Event point determination method, transient interval determination method and load event identification method

Abstract

The invention relates to the field of non-intrusive electric appliance load detection, and discloses an event point determination method, a transient interval determination method and a load event identification method. The event point determination method comprises the following steps: caching data of a target cycle length in a current connection waveform of a target circuit by adopting a first group of panes, and if current data in the first group of panes meet a target condition, preliminarily determining that an electrical appliance access event occurs; caching data of a target cycle length in the current connection waveform by adopting a second group of panes, screening out pseudo event points corresponding to the electrical appliance access event from a plurality of sampling points contained in the second group of panes, and obtaining current data of a complete cycle length at positions spaced by a second target cycle length before and after each pseudo event point, and screening real event points from the pseudo event points. According to the invention, the current waveform data is detected through two groups of three-pane sliding, so that noise interference in a circuit can be effectively handled, and the detection accuracy of an electrical appliance access event point is improved.

Description

Technical Field

[0001] The present application relates to the field of non-invasive electrical load detection, and in particular to an event point determination method, a transient interval determination method, and a load event identification method. Background Art

[0002] With the rapid development of smart home and IoT technologies, non-intrusive load monitoring (NILM) has become a research hotspot in power system monitoring and energy management. This technology uses a single sensor installed at the power inlet to monitor and analyze the power usage of multiple appliances within a home or building, eliminating the need for individual wiring or modifications to each appliance. NILM primarily analyzes electrical signals such as power signatures, current waveforms, and switching events to identify the operating status of different appliances.

[0003] In practical applications, non-intrusive event point detection of appliance connections is a key step in achieving accurate load decomposition and power usage analysis. By capturing the sudden power changes or current waveform changes generated when an appliance switches on and off, the appliance's start and stop times and operating cycles can be determined. However, due to the wide variety of appliances, similar power characteristics, and noise interference, existing technologies often suffer from errors in event point detection, making it difficult to accurately determine the transient intervals during which the appliance switches, thus affecting the accuracy of subsequent load identification and energy consumption analysis. Summary of the Invention

[0004] In view of this, in order to solve the technical problem of low accuracy in detecting electrical appliance access event points in the prior art, the present application provides an event point determination method, a transient interval determination method, and a load event identification method.

[0005] In a first aspect, the present application provides a method for determining an event point, comprising: Using a plurality of first groups of panes to cache data of a target cycle length in a current connection waveform of a target circuit, if the current data in any one or more of the first groups of panes meets a target condition, it is preliminarily determined that an electrical appliance connection event has occurred in the target circuit; Using a plurality of second groups of panes to cache data of a target cycle length in the current connection waveform, and filtering out pseudo event points corresponding to the electrical appliance connection event from a plurality of sampling points contained in each of the second groups of panes; Taking the pseudo event points as a reference, in the current continuous waveform, current data of a complete cycle length at positions separated by a second target cycle length before and after each of the pseudo event points is obtained. Based on the current data of each complete cycle length, a real event point corresponding to the electrical appliance connection event is screened out from each of the pseudo event points.

[0006] In an optional embodiment, each of the first group of panes includes first to third panes of the same size; The starting position and the ending position of each of the second group of panes are the starting position and the ending position corresponding to the first group of panes after the first group of panes is moved backward by a unit distance of one step in the current connection waveform; each of the second group of panes includes fourth to sixth panes of the same size; Each of the window panes contains current data of a plurality of sampling points.

[0007] In an optional embodiment, the target condition is: if the average current value of the third pane is the maximum value among the average current values ​​corresponding to the panes in the first group of panes; The average current value of each window pane is a value obtained by dividing the sum of the absolute values ​​of the current data of one cycle length in the corresponding window pane by the wavelength.

[0008] In an optional embodiment, for any of the second group of panes, filtering out pseudo event points from a plurality of sampling points included in the second group of panes includes: calculating an adjustment coefficient based on a ratio between an average current value of the sixth window pane and an average current value of the fourth window pane; Starting from the first sampling point in the fifth window pane, sequentially calculating the cumulative sum of the absolute values ​​of the currents of all sampling points between the first sampling point and the current sampling point, and then calculating the average value of the cumulative sums; If the average value of the accumulated sum is greater than the product of the adjustment coefficient and the average current value of the fourth window pane, the current sampling point is determined to be a pseudo event point.

