Power event identification using distributed computing

Abstract

Information about power events on the power grid may be determined by processing reports of power anomalies from power monitors installed at various points on the power grid, such as within the buildings of power end-users. Power monitors may process sensor readings on power lines to determine that a power anomaly has occurred. Power monitors may transmit information about the power anomaly, such as time and location, to the processing location. The processing location may select power anomaly reports with similar time and location to determine information about the power event that caused the power anomaly. Information about the power event includes the type, time, and location of the power event. Notifications about power events may be sent to facilitate repairs and mitigate risks such as the risk of electrical fires.

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims priority to U.S. Patent Application No. 18 / 320,785 (SAGE - 0009 - U01), filed on May 19, 2023, entitled "POWER EVENT IDENTIFICATION WITH DISTRIBUTED COMPUTING", and is a continuation thereof.

[0002] The above applications and / or patents are hereby incorporated by reference in their entirety for all purposes.

Background Art

[0003] Power companies maintain a power grid for power generation and power transmission to end - users such as enterprises and residential houses. The power grid may extend over long distances to supply power to all end - users of the power company. Events may occur at different points along the power grid that can interrupt power transmission to end - users or pose dangerous risks such as the risk of fire caused by the power grid.

[0004] Events may be reported by end - users, such as a report of a tree falling on a power line. Events may also be discovered by employees of the power company, such as employees who conduct inspections of the power grid. However, the power company may not receive timely notifications of some events, and the lack of timely notifications can cause service degradation (e.g., longer power outages) and an increase in risks such as the risk of fire.

Summary of the Invention

[0005] In some embodiments, the techniques described herein include receiving a first power anomaly report for a first power anomaly from a first power monitor, the first power monitor being located in a first building, and the first power anomaly report including (i) first location information for the first building and (ii) first time information for the first power anomaly; and receiving a second power anomaly report for a second power anomaly from a second power monitor, the second power monitor being located in a second building, and the second power anomaly report including (i) second location information for the second building and (ii) second time information for the second power anomaly. The present invention relates to a method comprising: determining a first power anomaly type for a first power anomaly by processing a first power anomaly report using a classifier; determining a second power anomaly type for a second power anomaly by processing a second power anomaly report using a classifier; selecting a subset of power anomaly reports from a plurality of power anomaly reports using location and time information, wherein the plurality of power anomaly reports include the first power anomaly report and the second power anomaly report; and determining the power event type and power event location of a power event using the subset of power anomaly reports.

[0006] In some embodiments, the techniques described herein relate to a method by which the first location information includes one or more of the following: an address, a postal code, an Internet Protocol address, or a media access control address.

[0007] In some embodiments, the techniques described herein relate to a method for determining a first power anomaly type, which includes processing meteorological information using a classifier.

[0008] In some embodiments, the techniques described herein relate to a method by which the first power anomaly type is one or more of a series high impedance fault, a series arc fault, a parallel arc fault, a parallel low impedance fault, an anomaly power, or a power loss.

[0009] In some embodiments, the techniques described herein relate to methods in which the classifier includes a recurrent neural network or a transformer neural network.

[0010] In some embodiments, the techniques described herein relate to a method in which the classifier includes a decision tree.

[0011] In some embodiments, the techniques described herein relate to a method for selecting a subset of power anomaly reports, which includes selecting power anomalies using time-difference thresholds and location-difference thresholds.

[0012] In some embodiments, the techniques described herein relate to methods that include selecting a subset of power anomaly reports using power anomaly types.

[0013] In some embodiments, the techniques described herein relate to methods by which the type of power event is one or more of the following: abnormal power from solar panels or batteries, abnormal power from consuming devices, abnormal power from grid supply, power loss, parallel faults due to plants, series faults due to loose or corroded connections, or power theft.

[0014] In some embodiments, the technology described herein relates to a system including at least one server computer having at least one processor and at least one memory, wherein the at least one server computer receives a first power anomaly report from a first power monitor for a first power anomaly, the first power monitor being located in a first building, and the first power anomaly report includes (i) first location information for the first building and (ii) first time information for the first power anomaly, and receives a second power anomaly report from a second power monitor for a second power anomaly, the second power monitor being located in a second building, and the second power anomaly report includes (i) second location information for the second building The system is configured to receive (i) a first power anomaly report including location information and (ii) second time information for a second power anomaly, to process the first power anomaly report using a classifier to determine a first power anomaly type for the first power anomaly, to process the second power anomaly report using a classifier to determine a second power anomaly type for the second power anomaly, and to select a subset of power anomaly reports from a plurality of power anomaly reports using location information and time information, wherein the plurality of power anomaly reports include the first power anomaly report and the second power anomaly report, and to use the subset of power anomaly reports to determine the power event type and power event location of a power event.

[0015] In some embodiments, the technology described herein relates to a system in which a subset of power anomaly reports includes power anomaly reports received from distribution transformers or substations.

[0016] In some embodiments, the technology described herein relates to a system in which a first power monitor is installed within a first electric meter that measures the power consumption of a first building.

[0017] In some embodiments, the technology described herein relates to a system in which a first power monitor is installed in a first electrical panel of a first building.

[0018] In some embodiments, the technology described herein relates to a system in which at least one server computer is configured to determine a first power anomaly type by processing data from a temperature or humidity sensor.

[0019] In some embodiments, the technology described herein relates to a system in which at least one server computer is configured to send power event notifications.

[0020] In some embodiments, the technology described herein relates to a system in which at least one server computer is configured to display first location information, second location information, and power event locations on a map.

[0021] In some embodiments, the technology described herein relates to a system configured such that at least one server computer determines a first power anomaly type for a first power anomaly by processing information about a change in the state of a device in a first building.

[0022] In some embodiments, the techniques described herein, when performed, include receiving a first power anomaly report for a first power anomaly from a first power monitor, the first power monitor being located in a first building, and the first power anomaly report including (i) first location information for the first building and (ii) first time information for the first power anomaly; receiving a second power anomaly report for a second power anomaly from a second power monitor, the second power monitor being located in a second building, and the second power anomaly report including (i) second location information for the second building and (ii) second time information for the second power anomaly; and processing the first power anomaly reports using a classifier. The present invention relates to one or more non-temporary computer-readable media comprising a computer-executable instruction causing at least one processor to perform an action including: determining a first power anomaly type for a first power anomaly; determining a second power anomaly type for a second power anomaly by processing a second power anomaly report using a classifier; and selecting a subset of power anomaly reports from a plurality of power anomaly reports using location and time information, wherein the plurality of power anomaly reports include the first power anomaly report and the second power anomaly report; and using the subset of power anomaly reports to determine the power event type and power event location of a power event.

[0023] In some embodiments, the techniques described herein relate to one or more non-temporary computer-readable media in which the location of a power event corresponds to a location inside a first building.

[0024] In some embodiments, the techniques described herein relate to one or more non-transient computer-readable media in which the location of a power event corresponds to the vicinity of a first building and a second building.

[0025] In some embodiments, the techniques described herein relate to one or more non-transient computer-readable media containing a sample of current or voltage from an electrical sensor, wherein the first power anomaly report comprises a sample of current or voltage from an electrical sensor.

[0026] In some embodiments, the technology described herein relates to one or more non - transitory computer - readable media in which first time information includes one or more of the start time, end time, or duration of a first power anomaly.

