EA037313B1 - Electrical event detection device and method of detecting and classifying electrical power usage - Google Patents

Abstract

Some embodiments can concern an apparatus configured to detect an electrical state of one or more electrical devices. The one or more electrical devices are coupled to an electrical power infrastructure and generate one or more high-frequency electrical signals on the electrical power infrastructure. The apparatus can include: (a) a processing module configured to run on a processor of a computational unit; and (b) a sensing device configured to be coupled to an electrical outlet. The sensing device can have: (a) a data acquisition receiver configured to receive the one or more high-frequency electrical signals via the electrical outlet and convert the one or more high-frequency electrical signals into one or more first data signals when the sensing device is coupled to the electrical outlet. The electrical outlet can be electrically coupled to the electrical power infrastructure. The sensing device is in communication with the computational unit. The processing module is further configured to identify the electrical state of the one or more electrical devices at least in part using the one or more first data signals. The high-frequency electrical signals comprise electrical signals in the 10 kHz to 3 MHz range. Other embodiments are disclosed.

Description

Cross-reference to related claims

This application is a partial continuation and claims the priority of US Application No. 12/283869, filed September 16, 2008, which claims the priority of US Provisional Application No. 60/973188, filed September 18, 2007. This application claims priority to U.S. Provisional Application No. 61/328122, filed April 26, 2010, and U.S. Provisional Application No. 61/426472, filed December 22, 2010. U.S. Application No. 12/283869 US Provisional Application No. 60/973188, US Provisional Application No. 61/328122 and US Provisional Application No. 61/426472 are incorporated herein by reference.

The technical field to which the invention relates

The present invention relates generally to electrical devices, and in particular to electrical event recording devices and methods for recording and classifying electricity consumption.

Prior art

Many modern approaches to recording and classifying the activity of electrical appliances use a distributed model, in which each electrical device contains a special sensor that monitors changes in the state of the device (for example, turning the device on and off). Determining the device level is conceptually straightforward, but time-consuming and costly to install and maintain. Indirect detection methods have also been used, in which microphones, accelerometers and video cameras are installed throughout the building to determine the activity of electrical appliances. Such methods are effective, but require expensive installation and maintenance, and can also compromise the privacy of a home installation. For example, one method of recording an electrical event involves listening indirectly to switch and motor activity using microphones distributed throughout the living area.

Thus, there is a need or a potential opportunity to benefit from an apparatus or method that can provide detailed information about the electrical status of electrical devices in a home or other building, but which also does not require relatively expensive deployment and professional installation.

The essence of the invention

Some embodiments may relate to apparatus for recording the electrical state of one or more electrical devices. One or more electrical devices are connected to the mains supply and generate high frequency electrical noise on the mains supply. The hardware may comprise (a) a processing module designed to operate on a processor of the computing unit; and (b) a sensor for connecting to an electrical outlet. The sensor may include (a) a data acquisition receiver for receiving high frequency electrical noise through an electrical outlet and converting the high frequency electrical noise into one or more first data signals when the sensor is connected to an electrical outlet. The electrical outlet can be electrically connected to the mains supply. The processing module is also designed to identify the electrical state of one or more electrical devices, at least in part using one or more of the first data signals. The sensor is connected to the computing unit. High frequency electrical noise refers to electrical noise in the frequency range 10 kHz to 3 MHz.

Other embodiments relate to a method for recording and classifying electricity consumption by one or more electrical devices. One or more electrical devices are connected to the power line. The method can include capturing two or more electrical signals on a power line, the two or more electrical signals comprising substantially continuous electrical noise; recording the occurrence of one or more electrical events on the power line using at least a substantially continuous electrical noise of two or more electrical signals; and associating one or more electrical events with a change in the electrical state of at least one of the one or more electrical devices. The substantially continuous electrical noise of two or more electrical signals includes electrical signals that can be identified on the power line over a duration of more than about 1 second.

In addition, other embodiments relate to an electrical event recording device for recording two or more electrical events on a building power line. An electrical event recording device may comprise (a) a receiving module for receiving and processing one or more electrical signals, where one or more electrical signals contain a high-frequency component, and the receiving module contains (1) an electrical interface for communication with a network line power supply; (2) one or more filter circuits connected to an electrical interface and designed to transmit one or more parts of one or more electrical signals; (3) converter module connected to the output of one or more filters

- 1 037313 control circuits and designed to convert one or more electrical signals into one or more data signals; wherein one or more data signals contain information about the high frequency component of one or more electrical signals; (b) a processing module for operating on a processor, the processing module comprising (1) an event registration module for using information about the high frequency component of one or more electrical signals to determine the occurrence of two or more electrical events; (2) a classification module designed to classify two or more electrical events; (3) a preparatory module for correlating the first type of event with the first event of two or more electrical events and for correlating the second type of event with the second event of one or more electrical events. Two or more electrical events include the activation of one or more electrical devices connected to the building power line and the activation of one or more electrical devices connected to the building power line. The high-frequency component of one or more electrical signals contains electrical signals with a frequency of more than 10 kHz.

List of figures of drawings and other materials

To facilitate further description of the embodiments, the following drawings are provided, in which FIG. 1 is an example of a general diagram of a typical electrical event recording device according to the first embodiment;

in fig. 2 is a block diagram of the electrical event recording device shown in FIG. 1 in accordance with the first embodiment;

in fig. 3 is an exemplary frequency band cascade diagram showing turning on and off electrical devices in accordance with an embodiment;

in fig. 4 is an incomplete connection diagram of a typical data acquisition receiver of the electrical event recorder shown in FIG. 1 in accordance with the first embodiment;

in fig. 5 is a general diagram of a typical electrical event recording device according to a second embodiment;

in fig. 6 is a block diagram of a sensor of the electrical event recording device shown in FIG. 5 in accordance with a second embodiment;

in fig. 7 is a flowchart of an exemplary embodiment of a method for providing an electrical event recording device in accordance with an embodiment;

in fig. 8 is a block diagram of an exemplary embodiment of a method for recording and classifying electricity consumption by one or more electrical devices connected to a power supply network, in accordance with an embodiment;

in fig. 9 is a flowchart of an exemplary operation of using first electrical signals to prepare a processing unit for correlating first data signals with a change in electrical state of a particular electrical device in accordance with an embodiment;

in fig. 10 is a plot of a typical fluctuation noise signature in a building in accordance with an embodiment;

in fig. 11 is a graph of a typical noise signature of a new device in accordance with an embodiment;

in fig. 12 is a graph of a typical noise signature of a new electrical device after removal of fluctuation noise in accordance with an embodiment;

in fig. 13 is a demographic table of buildings used for a typical installation of a typical electrical event recorder in accordance with an embodiment;

in fig. 14 is a table showing the performance of a typical electrical event recording system during a typical installation, in accordance with an embodiment;

in fig. 15 is a confusion matrix table showing the performance of a typical electrical event recording system during a typical installation in accordance with an embodiment;

in fig. 16 depicts another table summarizing the classification accuracy metrics using the minimum set of preparatory data of a typical system for recording an electrical event during a typical sweep in accordance with an embodiment;

in fig. 17 is a table showing the performance of a typical event recording system in classifying four electrical devices applied in various buildings using a 10-fold classification check, in accordance with an embodiment;

in fig. 18 is a graph showing temporal stability or signature variation over time for four randomly selected electrical devices by visualizing feature vectors in feature space in accordance with an embodiment;

- 2 037313 in Fig. 19 is a graph of a spectrum observed by an electrical event recording system showing the spectrum of noise generated by four CFLs (compact fluorescent lamps) of the same type, in accordance with an embodiment;

in fig. 20 is a plot of the spectrum of a portion of data from a test of an electrical device that was connected to two different wall outlets in a building, in accordance with an embodiment;

in fig. 21 is a graph of an EMI signal generated by a dimmer at different levels in accordance with an embodiment;

in fig. 22 is a graph showing the short burst of electromagnetic interference that CFLs produce when first turned on;

in fig. 23 depicts a computer that is suitable for implementing the embodiment of the computing unit shown in FIG. one;

in fig. 24 is a representative block diagram of an example of elements included in wiring diagrams within the casing of the computing unit of FIG. 23.

For simplicity and clarity of illustration, the figures depict a general view of the structure, and descriptions and details of well-known features and techniques may be omitted to avoid unnecessarily obscuring the invention. In addition, elements in the figures are not necessarily drawn to scale. For example, the dimensions of some of the elements in the figures may be exaggerated relative to other elements for a better understanding of embodiments of the present invention. The same reference numbers in different figures correspond to the same elements.

The terms first, second, third, fourth, and the like in the description and in the claims, if any, are used to distinguish between similar elements and are not required to describe a particular sequential or chronological order. It should be understood that the terms used in this manner are interchangeable under appropriate circumstances, such as the embodiments described herein, for example, may be used in sequences other than those depicted or otherwise described herein. In addition, the terms contain and have, and any variations thereof, are intended to denote a non-exclusive inclusion, such as a process, method, system, object, device, or apparatus that constitutes a list of items, not necessarily limited to these items, but may contain other items that are not are indicated in a direct form or are not inherent in such a process, method, system, object, device or apparatus.

The terms left, right, front, back, top, bottom, above, below, and the like in the description and in the claims, if any, are used for descriptive purposes and not necessarily to describe a permanent relationship. It should be understood that the terms used in this manner are interchangeable under appropriate circumstances, such as the embodiments of the invention described herein, for example, may be used in orientations other than those depicted or otherwise described herein.

The terms connection, connected, connection, connecting and the like should be understood in a broad sense and refer to the connection of two or more elements or signals electrically, mechanically and / or in another way. Two or more electrical elements may be electrically connected, but not mechanically or otherwise connected, two or more mechanical elements may be mechanically connected, but not electrically or otherwise connected, two or more electrical elements may be mechanically connected, but not electrically connected or in another way. The connection can be any time, for example, permanent, semi-permanent, or instantaneous.

Electrical connection and the like are to be understood broadly and include a connection in the presence of any electrical signal, be it a power signal, a data signal, and / or other types or combinations of electrical signals. Mechanical connection and the like are to be understood broadly to include any type of mechanical connection.

