Method, device and processor for non-intrusive load disaggregation of power consumption information

Abstract

Embodiments of the present application provide a method, device and processor for non-invasive load identification of power consumption information. The method comprises: obtaining power consumption information of a total power port corresponding to a power consumption device at each time point in a preset time period; processing the power consumption information to determine a load characteristic of the power consumption information; generating a first load curve according to characteristic values of the load characteristic at each time point in the preset time period; clustering and classifying the first load curve based on a preset clustering algorithm to obtain a second load curve after merging and perform analysis to determine an identification result of start-stop events of the plurality of power consumption devices in the preset time period; and matching the identification result with a load characteristic library to determine a device type of a user device for each start-stop event, and a number and time of start-stop events of each device type in the preset time period. The above technical solution can not only complete load identification in real time, but also has accurate load identification results.

Description

Technical Field

[0001] This application relates to the field of smart grid technology, and specifically to a method, apparatus, integrated fusion terminal, storage medium and processor for non-intrusive load identification of electricity consumption information. Background Technology

[0002] Currently, research on user electricity consumption behavior mainly involves mining the large amounts of collected electricity consumption data to analyze users' electricity consumption characteristics. By clustering the daily load curves of residents, analyzing the overall energy consumption type and electricity consumption characteristics, and analyzing users' electricity consumption behavior, specific information on the electricity consumption of each appliance within the user's premises and detailed electricity consumption behavior can be obtained.

[0003] Most existing load identification solutions are intrusive, requiring upgrades to the intrusive devices used in these solutions. Measuring devices must be installed at the point of entry into the home to obtain information on the operation of each device. A few non-intrusive load identification solutions still require the installation of load monitoring devices at the user's power inlet and the collection of various characteristics of the total load through additional wiring to analyze the status of each device. However, current technologies cannot achieve real-time and accurate load identification based on the collected power consumption information. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, integrated fusion terminal, storage medium, and processor for non-intrusive load identification of electricity consumption information.

[0005] To achieve the above objectives, a first aspect of this application provides a method for non-intrusive load identification of electricity consumption information, comprising: Obtain the power consumption information of the total power port corresponding to multiple electrical devices at various time points within a preset time period. The power consumption information includes the current and voltage of the total power port at various time points within the preset time period. Process electricity consumption information to determine the load characteristics of the electricity consumption information, including the effective value of the current; A first load curve is generated based on the characteristic values ​​of each time point within a preset time period according to the load characteristics. The first load curve is clustered and classified based on a preset clustering algorithm to obtain the merged second load curve; The second load curve is analyzed to determine the identification results of start-stop events of multiple electrical devices within a preset time period. The identification results include the total number of start-stop events of multiple electrical devices within the preset time period and the time of each start-stop event. The identification results are matched with the load feature database to determine the type of user equipment for each start-stop event, as well as the number of start-stop events and the time of each type of equipment within a preset time period.

[0006] In this embodiment of the application, the load characteristics also include current, voltage, power and harmonics. Generating a first load curve based on the characteristic values ​​of the load characteristics at each time point within a preset time period includes: normalizing the current, voltage, power and harmonics at each time point within the preset time period; comparing the characteristic values ​​corresponding to each time point within the preset time period for each load characteristic after normalization to determine the optimal characteristic among the load characteristics; and generating a first load curve based on the characteristic values ​​of the optimal characteristic at each time point within the preset time period.

[0007] In this embodiment of the application, the method further includes: after obtaining the power consumption information of the total power consumption port corresponding to multiple electrical devices at various time points within a preset time period, determining the power consumption parameters of the total power consumption port within the preset time period based on the power consumption information. The power consumption parameters include at least one of the following: effective current value, event occurrence time, event end time, and change in effective current value. The event occurrence time refers to the time when the current first reaches a preset current threshold within the preset time period, and the event end time refers to the time when the current first falls below the preset current threshold within the preset time period. A data recording table is generated to record the power consumption parameters.

