Automatic discovery of power supply network topology and phases

Abstract

A system for discovering the topology and phases of a power distribution system is provided. For example, a group of meters connected to the power distribution system can process sensor data acquired at the meters and generate descriptors based on the processed data, sending the descriptors to a headend system. The headend system can, after receiving descriptors from the individual meters in the system, group the meters to generate groups by applying a clustering algorithm to the descriptors of these meters. The headend system can also compare the current group with past groups to determine the confidence level of the current group, and assign segment identifiers or phase identifiers, or both, to one or more meters based on the confidence level.

Description

Technical Field

[0001] This disclosure generally relates to power supply networks, and more specifically, to the discovery of topology and phase information of power supply networks. Background Technology

[0002] Utility companies typically manually track the location of electrical meters installed in the field and their connections to distribution transformers. For large utilities, the number of meters can reach millions, with distribution transformers approaching one million, making this manual process time-consuming and error-prone. Furthermore, phase information is often not recorded due to technical complexity, labor, time constraints, cost, and equipment availability. Additionally, the actual phase designation of meters and associated upstream assets may change over time due to work performed on system assets by line workers. Similar problems may exist in three-phase distribution transformers with multiple meters connected to them. The phases of individual meters are typically not recorded but need to be determined for purposes such as managing loads on the grid. Summary of the Invention

[0003] Aspects and examples of apparatus and processes for discovering or identifying topology and phase information of assets in a power distribution system are disclosed. For example, a method for discovering the topological location and phase of one or more utility units in a resource distribution system includes receiving descriptors from a plurality of utility units connected to the power distribution system by a head-end system. The descriptors are generated at the respective utility units by processing sensor data obtained at those utility units. The method further includes: in response to determining that at least a threshold number of descriptors for a regular pattern have been received, grouping the plurality of utility units by the head-end system to generate a current group by applying a clustering algorithm to the descriptors of the plurality of utility units; comparing the current group with past groupings by the head-end system to determine a confidence level for the group; determining whether the confidence level exceeds a threshold confidence value; and, in response to determining that the confidence level exceeds the threshold confidence value, assigning at least one of a segment identifier or a phase identifier to one or more of the plurality of utility units by the head-end system.

[0004] In another example, a method performed by a utility device to generate descriptors for discovering the topology and phases of a power distribution system includes processing sensor data obtained by the utility device to generate processed data, and determining whether the utility device is operating in a fast mode or a normal mode. When the utility device operates in fast mode, it generates descriptors at a higher rate than when operating in normal mode. The method further includes: generating fast descriptors based on the processed data in response to determining that the utility device is operating in fast mode; and sending the fast descriptors to a headend system communicatively connected to the utility device in response to determining that at least a first threshold number of fast descriptors has been generated. The method also includes: generating normal descriptors based on the processed data in response to determining that the utility device is operating in normal mode; and sending the normal descriptors to the headend system in response to determining that at least a second threshold number of normal descriptors has been generated, wherein the second threshold number is higher than the first threshold number, and wherein the fast descriptors and normal descriptors are different.

[0005] In one additional example, a system includes multiple utility units and a headend system. The multiple utility units are connected to a power distribution system and communicatively connected to the headend system. Each of the multiple utility units is configured to process sensor data acquired at the utility unit to generate processed data, and to generate and transmit either a fast descriptor or a regular descriptor based on the processed data. The headend system is configured to receive the fast or regular descriptors from the multiple utility units, and determine whether the multiple utility units are operating in a fast mode or a regular mode. In fast mode, segment identifiers or phase identifiers are assigned to the utility units among the multiple utility units for a shorter time period than in regular mode. The headend system is also configured to, in response to determining that the multiple utility units are operating in the regular mode, group the multiple utility units by applying a clustering algorithm to the regular descriptors of the multiple utility units to generate a current group, compare the current group with past groups to determine a confidence level for the current group, and assign at least one of the segment identifiers or phase identifiers to one or more of the multiple utility units based on the confidence level.

[0006] These illustrative aspects and features are not intended to limit or restrict the subject matter described herein, but rather to provide examples to aid in understanding the concepts described in this application. Other aspects, advantages, and features of the subject matter will become apparent after reading the entire application. Attached Figure Description

[0007] These and other features, aspects and advantages of this disclosure should be better understood when reading the following detailed description with reference to the accompanying drawings.

[0008] Figure 1 This is a block diagram illustrating a power distribution system according to certain aspects of this disclosure.

[0009] Figure 2 This is a diagram illustrating a utility management system according to certain aspects of this disclosure, through which meters in a power distribution system can communicate with a headend system to facilitate the discovery of the topology and phases of the power distribution system.

[0010] Figure 3 This is an example of a process for generating descriptors at meters in a power distribution system, according to certain aspects of this disclosure, to facilitate topology and phase discovery of the head-end system.

[0011] Figure 4 This is an example of a process for processing sensor data at a meter in an electric power distribution system, according to certain aspects of this disclosure.

[0012] Figure 5 This is an example of a process for identifying the topology and phase of a power distribution system, according to certain aspects of this disclosure.

[0013] Figure 6 This is an example of a fast-pattern process for identifying the topology and phase of a power distribution system, based on certain aspects of this disclosure.

[0014] Figure 7 This is a block diagram illustrating an example of an electricity meter suitable for implementing various aspects of the technologies and sciences presented in this article.

[0015] Figure 8 It is a block diagram illustrating an example of a computing system suitable for implementing various aspects of the technologies and sciences presented in this article. Detailed Implementation

[0016] Systems and methods are provided for discovering or identifying topology and phase information of assets in a power distribution system. For example, assets in a power distribution system equipped with sensors (e.g., meters, transformers, generators) can be configured to collect sensor data (e.g., voltage, current, load impedance, temperature). Each asset can preprocess its sensor data before sending it to a headend system for topology and phase discovery. Preprocessing may include, for example, filtering the sensor data to remove or reduce noise or offset, transforming the sensor data to the frequency domain, or normalizing the sensor data to generate processed sensor data. In some cases, preprocessing may include enhancing noise that represents unique characteristics or signatures confined to a small group of meters. Preprocessing may also include detecting events such as disturbances or anomalies in the sensor data and determining the characteristics of the events, such as the duration, number, frequency, timing, and severity of the events.

[0017] Based on the processed sensor data and detected events, assets can be further descriptors generated. Descriptors may include, for example, value change descriptors describing variations in sensor data values, hierarchical descriptors identifying maximum or minimum values ​​in the sensor data, filtered descriptors including filtered values ​​of the sensor data, event-based descriptors including the timing, frequency, and amplitude of events, the frequency ratio of different types of events, and so on. The generated descriptors are sent to the headend system for topology and phase discovery.

[0018] After receiving descriptors from various assets in the power distribution system, the headend system can perform clustering to group the assets into different segments and phases. Reference assets with known phase IDs and segment IDs can be used to assign specific phase IDs and segment IDs to unknown assets. The determined assignment information can be sent to the individual assets for display or other purposes. Furthermore, the timing information of events contained in the descriptors can be used to synchronize the clocks of various assets in the power distribution system. In some implementations, assets in a network with idle computing and communication resources (e.g., edge processors) can act as proxies for the headend system to perform these operations, thereby reducing communication traffic to the headend system and / or reducing the computational load on the headend system.

