KR102483130B1 - Server for Collecting Power Usage Data - Google Patents

Abstract

A power usage data collection system according to the present invention includes a power measuring device installed at a power inlet point connected to at least one load device and measuring power usage information of the load device; a power usage analysis server configured to analyze the power usage information for transmitting an operation guide for inducing an operation of the load device; a user terminal interworking with the at least one load device and displaying the operation guide; and a database for storing reference data, wherein the power usage analysis server generates an event data set by detecting a power change event related to the operation indicated in the operation guide when the operation indicated in the operation guide is executed. The data set may be compared with the reference data, and if the event data set is new, it may be registered as new reference data in the database.

Description

Server for Collecting Power Usage Data}

The present invention relates to a power management scheme, and more particularly, to a system capable of managing power use by a registered load device by registering the load device in a power management server, a mobile terminal, and a power management method thereof.

Conventional techniques for obtaining power usage information of individual load devices for power management have been developed in various ways, such as a hardware approach and a software approach.

Since the hardware approach extracts energy consumption information of individual load devices by installing a plurality of individual energy measuring devices or using a plurality of sensors in a distribution panel, there are problems in that installation space is limited and equipment investment is increased.

To solve this problem, a software approach was devised that efficiently extracts the energy consumption information of each load device through cooperation with the server and the measuring device at the power entry point. In other words, a single measuring device collects various data such as current, voltage, and power, and the server interprets and collects the related data, and extracts the energy use information of each load device from the total energy use information. The power usage information calculation method of this approach is called NILM (Non-Intrusive Load Monitoring).

However, since load devices have very different characteristics for each product (for example, even a refrigerator of a certain manufacturer has different energy consumption characteristics depending on the year of release and whether or not there is a water purifier), in order to use NILM, energy consumption data of various individual load devices must be collected. It needs to be continuously acquired.

In order to solve the problems of the prior art described above, an object of the present invention is to provide a power data collection device using a power measurement device installed at a power input point without a separate additional device.

Another object of the present invention is to easily register a load device in a power management server using a power management application installed in a user terminal.

Another object of the present invention is to more accurately grasp the power consumption of each load device by allowing the power management server to register power consumption patterns and device information when registering load devices.

In order to solve the above technical problem, a power usage data collection system according to the present invention includes a power measuring device installed at a power inlet point connected to at least one load device and measuring power usage information of the load device; a power usage analysis server configured to analyze the power usage information for transmitting an operation guide for inducing an operation of the load device; a user terminal interworking with the at least one load device and displaying the operation guide; and a database for storing reference data, wherein the power usage analysis server generates an event data set by detecting a power change event related to the operation indicated in the operation guide when the operation indicated in the operation guide is executed. The data set may be compared with the reference data, and if the event data set is new, it may be registered as new reference data in the database.

In this case, the power usage analysis server may include an event detection unit, a data comparison unit, a recombination unit for each load device, and a user interface.

The data comparator includes an event power comparator, and the event power comparator compares the event data set based on at least one of visual data comparison, cross-correlation, and root mean square error between the event data set and the reference data. It may be determined whether the data is new compared to the reference data.

In addition, the event detector calculates the start and end points of the power change event based on the change in the amount of power within the analysis window while moving an analysis window including a specific time interval among the power usage information received from the power measurement device, The event data set may be generated such that the event data set includes power usage information from the starting point to the ending point.

Also, the operation guide may include an operation of turning on or off any one of the load devices.

In addition, the user terminal may include a power management application.

Also, the power usage analysis server may grant a predetermined reward for registration of the load device.

Meanwhile, a method for collecting power usage information according to an embodiment of the present invention includes requesting, at a user terminal, new registration of a load device connected to a voltage entry point; transmitting an operation guide for the load device to the user terminal interworking with the load device based on a new registration request of the load device connected to the voltage entry point; receiving power usage information from a power measurement device connected to the load device; detecting a power change event from the power usage information; comparing an event data set corresponding to the power data of the power change event with reference data stored in a database; and storing the event data set in the database as new reference data if the event data set is new based on the comparison result.

The comparing of the event data set and the reference data may include comparing the event data set and the reference data based on at least one of visual data comparison, cross-correlation, and root mean square error. .