[0027] The present invention and the following detailed description of certain embodiments thereof may be understood by reference to the following figures.

Brief Description of the Drawings

[0028] [Figure 1] A diagram showing an exemplary representation of a power grid.

[0029] [Figure 2] A diagram showing an exemplary network architecture that may be used to transmit power monitoring data from a power monitoring location to a processing location.

[0030] [Figure 3] A diagram showing an exemplary power monitoring location corresponding to an end - user.

[0031] [Figure 4] A diagram showing an exemplary anomaly reporting system that may be implemented by a power monitor.

[0032] [Figure 5A] A diagram showing an exemplary voltage waveform during the assumed operation of a power grid when no anomaly is present.

[0033] [Figure 5B] A diagram showing an exemplary voltage waveform for an anomaly caused by a series arc fault in the connection of an electrical panel between a backplane and a main breaker.

[0034] [Figure 6] A diagram showing an exemplary database table that may be used to store information about anomalies.

[0035] [Figure 7] This figure shows an exemplary system for clustering anomalous data.

[0036] [Figure 8] This figure shows an exemplary system for processing abnormal data to determine one or more of the following: whether a power event occurred, the type of power event, and / or the location of the power event.

[0037] [Figure 9] This diagram shows a flowchart illustrating an exemplary method for performing anomaly detection and reporting.

[0038] [Figure 10] This diagram shows a flowchart illustrating an exemplary method for processing power anomaly reports and determining the type and location of the power event corresponding to the power anomaly report.

[0039] [Figure 11] This figure illustrates components of one embodiment of a computing device for performing any of the techniques described herein. [Modes for carrying out the invention]

[0040] Power companies may use distributed computing and machine learning to improve the detection and identification of power events along the power grid. Power sensors may be installed at various points along the power grid, such as the locations of end-users of power (e.g., in electric meters or electric panels). These locations are referred to as power monitoring locations. Power monitoring locations may perform processing such as identifying anomalies in electrical signals and / or transmit information about electrical signals to processing locations such as server computers. Processing locations may receive information from monitoring locations and process the information to determine whether a power event has occurred, what type of power event it is (e.g., a tree falling on a power line), or the location of the power event.

[0041] Figure 1 is an exemplary representation of power grid 100. The example in Figure 1 is not limiting. Numerous variations of power grids are possible, and the techniques described herein may be applied to any type of power grid, including different power transmission standards used in different countries.

[0042] In Figure 1, power plant 110 generates electricity. Any suitable power generation technology may be used, such as coal, natural gas, hydroelectric, nuclear, wind, or solar.

[0043] Higher voltage transmission increases efficiency and reduces costs, so the high-voltage transformer 120 increases the voltage of power for power transmission over long distances.

[0044] The high-voltage transformer 120 may transmit power to one or more substations, such as substations 130-132. Each substation may supply power to a geographical area and may perform operations such as converting voltage to a lower voltage.

[0045] Each substation may further transmit power to one or more distribution transformers, such as distribution transformers 140-142. Each distribution transformer may supply the power provided by the substation to a subset of the geographical area. Distribution transformers may perform operations such as converting voltage to a lower voltage.

[0046] Each distribution transformer may further transmit power to one or more end users, such as end users 150-155. For clarity of explanation, the technology described herein may use “residence” as an exemplary end user, but the technology may be applied to any type of end user, such as a condominium or a business.

[0047] The electrical path from the distribution transformer to the end user may be mathematically represented as a tree structure. For example, a single set of wires (e.g., two live and neutral wires) may emerge from the distribution transformer 141 and supply power to end users 150-152. The power distribution may then branch into two sets of wires, with the first set supplying power to end user 150 and the second set supplying power to end users 151 and 152. The points where the wires branch may be referred to as common coupling points, such as common coupling point 160. Similarly, common coupling point 161 may be used to branch the wires to end users 151 and 152.

[0048] To monitor the performance of the power grid, electrical sensors may be installed at various locations. For example, sensors may be installed at any of the following locations: power plant 110, high-voltage transformer 120, substations 130-132, distribution transformers 140-142, and end-users 150-155. Any suitable sensor may be used. For example, the sensor may measure voltage and / or current levels. The measurements may be obtained using any suitable sensor, and the technique is not limited to any particular sensor or any particular type of value that may be obtained from a sensor.

[0049] Power monitoring data from different sources may be transmitted to a processing location, such as a server computer or other computing resource. In some implementations, raw sensor data may be transmitted, and in some implementations, the power monitoring data may be processed sensor data, and the results of processing (e.g., feature vectors or classifier outputs) may be transmitted. Any suitable network may be used to transmit the power monitoring data. For example, a power monitoring location may have a wired or wireless local area connection that enables the transmission of power monitoring data to the processing location over the Internet. In another example, a power monitoring location may have a cellular or mobile network connection that may be used to transmit power monitoring data to the processing location. In some implementations, a mesh network or a broadband network over power lines may be used.

[0050] Figure 2 shows an exemplary network architecture that may be used to transmit power monitoring data from a power monitoring location to a processing location. In Figure 2, dotted lines represent the transmission of data from one location to another. Different power monitoring locations may use different types of networks. For example, an end user 150 may transmit power monitoring data using a combination of a local Wi-Fi network and a cable modem to connect to the internet, and a distribution transformer 141 may have a cellular network connection.

[0051] In Figure 2, the computer network 220 may include any combination of network resources for transmitting data from one location to another, such as any resources used by an Internet service provider or a cellular network provider. The processing location 210 may be located in any suitable location, such as a building owned by a power company or a cloud computing provider. The processing location 210 may have any suitable resources for receiving power monitoring data via network connectivity and for processing the power monitoring data to determine information about power events. The processing location 210 may include, for example, one or more server computers.

[0052] Figure 3 shows an exemplary power monitoring location corresponding to an end user. For example, Figure 3 may correspond to end user 150. End user 300 may correspond to any suitable building, part of a building, or multiple buildings. For example, end user 300 may be the residence of a power residential end user.

[0053] An end-user 300 may have an electric meter 310 provided by the utility company to measure the end-user 300's electricity consumption. In the example in Figure 3, the power supply is shown as two live (thick) and neutral (thin) electric wires, corresponding to typical split-phase power in the United States. However, the techniques described herein may be used with any suitable power system, such as single-phase, two-phase, or three-phase power systems.

[0054] The electric meter 310 supplies power to the electric panel 320, which is compatible with any type of electric panel. The electric panel 320 may have multiple circuits, and the circuits may be connected to one or both of the live power lines. The electric panel 320 may distribute the power it receives to the circuits for use by devices on the multiple circuits. For example, the first circuit may supply power to devices 351-353, and the second circuit may supply power to devices 361-363.

[0055] The end user 300 may have a power monitor 330 that measures the electrical characteristics of the power supplied to the end user 300. For example, the power monitor 330 may have current sensors and / or voltage sensors attached to any combination of the first live line, the second live line, and the neutral line. The power monitor 330 may perform any appropriate processing of the sensor signal, such as any of the operations described in, for example, U.S. Nonprovisional Patent Application No. 14 / 707,665 (currently issued as U.S. Patent No. 9,443,195) (SAGE-0002-U01), filed 8 May 2015, U.S. Nonprovisional Patent Application No. 16 / 912,013 (currently issued as U.S. Patent No. 11,146,868) (SAGE-0005-U01-C01), filed 25 June 2020, and U.S. Nonprovisional Patent Application No. 16 / 179,567 (currently issued as U.S. Patent No. 10,878,343) (SAGE-0006-U01), which are incorporated herein by reference in their entirety.

