JP6213166B2 - Unknown parameter estimation method, unknown parameter program, and unknown parameter estimation device - Google Patents

Description

  The present invention relates to an unknown parameter estimation method, an unknown parameter program, and an unknown parameter estimation apparatus.

  In recent years, a distributed power source represented by a solar power generation system has been popularized in homes. While solar power generation systems have the advantage of being environmentally friendly, they also have the disadvantage of adversely affecting the quality of power supplied by the power company. In particular, there is a problem that the voltage in the distribution network may deviate from the specified voltage due to the introduction of the solar power generation system in the home (customer) in the distribution network. Note that the specified voltage means a voltage determined by the Electricity Business Law that must be maintained by the electric power company (in the case of 100V supply, a voltage in the range of 95V to 107V).

  As a method for controlling the voltage, for example, various methods such as changing the distribution voltage of the distribution substation, increasing the thickness of the electric wire, and reviewing the transformer tap can be considered. It is necessary to judge how to change. On the other hand, as a method for knowing the voltage of the distribution network, there is a method using a sensor installed on an electric wire, but at present, the number of sensors is very small with respect to the scale of the distribution network.

  On the other hand, as a method for obtaining the voltage in the distribution network and the current value in the distribution network, a technique called tidal current calculation is known. However, tidal current calculation is at least two of the voltage, current, active power, and reactive power of the consumer. It cannot be executed without entering more than one piece of information. However, smart meters currently being introduced can obtain values of active power and reactive power for some consumers, but in principle, only values of active power can be obtained. For this reason, in order to perform power flow calculation, it is important to estimate an unknown parameter (for example, reactive power) of a consumer.

  As a method for estimating an unknown parameter of a consumer, a method described in Patent Document 1 is known. This method is a combination of tidal current calculation and optimization method.

JP 2002-51464 A

  However, in the method of the above-mentioned patent document 1, when the number of unknown parameters is larger than the number of optimization objective functions, for example, all reactive powers of consumers are unknown, the value of the unknown parameter is uniquely determined. I can't.

  In one aspect, an object of the present invention is to provide an unknown parameter estimation method, an unknown parameter program, and an unknown parameter estimation apparatus that can estimate unknown parameters even when the number of unknown parameters is large.

  In one aspect, the unknown parameter estimation method includes processing for acquiring first data relating to power at a predetermined time interval from a first measurement device installed in a plurality of consumers, and upstream of the plurality of consumers. A process of acquiring second data related to power at a predetermined time interval from a second measuring device installed in the power supply, an unknown parameter related to power that cannot be acquired from the first measuring device, and the first data A process for determining a relational expression of the first parameter, a process for determining the probability of the relational expression using the first data and the second data, and a constraint on the unknown parameter using the probability of the relational expression. An unknown parameter estimation method in which a computer executes a process for determining a condition and a process for estimating the unknown parameter by an optimization process using the constraint condition.

  In one aspect, the unknown parameter estimation program acquires first data relating to power at a predetermined time interval from a first measurement device installed in a plurality of consumers, and is installed upstream of the plurality of consumers. The second data relating to the power is obtained at a predetermined time interval from the second measuring device, and the relational expression between the unknown parameter relating to the power which cannot be obtained from the first measuring device and the first data is obtained. Determining the probability of the relational expression using the first data and the second data, determining the constraint condition of the unknown parameter using the probability of the relational expression, and the constraint condition An unknown parameter estimation program for causing a computer to execute the process of estimating the unknown parameter by an optimization process using the.

  In one aspect, the unknown parameter estimation device includes a first acquisition unit that acquires first data related to power at predetermined time intervals from a first measurement device installed in a plurality of consumers, and the plurality of consumers. A second acquisition unit that acquires second data related to power at a predetermined time interval from a second measurement device installed upstream of the first measurement device, and an unknown parameter related to power that cannot be acquired from the first measurement device. A relational expression determination unit that determines a relational expression with the first data; a probability determination part that determines the probability of the relational expression using the first data and the second data; and the relational expression A constraint condition determination unit that determines the constraint condition of the unknown parameter using the certainty, and an estimation unit that estimates the unknown parameter by an optimization process using the constraint condition.

  Even if the number of unknown parameters is large, the unknown parameters can be estimated.

