JPH06276700A - Method and apparatus for using error data - Google Patents

Abstract

PURPOSE:To effectively execute computation for estimation by utilizing error data detected in the past in a method and apparatus for estimating power system condition. CONSTITUTION:A power system condition estimation apparatus is additionally provided with an error data storage unit 7 for storing error data detected in the past, a data cause analysis unit 6 for analyzing causes of error data detected in the past, an input data correction unit 8 for correcting data based on the causes of error data and an estimated value correcting unit 9.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a state estimation method and apparatus for estimating the internal state of a power system.

[0002]

2. Description of the Related Art For example, reference [TEDyliacco, State Esti
mation in Control Centers.Electrical Power & Ener
gy Systems, 5 (1983)], etc., when transmitting the measured value from a telemeter to the power supply command station in estimating the state of the power system, noise may be added during transmission,
Wrong data may be included in the observed value due to the deviation of the sampling cycle or the failure of the measuring instrument. These data are once subjected to estimation calculation and then determined as error data by using a statistical method. After that, the state estimation calculation is performed again with the error data removed. In the target power system, state estimation calculation can often be performed normally even if erroneous data is removed when there are many observed values being measured. However, if there are few observed values that are being measured, if error data is removed and an attempt is made to perform state estimation calculation, the estimation calculation may not be possible in a normal state. There are two reasons for this. One is that the redundancy of the system (the ratio of the state variable to be obtained and the observed value) becomes small. Another reason is that the measured observation data that have been removed are the observations that are indispensable to the target system for which estimation calculations are performed.

Up to now, in many cases, in all systems, error estimation cannot be performed to remove the state estimation calculation, and the target power system is divided, and the state estimation calculation is performed for each divided sub-system. There are two problems with this method. One is that a device that divides the target power system into a plurality of sub-systems is required, and furthermore, an estimation calculation must be performed for each of the plurality of sub-systems. This means that the state estimation calculation must be performed from the very beginning, which requires a long calculation time. Another point is that the cause of erroneous data is not clear.
In the operation of a power system that requires high reliability, it is essential to investigate the cause of such error data.
At present, the cause of erroneous data is investigated by humans, and an automatic cause finding method by a computer is desired. Therefore, it is possible to use error data generated in the estimation calculation process. The following are two prior arts that are one of the solutions to this point.

As a method of utilizing error data, Japanese Patent Laid-Open No. 60-
No. 113626, or JP-A-60-176425.
These are methods of utilizing error data generated in the estimation process by inverting or fixing the state information when the switch state information is erroneously recognized. This method does not take measures against the case where the observed data of the power flow value, power injection value, and voltage value of the entire power system contains error information.

[0005]

DISCLOSURE OF THE INVENTION The present invention has a problem of compensating for the lack of redundancy due to error data removal, which is a problem in the above-mentioned prior art, and whether an observed value necessary for executing estimation calculation is removed as error data. We try to solve the problem of preventing the error and the problem of automatically investigating the cause of the error data.

[0006]

[Means for Solving the Problems] The point of compensating for the lack of redundancy, which is the first point of the problems to be solved, and preventing the observation values that are indispensable for executing estimation calculation from being removed as error data. On the other hand, based on the case of error data generated in the past, when the degree of error data is large, the value is corrected to an appropriate value according to the degree of error data, and the estimation calculation function is executed again. When the degree of error is small, the estimation result is corrected based on the case of the error data that occurred in the past, so that the estimation result is converted into an appropriate estimation result. For the second point, automation of cause investigation,
Automatically identify the cause of error data by probabilistic statistical processing and the cause search, such as the situation where past error data occurred, the position of the error data that occurred, the degree of numerical value, the situation around the error data To analyze.

[0007]

In a power system that can normally perform estimation calculation, even if error data occurs, the operation of executing the estimation calculation function and the accuracy of the estimation result are guaranteed in a state close to normal.

