Separation of electricity data from distributed energy resources

Abstract

A system for separating electrical data from multiple distributed energy resource (DER) devices on a site. The system includes a multiport electric meter having a grid port configured for connecting the multiport electric meter to a power grid, and at least one DER circuit port configured for connecting the multiport electric meter to a circuit having two or more DER devices on it. The system also includes a data collector configured to collect data from multiple data channels, at least one of which includes an electrical data channel and at least one which includes a non-electric data channel, corresponding to one of the multiple DER circuit ports. The system further includes a processor configured to input the data collected from the multiple data channels into a classifier to classify the sampled electrical data into one or more classes, and to estimate the electricity usage and / or generated values ​​of one or more of the DER devices based on the classification output by the classifier.

Description

Technical Field

[0001] Technical Field The present invention relates to a system for separating electrical data from a plurality of distributed energy resource devices, a multi-port electricity meter, and a method for separating electrical data.

Background Art

[0002] Background The field of distributed energy resources (DER) has been growing rapidly in recent years due to the increasing number of buildings and homes equipped with renewable energy sources such as solar panels, wind turbines, and battery energy storage systems. These DERs have the potential to significantly reduce energy costs and carbon dioxide emissions, but they also pose new challenges to energy management and grid stability.

Summary of the Invention

Problems to be Solved by the Invention

[0003] One of the important issues in DER management is the need to collect and analyze data from multiple devices in real time. This data may include not only information related to energy generation, consumption, and storage, but also environmental factors such as temperature and humidity. Without accurate and timely data, it is difficult to optimize energy usage, predict demand, and ensure grid stability.

[0004] To address this issue, various systems for collecting and analyzing DER data have been developed. These systems range from simple data loggers to complex software platforms integrated with building automation systems and utility networks. However, many of these systems have limitations in terms of scalability, interoperability, and data accuracy.

[0005] Improvements are desired to overcome these shortcomings and enable more efficient and effective management of DERs. [Means for solving the problem]

[0006] overview In general, this disclosure relates to a system for collecting and analyzing electrical data from multiple distributed energy resource devices on a site. The system utilizes a multiport electric meter and a data collector to classify the data and estimate the use and generation of the devices. Advantageously, the invention utilizes a multiport meter to collect data at real-time or near real-time sampling rates across multiple data channels of the multiport meter, each port / channel corresponding to a circuit having one or more DER devices on it, as well as other data sources such as historical weather forecast data including temperature, cloud cover, and solar data, real-time weather data such as real-time temperature, cloud cover, and solar data, site location such as latitude and longitude, and one or more signals indicating the operating status of the devices, such as "on" or "off" status, received from one or more of the DER devices, or when the DER is on or off. The system also collects or receives data such as the charge level of the DER device, voltage or current waveform data of the DER device, and steady-state consumption of the DER device, measured in real time or near real time during the transition time while the system is tuned. By classifying the sampled data into one or more classes (e.g., presence or absence of a DER device, and / or estimation of power consumption or generated values ​​from the DER device and / or circuits on which the DER device is located), the system solves the problem of accurately isolating and monitoring the energy consumption and generation of various distributed energy resource devices such as solar panels, batteries, and electric vehicle (EV) chargers in a cost-effective and safe manner. By using not only data measured by a multiport electric meter, but also other data such as data provided by signals indicating the operating state of one or more DER devices, it becomes possible to isolate the energy flow to and from the DER devices with much greater accuracy.

[0007] According to one aspect of the present invention, a system is provided for separating electrical data from a plurality of distributed energy resource (DER) devices on a site. The system includes a multiport electric meter, which includes (i) a grid port configured for connecting the multiport electric meter to a power grid, and (ii) one or more DER circuit ports, each configured for connecting the multiport electric meter to a circuit having two or more DER devices on at least one DER circuit port; a data collector configured to collect data from a plurality of data channels, wherein at least one data channel includes an electrical data channel and corresponds to one of the plurality of DER circuit ports, and at least one data channel includes a non-electrical data channel; and a processor configured to input the data collected from the plurality of data channels to a classifier in order to classify the collected electrical data into one or more classes, and the processor to estimate the electricity usage and / or generated values ​​of one or more of the DER devices based on the classification output by the classifier.

[0008] Optionally, non-electric data channels include environmental data channels and / or geographical data channels.

[0009] Optionally, the collected non-electrical data may include one or more of the following: meteorological data and / or geographic location data.

[0010] The first sampling rate is optional and ranges from 1 Hz to 1 MHz.

[0011] Optionally, the processor is configured to perform input steps and estimation steps to provide continuous real-time or near-real-time estimations of the power consumption and / or generated values ​​of each DER device and / or DER circuit.

[0012] Optionally, one or more classes include at least one of the following: presence or absence of an electric vehicle charger, presence or absence of a battery system and / or solar inverter, presence or absence of an HVAC system, presence or absence of a swimming pool pump, presence or absence of a water heater, presence or absence of a dishwasher, presence or absence of a washing machine or dryer, and / or presence or absence of a freezer or refrigerator.