The absence of the word removable, removable and the like next to the word connected and the like does not mean that the connection in question, etc. is or is not removable.

Information confirming the possibility of carrying out the invention

Electricity in the power line may contain electrical noise. Electrical noise present on the power line can be caused by the operation of an electrical device that is electrically connected to the power line. This type of electrical noise is called conducted electromagnetic interference (EMI). EMI can be divided into two types: transient noise and continuous noise. In some embodiments, continuous or transient electrical noise that occurs when an electrical device is turned on does not have the same form of electrical noise after several AC electrical cycles (eg, one AC electrical cycle is 1/60 s in the United States). For example, the electrical noise of a compact fluorescent light bulb (CFL) has one shape for several AC electrical cycles until the CFL heats up and then the shape of the electrical noise changes to

- 3 037313 different shape after CFL warming up. In another example, DC motors are characterized by continuous noise, but the continuous noise of the DC motor can only last for microseconds, but can repeat every AC electric cycle as long as the DC motor is running. The electrical event recorder described below can record all of these types of electrical noise.

Short-term noise is characterized by a short duration, during which it can be observed, as a rule, from tens of nanoseconds to several milliseconds. On the other hand, continuous noise (i.e., substantially continuous noise) can be observed as long as the electrical device is in operation. In many embodiments, continuous noise, as used herein, can mean repetitive, constant, uninterrupted, or repetitive noise. In the same or other embodiments, the noise can be continuous if the noise spectrum repeats every AC cycle or an electrical noise signal is observed without interruption as long as the electrical device is in operation. Noise can still be considered continuous noise if AC cycle break occurs during noise.

In many examples, continuous electrical noise can be sensed on a power line for a longer period of time than an alternating current electrical cycle. In another example, continuous electrical noise may be detectable for less time than one AC cycle, but electrical signals are repeated in three or more AC electrical cycles. In another example, continuous electrical noise can be electrical signals that are recognizable on the power line for more than about 10 ms. In another example, continuous electrical noise can be electrical signals that are recognizable on the power line for more than about 50 ms. In other examples, continuous electrical noise may be electrical signals that are recognizable on the power line for more than about 1 second. In still further examples, continuous electrical noise can be electrical signals that are recognizable on the power line for more than about 10 seconds.

Both short-term and continuous noise can also be concentrated in a narrow frequency range or spread over a wide frequency range (ie, be broadband noise). CFL is an example of an electrical device that generates continuous noise that is conducted through the power line due to its electrical connection to the power line. Since the building's electrical distribution system is connected in parallel to the building's electrical panel, conductive EMI is widely distributed from a given electrical device throughout the building's power supply line.

Continuous noise is usually inherent in a working electrical device and internal electronics. Devices such as a grinding machine create electrical noise that is synchronous with the AC mains frequency (60 Hz in the US) and its harmonics (120 Hz, 180 Hz, etc.) due to the continuous closing and opening of the electrical contact by the motor bushings of this type of electrical devices.

A switching power supply (SMPS) is a power supply that contains a switching regulator to provide the output voltage required by electrical devices using SMPS. The function of the SMPS is to provide a stabilized output voltage, usually different in level from the output voltage obtained from the mains supply. Electrical devices using SMPS are becoming more common due to their higher efficiency, smaller size and lower cost compared to traditional power supplies. In addition, manufacturers are increasingly using SMPS in their products to meet minimum energy efficiency requirements (for example, the Energy Star Program of the US Department of Energy). For example, most personal computers as well as fluorescent lamps currently use SMPS. More than a decade ago, most consumer electronics devices were not SMPS due to the underdeveloped state of SMPS manufacturing technology and the lack of cheap single-chip SMPSs.

Modern SMPS-based electrical devices generate noise that is synchronized with their internal power supply. In addition, unlike traditional linear power regulators, the SMPS does not dissipate excess power in the form of heat, but instead stores energy with an inductor and transfers that stored energy from the power line to the load as needed and thus wastes much less energy. than traditional power supplies. The key to the smaller and more efficient SMPS is the use of a powerful transistor to transfer the stored electrical energy at a high frequency, also known as the switching frequency. The switching frequency is usually much higher than the 60 Hz AC mains frequency (in the US) because the higher frequency requires smaller inductors or converters. A typical SMPS operates at frequencies from

- 4037313 tens to hundreds of kilohertz (kHz). The switching circuit can be adjusted to meet the power requirements of the electrical device that is equipped with the SMPS. For example the source

The CFL is equipped with an SMPS to generate the high voltage required by the CFL. The switching action, which is the cornerstone of the SMPS principle, generates more EMI focused on the switching frequency.

In addition, most modern consumer electronic devices are moving towards using a program switch instead of a mechanical switch. Unlike a mechanical switch, a program switch uses a software-controlled toggle button that electronically transfers electrical power to an electrical device. In this type of switch, the indirect activation of an electrical device using a software-controlled electronic switch reduces the transient noise generated at the moment of activation. It has been found that several devices, such as LCD monitors (liquid crystal monitors) and DVD players, that use a softswitch do not generate noticeable intermittent noise. Software-controlled devices such as LCD monitors and DVD players are almost always SMPS-based and thus the electrical event recorder 100 (FIG. 1) can register changes in their power state by monitoring the electrical EMI devices produced.

Like modern electronic devices based on SMPS, dimmers also generate continuous noise due to the switching of at least one internal two-way thyristor switch. This continuous noise can be used to register and identify loads from a dimmer controlled incandescent lamp. In contrast to the narrowband noise produced by the SMPS, the dimmer produces broadband noise in the hundreds of kilohertz range, which can be modeled as a large deviation Gaussian distribution.

In addition, in the United States, the Federal Communications Commission (FCC) has established regulations (for example, Title 47 CFR Part 15/18: Consumer Exposure Limits) for any electrical device connected to a utility power line that specifies the maximum amount of EMI. which the electrical device can transfer back to the power line. The FCC limit is still 66 dBμV (decibels per microvolt) for a frequency range of 150 to 500 kHz, which is about -40 dBm (decibels per milliwatt) into a 50 ohm load.

Now let's turn to graphic materials. FIG. 1 is a general diagram of an exemplary electrical event recording apparatus 100 in accordance with the first embodiment. FIG. 2 is a block diagram of an exemplary system 200 including an electrical event recording apparatus 100 in accordance with a first embodiment. In some embodiments, electrical event recorder 100 may be configured to register the electrical state of one or more electrical devices 290 (FIG. 2) in system 200. Electrical event recorder 100 may also be configured to register one or more electrical events on a network line power 150 to a building in system 200. In many examples, electrical event recorder 100 may utilize continuous noise on power line 150 due to electrical devices 290 to record the electrical state of electrical devices 290 or to record electrical events on power line 150. The system 200 and the electrical event recorder 100 are exemplary only and not limited to the embodiments presented herein. Electrical event recorder 100 may be used in many different embodiments or examples not shown or depicted specifically in this application.

In some examples, electrical device 290 generates one or more high frequency electrical signals. In various embodiments, high frequency electrical signals (for example, electrical signals in the frequency range from 10 kHz to several megahertz) can be generated by one or more SMPS or other electrical components (for example, internal bi-directional thyristor switches or internal generators (other than SMPS)) of electrical devices 290. In these same or other embodiments, the high frequency electrical signals may be cyclic with respect to the AC electrical cycle of the power supply network 150. High frequency electrical signals can be cyclostationary signals in a broad sense. In some examples, high frequency electrical signals may exhibit cyclostationarity in the second, third, or fourth component of the rank statistics.

In many embodiments, the high frequency electrical signals are electrical signals in the frequency range from 10 kHz to 3 MHz. In other embodiments, the high frequency electrical signals are high frequency electromagnetic interference in the frequency range from about 10 kHz to about 1 MHz. In other embodiments, the high frequency electrical signals are high frequency electromagnetic interference in the 5,037313 frequency range of about 30 kHz to about 300 kHz.

In the same or other embodiments, the high frequency electrical signals are electrical signals with a frequency greater than about 10 kHz. In some embodiments, the electrical event recorder 100 may be sensitive enough to capture continuous noise of -100 to -10 dBm over a frequency range of about 10 kHz to about 3 MHz, or more specifically about 10 kHz to about 1 MHz.

In some examples, one or more electrical events recorded by the electrical event recorder 100 may comprise actuating and deactivating (i.e., turning on and off) electrical devices 290 connected to the power line 150. One or more electrical events may also be events in which the amount of electrical power supplied by a power supply of an electrical device to another electrical device (eg, turning a dimmer knob) is varied or limited. As used herein, each of the electrical devices 290 can have one of three power states: (a) a power on state; (b) a waiting state; (c) a completely power-off state. The power-on state includes all power states when the electrical device is turned on and the electrical device's power consumption exceeds the rated power (i.e., it is not in a standby or completely powered-off state).

A standby state is an electrical state where an electrical device is nominally off but still consuming power to provide one or more of the electrical device's functions. Those. Standby electrical devices are nominally off, but still consume power to provide one or more standard, continuous, or permanent functions. For example, after a user turns off a video recorder (VCR) or digital video recorder (DVR), the VCR and DVR may continue to draw power from the power line 150 to illuminate one or more displays on the device (such as a clock or one or more LEDs ( LEDs)) and / or performing one or more internal functions. In this case, although the user thinks the VCR or DVR is off, the VCR or DVR is actually in standby mode. In another example, after the user has turned off the electrical device, the electrical device can continue to consume power to charge the internal battery and thus be in a standby state.

The power off state is a state where the electrical device does not consume any power (i.e., there is actually zero power consumption) from the power line 150. In the case of a VCR or DVR that draws power even after the user has turned off the electrical device, the VCR or DVR can be put into a completely off state by disconnecting the electrical device from the mains 150 power line or connecting the VCR or DVR to an electrical switch that completely stops power consumption of VCR or DVD.

As used in this application, power-on and similar phrases refer to putting an electrical device into a powered-on state, either from a fully powered-off state or from a standby state. Just as used in this application, shutdown and similar phrases refer to the transfer of an electrical device from a power-on state, to a completely power-off state, or to a standby state. In addition, the power off state and similar phrases refer to either a completely power off state or a standby state. In addition, the power-on state and similar phrases refer to the power-on state.