[0008]

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

FIG. 1 shows a block diagram of a state estimation device to which the present invention is applied. The present invention comprises the following devices. Data input means 1, input data preprocessing device 2, estimation calculation device 3, error data detection device 4, database storage device 5,
Error data cause elucidation device 6, error data dedicated storage device 7, input data correction device 8, estimated value correction device 9, and control device 1 for controlling the power system using the state estimation result.
0, a result output device 11 that outputs an estimation result.

The data input means 1 is a part for fetching observed values sent from various facilities in the power system through transmission lines, system constants of the power system, and connection states of devices in the power system. The acquired observation value data is sent to the input data preprocessing device 2 through the communication line 101. The input data preprocessing device 2 determines whether or not to perform estimation calculation by dividing the system based on the observed value data of the target system, or performs so-called preprocessing such as reduction of the target system. Here, the processed system constant and observation value data are sent to the estimation calculation device 3 through the communication line 102. The estimation calculation device calculates the state quantity of the power system by performing state estimation calculation using a method based on the generally used weighted least squares method, the extended matrix method, or the linear programming method. The state quantity obtained here and the observed value preprocessed by the input data preprocessing device are sent to the error data detection device 4 through the communication line 103. If no error data is detected by the error data detection device, the result of the state estimation calculation is stored in the database storage device 5 through the communication line 104.
The saved result is sent to the control device 10 through the communication line 113 for use in the next stage of online control and the like. Simultaneously with the storage here, the estimation result is sent to the result output device 11 through the communication line 115, and the estimation result is displayed on the output device represented by the system supervisor. When the error data is detected, the error data is sent to the error data dedicated storage device 7 and the error data cause analysis device 6 through the communication line 105 and the communication line 106, respectively. The error data dedicated storage device 7 stores the error data detected in the past,
The result of analysis by the error data cause analysis device 6 is displayed on the communication line 112.
Sent through. In the error data cause analysis device 6, the error data detected through the communication line 106 and the communication line 1
The past cases taken in 07 are compared with the contents of the error data dedicated storage device 7 in which the past cases are accumulated, and the nature of the error data, for example, due to noise in the transmission line or due to a failure of the measuring instrument, or work Determine if it is due to a power outage. If the result is determined to be artificial due to work power outage or the like, the information is transmitted through the communication line 110 to the estimated value correction device 9
At, the correction amount of the data is determined. Based on this determination, a part of the estimation result of the database storage device 5 used for online control is corrected through the communication line 111. If the determination of the erroneous data is due to a large amount of noise or is caused by another cause, the information is sent to the input data correction device 8 through the communication line 108. When this information enters the input data correction device 8, the past relevant data held in the database storage device is fetched through the communication line 114, and the detected error data is collated with the past data. Based on the result, the input data is converted into a certain value, and the result is converted to the communication line 109.
To the input data preprocessing device 2 through. After changing the data in the input data preprocessing device 2, the estimation calculation is performed again. When the estimation result is corrected by the estimated value correction device 9,
The estimation result is displayed on the result output device 11 through the communication line 116.

Next, the details of each part will be described.

The data input means 1 is a part for taking in the data read by a device for measuring active power, reactive power and voltage value in the power system and sent to this device. The input data preprocessing device 2 reduces and converts the target power system into a model suitable for state estimation calculation. Here, information such as the impedance of the transmission line of the target power system, the in-house load for power generation, and the load at the substation is processed. Also, in order to facilitate estimation calculation of the observation values that have been separated for each tap, such as tap information between busbars, preprocessing such as making a group of observation values is performed. The input data preprocessed in this way is
As shown in FIG. 2, the constants 201 such as the impedance of the power system, the information 202 of the phase adjusting equipment, and the observed values 203 coming from the observer are arranged in this order.