[0019] The above operations can be performed when the asset operates in normal mode. In some scenarios, the asset can be configured to operate in fast mode, where segmentation and phase identification can be performed in a shorter time period than in normal mode. In fast mode, assets with unknown segments or phases and their neighboring assets can be configured to generate fast descriptors based on processed sensor data without detecting events or based on consecutively occurring minor events. The asset can send fast descriptors to the headend system or edge processor at a higher rate than in normal mode. Similarly, the headend system can determine segmentation and phase assignments faster than in normal mode by performing correlations on the fast descriptors. Segmentation and phase assignments can be determined by utilizing neighboring assets as reference assets.

[0020] The techniques described in this disclosure improve the efficiency and accuracy of asset segmentation and phase detection in power distribution systems, as well as communication between assets and headend systems. By configuring assets to generate and transmit descriptors instead of raw sensor data, bandwidth consumption can be significantly reduced, as descriptors are typically much smaller than raw sensor data. Furthermore, processing sensor data locally at the asset allows each asset to identify the timing of events and the relative timing of different events. This type of information makes asset grouping more accurate than with raw sensor data. Additionally, utilizing timing information, the headend system can determine the offset of an asset's clock relative to the clock of a high-precision asset, thereby performing high-precision time synchronization or adjustment between assets.

[0021] Exemplary operating environment

[0022] Figure 1 This is a block diagram illustrating a power distribution system 100 according to various aspects of this disclosure. Figure 1 In this system, power generation facility 110 can generate electricity. The generated electricity can be, for example, three-phase alternating current (AC). In a three-phase power system, each of the three conductors carries alternating current with the same frequency and voltage amplitude relative to a common reference, but with a phase difference of one-third of a cycle between each conductor. The electricity can be transmitted to power substation 120 via transmission line 115 at high voltage (e.g., approximately 140 kV to 750 kV).

[0023] At power substation 120, step-down transformer 130 reduces the high-voltage power to a voltage level more suitable for customer use. The stepped-down three-phase power is then transmitted via feeders 140a, 140b, and 140c to distribution transformer 150, which further reduces the voltage (e.g., 120-240V for residential customers). Each distribution transformer 150, 155 can supply single-phase and / or three-phase power to residential and / or commercial customers. Power from distribution transformers 150, 155 is delivered to the user via meter 160. Meter 160 may be supplied by an electricity utility and may be connected between the load (i.e., the customer's residence) and distribution transformers 150, 155. Three-phase transformer 155 may supply three-phase power to the customer's residence, for example, by supplying power to three lines on the street. In some areas, to obtain single-phase power, the customer's residence may be randomly connected to one of these lines. This random connection or tapping causes the utility company to lose track of which customer is on which phase. In addition to the three-phase power, single-phase power can be delivered from different phases of the three-phase power generated by the utility company to individual customers from the distribution transformer 150, resulting in uneven load distribution on the phases.

[0024] Sensor 180 can be distributed across various assets throughout the network, such as, but not limited to, feeder circuits and distribution transformers. Sensor 180 can sense various circuit parameters, such as frequency, voltage, current amplitude, and phase angle, to monitor the operation of the power distribution system 100. It should be understood that... Figure 1 The sensor locations shown are merely illustrative, and the sensors can be placed in other locations, and additional or fewer sensors can also be used.

[0025] As from Figure 1 As can be seen, each asset is connected to one or more phases and one or more segments of the power distribution system 100. The disclosure presented herein can automatically identify the segments and phases of assets in the power distribution system 100 and update information such as the topology of the power distribution system 100 and phase changes over time. The following description uses an electricity meter as an example of an asset. It should be understood that the described techniques are also applicable to other types of assets equipped with sensors, such as transformers, generators, contactors, reversing circuit breakers, fuses, switches, street lighting, ripple receivers, ripple generators, capacitor banks, batteries, synchronous condensers, etc. These assets, including electricity meters, are collectively referred to herein as "utility equipment".

[0026] Figure 2 This is a diagram illustrating a utility management system 200 according to various aspects of this disclosure. The utility management system 200 may include an electrical meter 160 (or simply meter 160), a headend system 210, and a storage depot 220. Although Figure 2For ease of explanation, one electrical meter 160 is shown, but it should be understood that multiple electrical meters 160 may also be included in the utility management system 200.

[0027] Electrical meter 160 can monitor and / or record various characteristics associated with the power distribution system, such as voltage, current, energy usage at customer residence 230, source impedance, and load impedance, and transmit this information to head-end system 210. For example, electrical meter 160 can continuously monitor and record voltage changes at customer residence 230, as well as the days of the week and times of day associated with these voltage changes, and transmit this information to head-end system 210. Furthermore, electrical meter 160 can act as a sensor to detect and / or record abnormal measurement results and / or events. It should be understood that other information can be monitored and transmitted via electrical meter 160, such as, but not limited to, average power consumption, peak power, etc.

[0028] Electrical meter 160 can communicate with headend system 210 via a wired or wireless communication interface using a communication protocol suitable for a specific communication interface. Different wired or wireless communication interfaces and associated communication protocols can be implemented on electrical meter 160 for communication with headend system 210. For example, in some cases, a wired communication interface can be implemented, while in others, a wireless communication interface can be implemented for communication between electrical meter 160 and headend system 210. In some examples, a wireless mesh network can be connected to electrical meter 160. Electrical meter 160 can send data to a collector (not shown) communicating with another network to send data to headend system 210. The collector can also act as an edge processor to process data (e.g., descriptors) sent from electrical meter 160, particularly in the fast mode described below, to offload some processing load from the headend system. Electrical meter 160 can communicate using radio frequency (RF), cellular, or power line communication. It should be understood that other communication methods can also be used.

[0029] The communication network connecting the electrical meter 160 and the head-end system 210 as described above can be used to transmit data for topology and phase identification (e.g., descriptors discussed below). The communication network can also be used to provide other types of communication from the electrical meter 160 to the head-end system 210, such as those used for reporting power consumption. In some examples, the communication network connecting the electrical meter 160 and the head-end system 210 as described above can cover… Figure 1 The diagram shows a power distribution network. For wireless communication, adjacent meters in the communication network can be different from adjacent meters in the distribution network.

[0030] The headend system 210 can also communicate with the storage library 220 via a wired or wireless communication interface. The storage library 220 can be implemented using, for example, but not limited to, one or more hard disk drives, solid-state storage devices, or other computer-readable storage media. Other storage configurations may also be used. The storage library 220 can be configured to store various data relating to the electricity meters 160 and other assets in the power distribution system 100. For example, the storage library 220 may include descriptors 226 transmitted for each asset, segments and phases 228 defined for each asset, etc.

[0031] The repository 220 may also include reference asset data 222 describing information associated with reference assets having known segmentation information or phase information, or both. In some examples, segmentation and phase information is represented using segment identifiers (IDs) and phase identifiers (IDs) that uniquely represent the individual segments and phases. Any asset sharing the same phase or segment ID but having unknown phases or network segments may have its phase or network segments inferred by the headend system 210. The known phases and segments of a reference asset may be “out-of-band information” set by processes other than the automated network topology discovery described herein. In some examples, most assets may have known phases and / or segments due to the loading of a large amount of out-of-band information. The processes proposed herein can be used to verify this data and flag any inconsistencies.

[0032] Various other information related to assets in the power distribution system 100 can be included in the storage 220, such as parameters generated for segments or assets. For example, a segment object can be created when the headend system 210 determines that a unique segment may exist. Segment parameters such as phase ID, segment ID, and segment type ID can be created and associated with segments.