[0008] In this embodiment, clustering and classifying the first load curve based on a preset clustering algorithm to obtain the merged second load curve includes: processing the load time series of the first load curve using a DTW-based K-Medoids clustering algorithm; arbitrarily selecting n time series segments; using the n time series segments as center points Medoids; treating each of the n time series segments as an initial cluster; initializing the number of clusters to make L=n; processing each cluster to determine the DTW distance from other time series segments of the first load curve to the center points to obtain an initial distance matrix; storing the distance values ​​and corresponding element indices in the initial distance matrix into an array dist; and sorting the values ​​in the array dist according to distance... The values ​​are sorted in ascending order. The two subsequences corresponding to the first value in the array `dist` are merged into one cluster. The remaining subsequences are assigned to the classes represented by the center points. Starting from the second value in the array `dist`, the criterion function of each of the L clusters is calculated. The point corresponding to the minimum criterion function is selected and used as the new center point `Medoids`. After each cluster is processed, the number of clusters L is decremented by 1 until all elements in the array `dist` have been processed. The steps for processing the load time series of the first load curve are repeated until the selected center point no longer changes or the set maximum number of iterations is reached. The center point selected based on no longer changing or reaching the set maximum number of iterations is used as the main feature point after merging to generate the second load curve.

[0009] In this embodiment of the application, determining the DTW distance from other time series segments of the load curve to the center point includes calculating the DTW distance according to formula (1): (1) Where P and Q refer to any two time series segments selected from n time series segments, dp(i,j) is the minimum distance at position (i,j) in the two-dimensional matrix, which is the DTW distance, min() represents taking the minimum value among the three, and d(i,j) represents Point and Euclidean distance of a point.

[0010] In this embodiment of the application, the point corresponding to the minimum of the criterion function is selected according to formula (2): (2) Where n refers to n time series segments, k refers to any point selected from n, P and Q refer to any two time series segments selected from n time series segments, and E is the criterion function.

[0011] In a second aspect of the application, a processor is provided, configured to perform the aforementioned method for non-intrusive load identification of electricity consumption information.

[0012] A third aspect of this application provides an integrated fusion terminal, including a processor configured to perform the above-described method for non-intrusive load identification of electricity consumption information.

[0013] A fourth aspect of this application provides an apparatus for non-intrusive load identification of electricity consumption information, comprising: The data acquisition module is used to acquire the power consumption information of the total power port corresponding to multiple electrical devices at various time points within a preset time period. The power consumption information includes the current and voltage of the total power port at various time points within the preset time period. The data processing module is used to process electricity consumption information to determine the load characteristics of the electricity consumption information. The load characteristics include the effective value of the current. Based on the characteristic values ​​of the load characteristics at each time point within a preset time period, a first load curve is generated. The first load curve is clustered and classified based on a preset clustering algorithm to obtain a merged second load curve. The load identification module is used to analyze the second load curve to determine the identification results of multiple electrical devices starting and stopping events within a preset time period. The identification results include the total number of start and stop events of multiple electrical devices within the preset time period and the time of each start and stop event. The classification module is used to match the identification results with the load feature library to determine the type of user equipment for each start-stop event, as well as the number of start-stop events and the time of each type of equipment within a preset time period.

[0014] A fifth aspect of this application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the aforementioned method for non-intrusive load identification of electricity consumption information.

[0015] The above technical solution collects power consumption information at the main port of the electrical equipment and identifies the load based on a preset algorithm to obtain the operating status and condition of each electrical device. Compared with existing technologies, it does not require modification or upgrade of each electrical device, nor does it require additional circuit wiring. By installing the device at the main port of the electrical equipment, it can not only complete load identification in real time and accurately, but also quickly detect and grasp detailed information on power supply anomalies of electrical equipment, and promptly handle abnormal situations.