In addition, the detecting of the power change event may include moving an analysis window including a specific time interval among the power usage information received from the power measurement device, and based on a change in the amount of power within the analysis window, the time point of the power change event and Calculating an endpoint; and generating the event data set so that the event data set includes power usage information from the starting point to the ending point.

Also, the operation guide may include an operation of turning on or off any one of the load devices.

In addition, the user terminal may include a power management application.

The method may further include providing a predetermined reward when the event data set is new and stored in the database.

The present invention seeks to improve the performance of extracting power usage information for individual load devices by acquiring data related to individual load devices from a power measuring device and characteristic information such as model name and operation mode of the corresponding load device from a user without adding a separate device. can do.

1 is a block diagram illustrating a power data collection system according to an embodiment of the present invention.
2 is a diagram showing the configuration of a power usage analysis server according to an embodiment of the present invention.
3 is a diagram illustrating a power data collection procedure according to an embodiment of the present invention.
4 to 6 show screens of a power data collection application according to an embodiment of the present invention.
7 and 8 are diagrams for explaining power data analysis according to an embodiment of the present invention.
9 and 10 are diagrams for explaining power data analysis according to an embodiment of the present invention.

The following merely illustrates the principles of the invention. Therefore, those skilled in the art can invent various devices that embody the principles of the invention and fall within the concept and scope of the invention, even though not explicitly described or shown herein. In addition, it should be understood that all conditional terms and embodiments listed in this specification are, in principle, expressly intended only for the purpose of making the concept of the invention understood, and not limited to such specifically listed embodiments and conditions. do.

The above objects, features and advantages will become more apparent through the following detailed description in conjunction with the accompanying drawings, and accordingly, those skilled in the art to which the invention belongs will be able to easily implement the technical idea of the invention. .

In addition, in describing the invention, if it is determined that a detailed description of a known technology related to the invention may unnecessarily obscure the subject matter of the invention, the detailed description will be omitted. Hereinafter, it will be described with reference to the accompanying drawings.

Hereinafter, a power data collection system according to an embodiment of the present invention will be described with reference to FIG. 1 .

A power data collection system according to an embodiment of the present invention includes a power usage measurement device 100 , a power usage analysis server 200 , a user terminal 300 and a database 500 . Here, the user terminal 300 is a client including an application program, and can be implemented in software or hardware. That is, the user terminal 300 may be an application program installed in a user device such as a mobile phone, tablet, or PC that functions as a client accessing the power usage analysis server 200 described above.

The power usage measuring device 100 is a device for measuring the amount of power consumed through a power entry point using a voltage/current sensor at the power entry point. In this case, the power entry point is a point (node) where power is introduced for a plurality of load devices, such as, for example, a switchboard or distribution board or power strip of each home or building. Therefore, the amount of power measured by the power usage measuring device 100 becomes the sum of the amounts of power used by a plurality of load devices connected to the power inlet point.

In more detail, the power consumption measuring device 100 measures the voltage and current consumed through the power input point using a voltage / current sensor, and uses it to measure various power information such as voltage, current, active / reactive power ( 200) may be transmitted to the power usage analysis server. At this time, the power consumption measurement device 100 measures the amount of power at a sampling rate of several KHz to MHz, and then stores and transmits information at a sampling rate required by the power consumption analysis server 200, for example, at about 5 Hz to 100 Hz. can This is because transmission at too high a data rate may cause a heavy load on the network, and transmission at too low a data rate may be disadvantageous in distinguishing home appliances when analyzing through NILM. Therefore, it is preferable to transmit information of the order of 5 Hz to 100 Hz. For example, the power usage measuring apparatus 100 may process power data sampled at 10 Khz into, for example, 10 Hz data according to settings and transmit the same.

The power usage analysis server 200 may analyze and process data received through the network 400 and transmit the power analysis data to the user terminal 300 interworking with the corresponding power usage measurement device 100 . In addition, the analyzed data may be stored in the database 500 and used as reference data when power data is analyzed in the future.

Referring to FIG. 2 , the power usage analysis server 200 may include an event detection unit 210, an energy calculation unit 220, a data comparison unit 230, and a user interface 240, and a database 400 and It is interlocked to store individual power usage pattern data.