1 is a diagram schematically illustrating a configuration of a power distribution network system according to an embodiment. It is a figure which shows the hardware constitutions of information processing apparatus. It is a functional block diagram of an information processor. It is a flowchart which shows the process of information processing apparatus. FIG. 5A is a diagram illustrating an example of the data structure of the sensor DB, and FIG. 5B is a diagram illustrating an example of the data structure of the meter DB. FIG. 6A and FIG. 6B are diagrams for explaining the processing in step S16 in FIG. FIG. 5 is a diagram (part 1) for explaining the processing in step S18 in FIG. 4; FIG. 5 is a diagram (part 2) for explaining the process of step S18 of FIG. 4; It is a figure for demonstrating the process of step S20 of FIG. It is a figure which shows an example of the data structure of unknown parameter DB.

  Hereinafter, an embodiment of a power distribution network system will be described in detail with reference to FIGS.

As shown in FIG. 1, the distribution network system 100 according to the present embodiment includes a distribution network (distribution system) 80 and an information processing apparatus 20 as an unknown parameter estimation apparatus. As an example, it is assumed that there are a distribution substation 10 and three consumers 1 to 3 in the distribution network 80. The consumers 1 to 3 are provided with a smart meter 12 as a first measuring device, and the consumers 14 to 1 are provided with a sensor 14 as a second measuring device. And In addition, the voltage at the time t of the distribution substation 10 is represented as V ₀ (t).

The smart meter 12 can acquire an active power value (hereinafter simply referred to as “active power”) as first data at a predetermined time interval (for example, every 30 minutes), and the acquired active power is used as the information processing apparatus. 20 is transmitted. However, the smart meter 12 cannot acquire reactive power. In FIG. 1, the active power acquired at the time t in the smart meter 12 of the customer 1 is P ₁ (t) [W], and the effective power acquired at the time t in the smart meter 12 of the customer 2. The power is P ₂ (t) [W], and the active power acquired at time t in the smart meter 12 of the consumer 3 is P ₃ (t) [W].

The sensor 14 can acquire values of passing active power (P _sensor (t) [W]), passing reactive power (Q _sensor (t) [Var]), and voltage (V _sensor (t) [V]). Among the acquired values, the values of the passing reactive power (Q _sensor (t) [Var]) and the voltage (V _sensor (t) [V]) are transmitted to the information processing apparatus 20.

  The information processing apparatus 20 executes a process for obtaining a voltage at each point in the power distribution network 80 using power flow calculation or the like. The voltage at each point determined by the information processing device 20 is sent to a voltage control device (not shown) and used for voltage control of the distribution network. In addition, since the smart meter 12 cannot obtain reactive power of each consumer, which is input data for power flow calculation, the information processing apparatus 20 also executes processing for estimating reactive power of each consumer.

  FIG. 2 shows a hardware configuration of the information processing apparatus 20. As shown in FIG. 2, the information processing apparatus 20 includes a CPU (Central Processing Unit) 90, a ROM (Read Only Memory) 92, a RAM (Random Access Memory) 94, and a storage unit (here, an HDD (Hard Disk Drive)) 96. , A communication unit 97, a display unit 93, an input unit 95, a portable storage medium drive 99, and the like. Each component of the information processing apparatus 20 is connected to a bus 98. The display unit 93 includes a liquid crystal display or the like, and the input unit 95 includes a keyboard and a mouse. The communication unit 97 communicates with the smart meter 12 and the sensor 14. In the information processing apparatus 20, a program (including an unknown parameter estimation program) stored in the ROM 92 or the HDD 96, or a program (including an unknown parameter estimation program) read from the portable storage medium 91 by the portable storage medium drive 99. As shown in FIG. 3, the CPU 90 executes the functions of the acquisition unit 50, the estimation unit 52, the voltage calculation / output unit 34, and the control unit 60. 3 also shows a sensor DB 42, a meter DB 44, a distribution network configuration DB 46, and an unknown parameter DB 48 stored in the HDD 96 or the like.

  The acquisition unit 50 includes a sensor data acquisition unit 22 as a second acquisition unit, and a meter data acquisition unit 24 as a first acquisition unit.

The sensor data acquisition unit 22 acquires the passing reactive power (Q _sensor (t) [Var]) and the voltage (V _sensor (t) [V]) from the sensor 14 at a predetermined time interval (for example, every 30 minutes), Store in the sensor DB 42. Here, the sensor DB 42 has a data structure as shown in FIG. Specifically, the sensor DB 42 stores “time”, “pass reactive power” for storing pass reactive power (Q _sensor (t) [Var]), and “V _sensor (t) [V]” for storing voltage (V _sensor (t) [V]). Each field of “Voltage” is included.