The estimation calculation device 3 is a device for estimating the state of the power system based on the input data such that Kirchhoff's law is consistently established. As a concrete method, a method based on the least square method or the linear programming method is generally used. The procedure for obtaining the solution of the estimation calculation is generally performed as follows. First, a state equation is calculated for each observation value data. Next, the state equations are rearranged in an order suitable for obtaining a solution by a method called a direct solution method. After that, the state of the power system is estimated by the direct solution method. The estimation calculation device spends a considerable amount of time in the total estimation time for rearranging the state equations.

The error data detection device 4 compares the observed value with the estimated value, and if there is a large difference between them, performs statistical processing to detect the data with the largest error as error data. Usually, the χ-square test is generally used for this process, but other statistical methods or other kinds of methods may be used. Here, how to determine the degree of error of the error data becomes a problem. The solution to this problem is to set a threshold value for the error data, and if the error data is detected with an error equal to or greater than the threshold value, it is determined that the error level is large. There are two ways.

In the database storage device 5, for example, FIG.
Save the result of the estimation calculation in the format as in. As an example of the form of storage, for example, regarding the amount expressing the internal state regarding the power transmission line, the power transmission line name 501, the active power value 502 at the A end of the power transmission line, and the reactive power value 50 at the A end of the power transmission line.
3, the active power value 504 at the B end of the power transmission line, the reactive power value 505 at the B end of the power transmission line are saved. For the quantity related to the substation, for example, substation name 506, injected active power value 5
07, injection reactive power value 508, substation bus voltage value 50
Save as 9. The accumulation format of the result of the estimation calculation is changed as needed depending on the purpose and usage of the control device 10 based on the database obtained by the estimation calculation.

The error data dedicated storage device 7 stores data as shown in FIG. The following is an example of the stored information. Error data detection location name 7
01, error data type 702 indicating whether the error data is active power, reactive power, or voltage value, and total error data indicating the frequency of the total error data during the past estimation calculation Occurrence frequency 703, Numerical value 704 indicating the type of error data case detected in the past, System maintenance work flag 705 indicating the status of the generated error data, Flag 706 indicating a normal time, Error in the past A data determination result 707, a degree of matching place 708 for error data, a column 709 indicating a past coping method when error data is matched, and a numerical value 710 indicating the frequency of occurrence of the same type of error data are shown. In the actual device, the data is stored in the format shown in FIG. The parts corresponding to 701 to 710 in FIG. 6 are 701 to 71 in FIG.
This is a part corresponding to 0.

In the error data cause analyzer 6, the communication line 1
The cause of the error data detected through 06 is analyzed based on the database accumulated in the error data dedicated storage device 7, and the analysis result is output. An example of this error data analysis process will be described with reference to the flowchart of FIG.

As an example, it is assumed that the detected error data is detected by an observer measuring the active power of the transmission line A. It is also assumed that the error data was detected during the system maintenance work. Regarding the location of the detected error data, that is, the active power portion of the power transmission line A in this case, it is assumed that a database regarding the error data as shown in FIG. 4 has been created. When the degree of detected error data is large, there are some cases in the past in which there is a complete match as shown in the column of n-5 in FIG. 4, so the error data occurrence frequency 710 for the corresponding case in FIG. Update (FIG. 5602). After that, the total error data occurrence frequency corresponding to the portion 703 in FIG. 4 is updated (FIG. 5).
603). Finally, a command to output the past coping method, correct the input data and perform the estimation calculation again, or terminate the estimation calculation by correcting a part of the estimation result is issued through the communication lines 109 and 110, respectively. It is sent to the data correction device and the estimated value correction device (604 in FIG. 5). Up to this point, there have been cases in the error database where exactly the same cases have existed in the past, but they do not always exist. Therefore, the following will describe a coping method when there is no identical case in the past. For example, suppose that the occurrence status of error data is detected during system maintenance work as in the previous example, and the degree of error data is small. In this way, if there are cases in which the error data generation status is the same, but to a different degree (Fig. 5605), what kind of remedy should be taken in similar cases of past error data generation. Find out if you were broken. The past solution is "correction of input data"
If it is (606 in FIG. 5), a command is sent to the input data correction device 8 to correct the input data and perform the estimation calculation again (607 in FIG. 5). After that, at the time of system maintenance work, a column of occurrence status that the degree of error data is small is added to the corresponding data in the error data dedicated database (FIG. 5).
608). Then, the total error data occurrence frequency is updated (609 in FIG. 5). After that, based on the result of the estimation calculation again, it is determined whether the corrective measure is correct (Fig.
10). When it is determined that the countermeasure is correct, the error data cause analysis device 6 ends the operation as it is. However, when it is determined that it is more appropriate to correct the state variable than to correct the input data, the previously added column of the past coping method for the error data occurrence state is updated (FIG. 56).
13).