[0016] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the following detailed description to explain the embodiments of this application, but do not constitute a limitation on the embodiments of this application. In the drawings: Figure 1 This illustration schematically depicts an application scenario of a method for non-intrusive load identification of electricity consumption information according to an embodiment of this application. Figure 2 The schematic diagram illustrates a process flow diagram of a method for non-intrusive load identification of electricity consumption information according to an embodiment of this application; Figure 3 A schematic diagram illustrating a comparison between a first load curve and a second load curve according to an embodiment of this application is shown. Figure 4 A schematic diagram of a DTW-based K-Medoids clustering algorithm according to an embodiment of this application is shown. Figure 5 This schematic diagram illustrates a structural block diagram of a non-intrusive load identification power consumption information device according to an embodiment of this application; Figure 6 The diagram illustrates the internal structure of a computer device according to an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] In one embodiment, the method for non-intrusive load identification of electricity consumption information provided in this application can be applied to, for example... Figure 1 In the application environment shown, the electricity meter 103 is connected to the main port of the electrical equipment (not shown in the figure), and the device 102 for non-intrusive load identification of electricity consumption information can also be connected to the main port of the electrical equipment, so the device 102 can be connected to the electricity meter 103. Furthermore, the device 102 communicates with the electrical equipment 101 via a circuit. Electrical equipment is a general term for devices that use electricity as their energy source; the electrical equipment 101 can be, but is not limited to, devices that use electricity as their energy source. Figure 1 The device 102 is used to collect real-time and accurate power consumption information of electrical equipment, analyze and process the power consumption information of electrical equipment, and realize non-intrusive load identification of electrical equipment to obtain the operating status and condition of each electrical device.

[0020] Figure 2 A schematic flowchart illustrating a method for non-intrusive load identification of electricity consumption information according to an embodiment of this application is shown. Figure 2 As shown in one embodiment of this application, a method for non-intrusive load identification of electricity consumption information is provided, specifically including: Step 201: Obtain the power consumption information of the total power consumption port corresponding to multiple electrical devices at various time points within a preset time period. The power consumption information includes the current and voltage of the total power consumption port at various time points within the preset time period.

[0021] Step 202: Process the electricity consumption information to determine the load characteristics of the electricity consumption information, including the effective value of the current.

[0022] Step 203: Generate the first load curve based on the characteristic values ​​of the load characteristics at each time point within a preset time period.

[0023] Step 204: Cluster and classify the first load curve based on a preset clustering algorithm to obtain the merged second load curve.

[0024] Step 205: Analyze the second load curve to determine the identification results of start-stop events of multiple electrical devices within a preset time period. The identification results include the total number of start-stop events of multiple electrical devices within the preset time period and the time of each start-stop event.

[0025] Step 206: Match the identification results with the load feature database to determine the type of user equipment for each start-stop event, as well as the number of start-stop events and the time of each type of equipment within a preset time period.

[0026] Technical personnel can customize the preset time period and the individual time points within that period according to specific business needs. For example, the preset time period could be from 9:00 AM to 3:00 PM. Then, the individual time points within that preset time period could be specific times within that period, such as 9:00 AM, 10:00 AM, 11:00 AM, 12:00 PM, 1:00 PM, 2:00 PM, and 3:00 PM.

[0027] The processor can acquire power consumption information for multiple electrical devices at various points in time within a preset time period, corresponding to the total power port. This power consumption information includes the current and voltage of the total power port at each point in time within the preset time period. The processor can process this acquired power consumption information to determine the load characteristics. These load characteristics represent the power consumption features or changes of the electrical devices within the preset time period. Specifically, the load characteristics include the effective value of the current. The effective value of the current refers to the amount of heat generated by the two currents passing through the resistor of the same resistance value within the same time interval when both currents are passed through it. Furthermore, the effective value of the current is not the same as the average value of the current. The effective value of alternating current is defined based on the thermal effect of the current, while the average value of alternating current refers to the magnitude of the average current over a certain period. The effective value and the average value of alternating current are two different physical quantities. After obtaining the load characteristics of the power consumption information, the processor can generate a first load curve based on the characteristic values ​​of the load characteristics at various points in time within the preset time period. The load curve refers to the curve showing the change of various electrical loads in a power system over time. Load changes in a power system are random, involving various users across a vast area. Each user's electricity consumption is very different, and it is impossible to know in advance when, where, or what type of load will be added. Therefore, load characteristics, such as the effective current value at various time points, can be used to generate a primary load curve to more intuitively show the current changes of multiple electrical devices within a selected preset time period. In other words, the load curve can intuitively reflect the characteristics and patterns of user electricity consumption. Therefore, the load curve can serve as a basis for dispatching power and planning power systems. Specifically, in a load curve graph, the vertical axis represents the load, and the horizontal axis represents the corresponding time of the load. In this technical solution, the load curve can refer to the curve showing the change of the effective current value over time.