[0013] Optionally, the system includes one or more classes of DER devices having estimated electricity consumption and / or generated values.

[0014] Optionally, at least one data channel includes data received from at least one of the DER devices via a communication channel.

[0015] Optionally, data received from at least one of the DER devices via a communication channel includes measurement data acquired by the DER device and / or data indicating the operating status of the DER device. The operating status of the DER device may refer to, for example, the on or off state of the device, the charge level of the DER device, and / or any other data indicating how the DER is operating or not operating at a given time.

[0016] Optionally, the processor is located on the aforementioned site.

[0017] Optionally, the multiport meter includes a housing, and the processor and data collector are housed within the housing.

[0018] Optionally, the processor is configured to generate a control signal based on the classification output by the classifier and to transmit the control signal to one or more of the DER devices.

[0019] Optionally, the control signals include instructions for turning one or more DER devices on or off, or for setting the devices to a predetermined power consumption or generated value.

[0020] Optionally, the processor is configured to transmit the classification results output by the classifier, the estimated electricity usage of one or more DER devices and / or DER circuits, and / or generated values ​​to the power network provider supplying the site.

[0021] According to one aspect of the present invention, a multiport electric meter is provided for separating electrical data from multiple distributed energy resource (DER) devices on a site. The multiport electric meter includes a grid port configured for connecting the multiport electric meter to a power grid, one or more DER circuit ports, each configured for connecting the multiport electric meter to a circuit having two or more DER devices thereon, a data collector configured to collect data from multiple data channels, the data collector having at least one data channel including an electrical data channel and corresponding to one of each of the multiple DER circuit ports, and at least one data channel including a non-electric data channel, and a processor configured to input the data collected from the multiple data channels to a classifier in order to classify the collected electrical data into one or more classes, and to estimate the electricity usage and / or generated values ​​of one or more of the DER devices based on the classification output by the classifier.

[0022] The first sampling rate is optional and ranges from 1 Hz to 1 MHz.

[0023] Optionally, the processor is configured to perform input steps and estimation steps to provide continuous real-time or near-real-time estimations of the power consumption and / or generated values ​​of one or more DER devices.

[0024] Optionally, the processor is configured to generate a control signal based on the classification output by the classifier and to transmit the control signal to one or more of the DER devices.

[0025] Optionally, the control signal includes instructions for turning on or off one or more of the DER devices or for setting the device to a predetermined electricity usage or generation value.

[0026] According to one aspect of the present invention, a method for separating electrical data from a plurality of distributed energy resource (DER) devices within a site is provided. The method uses a multi-port electric meter having a grid port connected to a power grid, one or more DER circuit ports each connected to a circuit having two or more DER devices thereon, and a data collector, and at least one includes an electrical data channel and a plurality of data channels corresponding to each of the plurality of DER circuit ports, wherein at least one data channel includes a non-electrical data channel, collecting data from the plurality of data channels, using a processor to input the data collected from the plurality of data channels into a classifier to classify the collected electrical data into one or more classes, and using a processor to estimate the electricity usage and / or generation value of one or more of the DER devices based on the classification output by the classifier.

[0027] Optionally, the first sampling rate is from 1 Hz to 1 MHz.

[0028] Optionally, the method includes performing the steps of inputting and estimating at a second sampling rate of 1 Hz to 1 MHz to provide a continuous real-time or near real-time estimate of the electricity usage and / or generation value of one or more of the DER devices and / or DER circuits.

[0029] Brief Description of the Drawings Here, these aspects and other aspects will be described with reference to the following figures.

Brief Description of the Drawings

[0030] [Figure 1]This disclosure exemplifies a system for separating electrical data from multiple distributed energy resource (DER) devices on a site. [Figure 2] The multiport electric meter described herein is illustrated in an example. [Figure 3] This disclosure exemplifies a method for separating electrical data from multiple distributed energy resource (DER) devices on a site. [Figure 4] An example of the architecture described herein is shown exemplarily. [Figure 5] An example of the architecture described herein is shown exemplarily. [Figure 6] An illustrative flowchart of the method described herein is shown. [Figure 7] An illustrative flowchart of the method described herein is shown. [Modes for carrying out the invention]

[0031] Detailed explanation Figure 1 illustrates a system 100 for isolating electrical data from multiple distributed energy resource (DER) devices 102a, 102b on a site DER circuit (i.e., a circuit in which multiple DER devices are connected) as described herein. The system 100 includes a multiport electric meter 103 having one or more DER circuit ports 105a, a grid port 104 for connecting the multiport electric meter to a grid 104a of a power network, a data collector 106 configured to collect not only electrical data from one or more DER circuit ports 105a but also other non-electrical data (environmental data and geographical data, e.g., weather data and / or location data) from one or more other data sources, and a processor 107. The multiport electric meter 103 in Figure 1 also has a non-DER device port 105b, through which one or more non-DER devices 102c may be connected. The processor 107 is configured to input the collected electrical data into a classifier in order to classify the collected electrical data into one or more classes, and to estimate the electricity consumption and / or generated values ​​of one or more of the DER devices 102a, 102b based on the classification output by the classifier.