In some examples, which will be discussed below in relation to dimmers and televisions, the electrical event recorder 100 may also register intermediate states of electrical devices 290. electrical event recorder 100 may record various power-on states of electrical devices 290.

FIG. 3 is an exemplary frequency band waterfall diagram 300 showing on and off electrical devices in accordance with an embodiment. As shown in FIG. 3, if a typical electrical device is turned on, a narrowband continuous noise signature is observed that is continuous during operation of the device. In addition, in FIG. 3 it can be seen that the noise has the greatest intensity in the center of the noise (for example, at the switching frequency of the SMPS of electrical devices) and then propagates to the lower and upper frequencies with attenuation of the intensity. Intensity decay can be roughly modeled using a Gaussian function whose mathematical expectation is located at the switching frequency. This allocation can be related to the error tolerance of the components that make up the core of the switching circuit, as well as the load characteristics of the power supply. If all electrical devices and their components were ideal, then a single narrowband signal peak at the switching frequency would be visible in FIG. 3. The error tolerance of SMPS components is also

- 6 037313 may differ between other identical devices, such as different blocks of the same CFL model. Finally, the power line itself can perform a transmission function (i.e., change the inductance between the source of information being collected and the electrical device) and can add further differences to a variety of such electrical devices.

Returning to FIG. 1-2, the electrical event recording device 100 may comprise (a) at least one sensor 110 for connecting to at least one electrical outlet 151 of a power line 150 (i.e., a power line in a building); and (b) at least one computing unit 120. In some embodiments, the electrical event recorder 100 does not include a power line 150, an electrical outlet 151, or electrical devices 290. In various embodiments, the electrical event recorder 100 also does not include a computing unit 120. In some examples, electrical event recorder 100 includes a processing unit 222 (FIG. 2), but does not include a computing unit 120.

The sensor 110 may comprise (a) at least one receiving module or receiver 211 for collecting data; (b) controller 215; (c) a transmitter communication device 216; (d) a power supply 217 for providing electrical power to the data acquisition receiver 211, the controller 215, and the communication device 216. Computing unit 120 may comprise (a) receiver communication device 221; (b) a processing unit 222; (c) a data storage unit 230.

Without considering the following as limiting, a simple example of using the electrical event recorder 100 comprises electrical devices 290 (FIG. 2) generating one or more high frequency electrical signals (eg, EMI) on a power line 150. The sensor 110 can detect high frequency electrical signals (eg, continuous noise) on the power line 150 and generate one or more data signals that contain information about the high frequency electrical signals. Sensor 110 may transmit data signals to computing unit 120 using wired and / or wireless communications. Computing unit 120 can identify the electrical state of electrical devices 290 using data signals at least in part.

The data acquisition receiver 211 may be configured to receive and process one or more electrical signals from the power line 150. Electrical signals can contain high frequency components (eg EMI). Those. the data acquisition receiver 211 may be configured to receive electrical signals with high frequency components and convert the electrical signals, and in particular a high frequency component, into one or more data signals.

FIG. 4 is an incomplete schematic diagram of a typical data acquisition receiver 211 in accordance with the first embodiment. As shown in FIG. 2 and 4, in various embodiments, the data acquisition receiver 211 may comprise (a) at least one electrical interface 212 for connecting to an electrical outlet 151 (FIG. 1) of a power line 150; (b) one or more filter circuits 213; (c) at least one converter module 214. In various embodiments, electrical interface 212 may include a two-pin or three-pin power connector.

In some examples, filter circuits 213 may be electrically connected to electrical interface 212 and are designed to filter portions of incoming electrical signals from the mains supply. The filter circuits can be designed to transmit high frequency electrical noise. For example, the data acquisition receiver 211 can filter the AC mains frequency (60 Hz in the US) such that the converter module 214 is not overloaded by the 60 Hz frequency reference. In the same or other examples, the filter circuits 213 may comprise a high-pass filter. In some embodiments, the high-pass filter may have a substantially flat frequency response in the 50 kHz to 30 MHz frequency range. A 3 dB kink in the high-pass filter can be observed at 36.7 kHz. A 3 dB kink gives a wide enough bandwidth to observe the full range of EMI and other high frequency continuous noise in electrical signals. In some examples, the filter circuits 213 may also include a 10 dB attenuator such that a constant 50 ohm load is present at the input of the receiver 211 for data acquisition regardless of signal frequency or AC line condition. In addition, high voltage capacitors may be used in some examples for safety and line voltage isolation. The same or other examples may use the polarity shown in FIG. 4. Ie in one embodiment, the line and zero lines should not be reversed and the isolation capacitors should be AC polyester film capacitors for safety reasons.

In some examples, the filter circuit may include capacitors 461, 462 and resistors 463, 464, 465. Capacitors 461 and 462 may be 0.1 μF (450 volt polyester) capacitors. Resistors 463 and 465 may be 100 ohm 1 watt resistors. Resistor 464 may be a 75 ohm resistor with a power

- 7 037313 in 1 W.

Converter module 214 may be electrically connected to filter circuits 213 and configured to receive the filtered signal from filter circuits 213. Converter module 214 may be configured to convert one or more filtered signals to one or more data signals. One or more data signals may contain information about the high frequency component of one or more electrical signals. In some examples, the converter module 214 may include an analog-to-digital converter. In some examples, the analog-to-digital converter may sample the filtered electrical signal at a specified frequency (eg, 1 MHz). In one example, the converter module 214 may comprise a USRP (Universal Software Radio Peripheral).

In some examples, communications apparatus 216 may comprise a wireless transmitter and communications apparatus 221 may be a wireless transmitter. In some examples, electrical signals may be transmitted using WI-FI (wireless communication), IEEE (Institute of Electrical and Electronics Engineers) 802.11 wireless protocol, or Bluetooth 3.0 + HS (high speed) wireless protocol. In additional examples, these signals can be transmitted over Zigbee (802.15.4), Z-Wave, or a custom wireless standard. In other examples, communications device 216 may transmit electrical signals using cellular communications or a wired connection (eg, using a wire).

In North America and many other parts of the world, three-phase power is commonly used in most buildings. In some buildings, a strong noise signal may be detected in the phases of the power line 150. However, for some parts of some buildings that are not in phase with the electrical outlet 151 to which the electrical interface 212 is connected, it can be difficult to detect electrical events. In some examples, this problem can be solved by installing a sensor 110 on each phase (i.e. installing two sensors for a building using a three-phase power supply) or installing the sensor 110 in an available 240 V outlet if both phases are present (for example, in an electrical dryer connector). Installing two sensors 110 will allow the electrical event recorder 100 to capture events in both phases, but will also increase the likelihood of similar signatures occurring for two similar devices. However, this kind of signature problem can be solved by knowing which of the two sensors 110 has registered the event.

In some examples, processing module 222 may include software and may comprise (a) a prep module 223 for correlating one or more electrical events with one or more electrical events of different types; (b) an event registration module 224 for detecting the occurrence of one or more electrical events using data signals at least in part; (c) a classification module 226 for classifying the type of one or more electrical events; (d) a communication module 225 for communicating with a user.

Computing unit 120 or processing unit 222 may be configured to identify the electrical state of electrical devices 290 using at least in part one or more first data signals (eg, information about the high frequency component of one or more electrical signals). In various embodiments, computing unit 120 and / or processing unit 222 may be configured to measure the amount of electricity consumed by one or more electrical devices, at least in part using one or more data signals from communication device 216. For example, processing unit 222 may be designed to identify when the first of the one or more electrical devices 290 is turned on or off, at least in part using data signals from the communication device 216.

As will be described in more detail below, in some examples, in addition to data signals received from sensor 110, processing unit 222 may use the following types of data to identify the electrical state of electrical devices 290: (a) data from a database from a control; (b) data from one or more databases that contain data about previously observed data signals; (c) data from one or more brands of electrical devices; and / or (d) data from the user regarding the identification of one or more electrical devices.

As will be described in more detail below, the event registration module 224 is designed to determine the occurrence of one or more electrical events, at least in part using data signals from the communication device 216. Classification module 226 may be configured to identify a specific electrical event on a line of power supply network 150. Classification module 226 may be configured to correlate electrical events on power line 150 with changes in power state of specific electrical devices.

The preparatory module 223 may be configured to correlate specific electrical signals on the power line 150 with specific electrical events. For example, preparatory module 223 can be designed to determine the suitability of a particular

- 8 037313 electrical event on the line of the power supply network 150 turning on or off a specific electrical device (eg, a fluorescent lamp, computer, or washing machine).

In some examples, preparation module 223 may be configured to perform a preparation or calibration sequence for correlating electrical events detected by data acquisition receiver 211 with specific electrical events. After performing the calibration sequence, the preparatory module 223 can transmit the preparatory correlation data to the classification module 226 such that the classification module 226 can correlate electrical events detected by the data acquisition receiver 211 with a specific change in the state of a particular electrical device. Typical prep or calibration sequences are described in relation to the operation 835 of method 800 depicted in FIG. eight.

The communication module 225 can be used to transmit and receive information from one or more electrical event recording devices 100. For example, the user can use the communication module 225 to enter information during the execution of the preparation or calibration sequence. In addition, the communication module 225 can inform the user of the occurrence of an electrical event. In some embodiments, communications module 225 may utilize monitor 2306, keyboard 2304, and / or mouse 2310, as depicted in FIG. 23.

The data storage module 230 may store information and data used by the preparatory module 223, the event logging module 224, the communication module 225, and the classification module 226. In some examples, storage module 230 may include a USB device connected via USB port 2312 (FIG. 23), CD-ROM drive or DVD in CD-ROM drive and / or DVD drive 2316 (FIG. 23), hard disk 2314 (Fig. 23) or memory 2408 (Fig. 24).

Processing unit 222 may be configured to operate on a processor (e.g., Central Processing Unit (CPU) 2410, as depicted in FIG. 24) of computing unit 120. Computing unit, as used herein, may mean a single computer, a single server, a group, or a series of computers and / or servers located in one or more locations. In one example, computing unit 120 may comprise the computer of FIG. 23 and 24. In some examples, computing unit 120 may be located at least partially adjacent to the user. In other examples, users may access computing unit 120 over the Internet or other networks.