[0009] In an optional embodiment, the filtering out a real event point from each of the pseudo event points based on the current data of each complete cycle length includes: In the current connection waveform, current data of a complete cycle length is intercepted at a first target distance before and after the pseudo event point to obtain two current cycle data; Obtaining the maximum absolute value of the current value at each sampling point from the two current cycle data to obtain two maximum current absolute values; If the difference between the two maximum current absolute values ​​is greater than a preset threshold, the pseudo event point is determined to be a true event point.

[0010] In a second aspect, the present application provides a method for determining a transient interval, comprising: In the current connection waveform of the target circuit, a steady-state current waveform of a full cycle length is intercepted at a second target distance before and after the true event point; the true event point is determined based on the aforementioned event point determination method; Phase alignment is performed based on each of the steady-state current waveforms, and a start point and an end point of a transient interval are determined according to the aligned steady-state current waveforms.

[0011] In an optional embodiment, if the steady-state current waveform is a previous steady-state current waveform before the actual event point in the current continuous waveform; performing phase alignment based on each of the steady-state current waveforms and determining the starting point of the transient interval according to the aligned steady-state current waveform includes: In the voltage continuous waveform of the target circuit, taking the actual event point as a reference point, obtaining a first transient voltage waveform before the actual event point and including different voltage phases, and obtaining a first steady-state voltage waveform at a corresponding moment of the previous steady-state current waveform; aligning the phases of the first steady-state voltage waveform and the first transient voltage waveform to obtain a first identical voltage phase; Obtaining a first current waveform at a time corresponding to the first same voltage phase from the current continuous waveform; subtracting the previous steady-state current waveform from the first current waveform to obtain a first new current waveform; The first point in the first new current waveform that deviates from the zero axis by a target proportional value is obtained from left to right as the starting point of the transient interval.

[0012] In an optional embodiment, if the steady-state current waveform is a post-steady-state current waveform after the real event point in the current continuous waveform; performing phase alignment based on each of the steady-state current waveforms and determining the end point of the transient interval according to the aligned steady-state current waveform includes: In the voltage continuous waveform of the target circuit, taking the actual event point as a reference point, obtaining a second transient voltage waveform before the actual event point and including a different voltage phase, and obtaining a second steady-state voltage waveform at a corresponding moment of the post-steady-state current waveform; aligning the phases of the second steady-state voltage waveform and the second transient voltage waveform to obtain a second identical voltage phase; Obtaining a second current waveform at a time corresponding to the second same voltage phase from the continuous current waveform; subtracting the post-steady-state current waveform from the second current waveform to obtain a second new current waveform; The first point in the second new current waveform that deviates from the zero axis by a target proportional value is obtained from right to left as the end point of the transient interval.

[0013] In a third aspect, the present application provides a load event identification method, comprising: Extracting transient features according to a transient interval; the transient interval is determined based on the aforementioned transient interval determination method; The start and stop states of each electrical appliance in the target circuit are determined according to the transient characteristics.

[0014] In a fourth aspect, the present application provides a computer device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is used to execute the computer program to implement the event point determination method before implementation, or the aforementioned transient interval determination method, or the aforementioned load event identification method.

[0015] The embodiments of the present application have the following beneficial effects: An embodiment of the present application provides a method for determining an event point. This method can effectively cope with noise interference in the circuit by slidingly detecting current waveform data using two groups of three panes, thereby more accurately determining the specific moment when the appliance is connected to the target circuit (i.e., the event point). This improves the detection accuracy of the appliance connection event point, facilitates the subsequent effective elimination of noise interference and accurate positioning of the transient interval of appliance connection based on the event point, and thus improves the accuracy of load identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solution of this application, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of this application and should not be considered as limiting the scope of protection of this application. Those skilled in the art can also derive other relevant drawings based on these drawings without inventive effort.