[0033] Now for reference Figure 3 , Figure 3 An example is shown of a process, according to certain aspects of this disclosure, for generating a descriptor 226 by a meter 160 of a power distribution system 100 to facilitate topology and phase discovery of a head-end system 210. The meter 160 can be implemented by executing appropriate programming code. Figure 3 The operation is shown in the figures. For illustrative purposes, process 300 is described with reference to certain examples depicted in the accompanying drawings. However, other implementations are also possible.

[0034] At box 302, process 300 involves acquiring sensor data at meter 160 or other assets and preprocessing the sensor data. Depending on the type of asset and the sensors installed on the asset, the sensor data may include, but is not limited to, power supply voltage, current, power, source impedance, load impedance, temperature, light level, humidity, pressure, sound, vibration, and other distributed network or environmental quantities. Preprocessing the sensor data may include filtering and normalizing the sensor data to remove or reduce noise, offset, etc., to generate processed sensor data. Preprocessing may also include detecting events such as disturbances or anomalies in the sensor data and determining the characteristics of the events, such as the duration of the events, the number of events, the time of the events, and the severity of the events. For meter 160, the sensor data may include power voltage, current, power usage, etc. Additional details regarding acquiring and preprocessing sensor data are referenced below. Figure 4 Presented.

[0035] At box 304, process 300 involves determining whether meter 160 is in fast mode. When meter 160 is in fast mode, meter 160 can generate descriptors (also called "fast descriptors") at a higher rate than in normal mode and send them to headend system 210, so that segments or phases of meter 160 can be identified by headend system 210 within a short period of time. If meter 160 is not in fast mode (i.e., meter 160 operates in normal mode), meter 160 can generate descriptors 226 (also called "normal descriptors") at a normal rate.

[0036] If it is determined that meter 160 is in normal mode, process 300 involves generating a set of normal descriptors 226 based on the processed sensor data. Meter 160 can be configured to generate a set of normal descriptors 226 at descriptor intervals (such as 30 minutes, one hour, or two hours). Normal descriptors 226 can be generated to include filtered or normalized sensor data, detected events, and characteristics associated with those events. In some examples, descriptors 226 may include vector or scalar values ​​describing the sensor data and detected events.

[0037] The meter 160 can be configured to generate and send different types of descriptors, including but not limited to raw descriptors, raw statistical descriptors, normalized change descriptors, hierarchical descriptors, filtered descriptors, disturbance characterization descriptors, distribution characterization descriptors, signal frequency analysis descriptors, event timing analysis descriptors, event frequency analysis descriptors, and event amplitude analysis descriptors.

[0038] For example, raw descriptors may include descriptors with fewer preprocessing requirements. Due to high bandwidth consumption and limited contribution to the packetization process performed at headend system 210, raw descriptors are typically not sent to headend system 210 in a conventional manner. In some examples, sensor data in the raw descriptor is not normalized. Instead, the sensor data values ​​may be raw values ​​and may include averages within the descriptor interval. The raw descriptor may include the average root mean square (RMS) voltage during that interval, the average voltage imbalance during that interval, the average supply impedance, the average temperature, the average optical level, the average load impedance, the average output power, the average vibration, the instantaneous communication channel address or name, and so on.

[0039] Raw statistical descriptors can provide standard statistics of sensor data, such as maximum, minimum, median, range, standard deviation, and variance. Raw statistical descriptors can be determined based on the raw values ​​of the sensor data rather than normalized values. During normal mode, raw statistical descriptors are typically not sent to the headend system 210.

[0040] A normalized variation descriptor can represent a change in a fundamental quantity. In this type of descriptor, the value of the sensor data is normalized or scaled using the average of the number of sensor data points from previous descriptor intervals. The normalized variation descriptor may include the normalized maximum RMS positive voltage change of the metering window associated with meter 160, the normalized maximum RMS positive voltage change of half a cycle of meter 160, the normalized maximum RMS positive voltage change of a single sample, the normalized minimum RMS negative voltage change of the metering window, and the change in power supply frequency. Typically, the normalized maximum value is sent to headend system 210 in normal mode.

[0041] The hierarchical descriptor may include values ​​such as the Nth highest maximum or Nth lowest minimum value during an interval period. A single Nth maximum or minimum value, or a sequence of N highest or lowest values, may be included in descriptor 226 and may be sent to headend system 210. Examples of hierarchical descriptors include, but are not limited to, the normalized Nth highest maximum RMS voltage change of a measurement window, the normalized Nth highest maximum RMS voltage change of a half-cycle, and the normalized Nth highest maximum voltage change of a single sample.

[0042] Filter descriptors can include values ​​generated by applying digital filters to sensor data. Examples of filter descriptors include, but are not limited to, the normalized maximum RMS positive voltage change of a single sample filtered to exclude pulses, the normalized maximum RMS positive voltage change of a single sample filtered to exclude the fundamental power supply frequency, the normalized Nth highest maximum RMS voltage change of a metering window filtered over a period of time (e.g., one minute), and the normalized Nth highest maximum RMS voltage change of a metering window filtered over a period of time (e.g., five minutes). Filters can also include finite impulse response (FIR) and infinite impulse response (IIR) digital filters to form bandpass, low-pass, high-pass, or notch filters, etc. These filters can operate on metric outputs that are raw sampled at high rates up to 1000 Hz or processed at low rates up to 1 Hz.

[0043] Event characterization descriptors include descriptors that characterize the properties of events such as disturbances or anomalies. For example, the maximum disturbance in a descriptor interval can be analyzed and included as a value of the event characterization descriptor. Examples of event characterization descriptors may include, but are not limited to, the duration of a ringing disturbance, the frequency of a ringing disturbance, the normalized peak amplitude of a ringing disturbance, the normalized amplitude of a pulse, the normalized amplitude of a voltage step change, the phase of a voltage step change, the duration of a short-term voltage drift, the normalized voltage change of a short-term voltage drift, etc.

[0044] A distribution descriptor can characterize a signal in sensor data as a distribution of samples to N distinct classes. Normalization can be performed such that the number of distinct values ​​is in the range of, for example, 5 to 10. The distribution descriptor can be the number of samples in a particular class or a statistical property of the distribution, such as the median or mean. Examples of distribution descriptors can include the total number of samples residing in one of N distinct classes of normalized voltage amplitude, the total number of samples residing in one of M distinct classes of normalized voltage amplitude variation, or the average duration of the persistence of voltage amplitude residing in one of N distinct classes of voltage amplitude.

[0045] Signal frequency analysis descriptors can be used to describe the frequency components in sensor data. For example, a Fast Fourier Transform (FFT) can be used to determine the frequency components of sensor data. Examples of signal frequency analysis descriptors may include, but are not limited to, the normalized amplitude and frequency of the second maximum frequency detected within a metering window during an interval, the frequency of the maximum frequency detected in the longest operating ringing disturbance, the frequency of the maximum frequency detected in an RMS voltage sample during an interval, the frequency and amplitude of the longest operating disturbance with an amplitude exceeding p% (e.g., p=2) of the rated voltage, the total harmonic distortion (THD) of the sensor data, the variation in THD, the amplitude variation of the most significant harmonic, etc.

[0046] Event timing analysis descriptors, event frequency analysis descriptors, and event amplitude analysis descriptors characterize different categories of events. Examples of these types of descriptors include the time of an event, the time it takes for the output of one filter to become higher than or equal to the output of another filter (e.g., a one-minute moving average crossing a five-minute moving average of the RMS voltage), the duration between two events, the average duration between multiple events of the same type, the minimum / maximum / median / variance of the amplitude / duration / time / frequency of multiple events of the same type, the number of times a threshold has been breached, the number of times a value falls within the valid range, and the amplitude of other filter quantities when a threshold is breached. Other examples may include the normalized amplitude of the N highest maximum differences in RMS values, the time interval between adjacent occurrences, the normalized amplitude of the N highest maximum differences in RMS values, and the cumulative sum of the voltage amplitudes of all breach events. It should be noted that the above times can be relative to an absolute timescale or relative to the occurrence time of some other event. Absolute timing may require a precise timer, while timing relative to some other event does not require a precise timer.