In some examples, computing unit 120 may be the first server. The first server may be the home computer of the user of the electrical event recorder 100, or a computer owned or operated by the building owner in which the electrical event recorder 100 is installed. In other examples, the first server may be another electrical device (with a processor) located in the building (eg, home control or automated control system, security system, air conditioning system). The first server may include a first portion of a communication device 221, a storage module 230, a preparation module 223, an event registration module 224, a communication module 225, and a classification module 226. One or more second servers (eg, a computer or server owned or operated by the manufacturer or distributor of electrical event recorder 100, utility, or security company) may contain a second, possibly redundant, portion of these modules. In these examples, computing unit 120 may comprise a combination of a first server and one or more second servers.

FIG. 5 is a general diagram of an exemplary electrical event recording apparatus 500 in accordance with a second embodiment. FIG. 6 is a block diagram of a sensor 510 of an electrical event recording device 500 in accordance with a second embodiment. In some embodiments, electrical event recorder 500 may be configured to record the electrical state of one or more electrical devices 290 (FIG. 2) of system 501. Electrical event recorder 500 may also be configured to register one or more electrical events on a network line 150 1 and 5) of system 501. In many examples, electrical event recorder 500 may use continuous noise on power line 150 due to electrical devices 290 (FIG. 2) to record the electrical state of electrical devices 290 or to register electrical events in power supply network 150 lines. The electrical events recorded by the electrical event recorder 500 may comprise actuating and deactivating (ie, turning on and off) the electrical devices 290 (FIG. 2) connected to the power line 150. One or more electrical events may also be events in which the amount of electrical power supplied by the power supply of the electrical device to other electrical devices (eg, turning a dimmer knob) is varied or limited. The system 501 and the electrical event recorder 500 are exemplary only and are not limited to the embodiments presented herein. System 501

9 037313 and the electrical event recorder 500 may be used in many different embodiments or examples not shown or described specifically in this application.

As shown in FIG. 5 and 6, electrical event recorder 500 may comprise (a) a sensor 510; (b) a hub 540 for connecting to an electrical outlet 151 of a power line 150; (c) a router 505 for connecting the hub 540 to the computing unit 120; (d) computing unit 120. In other examples, hub 540 may be connected to computing unit 120 without using router 505, in which case electrical event recorder 500 and system 501 do not include router 505.

In some examples, sensor 510 is coupled to at least one electrical box 552 and utility wattmeter 553. In other examples, sensor 510 may be integrated with or be part of at least one electrical box 552 or utility wattmeter 553. In one embodiment, recorder 500 electrical event does not include an electrical box 552, utility meter 553, utility line 150, electrical outlet 151, or electrical devices 290. In another embodiment, electrical event recorder 500 also does not include computing unit 120 or router 505. In some examples, recorder 500 electrical event may include a processing unit 222, but not a computing unit 120.

Without considering the following as limiting, a simple example of using the electrical event recorder 500 comprises electrical devices 290 (FIG. 2) that generate one or more high frequency electrical signals. The sensor 510 can register high frequency electrical signals on the power line 150 and generate one or more data signals that contain information about the high frequency electrical signals. The sensor 510 can transmit data signals to the concentrator 540 using electrical grid communication (PLC) technology over the power grid 150 line. Hub 540 may transmit data signals to computing unit 120 using a wired or wireless communication method. Computing unit 120 can identify the electrical state of one or more electrical devices using data signals at least in part.

As shown in FIG. 5 and 6, the sensor 510 may comprise (a) a receiver 611 for collecting data; (b) controller 615; (c) a communication device 616; (d) power supply 617. In some examples, controller 615 and power supply 617 may be the same or the same as controller 215 and power supply 217, respectively, in FIG. 2. Similar to sensor 110 (FIGS. 1 and 2), sensor 510 may be configured to detect continuous noise on the power line 150 by processing one or more electrical signals on the power line 150. In some embodiments, the sensor 510 may be configured to receive electrical signals with a high frequency component and convert the electrical signals, and in particular the high frequency component of electrical signals, into one or more data signals.

The receiver 611 for data collection may include (a) one or more current transformers 628 and 629; (b) one or more filter circuits 613; (c) at least one converter module 614.

Current transformers 628 and 629 can be used to measure the amount of AC current passing through a conductor. Current transformers 628 and 629 can produce a weak, easily measurable signal that is proportional to the greater amperage in the conductor. In particular, current transformers 628 and 629 measure the amount of current passing through two conductors connecting electrical panel 552 to utility wattmeter 553 using one current transformer for each AC phase. The total electricity consumed by a building is the total electricity transmitted by each AC phase.

In other examples, current transformers 628 and 629 are not current transformers. Current transformers 628 and 629 can be any device that can measure current, such as Hall sensors, Rogowski coils, current transformers, or shunt resistors. For example, current transformers 628 and 629 may be shunt resistors. Those. the AC current can flow through a low resistance, and the data acquisition receiver 611 can measure a low voltage across the low resistance resistor. This measured low voltage is proportional to the amount of current flowing through the two conductors connecting the 552 electrical box to the 553 utility wattmeter.

In some examples, filter circuits 613 are designed to filter portions of incoming electrical signals from the mains power line. For example, filter loops 613 may filter the fundamental frequency of the AC line (60 Hz in the US) so that the converter module 614 is not overloaded by the 60 Hz frequency reference. In the same or other examples, the filter circuits 613 may include a high-pass filter. In some embodiments, the filter circuits 613 may be the same or the same as the filter circuits 213 depicted in FIG. 2.

The converter module 614 can be electrically connected to the filter circuits 613 and

10 037313 may be configured to receive a filtered signal from filter circuits 613. Converter module 614 may be configured to convert one or more filtered signals into one or more data signals. The data signals can contain information about the high frequency component of the electrical signals. In some examples, the converter module 614 may comprise: (a) an analog-to-digital converter 619; (b) a digital signal processor 618.

In various embodiments, communications device 616 may comprise a PLC transmitter. The PLC transmitter can be a transmitter for transmitting data through a conductor used to transmit electrical power (eg, power line 150). The communication device 616 may be electrically coupled to the line of the power supply network 150 and may be configured to transmit data signals to the hub 540 via the line of the power supply network 150. In the same or other PLC examples, communication between communications device 616 and hub 540 may be bidirectional. In other examples, communications device 616 may be similar or identical to communications device 216 and may transmit a data signal directly to router 505 and / or computing unit 120.

The hub 540 may comprise (a) an electrical interface 641 for connecting to an electrical outlet 151 (FIG. 5) of a power line 150 (FIG. 5); (b) communication device 643.

In various embodiments, electrical interface 641 may include a two-pin or three-pin power connector. In some examples, electrical interface 641 may also include a PLC receiver. The PLC receiver of the electrical interface 212 may be configured to receive data signals from the communication device 616 of the sensor 510 via the power line 150. In some examples, electrical interface 641 may also include a PLC transmitter.

Communications device 643 may be electrically coupled to electrical interface 641 and receive a data signal from electrical interface 641. In some embodiments, communications device 643 may comprise a wireless transmitter. The wireless transmitter can transmit data signals to computing unit 120 via router 505 (or directly to computing unit 120). In other examples, communications device 643 may be wired to router 505 and / or computing unit 120 and transmit data signals over the wired. In other examples, hub 540 may be integrated with or part of computing unit 120 or router 505.

FIG. 7 is a flow diagram of an exemplary embodiment of a method 700 for providing an electrical event recording device in accordance with an embodiment. Method 700 is exemplary only and is not limited to the embodiments presented herein. The method 700 can be applied in many different embodiments or examples not shown or described specifically in this application. In some embodiments, the actions, operations, and / or processes of method 700 may be performed in accordance with the presented order. In other embodiments, the actions, operations, and / or processes of method 700 may be performed in any other suitable order. In other embodiments, one or more actions, operations, and / or events of method 700 may be combined or omitted.

As shown in FIG. 7, method 700 includes an act 710 of providing at least one receiver for data collection. In some embodiments, the data acquisition receiver may be configured to receive and process one or more electrical signals from the power line and convert the electrical signals to first data signals. For example, the data acquisition receiver may be the same or identical to the data acquisition receiver 211 shown in FIG. 2, or the data acquisition receiver 611 shown in FIG. 6. In the same or other examples, act 710 may include providing a second receiver for data collection. In some examples, a first data acquisition receiver may be connected to a first phase of a building power line, and a second data acquisition receiver may be connected to a second phase of a building power line.

The method 700 depicted in FIG. 7 continues with act 715 to provide a communication device. In some embodiments, the communication device may comprise a transmitter and may be configured to transmit the first data signals. For example, the communication device may be similar or identical to the communication device 216 depicted in FIG. 2, or the communication device 616 shown in FIG. 6.

Next, the method 700 depicted in FIG. 7, provides for an act 720 to grant a controller. For example, the controller may be similar or identical to controller 215 shown in FIG. 2, or the controller 615 shown in FIG. 6.

Next, the method 700 depicted in FIG. 7 includes an act 725 of electrically connecting a communications device, a data acquisition receiver, and a controller. For example, the electrical connection of the communication device, data acquisition receiver, and controller may be similar or identical to the electrical connection of the communication device 216, data acquisition receiver 211, and controller 215 shown in FIG. 2, or electrical connection of communication device 616, receiver 611 for failure

11 037313 data and controller 615 shown in FIG. 6.

The method 700 depicted in FIG. 7 continues with action 730 to provide a computing unit. In some embodiments, the computing unit may be configured to determine a change in electrical states of one or more electrical devices using the first data signals. For example, the computing unit may be similar or identical to the computing unit 120 shown in FIG. 1 and / or 5.

In some examples, act 730 may include providing a computing unit comprising: (a) a communications device for receiving data signals from the communications device; (b) a preparatory module for correlating one or more electrical events with one or more different types of electrical events; (c) an event registration module for detecting the occurrence of electrical event (s); (d) a classification module for classifying the type of electrical event (s); and (e) a communication module for communicating with a user. For example, the communication device, preparatory module, event logging module, classification module, and / or communication module may be similar or identical to communication device 221, preparatory module 223, event logging module 224, classification module 226, and communication module 225, respectively, shown in FIG. 2.