[0017] Figure 1 A schematic diagram of a flow chart of a method for determining an event point in an embodiment of the present application is shown; Figure 2 A schematic diagram showing a principle logic of a method for determining an event point in an embodiment of the present application is shown; Figure 3 A schematic flow chart of a method for determining a transient interval in an embodiment of the present application is shown; Figure 4a A waveform diagram of the front steady-state current waveform in an embodiment of the present application is shown; Figure 4b A waveform diagram of the post-steady-state current waveform in an embodiment of the present application is shown; Figure 5a A waveform diagram of a first current waveform in an embodiment of the present application is shown; Figure 5b shows a waveform diagram of the second current waveform in an embodiment of the present application; Figure 6a A waveform diagram of a first new current waveform in an embodiment of the present application is shown; Figure 6b shows a waveform diagram of a second new current waveform in an embodiment of the present application; Figure 7 A schematic diagram of a current waveform including a transient interval in an embodiment of the present application is shown; Figure 8 A flow chart of a load event identification method in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments.

[0019] The components of the embodiments of the present application generally described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed application, but rather merely represents selected embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort are within the scope of protection of the present application.

[0020] Hereinafter, the terms "including", "having" and their cognates, which may be used in various embodiments of the present application, are intended only to indicate specific features, numbers, steps, operations, elements, components or combinations of the foregoing items, and should not be understood as first excluding the existence of one or more other features, numbers, steps, operations, elements, components or combinations of the foregoing items or the possibility of adding one or more features, numbers, steps, operations, elements, components or combinations of the foregoing items.

[0021] Furthermore, the terms “first,” “second,” “third,” etc., are merely used for distinguishing descriptions and are not to be understood as indicating or implying relative importance.

[0022] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by those skilled in the art to which the various embodiments of the present application belong. The terms (such as those defined in generally used dictionaries) will be interpreted as having the same meaning as in the context of the relevant technical field and will not be interpreted as having an idealized meaning or an overly formal meaning unless clearly defined in the various embodiments of the present application.

[0023] The following describes some embodiments of the present application in detail with reference to the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments may be combined with each other.

[0024] The transient period refers to the duration of instantaneous changes in current, voltage, and other signals when power equipment is started, shut down, or switching between different states. This change typically contains a wealth of information, such as the magnitude and waveform characteristics of the starting current, the frequency components of the transient process, and the changing patterns of energy consumption. These characteristics can be used to distinguish different types of electrical equipment.

[0025] Load event identification (NILM) is a technology that decomposes the operating status and energy consumption of each electrical device by monitoring the power signals (such as voltage and current) on the bus.

[0026] Generally speaking, each device generates unique transient signals (such as inrush current, power factor changes, etc.) when it starts or shuts down. These signals can be regarded as the "fingerprint" of the device and are used to distinguish different types of devices. The transient interval usually corresponds to the startup or shutdown event of the device and can therefore be used as a timestamp to help determine the operating time of the device. In other words, the transient interval provides key feature information for non-invasive load identification, especially playing an important role in device fingerprinting, timestamp labeling, and fine-grained status monitoring. However, due to the challenges of noise interference, signal resolution limitations, and device diversity, it is often difficult to find the exact transient interval.

[0027] Based on this, an embodiment of the present application provides a method for determining an event point, which uses two sets of three-pane sliding detection current waveform data to determine the specific moment (i.e., the event point) when the electrical appliance is connected to the target circuit, thereby facilitating the subsequent effective elimination of noise interference and accurate positioning of the transient interval where the electrical appliance is connected based on the event point, thereby improving the accuracy of load identification.

[0028] like Figure 1 As shown, exemplarily, the event point determination method includes the following steps: S110, using a plurality of first groups of panes to cache data of a target cycle length in a current connection waveform of a target circuit, if the current data in any one or more of the first groups of panes meets a target condition, it is preliminarily determined that an electrical appliance connection event has occurred in the target circuit.

[0029] In this embodiment, real-time current, real-time voltage and other data in the monitoring target circuit are obtained to identify the electrical appliance connection event in the target circuit based on the waveform changes corresponding to the real-time current and real-time voltage.

[0030] For reference, use the sliding pane technology to collect the current and voltage data of the target circuit in real time. Figure 2Define the first group of panes and set the sliding step size. Each first group of panes includes the first, second, and third panes of equal size. Each pane contains current data from multiple sampling points. This first group of panes is used to preliminarily determine whether a new appliance has been connected, that is, to determine whether an appliance connection event has occurred in the target circuit.

[0031] The size of each window is generally 1 to 2 cycles. In a noisy scene, the window size is suitable to be larger. In this embodiment, the window size is 1 cycle, and a current cycle and a voltage cycle each contain 256 sampling points. Then, three windows contain data of 768 sampling points. The sliding step size is The size of a cycle is 128 sampling points of data.