  The meter data acquisition unit 24 acquires active power data measured at predetermined time intervals (30 minute intervals) from the smart meter 12 and stores the data in the meter DB 44. Here, the value of active power every 30 minutes for each of the consumers 1 to 3 is stored in the meter DB 44. In FIG. 5B, the active power of the customer 1 is “customer first active power”, the active power of the customer 2 is “customer second active power”, and the active power of the customer 3 is “demand”. Stored in each field of “house third active power”.

  The estimation unit 52 includes an unknown parameter relational expression determination unit 26 as a relational expression determination unit, a relational expression reliability determination unit 28 as a probability determination unit, an unknown parameter range determination unit 30 as a constraint condition determination unit, and an estimation And an optimization technique execution unit 32 as a unit.

  The unknown parameter relational expression determination unit 26 determines a relational expression between the value of active power of each consumer and the value of reactive power of each consumer as an unknown parameter.

  The relational expression reliability determination unit 28 determines the reliability (probability) of the relational expression determined by the unknown parameter relational expression determination unit 26.

  The unknown parameter range determination unit 30 uses the reliability (probability) of the relational expression to determine the constraint condition of reactive power of each consumer as an unknown parameter.

  The optimization technique execution unit 32 estimates the reactive power of each customer as an unknown parameter by an optimization process using the constraint condition determined by the unknown parameter range determination unit 30. Note that the optimization technique execution unit 32 refers to the distribution network configuration DB 46 in which circuit information and the like in the distribution network 80 are stored when estimating the reactive power of each consumer. Further, the optimization technique execution unit 32 stores the estimated unknown parameter in the unknown parameter DB 48. As shown in FIG. 10, the unknown parameter DB 48 has fields of “time”, “customer first reactive power”, “customer second reactive power”, and “customer third reactive power”. The field of “customer first reactive power” stores the estimated value of reactive power of customer 1, and the field of “customer second reactive power” stores the estimated value of reactive power of customer 2. Then, the estimated value of the reactive power of the customer 3 is stored in the field of “the third consumer reactive power”.

  The voltage calculation / output unit 34 refers to the distribution network configuration DB 46 and performs power flow calculation based on the reactive power value of each consumer stored in the unknown parameter DB 48 and the values of the sensor DB 42 and the meter DB 44. The voltage calculation / output unit 34 calculates the voltage at each point in the distribution network 80 obtained by power flow calculation, and outputs it to a voltage control device (not shown). Thereby, the voltage control apparatus can maintain the voltage in the power distribution network 80 within the specified voltage (95 to 107 V).

  The control unit 60 controls the acquisition unit 50, the estimation unit 52, and the voltage calculation / output unit 34 in an integrated manner.

(Processing of information processing apparatus 20)
Next, details of processing by the information processing apparatus 20 of the present embodiment will be described along the flowchart of FIG. 4 with reference to other drawings as appropriate.

In the process of FIG. 4, first, in step S10, the control unit 60 sets the data acquisition count T and sets the bootstrap sample count B. For example, the control unit 60 sets T = 48 points / 1 day × 30 days = 1440 as T when the data acquisition period is 30 days and the data acquisition frequency is 30 minutes. In the present embodiment, if the time t ₁ is 0:00 on the first day of data acquisition, the time t ₂ is 0:30 on the first day of data acquisition, and the time t _T The day will be 23:30. Further, the control unit 60 sets 1000 as the bootstrap sample number B, for example.

Next, in step S12, under instruction of the control unit 60, sensor data acquisition unit 22, the time t ₁ obtained by the sensor 14, ..., reactive power passing in t _T Q _sensor (t) and the voltage V _sensor (t) Is stored in the sensor DB. The sensor DB 42 stores data as shown in FIG.

In step S14, under the instruction of the control unit 60, the meter data acquisition unit 24 obtains the active power P at times t ₁ ,..., _{T T} obtained by the smart meter 12 of the customer j (j = 1 to 3). (t) is acquired and stored in the meter DB 44. The meter DB 44 stores data as shown in FIG. Note that steps S12 and S14 are actually performed at the same timing.

Next, in step S16, under the instruction of the control unit 60, the unknown parameter relational expression determination unit 26 determines the relational expression between the active power P _j (t) and the reactive power Q _j (t) of the customer j as Q _j (t ) = k _j P _j (t), and based on the data obtained at time t ₁ ,..., t _T , the proportionality constant k _j of the relational expression Q _j (t) = k _j P _j (t) is minimized. Obtained by the square method.

  Specifically, in step S16, the following processing is executed.

  FIG. 6A shows a table summarizing each data (excluding “voltage” in FIG. 5A) of the sensor DB 42 in FIG. 5A and the meter DB 44 in FIG. 5B. Yes.