[0028] After obtaining the first load curve, the processor can cluster and classify it based on a preset clustering algorithm to obtain the merged second load curve. The preset clustering algorithm can be a K-Medoids clustering algorithm based on Dynamic Time Warping (DTW). DTW distance is used to measure the similarity between two sequences and is suitable for time series of different lengths and rhythms. When two time series are of equal length, we can use Euclidean distance to measure their similarity. However, when two time series are of unequal length, Euclidean distance is insufficient to measure their similarity. Therefore, DTW distance can be used to replace traditional Euclidean distance to describe the similarity between two time series. As a new similarity measurement method, DTW can find the optimal matching path between data in two arbitrary long-term series by adjusting the correspondence between time points, thus more effectively measuring the similarity of time series. The K-Medoids clustering algorithm, where Medoids represent the centroids, is the unique center of each cluster. Therefore, K-Medoids clustering is also known as K-centroid clustering, and it's an improvement on K-Means clustering. In the K-Means algorithm, initial centroids are randomly selected first. Only the first randomly selected initial centroids are the points in the actual cluster set. Subsequent assignments of non-centroids to corresponding centroids result in recalculated centroids that are not necessarily in the actual cluster set. Furthermore, if some non-centroids are outliers, the recalculated centroids may deviate from the overall cluster. To address this issue, the improved K-Medoids clustering algorithm was proposed. The biggest difference between K-Medoids and K-Means clustering algorithms is that K-Means uses the mean of the clusters as the centers, while K-Medoids, after finding the mean, selects the actual point closest to that mean as the center. In other words, K-Medoids uses the truly optimal point in the dataset as its center. K-Medoids has the advantages of being fast and having a fixed number of convergence steps.

[0029] Furthermore, the processor can analyze the second load curve to determine the identification results of start-stop events occurring in multiple electrical devices within a preset time period. The identification results include the total number of start-stop events occurring in the multiple electrical devices within the preset time period and the time of each start-stop event. A start-stop event can refer to an operational event that allows an electrical device to start or stop. Analyzing the second load curve can determine the number of times all electrical devices start or stop within the pre-selected preset time period, as well as the time of each start-stop event. At this point, the specific start-stop data for each electrical device is not yet available. For example, assuming the preset time period is from 9:00 AM to 3:00 PM, and the multiple electrical devices include a computer, washing machine, and air conditioner, then the identification results here refer to the total number of start-stop events occurring jointly by the computer, washing machine, and air conditioner within this time period, as well as the time of each start-stop event. At this point, it is impossible to determine the specific time point at which each electrical device starts or stops.

[0030] Therefore, the identification results can be further matched with the load characteristic database to determine the type of user equipment that experiences each start-stop event, as well as the number and timing of start-stop events for each equipment type within a preset time period. After matching, the specific start-stop status of each equipment type can be determined. For example, after matching the identification results with the load characteristic database, the start-stop status of each electrical device can be determined between 9:00 AM and 3:00 PM. For instance, the start-stop status of a computer is: 10:00 AM on, 10:30 AM off, 10:50 AM on, and 1:00 PM off. The start-stop status of a washing machine is: 10:10 AM on and 10:40 AM off. The start-stop status of an air conditioner is: 10:20 AM on and 2:00 PM off.

[0031] The above technical solution collects real-time power consumption information from the main ports of electrical equipment and clusters and classifies the load curves based on a preset clustering algorithm. This allows for accurate determination of the number of starts and stops for each piece of electrical equipment during the power consumption period, as well as the specific time of each start / stop event. This technical solution enables real-time and accurate load identification based on the collected power consumption information.