[0032] In other words, each pane in the first group of panes is used to cache data for a continuous current waveform corresponding to one cycle of real-time current data (i.e., current data corresponding to 256 sampling points), and the continuous current waveforms cached in different first groups of panes do not overlap. The first first group of panes is slid backward by half a cycle (i.e., by one step size) to obtain the second first group of panes. That is, the starting or ending positions of two adjacent first group of panes are separated by half a cycle.

[0033] If the current data in at least one of the first group of panes meets the following target condition, it can be preliminarily determined that an electrical device has been connected to the target circuit, and the event point can be further determined. The target condition is: if the average current value of the third pane in the first group of panes is the maximum of the average current values ​​corresponding to each pane in the first group of panes. The average current value of each pane is the sum of the absolute values ​​of the current data for one cycle within the corresponding pane divided by the wavelength.

[0034] For example, in the first group of window panes (including window panes Window a, Window b, and Window c), the accumulated value of the current data of one cycle in a window pane is first calculated after taking the absolute value, and the value obtained by dividing the accumulated value by the wavelength is used as the average current value of the window pane, that is, the average current value of window pane c is the value obtained by taking the accumulated value of the current data of one cycle in the window pane after taking the absolute value and dividing it by the wavelength.

[0035] If the average current value of window c (i.e., I_mean c) is greater than the average current value of window a and greater than the average current value of window b (i.e., I_mean c>I_mean a&I_mean c>I_meanb), it means that an electrical appliance has been connected to the target circuit.

[0036] S120 , using a plurality of second groups of panes to cache data of a target cycle length in the current connection waveform, and filtering out pseudo event points corresponding to the appliance connection event from a plurality of sampling points included in each second group of panes.

[0037] After determining that an electrical appliance connection event has occurred, a second set of three panes is used to cache data on the target cycle length in the continuous current waveform. In the continuous current waveform, the starting position of the second set of panes lags behind the starting position of the first set of panes by one step length. That is, the starting position of the second set of panes should be the position of the first set of three panes moved back by one step length.

[0038] In other words, the starting and ending positions of each second group of panes are the starting and ending positions of the corresponding first group of panes after moving backward a unit distance of one step in the current connection waveform. Each second group of panes includes the fourth to sixth panes of the same size, and the size of each pane in the second group of panes is the same as the size of each pane in the first group of panes. In other words, the second group of panes can be viewed as the first group of panes moved backward a unit distance in the current connection waveform.

[0039] In this embodiment, the first set of panes is used to determine whether an appliance connection event has occurred, and the second set of panes is used to further determine the specific event point at which the appliance connection event occurred. In other words, the two sets of three panes are used to gradually analyze the changes in current data to accurately locate the event point and transient interval.

[0040] That is, after determining that the current in a pane in the first group of three panes has increased significantly, it is preliminarily determined that an electrical appliance connection event has occurred, and the second group of three panes is used for further confirmation. According to the current change law and combined with specific parameters, the actual event point is determined.

[0041] In one example, for any second group of panes, this embodiment first filters out pseudo event points from a plurality of sampling points included in any second group of panes, and then determines the real event points through the pseudo event points.

[0042] Furthermore, by comparing the current values ​​and proportional relationships between different panes in the second group of panes, the pseudo event point is further confirmed and it is determined whether the pseudo event point is a real event point.

[0043] In this embodiment, based on existing experience from multiple measurements, it can be found that the pseudo event point generally falls within the second pane of the second group of panes (i.e., the fifth pane). Based on the relationship between the current values ​​of each sampling point in the fifth pane and the current values ​​of other panes in the second group of panes, the pseudo event point can be screened out from the multiple sampling points in the fifth pane, and the real event point can be determined based on the pseudo event point.

[0044] An adjustment coefficient is calculated based on the ratio of the average current value of the sixth window pane to the average current value of the fourth window pane. Starting from the first sampling point in the fifth window pane, the cumulative sum of the absolute values ​​of the currents of all sampling points between the first sampling point and the current sampling point is calculated in sequence, and then the average of the cumulative sums is calculated. If the average of the cumulative sums is greater than the product of the adjustment coefficient and the average current value of the fourth window pane, the current sampling point is determined to be a pseudo event point.