FIG. 8 is a flow diagram of an embodiment of a method 800 for recording and classifying electricity consumption by one or more electrical devices connected to a power supply network, in accordance with an embodiment. Method 800 is exemplary only and is not limited to the embodiments presented herein. Method 800 can be applied to many different embodiments or examples not shown or described specifically in this application. In some embodiments, the steps and / or processes of method 800 may be performed in the order presented. In other embodiments, the steps and / or processes of method 800 may be performed in any other suitable order. In other embodiments, one or more acts, operations, and / or events in method 800 may be combined or omitted.

As shown in FIG. 8, method 800 comprises an act 810 of connecting a receiver to collect data to a utility power line. In some examples, the data acquisition receiver may include an electrical interface. The electrical interface can be designed to be connected to the mains power line. For example, the electrical interface may be a two-prong or three-pronged electrical plug, and the electrical plug may be connected to an electrical wall outlet on a power line. In other examples, the data acquisition receiver may be connected to an electrical panel and / or a building's utility wattmeter. In some embodiments, act 810 may include connecting a single data acquisition receiver to each phase of a two-phase power line. In some examples, the data acquisition receiver and electrical interface may be similar or identical to the data acquisition receiver 211 and electrical interface 212, respectively, shown in FIG. 2. The data acquisition receiver may also be similar or identical to the data acquisition receiver 611 shown in FIG. 6.

The method 800 depicted in FIG. 8 continues with act 815 to capture one or more of the first electrical signals from the power line. In some examples, the captured electrical signal contains electrical noise, more specifically continuous electrical noise at a fundamental frequency. In various examples, continuous electrical noise may include high frequency electromagnetic interference (eg, electrical signals in the frequency range of ten to hundreds of kilohertz). In many embodiments, the acquisition receiver can capture the first electrical signals.

Next, the method 800 depicted in FIG. 8, provides for an act 820 of converting the first electrical signals to one or more first data signals. In some examples, converting the first electrical signals may first include filtering the electrical signals using filter loops and then converting the filtered signals to digital signals using a converter module. In various examples, a high-pass filter can be applied to the first electrical signals to filter, for example, the AC frequency, harmonics of the AC frequency, and / or other frequencies. The filtered electrical signals can then be converted to first data signals by supplying the filtered electrical signals to a converter module (eg, an analog-to-digital converter) which, for example, can sample the filtered electrical signals at 1 MHz. In some examples, the filter circuits and converter module may be similar or identical to the filter circuits 213 and converter module 214 shown in FIG. 2 and / or filter circuits 613 and converter module 614 shown in FIG. 6.

Next, the method 800 depicted in FIG. 8, provides for an act 825 to transmit the first data signals. In some embodiments, the implementation of the first electrical signals may be transmitted by the transmitter. For example, the transmitter can be similar or identical to the device

- 12 037313 communication line 216 shown in FIG. 2, or the communication device 616 shown in FIG. 6. In some examples, the transmitter may be a wireless transmitter. In other examples, the first data signals may be transmitted via a wired connection. In other examples, the first data signals may be transmitted via a combination of wired and wireless communications. For example, a signal can be transmitted from a data collection receiver to a hub via power lines and wirelessly transmitted from the hub to a computing unit.

The method 800 depicted in FIG. 8 continues at 830 to receive the first data signals. In some embodiments, the first electrical signals may be received by the receiver. For example, the receiver may be similar or identical to the receiver of communications apparatus 221 depicted in FIG. 2. In some examples, the receiver may be a wireless receiver. In other examples, the first data signals may be received by a wired receiver via a wire, or received by a combination of wired and wireless receivers.

Next, the method 800 depicted in FIG. 8, provides for an act 835 of using the first electrical signals to prepare a processing module operating on the computing module to correlate the first data signals with a change in the electrical state of a particular electrical device. FIG. 9 is an exemplary flow diagram of an operation 835 of using first electrical data signals to prepare a processing unit for correlating the first data signals with a change in electrical state of a particular electrical device in accordance with an embodiment.

In a first step of step 835 in FIG. 9 is an operation 911 for preparing the first data signals. In some examples, the first data signals are buffered as 2048-point vectors and fast Fourier transforms (FFTs) are performed on the first data signals to obtain a frequency domain signal. 2048 points can be evenly spaced across a 500 kHz spectrum width, which gives a resolution of 244 Hz per bin in FFT. The FFT vector or frequency vector is calculated 244 times / s. In other examples, other FFT bin sizes and number of point vectors may be used. In some embodiments, the preparation of the first data signals is performed by the event registration module. For example, the event registration module may be similar or identical to the event registration module 224 shown in FIG. 2.

By the next step of step 835 shown in FIG. 9 is an operation 912 of registering the occurrence of one or more electrical events on the power supply network. Most SMPS devices generate noise peaks that are 8-60 dB above the low fluctuation noise level of the first data signals. In most buildings, the fluctuation noise floor can vary unpredictably from -90 to -70 dBm across the entire spectrum. Due to the high instability of the fluctuation noise, the input frequency vector can be averaged over time to obtain a stable noise level. In some examples, an average sliding window with a predefined window size (eg, twenty-five) may be used. Using a window size that is too small can lead to an increase in false positives, while a large window size can increase the amount of time it takes for the system to register nearby events occurring simultaneously as separate events.

In some examples, the event registration module may calculate the mathematical expectation of a given number of frequency vectors and store it as a fluctuation noise signature in the data storage unit 230 (FIG. 2). FIG. 10 is a graph 1000 of an example of a fluctuation noise signature of a building in accordance with an embodiment.

Then a new window is calculated through each predetermined number of frequency vectors and a difference vector with the signature of the new fluctuation noise is calculated. If the electrical device is turned on and new noise in the frequency range is transmitted to the power line, the difference vector reflects this change. Thus, this vector of difference allows segmentation of the event. FIG. 11 depicts an example of a noise signature plot 1100 of a new device in accordance with an embodiment. FIG. 12 depicts an example of a noise signature plot 1200 of a new electrical device after fluctuation noise has been removed in accordance with an embodiment.

In various embodiments, the event registration module may scan the difference vector to find any values above a predetermined threshold. The event registration module can use a global threshold, which is set once and used in different buildings. In some examples, a threshold 8 dB above the fluctuation noise may be a sufficient threshold for the power level. Due to the fact that the vectors in the window are averaged, if the window only partially matches the event, the event registration module can still register it correctly, but the difference vector will reflect a smaller value. To mitigate this potential problem, if an event is logged, a new difference vector is computed using the next window cycle. After the event is registered, the fluctuation noise vector is updated to reflect the new noise level of the power supply network.

In many examples, the event registration module can determine amplitude or frequency

- 13 037313 components above the threshold using the difference vector and can use the Gauss function to extract the mean, amplitude and variance parameters. For example, in FIG. 12 depicts a typical amplitude (A), mathematical expectation (μ) and variance (σ) of an electrical signal from a newly turned on device. The amplitude change can be positive or negative depending on whether the device has been turned on or off. The noise signatures are reversed with respect to each other on transitions to the opposite state (ie, on and off have opposite noise signatures).

The feature vector for a potential event can be generated by the event registration module using the parameters of the center frequency, which usually corresponds to the global maximum of the frequency component. Other peaks can also be represented as harmonics. In many examples, the event registration module can apply the KNN (k-nearest neighbors) method (k = 1) using the Euclidean metric system and inverse weighting to generate a feature vector.

Next, action 835 shown in FIG. 9 includes an operation 913 to associate specific electrical devices with specific electrical events. In some examples, the prep module can perform a prep or calibration process with user input to create a given association.

If the electrical event recording system uses a fingerprint method to identify the device, then it may need a preparatory process to obtain information about the parameters of the various devices in the building. In some examples, this provisioning is typically accomplished by enabling each device by the user at least once.

In other examples, another source of information for the electrical noise signature database may be a cloud information source. A central database can be created that can be shared by all electrical event recorders. If the user marks a new device, then it becomes part of the shared storage, allowing future requests to use the information to identify the new electrical device. For example, a user may have access to an online database of popular consumer electronic device signatures submitted by other users who have gone through the calibration process.

In addition, since the received EMI is a simulated signal, it is possible to create such signature databases without physically turning on the device or without physically accessing it. This can be accomplished by obtaining information from the United States Government's FCC (Federal Communications Agency) compliance report and other technical specifications, which contain the frequency and emission value in dBμV for various peaks of noise emitted by specific electrical devices.

For example, consumer electronic devices must pass certain certifications, such as the FCC, to ensure that the EMI they generate are within certain acceptable limits. The result of the certification is a publicly available document that describes the electrical specifications and noise performance of equipment. Information can be obtained from publicly available documents that can be used to create a database that contains calculated resonant frequencies for various devices. Other sources of this information can be obtained from the specifications of the internal integrated circuits and oscillators used in consumer electronic devices.

As shown in FIG. 2, computing unit 120 (FIG. 1) may store the results of the association or calibration process in data storage unit 230 (FIG. 2). This information can be further used to associate one or more noise signatures with specific electrical events.

In the same or other examples, the storage unit 230 may comprise a database of electrical noise signatures of electrical devices. In some examples, electrical noise signatures of electrical devices may be provided by a third party (eg, the manufacturer of the device 100 (FIG. 1) and / or 500 (FIG. 5)) recording the electrical event.

In the same or other embodiments, the calibration process may include a marking process if the user of the electrical event recorder 100 helps associate the electrical device with a specific electrical event. In some embodiments, the preparatory sequence takes into account the participation of the user of the electrical event recorder 100 turning each of the electrical devices on and off (or at least turning on and off electrical devices using SMPS in a home or other building) until module 223 The preparation works and records electrical events on the power line 150.

To identify electrical events, preparation module 223 (or event registration module 224) may employ a stepping sliding window detector that continuously searches for changes in the value of the segmented data. Preparation module 223 (or event registration module 224)

- 14 037313 can activate the step detector if it encounters a monotonically increasing or decreasing signal with a rate of change above the received or predetermined threshold.