[0032] In one embodiment, the load characteristics further include current, voltage, power, and harmonics. Generating a first load curve based on the characteristic values ​​of the load characteristics at each time point within a preset time period includes: normalizing the current, voltage, power, and harmonics at each time point within the preset time period; comparing the characteristic values ​​corresponding to each time point within the preset time period for each processed load characteristic to determine the optimal characteristic among the load characteristics; and generating a first load curve based on the characteristic values ​​of the optimal characteristic at each time point within the preset time period.

[0033] The processor processes the acquired electricity consumption information to determine its load characteristics. These characteristics include not only the RMS current value but also current, voltage, power, and harmonics. Harmonics refer to the electrical quantities in the current whose frequencies are integer multiples of the fundamental frequency. Generally, this refers to the electrical quantities generated by currents with frequencies higher than the fundamental frequency after performing Fourier series decomposition on periodic non-sinusoidal electrical quantities. After obtaining the current, voltage, power, and harmonics at various time points within a preset time period, the processor normalizes these parameters. Normalization involves limiting the processed data (using an algorithm) to a specific range, aiming to standardize the statistical distribution of the samples. Data standardization involves scaling the data proportionally to fit it into a small, specific interval. The processor compares the processed power, harmonics, and RMS current data to determine the optimal characteristics and generates a first load curve based on these optimal characteristics at various time points within the preset time period. The optimal characteristics are those that best represent the electricity consumption patterns or changes of the electrical equipment within the preset time period.

[0034] In one embodiment, such as Figure 3 As shown, a schematic comparison diagram of a first load curve and a second load curve according to an embodiment of this application is illustrated. Figure 3 As shown in one embodiment of this application, a graph of the first load curve changing over time is provided, specifically including: After normalizing the current, voltage, power, and harmonics, the processor compares the obtained current, voltage, power, and harmonic data with the RMS current data to determine the optimal feature as the RMS current. If the optimal feature is the RMS current, the processor then generates data such as... Figure 3 The effective value curve of phase A current in the figure.

[0035] In one embodiment, the method further includes: after acquiring the power consumption information of the total power consumption port corresponding to multiple electrical devices at various time points within a preset time period, determining the power consumption parameters of the total power consumption port within the preset time period based on the power consumption information. The power consumption parameters include at least one of the following: effective current value, event occurrence time, event end time, and change in effective current value. The event occurrence time refers to the time when the current first reaches a preset current threshold within the preset time period, and the event end time refers to the time when the current first falls below the preset current threshold within the preset time period. A data recording table is generated to record the power consumption parameters.

[0036] The processor can acquire power consumption information for multiple electrical devices at various points in time within a preset time period, corresponding to the total power port. This power consumption information includes not only the current and voltage at each point in time within the preset time period, but also the effective value of the current, the event occurrence time, the event end time, and the change in the effective value of the current. Technicians can select the required power consumption information from these current-related parameters based on specific business needs, and generate a data recording table to record the data. The event occurrence time refers to the time when the current first reaches a preset current threshold within the preset time period, and the event end time refers to the time when the current first falls below the preset current threshold within the preset time period. An event occurrence can refer to the electrical device just starting up, and an event end can refer to the electrical device operating smoothly. The current threshold can refer to the rated current, which represents the ideal current operating state. When electrical equipment starts, a starting current is generated, also known as inrush current. The starting current is the change in current during the brief period from the moment the motor or inductive load is energized to when it reaches stable operation. This current is typically 4 to 7 times the rated current. When the equipment is running smoothly, the current is the operating current. Under normal circumstances, due to various external factors and equipment-specific factors, the actual operating current cannot reach the rated current operating state; therefore, the actual current is usually smaller than the rated current.

[0037] In one embodiment, a user load change information table is generated based on the user number, data recording time, event start time, event end time, phase sequence, change in effective current value, and change in fundamental current amplitude to record power consumption parameters, as shown in Table (1):

[0038] Table (1) In one embodiment, such as Figure 4 As shown, a schematic diagram of the DTW-based K-Medoids clustering algorithm according to this application is illustrated. Figure 4 As shown, in one embodiment of this application, a K-Medoids clustering algorithm based on DTW is provided, specifically including: S401, segmented load time series and number of segments n.