[0056] The power monitor 330 may perform anomaly detection and / or event detection as described herein. In some examples, the power monitor 330 may be a device installed separately from the electric meter 310 and the electric panel 320. In some examples, the power monitor 330 may be part of the electric meter 310 or part of the electric panel 320. The power monitor 330 may receive power directly from the power lines or from the circuitry of the electric panel 320.

[0057] The power monitor 330 may have a connection to a network device 340 to transmit information to another location, such as the processing location 210. Any suitable network connection may be used. For example, the power monitor 330 may have a Wi-Fi connection to the network device 340, and the network device 340 may be a cable modem connected to the internet. As another example, the power monitor 330 may have a cellular connection and may not use the network device 340 at all.

[0058] Although not shown in Figure 3, the end user 300 may have other electrical components. For example, the end user 300 may have solar panels or a generator, and the power monitor 330 may have additional sensors to measure the electrical characteristics of the power generated by the solar panels or generator. As another example, the end user 300 may have a residential battery, and the power monitor 330 may have sensors to measure the amount of power stored and discharged from the battery.

[0059] The power monitor may process sensor signals from power lines to detect anomalies that may occur in the lines. The cause of the anomaly may be inside or outside the end-user's building. For example, the anomaly may be caused by an arc fault, a failure of a solar panel or battery (e.g., an inverter failure), a failure of a device in the dwelling, or a transformer failure. When an anomaly occurs, the power monitor may transmit a notification and / or other information to another location, such as a processing location.

[0060] Figure 4 shows an exemplary anomaly reporting system 400, which may be implemented by a power monitor.

[0061] In Figure 4, the anomaly detection component 410 may sequentially process sensor signals, such as any combination of current and / or voltage sensor signals from two live and neutral lines. The anomaly detection component 410 may process the sensor signals to determine whether an anomaly has occurred. In some implementations, the anomaly detection component 410 may compute a streaming digital sample of the power monitoring signal, compute a streaming sequence of feature vectors from the sensor signals, and process the feature vectors using an anomaly detector. For example, a detector that outputs a detection score may be implemented, and an anomaly may be determined to have occurred when the detection score exceeds a threshold.

[0062] The anomaly detection component 410 may also process other types of sensor signals, such as from local temperature or humidity sensors. Temperature or humidity near the power monitor may differ from local weather information near events such as overheating, fire, water damage, and flooding, which may point to the cause of the power anomaly.

[0063] The anomaly classification component 420 may also process the sensor signal (and / or the output of the anomaly detection component 410, such as a feature vector) to determine the type of anomaly that has occurred. For example, a classifier may be implemented that outputs a vector of classification scores, each of which corresponds to a type of anomaly, and the type of anomaly may be determined as corresponding to the highest classification score. The anomaly classification component 420 may be configured to classify any appropriate power anomaly type, such as one or more of series high impedance faults, series arc faults, parallel arc faults, parallel low impedance faults, anomaly power, or power loss.

[0064] In some implementations, the anomaly detection component 410 and the anomaly classification component 420 may be combined into a single detection and classification component (one of which is the classification in which no anomalies have occurred), while in some implementations, they may be separate components as described in Figure 4.

[0065] Detection and classification may utilize information other than sensor signals to improve the accuracy of detection and classification. For example, current weather conditions may be an important factor in determining whether an anomaly has occurred and what type of anomaly it is (for example, strong winds may increase the likelihood of an anomaly occurring due to trees falling onto power lines). Weather data may include, for example, current temperature, recent low or high temperatures, or average temperatures over time. Similar data may be obtained for wind speed, sunshine, humidity, rainfall, or snowfall.

[0066] As another example, information about a building can be an important factor in determining whether and what type of anomaly has occurred. This information may include any appropriate information such as the type of building (e.g., commercial, residential, apartment, or single-family home), the building's area, the building's construction date (which may include, for example, overhead lines), the building's electricity supply (e.g., amperage), whether the building has solar panels, batteries, or a generator, and whether the power lines to the building are above ground or underground.

[0067] The anomaly detection component 410 and the anomaly classification component 420 (or a combined version thereof) may be implemented using similar techniques such as the following:

[0068] In some implementations, features may be calculated from the sensor signal. For example, features may be calculated for every 60Hz power cycle, for every specified number of 60Hz power cycles, or for every window of the sensor signal corresponding to a specified time (e.g., 1 second). The calculated features may include any of the following: the number of cycles above or below a specified root-mean-square (RMS) voltage, the percentage of power in the bandwidth around the fundamental frequency or other frequencies, total harmonic distortion, distortion rate, features calculated by a recurrent neural network, or waveform comparisons (e.g., to capture waveform asymmetry).

[0069] In some implementations, features may include information about individual devices within a house, which may be obtained by performing decomposition techniques on sensor signals, such as any of the techniques described in the patent incorporated by reference. For example, if a power anomaly frequently occurs simultaneously with a change in the state of a device within a building, the power anomaly may be caused by a device within the building.

[0070] The features may be processed using any suitable detector or classifier. For example, the detector or classifier may be implemented using any of the following: a neural network (e.g., a recurrent neural network, a convolutional neural network, a graph neural network, or a transformer neural network), a self-organizing map, a support vector machine, a decision tree, a random forest, a Gaussian mixture model, or gradient boosting. The detector or classifier may take any suitable data as input, such as features computed from sensor signals or weather data.

[0071] The anomaly reporting component 430 may receive detection / classification results and may decide whether to report an anomaly and how to report the anomaly. In some implementations, information about the anomaly may be recorded locally and not transmitted to any other location. In some implementations, information about the anomaly may be transmitted to another location, such as a server computer at the processing location. The recipient of the anomaly report may be an end user, a power company, or another recipient.

[0072] The report may include any appropriate information about the anomaly, such as, namely, one or more of the following: a power monitor or end-user identifier (e.g., Internet Protocol address, Media Access Control address, or unique identifier), a sample of the sensor signal or power monitoring signal, the time corresponding to the anomaly (e.g., start time, end time, or duration), the location corresponding to the power monitor or end-user (e.g., postal code or full address), or one or more of the power anomaly types.

[0073] Various factors may be used to determine whether and how an anomaly should be reported. Anomaly reporting settings may be configured by the end user, the power company, or another entity. Settings may be based on a desired level of privacy, detection and / or classification scores (e.g., reporting only highly reliable anomalies), or any other appropriate factors. In some examples, duplicate anomalies may not be reported. For example, if an anomaly is detected every minute for an hour, only the first anomaly may be reported. In some examples, anomalies may be reported for events occurring outside the end user's building (e.g., a tree falling onto a power line), but not for events occurring inside the end user's building (e.g., caused by an electrical appliance inside the building). In some examples, anomaly reporting settings may be based on the available computer network and / or bandwidth. For internet connections with higher bandwidth, more anomalies may be reported with more information about the anomalies (e.g., sensor waveform data). For mesh networks with lower bandwidth, minor anomalies may be reported with less information about the anomalies (e.g., only that an anomaly occurred).