In some examples, the resulting threshold is set at an arbitrarily large value and decreased in small steps. For each step, preparation module 223 (or event registration module 224) may segment a random subset of events that occur in isolation. If the correct number of events is calculated, then the threshold is accepted. If not, then the threshold is decreased and the process is repeated. For example, if a subset contains four events, then there must be four increments and four decrements. Preparatory module 223 (or event registration module 224) can monotonically decrease the detection threshold until a given picture appears. In this case, the resulting threshold is set with minimal control or human intervention. In other embodiments, preparatory module 223 may set the threshold to a predetermined value (eg, eight dB).

After energizing each electrical device, the user can mark each electrical event registered by the preparation unit 223 using the communication unit 225. For example, if the user (1) turns the TV on or off; (2) turns on or off each of the four fluorescent lamps in the room, the first two electrical events recorded by the event recorder 100 can be marked as turning the TV on and off, and the subsequent eight electrical events can be marked as turning on and off each light bulb. Likewise, the user can turn on and off all electrical devices in the house or building and perform a similar marking operation.

In other examples, communications device 225 may include a portion that can operate on a mobile electrical device (such as an iPhone® or iPad® developed by Apple Computers, Inc., Cupertino, Calif.) That allows a user to time-stamp the occurrence of a particular electrical events. In these examples, the user can turn on and off the mobile power device on all electrical devices in the building as long as he carries the electrical device using a portion of the communication module 225. The user can use the communication module 225 to flag when an electrical event occurs. For example, while the preparatory module 223 is operating and recording electrical events, the user can turn on the laptop and press a button on the electrical device, causing the electrical device to begin recording a description of the electrical event and the time the event occurred.

The preparatory module 223 may correlate data recorded by the mobile electrical device and electrical signals recorded by the receiver 211 for data acquisition. In some examples, the mobile electrical device can transmit data instantly (eg, in real time) to the computing unit 120, and in other examples, the data may be transmitted to the computing unit 120 after the end of the preparatory event (eg, in burst mode).

In additional examples, a user can turn on or off all electrical devices in a building, and preparation module 223 can access electrical noise signatures of electrical devices that are stored in storage module 230 (or online database) and automatically associate electrical events with signatures. electrical noise of specific electrical devices. In various embodiments, a user may enter information about the building's electrical devices (eg, manufacturer, module number, and / or serial number) to assist prep module 223 in associating electrical events with electrical noise signatures of building electrical devices. Those. preparation module 223 may compare electrical noise signatures determined by data acquisition receiver 211 (FIG. 2) with stored electrical noise signatures of building electrical devices to associate electrical events with specific electrical devices.

In various embodiments, instead of running the preparatory sequence, the preparatory module 223 or classification module 226 may access the electrical noise signature database if a new electrical device is first powered on during the operation of the electrical event recorder 100. In this case, the electrical noise signature of the new electrical device can be compared with the electrical noise signatures from the database to find a matching signature.

In the same or other examples, if the electrical event cannot be identified using the electrical noise signature database, the electrical event can be manually flagged. In some examples, preparation module 223 can classify an electrical device into a device category (eg, TV) and the user can accurately label (eg, Sony KDL-40V TV). If the user has a large number of such electrical devices, the user can turn the electrical devices on or off, and the electrical event recorder 100 can generate a list of category markings that can be updated based on more accurate information. After you have finished associating specific electrical devices with specific noise signatures as part of operation 911 depicted in FIG. 9, action 835 shown in FIG. 9 ends.

Referring again to FIG. 8, method 800 continues with act 840 to capture second electrical signals from the mains supply. Act 840 may be similar or identical to act 815.

The method 800 depicted in FIG. 8 continues with act 845 to convert the second electrical signals to second data signals. Act 845 may be similar or identical to act 820.

Next, the method 800 depicted in FIG. 8 provides for act 850 to transmit second data signals. Act 850 may be similar or identical to Act 825.

Next, the method 800 depicted in FIG. 8 provides an act of 855 to receive the second data signals. Act 855 may be similar or identical to act 830.

The method 800 depicted in FIG. 8 continues with action 860 to register the occurrence of one or more electrical events in the power supply network. Act 860 may be similar or identical to step 912 depicted in FIG. 9. In this action, the event recorder can detect peaks in electrical signals above a predefined threshold.

Next, the method 800 depicted in FIG. 8, provides for an act 865 to associate one or more electrical events with a change in the electrical state of at least one electrical device of the one or more electrical devices. In some examples, a change in the electrical state of an electrical device may be a transition from one of: (a) from a completely power off state (or standby state) to a power on state; or (b) a power-on state to a fully powered-off state (or standby state).

In the same or other embodiments, electrical events may result from a transition from a first state providing a first power state for an electrical device to a second state providing a second power state for an electrical device. In this example, the first power state is different from the second power state. In some embodiments, the first power state and the second power state are not power off states.

Event registration module 224 (FIG. 2) can determine peaks above the threshold using the difference vector and adjusts the Gaussian function to extract the mean, amplitude and variance parameters. Feature vectors are generated for each segmented event and can subsequently be used to build a model based on the KNN method. The KNN model can be used to automatically identify the source of electrical events. The event registration module 224 (FIG. 2) may apply the KNN (k = 1) method using Euclidean metric and inverse weighting.

In the same or other examples, after the event registration module 224 (FIG. 2) detects the peak of the electrical signal from the data acquisition receiver 211 (FIG. 2), the event registration module 224 may extract three features from the signal:

(1) the relative size of the step of change; (2) deviation in flow change; and (3) increasing or decreasing the time of the unknown event. These three features can be useful for disambiguating electrical devices.

The event registration module 224 (FIG. 2) can use the information collected about the electrical event to associate it with a particular electrical device. For example, event registration module 224 (FIG. 2) may compare the signature of the electrical event with the signatures stored due to the preparatory event to determine which electrical device is causing the event and the type of event.

Next, the method 800 depicted in FIG. 8, provides for an act 870 to display information about one or more electrical events to a user. In some embodiments, communication module 225 (FIG. 2) may display information about one or more electrical events to a user. Information can be displayed in various forms. In some embodiments, the electrical state of two or more electrical devices in a building for a specific time span may be presented to the user in the form of a diagram. In some embodiments, the implementation of information about one or more electrical events can be displayed in real time. After action 870, actions 840, 845, 850, 855, 860, 865, and 870 can be repeated each time a new electrical event occurs to identify the electrical event. Using these actions, the system can record and classify the power consumption of one or more electrical devices.

Test results for a typical embodiment of an electrical event recording device 100 are disclosed herein. As will be described in detail below, the test result of a typical embodiment of an electrical event recording device 100 (FIG. 1) had an average classification accuracy of 93.82% in identifying and classifying continuous electrical event noise at the level of an individual electrical device (e.g., a specific

- 16 037313 TV, laptop or CFL lamp).

An exemplary embodiment of an electrical event recorder 100 was tested on a 120 V, 60 Hz electrical network. However, the electrical event recorder 100 (FIG. 1) can be easily applied to an electrical network using a different frequency and operating voltage with minimal hardware and software changes. For buildings that have split phase wiring (i.e., two 120 V branches that are 180 ° out of phase), in some examples, the crosstalk between the two phases allows the electrical event recorder 100 (FIG. 1) to continue observing in one place and capture events from both phases.

The devices were housed in seven different buildings using a typical embodiment of electrical event recorder 100 (FIG. 1). Data were collected from one building over a longer period of time (i.e., six months), and from the rest of the buildings over a shorter period of time. This approach shows the general applicability of the electrical event recorder 100 (FIG. 1) to a variety of buildings, as well as the long term temporal stability of the electrical event recorder 100 (FIG. 1). FIG. 13 depicts a demographic table 1300 for buildings used for a typical installation of a typical electrical event recorder 100, in accordance with an embodiment.

For each building, an available electrical outlet with two electrical outlets was randomly selected. A typical sensor 110 was connected to one electrical outlet 151 and another electrical outlet was used to provide power to the computing unit 120 (FIG. 1). Computing unit 120 (FIG. 1) generates EMI noise, but this EMI noise is subtracted as part of the fluctuation noise. After installation, notes were made on each electrical device in the building that is equipped with a switching power supply. This list included incandescent bulbs that were controlled by a dimmer control, in addition to any CFL lighting fixtures. For dimmers, events were only collected at 0 and 100% lighting levels due to the problem of accurate and repeatable setting of intermediate dimmer levels. A detailed analysis of dimmers will be presented below. The collected markings were then fed into a marking program using ground-based data control.

The electrical devices of the buildings were tested in two stages. In the first step, each device was individually turned on and off five to six times to ensure that a signature was captured for each device. Each time an event was recorded by the electrical event logging system (i.e., when the electrical device was turned on), the tags were automatically retrieved and passed to the waiting list of the marking program using ground-based data control. Once events were captured, those events were flagged using a marking program using ground-based data control. This process was carried out for all electrical devices in the building.

In the second phase of data collection in each building, the goal was to collect data within the natural environment under which overlapping and concurrent events could occur. To create a more natural environment, building owners were asked to perform a certain kind of activity, such as watching TV, cooking, etc., while random devices and / or logical groups of devices were turned on. For example, a DVD player, digital cable TV, and / or game console was turned on and then turned on to simulate watching TV with a universal remote control.

In addition to the electrical devices already in each of the test buildings, data was collected from the same laptop power adapter, the same two CFLs, and the same camera charger in each building. To mimic the natural use of each device, these electrical devices were moved around the building and plugged into any available outlet. This envisaged analysis of the stability of these noise signatures has been performed in various buildings.

The data collected by the marking program using ground control data was time-stamped, tagged, and stored in an XML (Extended Document Description Language) database. Since more features were collected during the tests than necessary, the XML database makes it easy to filter and parse the data to generate output in a format that can be directly imported or processed.