[0039] S402, initialize the number of clusters L=n.

[0040] S403, calculate DTW distance.

[0041] S404 sorts the array dist in ascending order.

[0042] S405, process the cluster.

[0043] S406, determine if the number of clusters L is greater than 1. If yes, execute S401; otherwise, execute S407.

[0044] S407, Select the center point.

[0045] S408 generates the second load curve.

[0046] The processor uses the DTW-based K-Medoids clustering algorithm to process the load time series of the first load curve. It arbitrarily selects n time series segments, using these segments as center points (Medoids), and treats each of these n segments as an initial cluster. The number of clusters is initialized to L=n. Each cluster is then processed to determine the DTW distance from other time series segments of the first load curve to the center points, resulting in an initial distance matrix. The distance values ​​and corresponding element indices from the initial distance matrix are stored in the array `dist`, and the values ​​in `dist` are sorted in ascending order by distance. The two subsequences corresponding to the first value in `dist` are merged into one cluster, and the remaining subsequences are assigned to the clusters represented by the center points. Starting from the second value in `dist`, the criterion function for each of the L clusters is calculated, and the point corresponding to the minimum criterion function is selected as the new center point (Medoids). After processing each cluster, the number of clusters L is decremented by 1 until all elements in the array `dist` have been processed. Repeat the steps of processing the load time series of the first load curve until the selected center point no longer changes or the set maximum number of iterations is reached. The center point selected based on no longer changing or reaching the set maximum number of iterations is used as the main feature point after merging to generate the second load curve.

[0047] The K-Medoids clustering algorithm based on DTW distance metric can cluster and classify load curves, which can improve the accuracy of similarity measurement algorithm and reduce the time complexity of the algorithm.

[0048] In one embodiment, such as Figure 3 As shown, a schematic comparison diagram of a first load curve and a second load curve according to an embodiment of this application is illustrated. Figure 3 As shown in one embodiment of this application, a graph of the first load curve and the second load curve changing over time is provided, specifically including: After normalizing the current, voltage, power, and harmonics, the processor compares the obtained current, voltage, power, and harmonic data with the current RMS data, determines the optimal feature as the current RMS value, and generates data such as... based on the current RMS value at various time points within a preset time period. Figure 3The processor then applies a K-Medoids clustering algorithm based on the DTW distance metric to the effective value curve of the phase A current, resulting in the following: Figure 3 The combined current RMS curve is shown in the figure.

[0049] In one embodiment, determining the DTW distance from other time series segments of the load curve to the center point includes calculating the DTW distance according to formula (1): (1) Where P and Q refer to any two time series segments selected from n time series segments, dp(i,j) is the minimum distance at position (i,j) in the two-dimensional matrix, which is the DTW distance, min() represents taking the minimum value among the three, and d(i,j) represents Point and Euclidean distance of a point.

[0050] In one embodiment, the point corresponding to the minimum of the criterion function is selected according to formula (2): (2) Where n refers to n time series segments, k refers to any point selected from n, P and Q refer to any two time series segments selected from n time series segments, and E is the criterion function.

[0051] In one embodiment, such as Figure 5 As shown, a device 500 for non-intrusive load identification of electricity consumption information is provided, including a data acquisition module 502, a data processing module 504, a load identification module 506, and a classification module 508, wherein: Data acquisition module 502 is used to acquire power consumption information of the total power port corresponding to multiple electrical devices at various time points within a preset time period. The power consumption information includes the current and voltage of the total power port at various time points within the preset time period. The data processing module 504 is used to process electricity consumption information to determine the load characteristics of the electricity consumption information. The load characteristics include the effective value of the current. Based on the characteristic values ​​of the load characteristics at each time point within a preset time period, a first load curve is generated. The first load curve is clustered and classified based on a preset clustering algorithm to obtain a merged second load curve. The load identification module 506 is used to analyze the second load curve to determine the identification results of multiple electrical devices starting and stopping events within a preset time period. The identification results include the total number of start and stop events of multiple electrical devices within the preset time period and the time of each start and stop event. The classification module 508 is used to match the identification results with the load feature library to determine the equipment type of the user equipment for each start-up and shutdown event, as well as the number of start-up and shutdown events and the time of each equipment type within a preset time period.