[0045] For example, Figure 2 As shown, the three panes in the second group of panes are Window 1, Window 2, and Window 3. An adjustment coefficient α is set, which is a parameter determined according to the ratio between the average current value of each sampling point in pane Window 3 and the average current value of each sampling point in pane Window 1.

[0046] In one example, if the ratio of the average current value (denoted as I_mean3) of each sampling point in the sixth window (i.e., window 3) to the average current value (denoted as I_mean1) of each sampling point in the fourth window (i.e., window 1) is greater than a first threshold, the adjustment coefficient α is determined to be a first target value; if the ratio of the average current value of each sampling point in the sixth window to the average current value of each sampling point in the fourth window is greater than a second threshold, the adjustment coefficient α is determined to be a second target value; if the ratio of the average current value of each sampling point in the sixth window to the average current value of each sampling point in the fourth window is less than or equal to the second threshold, the adjustment coefficient α is determined to be a third target value; wherein the average current value of each window in the second group of windows is the value obtained by dividing the sum of the absolute values ​​of the current data of one cycle length in the corresponding window by the wavelength; the first threshold is greater than the second threshold; and the first target value, the second target value, and the third target value decrease in sequence.

[0047] Exemplarily, the first threshold value may be 1.3, the first target value may be 1.05, the second threshold value may be 1.1, the second target value may be 1.03, and the third target value may be 1. That is, if I_mean 3 > (1.3*I_mean 1), α is 1.05; if I_mean 3 > (1.1*I_mean 1), α is 1.03; otherwise, α is 1. In other words, if the ratio of the average current value of the sixth window pane to the average current value of the fourth window pane falls within the interval (1.3, +∞), the value of α is 1.05; if the ratio of the average current value of the sixth window pane to the average current value of the fourth window pane falls within the interval (1.1, 1.3), the value of α is 1.03; and if the ratio of the average current value of the sixth window pane to the average current value of the fourth window pane falls within the interval (0, 1.1), the value of α is 1.

[0048] Next, the cumulative sum of the absolute current values ​​from the first sampling point to the current sampling point in the fourth window is calculated, and the mean of these cumulative sums (denoted as add_mean) is calculated. Since the fourth window contains 256 sampling points, 256 cumulative sums are obtained, and the mean of these 256 cumulative sums is calculated.

[0049] Furthermore, if the mean of any accumulated sum is greater than the product of the adjustment coefficient α and the average current value of each sampling point in the fourth window pane (ie, add_mean>I_mean 1*α), the current sampling point is determined to be a pseudo event point.

[0050] For example, starting from the first sampling point in the fifth window pane, the average of the accumulated sum values ​​corresponding to the first sampling point is the absolute current value of the first sampling point. If the absolute current value of the first sampling point is less than or equal to the product of the adjustment coefficient and the average current value of all sampling points in the fourth window pane, the process moves to the second sampling point. The average of the accumulated sum values ​​corresponding to the second sampling point is the average of the sum of the absolute current values ​​of the first sampling point and the second sampling point. If this value is greater than the product of the adjustment coefficient and the average current value of all sampling points in the fourth window pane, the second sampling point can be determined to be a pseudo event point.

[0051] In short, starting from the first sampling point in the fifth window, the absolute values ​​of the currents are added and averaged sequentially until the cumulative average value of the current is greater than α times I_mean 1 (i.e. add_mean>I_mean 1*α). The position of this sampling point at this time is the pseudo-event point found (denoted as k').

[0052] S130, taking the pseudo event points as a reference, in the current continuous waveform, obtain current data of a complete cycle length at positions separated by a second target cycle length before and after each pseudo event point, and based on the current data of each complete cycle length, select a real event point corresponding to an appliance connection event from each pseudo event point.

[0053] Taking the pseudo event point as the reference point, the before and after steady-state waveforms can be obtained in the current continuous waveform, so as to finally confirm the real event point when the appliance is connected based on the before and after steady-state waveforms.

[0054] Among them, in the current connection waveform, the current data of a complete cycle length is intercepted at the first target distance before and after the pseudo event point, respectively, to obtain two current cycle data; the maximum absolute value of the current value of each sampling point is obtained from the two current cycle data to obtain two maximum current absolute values; if the difference between the two maximum current absolute values ​​is greater than the preset threshold (denoted as β), the pseudo event point is determined to be a real event point.