[0074] In some examples, the anomaly reporting component 430 may receive and respond to requests for information. The processing location may request a specific power monitoring location to provide information about an anomaly, or to provide additional information about a previously reported anomaly.

[0075] In some implementations, the anomaly reporting system 400 may be configured to determine that power is about to be lost or that the main power supply has been lost and the system is operating on a backup power supply. In such examples, the anomaly reporting system 400 may be configured to store information about the current or recent operation of the power grid and to transmit the information at a later date when power is restored.

[0076] The anomaly reporting system 400 may use any appropriate technique to determine that a power loss is imminent. In some implementations, a separate detector or classifier may be used to predict an upcoming power loss, and the detector or classifier may be implemented using any of the techniques described herein. In some implementations, the output of the anomaly classification component 420 may indicate an upcoming power loss. If the anomaly reporting system 400 loses its main power, any appropriate backup power source, such as a battery or capacitor, may be used.

[0077] The anomaly reporting system 400 may store any appropriate information, such as current or voltage sensor signals, anomaly detector outputs, or anomaly classifier outputs, when it is determined that power is about to be lost or has been lost. When power is restored, the stored information may be transmitted using any of the techniques described herein.

[0078] In some implementations, the anomaly reporting system 400 may be extended to report a wider variety of information, such as periodic status updates or non-anomalous events. For example, the anomaly reporting system 400 may have one or more of the following components: an event detection component, an event classification component, or an event reporting component.

[0079] Figures 5A and 5B are exemplary waveforms that may be obtained from sensors of a power monitor. Figure 5A is an exemplary voltage waveform during assumed operation of the power grid when no anomalies are present. Figure 5B is an exemplary voltage waveform for an anomaly caused by a series arc fault in the electrical panel connection between the backplane and the main breaker. The anomaly in Figure 5B may be detected by an anomaly detection component 410, classified by an anomaly classification component 420, and reported by an anomaly reporting component 430.

[0080] The processing location may receive anomaly reports from numerous power monitoring locations spanning various geographical areas, such as a county, state, or multiple states. In some examples, the processing location may be based in a state and receive anomaly reports from any power monitoring location within that state.

[0081] The processing location may store anomaly reports and process them to determine whether a power event has occurred, the type of power event, and the location of the power event. As used herein, a power event may refer to any type of event that may be of interest to a power company or any other entity providing products or services related to the power grid. For example, a power event may include any of the following: equipment damaged by fire or high temperature, transformer failure (which may be further divided into specific types of transformer failures such as winding failure, tap changer failure, and cooling failure), arc discharge by plants, overhead line failure, grounding, neutral line energization, arc failure (e.g., series and / or parallel), power theft, power generation failure by solar panels, power generation failure by batteries, power generation failure by transformers, and abnormal power due to a faulty or out-of-spec device (e.g., a device not configured to receive split-phase power being connected).

[0082] Power event types may be the same as or different from anomaly types. For example, an arc fault may be both an anomaly type and a power event type. In some examples, power event types may be broader than anomaly types (for example, multiple anomaly types may correspond to the same event type). In some examples, anomaly types may be broader than power event types (for example, multiple power event types may correspond to the same anomaly type).

[0083] The processing location may store anomaly information in a database. Figure 6 shows an exemplary database table that may be used to store information about anomalies. The database table may store any appropriate information about anomalies, such as the identifier of the end user or power monitor that reported the anomaly, the location information of the end user or power monitor that reported the anomaly (e.g., postal code), the time corresponding to the anomaly, the type of anomaly (e.g., as determined by the power monitor that reported the anomaly or as determined at the processing location), a time cluster identifier, or an anomaly cluster identifier.

[0084] Determining whether a power event has occurred, the type of power event, and the location of the power event may be improved by processing anomalies identified at multiple power monitoring locations. For example, suppose a tree has fallen onto the power line between distribution transformer 141 and common coupling 160. In this situation, it may be predicted that an anomaly will occur at end users 150-152, but not at end users 153-155. Conversely, the location (or approximate location) of a power event may be determined by processing power anomalies received from multiple end users.

[0085] To identify power events, stored anomalies may be clustered or grouped according to various criteria such as anomaly time, anomaly location, or anomaly type.

[0086] For some power events, multiple power monitoring locations may experience anomalies at the same time. To identify power events more efficiently and quickly, anomalies may be assigned to one or more time clusters. Including time cluster identifiers in the database can improve the efficiency of event retrieval and processing. Any suitable technique may be used to assign one or more time clusters. For example, a time cluster may be created each time a threshold number of anomalies occur within a time threshold (e.g., more than three anomalies in less than five minutes). Alternatively, a cluster may be created for all anomalies occurring within a time period, such as creating a cluster every minute (e.g., a first cluster for anomalies occurring between 1:01:00 PM and 1:01:59 PM, and a second cluster for anomalies occurring between 1:02:00 PM and 1:02:59 PM).

[0087] For certain power events, multiple power monitoring locations may experience anomalies within their geographical area. For example, if a tree falls onto a power line, all end users receiving power through that line may experience an anomaly, and these end users may be geographically close to each other.

[0088] Anomalies may also be clustered by anomaly type. If two anomalies occur at the same time and location but have different anomaly types, the two anomalies may be unrelated to each other and may have occurred coincidentally at the same time and location.

[0089] Therefore, to facilitate anomaly processing, one or more cluster identifiers may be stored along with the anomaly. In the example in Figure 6, time clusters are stored for each anomaly, such that anomalies from similar time periods are in the same cluster. Additionally, the entire anomaly cluster is stored for anomalies that are in the same time cluster, occur in similar geographical locations, and have the same anomaly type.

[0090] Computational resources may process data from an anomaly database to determine whether a power event occurred, the type of power event, and / or the location of the power event.

[0091] Figure 7 shows an exemplary system 700 for clustering anomalous data. System 700 may be implemented using any suitable computing resources, such as on-premises computing resources and / or computing resources provided by third parties.

[0092] In Figure 7, the anomaly clustering component 710 may perform anomaly clustering using any of the techniques described herein. The anomaly data store 720 may store information about anomalies both before and after clustering. In some implementations, the first anomaly data store may store information about unclustered anomalies, and the second anomaly data store may store information about clustered anomalies.

[0093] The anomaly clustering component 710 may perform anomaly clustering using any suitable technique, such as any of the techniques described herein. Clustering may be performed at any suitable time interval. For example, clustering may be performed at regular time intervals, such as every minute or every hour, or after the number of unclustered anomalies exceeds a threshold. Clustering may be performed in response to (and / or in anticipation of / preparation for) one or more events and / or a set of events.

[0094] The anomaly clustering component 710 may cluster based on any appropriate criterion, such as location, time, or one or more anomaly types. A time difference threshold may be used to cluster anomalies by time, and a location difference threshold may be used to cluster anomalies by location. The anomaly clustering component 710 may retrieve anomalies from the anomaly data store 720, perform clustering on those anomalies, and then write the clustered anomaly information back to the anomaly data store 720.

[0095] Figure 8 shows an exemplary system 800 for processing abnormal data to determine one or more of the following: whether a power event occurred, the type of power event, and / or the location of the power event. System 800 may be implemented using any suitable computing resources, such as on-premises computing resources and / or computing resources provided by third parties.

