JP2022131016A - State estimation device, state estimation program, and state estimation method - Google Patents

Abstract

To provide a state estimation device, a state estimation program, and a state estimation method for estimating the state of a system with high accuracy even when the measurement result of a sensor includes an abnormal value.SOLUTION: A state estimation device includes a first state estimation unit that estimates the state of a system on the basis of system information measured by a plurality of tools and a system model, a determination unit that determines whether there is an abnormality in the system information on the basis of the plurality of pieces of system information and a first estimation result estimated by the first state estimation unit, and a second state estimation unit that outputs a second estimation result when it is determined that there is an abnormality in the system information. The first state estimation unit estimates the state of the system using a determination variable that maximizes the evaluation value in an objective function that increases the evaluation value as the difference between the system information and the first estimation result decreases, and reduces the influence of abnormal values included in the system information on the evaluation value as the value of the predetermined variable decreases. The second state estimation unit outputs the second estimation result on the basis of variations in the estimation results for the respective value of the predetermined variable.SELECTED DRAWING: Figure 7

Description

The present invention relates to a state estimation device, a state estimation program, and a state estimation method.

For example, there is known a technique of estimating the state of a system that supplies electric power, gas, and water based on measured system information (see Patent Documents 1 and 2, for example).

JP-A-2005-62171 Japanese Unexamined Patent Application Publication No. 2013-132106

By the way, when a sensor for measuring system information provided in a system fails, an abnormal value (or an outlier) is included in the measurement result of the sensor. If the state of the system is estimated by removing such sensor abnormal values, the accuracy of the estimation result generally deteriorates.

The present invention has been made in view of the conventional problems as described above, and is capable of estimating the state of the system with high accuracy even when the sensor measurement results include abnormal values. The purpose is to provide technology.

The main invention for solving the above-mentioned problems indicates the state of a system that supplies any one of electric power, gas, and liquid, and the plurality of system information measured at each of a plurality of measurement points, and the system a simulated model; a first state estimating unit that estimates the state of the system based on; the plurality of system information; and a first estimation result estimated by the first state estimating unit. a determination unit that determines whether or not there is an abnormality in a plurality of system information; and a second state that outputs a second estimation result of estimating the state of the system when it is determined that there is an abnormality in the plurality of system information. and an estimating unit, wherein the first state estimating unit increases the evaluation value as the difference between the system information and the first estimation result decreases, and increases the evaluation value as the value of a predetermined variable decreases. In the objective function that reduces the influence of the abnormal value included in the evaluation value, a decision variable indicating a value corresponding to the change in any one of the predetermined positions in the system is changed, and the evaluation value is maximized The second state estimating unit estimates the state of the system using the decision variable when the state is different, and the second state estimating unit estimates the evaluation using the optimization method under a plurality of different conditions for each value of the plurality of predetermined variables The state estimating device outputs the second estimation result based on variations in the plurality of estimation results for each value of the predetermined variable after obtaining a plurality of estimation results having the maximum value.

ADVANTAGE OF THE INVENTION According to this invention, the technique which can estimate the state of a system|strain with high precision even when the measurement result of a sensor contains an abnormal value can be provided.

1 is a diagram showing a configuration of an information processing device 10; FIG. 3 is a diagram for explaining a power system 30; FIG. 2 is a diagram for explaining the information processing apparatus 10; FIG. 4 is a diagram for explaining a voltage distribution (estimation result) of the distribution line 41; FIG. 4 is a flowchart showing an example of processing executed by the information processing apparatus 10; 4 is a diagram for explaining an example of voltage distribution of distribution line 41. FIG. FIG. 10 is a diagram for explaining variations in voltage distribution when the kernel size is changed;

At least the following matters will become apparent from the descriptions of this specification and the accompanying drawings.

=====This Embodiment=====
<<<configuration of information processing apparatus 10>>>
FIG. 1 is a diagram showing the configuration of an information processing apparatus 10 that is an embodiment of the present invention. The information processing device 10 is, for example, a device for estimating the state of a power system 30 (described later) shown in FIG. and a computer including a communication device 25 . The information processing device 10 corresponds to a "state estimation device".

The CPU 20 implements various functions of the information processing apparatus 10 by executing programs stored in the memory 21 and the storage device 22 (storage unit).

The memory 21 is, for example, a RAM (Random-Access Memory) or the like, and is used as a temporary storage area for programs, data, and the like.