[0052] The device for non-intrusive load identification of electricity information includes a processor and a memory. The data acquisition module and the load identification module are stored in the memory as program units, and the processor executes the program modules stored in the memory to implement the corresponding functions.

[0053] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters enables methods for non-intrusive load identification of electricity consumption information.

[0054] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0055] This application provides a storage medium storing a program that, when executed by a processor, implements the above-described method for non-intrusive load identification of electricity consumption information.

[0056] This application provides a processor for running a program, wherein the program executes the above-described method for non-intrusive load identification of electricity consumption information.

[0057] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6As shown in the figure, the computer device includes a processor A01, a network interface A02, a display screen A04, an input device A05, and a memory (not shown) connected via a system bus. The processor A01 provides computing and control capabilities. The memory includes internal memory A03 and a non-volatile storage medium A06. The non-volatile storage medium A06 stores an operating system B01 and a computer program B02. The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 stored in the non-volatile storage medium A06. The network interface A02 is used for communication with external terminals via a network connection. When the computer program is executed by the processor A01, it implements a method for non-intrusive load identification of electricity consumption information. The display screen A04 can be a liquid crystal display (LCD) or an e-ink display. The input device A05 can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0058] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0059] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform a program having method steps for non-intrusive load identification of electricity consumption information.

[0060] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0061] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0062] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0063] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0064] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0065] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0066] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0067] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0068] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for non-intrusive load identification of electricity consumption information, characterized in that, include: Obtain the power consumption information of the total power port corresponding to multiple electrical devices at various time points within a preset time period. The power consumption information includes the current and voltage of the total power port at various time points within the preset time period. The electricity consumption information is processed to determine the load characteristics of the electricity consumption information, the load characteristics including the effective value of the current; A first load curve is generated based on the characteristic values ​​of the load characteristics at each time point within the preset time period. The first load curve is clustered and classified based on a preset clustering algorithm to obtain the merged second load curve; The second load curve is analyzed to determine the identification results of start-stop events of the plurality of electrical devices within the preset time period. The identification results include the total number of start-stop events of the plurality of electrical devices within the preset time period and the time of each start-stop event. The identification results are matched with the load feature database to determine the type of user equipment for each start-stop event, as well as the number of start-stop events and the time of each type of equipment within the preset time period. The step of clustering and classifying the first load curve based on a preset clustering algorithm to obtain the merged second load curve includes: The load time series of the first load curve is processed using the K-Medoids clustering algorithm based on DTW. Arbitrarily select n time series segments of the time series, take the n time series segments as center points Medoids, and regard each of the n time series segments as an initial cluster; Initialize the number of clusters to make the number of clusters L=n, and process each of the clusters; Determine the DTW distances from other time series segments of the first load curve to the center point to obtain an initial distance matrix; Store the distance values ​​and corresponding element indices from the initial distance matrix into the array dist; Then sort the values ​​of the array dist in ascending order based on the distance value; The two time series segments corresponding to the first value in the array dist are merged into one cluster, and the remaining time series segments are assigned to the cluster represented by the center point. Starting from the second value in the array dist, calculate the criterion function for each of the L clusters; Select the point corresponding to the minimum of the criterion function, and use the point corresponding to the minimum as the new center point Medoids; After each cluster is processed, the number of clusters L is decremented by 1 until all elements in the array dist have been processed. Repeat the steps of processing the load time series of the first load curve until the selected center point no longer changes or the set maximum number of iterations is reached; The center point selected based on whether it no longer changes or reaches the set maximum number of iterations is used as the main feature point after merging to generate the second load curve.

2. The method for non-intrusive load identification of electricity consumption information according to claim 1, characterized in that, The load characteristics also include current, voltage, power, and harmonics. Generating a first load curve based on the characteristic values ​​of the load characteristics at various time points within the preset time period includes: The current, voltage, power and harmonics at each time point within the preset time period are normalized. The feature values ​​corresponding to each load feature at each time point within the preset time period are compared to determine the optimal feature among the load features. A first load curve is generated based on the feature values ​​at each time point within the preset time period according to the optimal features.