Here, the unknown parameter relational expression determining unit 26 is a proportional constant of the relational expression Q _j (t) = k _j P _j (t) of the active power P _j (t) of the customer j and the reactive power Q _j (t). Assuming that k _j is known, the table of FIG. 6A is rewritten as shown in FIG. Note that the relationship between the active power P _j (t) and the reactive power Q _j (t) of the consumer j is Q _j (t) = k _j P _j (t) in general. This is because there is a high possibility that the active power P _j (t) and the reactive power Q _j (t) are proportional to each other. However, the present invention is not limited to this, and other relational expressions may be used according to the actual situation.

Here, since the passing reactive power Q _sensor (t) is the sum of reactive power consumed on the downstream side (the consumer side) from the sensor 14, the unknown parameter relational expression determination unit 26 calculates the following formula (1): A simultaneous equation such as

In addition, Q _loss of the above formula (1) is reactive power consumed by the distribution line, and ε is an error.

  The unknown parameter relational expression determination unit 26 calculates the sum of squared errors.

を最小とする定数k₁，k₂，k₃を最小二乗法で求めることで、各需要家における有効電力と無効電力の関係式を特徴づける定数k₁，k₂，k₃を導出する。

By obtaining the constants k ₁ , k ₂ , and k ₃ that minimize the value by the least square method, the constants k ₁ , k ₂ , and k ₃ that characterize the relational expression between the active power and the reactive power in each consumer are derived.

That is, the unknown parameter relational expression determination unit 26 derives the following relational expressions (the following expressions (3) to (5)).
Q ₁ (t) = k ₁ P ₁ (t) (3)
Q ₂ (t) = k ₂ P ₂ (t) (4)
Q ₃ (t) = k ₃ P ₃ (t) (5)

Next, in step S18, under instruction of the control unit 60, the relational expression reliability determining unit 28, the time t _1, ..., data obtained in t _T (FIG. 6 (b)) was B times resampling, Create a set of B samples.

Specifically, first, the relationship reliability determining unit 28, in FIG. 6 (b), the set of data obtained at time _{t i {Q sensor (t i} ), P 1 (t i), P 2 ( t _i ), P ₃ (t _i )} are defined as Z _i (see FIG. 7).

Next, as shown in FIG. 8, the relational expression reliability determination unit 28 randomly selects T data from the set {Z ₁ ,..., Z _T } with duplication, and creates B sets of samples. . As a result, for example, a plurality of Z ₂ may be selected and Z ₃ may not be selected as in the first set of samples in FIG. 8, or Z ₄ may be selected as in the second set of samples in FIG. It may happen that multiple selections are made and Z ₁ is not selected. Each of these sets of samples is called a bootstrap specimen.

Next, in step S20, under the instruction of the control unit 60, the relational expression reliability determination unit 28 sets the proportional constant k _j of Q _j (t) = k _j P _j (t) for each of the B sets of bootstrap samples. Executes processing obtained by the least square method.

In this case, the relational expression reliability determination unit 28 performs the same processing as step S16 on each of the B sets of samples, and determines the active power and reactive power of the consumer j obtained for the b set of bootstrap samples. The proportionality constant k _{j (b group)} that characterizes the relationship is obtained. FIG. 9 schematically shows the process of step S20.

Next, in step S22, under the instruction of the control unit 60, the relational expression reliability determination unit 28 obtains the variance of k _j based on k _{j (b set)} obtained for the B set.

For example, the k _{1 (b group)} obtained for the set B for the customer 1 is k _{1 (first group)} , k _{1 (second group)} ,..., K _{1 (B group)} . The relational expression reliability determination unit 28 calculates the variance V (k ₁ ) of the B pieces of k ₁ data from the following expression (6).

The relational expression reliability determination unit 28 also calculates the variances V (k ₂ ) and V (k ₃ ) of the proportionality constants k ₂ and k ₃ of the relational expressions of the consumers ₂ and _{3 in} the same manner as described above. .

Next, in step S24, under the instruction of the control unit 60, the unknown parameter range determination unit 30 determines a range in which Q _j (t) moves based on the variance of k _j .