[0032] Although only one DER circuit is shown in Figure 1, this can be extended to multiple DER circuits, and accordingly, it is assumed that the same advantages described herein are applicable to systems having multiple DER circuits, and therefore multiple DER circuit ports on a multiport electric meter.

[0033] This disclosure is intended to be used not only for isolating data from DER devices on a DER circuit via a DER circuit port, but also for isolating data from non-DER devices on non-DER device ports of a multiport meter.

[0034] Advantageously, because the isolation of electrical data is performed within a computer-enabled power network edge intelligence device (i.e., multiport meters are provided with a microprocessor unit (MPU) or microcontroller unit (MCU)), the system can easily utilize and process data collected from multiple data channels locally without relying on cloud infrastructure. Additionally or alternatively, the system's functionality can be enhanced by using cloud infrastructure to collect, store, stream, and / or process data from external sources, for example, as data channels that can be used as input to a classifier. Furthermore, because multiport electricity meters enable monitoring of multiple channels, unlike conventional single-channel meters, they not only have one channel for conventional household electricity use and additional channels for DER devices such as solar panels, batteries and electric vehicle (EV) chargers, HVAC systems, pool pumps, water heaters, dishwashers, washer / dryer, freezers, refrigerators, and any other large household loads, but also channels for non-electrical data (such as weather data and location data) that can correlate with electrical data. This allows for more accurate estimations than systems that rely solely on electrical data, thanks to the diverse and numerous data channels and sources, enabling the system to more accurately estimate the electricity consumption and / or generated values ​​of each DER device 102a, 102b, and non-DER device 102c.

[0035] In some embodiments, the first sampling rate is 1 Hz to 1 MHz. In some embodiments, the processor 107 is configured to perform the input step and the estimation step to provide a continuous real-time or near real-time estimation of the power consumption and / or generated values ​​of each of the DER devices 102a, 102b on the DER circuit.

[0036] Advantageously, compared to the sampling rates of existing technologies that use sampling and inference rates ranging from minutes to days, weeks to months, the higher sampling rate of System 100 allows for real-time or near-real-time isolation of electrical data. This enables the system to detect changes in electrical usage, such as a 4kW increase that may indicate an EV charger is in use, at intervals of seconds or faster, by analyzing data from the DER data channel and other data channels. It also becomes easier to incorporate more granular real-time weather data into data channels, which can improve the accuracy of the output classification that is not possible with lower sampling rates such as daily, weekly, monthly, etc. In certain non-limiting examples, solar panels and inverters may be present in the DER circuit, and cloudy weather conditions with intermittent sunshine may cause peaks and dips in the output of the solar inverter. Electrical and weather data sampled at the above rates of 1Hz to 1MHz can better capture fine changes in real time.

[0037] It will be understood that the classifiers envisioned herein are trained on training datasets containing data that matches the sampling rate and data type they will receive when performing inference. That is, if input to a classifier sampled at 1 Hz, the classifier is assumed to have been trained on training data that is also sampled at 1 Hz (or synthetically generated to simulate such data). Conversely, if input to a classifier sampled at 1 MHz, the classifier is assumed to have been trained on 1 MHz sampled (or synthetic) data, etc. Similarly, the training datasets contain the same types and number of non-electric data channel types (e.g., weather, location, etc.) that are assumed to be used during inference.

[0038] In some embodiments, one or more classes include the presence or absence of an electric vehicle charger, a battery system, and / or a solar inverter, an HVAC system, a swimming pool pump, a water heater, a dishwasher, a washing machine or dryer, and / or a freezer or refrigerator. In some embodiments, system 100 includes one or more classes of estimated electricity usage and / or generated values ​​from one or more DER devices 102a, 102b.

[0039] Advantageously, system 100 can classify electrical data into various classes, corresponding to the presence or absence of specific DER devices such as electric vehicle chargers, battery systems, and solar inverters. This allows system 100 to more accurately estimate the electrical consumption and / or generated values ​​of each DER device 102a, 102b, thereby increasing the overall efficiency and effectiveness of the system. Specifically, by making a simple classification of the presence or absence of one or more DER device types that typically constitute the majority of use or generation in a DER circuit (e.g., EV chargers, solar inverters, and / or battery systems), it becomes possible to infer further information. For example, if a classification is made that solar inverters and EV chargers are present on the DER circuit but battery systems are absent, it becomes possible to infer that any use or generation is entirely attributable to the solar inverters and / or electric vehicle chargers, and any signals that might otherwise have been incorrectly characterized as indicating the presence of a battery system can be ignored when separating the data. Conversely, if solar inverters, electric vehicle chargers, and battery systems are classified as being present on the DER circuit, it can be inferred that uniquely identifiable signal features associated with such devices may be present in the data channel and could be used accordingly to isolate the data.