The storage device 22 is a nonvolatile storage device that stores various data to be executed or processed by the CPU 20 .

The input device 23 is a device that receives commands and data input by a user, and includes an input interface such as a keyboard and a touch sensor that detects a touch position on a touch panel display.

The display device 24 is, for example, a device such as a display, and the communication device 25 exchanges various programs and data with other computers via a network (not shown).

<<<Example of Power System 30>>>
FIG. 2 is a diagram showing an example of a power system 30 in which the information processing device 10 performs state estimation. The electric power system 30 is, for example, a 6.6 kV distribution line system, and includes a transformer 40, a distribution line 41, consumers 42 to 45, and sensors S1 to S4. A general power system also includes distribution lines, consumers, sensors, and the like, but for the sake of convenience, a simplified power system 30 is shown here as an example.

The transformer 40 transforms a voltage supplied from a transmission line (not shown) and outputs a voltage of 6.6 kV to the distribution line 41 .

The consumer 42 is a load (for example, a factory) that consumes power supplied from the distribution line 41 , and includes power generating equipment (not shown) such as an inverter that supplies power to the distribution line 41 . The consumers 43 to 45 are similar to the consumer 42. Therefore, the distribution line 41 is connected to a load that consumes power from the distribution line 41 and a power generation facility that supplies power to the distribution line 41 .

The sensor S<b>1 is a measuring instrument provided at “measurement point A” of the distribution line 41 to measure “system information Za” of the distribution line 41 . Specifically, the sensor S1 measures the “voltage Va” and the “current Ia” at the “measurement point A” of the distribution line 41 .

The sensors S2 to S4 are similar to the sensor S1, and are measuring instruments provided at the "measurement points B" to "measurement points D" of the distribution line 41, respectively. System information Zd'' is measured.

The "system information Zi" (i=a to d) in this embodiment is the "voltage Vi" and the "current Ii", but for the sake of convenience, the following description will focus on the "voltage Vi".

Further, although the details will be described later, the information processing apparatus 10 changes the active power P and the reactive power Q of the consumers 42 to 45, and for example, A power flow calculation is performed so that each estimation result matches the measurement result of the sensors S1 to S4. As a result, the information processing apparatus 10 can calculate an estimation result (for example, voltage distribution of the distribution line 41) as shown in FIG.

<<<Function Blocks of Information Processing Apparatus 10>>>
FIG. 3 is a diagram showing functional blocks and the like of the information processing apparatus 10. As shown in FIG. A system model 60 is stored in the storage device 22 of the information processing device 10 . Further, the information processing apparatus 10 implements a first state estimation unit 70, a determination unit 71, and a second state estimation unit 72 by the CPU 20 executing a predetermined program.

The system model 60 is, for example, a model simulating the power system 30 represented by a state equation. Although details will be described later, when the power flow calculation is executed using the system model 60 in a state where the active power P and the reactive power Q of the consumers 42 to 45 are specified, the state of the power system 30, that is, the distribution line 41 voltages, currents, phase angles, etc. are obtained.

==First state estimator 70==
The first state estimating unit 70 calculates the electric power based on the “system information Za” to “system information Zd” measured at each of the four “measurement points A” to “measurement points D” and the system model 60. The state of system 30 is estimated. Specifically, the first state estimation unit 70 estimates the voltage distribution of the distribution line 41 using the measured “voltage VA” to “voltage VD”.

Here, the details of the processing executed by the first state estimation unit 70 will be described with reference to FIG. 4 . FIG. 4 is a diagram showing an example of an estimation result of the electric power system 30 calculated by the "least squares method" method and by the "colentropy" method.

In FIG. 4, among the measurement results of "measurement point A" to "measurement point D" ("black circles" in FIG. 4), it is assumed that only the measurement result of "measurement point C" is abnormal.

When the voltage distribution of the distribution line 41 is estimated, hypothetically, the objective function based on the least squares method that minimizes the difference between the measurement results of “measurement point A” to “measurement point D” and the estimation result is When used, the "estimation result R1" indicated by the dashed line deviates greatly from the correct voltage distribution indicated by the dotted line. This is because the value of the objective function of the least-squares method is minimized by bringing the "estimation result R1" close to the "measurement point C", which is an abnormal value.

On the other hand, the "colentropy" method is a method capable of suppressing the influence of "measurement point C", which is an abnormal value. In the "colentropy" method, the following objective function J(x), which is the normal distribution of the formula for the difference between the estimated result and the measured result, is used, and the objective function J(x) is determined to be the maximum. A variable x is searched.