Here, the large variance of k _j , that is, the relational expression Q _j (t) = k _j P _j (t) between the active power and the reactive power related to customer j is difficult to hit (the reliability is low, Not). Therefore, the unknown parameter range determination unit 30 represents the range in which the reactive power moves in each customer j by the following equations (7) to (9), and increases the value of l _j as the variance of k _j increases. .
| Q ₁ (t) -k ₁ P ₁ (t) | <s · l ₁ (t) (7)
| Q ₂ (t) -k ₂ P ₂ (t) | <s · l ₂ (t) (8)
| Q ₃ (t) -k ₃ P ₃ (t) | <s · l ₃ (t) (9)

Note that s is a coefficient. Here, as described above, in order to increase the value of l _j as the variance of k _j increases, the unknown parameter range determination unit 30 obtains the value of l _j (t) in step S22. V (k _j ), (V (k _j )) ^1/2 or the like can be used. However, not limited to this, as the value of l _j (t), other expressions using V (k _j ) (however, the value of l _j increases as the variance of k _j increases) are adopted. It is good as well.

Next, in step S26, under the instruction of the control unit 60, the optimization method execution unit 32 estimates the value of the reactive power Q _j (t) by the optimization method using the range in which the reactive power Q _j (t) moves. Execute the process.

In this process, first, the optimization method execution unit 32 performs reactive powers Q ₁ (t), Q ₂ (t), Q ₃ (t) of consumers 1, 2, 3 and consumers 1, 2, 3 The sensor 14 voltage calculated based on the active power P ₁ (t), P ₂ (t), and P ₃ (t) is V _{sensor_calc} (t), and this value is actually measured by the sensor 14. The total M of all the differences in V _sensor (t) is expressed by the following equation (10).

Then, the optimization technique execution unit 32 determines unknown parameters (reactive power) Q ₁ (t), Q ₂ (t), and Q ₃ (t) so that the value of M is minimized, and uses it as an estimated value. . However, the optimization technique executing unit 32 reacts the reactive power Q ₁ (t), Q ₂ (t), and Q ₃ (t) so that the constraints of the above equations (7) to (9) are satisfied. _It is necessary to estimate ₁ (t), Q ₂ (t), and Q ₃ (t).

Next, in step S28, the optimization technique execution unit 32 determines whether or not the unknown parameters (reactive power) Q ₁ (t), Q ₂ (t), and Q ₃ (t) have been estimated. If the determination here is negative, that is, if reactive powers Q ₁ (t), Q ₂ (t), and Q ₃ (t) cannot be estimated, the process proceeds to step S30 and the optimization method is executed. The unit 32 changes s in the above formulas (7) to (9). For example, if the constraint condition is too strict and at least one of reactive powers Q ₁ (t), Q ₂ (t), and Q ₃ (t) cannot be estimated, the optimization technique execution unit 32 sets the constraint condition. Increase s to loosen. If the constraint condition is too loose and at least one solution of the reactive powers Q ₁ (t), Q ₂ (t), and Q ₃ (t) cannot be determined, the optimization technique execution unit 32 sets the constraint condition. Reduce s to make it more strict. Thereafter, the process returns to step S24, and steps S24 to S30 are repeated.

Then, at the stage where the determination in step S28 is affirmed, that is, at the stage where all of the reactive powers Q ₁ (t), Q ₂ (t), and Q ₃ (t) can be estimated uniquely, the optimization technique execution unit 32 Stores the values of the reactive powers Q ₁ (t), Q ₂ (t), and Q ₃ (t) in the unknown parameter DB 48, and completes the entire process of FIG.

  As described above, at the stage where the process of FIG. 4 is completed, the estimation result of the reactive power is stored in the unknown parameter DB 48 as shown in FIG. This reactive power estimation result is used for power flow calculation by the voltage calculation / output unit 34. Thereby, the voltage calculation / output unit 34 can calculate the voltage at each point in the power distribution network 80 with high accuracy, and outputs the calculated voltage value at each point to a voltage control device (not shown). Can do. Therefore, in the voltage control device, based on the voltage at each point in the distribution network 80 acquired from the voltage calculation / output unit 34, the voltage management is performed so as to maintain the voltage at each point within the specified voltage (95 to 107V). Can be done.

As described above in detail, according to the present embodiment, the meter data acquisition unit 24 sends the active power from the smart meter 12 installed in a plurality of consumers j at predetermined time intervals (for example, every 30 minutes). Obtaining (S14), the sensor data obtaining unit 22 obtains the passing reactive power and voltage values at predetermined time intervals (for example, every 30 minutes) from the sensors 14 installed upstream of the plurality of consumers ( S12). Then, the unknown parameter relational expression determination unit 26 determines a relational expression between the unknown parameter (reactive power) and the value of the active power (S16), and the relational expression reliability determination unit 28 determines the active power, the passing reactive power, Using the voltage value, the probability of the relational expression (for example, k _j variance) is obtained (S18 to S22). Further, the unknown parameter range determination unit 30 determines the constraint conditions of the unknown parameters (the above formulas (7) to (9)) using the certainty of the relational expression (for example, the variance of k _j ) (S24), The optimization technique execution unit 32 estimates the unknown parameter by the optimization process using the constraint condition (S26). Thus, in the present embodiment, by providing a constraint condition that reflects the probability of the relational expression (for example, the variance of k _j ) in the unknown parameter, the number of unknown parameters is an optimization objective function (formula (10)). Even when the number is larger than the number of unknown parameters, it is possible to uniquely determine the value of the unknown parameter by narrowing down the unknown parameter according to the constraint condition.