[0040] In some embodiments, the processor 107 is further configured to receive weather data associated with the geographical location of the system for input to the classifier, thereby the weather data being an additional data channel.

[0041] Advantageously, as described above, system 100 can improve the accuracy of separation by incorporating meteorological data such as temperature, cloud cover, and solar radiation forecasts. By combining this meteorological data with sampled electrical data, the system can better estimate the electricity consumption and / or generated values ​​of each DER device 102a, 102b and take into account the impact of meteorological conditions on the performance of these devices. This makes it possible to manage electrical data from DER devices more accurately and efficiently. Specifically, certain signals in the data collector data, such as cloud cover and decreased solar inverter output, or freezing temperature and increased power consumption of electric vehicle chargers, are likely to be correlated. By incorporating this data as a data channel into the system, information from these correlations can be captured during classifier training, and accordingly, the classifier can classify more accurately when it encounters such data during inference.

[0042] In some embodiments, the processor 107 is configured to receive state data and / or measurement data from and acquired by one or more of the DER devices for input to the classifier, thereby this state data and / or measurement data is a further data channel. Advantageously, this allows the system to incorporate additional information from the DER devices themselves, such as on / off indicators, charge level indicators (e.g., charge status), or signals corresponding to solar output (e.g., irradiance). By combining this DER state data and / or measurement data with sampled electrical data, the system can improve the accuracy of separation and better estimate the electrical consumption and / or generated values ​​of each DER device 102a, 102b. This is particularly useful in scenarios where the DER device has multiple charge levels or variable output levels, and / or its own weights and measures, because additional information provided through further data channels, such as the operating status of the DER device and / or any other data received from the DER device, enhances the accuracy of estimating isolation.

[0043] In some embodiments, the processor 107 is located on the premises where the system 100 is situated, rather than in the cloud. Advantageously, this not only enables faster processing and communication between the DER device and the processor 107, but also reduces system latency more broadly, as it eliminates the need to rely on cloud-based computing resources. By locating the processor 107 on-premises, the system can, accordingly, more efficiently manage electrical data from the DER device and provide real-time or near-real-time estimates of electricity usage and / or generated values. However, in some alternative embodiments, it is also conceivable that one or more additional processors may be provided, located in the cloud or elsewhere on the network with the system, allowing computing resources to be distributed across several devices.

[0044] In some embodiments, the multiport electric meter 103 includes a housing, and the processor 107 and data collector 106 are housed within the housing. Advantageously, this not only provides a compact, integrated, single-part device that allows for easier installation and maintenance of the system 100, but also improves communication between the processor 107 and the data collector 106, as the main functional components of the system 100 are housed in a single, convenient device (the multiport meter 103) rather than being scattered across different locations and servers in the cloud. Specifically, by housing the processor 107 and the data collector 106 together, the system 100 can more efficiently sample and process electric data from the DER device, leading to more accurate separation and estimation of electricity usage and / or generated values.

[0045] In some embodiments, the processor 107 is further configured to generate control signals based on the classification output by the classifier and to transmit the control signals to one or more of the DER devices. The control signals may include, for example, instructions to turn one or more of the DER devices on or off, or instructions to set the devices to predetermined power consumption or generated values. Advantageously, this enables the system to proactively manage the DER devices based on isolated electrical data and classification results, without relying on external control signals received via the cloud or other communication channels. By generating and transmitting control signals to the DER devices, the system can optimize power usage and / or generation, thereby increasing the overall efficiency and effectiveness of the system.

[0046] In some embodiments, the processor 107 is further configured to transmit the classification results output by the classifier and / or the estimated electricity usage and / or generated values ​​of one or more of the DER devices 102a, 102b and / or DER circuits to the provider of the power network 104a supplying the site. Advantageously, this allows the power network provider to access detailed information about the electricity usage and / or generation of the DER devices, enabling better management and optimization of the power network. By sharing this information with the power network provider, the system can contribute to a more efficient and reliable power grid, benefiting both the site and the wider community. For example, the power network provider can override any control signals generated by the processor to enable manual management of the power network. Further conceivable use cases include using the separated DER data to facilitate billing / payment at specific rates for different DER devices (e.g., for EV charging, solar power generation, etc.), as well as calculating any applicable renewable energy credits if such schemes are operational at the location where the system is located. More specifically, by relying on multiple electrical and non-electrical data channels, the accuracy of DER data isolation can exceed a predetermined percentage (e.g., 2%), making it possible to use it for billing purposes. This is often impossible with existing isolation techniques that rely solely on electrical data channels. Furthermore, if a DER device has its own weights and measures controlled by the DER device provider, access to these weights and measures can present some difficulties for utility providers, often in terms of cost, security issues, and inability to guarantee accuracy. One existing solution to this is to provide each DER with its own high-accuracy (e.g., revenue grade, accuracy exceeding 2%) dedicated meter to measure its consumption and / or generation. However, this solution is expensive and complex.Accordingly, the separation described in this disclosure facilitates the realization of highly accurate consumption and / or generation estimations without requiring separate, dedicated meter hardware for each device. However, in some embodiments, it is also conceivable that the multiport meter of the present invention may have a dedicated DER port for each device. For example, if there are two DER devices, the meter may have two DER ports, each providing accurate, direct consumption and generation measurements related to the connected DER device.