[0096] In Figure 8, the anomaly selection component 810 may select a subset of anomalies to process. The anomaly selection by the anomaly selection component 810 may be performed at any appropriate interval, such as the interval described above, with respect to the anomaly clustering component 710. In some implementations, the anomaly selection component 810 may also perform clustering operations, and in some implementations, the anomaly selection component 810 may select previously clustered anomalies.

[0097] Any suitable subset of anomalies may be selected. A subset of anomalies may be referred to as an anomaly cluster or group. In some implementations, each anomaly may be part of a cluster, and in some implementations, a cluster may consist of a single anomaly. In some implementations, only some anomalies may be part of a cluster, or only clusters that meet specified criteria (e.g., clusters containing the minimum number of anomalies) may be processed further. In some implementations, an anomaly cluster may contain anomalies that are within a specified time and within a specified distance from each other. In some implementations, an anomaly cluster may contain anomalies that are within a specified time and distance and have the same or similar anomaly types. Each selected subset of anomalies may then be processed further.

[0098] The power event classification component 820 may receive information about a subset of anomalies from the anomaly selection component 810, process the information, and determine the power event type corresponding to the subset of anomalies. Any appropriate power event type may be determined, namely, one or more of the following: anomaly power from solar panels or batteries, anomaly power from consuming devices, anomaly power from grid supply, power loss, parallel faults due to plants, series faults due to loose or corroded connections, or power theft. The power event classification component 820 may use any appropriate technique to determine the power event type corresponding to the subset of anomalies.

[0099] The power event classification component 820 may process any appropriate data relating to a subset of anomalies, including, but not limited to, waveform data from power monitoring locations (e.g., current and / or voltage waveforms from any of the power monitoring locations in Figure 1), anomaly types (e.g., as determined at the power monitoring location or at the processing location), information from nearby power monitoring locations that did not report anomalies (e.g., waveform data, or the number of locations that reported or did not report anomalies), and meteorological data (e.g., any of the meteorological data described herein). In some implementations, one or more feature vectors may be created from the above, and these feature vectors may be processed by a classifier to determine the power event type.

[0100] In some implementations, the processing location may send requests for additional data to the power monitoring location. For example, if the power monitoring location reports an anomaly and provides some data, the processing location may request that the power monitoring location provide additional data (e.g., higher resolution waveform data). In another example, the processing location may request data from a power monitoring location that is geographically close to the power monitoring location that reported the anomaly, even though the power monitoring location has not reported an anomaly.

[0101] The power event classification component 820 may use any suitable classification method, such as one of the classification techniques described herein for anomaly classification. In some implementations, the power event classification component 820 may compute a vector of scores, likelihoods, or probabilities, where each element of the vector corresponds to a power event type. The power event type may be selected to correspond to the maximum value of the vector. For example, the vector may be the output of a softmax layer of a neural network.

[0102] The power event localization component 830 receives information about a subset of anomalies, processes the information, and determines the location corresponding to the power event. The processing of the power event classification component 820 and the power event localization component 830 may be performed in any order or in parallel. The power event localization component 830 may use the output of the power event classification component 820, and vice versa.

[0103] The determined location of a power event may be specified using any appropriate technique. In some implementations, the location may be specified geographically, for example, within a specified distance of one or more power monitoring locations, or within the vicinity of one or more power monitoring locations (e.g., within a circle surrounding one or more power monitoring locations). In some implementations, the location may be specified electrically to correspond to a portion of the power grid. For example, a power event may be determined to have occurred between the distribution transformer 141 and the common coupling 160 in Figure 1. In some implementations, the location may be probabilistic, having a probability of occurring at a geographical and / or electrical location. In some examples, the location may be determined to be inside a building corresponding to a power monitoring location (e.g., caused by a faulty piece of equipment located inside the building).

[0104] The power event localization component 830 may determine the location of a power event using any suitable technique. Location determination may be based on the power event type and / or anomaly type. In some implementations, location may be determined using one or more of the following: neural networks (e.g., recurrent neural networks, convolutional neural networks, graph neural networks, or transformer neural networks), self-organizing maps, support vector machines, decision trees, random forests, Gaussian mixture models, or gradient boosting.

[0105] Information about power events, such as type, location, and time, may be stored in any suitable location, such as a power event data store.

[0106] The power event notification component 840 may receive information about the type and / or location of a power event and may decide whether and how to issue a notification. For example, if a power event does not involve any associated risks and can be rectified by routine maintenance, a notification may not be issued. As another example, if it is determined that a power event occurred inside the end user's building, a notification may be sent to the end user (e.g., via postal mail, email, or an application such as a smartphone application). As yet another example, if it is determined that an event occurred outside the end user's building at a location on the power grid, a notification may be sent to the appropriate person in charge at the power company to take action.

[0107] In some implementations, notifications may be presented on a map. For example, a web application or other application may present a map of a portion of the service area. Markers may be placed at power monitoring locations reporting anomalies and / or at power events. Markers may be presented using any appropriate technique, such as color coding corresponding to different event types or size corresponding to the severity of the event.

[0108] Figure 9 is a flowchart illustrating an exemplary method for performing anomaly detection and reporting. Figure 9 may be performed by any suitable device, such as a power monitor located at a power monitoring location (e.g., an end-user building, a distribution transformer, or a substation).

[0109] In step 910, a power monitoring signal is obtained from a sensor that measures the electrical characteristics of the wire. The power monitoring signal may be obtained using any suitable technique, by measuring any suitable electrical characteristics (e.g., current or voltage), or by using any suitable sensor. The power monitoring signal may be streamed in real time or near real time.

[0110] In step 920, features are calculated from the power monitoring signal. Any suitable features may be calculated, such as any of the features described herein. The feature vector may be streamed in real time or near real time. In some implementations, the feature vector may not be calculated at this time and may be calculated later (for example, for power anomaly classification).

[0111] In step 930, it is determined that a power anomaly has occurred. Any suitable technique, such as one of the techniques described herein, may be used to determine that a power anomaly has occurred. In some implementations, a power monitoring signal, or a feature vector calculated from a power monitoring signal, may be processed by an anomaly detector to calculate a detection score indicating the probability or likelihood that an anomaly has occurred.

[0112] In step 940, the power anomaly type corresponding to the power anomaly is determined. Any suitable technique, such as one of the techniques described herein, may be used to determine the power anomaly type. In some implementations, the power monitoring signal or the feature vector computed by the power monitoring signal may be processed by an anomaly classifier to compute a classification score that points to the probability or likelihood for different anomaly types.

[0113] In some implementations, steps 930 and 940 may be combined into a single step that determines both whether an anomaly has occurred and the type of anomaly. For example, the classifier may be used if one of its outputs corresponds to no anomaly occurring.

[0114] In step 950, it is decided to report a power anomaly. Any suitable technique, such as one of the techniques described herein, may be used to determine whether to report an anomaly and where and how to report the anomaly.

[0115] In step 960, a power anomaly report is transmitted. Any suitable technique, such as any of the techniques described herein, may be used to transmit the power anomaly report.

[0116] Figure 10 is a flowchart illustrating an exemplary method for processing power anomaly reports and determining the type and location of the power event corresponding to the power anomaly report.