Further, in the present embodiment, as the probable relational expression, for example, by narrowing the range of values distributed can take more unknown parameters is small k _j, as relational expression not like Indeed, a large dispersion of example k _j The range of values that unknown parameters can take is increased. Thereby, in this embodiment, the value which an unknown parameter can take can be set appropriately according to the probability of a relational expression.

In this embodiment, the active power value acquired from the smart meter 12 and the pass reactive power and voltage values acquired from the sensor 14 are resampled a plurality of times, and a plurality of bootstrap samples obtained by resampling are used. Thus, the variance of the proportionality constant k _j of the relational expression, that is, the probability of the relational expression is determined. Thereby, the probability of the relational expression can be calculated statistically with high accuracy.

In this embodiment, the voltage value V _sensor (t) acquired from the sensor 14 and the estimation of the voltage obtained by the sensor 14 expressed as a function of the unknown parameter under the constraint condition of the unknown parameter (reactive power). An unknown parameter is determined by applying an optimization method using the difference from the value V _{sensor_calc} (t) as an objective function. Thereby, an unknown parameter can be determined by an appropriate method.

In the above embodiment, for convenience of explanation, the case where the voltage calculation / output unit 34 calculates the voltage at each point after estimating Q _i (t) in step S26 has been described. However, the present invention is not limited to this. Absent. That is, the voltage calculation / output unit 34 may output the voltage value of each consumer calculated when Q _i (t) is estimated in step S26 to the voltage control device.

  In addition, although the said embodiment demonstrated the case where there were three consumers in the power distribution network 80, it is not restricted to this, A consumer may be the number of other houses. When reactive power can be measured with a smart meter installed in some consumers, only the reactive power of other consumers may be used as an unknown parameter.

  In the above embodiment, the reactive power value at the consumer is estimated as an unknown parameter from the value of the passing reactive power and voltage obtained from the sensor 14 and the value of the active power obtained from the smart meter 12. explained. However, the present invention is not limited thereto, and the data obtained from the sensor 14 and the smart meter 12 may be data related to other power, and the data estimated as the unknown parameter may be data related to power other than reactive power. The data output from the information processing apparatus 20 is not limited to the voltage data at each point in the distribution network, and may be other data related to power such as current data.

  Moreover, each process demonstrated by the said embodiment is an example. For example, the process for determining the relational expression between the unknown parameter that cannot be acquired from the smart meter 12 and the data obtained by the smart meter 12 is not limited to the method described in step S16 described above, and may be related by other methods. It is good also as determining a formula. Further, the process for determining the certainty of the relational expression using the data obtained by the smart meter 12 and the data obtained by the sensor 14 is not limited to the method described in steps S18 to S22 described above, and other methods. The probability of the relational expression may be determined by (for example, a method using a value other than variance). Further, the process for determining the constraint condition of the unknown parameter using the certainty of the relational expression is not limited to the method described in step S24 described above, and the constraint condition of the unknown parameter may be determined by other methods. Good. Furthermore, the process for estimating the unknown parameter by the optimization process using the constraint condition is not limited to the method described in step S26 described above, and the unknown parameter may be estimated by another optimization process.

  The above processing functions can be realized by a computer. In that case, a program describing the processing contents of the functions that the processing apparatus should have is provided. By executing the program on a computer, the above processing functions are realized on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium (except for a carrier wave).

  When the program is distributed, for example, it is sold in the form of a portable recording medium such as a DVD (Digital Versatile Disc) or a CD-ROM (Compact Disc Read Only Memory) on which the program is recorded. It is also possible to store the program in a storage device of a server computer and transfer the program from the server computer to another computer via a network.

  The computer that executes the program stores, for example, the program recorded on the portable recording medium or the program transferred from the server computer in its own storage device. Then, the computer reads the program from its own storage device and executes processing according to the program. The computer can also read the program directly from the portable recording medium and execute processing according to the program. Further, each time the program is transferred from the server computer, the computer can sequentially execute processing according to the received program.