First, the event detector 210 of the power usage analysis server 200 may detect a power change event from power data received in real time. Such a change in power amount can be, for example, turned on/off of a load device (eg, a TV, refrigerator, etc.), a specific part (eg, a defroster of a refrigerator, an outdoor unit of an air conditioner), or a specific mode (eg, a warm mode of a rice cooker). ) is due to the action of

The data comparator 220 compares the power data with power usage data (reference data) stored in a conventional database based on the detected event, and distinguishes and labels which device is using power.

In this case, the data comparator 220 may include an event power comparator 222 and an event pattern comparator 224 .

The event power comparator 222 finds which device generated the power change event detected by the event detector 210 . For example, the event power comparator 222 may find a load device in which an event has occurred based on the amount of change (Δp, Δq) in the amount of active/reactive power. The event power comparison unit 222 may add meta data indicating which device the event occurred in to the event detected by the event detection unit 210 . The function of the event power comparator 222 will be described later in detail.

The event pattern comparator 224 may store and compare pattern data generated according to each device and over time. For example, in the case of a rice cooker, there is a time period in which water is heated until the first water boils to cook rice, and there is a period in which heating is performed at short intervals to maintain the temperature until the rice is cooked after the water boils. After that, there may be a section in which heat is periodically applied for keeping warm. In this way, each load device has a power usage pattern (eg, time when a specific part is turned on/off) according to time corresponding to a specific operation of the load device, and the event pattern comparator 224 describes this pattern data as follows. By comparing the recombined data for each load device, the power consumption for each load device can be distinguished.

The event pattern comparator 24, when the event detector 210 and the event power comparator 222 cannot find an event that occurred in a device, event pattern data (reference data) of individual load devices stored in the database 400. ), it is possible to distinguish the power use of individual load devices from the total amount of power. Meanwhile, the event pattern comparator 24 may update and store event pattern data of individual load devices through machine learning.

The recombination unit 230 for each load device may recombine the power consumption for each load device on the time axis based on the metadata added by the event power comparison unit 222 to calculate the power consumption for each load device.

For example, when an operation of turning on the TV that occurred at time A and an operation of turning off the TV that occurred at time B are discovered through the event detection unit 210 and the event power comparison unit 222, the recombination unit 230 for each load device It is possible to calculate the amount of power of the TV during a specific time by recombining the amount of change in power caused by the operation of turning on the TV and the operation of turning off the TV on the time axis on the axis.

Meanwhile, the user interface 240 may transmit data provided from the power usage analysis server 200 and the calculated amount of power to the user terminal 300 . In this case, the user interface 240 may provide the power usage analysis result to the user terminal through a mobile web page, or to the user terminal through an application program (hereinafter referred to as “app”) operating in the mobile device. may be

Accordingly, the power usage analysis server 200 according to the present invention may distinguish and calculate the power usage of each individual load device connected to the power entry point based on the total amount of power sensed at the voltage entry point through the above-described operations. In addition, the user's convenience may be provided by transmitting the calculated power consumption of each individual load device to the user terminal.

However, an important point in the above-described power consumption analysis is data that is the basis for the event power comparison analysis of individual load devices, for example, on/off data of individual load devices or power amount change data according to specific operations of individual load devices. to be. In order to obtain such data, in the past, an energy measurement device was separately attached to each individual load device, the power value was measured, and then the data was transmitted to the server. Basic data for NILM analysis can be easily obtained.

Hereinafter, a power data collection method according to the present invention, in particular, a method of registering a load device in a power usage analysis server through the user terminal 300 and the power measurement device 100 will be described in detail with reference to FIGS. 3 to 6. do.

According to FIG. 3 , first, the power data collection method according to the present invention starts by running a power management application in the user terminal 300 (S302).

When the power management application is initially run, a login screen as shown in FIG. 4 (a) appears, and the user can request authentication from the server by inputting an ID and password and pressing an OK button, for example (S304).

At this time, the power usage analysis server 200 authenticates the user (S306) and transmits an authentication signal to the user terminal (S308). After user authentication, an initial screen as shown in FIG. 4(b) is displayed. For example, information required by the user, such as real-time power consumption or real-time electricity cost, is displayed on the initial screen.

