index becomes a predetermined value or less.
The reliability processing unit includes:
The information indicating that the measurement value belonging to the second section exists is transmitted to another control device when the measurement value belonging to the second section exists. Power system control system described in 1.
The control processing unit in the control device that has received the information that the measurement value belonging to the second section exists,
The electric power system control system according to claim 3, wherein the temporary control amount is calculated and the control based on the temporary control amount is performed.
The power system control system according to claim 1, wherein the calculation model is generated for each missing state of the measurement value.
The power system control system according to claim 1, wherein the calculation model is generated for each predetermined living condition.
A management device that controls the control device
Based on the measurement values input from multiple measurement devices, generate a calculation model for estimating the control amount in the power system,
Send the generated calculation model to the control device,
A control device that receives the calculation model,
Information related to the first interval between the predetermined upper limit value and the lower limit value and information related to the second interval other than the first interval are stored in the storage unit as interval information,
The measurement value occurring in the first interval is more reliable than the measurement value occurring in the second interval,
Based on the section information, determine whether a measurement value transmitted from the measuring device belongs to the first section or the second section,
The amount of communication with the management device is monitored, and when the amount of communication is equal to or less than a predetermined value and the measured value is the second interval, the calculation is performed using the calculation model in the first interval. Calculate the temporary control amount based on the control amount, perform own control based on the temporary control amount,
When the communication amount is equal to or less than a predetermined value and the measured value is the first section, the control amount is calculated based on the calculation model, and the control is performed based on the control amount. A power system control method characterized in that:
The control device includes:
Control the own device based on the temporary control amount,
Calculating a target deviation index that is a difference between an output value as a result of the control and a target output value in the device itself;
Based on the target deviation index, the temporary control amount is newly calculated,
The power system control method according to claim 7, wherein the control of the device based on the temporary control amount is repeated until the target deviation index becomes a predetermined value or less.
The control device includes:
The information indicating that the measurement value belonging to the second section exists is transmitted to another control device when the measurement value belonging to the second section exists. The electric power system control method as described in.
The control device that has received the information indicating that the measurement value belonging to the second section exists,
The power system control method according to claim 9, wherein the temporary control amount is calculated and the control is performed based on the temporary control amount.