[0047] These types of descriptors are not mutually exclusive, and descriptors can be classified into more than one type. Additionally, meter 160 can be configured to generate and transmit a range of descriptors at each interval based on the descriptor configuration. For example, for assets with multiphase power supply and current measurement capabilities, the descriptor configuration may include the average RMS voltage during the interval, the average voltage imbalance during the interval, the average supply impedance, the time period between the M longest intervals where the one-minute and five-minute moving averages do not cross, a count of the number of one-minute and five-minute crosses, a count of the number of voltage changes exceeding 10% of the average RMS voltage of the previous interval, the time between a zero cross and the start of a maximum synchronization event, the number of ripple telegraph pulses, and the average duration of voltage amplitude residing within one of 10 different voltage amplitude categories.

[0048] At box 308, procedure 300 involves appending the generated regular descriptors to a regular descriptor block. In some examples, regular descriptors can be organized into regular descriptor blocks of a fixed size. Newly generated descriptors can be appended to the current descriptor block until the descriptor block is full. The descriptor block full of regular descriptors can then be appended to a descriptor profile associated with meter 160. The descriptor profile can be configured to maintain regular descriptor blocks to be sent to headend system 210 and the most recently sent regular descriptor blocks to headend system 210. In other implementations, regular descriptors can be maintained in the descriptor profile without involving descriptor blocks.

[0049] At box 310, process 300 involves determining whether the generated regular descriptors are ready to be sent to headend system 210. In some examples, a regular descriptor is ready to be sent to headend system 210 when there are more than K descriptor blocks in the descriptor profile and K is a natural number. In other examples, a regular descriptor is ready to be sent to headend system 210 if regular descriptors for L descriptor intervals have been generated. For example, meter 160 can be configured to send regular descriptors to headend system 210 at a time, every 4 descriptor intervals. Other criteria can be used to determine whether regular descriptors can be sent to headend system 210. At box 312, process 300 involves sending regular descriptors, such as the oldest N descriptor blocks in the descriptor profile or descriptors from the past L descriptor intervals, to headend system 210. At box 314, process 300 involves deleting the oldest descriptors in the descriptor profile, such as the descriptors sent in box 312, to free up storage space for new descriptors. Process 300 then returns to box 302 to collect more sensor data for generating a new descriptor.

[0050] If it is determined at block 304 that meter 160 is in fast mode, process 300 involves generating a fast descriptor at block 320. In some cases, meter 160 can be manually configured to operate in fast mode. For example, a technician may be installing a new meter or repairing an existing meter at a customer's premises 230 and needs to quickly determine the segment and phase identifiers of the new or existing meter. To obtain such information, the technician can configure a meter with unknown or unverified segment and phase identifiers (also referred to as an "unknown meter") to operate in fast mode via user input. Additionally, the technician can configure neighboring meters of an unknown meter with known segment or phase information to operate in fast mode. These neighboring meters can be used as reference meters during the phase and segment identification process. In other examples, headend system 210 may, in response to a technician's request or other actions initiated at the headend, send instructions to the unknown meter and its neighboring meters to operate in fast mode in order to determine the segment and phase of the unknown meter within a short timeframe.

[0051] To accelerate the process, fast descriptors can be generated and sent to headend system 210 at a higher rate than regular descriptors, such as twice per minute. This allows headend system 210 to determine the phase and segment of an unknown meter and provide that information back to the meter and / or technicians in the field almost in real time. Fast descriptors can include data or descriptors requiring fewer computational resources, such as raw descriptors or raw statistical descriptors, where time-consuming processing such as FFT is not involved. Furthermore, the meter does not need to wait for simultaneous power outages when generating fast descriptors. However, due to the potential lack of time synchronization between meters, the headend system may need to perform additional processing in a time-dependent manner from fast descriptors from multiple meters.

[0052] At box 322, process 300 involves appending fast descriptors to a fast descriptor block or other type of data structure. At box 324, process 300 involves determining whether enough fast descriptors have been generated, for example, whether a predetermined number of fast descriptors have been generated, whether the fast descriptor block is full, or whether fast descriptors for the fast mode interval have been generated. If not, process 300 involves obtaining more sensor data at box 302 for generating additional fast descriptors. If enough fast descriptors have been generated, process 300 involves sending the generated fast descriptors to headend system 210 at box 326. In some examples, the fast mode interval is much shorter than the descriptor interval in a regular mode. At box 328, process 300 involves deleting fast descriptors to free up space for new descriptors. This process also involves obtaining additional sensor data at box 302.

[0053] Now for reference Figure 4 , Figure 4 An example of a process for processing sensor data at a meter 160 in an electricity distribution system 100, according to certain aspects of this disclosure, is shown. The meter 160 can be implemented by executing appropriate programming code. Figure 4 The operation is shown in the figures. For illustrative purposes, process 400 is described with reference to certain examples depicted in the accompanying drawings. However, other implementations are also possible.

[0054] At block 402, process 400 involves obtaining reading samples from a sensor associated with meter 160 and configured to measure and generate sensor data. At block 404, process 400 involves performing data processing on current and past samples. Examples of preprocessing include, but are not limited to, statistical processing, mathematical operations, and digital signal processing (DSP). Statistical processing may include determining, for example, instantaneous values, maximum values, minimum values, average values, median values, ranges of values, standard deviations, a tiered list of N maximum / minimum values, and sample counts. Mathematical operations may include RMS operations, dedifferentiation, and / or integration. DSP operations may include FFT, finite impulse response filtering (FIR), infinite impulse response filtering (IIR), wavelet transform, normalization, and / or resampling.

[0055] At box 406, process 400 involves normalizing the processed value and identifying events, such as disturbances or anomalies. For example, normalization can be performed by dividing the processed value by the average voltage amplitude from previous descriptor intervals. Event identification can be performed by comparing the processed value to one or more thresholds or by matching the value to a reference value. Thresholds can be fixed static values, configurable thresholds controlled, for example, by head-end system 210, or dynamic variables that are a function of the sensor data itself (e.g., instantaneous sensor values ​​can be compared to short-term averages of sensor values). An event can be identified if the processed value is above a corresponding threshold (or below a threshold, depending on the type of threshold), outside the valid range, or matches a reference value. For example, a voltage anomaly event can be identified if the average voltage value is above a high voltage threshold, below a low voltage threshold, or outside the valid range. Multiple events can be identified similarly.

[0056] At box 408, process 400 involves determining characteristics or attributes of the identified event. For example, the characteristics of an event may include the time when the event occurred, the severity of the disturbance in the event (e.g., measured by the amount by which a processed value exceeds a threshold or deviates from a valid range), the number of events in an event category, the time intervals between various events, etc. These characteristics of the event may be associated with the corresponding event and used by meter 160 to generate a descriptor, as referenced above. Figure 3 As stated above.

[0057] At box 410, process 400 involves adjusting the normalization factor and reference time. For example, normalization can be performed based on the mean and standard deviation of the sensor data. As more sensor data is collected, the mean can be updated to include the new sensor data. The standard deviation can be updated similarly. Additionally, if head-end system 210 instructs meter 160 to adjust the reference time, meter 160 can further perform reference time adjustment during preprocessing.