Meanwhile, as shown in the lower screen of FIG. 4(b), the application program according to an embodiment of the present invention may display an icon 410 for registering a new device. When the user presses the new device registration icon 410, the user terminal 300 may request new device registration from the power usage analysis server 200 (S312).

However, in this embodiment, a new device registration button is displayed on the user interface and the user sends a new device registration request, but the power usage analysis server 200 detects a new event while the app is running on the user terminal. Then, the power usage analysis server 200 may request registration of a new device from the user, receive confirmation from the user, and proceed with a load device registration procedure.

Meanwhile, upon receiving a new device registration request from the user terminal 300, the power usage analysis server 200 may request the user terminal to input information (S314). In this case, the user terminal 300 is shown in FIG. The same information input request screen can be displayed.

At this time, the user can input items requested by the power consumption analysis server 200 in the input window 510, and selects a device to be registered from among a list of multiple devices displayed in a multiple choice method through the selection button 520. You can also select and enter.

After input, the power usage analysis server 200 may display an operation guide for inducing the operation of a device to be registered on the user terminal. For example, as shown in FIGS. 5(b) and 6(a), a request to turn on/off a device to be registered is displayed on a user screen through a user terminal, or a specific mode such as selecting a cook button of a rice cooker A guide leading to entry may be displayed on the user's screen.

In this case, the power usage analysis server 200 may transmit an error message to the user terminal 300 when a power change event is not detected in data through the interlocked power measuring device 100 for a predetermined time.

Meanwhile, when a power change event occurs in data received from the interlocked power measurement device 100 for a predetermined time, the power usage analysis server 200 compares and analyzes the power data (S324). More specifically, the power usage analysis server 200 performs a comparative analysis with the reference data to see whether a device to be registered by the user generates the same power change event as the event data (reference data) stored in the database 500. . Then, as shown in FIG. 6(b), the comparison result is transmitted to the user terminal 300 and displayed on the screen (S326), and the user's confirmation is requested.

In this case, the power consumption analysis server 200 may display more detailed information than the user input part, for example, a detailed model name, based on the previously stored reference data, or may correct and display an incorrect part of the user input according to the comparison result. there is.

Meanwhile, the user checks the data displayed on the user terminal 300 and presses the check button to inform the power consumption analysis server 200 of the confirmation intention. Then, the corresponding data is registered in the database 400.

As a result of the comparison, if a device to be newly registered does not exist in the existing database, user input information is newly registered in the database 400 . In this case, the power usage analysis server 200 may request more detailed additional data than data input through the user terminal 300 from the user.

Through such a user's device registration procedure, the power usage analysis server 200 may collect basic data required for more accurate analysis.

Meanwhile, when such a new device registration procedure is performed, the power usage analysis server 200 may provide a reward for this to the user. For example, the power usage analysis server 200 may provide rewards in various ways, such as cyber money or points.

Hereinafter, an event detection and comparison method according to an embodiment of the present invention will be described in detail with reference to FIGS. 7 to 10 .

As described above, in order to register a new device in the power usage analysis server 200, the power usage analysis server 200 performs an analysis window ( While moving 710 (730), it is detected whether there is a rising change (715) or a falling change (717) of the power amount data.

In more detail, the event detection unit 210 of the power usage analysis server 200 may detect the maximum value and the minimum value within the analysis window 710, and indicate that an event has occurred if the error is larger than a predetermined value. However, the analysis window 710 may check the start and end points of the event while continuing to move until the event is stopped. Then, the event detection unit 210 transmits the event data set 770 including power usage data between the start and end points of the event to the event power comparison unit 222 .

The event power comparator 222 compares the event data 810 , 820 , and 830 in the database 400 with the transmitted event data set 770 to determine which device generated the data.

In this case, data similarity determination may use an algorithm such as visual shape comparison, cross correlation coefficient comparison, and/or root mean square error.

For example, the received event data set 770 may be visually compared with events of individual load devices in the database by artificial intelligence performing deep-running machine learning. The above-described event data set 770 and event data of individual load devices stored in the database 400 are recognized as images, and visual shapes (eg, overshoot lobe shapes, etc.) are compared to determine whether an event occurred in the same load device. can be checked by comparison. (See Fig. 9)

Power change events can be compared in the following process.