[0047] Figure 2 illustrates a multiport electric meter 203 according to the present disclosure. The multiport electric meter 203 includes a grid port 204, a non-DER device port 205a, a DER circuit port 205b, a data collector 206, and a processor 207. The grid port 204 is configured to connect the multiport electric meter 203 to a power grid. The DER circuit port 205b is configured to connect the multiport electric meter 203 to a circuit having two or more DER devices 102b, 102c on it. The data collector 206 is configured to collect data from multiple data channels, at least one of which includes an electrical data channel corresponding to each of the DER circuit ports 205b, and at least one of which includes a non-electrical data channel. The processor 207 is configured to input data collected from multiple data channels into a classifier in order to classify the collected electrical data into one or more classes, and to estimate the power consumption and / or generated values ​​of one or more DER devices and / or DER circuits based on the classification output by the classifier.

[0048] Advantageously, as described above, the multi-port electricity meter 203 allows for monitoring of multiple data channels, including electrical and non-electrical data (e.g., weather, location, and operating status of DER devices, as mentioned above), in contrast to conventional single-channel meters. This enables the meter 203 to more accurately estimate the electricity consumption and / or generated values ​​of each DER device, such as solar panels, batteries, and electric vehicle (EV) chargers. As a result, the system can more accurately isolate electrical data, improving the overall efficiency and effectiveness of the system.

[0049] In some embodiments, the first sampling rate is 1 Hz to 1 MHz. In some embodiments, the processor is configured to perform the input step and the estimation step to provide a continuous real-time or near real-time estimation of the power consumption and / or generated values ​​of one or more DER devices.

[0050] Advantageously, as mentioned above, this sampling rate facilitates real-time or near-real-time isolation of electrical data and is processed locally, eliminating the need to rely on cloud computing power.

[0051] In some embodiments, the processor 207 is further configured to generate a control signal based on the output classified by the classifier and to transmit the control signal to one or more of the DER devices. In some embodiments, the control signal includes instructions to turn one or more of the DER devices on or off, or instructions to set the devices to a predetermined power consumption or generated value.

[0052] Advantageously, the system can generate control signals based on the classification output by the classifier, enabling more efficient management of the DER device. This may include turning the device on or off, or setting the device to a predetermined power consumption or generated value. By incorporating the various inputs enumerated in this disclosure, the system can control the DER device more precisely, leading to improved energy management and potential cost savings for the user.

[0053] Figure 3 illustrates a flowchart of a method 300 for separating electrical data from multiple distributed energy resource (DER) devices on a site, as described herein. The method 300 includes using a multiport electric meter having a grid port connected to a power grid, at least one DER circuit port, each connected to a circuit having two or more DER devices thereon, and a data collector. The data collector collects data from data channels, each having multiple data channels, at least one of which includes an electrical data channel and corresponding to one of the multiple DER circuit ports, and at least one of which includes a non-electric data channel 301. The method 300 then includes inputting the data collected from the multiple data channels into a classifier 302 to classify the collected electrical data into one or more classes, and estimating the electricity usage and / or generated values ​​of one or more of the DER devices based on the classification output by the classifier 303.

[0054] In some embodiments, the first sampling rate is 1 Hz to 1 MHz. In some embodiments, the method includes performing input and estimation steps at a second sampling rate of 1 Hz to 1 MHz to provide a continuous real-time or near real-time estimation of the power consumption and / or generated values ​​of one or more DER devices.

[0055] Advantageously, a method of separating electrical data using a multiport electric meter and multiple data channels (including electrical data channels and non-electrical data channels) enables more accurate estimation of the electricity usage and / or generated values ​​of DER devices. This method facilitates the detection of changes in electricity usage, such as a 4kW increase that may indicate that an EV charger is in use, by analyzing data from one or more of the data channels. The method can, accordingly, incorporate solar data and other inputs such as historical data, weather forecasts, real-time cloud cover, solar radiation forecasts, house location, input signals from DER devices, on / off indicators from EV chargers or battery system chargers, charge % indicators from EV chargers or battery system chargers, and signals from solar inverters corresponding to solar output to improve the accuracy of separation. This method can be applied in a variety of scenarios, as described in the examples of this disclosure, to better estimate the electricity usage and / or generated values ​​of each DER device and improve the overall efficiency and effectiveness of the system.