[0058] In some examples, meter 160 may further perform time-series analysis related to events during preprocessing. Time-series analysis may include recording the time of the event and / or the time elapsed since the previous event. Determining and reporting the duration between events, rather than the time of the events, may be beneficial because the duration between events does not require precise time synchronization between meters 160. In other words, the relative time between events is insensitive to the timing inaccuracies of the meters themselves and the headend system 210.

[0059] Preprocessing sensor data can reduce the size of the data sent to the headend system 210. In some systems, the raw sensor data rate is approximately 13 KB / h or 32 Mbps, while the descriptor data can have a lower rate of approximately 50 KB / h. Therefore, preprocessing can provide the benefit of allowing high-bandwidth data to be processed without having to send it over the communication channel to the headend system 210, thereby reducing network bandwidth consumption.

[0060] Now for reference Figure 5 , Figure 5 An example of a process 500 for identifying the topology and phases of a power distribution system 100, according to certain aspects of this disclosure, is shown. The head-end system 210 can implement this by executing appropriate program code in the meter 160. Figure 5 The operations depicted herein are described with reference to certain examples depicted in the accompanying drawings for illustrative purposes. However, other implementations are also possible.

[0061] At block 502, process 500 involves receiving descriptors from meters 160 in the power distribution system 100. Depending on the operating mode of the meters 160 in the power distribution system 100, descriptors can be received from a small set of meters 160 configured to operate in fast mode or from meters 160 in the power distribution system 100 configured to operate in normal mode. At block 504, process 500 involves determining whether the received descriptor is a fast descriptor. In some examples, a fast descriptor may include a flag indicating that the descriptor was generated by the meter 160 when operating in fast mode. Head-end system 210 can determine whether the received descriptor is a fast descriptor by detecting this flag. In other examples, head-end system 210 may determine a descriptor as a fast descriptor based on the size and frequency of the received descriptor. For example, fast descriptors may be transmitted using smaller packets and at a higher frequency.

[0062] If the headend system 210 determines that the descriptor is a fast descriptor, then procedure 500 involves performing fast mode processing. Details regarding fast mode processing are provided below. Figure 6Provided. If headend system 210 determines that the descriptor is not a fast descriptor, process 500 involves performing regular mode processing starting from block 506. At block 506, headend system 210 may store the received descriptors for the associated meter 160. At block 508, headend system 210 may determine whether a sufficient number of descriptors have been received (e.g., the number of received descriptors exceeds a threshold number of descriptors). In some implementations, headend system 210 is configured to identify segments and phases of the meter on an interval-by-interval basis. In other words, headend system 210 identifies segments and phases based on descriptors collected over the entire detection interval. The detection interval may be set to an integer number of descriptor intervals (e.g., 4 hours or 8 hours for a 1-hour descriptor interval).

[0063] If the number of received descriptors is insufficient, process 500 involves receiving more descriptors at block 502. If a sufficient number of descriptors have been received, process 500 involves grouping the meters 160 at block 510 based on the received descriptors (e.g., descriptors received for the current detection interval). Grouping can be performed using, for example, K-means clustering, PAM (Peripheral K-class clustering), hierarchical clustering, fuzzy clustering, model-based clustering, density-based clustering, hybrid clustering, etc. Headend system 210 can assess clustering trends, determine the number of clusters, and evaluate clustering quality. Grouping or clustering results can be used to automatically customize the clustering method for grouping descriptors.

[0064] At box 512, process 500 involves comparing the current group with past groups to determine the confidence level of the group. For example, if the current group is consistent with past groups (e.g., the same meters are continuously grouped into the same group), the quality of the group can be determined to be high, and a high confidence level can be assigned to the group; otherwise, a low confidence level is assigned.

[0065] In some examples, confidence levels can be determined based on grouping using different types of descriptors. As described above, meter 160 can generate and send multiple descriptors of different types based on descriptor configuration. For each descriptor type, headend system 210 can perform clustering to assign meters into multiple groups. Groups of one type of descriptor may overlap with groups of another type forming a union. This union may contain a group of meters connected to the same segment. However, this union may also contain meters that do not belong to the same segment or omit meters that should be included. This problem can be mitigated by determining and comparing unions over multiple detection intervals. If, for most cases across multiple detection intervals, the meters exist in the same union, they can be declared to belong to the segment associated with that union. A high confidence level can be assigned to such grouping. Otherwise, a low confidence level is assigned.

[0066] At box 514, process 500 involves comparing a confidence level to a confidence threshold. If the confidence level exceeds the threshold, the meter can be determined to belong to a defined group, and the segment ID and phase ID of a reference meter in that group can be assigned to the meters in that group, as shown in box 518. If the confidence level does not exceed the threshold, then at box 516, head-end system 210 can select a new descriptor configuration and send it to meter 160, allowing meter 160 to generate a different set of descriptors to increase the confidence level of the group. For example, if the existing descriptor configuration without a signal frequency descriptor cannot provide a confidence level exceeding the threshold, head-end system 210 can select a new descriptor configuration to include one or more signal frequency descriptors. If disturbances with certain frequencies have been observed in power distribution system 100, head-end system 210 can select this new descriptor configuration. The descriptor configuration can be selected differently for meters 160 at different locations in power distribution system 100. In this way, head-end system 210 can customize the descriptors sent by meter 160 to match the types of grid disturbances and parameters prevalent at a given time and / or at a given point in the power distribution system 100. When descriptors are selected correctly, a high signal-to-noise ratio of the descriptor values ​​will exist, and clustering quality can be improved. Process 500 can then proceed to block 502 to receive additional descriptors. In some examples, block 516 can be performed after multiple rounds of operation and determining that the confidence level of the grouping is below a threshold. At this point, head-end system 210 can be configured to generate a new descriptor configuration and send it to meter 160. In some examples, if after several iterations, none of the neighboring meters has a phase ID and segment ID, a new phase ID and segment ID can be created and assigned to the meter. Similarly, if no neighboring meters are found, a new phase ID and segment ID can be created and assigned to the meter.

[0067] In some implementations, after the meters have been identified and classified, the head-end system 210 can instruct the meters 160 to disable or reduce descriptor transmission activity to reduce communication network bandwidth consumption. The head-end system 210 can also choose to do this for unclassified meters, allowing it to focus its computational work on specific areas within the power distribution system 100. In other implementations, the head-end system 210 can temporarily disable descriptor generation and transmission behavior in all or some meters, allowing other meter functions to prioritize or avoid collecting data known to be erroneous at times.

[0068] Now for reference Figure 6 , Figure 6 An example of a fast-mode process 600 for identifying the topology and phases of a power distribution system 100, according to certain aspects of this disclosure, is shown. The head-end system 210 can implement this by executing suitable programming code. Figure 6The operations depicted herein are described with reference to certain examples depicted in the accompanying drawings for illustrative purposes. However, other implementations are possible.

[0069] At block 602, process 600 involves storing the received fast descriptors of the associated meter. At block 604, process 600 involves determining whether enough fast descriptors have been received. For example, headend system 210 can be configured to identify the segment and phase of the meter for every Y fast interval, where Y is a positive integer. The fast interval is greater than the one referenced above. Figure 5 The detection interval for the regular mode discussed in box 508 is much shorter. For example, the fast interval can be set to one minute, 30 seconds, or even shorter. If no fast descriptors from the past Y fast intervals are received, process 600 proceeds to box 614 to collect more fast descriptors (which can be...). Figure 5 (See box 502). If fast descriptors for the past Y fast intervals have been received, process 600 involves grouping the meters in box 606 by performing correlation analysis between the fast descriptors of the meters operating in fast mode. Because the intervals for collecting sensor data from the meters are short, fast descriptors may not contain events such as disturbances and anomalies as in regular descriptors. Instead, fast descriptors primarily contain small variations in sensor data values, such as voltage changes or total harmonic distortion (THD), temperature, source impedance, current, etc. For this type of data, correlation-based grouping can provide more accurate grouping results.