1) First, a graph P is drawn with the x-axis as time and the y-axis as active power for the data of individual load devices stored in the database 400, and this is saved as a picture.

2) Then, label the figure (P) with the individual device name (X).

3) Create an automatic image classifier (e.g., a CNN-based classifier) using deep learning. In this case, the input becomes the figure (P) and the output becomes the device name X.

4) For the received event data set 770, a graph (P') with time as the x-axis and active power as the y-axis is drawn, and individual device names (X') for this are entered into the image classifier created in 3) get

5) If the event data set 770 is classified as an individual device existing in the database 400 with a high probability through 4), the event data set 770 is recognized as an event generated in the individual load device recorded in the database 400 do.

Meanwhile, two classifiers can be generated by configuring the same algorithm for reactive power. In addition, recognition as the same load device may be limited to when two classifiers present the same answer.

Alternatively, the cross-correlation between the transmitted event data set 770 and the data stored in the database is obtained, and when the cross-correlation coefficient is within a predetermined range, the two event data are events generated in the same load device. cognition can be compared.

Alternatively, the root-mean-square error of the transmitted event data set 770 and the data stored in the database is obtained, and when the error is within a predetermined range, the two event data are generated from the same load device. You can check whether it is an event or not.

The allowable ranges of the cross-correlation coefficient and root mean square error may be adaptively changed by machine learning according to data accumulation.

The allowable range of the cross-correlation coefficient and root mean square error is adaptively changed in the following way (see FIG. 10).

1) A cross-correlation coefficient (or root mean square) (P) between X and all other individual load device events is calculated for an event X of an individual load device in the database 400.

2) In the case of the cross-correlation coefficient in 1), draw a histogram for 1-abs(P). In the root mean square case, draw a histogram for P.

3) 2) is decomposed into clusters of the number (k) of individual load devices classified in the database 400.

4) Among the clusters decomposed in 3), the center of the cluster closest to 0 is called c. In this case, c becomes, for example, the median value of the cluster.

5) If c' is the center of the next cluster, the allowable range is set to, for example, (c+c')/2.

Therefore, according to the present invention, meta data on power change events of individual load devices in a home or building where a user resides can be easily obtained through the user terminal 300, which is great for power consumption analysis through NILM. This can help.

The above description is merely an example of the technical idea of the present invention, and those skilled in the art can make various modifications, changes, and substitutions without departing from the essential characteristics of the present invention. will be.

Therefore, the embodiments disclosed in the present invention and the accompanying drawings are not intended to limit the technical idea of the present invention, but to explain, and the scope of the technical idea of the present invention is not limited by these embodiments and the accompanying drawings. . The protection scope of the present invention should be construed according to the following claims, and all technical ideas within the equivalent range should be construed as being included in the scope of the present invention.

Claims (13)

When the operation indicated in the operation guide for inducing the operation of at least one load device is executed, a power measurement device installed at the power inlet point connected to the load device detects a change event of power consumption information for the load device and changes the event. an event detection unit generating a data set;
An event power comparison unit registering the event data set as new reference data in the database when it is determined that the event data set is new based on a cross-correlation between the event data set and reference data previously stored in the database or an allowable range of root mean square error. include
Wherein the event power comparison unit machine-learns the allowable range based on cross-correlation between the event data set and the reference data or clustering of root mean square errors.
According to claim 1,
A power consumption information collection server further including a recombination unit for each load device and a user interface.
delete According to claim 2,
The event detection unit,
While moving an analysis window including a specific time interval among the power usage information received from the power measuring device, the start point and end point of a power change event are calculated based on the change in the amount of power within the analysis window, and the event data set is the starting point. and generating the event data set to include power usage information from the end point to the end point.
According to claim 1,
The operation guide includes an on or off operation of any one of the load devices.
According to claim 2,
The user interface transmits an operation guide for inducing an operation of the load device to a user terminal;
The user terminal includes a power management application, interworks with at least one load device, and displays the operation guide.
According to claim 2,
The user interface provides a predetermined reward for registering a load device, the power usage information collecting server.
According to claim 1,
The event power comparison unit,
A power usage information collection server that clusters cross-correlation or root-mean-square error of the reference data using a histogram based on cross-correlation or root-mean-square error between the event data set and the reference data. delete delete delete delete delete

Images