[0056] Figure 4 illustrates an illustrative flowchart of a classifier architecture 400 that may be used in conjunction with this disclosure.

[0057] Architecture 400 includes multiple input data channels 401a, 401b, and 401c, such as those described above, for example, those corresponding to each of the DER circuit ports of a multiport meter, and the operating state of one or more DER devices connected to the DER circuit ports, waveform data measured in real time or near real time during the transition time while the DER devices are tuned to on or off, as well as those corresponding to other input data channels, such as the steady-state consumption of the DER devices, weather data, past usage and generation. Data from the input data channels is passed to a low-level feature detection component 402 configured to extract low-level features from input data channels 401a, 401b, and 401c. The extracted low-level features are then passed to a high-level feature detection component 403 configured to extract high-level features from input data channels 401a, 401b, and 401c. The extracted high-level features are then passed to a temporal ordering component 404 before being passed to a classification component 405. The classification component 405 outputs classification results, for example, an output class indicating whether a given DER device type exists or not, and / or a subtype of that device, and / or an estimated usage or generated value of the device during a given sampling period.

[0058] The low-level feature detection component 402, the high-level feature detection component 403, the temporal ordering component 404, and the classification component 405 are separately trained standalone models, where the output of one model may be passed as input to the next model, or they may be combined into a single model, thereby allowing all the weights, biases, and other parameters of that single model to be collectively and iteratively adjusted. The components may include one or more of a recurrent neural network and a convolutional neural network, and the final output layer of the classification component 405 may be a softmax layer.

[0059] Figure 5 illustrates an exemplary architecture 500 of a machine learning model for energy forecasting according to this disclosure. Several inputs are illustrated, which may be in the form of n-dimensional vectors (e.g., embeddings) that may include one or more of the following: one or more vectors 501a of recent historical energy values; one or more vectors 501b of weather forecast data (e.g., temperature, solar radiation, and other meteorological data); one or more vectors of temporal information (e.g., day of the week, week of the year, etc.); and one or more vectors of geographical location (e.g., latitude and / or longitude data). As stated above, it is assumed that other input data sources and types may also be provided, namely the operating state of one or more DER devices connected to the DER circuit port, waveform data measured in real time or near real time during transition time while the DER devices are tuned on or off, and steady-state consumption of the DER devices. The input vectors are fed to a machine learning model 502 suitable for regression. This could be, for example, one or more of the following models: a decision tree model, a random forest model, a support vector machine (SVM) model, an XGBoost model, a long-term short-term memory (LSTM) model, or an autoregressive integrated moving average (ARIMA) model. In any case, it will be understood that the model used is trained on a training dataset of appropriate size and, as appropriate, achieves the desired performance with respect to one or more of the following machine learning model performance metrics: accuracy, precision, recall, and / or other metrics. The output of the model is expected to include a vector containing predicted energy values, i.e., values ​​indicating energy consumption and / or generation.

[0060] Figure 6 illustrates a flowchart of the method according to the disclosure that can determine or classify energy consumption and / or generation when there are two DER devices (a battery system such as an EV or stationary battery system and a solar inverter) connected to a single DER circuit port of an electric meter. In this example, from one input data channel (e.g., weather data, time data, and / or geographical data), it is determined that the sun has set and therefore the solar inverter is not outputting energy, and from another input data channel, it is determined that the energy consumption is below a threshold.601 In this case, it can be simply determined that the battery system is not being charged. Specifically, it can be determined that all DER devices are off. In this case, the power consumption of the system can be set to the nominal power consumption of the devices (e.g., the always-on load value).602

[0061] Figure 7 illustrates the method according to the present disclosure, which allows for a determination or classification of energy consumption and / or generation when there are two DER devices (a battery system such as an EV or stationary battery system, and a solar inverter) connected to a single DER circuit port of an electric meter. In this example, it is determined from one data channel (e.g., an input signal from a DER device indicating a change in the operating state of a DER device) that the state of one or more DER devices has changed.701 The multiport electric meter simultaneously detects and measures an increase in consumption at the DER circuit port indicating that the battery system is consuming power.702 It is then determined whether it is possible to determine which device on the DER circuit is off (e.g., by using weather data, time data, and / or geographical data as described above, in the case of a solar inverter) using the method of Figure 6.703 Once this determination is made, the nominal value (e.g., the value of a constantly on load) is then subtracted from the consumption measurement of the multiport electric meter.704 The resulting value is a power value that can be assigned to the power consumption of the (charging) battery system.705 Next, another input data channel (e.g., the presence or absence of a signal indicating the operating status from one or more DER devices) is used to determine whether a predetermined percentage of charge is provided to the charging battery system 706. If provided, the amount intended to be provided according to the signals received from the DER devices is correlated with the actual charge rate provided 707, and the signature, including waveform data of the DER devices during power-up, is assigned as the power consumption value to the battery system 708. If not provided, it is determined whether the battery system has two or more charge levels 709. If multiple charge levels are available, the resulting power consumption value is assigned to one of the charge levels of the battery system 710, and the signature, including waveform data of the DER devices during power-up, is assigned as the power consumption value to the battery system 709.If no multiple levels exist, the resulting value is assigned to the power consumption of the battery system 711, and the signature containing waveform data during power-up of the DER device is assigned to the power consumption value to the battery system 708.