[0070] At block 608, process 600 involves assigning a segment ID and / or phase ID to an unknown meter. For example, the segment ID and / or phase ID of an adjacent meter in the same group as the unknown meter can be assigned to the unknown meter. The assigned segment ID and / or phase ID can be stored in asset data 224 of storage 220. At block 610, process 600 involves sending the assigned information to the unknown meter. The unknown meter can display the assigned information on a display device. In an example where a field technician configures the meter in fast mode, the technician can read the assigned information and proceed with the installation or repair process accordingly. Alternatively or additionally, the meter's assignment information can be obtained from head-end system 210 when needed. At block 612, process 600 involves sending a command to the meter to exit fast mode. The meter can then be accessed as described above. Figure 3 and 4 The descriptor is generated and sent in the normal mode. In some examples, the meter can exit fast mode by receiving a command via user input at the corresponding meter location.

[0071] It should be understood that although the above disclosure pertains to electricity meters, the operations performed by electricity meters to generate descriptors can also be performed by other types of assets to facilitate phase and segment operations. Furthermore, if no reference asset information exists for assigning phase IDs and segment IDs, the headend system 210 can generate or create different phase IDs and / or segment IDs for different asset groups, allowing different asset groups to be distinguished from each other.

[0072] It should also be understood that while the above disclosure pertains to a single headend system 210 receiving all descriptors, other arrangements are possible. For example, the power distribution system 100 may be split into separate zones, each with its own dedicated headend system. This zone headend system may be independent of any other headend system, or it may coordinate its activities with other headend systems. In a multi-headend environment, a hierarchical arrangement may exist, where a single top-level headend system coordinates the activities of all other zone headend systems. Each zone headend system may undertake the task of processing large amounts of descriptor data and forward limited descriptor data to that top-level headend system. This limits the data processing requirements of any single headend system and reduces bandwidth requirements. A zone headend system may be optionally activated to process fast descriptor data from nearby assets. In some examples, edge processors may be used to process descriptors in individual zones and forward processed descriptor data to the top-level headend system.

[0073] Furthermore, as described above, the descriptors sent to the headend system 210 are not permanently stored at the meter 160. In some examples, only data used to make current determinations about which segments are associated with which assets is stored. This data is continuously refreshed, and old data is continuously deleted. For example, descriptors from the last N intervals are stored (e.g., N ranges from 10 to 200). Additionally, descriptors for assets from intervals around different critical event classes can be stored. Critical events can include rare disturbances on the power distribution system detected by a large number of assets. For each critical event, the intervals before, during, and after the event can be stored, and the intervals of the last M critical events for each category can be stored (e.g., M=10).

[0074] As described above, the meter generates descriptors at several time intervals, and the headend system 210 performs identification based on these descriptors. This requires time synchronization across all assets that send descriptors to the headend system 210. In some power distribution systems 100, time errors can be as high as + / - 2 minutes. Since the descriptors have different values, or otherwise these values ​​would be the same, timing inaccuracies will negatively impact grouping activities performed in the headend system 210. To mitigate this problem, the time intervals can be chosen to be relatively large relative to the amount of time error. For example, a one-hour interval can be chosen to overcome a 2-minute timing error.

[0075] However, this mechanism may not work for fast mode, where a small number of physically adjacent assets send short-interval data over a short period (e.g., approximately 10 seconds). For these cases, the headend system 210 can perform correlation operations across multiple time intervals on a finite number of descriptors to determine the time offset of each asset. Asset sensors will experience similar electrical noise; therefore, those patterns can be used as correlation-detectable synchronization events. Once the time offset of each asset is known, the descriptor grouping process described above can be performed.

[0076] The disclosure presented herein can also be used to clock assets in a power distribution system 100. Depending on the asset's hardware capabilities, sub-cycle preprocessing (e.g., 20 ms @ 50 Hz) is possible in high-performance assets. Therefore, these assets can detect short-duration disturbances, such as high-frequency ringing of a few milliseconds or large voltage changes. Although propagation delays may exist due to transmission lines, transformers, and capacitor banks, it can be assumed that assets in a small area have the same detection time for these types of events. Therefore, such events can be used to synchronize clocks across multiple assets that have already detected the event with an accuracy close to the sensor sampling period (e.g., + / - 2 ms). For example, these synchronization events can be included in descriptors so that the head-end system 210 can identify these synchronization events and use the occurrence time of the synchronization events as a reference time for all other events in the descriptors received by the head-end system 210 to adjust the timing of those events so that all events are synchronized. The head-end system 210 can also calculate the time offset of each asset from the reference clock and send time correction instructions to these assets to correct their respective clocks. Alternatively or additionally, the headend system 210 may also send a reference time of the synchronization event to the asset. The asset can determine the difference between the reference time and the corresponding time when it observes the synchronization event. The asset can then adjust its local clock based on this difference.

[0077] Longer-duration synchronization events can exist in the form of high-frequency tones or pulses, which are conducted disturbances superimposed on the supply voltage. Tones have a longer disturbance duration (e.g., 50ms-5000ms) than pulses (e.g., shorter than 50ms). Pulses may suffer more attenuation compared to tones due to impedance and filtering delay. These can be unintentional events from sources such as grid-connected inverters, variable speed drives, and switch-mode power supplies, or intentional events such as injected ripple telegraphs used for load control. A series of pulses or tones can last only a few milliseconds, 10 seconds (in the case of ripple telegraphs), or even continuously (in the case of equipment failure). These signals can last for several power cycles. Detection of these events can take a long time due to filtering delays within the meter itself, but synchronization is still possible because the delay is a known constant. Different areas of the network may experience long-duration disturbances at different frequencies. If searching all frequencies is chosen due to processing limitations rather than as an option, it may be necessary to configure assets to search for specific frequencies. The headend system 210 can automatically configure assets to focus on specific frequencies based on disturbances present in the area.

[0078] Assets whose timing is synchronized using power grid events with higher precision, such as better than + / - 3.3 ms, can be used to determine their phases. Many types of power grid events occur simultaneously on all phases. A positive zero-crossing of the voltage signal occurs every 20 ms, and each phase is delayed by 6.666 ms. The phase can be directly determined by calculating the time interval between the positive zero-crossing and the power grid synchronization event, since each phase A, B, or C will have a time interval that is a multiple of 6.666 ms. This time interval can be represented by a descriptor and sent to the headend system 210, where it can be determined by comparing the time interval of assets with unknown phases with the time interval of assets with known phases.

[0079] Synchronization events in the power distribution system 100 are useful for identifying segments and phases. Some categories of synchronization events can propagate across a large number of assets and segments. The reception time can serve as a reference point for the local event occurrence time. Assets on the same segment will have similar time periods between the synchronization event time and the local disturbance time. This time period can be sent to the headend system 210 in the form of descriptors, which can be grouped with other similar descriptors to identify the asset's segment and phase. Furthermore, synchronization events can trigger the sending of time adjustment messages from the headend system 210 to the assets. This can occur when some assets contain high-precision clocks. When a high-precision asset reports a synchronization event to the headend system 210, the reporting message can also include when the event occurred relative to its precise clock. In this way, the headend system 210 can know the precise time the event occurred. Using this information, the headend system 210 can notify other assets that have also detected the event of the precise time when the event occurred, thereby causing the other assets to adjust their clocks accordingly.