[0117] In step 1010, a first power anomaly report is received from a first power monitor located in the first building. The first power anomaly report may include any appropriate information about the first power anomaly, such as first location information corresponding to the first building, first time information corresponding to the first power anomaly, first power anomaly type of the first power anomaly, and any other power anomaly information described herein.

[0118] In step 1020, a second power anomaly report is received from a second power monitoring device located in the second building. This step may be carried out as described in step 1010.

[0119] In step 1030, a first power anomaly type is determined for the first power anomaly by processing the first power anomaly information using a classifier. This step may be performed at a power monitoring location, transmitted to a processing location (e.g., in a power anomaly report), or performed at a processing location. In some implementations, the power monitoring location may determine the initial power anomaly type, and the processing location may calculate the revised power anomaly type (e.g., using larger computing resources or data).

[0120] In step 1040, the second power anomaly type is determined for the second power anomaly by processing the second power anomaly information using a classifier. This step may be carried out as described in step 1030.

[0121] The above steps may be additionally performed for any number of power anomaly reports from any number of locations. For example, a third power anomaly report may be received from a third power monitoring device located in a third building, and a third power anomaly type may be determined for the third power anomaly. Power anomaly reports may be received from any suitable location, such as any of the locations in Figure 1.

[0122] In step 1050, a subset of power anomaly reports is selected from the available power anomaly reports. The available power anomaly reports may include first and second power anomaly reports. Any suitable technique may be used to select the subset of power anomaly reports, such as any of the techniques described herein. In some implementations, power anomalies may be clustered, and the subset may correspond to the clusters. In some implementations, the subset may correspond to power anomalies having similar locations, similar times, and / or power anomaly types.

[0123] In step 1060, the power event type and / or power event location are determined for power events corresponding to a subset of power anomaly reports. Any suitable technique, such as one of the techniques described herein, may be used to determine the power event type and / or the location of the power event.

[0124] In step 1070, a power event notification is issued in response to a power event. The power event notification may be issued using any suitable technique, such as one of the techniques described herein.

[0125] Figure 11 illustrates a component of one embodiment of a computing device 1100 for carrying out any of the techniques described herein. In Figure 11, the component is shown as being on a single computing device, but the component may be distributed across multiple computing devices, such as a system of computing devices, including, for example, end-user computing devices (e.g., smartphones or tablets) and / or server computers (e.g., cloud computing).

[0126] The computing device 1100 may include any components typical of a computing device, such as volatile or non-volatile memory 1110, one or more processors 1111, and one or more network interfaces 1112. The computing device 1100 may also include any input and output components, such as a display, keyboard, and touchscreen. The computing device 1100 may also include various components or modules that provide specific functions, and these components or modules may be implemented in software, hardware, or a combination thereof. The computing device 1100 may include one or more non-temporary computer-readable media containing computer-executable instructions that, when executed, cause a processor to perform an action corresponding to one of the techniques described herein. Some examples of components are described below as an exemplary embodiment, but other embodiments may include additional components and may exclude some of the components described below.

[0127] The computing device 1100 may have an anomaly detection component 1120 which may detect power anomalies using any of the techniques described herein. The computing device 1100 may have an anomaly classification component 1121 which may classify power anomalies into power anomaly types using any of the techniques described herein. The computing device 1100 may have an anomaly reporting component 1122 which may report power anomalies using any of the techniques described herein. The computing device 1100 may have an anomaly clustering component 1123 which may cluster power anomalies using any of the techniques described herein. The computing device 1100 may have an anomaly selection component 1124 which may select one or more power anomalies using any of the techniques described herein. The computing device 1100 may have a power event classification component 1125 which may classify power events into power event types using any of the techniques described herein. The computing device 1100 may have a power event location component 1126 which may determine the location corresponding to a power event using any of the techniques described herein. The computing device 1100 may have a power event notification component 1127 that may provide notifications regarding power events using any of the techniques described herein.

[0128] The computing device 1100 may include or access various data stores. The data stores may use any known storage technology, such as files, relational databases, non-relational databases, or any non-temporary computer-readable media. The computing device 1100 may have an anomaly data store 1130 for storing information about reported power anomalies. The computing device 1100 may have a power event data store 1131 for storing information about power events.

[0129] Although only a few embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that many changes and modifications may be made thereto without departing from the spirit and scope of this disclosure, as described in the following claims. All foreign and national patent applications and patents, as well as all other publications referenced herein, are incorporated herein in their entirety to the maximum extent permitted by law.

[0130] The methods and systems described herein may be partially or entirely arranged through computer-readable instructions, program code, computers, computing devices, processors, circuits, and / or servers having machines that execute instructions, and / or include hardware configured to functionally perform one or more operations of the methods and systems described herein. The terms computer, computing device, processor, circuit, and / or server ("computing device") should be understood broadly when used herein.

[0131] An exemplary computing device includes any type of computer capable of communicating and accessing instructions stored in a non-temporary computer-readable medium, such as a non-temporary computer-readable medium, and the computer performs the operations of the computing device when it executes an instruction. In certain embodiments, such instructions themselves constitute the computing device. Additionally or alternatively, the computing device may include separate hardware devices, one or more computing resources distributed across hardware devices, and / or logic circuits, embedded circuits, sensors, actuators, input and / or output devices, network and / or communication resources, any type of memory resource, any type of processing resource, and / or hardware devices configured to functionally perform one or more operations of the systems and methods herein in response to determined conditions.

[0132] Networks and / or communication resources include, but are not limited to, local area networks, wide area networks, wireless, the Internet, or other known communication resources and protocols. Exemplary and non-limiting hardware and / or computing devices include, but are not limited to, general-purpose computers, servers, embedded computers, mobile devices, virtual machines, and / or emulated computing devices. A computing device may be a distributed resource, included as several device embodiments such that distributed resources work together to perform the operation of the computing device, and also as an interoperable set of resources for performing the described function of the computing device. In certain embodiments, each computing device may be on separate hardware, and / or one or more hardware devices may include multiple computing device embodiments, for example, as individually executable instructions stored on the device, and / or as logically divided embodiments of an executable instruction set, some embodiments including part of a first computing device, and some embodiments including part of another computing device.

[0133] A computing device may be a server, client, network infrastructure, mobile computing platform, stationary computing platform, or part of another computing platform. A processor may be any type of computing or processing device capable of executing program instructions, code, and binary instructions, etc. A processor may be any variation of a signal processor, digital processor, embedded processor, microprocessor, or coprocessor (such as a numerical coprocessor, graphics coprocessor, and communications coprocessor) that may directly or indirectly facilitate the execution of stored program code or program instructions, or may include such a processor. A processor may also enable the execution of multiple programs, threads, and code. Threads may run concurrently to enhance the performance of the processor and facilitate the concurrent operation of applications. As a method of implementation, the methods, program code, and program instructions described herein may be implemented by one or more threads. Threads may spawn other threads which may have associated assigned priorities, and the processor may execute these threads based on priority or on any other order based on instructions provided in the program code. A processor may include memory for storing methods, code, instructions, and programs as described herein and elsewhere. The processor may access, through an interface, a storage medium that may store methods, code, and instructions as described herein and elsewhere. The storage medium associated with the processor for storing methods, programs, code, program instructions, or other types of instructions executable by the computing or processing device may include, but are not limited to, one or more of the following: CD-ROMs, DVDs, memory, hard disks, flash drives, RAM, ROM, and caches.