[0062] As used herein, the term “classifier” may refer to an algorithm, software, or hardware component within a system that processes electrical data collected from distributed energy resource devices and categorizes it into one or more classes.

[0063] As used herein, the term “control signal” may refer to electrical or digital signals transmitted to a distributed energy resource device to adjust its operation, such as regulating power output, switching it on or off, or modifying other operating parameters.

[0064] As used herein, the term “estimated usage / generation” may refer to a calculated amount of electrical energy used or generated by distributed energy resource devices on a site, based on data collected and analyzed from multiport electric meters and / or other data sources.

[0065] As used herein, the term “multiple data channels” may refer to multiple communication paths or connections through which electrical and other data from various distributed energy resource devices are collected, transmitted, and analyzed.

[0066] Where described herein, it will be understood that the data is available to the utility provider. This does not have to be limited to data at the point of supply or consumption, and may also include other data sources such as local carbon intensity scores and / or other metrics that may be available to the utility provider in certain locations and jurisdictions. [Examples]

[0067] Here, we will describe some illustrative use cases of this disclosure.

[0068] Example 1: A solar inverter and an EV charger are connected to a single measuring port on an electric meter. There are no signaling (communication) connections between the DER devices and / or between the meters. The EV charger charges at only one level when charging. This system identifies the time when the inverter was unable to output energy when the EV charger was on (indicated by the delivered load) (determined by using location data and other yearly data to determine the daily time when it is dark outside and the solar panels are not outputting energy). If the power flow is in the "delivering" direction (i.e., power is flowing from the utility), it indicates that the load should be on. If the power flow exceeds the threshold of the nominal power consumption of the connected devices (defined as the power consumption of the inverter + EV charger itself, with no solar input and no charging output to the EV), the EV charger should be charging. If the power flow is in the "receiving" direction (energy is flowing towards the utility), the solar inverter should be receiving power from the solar panels, or the EV charger may be discharging using V2X technology. Accordingly, the model of this disclosure can learn an EV charger signature, including what the startup is like and what the actual load is when it is on, assign this to the EV, leaving the rest assigned to solar power. The EV charger signature includes one or more of the following: recognizable voltage and current waveform shapes when the charger starts charging, steady-state power consumption during charging (one or more levels, depending on whether the charger supports only one label or two or more levels), and waveform shapes when the charger stops charging.

[0069] Example 2: The Disclosure includes a signal from an EV charger that indicates the operating status of the EV charger (e.g., charging). In this case, the signal from the EV charger simply indicates whether or not the EV charger is charging. This signal is used by the Disclosure to determine whether or not the EV charger is charging, and the Disclosure determines how much power (in watts or kilowatts) the EV charger is consuming at a given time interval, based on learning that has been done during the time the sun is down.

[0070] Example 3: Similar to Example 2, but the EV charger has multiple charge levels and has either an on / off signal or a signal indicating a charge level that the Disclosure considers as input. The Disclosure can correlate the charge level with the power consumption value of the EV charger by using data from the time of sunset and possibly some interpolation.

[0071] Example 4: Similar to Example 1, but this disclosure uses meteorological data and historical learning to estimate solar output and better determine when an EV charger is being charged. For example, meteorological data may include temperature and cloud cover. An algorithm can use an existing model to determine solar radiation based on these inputs and the angle of the sun at a given time and day for a given location. This disclosure can also further train the model to improve its accuracy over time by comparing the actual power output with the predicted power output and predicted solar radiation. When the power fluctuates from this predicted solar inverter power output, which corresponds to one or more possible charge levels of the EV charger, this power is attributed to the EV charger.

[0072] Example 5: The system may also include stationary batteries used for energy storage, which can be added to the system. The signature for charging the batteries would look similar to the signature for charging.

[0073] Example 6: An EV or stationary battery charger and / or solar inverter may have its own measuring instrument, and this disclosure can improve the accuracy of the measuring instrument. The measuring instrument from the EV or stationary battery charger and / or solar inverter sends its real-time readings to a meter, and an algorithm running inside the meter either directly takes this measuring instrument data or uses the data as input to an algorithm that can correct the actual measuring instrument readings to provide a power breakdown for various devices.

[0074] Those skilled in the art will understand that various modifications can be made to the above examples without departing from the scope of the invention as defined by the appended claims.