Here, "N" is the number of measurement points, "i" indicates the i-th measurement point among the measurement points, and "σ" is, for example, the kernel size set by the user (or standard deviation). Also, "yi(x)" indicates the estimation result of the i-th measurement point, and "zi" indicates the measurement result of the i-th measurement point. The kernel size σ corresponds to a "predetermined variable", and the value of J(x) corresponds to an "evaluation value".

In such a formula, when the estimated result "yi(x)" deviates from the measured result "zi", the value of the objective function J(x) decreases and approaches "0" (zero).

On the other hand, when the estimation result "yi(x)" approaches the measurement result "zi", the value of the objective function J(x) increases.

For this reason, for example, rather than bringing the estimation result closer to the measurement result of the erroneous "measurement point C" in FIG. The closer the results, the larger the value of the objective function J(x).

Here, the first state estimation unit 70 of the present embodiment uses the active power P and the reactive power Q of the consumers 42 to 45 as the decision variable x, and as "i", "measurement point A" to "measurement point "4", which is the number of D", is used.

Then, the first state estimating unit 70 uses the decision variable x (that is, the active power P and the reactive power Q of the consumers 42 to 45) when the value of the objective function J(x) is maximized to determine the power system Estimate 30 states. As a result, by maximizing the objective function J(x), the first state estimator 70 can calculate, for example, the “estimation result R2” indicated by the dashed line, which is close to the correct voltage distribution indicated by the dotted line. .

==Determination Unit 71==
The determination unit 71 determines whether or not there is an abnormality in the output of the sensor Si at the i-th measurement point. Specifically, the determination unit 71 determines that the difference between the “voltage Vi” that is the measurement result of the sensor Si and the estimation result of the “measurement point i” estimated by the first state estimation unit 70 is greater than a predetermined value. In this case, it is determined that there is an abnormality in the measurement result of the sensor Si. A measurement point determined to be abnormal is hereinafter referred to as an "abnormal point". A measurement result at an abnormal point is called an "abnormal value" or an "outlier."

==Second state estimator 72==
When it is determined that there is an abnormal value, the second state estimator 72 changes the kernel size of the objective function J(x) and performs state estimation of a plurality of power systems 30 for each kernel size. Specifically, the second state estimator 72 uses a plurality of optimization methods with different conditions (for example, a multipoint search optimization method) for each kernel size, and the value of the objective function J(x) is maximized. Calculate multiple estimation results.

In this case, a multi-point search optimization method is used. may be changed to obtain a plurality of estimation results for each kernel size.

In addition, although the details will be described later, the second state estimator 72 calculates variations in a plurality of estimation results for each kernel size, and selects the kernel size that minimizes the variations. Then, the second state estimator 72 outputs an estimation result of the power system 30 based on a plurality of estimation results calculated with the selected kernel size.

==Example of Process Executed by Information Processing Apparatus 10==
FIG. 5 is a flowchart showing an example of the process S10 executed by the information processing device 10. As shown in FIG. First, the first state estimator 70 acquires the measurement results of the sensors S1 to S4 (S20). Then, the first state estimation unit 70 estimates the state of the power system 30 based on the measurement result and the system model 60 (S21: first state estimation process, first state estimation step).

Here, the first state estimating unit 70 maximizes the value of the objective function J(x) using the above-described colentropy method, the active power P and the reactive power of the consumers 42 to 45, which are the decision variable x. Determine Q. Then, the first state estimator 70 estimates the state of the electric power system 30 based on the defined active power P and reactive power Q using the system model 60 representing the state equation. At this time, the kernel size of the objective function J(x) is the initial value (for example, 0.01) set by the user. As a result, for example, the voltage distribution (solid line) of the distribution line 41 in FIG. 6 is obtained.

Then, the determination unit 71 determines whether there is an abnormality in the measurement results based on the measurement results of the sensors S1 to S4 and the estimation results of the electric power system 30 (S22: determination processing, determination step). Specifically, the determination unit 71 determines whether the difference between the measurement results of the sensors S1 to S4 and the estimation results of the power system 30 corresponding to the positions of the sensors S1 to S4 is equal to or greater than a predetermined value. .

If the difference between the measurement result and the estimation result is less than the predetermined value and it is determined that there is no abnormality in the measurement result (S22: No), the first state estimator 70 converts the estimation result obtained in the process S21 into For example, it outputs to the display device 24 (S23). As a result, the estimation result of the power system 30 is displayed on the display device 24 . Note that the estimation result estimated in the process S21 by the first state estimation unit 70 of the present embodiment corresponds to the "first estimation result".