[0055] In one embodiment, two non-directly connected current cycle data points are captured at positions separated by half a cycle length (i.e., 128 points) before and after the pseudo-event point k'. The maximum absolute current value (denoted as |max1|) and the maximum absolute current value (denoted as |max2|) at each sampling point in these two current cycle data points are calculated. If |max2| - |max1| > β, the pseudo-event point k' is the true event point k. The specific value of β can be set according to actual needs. For example, the value range of β can be between 0.7 and 0.9 A.

[0056] In this embodiment, a set of three consecutive windows (Window a, Window b, and Window c) is first used to analyze the current data in the target circuit in real time. If the average current value of the third window (Window c) is significantly higher than that of the first two windows (Window a and Window b), it is considered that an appliance connection event has occurred. The current data is initially analyzed across three windows to filter out interference from background appliances and focus on the changes in the newly connected appliance. After initially detecting a possible appliance connection event, another set of three windows (Window 1, Window 2, and Window 3) is used for more precise analysis. This analysis identifies possible false event points based on the current variation patterns between the windows. Two current waveform segments are then taken before and after the false event point, and the difference between their maximum values ​​is compared to filter out true event points from false event points.

[0057] This embodiment uses two sets of three-pane detection (three sliding panes per set) to monitor current changes in real time. If the average current value of a pane is significantly higher than that of the other panes, it is considered a possible event point. The second set of three-pane detections is then used to further confirm the event point. Accuracy is improved by dynamically adjusting the threshold parameter α, effectively eliminating the influence of noise disturbances, thereby accurately confirming the state of the appliance connection event and locating the event point. Furthermore, the two-step three-pane detection method effectively eliminates the influence of noise disturbances, thereby precisely locating the event point. Compared to existing technologies that are susceptible to false detections due to interference from circuit noise waveforms, this application uses two-step three-pane detection to effectively eliminate the influence of noise disturbances and improve the accuracy of event point detection.

[0058] It should be noted that, in this embodiment, the process of determining the event point is implemented based on the current changes in a first group of panes and a second group of panes corresponding to the first group of panes (i.e., two three-pane detections); if no event point is detected in the first group of panes and the second group of panes, the event point can be searched from the next first group of panes and its corresponding second group of panes until one or more event points are finally found. The process of determining the event point is the same as that described above, so it will not be repeated here.

[0059] Furthermore, the determination of the event point can facilitate the subsequent determination of the transient interval. Furthermore, the present application provides a transient interval determination method. Based on the event point determined by the above-mentioned event point determination method, the steady-state current waveform before and after the real event point is used as a benchmark, and after removing the influence of the background electrical appliances and the steady-state part, the starting point and end point of the transient interval are determined, and then the transient interval range corresponding to the connection of the electrical appliance is determined from the current data.

[0060] For example, Figure 3 As shown, the transient interval determination method includes the following steps: S310 , in the current connection waveform of the target circuit, intercept a steady-state current waveform of a full cycle length at a second target distance before and after the actual event point.

[0061] S320 , performing phase alignment based on the steady-state current waveforms, and determining a start point and an end point of the transient interval according to the aligned steady-state current waveforms.

[0062] See also 3. Figure 4a and Figure 4b Based on the actual event point determined by the aforementioned event point determination method, the steady-state current waveform of the complete cycle length at the second target distance before and after the actual event point is respectively intercepted. In the current continuous waveform, the actual event point is used as a reference, and the steady-state current waveform intercepted before it is recorded as the pre-steady-state current waveform, and the steady-state current waveform intercepted after the actual event point is recorded as the post-steady-state current waveform.

[0063] To ensure consistency in the time base and eliminate noise interference caused by signal distortion or errors due to phase differences, the phases of the preceding and following steady-state current waveforms are aligned to determine the start and end points of the transient interval. Since the continuous current waveform is unstable, while the continuous voltage waveform is more stable, phase alignment can be performed based on the corresponding continuous voltage waveform at the same moment as the continuous current waveform.