To test the temporal stability of the long-term operability of the electrical event recorder, a typical embodiment of the electrical event recorder was housed in one of the buildings for more than six months. Events were collected and noted throughout this time period using a minor modification of the marking program using ground-based data control applied in other buildings. Each time an event is logged, the logging program places the extracted features on a pending list and generates a beep. The residents of the building were asked to correlate the events with an electrical device that they turned on using the tagging program.

- 17 037313

The marking software has been designed with a greatly simplified interface requiring only the selection of device brands from a selection list on the screen. If a user did not mark an event in the waiting list for more than six minutes, then he was marked as unknown and removed from the waiting list. This feature allowed residents to avoid any misrecorded events and to ignore events if residents were unable to use the marking device promptly. While this feature is convenient, it also resulted in the loss of real-life events that residents missed or chose to ignore. However, the goal of this device was to collect as many ground control data marks as possible over a long period of time.

A total of 2,576 electrical events were collected from seven buildings. Most of the events came from lighting using either CFLs or dimmers. Most of the other detectable devices were common consumer electronic devices such as LCD (liquid crystal) or LED (light-emitting diode) TVs, game consoles, personal computers, power adapters, etc.

Devices such as dryers and electric stoves did not generate events in some test buildings. Typically such devices have a large resistive load and therefore do not emit high frequency noise. Continuous noise events were observed from the washing machine in House H7, as only House H7 had a modern, highly efficient Energy Star compliant washing machine that generated continuous noise. Continuous electrical noise came from her electronically controlled DC motor powered by an SMPS. In addition, for most large devices, the use of SMPS is considered impractical compared to the overall power consumption of the device.

To determine the classification performance of the actuation of electrical devices in a building, the classification method was evaluated using two different operations. In the first operation, the performance of our classifier using the KNN method was estimated using a tenfold cross-validation for each building. In the second test, a minimal pre-set (a single pre-case for each device of interest) was used to simulate a more practical real-world situation. FIG. 14 is a table 1400 showing the performance of a typical electrical event recording system during a typical installation, in accordance with an embodiment. When using tenfold cross-validation, an overall mean accuracy of 91.75% was observed.

When analyzing the matrices of uncertainties for each of the buildings, it was found that for H5 and H6 there was a mismatch between some lighting devices. FIG. 15 depicts a confusion matrix table 1500 showing the performance of a typical electrical event recording system during a typical installation, in accordance with an embodiment. Based on the data, it was noted that the processing module correctly identified similar lighting fixtures (i.e. similar model and brand) that were located in different rooms. However, houses H5 and H6 contained rooms in which similar models of fluorescent lighting fixtures were installed spatially adjacent to each other (1-2 feet apart), which produced very similar noise signatures. Thus, it was these lamps that did not differ enough, being not far enough along the power line to distinguish between these lamps.

Using hardware with a higher frequency resolution can partially solve this problem. It may also be acceptable for some devices to logically combine or group lighting that is spatially aligned. Using this latter approach, the performance of a typical electrical event recording system during a typical deployment has been improved by grouping similar lamp or device models that are 1 to 2 feet apart as a single event. This approach gives an increase in the classification accuracy for H5 (92.4%) and H6 (91.8%) with an average overall accuracy of 93.82%.

Low accuracy for home H2 was observed due to a mismatch between two devices of the same brand that were in the opposite phase of the installation point of the device 100 for recording an electrical event. Part of the second floor in house H2 showed very weak coupling between the electrical phases, which forced us to connect the electrical event recorder 100 in that phase and thus caused some of the signatures to look very similar. This problem can be solved, for example, by using two different electrical event recorders 100, one on each of the two phases, or by installing one electrical event recorder 100 on a split-phase 240 volt outlet.

Since N-fold cross-validation is usually optimistic and is not a true measure of the expected classification performance for a real system, the benchmark analysis was performed using the minimum pre-set, i.e. one event signature for each device to simulate and then apply to the test set. For example, a building owner is likely to be willing to provide only a few preparatory events for each device. FIG. 16 depicts another table 1600 summarizing the classification accuracy metrics using the minimum set of preparatory data of a typical electrical event recording system during a typical installation, in accordance with an embodiment. Using this approach, for verification, the accuracy was established

89.25%.

Since there are no two buildings with exactly the same electrical network and with completely different fluctuation noise presented in the power line, the plausibility of noise signatures across all buildings was also considered.

The use of two devices has been implemented, which together led to the assumption that the detected signal for an electrical device in one building could be used to classify a similar electrical device in another building. When using the first device, the EMI signal of the device confirmed independence from the building in which it was used. Thus, the signal is authentic with respect to the functioning of the electrical device. With the second device, the EMI signal has confirmed dispersion compatibility with several but similar devices. Those. EMI signals from devices of the same make and model are substantially similar.

The application of the first device involved collecting data from each of four preselected devices in each of the seven buildings. The device validated the plausibility of the signal across all buildings based on the classification results. The average accuracy of ten trials for the ten-fold cross-validation test was 96.87%. For three of the four devices, the classification accuracy was 100%. This result strongly suggests that these devices generated similar signals. Only one device - the laptop power adapter - performed poorly. FIG. 17 depicts a table 1700 of the performance of a typical event recording system in classifying four electrical devices used in different buildings using a 10-fold classification check, in accordance with an embodiment;

The laptop power adapter was more difficult to identify because the specific extracted feature vector for a given device in H6 did not differ significantly from other buildings. However, upon closer inspection, the noise generated by the laptop power adapter had a harmonic peak that was very close in amplitude to the carrier frequency. This marginal discrepancy caused the processing module to accept the first harmonic as the carrier frequency in H6 and thus extract a different feature vector. A simple approach to partially solve this problem is to use a classification module that classifies an event only if the distance between the event feature vector and the nearest neighbor, using the KNN method, exceeds a certain threshold. If the threshold is not exceeded, then the classification module can construct a new feature vector using the next largest peak. With this new method, the classification accuracy for a laptop was 100%.

The second unit collected data on eight Dell®-branded 20-inch LCD monitors that are similar models. This model of LCD monitor was also simultaneously found in houses H5 and H7. The EMI signature for one of the monitors has been replaced with the EMI signature of the other nine monitors obtained from a different building or home. For example, a processing module was prepared using a signature from house H7 and tested in house H5. This test confirmed that if the signatures of any LCD monitor were different, it would be incorrectly classified as a different device. All tests showed 100% accuracy, leading to the assumption that similar brands and models of devices produce similar signatures.

In addition, for any signature or fingerprint-based classification system, temporal stability is important. The classification module should function properly for months (or, ideally, years) without the mandatory, often repeated training that requires the underlying characteristics to remain stable over an extended period of time.

To confirm the stability of noise signatures over an extended period of time, electrical devices were selected from our long-term placement dataset that met two criteria: (1) the devices were fixed in their installation locations at the time of placement, and (2) the devices were not replaced in any way ( for example, lamp replacement was not carried out). For these devices, randomly selected vectors of EMI signatures were extracted, distributed over a six-month verification period.

FIG. 18 is a graph 1800 showing temporal stability or signature variation over time for four randomly selected electrical devices by visualizing feature vectors in feature space in accordance with an embodiment. It was noted that the long-term temporary change was similar to that observed with the short-term temporary change of these electrical devices, and that none of the clusters were overlapped.

In order to better understand how a temporary change affects classification accuracy over an extended period of time, test kits were created for each electrical device, including all events that occurred more than one week before the last

- 19 037313 posting day, and a prep kit was created containing electrical events from all electrical devices that occurred in the last week. This setup confirmed that the processing engine would detect misclassifications if the continuous noise signatures in the test set deviated more than the distance between the clusters of the electrical device. Tests have shown 100% accuracy using the KNN classifier, which shows that electrical devices are largely stable over an extended period of time. It is important to note that this long term application covered the summer and winter periods.

It is common for a building to have a large number of such devices, such as having many televisions or, more commonly, lamps, all of which use the same brand of CFL lamps. This many identical devices can cause problems, especially if such devices cannot be grouped into a single unified group. For example, grouping two ceiling lights in a bedroom may be acceptable, but grouping lamps that are in different rooms or on different floors may not be acceptable.

Component tolerances in the components that make up the switching circuit of electrical devices usually give sufficient variability in the switching frequency such that the mathematical expectation of a Gaussian distribution observed on the power line is also variable. FIG. 19 is a graph 1900 of a spectrum observed by an electrical event recording system showing the spectrum of noise generated by four CFL lamps of the same model purchased in a package of four lamps, in accordance with an embodiment. When using lamps purchased in a package, check that the lamps are from the same production batch. Note that the spectra do not overlap even for CFLs from the same batch.

An electrical event registration system is able to distinguish these signs if the observation is carried out in isolation, i.e. a line isolation transformer was used to create a silent power line for this particular application. With a higher resolution of the analog-to-digital converter and more FFT, this disadvantage can be eliminated. Thus, increased resolution can provide better differentiation.

Since conducted EMIs propagate through the power line, they are influenced in several ways, but most significantly the electrical signal is attenuated depending on the network inductance between the noise source and the measuring point. Thus, two identical devices generating identical EMI may look different at the reference point of measurement depending on where the devices are connected along the power line, which has been observed in building applications.

To verify this, an electrical device was connected at two different locations in the building, and raw spectral data was loaded and recorded by the electrical event recording system. FIG. 20 is a plot of the 2000 spectrum of a piece of data from a test of an electrical device that was connected at two different locations in a building, in accordance with an embodiment. The confidence in the mean or peak location, together with the corresponding Gaussian parameters extracted from the form, made it possible to accurately identify a mobile device, such as a laptop adapter, despite changes in the amplitude of the observed features. In addition, within a building, similar devices located in different parts of the building can be differentiated by observing the shape of the EMI spectrum characteristic of the electrical device. This result confirms that the number of installed devices present in the building (i.e. the number of CFL lamps in the house or the number of televisions of a certain type) can be determined.

The likelihood of two different types of devices having the same Gaussian mean and variance is very low. At best, the likelihood can be 1 / (FFT size), which is the likelihood of two electrical devices having the same center frequency. With an FFT size of 2048, this likelihood is 0.05%. In practice, due to the location and use of Gaussian distributed parameters, the actual likelihood is much lower.