[0134] A processor may include one or more cores that can enhance the speed and performance of a multiprocessor. In practice, a processor may be a dual-core processor, a quad-core processor, or other chip-level multiprocessor, combining two or more independent cores (called dies).

[0135] The methods and systems described herein may be partially or entirely configured through machines that execute computer-readable instructions on servers, clients, firewalls, gateways, hubs, routers, or other computers and / or networked hardware. Computer-readable instructions may be associated with servers, which may include file servers, print servers, domain servers, internet servers, intranet servers, and other variations such as secondary servers, host servers, and distributed servers. A server may include one or more of the following: memory, processors, computer-readable temporary and / or non-temporary media, storage media, ports (physical and virtual), communication devices, and interfaces that allow access to other servers, clients, machines, and devices via wired or wireless media. Methods, programs, or code described herein and elsewhere may be executed by a server. Other devices necessary for executing methods described herein may also be considered part of the infrastructure associated with the server.

[0136] The server may provide interfaces to other devices, including but not limited to clients, other servers, printers, database servers, print servers, file servers, communication servers, and distributed servers. This coupling and / or connection may also facilitate the remote execution of instructions across a network. Networking some or all of these devices may, without departing from the scope of this disclosure, facilitate the parallel processing of program code, instructions, and / or programs at one or more locations. Furthermore, all devices attached to the server through the interface may include at least one storage medium capable of storing methods, program code, instructions, and / or programs. A central repository may provide program instructions to be executed on different devices. In this implementation, a remote repository may act as a storage medium for methods, program code, instructions, and / or programs.

[0137] Methods, program code, instructions, and / or programs may be associated with clients, which may include file clients, print clients, domain clients, internet clients, intranet clients, and other variations such as secondary clients, host clients, and distributed clients. A client may include one or more of the following: memory, processors, computer-readable temporary and / or non-temporary media, storage media, ports (physical and virtual), communication devices, and interfaces that allow access to other clients, servers, machines, and devices via wired or wireless media. Methods, program code, instructions, and / or programs as described herein and elsewhere may be executed by a client. Other devices necessary for executing methods as described herein may also be considered part of the infrastructure associated with the client.

[0138] The client may provide interfaces to other devices, including but not limited to servers, other clients, printers, database servers, print servers, file servers, communication servers, and distributed servers. This coupling and / or connection may also facilitate the remote execution of methods, program code, instructions, and / or programs across a network. Networking some or all of these devices may, without departing from the scope of this disclosure, facilitate the parallel processing of methods, program code, instructions, and / or programs at one or more locations. Furthermore, all devices attached to the client through the interface may include at least one storage medium capable of storing methods, program code, instructions, and / or programs. A central repository may provide program instructions to be executed on different devices. In this implementation, a remote repository may act as a storage medium for methods, program code, instructions, and / or programs.

[0139] The methods and systems described herein may be partially or entirely deployed through a network infrastructure. The network infrastructure may include elements such as computing devices, servers, routers, hubs, firewalls, clients, personal computers, communication devices, routing devices, and other active and passive devices, modules, and / or components known in the art. Computing and / or non-computing devices associated with the network infrastructure may include storage media such as flash memory, buffers, stacks, RAM, and ROM, in addition to other components. The methods, program code, instructions, and / or programs described herein and elsewhere may be executed by one or more of the network infrastructure elements.

[0140] The methods, program code, instructions, and / or programs described herein and elsewhere may be implemented on a cellular network having multiple cells. The cellular network may be either a frequency division multiple access (FDMA) network or a code division multiple access (CDMA) network. The cellular network may include mobile devices, cell sites, base stations, repeaters, antennas, and towers, etc.

[0141] The methods, program code, instructions, and / or programs described herein and elsewhere may be implemented on or through a mobile device. Mobile devices may include navigation devices, mobile phones, personal digital assistants, laptops, palmtops, netbooks, pagers, e-readers, and music players. These devices may include, in addition to other components, storage media such as flash memory, buffers, RAM, and ROM, and one or more computing devices. The computing devices associated with the mobile device may be capable of executing the stored methods, program code, instructions, and / or programs. Alternatively, the mobile device may be configured to execute instructions in conjunction with other devices. The mobile device may communicate with a base station interfaced with a server and configured to execute the methods, program code, instructions, and / or programs. The mobile device may communicate over a peer-to-peer network, a mesh network, or other communication network. The methods, program code, instructions, and / or programs may be stored on storage media associated with the server and executed by a computing device embedded in the server. The base station may include a computing device and storage media. The storage device may store methods, program code, instructions, and / or programs executed by computing devices associated with the base station.

[0142] Methods, program code, instructions, and / or programs may be stored on and / or accessed on machine-readable non-temporary and / or non-temporary media, which may include computer components, devices, and recording media that hold digital data used for calculations over some time intervals, semiconductor storage known as random access memory (RAM), mass storage typically for more persistent storage such as optical disks, hard disks in the form of magnetic storage devices, tapes, drums, cards, and other types, processor registers, cache memory, volatile memory, non-volatile memory, optical storage such as CDs, DVDs, and removable media such as flash memory (USB sticks or keys), floppy disks, magnetic tape, paper tape, punch cards, standalone RAM disks, Zip drives, removable mass storage, and offline, and other computer memory such as dynamic memory, static memory, read / write storage, modifiable storage, read-only, random access, sequential access, location addressable, file addressable, content addressable, network-attached storage, storage area networks, barcodes, and magnetic ink.

[0143] Certain operations described herein include interpreting, receiving, and / or determining one or more values, parameters, inputs, data, or other information ("receiving data"). Operations for receiving data include, but are not limited to, receiving data via user input, receiving data over any type of network, reading data values ​​from memory locations communicating with a receiving device, using default values ​​as received data values, estimating, calculating, or deriving data values ​​based on other available information by the receiving device, and / or updating any of these in accordance with data values ​​received later. In certain implementations, data values ​​may be received by a first operation as part of receiving data values ​​and subsequently updated by a second operation. For example, a first receiving operation may be performed when communication is down, intermittent, or interrupted, and an updated receiving operation may be performed when communication is restored.

[0144] Certain logical groupings of operations as described herein, for example, methods or procedures disclosed herein, are provided to illustrate aspects of the disclosure. The operations described herein are described and / or depicted in a general manner, and the operations may be combined, divided, reordered, added, or deleted in a manner consistent with the disclosure herein. The context of the description of operations may require ordering of one or more operations, and / or the order of one or more operations may be explicitly disclosed, but the order of operations should be understood broadly, and any equivalent grouping of operations that yield equivalent results of operations is specifically assumed herein. For example, if a value is used in an operation step, the determination of the value may be required before that operation step in certain contexts (e.g., when a time delay in data is important for the operation to achieve a certain effect), but not before that operation step in other contexts (e.g., when using a value from a previous execution cycle of the operation would be sufficient for those purposes). Accordingly, in certain embodiments, the sequence and grouping of operations described herein are explicitly considered, and in certain embodiments, changes in the sequence, subdivision, and / or different groupings of operations are explicitly considered herein.