3. The method for non-intrusive load identification of electricity consumption information according to claim 1, characterized in that, The method further includes: After acquiring the power consumption information of the total power consumption port corresponding to multiple electrical devices at various time points within a preset time period, the power consumption parameters of the total power consumption port within the preset time period are determined based on the power consumption information. The power consumption parameters include at least one of the following: effective current value, event occurrence time, event end time, and change in effective current value. The event occurrence time refers to the time when the current first reaches a preset current threshold within the preset time period, and the event end time refers to the time when the current first falls below the preset current threshold within the preset time period. Generate a data logging table to record the power consumption parameters.

4. The method for non-intrusive load identification of electricity consumption information according to claim 1, characterized in that, Determining the DTW distance from other time series segments of the first load curve to the center point includes calculating the DTW distance according to formula (1): （1） Where P and Q refer to any two time series segments selected from the n time series segments, dp(i,j) refers to the minimum distance at position (i,j) in the two-dimensional matrix, which is the DTW distance, min() represents taking the minimum value among the three, and d(i,j) represents Point and Euclidean distance of a point.

5. The method for non-intrusive load identification of electricity consumption information according to claim 1, characterized in that, The points corresponding to the minimum selection criterion function include those selected according to formula (2): （2） Where n refers to the n time series segments, k refers to any point selected from n, P and Q refer to any two time series segments selected from the n time series segments, and E is the criterion function.

6. A processor, characterized in that, It is configured to perform the method for non-intrusive load identification of electricity consumption information as described in any one of claims 1 to 5.

7. An integrated fusion terminal, characterized in that, Includes the processor according to claim 6.

8. A device for non-intrusive load identification of electricity consumption information, characterized in that, include: The data acquisition module is used to acquire the power consumption information of the total power port corresponding to multiple electrical devices at various time points within a preset time period. The power consumption information includes the current and voltage of the total power port at various time points within the preset time period. A data processing module is used to process the electricity consumption information to determine the load characteristics of the electricity consumption information, the load characteristics including the effective value of current. Based on the load characteristics, a first load curve is generated at each time point within a preset time period. The first load curve is then clustered and classified using a preset clustering algorithm to obtain a merged second load curve. The clustering and classification of the first load curve using the preset clustering algorithm to obtain the merged second load curve includes: processing the load time series of the first load curve using a K-Medoids clustering algorithm based on DTW; arbitrarily selecting n time series segments, using the n time series segments as center points (Medoids), and considering each of the n time series segments as an initial cluster; initializing the number of clusters, such that the number of clusters L=n, and processing each cluster; determining the DT values ​​of other time series segments of the first load curve to the center points. The distance W is used to obtain an initial distance matrix; the distance values ​​and corresponding element indices in the initial distance matrix are stored in the array dist; the values ​​in the array dist are sorted in ascending order by distance value; the two time series segments corresponding to the first value in the array dist are merged into one cluster, and the remaining time series segments are assigned to the clusters represented by the center point; starting from the second value in the array dist, the criterion function of each of the L clusters is calculated; the point corresponding to the minimum of the criterion function is selected, and the point corresponding to the minimum is taken as the new center point Medoids; after each cluster is processed, the number of clusters L is decremented by 1 until all elements in the array dist are processed; the steps of processing the load time series of the first load curve are repeated until the selected center point no longer changes or the set maximum number of iterations is reached; the center point selected based on no longer changing or the set maximum number of iterations is taken as the main feature point after merging to generate the second load curve; The load identification module is used to analyze the second load curve to determine the identification results of start-stop events of the plurality of electrical devices within the preset time period. The identification results include the total number of start-stop events of the plurality of electrical devices within the preset time period and the time of each start-stop event. The classification module is used to match the identification results with the load feature library to determine the device type of the user equipment for each start-stop event, as well as the number of start-stop events and the time of each device type within the preset time period.

9. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, the instruction causes the processor to be configured to perform the method for non-intrusive load identification of electricity consumption information according to any one of claims 1 to 5.

Images