In FIG. 5 606, when the past countermeasure is not “correction of input data”, that is, “corresponding state variable is x (%)
If it is “correct”, the degree of similarity between the similar case and the detected error data case is calculated using a certain method. As the calculation method here, a mathematical calculation method represented by the goodness of fit of the fuzzy theory, the degree of coincidence of Dempster-Shafer, or the like, or a calculation method using another model is used. Based on the degree of similarity obtained here, it is determined which state variable and how much to correct (FIG. 5).
611). After that, add case, update error frequency,
The determination and correction of the corrective method of the coping method is the same as in the case where the past coping method is “correction of input data”.

If there is no same case in the past and there is a completely new error data generation situation, an unconditional “correction of input data” measure is output from the error data cause analysis device 6 (612 in FIG. 5). . After that, the procedure for adding a new error data generation status, updating the total error data generation frequency, and determining the correctness of the remedy is the same as that described above.

The operation of the error data cause analysis device 6 up to this point has been analyzed based on the state of occurrence. However, when the amount of information about such error data increases over time, FIG. It is also possible to perform an analysis focusing on the frequency of each occurrence situation in. With this method, it becomes possible to analyze the cause of generation of error data more efficiently. Error data cause analyzer 6 in this case
The flow of operations inside is 601-602-603-604.
It is done in order. By using the error data cause analysis device 6 that pays attention to the frequency of occurrence, it is easy to find a failure or a decrease in accuracy due to secular change of the measuring instrument by looking at this frequency column.

The estimation result output device 11 outputs the presence / absence of error data detection, or when error data occurs, the degree of error data, the measures taken to deal with the error data, and the like. A system supervisor can be considered as an example of the first output device. The system supervisor displays the estimation results, the observation value data obtained by the data input means, and the frequency of past error data. If the occurrence rate of error data is more than a certain frequency, the possibility of failure of the measuring instrument is proposed and the operator is notified through the system supervisor. Also, for the observation values that are indispensable for the estimation calculation, for example, by using a different color from the normal observation values and displaying them on the system controller, changing the font, etc. Let them know what's going on.

Recording on a printer or the like is an example of the second output device. Even when leaving records in this way, the frequency of occurrence of estimation results, observation value data, and error data is recorded as in the case of the system supervisor. When error data frequently occurs in the future, the error data cause analyzer analyzes the measures that should be taken in advance, and records the countermeasures.

Next, the estimation result is corrected based on the information output from the error data cause analyzer. As described above, there are two major methods of modification, as described above. One is when the degree of error data is very large. Specifically, it is the case where it is determined that it is based on maintenance work of the power system or due to a sudden failure of the measuring instrument. In this case, in the input data correction device 8, information of the latest estimation result in the past, or the estimation result closest to the estimation execution time, is transmitted from the database storage device 5 through the communication line 114 to the corresponding data in FIG. Change the data to plausible data in the past by replacing parts or using moving averages at some points in the past. The changed observation value is sent to the input data preprocessing device 2 through the communication line 109, and the estimation calculation is executed again. When performing the estimation calculation here, the structure of the state equation for finding the solution is the same as before the error data was detected, only the corresponding numerical value changed, so the state equation that spends most of the calculation time is Therefore, the state estimation calculation can be executed at high speed.