  The above-described embodiment is an example of a preferred embodiment of the present invention. However, the present invention is not limited to this, and various modifications can be made without departing from the scope of the present invention.

In addition, the following additional remarks are disclosed regarding description of the above embodiment.
(Additional remark 1) The process which acquires the 1st data regarding electric power at a predetermined time interval from the 1st measuring device installed in a plurality of consumers,
From the second measuring device installed on the upstream side of the plurality of consumers, a process of acquiring second data related to power at a predetermined time interval;
A process for determining a relational expression between an unknown parameter relating to power that cannot be obtained from the first measurement device and the first data;
A process of determining the probability of the relational expression using the first data and the second data;
A process for determining a constraint condition of the unknown parameter using the certainty of the relational expression;
An unknown parameter estimation method, wherein a computer executes a process of estimating the unknown parameter by an optimization process using the constraint condition.
(Supplementary Note 2) In the process of determining the constraint condition of the unknown parameter,
The unknown parameter according to claim 1, wherein the range of values that can be taken by an unknown parameter is narrowed as the relational expression is more probable, and the range of values that can be taken by an unknown parameter is widened so that the relational expression is less accurate. Estimation method.
(Additional remark 3) When the unknown parameter cannot be estimated in the process of estimating the unknown parameter, the computer executes the process of estimating the unknown parameter again by changing the constraint condition. Or the unknown parameter estimation method of 2.
(Supplementary Note 4) In the process of determining the certainty of the relational expression,
Resampling the first data acquired from the first measurement device and the second data acquired from the second measurement device a plurality of times, using a plurality of samples obtained by resampling, 4. The unknown parameter estimation method according to any one of appendices 1 to 3, wherein the probability of the relational expression is determined.
(Supplementary Note 5) In the process of estimating the unknown parameter,
By applying an optimization method using the difference between the second data and the estimated value of the second data expressed as a function of the unknown parameter under the constraint condition as an objective function, the unknown parameter is The unknown parameter estimation method according to any one of supplementary notes 1 to 4, wherein the parameter is determined.
(Supplementary Note 6) First data related to power is acquired at predetermined time intervals from a first measuring device installed in a plurality of consumers.
From the second measuring device installed on the upstream side of the plurality of consumers, second data on power is acquired at predetermined time intervals,
Determining a relational expression between an unknown parameter relating to power that cannot be obtained from the first measurement device and the first data;
Using the first data and the second data to determine the probability of the relational expression;
Using the certainty of the relational expression, determine the constraint condition of the unknown parameter,
An unknown parameter estimation program for causing a computer to execute a process of estimating the unknown parameter by an optimization process using the constraint condition.
(Supplementary note 7) In the process of determining the constraint condition of the unknown parameter,
The unknown parameter according to appendix 6, wherein the range of values that the unknown parameter can take is narrowed as the relational expression is more probable, and the range of values that the unknown parameter can take is widened so that the relational expression is not as accurate. Estimation program.
(Supplementary note 8) When the unknown parameter cannot be estimated in the process of estimating the unknown parameter, the computer is caused to execute a process of estimating the unknown parameter again by changing the constraint condition. Or the unknown parameter estimation program according to 7;
(Supplementary Note 9) In the process of determining the certainty of the relational expression,
Resampling the first data acquired from the first measurement device and the second data acquired from the second measurement device a plurality of times, using a plurality of samples obtained by resampling, The unknown parameter estimation program according to any one of appendices 6 to 8, wherein the probability of the relational expression is determined.
(Supplementary Note 10) In the process of estimating the unknown parameter,
By applying an optimization method using the difference between the second data and the estimated value of the second data expressed as a function of the unknown parameter under the constraint condition as an objective function, the unknown parameter is The unknown parameter estimation program according to any one of appendices 6 to 9, wherein the unknown parameter estimation program is determined.
(Additional remark 11) The 1st acquisition part which acquires the 1st data about electric power from a 1st measuring device installed in a plurality of consumers at predetermined time intervals,
A second acquisition unit that acquires second data relating to power at predetermined time intervals from a second measurement device installed on the upstream side of the plurality of consumers;
A relational expression determining unit that determines a relational expression between an unknown parameter relating to power that cannot be obtained from the first measurement device and the first data;
A probability determining unit that determines the probability of the relational expression using the first data and the second data;
A constraint condition determination unit that determines the constraint condition of the unknown parameter using the probability of the relational expression;
An unknown parameter estimation device comprising: an estimation unit that estimates the unknown parameter by an optimization process using the constraint condition.
(Supplementary Note 12) The constraint condition determining unit
The unknown parameter according to appendix 11, wherein the range of values that can be taken by an unknown parameter is narrowed as the relational expression is more probable, and the range of values that can be taken by an unknown parameter is widened so that the relational expression is less accurate. Estimating device.
(Additional remark 13) The said estimation part changes the said constraint, and estimates the said unknown parameter again, when the said unknown parameter cannot be estimated, The unknown parameter estimation apparatus of Additional remark 11 or 12 characterized by the above-mentioned.
(Supplementary Note 14) The probability determining unit
Resampling the first data acquired from the first measurement device and the second data acquired from the second measurement device a plurality of times, using a plurality of samples obtained by resampling, 14. The unknown parameter estimation device according to any one of appendices 11 to 13, wherein the probability of the relational expression is determined.
(Supplementary Note 15) The estimation unit
By applying an optimization method using the difference between the second data and the estimated value of the second data expressed as a function of the unknown parameter under the constraint condition as an objective function, the unknown parameter is The unknown parameter estimation device according to any one of appendices 11 to 14, wherein the unknown parameter estimation device is determined.