[0145] Methods and systems described herein may transform physical and / or intangible items from one state to another. Methods and systems described herein may also transform data representing physical and / or intangible items from one state to another.

[0146] The methods and / or processes described above, and their steps, may be implemented in hardware, program code, instructions, and / or programs, or any combination of hardware and methods, program code, instructions, and / or programs, that are appropriate for a particular application. The hardware may include a dedicated computing device or a specific computing device, a specific aspect or component of a specific computing device, and / or an arrangement of hardware components and / or logic circuits to perform one or more of the operations of the methods and / or systems. The processes may be implemented in one or more microprocessors, microcontrollers, embedded microcontrollers, programmable digital signal processors, or other programmable devices, together with internal and / or external memory. The processes may also, or instead, be implemented in application-specific integrated circuits, programmable gate arrays, programmable array logic, or any other devices or combinations of devices that may be configured to process electronic signals. It will be further noted that one or more processes may be implemented as computer executable code that can be run on machine-readable media.

[0147] Computer executable code may be written using structured programming languages ​​such as C, object-oriented programming languages ​​such as C++, or any other high-level or low-level programming languages ​​(including assembly languages, hardware description languages, and database programming languages ​​and technologies), and may be stored, compiled, or interpreted for execution not only on the devices mentioned above, but also on heterogeneous combinations of processors, processor architectures, or combinations of different hardware and computer-readable instructions, or any other machine capable of executing program instructions.

[0148] Accordingly, in one embodiment, each of the methods described above, and combinations thereof, may be embodied in computer executable code that performs those steps when executed on one or more computing devices. In another embodiment, the methods may be embodied in a system that performs those steps, or they may be distributed across devices in several ways, or all of the functionality may be integrated into a dedicated standalone device or other hardware. In yet another embodiment, the means for performing the steps related to the processes described above may include any of the hardware and / or computer-readable instructions described above. All such permutations and combinations are intended to be within the scope of this disclosure.

[0149] All documents referenced herein are incorporated herein by reference.

Claims

1. Receiving a first power anomaly report for a first power anomaly from a first power monitor, The first power monitor is located in the first building. The first power anomaly report includes (i) first location information for the first building and (ii) first time information for the first power anomaly. Receiving and Receiving a second power anomaly report regarding a second power anomaly from a second power monitor, The second power monitor is located in the second building. The second power anomaly report includes (i) second location information for the second building and (ii) second time information for the second power anomaly. Receiving and The first power anomaly report is processed using a classifier to determine the first power anomaly type for the first power anomaly, The second power anomaly report is processed using the classifier to determine the second power anomaly type for the second power anomaly, The selection involves using location and time information to select a subset of power anomaly reports from a plurality of power anomaly reports, wherein the plurality of power anomaly reports include the first power anomaly report and the second power anomaly report. Using the aforementioned subset of power anomaly reports, determine the power event type and power event location of the power event. Methods that include...

2. The method according to claim 1, wherein the first location information includes one or more of an address, a postal code, an Internet Protocol address, or a media access control address.

3. The method according to claim 1, wherein determining the first power anomaly type includes processing weather information using the classifier.

4. The method according to claim 1, wherein the first power anomaly type is one or more of a series high impedance fault, a series arc fault, a parallel arc fault, a parallel low impedance fault, an anomaly, or a power loss.

5. The method according to claim 1, wherein the classifier comprises a recurrent neural network or a transformer neural network.

6. The method according to claim 1, wherein the classifier comprises a decision tree.

7. The method according to claim 1, wherein selecting the subset of power anomaly reports includes selecting power anomalies using a time difference threshold and a location difference threshold.

8. The method according to claim 1, wherein selecting the subset of power anomaly reports includes using power anomaly types.

9. The method according to claim 1, wherein the power event type is one or more of the following: abnormal power from a solar panel or battery, abnormal power from a consuming device, abnormal power from a grid supply, power loss, parallel fault due to vegetation, series fault due to loose or corroded connections, or power theft.

10. A system comprising at least one server computer having at least one processor and at least one memory, wherein the at least one server computer is Receiving a first power anomaly report for a first power anomaly from a first power monitor, The first power monitor is located in the first building. The first power anomaly report includes (i) first location information for the first building and (ii) first time information for the first power anomaly. Receiving and Receiving a second power anomaly report regarding a second power anomaly from a second power monitor, The second power monitor is located in the second building. The second power anomaly report includes (i) second location information for the second building and (ii) second time information for the second power anomaly. Receiving and The first power anomaly report is processed using a classifier to determine the first power anomaly type for the first power anomaly, The second power anomaly report is processed using the classifier to determine the second power anomaly type for the second power anomaly, The selection involves using location and time information to select a subset of power anomaly reports from a plurality of power anomaly reports, wherein the plurality of power anomaly reports include the first power anomaly report and the second power anomaly report. Using the aforementioned subset of power anomaly reports, determine the power event type and power event location of the power event. A system configured to do so.

11. The system according to claim 10, wherein the subset of power anomaly reports includes power anomaly reports received from distribution transformers or substations.

12. The system according to claim 10, wherein the first power monitor is installed in a first electric meter for measuring the amount of electricity used by the first building.

13. The system according to claim 10, wherein the first power monitor is installed in the first electrical panel of the first building.

14. The system according to claim 10, wherein the at least one server computer is configured to determine the first power anomaly type by processing data from a temperature sensor and / or a humidity sensor.

15. The system according to claim 10, wherein the at least one server computer is configured to send power event notifications.

16. The system according to claim 10, wherein the at least one server computer is configured to display the first location information, the second location information, and the power event location on a map.

17. The system according to claim 10, wherein the at least one server computer is configured to determine the first power anomaly type for the first power anomaly by processing information about a change in the state of a device in the first building.

18. One or more non-temporary computer-readable media containing computer-executable instructions, wherein, when the computer-executable instructions are executed, Receiving a first power anomaly report for a first power anomaly from a first power monitor, The first power monitor is located in the first building. The first power anomaly report includes (i) first location information for the first building and (ii) first time information for the first power anomaly. Receiving and Receiving a second power anomaly report regarding a second power anomaly from a second power monitor, The second power monitor is located in the second building. The second power anomaly report includes (i) second location information for the second building and (ii) second time information for the second power anomaly. Receiving and The first power anomaly report is processed using a classifier to determine the first power anomaly type for the first power anomaly, The second power anomaly report is processed using the classifier to determine the second power anomaly type for the second power anomaly, The selection involves using location and time information to select a subset of power anomaly reports from a plurality of power anomaly reports, wherein the plurality of power anomaly reports include the first power anomaly report and the second power anomaly report. Using the aforementioned subset of power anomaly reports, determine the power event type and power event location of the power event. One or more non-temporary computer-readable media that cause at least one processor to perform an action including the following.

19. The power event location corresponds to a location inside the first building, one or more non-temporary computer-readable media according to claim 18.

20. The power event location corresponds to the vicinity of the first building and the second building, one or more non-temporary computer-readable media according to claim 18.

21. The first power anomaly report comprises one or more non-transient computer-readable media according to claim 18, including a sample of current or voltage from an electrical sensor.

22. The first time information includes one or more of the start time, end time, or duration of the first power anomaly, according to one or more non-temporary computer-readable media of claim 18.

Images