The other case is when the degree of error is relatively small, which is determined to be caused by the deviation of the sampling cycle of the transmission system or noise. In this case, by utilizing the fact that the internal state of the power system does not fluctuate significantly from moment to moment, when the error data cause analyzer 6 determines that the cause is, the error data cause analyzer 6 surrounds the error data. The amount of change in the estimated value of is also output at the same time. In the estimated value correction device 9, the corrected estimated value is sent to the database storage device 5, and the result of the estimated value corresponding to the corresponding part in FIG. 3 is changed. In this case, since processing that requires a large amount of calculation, such as estimation calculation, is not performed, processing speed is hardly required.

[0026]

According to the present invention, when the state estimation of the power system is executed in the observer system in which error data exists, the processing can be performed faster than before, and the conventional power system division is not performed. Since it is assured that the estimation calculation can be executed at high reliability, the system state can be estimated with high reliability. Further, it becomes possible to monitor the deterioration of accuracy due to the secular change of the measuring device in the power system, and to estimate the internal state of the power system with higher accuracy.

[Brief description of drawings]

FIG. 1 is a block diagram of an error data utilization method and apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing an example of a structure of data preprocessed by an input data preprocessing device.

FIG. 3 is an example of a storage state of estimated calculation results in a database storage device.

FIG. 4 is a diagram showing an example of a method of storing an error data generation case in an error data storage device.

FIG. 5 is a flowchart showing an embodiment in the error data cause analysis apparatus.

FIG. 6 is a diagram showing an image of the error data generation case of FIG. 4 in the apparatus.

[Explanation of symbols]

1 ... Data input means, 2 ... Input data preprocessing device, 3 ...
Estimate calculation device, 4 ... Error data detection device, 5 ... Database storage device, 6 ... Error data cause analysis device, 7 ... Error data dedicated storage device, 8 ... Input data correction device, 9 ... Estimated value correction device, 10 ... Control Device, 11 ... Result output device,
101 ... Communication line for sending system data to input data pre-processing device, 102 ... Communication line for sending pre-processed system data to the estimation calculation device, 103 ... Communication line for sending estimation result to error data detection device, 104 ... Estimation result To the database storage device, 105 ... Communication line for sending the detected error data to the error data storage device, 106 ... Communication line for sending the detected error data to the error data cause analysis device, 107 ... Past error data To the error data cause analysis device, 108, 110 ... Communication line for sending the result of the error data cause analysis device, 109 ... Communication line for sending information of input data to be corrected to the input data preprocessing device, 111 ... A communication line that pretends to be a database storage device with information on estimated results to be
112 ... Communication line for sending error data cause analysis result to storage device dedicated to error data, 113 ... Communication line for sending data used for control device, 114 ... Communication line for sending data in database storage device to input data correction device, 115 ... A communication line that sends the detected error data to the result output device, 116 ... A communication line that sends the data in the database storage device to the result output device.

Claims (3)

[Claims] 1. An input data preprocessing device, an estimation calculation device, an error data detection device including a data input means, an observability determination function, for estimating a power flow value, an injection value and a voltage value of a power system,
A device for displaying and outputting the estimation result obtained by the estimation calculation device, a state estimation method and a device having a database storage device for storing the estimation result, and when error data is detected by using the estimation device, Has a device that analyzes the cause of error data based on the data such as the frequency of occurrence of error data, the status of occurrence, and the degree of error detected during the estimation calculation, and makes an error based on the result from the device. An electric power system state estimation method and an apparatus having a device for correcting data or an estimation result. 2. The power system state estimating method and apparatus according to claim 1, wherein when the error data is corrected or the estimation result is corrected, the information is clearly indicated to the system supervisor. Estimation method and apparatus. 3. The power system state estimation method and apparatus according to claim 1, when the error data is corrected or the estimation result is corrected, the information is output to a printing device such as a printer and detected in the past. A power system state estimation method and apparatus characterized by recording a history of error data.