On the other hand, if the difference between the measurement result and the estimation result is greater than or equal to the predetermined value, it is determined that there is an abnormality in the measurement result (S22: Yes). In addition, in the example of FIG. 6, for example, the difference between the measurement result of the sensor S4 and the estimation result is equal to or greater than a predetermined value. Therefore, here, the determination unit 71 determines that there is an abnormality in the power system 30 (the measurement point of the sensor S4 is the abnormal point).

When the determination unit 71 determines that there is an abnormality in the measurement result (S22: Yes), the second state estimation unit 72 obtains a plurality of estimation results using a multi-point search method for each of a plurality of kernel sizes (S24 ). Specifically, the second state estimation unit 72 of the present embodiment changes the kernel size from the initial value (=0.01) to the maximum value (eg, 0 .1) is changed in 10 steps.

Then, the second state estimator 72 obtains, for example, three estimation results using a multi-point search method for every ten kernel sizes. In the present embodiment, the number of estimation results obtained by the second state estimation unit 72 for each kernel size is three as an example, but the number is not limited to this, and may be four or more.

By the way, when estimating the state of the power system 30, if the kernel size is made very small and the effects of abnormal values are almost eliminated, there will be no values (abnormal values) to be compared with the estimation results. , the accuracy of the estimation result at anomalous points is greatly deteriorated. Further, if the kernel size is made very large and abnormal points are treated in the same way as normal points to estimate the state of the power system 30, the estimation results at the abnormal points are very close to the abnormal values. Accuracy is greatly degraded. Therefore, in the present embodiment, by changing the kernel size, the influence of the abnormal point is considered while being suppressed, and the probable state of the electric power system 30 is estimated.

FIG. 7 is a diagram showing an example of three estimation results with a small kernel size (eg, 0.02) and three estimation results with a large kernel size (eg, 0.08). When the kernel size is small, the influence of anomalous points among the measurement points is generally considered from the colentropy equation (1), but the influence is relatively small. Therefore, in such a case, there is a tendency for the three estimation results to have large variations in the estimation results, particularly at the abnormal point. Note that the variation in the three estimation results when the kernel size is, for example, 0.02 is generally smaller than the variation in the estimation results when no abnormal points are used.

On the other hand, if a large value (for example, 0.08) is used as the kernel size, the influence of abnormal points among the measurement points will generally gradually increase according to Equation (1). Therefore, in such a case, as shown by the two-dot chain line when the kernel size is 0.02, the estimation result may be far from the abnormal point (that is, the actual estimation result may deviate greatly). high estimation results) can be eliminated.

In this way, when the estimation result of the power system 30 is calculated while changing the kernel size, the estimation result when the influence of the abnormal value is completely eliminated, or the estimation result when the influence of the abnormal value is completely considered , it is possible to obtain a more probable estimation result in the actual electric power system 30 . Then, in the present embodiment, the estimation result for the kernel size that minimizes the variation in the estimation result of the abnormal point is assumed to be the probable estimation result in the actual electric power system 30, and the process is executed.

Specifically, the second state estimator 72 calculates variations in estimation results of three abnormal points for every ten kernel sizes, and selects the kernel size that minimizes the variations (S25). For example, out of ten kernel sizes (0.01 to 0.1), if a kernel size of 0.08 has the smallest variation, the kernel size of 0.08 is selected.

Also, the second state estimator 72 selects the estimation result that maximizes the objective function J(x) from among the three estimation results for the kernel size selected in step S25 (S26). For example, in the example of FIG. 7, among the three estimation results (dotted line, one-dot chain line, two-dot chain line), the objective function J(x) of the estimation result indicated by the dotted line is lower than the other two objective functions J(x). If so, the dotted line estimation result is chosen.

Then, the second state estimator 72 outputs the estimation result selected in step S26 to, for example, the display device 24 (S27). As a result, the estimation result of the power system 30 is displayed on the display device 24 . Note that the estimation result output by the second state estimator 72 of the present embodiment in processing S27 corresponds to the "second estimation result". Further, the processes S24 to S27 executed by the second state estimating section 72 correspond to the "second state estimating process" and the "second state estimating step".