Unlike SMPS-based devices, dimmers generate a broadband signal with their internal bi-directional thyristor switches. The electrical event recording system can be configured to use a Gaussian distribution in the feature extraction phase.

Thus, the broadband noise emitted by the dimmer is modeled on the basis of a Gaussian distribution, even if bandwidth constrained uniform distribution would be more appropriate. It was found that for devices controlled by a dimmer, a Gaussian distribution with high variance was observed. FIG. 21 is a graph 2100 of the EMI signals generated by the dimmer at different levels, in accordance with an embodiment.

The difference in the signature of the noise generated at the different dimmer levels allows the electrical event recorder 100 (FIG. 1) to identify the dimmer as well as set the dimmer level. Since it is not possible to prepare data for different levels, a noise performance model can be built, and how the noise performance depends on the level can be determined.

- 20 037313 nya dimmer or cutoff angle of the two-way thyristor switch.

The performance and reliability of the classification module depends only on the features used, so it may be desirable to investigate and extract more advanced features from the underlying signals. There are many other potentially useful features that can be used by the classification module. For example, some devices produce a characteristic EMI pattern that persists for a short period of time after they are turned on. Those. these electrical devices produce narrow-band transient EMI during SMPS startup. These short term disturbances have a low frequency. FIG. 22 is a graph 2200 showing the short burst of EMI that typical CFLs produce when first powered up. The momentary burst comes from the CFL trigger circuitry, which is needed to warm up the lamp for operation. Likewise, other devices, such as modern televisions and DVD players, which contain multiple power supplies and provide electronics, also generate power-on noise signatures that can provide additional information to categorize electrical devices. In another example, SMPS modulation controls backlighting that can be used to display channel changes and TV content.

In some examples of electrical event recorder 100 (FIG. 1), computing unit 120 (FIG. 1) assumes that the mathematical expectation or location of the EMI peaks does not change during operation of the device. This assumption may be false for some types of devices. For example, in a certain brand of LCD TVs, the switching frequency of their power supply depends on the brightness of the screen. Thus, for a given brand of LCD TV, the expectation of the EMI peaks changes as the screen content changes. In some examples, computing unit 120 (FIG. 1) may temporarily block the noise peak from changing and extract temporal cues or patterns from this data.

In addition, some motorized devices, such as a washing machine or dishwasher, generate low frequency intermittent noise patterns in their motor control systems that can be used to identify the device and its status. For example, at home H7, it was observed that the washing machine in its wash cycle produces intermittent noise at a frequency of about 0.1 Hz, as opposed to constant noise during its rotational cycle. Such features can be used by a computational module to build a finite state machine or a static model for a more detailed classification.

Various examples of an electrical event recording device can record nearly simultaneous events, i.e. events with a difference of 102 ms. Those. two events that elapse more than 102 ms between occurrence can be successfully recorded as separate events by various examples of electrical event recorder using one separate sampling rate and average window size. If simultaneous events occur in a time period of less than 102 ms using these various examples of an electrical event recorder, then the individual electrical events are recorded as a single event and the extracted features are a collection of features from a large number of devices. Thus, the fundamental features remain unchanged, but are described as one event. The computational module can use different classification approaches to isolate these mixed features and identify individual devices.

FIG. 23 depicts an exemplary embodiment of a computer system 2300 that is suitable for implementing an embodiment of at least a portion of the electrical event recording apparatus 100 (FIG. 1). Computing unit 120 (FIG. 1) may contain one or more computer systems 2300. Computer system 2300 includes a chassis 2302 containing one or more wiring diagrams (not shown), a USB (universal serial bus) port 2312, a CD drive 2316 for write-once data (CD-ROM) and / or digital video disc (DVD) and hard disk 2314. An exemplary block diagram of elements included in wiring within chassis 2302 is shown in FIG. 24. The central processing unit (CPU) 2410 shown in FIG. 24 is connected to the system bus 2414 of FIG. 24. In various embodiments, the CPU 2410 architecture may correspond to any of a variety of commercially available processor architecture families.

System bus 2414 is also coupled to memory 2408, which contains both read only memory (ROM) and random access memory (RAM). Non-volatile portions of memory 2408 or ROM may be encoded with a boot code sequence suitable for restoring computer system 2300 (FIG. 23) to a functional state after a system reset. In addition, memory 2408 may contain microcode, such as a basic input / output system (BIOS). In some examples, the storage module 230 (FIG. 2) may include a memory device 2408, a USB (universal serial bus) port 2312, a hard disk 2314, and / or a CD-ROM or DVD drive 2316.

- 21 037313

In the embodiment shown in FIG. 24 various I / O devices such as disk controller 2404, graphics adapter 2424, video controller 2402, keyboard adapter 2426, mouse adapter 2406, network adapter 2420, and one or more other I / O device adapters 2422 may be connected to system bus 2414. Keyboard adapter 2426, disk controller 2404 and mouse adapter 2406 are connected, respectively, to keyboard 2304 (FIGS. 23 and 24), USB port 2312 (FIGS. 23 and 24) and mouse 2310 (FIGS. 23 and 24) of the computer system 2300 (Fig. 23). While graphics adapter 2424 and video controller 2402 are designated as discrete units in FIG. 24, video controller 2402 may be integrated with graphics adapter 2424, or vice versa in other embodiments. The video controller 2402 is suitable for updating the monitor 2306 (FIGS. 23 and 24) when displaying pictures on the screen 2308 (FIG. 23) of the computer system 2300 (FIG. 23). A disk controller 2404 can control a hard disk 2314 (FIGS. 23 and 24) and a CD-ROM or DVD drive 2316 (FIGS. 23 and 24). In other embodiments, a discrete unit may be used to control each of these devices separately.

Although many other components of the computer system 2300 (FIG. 23) are not shown, such components and their interrelationships are well known to those skilled in the art. Accordingly, in this application, there is no need to discuss additional details regarding the construction and composition of the computer system 2300 and the wiring diagrams within the chassis 2302 (Fig. 23).

When operating the computer system 2300 of FIG. 23, program instructions stored on a USB device in a USB port 2312, on a CD-ROM, or a DVD in a CD-ROM and / or DVD drive 2316, on a hard disk 2314 or in a memory device 2408 (FIG. 24) are executed by the CPU 2410 (Fig. 24). A portion of the program instructions stored in these devices may be suitable for implementing the methods 800 depicted in FIG. eight.

Although the invention has been described in accordance with specific embodiments, those skilled in the art will appreciate that various changes may be made without departing from the spirit and scope of the invention. Accordingly, the disclosure of the embodiments is intended to show the scope of the invention and is not intended to be limiting. It is intended that the scope of the invention be limited only by the limits required by the appended claims. For example, one of ordinary skill in the art will appreciate that acts 710, 715, 720, 725, and 730 depicted in FIG. 7, actions 810, 815, 820, 825, 830, 835, 840, 845, 850, 855, 860, 865, and 870 depicted in FIG. 8 and steps 911-913 shown in FIG. 9 can consist of many different actions, operations and be performed by many different modules in many different orders, that any element depicted in FIG. 1-2 and 4-6 may be modified and that the above description of some of these embodiments does not necessarily represent a complete description of all possible embodiments.

All elements claimed in each individual claim are essential to the embodiment claimed in a single claim. Consequently, replacement of one or more of the claimed items constitutes refurbishment and not repair. In addition, beneficial effects, other advantages, and solutions to problems have been described with reference to specific embodiments. Benefits, advantages and solutions to problems and any element or elements that may lead to any beneficial effect, advantage or solution, arising or becoming more pronounced, nevertheless, should not be considered as critical, required or essential features or elements of any or all claims, unless such beneficial effects, advantages, solutions or elements are indicated in such claim.

In addition, the embodiments and limitations disclosed in this application are not intended to be made public in accordance with the doctrine of copyright disclaimer unless the embodiments and / or limitations: (1) are not explicitly stated in the claims; and (2) are equivalents or potential equivalents of the expressed elements and / or limitations in the claims in accordance with the doctrine of equivalents.

Claims (6)

CLAIM 1. A device for detecting electricity consumption, containing a computing unit; a sensor configured to be connected to an electrical outlet, the sensor comprising a data collection receiver configured to receive electrical noise through the electrical outlet and convert the electrical noise into one or more first data signals when the sensor is connected to the electrical outlet; while the electrical outlet is electrically connected to the power supply network; the sensor is exchanging data with the computing unit; and the computing unit is further configured to determine the use of one or more electrical devices connected to the power supply network at least in part using one or more of the first data signals. 2. The apparatus of claim 1, wherein the computing unit is further configured to determine the electrical power consumed by the one or more electrical devices at least in part using one or more of the first data signals. 3. The device according to claim 1 or 2, in which the computing unit is further configured to determine the signature of the fluctuation noise of one or more of the first data signals; the computing unit is further configured to determine one or more amplitude or frequency components in one or more of the first data signals that are a predetermined threshold value higher in the value of the fluctuation noise signature of the one or more first data signals; the computing unit is further configured to link at least one electrical device from one or more electrical devices with one or more amplitude or frequency components in one or more of the first data signals. 4. The device according to claim 1, 2 or 3, in which the receiver for collecting data is configured to receive electrical noise, which contains essentially continuous electrical noise in the power supply network; and the substantially continuous electrical noise comprises a first electrical noise that is identifiable in the mains supply for a first length of time that is longer than one AC electrical cycle, or a second electrical noise that is identifiable in the mains for a second time that is shorter than one AC electrical cycle, but the second electrical noise is repetitive in three or more AC electrical cycles. 5. The device according to claim 1, 2, 3 or 4, in which the computing unit comprises an event registration module configured to use one or more of the first data signals to determine that one or more electrical events have occurred; a classification module configured to determine the use of one or more electrical devices using one or more electrical events; a preparation module configured to establish a relationship between an event of the first type with the first event of one or more electrical events, as well as an event of the second type with a second event of one or more electrical events. 6. The device according to claim 1, 2, 3, 4 or 5, in which the receiver for collecting data is configured to receive electrical noise, which contains essentially continuous electrical noise in the power supply network; the substantially continuous electrical noise contains high frequency signals that are cyclically stationary with respect to the AC electrical cycle in the power supply network.