12 Smart meter (first measuring device)
14 Sensor (second measuring device)
20 Information processing device (unknown parameter estimation device)
22 Sensor data acquisition unit (second acquisition unit)
24 Meter data acquisition unit (first acquisition unit)
26 Unknown parameter relational expression determination unit (relational expression determination unit)
28 Relational Expression Reliability Determining Unit (Probability Determining Unit)
30 Unknown parameter range determination unit (constraint condition determination unit)
32 Optimization method execution part (estimation part)

Claims (7)

A process of acquiring first data relating to power at a predetermined time interval from a first measuring device installed in a plurality of consumers;
From the second measuring device installed on the upstream side of the plurality of consumers, a process of acquiring second data related to power at a predetermined time interval;
A process for determining a relational expression between an unknown parameter relating to power that cannot be obtained from the first measurement device and the first data;
A process of determining the probability of the relational expression using the first data and the second data;
A process for determining a constraint condition of the unknown parameter using the certainty of the relational expression;
An unknown parameter estimation method, wherein a computer executes a process of estimating the unknown parameter by an optimization process using the constraint condition. In the process of determining the constraint condition of the unknown parameter,
2. The unknown range according to claim 1, wherein the range of values that can be taken by an unknown parameter is narrowed as the relational expression is more probable, and the range of values that can be taken by an unknown parameter is widened as the relational expression is less accurate. Parameter estimation method.   3. The computer according to claim 1, wherein, when the unknown parameter cannot be estimated in the process of estimating the unknown parameter, the computer executes a process of changing the constraint condition and estimating the unknown parameter again. The unknown parameter estimation method described. In the process of determining the certainty of the relational expression,
Resampling the first data acquired from the first measurement device and the second data acquired from the second measurement device a plurality of times, using a plurality of samples obtained by resampling, 4. The unknown parameter estimation method according to claim 1, wherein the probability of the relational expression is determined. In the process of estimating the unknown parameter,
By applying an optimization method using the difference between the second data and the estimated value of the second data expressed as a function of the unknown parameter under the constraint condition as an objective function, the unknown parameter is The unknown parameter estimation method according to any one of claims 1 to 4, wherein the parameter is determined. From the first measuring device installed in a plurality of consumers, first data related to power is acquired at predetermined time intervals,
From the second measuring device installed on the upstream side of the plurality of consumers, second data on power is acquired at predetermined time intervals,
Determining a relational expression between an unknown parameter relating to power that cannot be obtained from the first measurement device and the first data;
Using the first data and the second data to determine the probability of the relational expression;
Using the certainty of the relational expression, determine the constraint condition of the unknown parameter,
An unknown parameter estimation program for causing a computer to execute a process of estimating the unknown parameter by an optimization process using the constraint condition. A first acquisition unit that acquires first data relating to electric power at predetermined time intervals from a first measurement device installed in a plurality of consumers;
A second acquisition unit that acquires second data relating to power at predetermined time intervals from a second measurement device installed on the upstream side of the plurality of consumers;
A relational expression determining unit that determines a relational expression between an unknown parameter relating to power that cannot be obtained from the first measurement device and the first data;
A probability determining unit that determines the probability of the relational expression using the first data and the second data;
A constraint condition determination unit that determines the constraint condition of the unknown parameter using the probability of the relational expression;
An unknown parameter estimation device comprising: an estimation unit that estimates the unknown parameter by an optimization process using the constraint condition.

Images