<<Other Application Examples of Information Processing Apparatus 10>>
The information processing device 10 is used, for example, when estimating the state of the power system 30, but is not limited to this. For example, instead of the power system 30 that supplies "electricity", it may be used for pipe network analysis to estimate (or analyze) the state of a system (not shown) that supplies "gas" or "water". .

For example, when "gas" is supplied to the system, the transformer 40 in FIG. 2 corresponds to, for example, the gas supply facility, and the distribution line 41 corresponds to the supply pipe that supplies "gas". The consumers 42 to 45 consume the supplied "gas", and the sensors S1 to S4 correspond to measuring instruments for measuring the pressure (for example, gas pressure) of the supply pipe as system information. As in the present embodiment, the information processing apparatus 10 can estimate and analyze the state of the "gas" supply system including these configurations.

Although "gas" has been described here, the same applies to, for example, a piping network for supplying "compressed air" to compressors and the like in factories. Therefore, the information processing apparatus 10 may be used in a system that supplies "gas" including gas and air.

For example, when "water supply (or water)" is supplied to the system, the transformer 40 in FIG. corresponds to the tube. The consumers 42 to 45 use the supplied "water", and the sensors S1 to S4 correspond to measuring instruments for measuring the pressure (for example, water pressure) of the supply pipe as system information. As in the present embodiment, the information processing apparatus 10 can estimate and analyze the state of the "water supply" system including these configurations.

Although "water" has been described here, the same applies to, for example, a piping network that supplies "oil". Therefore, the information processing apparatus 10 may be used in a system that supplies "liquid" including water and oil.

=====Summary=====
The information processing apparatus 10 of the present embodiment has been described above. When there is an abnormality in the measurement results of the sensors S1 to S4, the second state estimation unit 72 maximizes the objective function J(x) for each of a plurality of kernel sizes. A plurality of active powers P and reactive powers Q are searched for. Then, the second state estimator 72 estimates the state of the power system 30 based on variations in estimation results for each of the plurality of kernel sizes. As a result, in this embodiment, the state of the power system 30 can be estimated with high accuracy. In the objective function J(x), the smaller the kernel size value, the smaller the influence of the abnormal value (or outlier) J(x) on the value (evaluation value).

Further, the second state estimator 72 may, for example, select a kernel size from a plurality of kernel sizes that makes the variation of the estimation result smaller than a predetermined value, and use the estimation result of the selected kernel size. However, in general, the smaller the variation in the plurality of estimation results, the higher the possibility that the estimation results in the actual power system 30 are reflected. Since the second state estimator 72 of the present embodiment selects a kernel size that minimizes variation in estimation results, the state of the power system 30 can be estimated with higher accuracy.

Further, when calculating the variation of the estimation result, the second state estimating unit 72 may calculate the variation of the estimation result in a predetermined range (eg, the range from measurement point C to measurement point D) of the distribution line 41, for example. . However, in the present embodiment, the variation in estimation results at a predetermined position (for example, measurement point D) is calculated, so the amount of calculation can be reduced.

In addition, the second state estimating section 72 calculates the variation of the estimation result especially at the measurement point D which is an abnormal point. In general, for example, the variation in the estimation result of the measurement point A, which is far from the abnormal point, does not change significantly even when the kernel size is changed. Therefore, by using the variation of the estimation result at the abnormal point, the variation can be evaluated with higher accuracy.

Also, among the three estimation results when the kernel size is the smallest, for example, the average value of the three estimation results or the median value of the three estimation results (for example, in the case of FIG. 7, the middle value of the three at the measurement point D A certain dotted line estimation result) may be output as the estimation result of the power system 30 . However, in the present embodiment, the estimation result with the maximum objective function J(x) is selected from among the three estimation results, so that a more probable estimation result of the power system 30 can be obtained.

Further, the information processing apparatus 10 of the present embodiment is applied to the state estimation of the electric power system 30, but as described above, for example, the state of the "gas" or "water" system can also be estimated.

The above-described embodiments are intended to facilitate understanding of the present invention, and are not intended to limit and interpret the present invention. Further, the present invention can be modified and improved without departing from its spirit, and it goes without saying that the present invention includes equivalents thereof.

For example, as the system model 60, a model based on the equation of state is used, but it is not limited to this. The system model 60 may be a model that simulates the power system 30 with hardware.