[0080] Example electricity meter

[0081] Figure 7 An exemplary electricity meter 700, such as electricity meter 160, that can be used to implement the sensor data collection and descriptor generation described herein is illustrated. The electricity meter 700 includes a communication module 716 and a metering module 718 connected via a local or serial connection 730. These two modules can be housed in the same unit on separate boards, thus the local connection 730 can be an onboard socket. Alternatively, these two modules can be housed in the same unit having a single processor and memory block performing multiple tasks including sensing / metering and communication coordination. Alternatively, the modules can be housed separately, and therefore the local connection 730 can be a communication cable such as a USB cable or other conductor.

[0082] The communication module 716 functions to send descriptor 226 and other data to other meters or head-end system 210, and to receive data from head-end system 210 or other meters. The metering module 718 functions to manage resources, particularly those necessary for accessing and measuring the resources used. The communication module 716 may include a communication device 712, such as an antenna and a radio. Alternatively, the communication device 712 may be any device that allows wireless or wired communication. The communication module 716 may also include a processor 713 and a memory 714. The processor 713 controls the functions performed by the communication module 716. The memory 714 may be used to store data used by the processor 713 to perform its functions. The memory 714 may also temporarily store other data for the meter 700, such as descriptor 226.

[0083] The measurement module 718 may include a processor 721, a memory 722, and a measurement circuit 723. The measurement circuit 723 processes resource measurements and can be used as a sensor to collect sensor data. The processor 721 in the measurement module 718 controls the functions performed by the measurement module 718. For example, the processor 721 is configured to calculate a descriptor 226 based on sensor data obtained by the measurement circuit 723, as described above. The memory 722 stores data required by the processor 721 to perform its functions. The memory 722 also stores the descriptor 226 calculated by the processor 721. The communication module 716 and the measurement module 718 communicate with each other via a local connection 730 to provide data required by the other module. Both the communication module 716 and the measurement module 718 may include computer-executable instructions stored in memory or another type of computer-readable medium, and one or more processors within the module may execute the instructions to provide the functions described herein.

[0084] Examples of headend systems used to implement certain embodiments

[0085] Any suitable computing system or group of computing systems can be used to perform the operations described herein. For example, Figure 8 An example of a computing system 800 is depicted. An implementation of the computing system 800 can be used in a headend system 210.

[0086] The depicted example of computing system 800 includes a processor 802 communicatively coupled to one or more memory devices 804. Processor 802 executes computer-executable program code stored in memory device 804, accesses information stored in memory device 804, or both. Examples of processor 802 include microprocessors, application-specific integrated circuits (“ASICs”), field-programmable gate arrays (“FPGAs”), or any other suitable processing device. Processor 802 may include any number of processing devices, including a single processing device.

[0087] Memory device 804 includes any suitable non-transitory computer-readable medium for storing program code 805, program data 807, or both. Computer-readable media can include any electronic, optical, magnetic, or other storage device capable of providing computer-readable instructions or other program code to a processor. Non-limiting examples of computer-readable media include disks, memory chips, ROM, RAM, ASICs, optical storage, magnetic tape or other magnetic storage, or any other medium from which instructions can be read by a processing device. Instructions can include processor-specific instructions generated by a compiler or interpreter from code written in any suitable computer programming language, such as C, C++, C#, Visual Basic, Java, Python, Perl, JavaScript, and ActionScript.

[0088] The computing system 800 executes program code 805, which configures the processor 802 to perform one or more operations described herein. In various embodiments, examples of program code 805 include program code for identifying segments and phases of an asset based on conventional and fast descriptors, or other suitable applications for performing one or more operations described herein. The program code may reside in memory device 804 or any suitable computer-readable medium and may be executed by processor 802 or any other suitable processor.

[0089] In some embodiments, one or more memory devices 804 store program data 807 including one or more datasets described herein. Examples of such datasets include past security tokens, global token tables, etc. In some embodiments, one or more of the datasets, models, and functions are stored in the same memory device (e.g., one of the memory devices 804). In additional or alternative embodiments, one or more of the programs, datasets, models, and functions described herein are stored in different memory devices 804 accessible via a data network. The computing system 800 also includes one or more buses 806. Buses 806 communicatively couple one or more components of a corresponding one of the computing systems 800.

[0090] In some embodiments, the computing system 800 further includes a network interface device 810. The network interface device 810 includes any device or group of devices adapted to establish a wired or wireless data connection to one or more data networks. Non-limiting examples of the network interface device 810 include Ethernet adapters, modems, etc. The computing system 800 is able to communicate with one or more other computing devices via a data network using the network interface device 810.

[0091] The computing system 800 may also include multiple external or internal devices, input devices 820, presentation devices 818, or other input or output devices. For example, the computing system 800 is shown having one or more input / output (“I / O”) interfaces 808. The I / O interface 808 can receive input from an input device or provide output to an output device. Input devices 820 may include any device or group of devices adapted to receive visual, auditory, or other suitable input that controls or influences the operation of the processor 802. Non-limiting examples of input devices 820 include touchscreens, mice, keyboards, microphones, stand-alone mobile computing devices, etc. Presentation devices 818 may include any device or group of devices adapted to provide visual, auditory, or other suitable sensory output. Non-limiting examples of presentation devices 818 include touchscreens, monitors, speakers, stand-alone mobile computing devices, etc.

[0092] although Figure 8 Input device 820 and presentation device 818 are depicted as being local to the computing device performing head-end system 210, but other implementations are possible. For example, in some embodiments, one or more of input device 820 and presentation device 818 may include a remote client computing device that communicates with computing system 800 via network interface device 810 using one or more data networks described herein.

[0093] General considerations

[0094] Numerous specific details are set forth herein to provide a thorough understanding of the claimed subject matter. However, those skilled in the art will understand that the claimed subject matter can be practiced without these specific details. In other instances, methods, apparatus, or systems that a person skilled in the art should understand have not been described in detail to avoid obscuring the claimed subject matter.

[0095] The features discussed herein are not limited to any particular hardware architecture or configuration. A computing device can include any suitable arrangement of components that provide results modulated on one or more inputs. Suitable computing devices include microprocessor-based multipurpose computer systems that access stored software (i.e., computer-readable instructions stored in the memory of a computer system) that programs or configures the computing system from a general-purpose computing device to a dedicated computing device that implements one or more aspects of this subject. Any suitable programming, scripting, or other type of language or combination of languages ​​can implement the teachings contained herein in software that will be used to program or configure the computing device.

[0096] The aspects of the methods disclosed herein can be executed in the operation of such a computing device. The order of the boxes presented in the above examples can be varied; for example, the boxes can be reordered, combined, and / or decomposed into sub-boxes. Some boxes or procedures can be executed in parallel.

[0097] The use of “suitable for” or “configured to” here implies open-ended and inclusive language, which does not exclude devices suitable for or configured to perform additional tasks or steps. Similarly, the use of “based on” implies open-ended and inclusive language, because a process, step, calculation, or other action “based on” one or more of the stated conditions or values ​​may also be based on additional conditions or values ​​beyond those stated. The headings, lists, and numbering included herein are for illustrative purposes only and are not restrictive.