Claims

1. A system for separating electrical data from multiple distributed energy resource (DER) devices on a site, A multiport electric meter comprising (i) a grid port configured for connecting the multiport electric meter to a power grid, and (ii) one or more DER circuit ports, each comprising one or more DER circuit ports configured for connecting the multiport electric meter to a circuit having two or more DER devices on at least one DER circuit port, A data collector configured to collect data from multiple data channels, wherein at least one data channel includes an electrical data channel and corresponds to one of the DER circuit ports, and at least one data channel includes a non-electrical data channel, It is a processor, The system is configured to input the collected electrical data from the multiple data channels into a classifier in order to classify the collected electrical data into one or more classes. A processor that estimates the power consumption and / or generated values ​​of one or more of the DER devices based on the classification output by the classifier, A system that includes this.

2. The system according to claim 1, wherein the at least one data channel, which includes a non-electric data channel, includes an environmental data channel and / or a geographical data channel.

3. The system according to claim 2, wherein the collected non-electrical data includes one or more of meteorological data and / or geographic location data.

4. The system according to claim 1, wherein the data collector collects data in the electrical data channel at a first sampling rate, the first sampling rate being between 1 Hz and 1 MHz.

5. The system according to claim 4, wherein the processor is configured to perform the input step and the estimation step in order to provide a continuous real-time or near real-time estimation of the power consumption and / or generated values ​​of each of the DER devices.

6. The system according to claim 1, wherein one or more of the above classes include at least one of the following: the presence or absence of an electric vehicle charger, the presence or absence of a battery system, the presence or absence of a solar inverter, the presence or absence of an HVAC system, the presence or absence of a swimming pool pump, the presence or absence of a water heater, the presence or absence of a dishwasher, the presence or absence of a washing machine or dryer, and / or the presence or absence of a freezer or refrigerator.

7. The system according to claim 1, wherein the one or more classes include one or more estimated electricity consumption and / or generated values ​​of the DER devices.

8. The system according to claim 1, wherein at least one data channel includes data received from at least one of the DER devices via a communication channel.

9. The system according to claim 8, wherein the data received from at least one of the DER devices via the communication channel includes measurement data acquired by the DER device and / or data indicating the operating state of the DER device.

10. The system according to claim 1, wherein the processor is located on the site.

11. The system according to claim 10, wherein the multiport meter includes a housing, and the processor and the data collector are provided within the housing.

12. The system according to claim 1, wherein the processor is further configured to generate a control signal based on the classification output by the classifier and to transmit the control signal to one or more of the DER devices.

13. The system according to claim 12, wherein the control signal includes an instruction to turn on or off one or more of the DER devices, or an instruction to set the devices to a predetermined amount of electricity consumption or generated value.

14. The system according to claim 1, wherein the processor is further configured to transmit the classification results output by the classifier and / or the estimated electricity usage and / or generated values ​​of one or more of the DER devices to a provider of the power network supplying the site.

15. A multi-port electric meter for separating electrical data from multiple distributed energy resource (DER) devices on a site, A grid port configured to connect the aforementioned multiport electric meter to the power grid, One or more DER circuit ports, each configured to connect the multiport electric meter to a circuit having two or more DER devices on at least one DER circuit port, A data collector configured to collect data from multiple data channels, wherein at least one data channel includes an electrical data channel and corresponds to one of the multiple DER circuit ports, and at least one data channel includes a non-electrical data channel, It is a processor, The system is configured to input the collected electrical data from the multiple data channels into a classifier in order to classify the collected electrical data into one or more classes. A processor that estimates the power consumption and / or generated values ​​of one or more of the DER devices based on the classification output by the classifier, A multi-port electric meter, including...

16. The multiport electric meter according to claim 15, wherein the first sampling rate is between 1 Hz and 1 MHz.

17. The multiport electric meter according to claim 16, wherein the processor is configured to perform the input step and the estimation step in order to provide a continuous real-time or near real-time estimation of the electricity consumption and / or generated values ​​of one or more of the DER devices.

18. The multiport electric meter according to claim 15, wherein the processor is further configured to generate a control signal based on the classification output by the classifier and to transmit the control signal to one or more of the DER devices.

19. The multiport electric meter according to claim 18, wherein the control signal includes an instruction to turn on or off one or more of the DER devices, or an instruction to set the devices to a predetermined amount of electricity consumption or generated value.

20. A method for separating electrical data from multiple distributed energy resource (DER) devices on a site, A multiport electric meter having a grid port connected to a power grid, one or more DER circuit ports, each connected to a circuit having two or more DER devices on at least one DER circuit port, and a data collector, to collect data from multiple data channels, wherein at least one of the data channels includes an electrical data channel and corresponds to one of each of the multiple DER circuit ports, and at least one of the data channels includes a non-electrical data channel. Using a processor, the collected electrical data is input to a classifier from the multiple data channels in order to classify the collected electrical data into one or more classes. Using the aforementioned processor, the power consumption and / or generated values ​​of one or more of the DER devices are estimated based on the classification output by the classifier. Methods that include...

Images