10 information processing device 20 CPU
21 memory 22 storage device 23 input device 24 display device 25 communication device 30 power system 40 transformer 41 distribution lines 42 to 45 consumer 60 system model 70 first state estimator 71 determination unit 72 second state estimator

Claims (8)

showing the state of a system that supplies any one of electric power, gas, and liquid, and based on the plurality of system information measured at each of a plurality of measurement points and a model that simulates the system, a first state estimator for estimating the state of
a determination unit that determines whether or not there is an abnormality in the plurality of system information based on the plurality of system information and the first estimation result estimated by the first state estimation unit;
a second state estimation unit that outputs a second estimation result of estimating the state of the system when it is determined that the plurality of system information is abnormal;
with
The first state estimator,
The smaller the difference between the system information and the first estimation result, the larger the evaluation value, and the smaller the value of the predetermined variable, the smaller the influence of the abnormal value included in the system information on the evaluation value. In the objective function, a decision variable indicating a value corresponding to a change in any one of the above at a predetermined position of the system is changed, and the state of the system is determined using the decision variable when the evaluation value becomes maximum. presume,
The second state estimator,
After obtaining a plurality of estimation results that maximize the evaluation value using an optimization method under a plurality of different conditions for each of the plurality of predetermined variable values, the plurality of estimation results for each of the predetermined variable values outputting the second estimation result based on the variation of the results;
State estimator. The state estimation device according to claim 1,
The second state estimator,
Outputting the second estimation result based on the plurality of estimation results estimated with the value of the predetermined variable that minimizes the variation among the values of the plurality of predetermined variables;
State estimator. The state estimation device according to claim 2,
The second state estimating unit selects a predetermined variable value that minimizes the variation based on variations in the plurality of estimation results at predetermined positions in the system.
State estimator. The state estimation device according to claim 3,
The predetermined position is
It is the position of the measurement point where the system information determined to be abnormal was measured,
State estimator. The state estimation device according to any one of claims 2 to 4,
The second state estimator,
outputting, as the second estimation result, an estimation result that maximizes the value of the objective function J(x) among the plurality of estimation results estimated with the values of the predetermined variables that minimize the variation;
State estimator. The state estimation device according to any one of claims 1 to 5,
The system is a power system that supplies electric power to consumers,
The first and second state estimators are
estimating the state of the system by performing a power flow calculation;
State estimator. to the computer,
showing the state of a system that supplies any one of electric power, gas, and liquid, and based on the plurality of system information measured at each of a plurality of measurement points and a model that simulates the system, a first state estimation process for estimating the state of
a determination process for determining whether or not there is an abnormality in the plurality of system information based on the plurality of system information and the first estimation result estimated in the first state estimation process;
A second state estimation process for outputting a second estimation result of estimating the state of the system when it is determined that the plurality of system information is abnormal;
and
The first state estimation process includes:
The smaller the difference between the system information and the first estimation result, the larger the evaluation value, and the smaller the value of the predetermined variable, the smaller the influence of the abnormal value included in the system information on the evaluation value. In the objective function, a decision variable indicating a value corresponding to a change in any one of the above at a predetermined position of the system is changed, and the state of the system is determined using the decision variable when the evaluation value becomes maximum. presume,
The second state estimation process includes:
After obtaining a plurality of estimation results that maximize the evaluation value using an optimization method under a plurality of different conditions for each of the plurality of predetermined variable values, the plurality of estimation results for each of the predetermined variable values outputting the second estimation result based on the variation of the results;
State estimation program. showing the state of a system that supplies any one of electric power, gas, and liquid, and based on the plurality of system information measured at each of a plurality of measurement points and a model that simulates the system, a first state estimation step of estimating the state of
a determination step of determining whether or not there is an abnormality in the plurality of system information based on the plurality of system information and the first estimation result estimated in the first state estimation step;
a second state estimation step of outputting a second estimation result of estimating the state of the system when it is determined that the plurality of system information is abnormal;
including
The first state estimation step includes:
The smaller the difference between the system information and the first estimation result, the larger the evaluation value, and the smaller the value of the predetermined variable, the smaller the influence of the abnormal value included in the system information on the evaluation value. In the objective function, a decision variable indicating a value corresponding to a change in any one of the above at a predetermined position of the system is changed, and the state of the system is determined using the decision variable when the evaluation value becomes maximum. presume,
The second state estimation step includes:
After obtaining a plurality of estimation results that maximize the evaluation value using an optimization method under a plurality of different conditions for each of the plurality of predetermined variable values, the plurality of estimation results for each of the predetermined variable values outputting the second estimation result based on the variation of the results;
State estimation method.

Images