[0098] Although the subject matter has been described in detail with respect to specific aspects, it should be understood that those skilled in the art, upon gaining an understanding of the foregoing, can readily make changes, variations, and equivalents to these aspects. Therefore, it should be understood that this disclosure has been presented for illustrative purposes rather than for limitation, and does not exclude such modifications, variations, and / or additions to the subject matter that will be readily apparent to those skilled in the art.

Claims

1. A method for discovering the topological location and phase of one or more utility installations in a power distribution system, comprising: The head-end system receives descriptors from multiple utility devices connected to the power distribution system, the descriptors being generated at the respective utility devices by processing sensor data obtained at the respective utility devices; In response to determining that at least a threshold number of descriptors for the normal mode have been received, The head-end system groups the plurality of utility devices by applying a clustering algorithm to the descriptors of the plurality of utility devices to generate the current group; The headend system compares the current packet with past packets to determine the confidence level of the current packet; The head-end system determines whether the confidence level exceeds a threshold confidence value; as well as In response to determining that the confidence level exceeds the threshold confidence value, the headend system assigns at least one of a segment identifier or a phase identifier to one or more of the plurality of utility devices.

2. The method according to claim 1, further comprising: Determine whether the plurality of utility devices are operating in a fast mode, wherein in the fast mode, the segment identifier or the phase identifier is assigned to the utility device within a shorter time period than in the normal mode; In response to determining that the plurality of utility devices are operating in fast mode and that sufficient descriptors for the fast mode have been received, the plurality of utility devices are grouped by relating the received descriptors for the fast mode. Allocation information is generated by assigning at least one of segment identifiers or phase identifiers to an unknown utility device among the plurality of utility devices; Send the allocation information to the unknown utility device; and Send instructions to the plurality of utility devices to exit the fast mode.

3. The method of claim 1, further comprising, in response to determining that the confidence level does not exceed the threshold confidence value, the headend system generating a new descriptor configuration and sending it to one or more of the plurality of utility units.

4. The method according to claim 1, wherein, Comparing the current group with past groups to determine the confidence level of the group includes assigning a high confidence level in response to determining that a group of utility devices are grouped together in the current group and the past group; In addition, in response to determining that a group of utility devices is grouped differently in the current group than in the past, a low confidence level is assigned.

5. The method according to claim 1, further comprising: It is determined that the descriptor contains a synchronization event; as well as Use the occurrence time of the synchronization event as a reference time to adjust the timing of other events in the descriptor.

6. The method according to claim 5, further comprising: Based on the synchronization event, determine the time offset of one or more of the plurality of utility devices; as well as Instructions are sent to one or more of the plurality of utility devices to correct the clock according to the corresponding time offset.

7. A method performed by a utility device to generate descriptors for discovering the topology and phases of a power distribution system, comprising: The sensor data obtained from the utility equipment is processed to generate processed data; Determine whether the utility equipment is operating in fast mode or normal mode, wherein the utility equipment generates descriptors at a higher rate when operating in fast mode than when operating in normal mode; In response to the determination of utility equipment to operate in fast mode, Generate fast descriptors based on processed data; and In response to determining that at least a first threshold number of fast descriptors have been generated, the fast descriptors are sent to a headend system communicatively connected to the utility equipment; and In response to the determination that utility equipment should be operated in a normal mode, Generate regular descriptors based on the processed data; and In response to determining that at least a second threshold number of regular descriptors has been generated, the regular descriptors are sent to the headend system, wherein the second threshold number is higher than the first threshold number, and wherein the fast descriptors and the regular descriptors are different.

8. The method according to claim 7, wherein, Processing the sensor data to generate processed data includes filtering the sensor data, transforming the sensor data to the frequency domain, or normalizing one or more of the sensor data.

9. The method according to claim 8, wherein, Processing the sensor data also includes: Disturbances are identified by determining that the values ​​of the processed data are outside a predetermined range; and Determine the characteristics of the disturbance.

10. The method according to claim 9, wherein, The regular descriptor is generated to include one or more of the characteristics of the disturbance.

11. The method of claim 7, further comprising adjusting the reference time of the utility equipment in response to receiving an instruction from the headend system to correct the reference time of the utility equipment.

12. The method of claim 7, further comprising: In response to receiving a first user input at the utility facility or a first instruction from the headend system, the utility facility is operated in the fast mode; as well as In response to receiving a second user input at the utility facility or a second instruction from the headend system, exit the fast mode.

13. The method of claim 7, further comprising: Receive allocation information for the utility equipment from the headend system; as well as The allocation information is displayed on a display device associated with the utility equipment.

14. A system for generating descriptors for discovering the topological location and phase of a power distribution system, comprising: Multiple utility units, connected to a power distribution system and communicatively connected to a headend system, each of the multiple utility units is configured to: Sensor data obtained at the utility facility is processed to generate processed data; Generate and send fast or regular descriptors based on the processed data; and The headend system is configured to: Receive the fast descriptor or the regular descriptor from the plurality of utility units; Determine whether the plurality of utility devices are operating in fast mode or normal mode, wherein, in fast mode, segment identifiers or phase identifiers are assigned to the utility devices among the plurality of utility devices within a shorter time period than in normal mode; In response to the determination that multiple utility facilities are operating in normal mode, The plurality of utility units are grouped by applying a clustering algorithm to the conventional descriptors of the utility units to generate the current group; The current group is compared with past groups to determine the confidence level of the current group; and At least one of the segment identifiers or phase identifiers is assigned to one or more of the plurality of utility devices based on the confidence level.

15. The system according to claim 14, wherein, Each of the plurality of utility devices is further configured to determine whether the utility device is operating in the fast mode or the normal mode, wherein the normal descriptor is generated and sent in response to determining that the utility device is operating in the normal mode.

16. The system according to claim 15, wherein, Each of the plurality of utility units is also configured to respond to determining that the utility unit is operating in the fast mode. Generate fast descriptors based on processed data; Determine that at least a first threshold number of fast descriptors have been generated; as well as The fast descriptor is sent to the headend system.

17. The system according to claim 16, wherein, Each of the plurality of utility devices is further configured to, in response to determining that the utility device is operating in the normal mode, determine that at least a second threshold number of normal descriptors has been generated, wherein, in response to determining that at least a second threshold number of normal descriptors has been generated, send the normal descriptors to the headend system.

18. The system according to claim 17, wherein: Determining that at least a first threshold number of fast descriptors has been generated includes determining that the at least first threshold number of fast descriptors has been generated for the fast mode interval; Determining that at least a second threshold number of regular descriptors has been generated includes determining that at least a second threshold number of regular descriptors has been generated for a predetermined number of descriptor intervals; as well as The fast mode interval is shorter than the descriptor interval.

19. The system according to claim 14, wherein, The headend system is also configured to: Determine whether the confidence level exceeds the threshold confidence value; as well as In response to determining that the confidence level does not exceed the threshold confidence value, a new descriptor configuration is generated and sent to one or more of the plurality of utility units. In response to determining that the confidence level exceeds the threshold confidence value, at least one of the segment identifiers or phase identifiers is assigned to one or more of the plurality of utility devices.

20. The system according to claim 14, wherein, The headend system is also configured to respond to determining that the plurality of utility units are operating in fast mode. The multiple utility units are grouped by associating the fast descriptors; Allocation information is generated by using adjacent utility devices of the unknown utility device among the plurality of utility devices as a reference, and assigning at least one of segment identifiers or phase identifiers to the unknown utility device among the plurality of utility devices. Send the allocation information to the unknown utility device; as well as Send instructions to the plurality of utility devices to exit the fast mode.

Images