[0064] Furthermore, the phase alignment of the pre-steady-state current waveform and the post-steady-state current waveform is performed based on the pre-steady-state voltage waveform and the post-steady-state voltage waveform at the same moment corresponding to the pre-steady-state current waveform and the post-steady-state current waveform, so as to determine the starting point and the end point of the transient interval; wherein the pre-steady-state current waveform is used to determine the starting point of the transient interval, and the post-steady-state current waveform is used to determine the end point of the transient interval.

[0065] In one embodiment, if the steady-state current waveform is a previous steady-state current waveform before the actual event point in the current continuous waveform, that is, for the previous steady-state current waveform, in the voltage continuous waveform of the target circuit, with the actual event point as a reference point, a first transient voltage waveform before the actual event point and including a different voltage phase is obtained, and a first steady-state voltage waveform at the same moment corresponding to the previous steady-state current waveform is obtained; the first steady-state voltage waveform is phase-aligned with the first transient voltage waveform to obtain a first identical voltage phase; as shown in FIG. Figure 5a As shown, a first current waveform at a moment corresponding to the first same voltage phase is obtained from the continuous current waveform.

[0066] like Figure 6a As shown, the first current waveform is subtracted from the previous steady-state current waveform to obtain a first new current waveform; the first point in the first new current waveform that deviates from the zero axis to the target proportional value is obtained from left to right as the starting point of the transient interval.

[0067] In one embodiment, if the steady-state current waveform is a post-steady-state current waveform after the real event point in the current continuous waveform, that is, for the post-steady-state current waveform, in the voltage continuous waveform of the target circuit, with the real event point as a reference point, a second transient voltage waveform before the real event point and including a different voltage phase is obtained, and a second steady-state voltage waveform corresponding to the post-steady-state current waveform is obtained; the second steady-state voltage waveform is phase-aligned with the second transient voltage waveform to obtain a second identical voltage phase; as shown in FIG. Figure 5b As shown, a second current waveform at a moment corresponding to the second same voltage phase is obtained from the continuous current waveform.

[0068] like Figure 6b As shown, the second current waveform is subtracted from the post-steady-state current waveform to obtain a second new current waveform. The first point in the second new current waveform that deviates from the zero axis by a target ratio value is obtained from right to left and is used as the end point of the transient interval. The value of the target ratio value is not limited here; for example, the target ratio value can be 0.95, 0.1, 0.15, etc.

[0069] like Figure 7 As shown, based on the starting point and end point of the transient interval, the transient interval (such as Figure 7 (shown as the dotted waveform in the figure).

[0070] It is worth noting that in order to obtain an accurate electrical transient interval, it is necessary to eliminate the influence of noise disturbance waveforms on event detection as much as possible and locate the electrical transient interval as accurately as possible. Furthermore, this embodiment uses event point k as a reference point on the continuous current waveform to acquire current waveform data for a complete cycle length of the preceding steady state and the following steady state (a cycle length of 256 points). For example, the preceding steady-state current waveform is intercepted 5 cycles forward of the event point, and the following steady-state current waveform is intercepted 300 cycles backward of the event point. The voltage waveforms corresponding to the preceding and following steady-state current waveforms are then aligned with the voltage waveforms before and after the event point k. The preceding and following steady-state current waveforms are then subtracted from the current waveforms that maintain the same voltage phase to obtain new current waveforms. The first point of the new current waveform corresponding to the preceding steady-state current waveform that deviates from the zero axis by 0.1 times the peak value of the preceding steady-state waveform from left to right is taken as the transient starting point, and the first point of the new current waveform that deviates from the zero axis by 0.1 times the peak value of the preceding steady-state waveform from right to left is taken as the transient end point. This determines the transient range of the electrical device connected this time.

[0071] In short, after determining event point k, this embodiment extracts the steady-state current waveforms before and after the appliance connection event, aligns these steady-state waveforms with the continuous waveforms before and after the event point, and performs a subtraction operation. Finally, the first point in the resulting new waveforms that deviates from the zero axis by a certain ratio (for example, 0.1 times the peak value of the previous steady-state waveform) is found as the starting and ending points of the transient.

[0072] Compared with the existing technology, which is often not accurate enough in determining the transient interval and is easily affected by the normal operation fluctuations of the electrical appliance, the embodiment of the present application uses the steady-state waveforms before and after the event point for comparative analysis, so that the transient interval can be determined efficiently and accurately. After finding the event point, a section of the steady-state current waveform before and after the event is extracted as a benchmark, and the steady-state voltage waveform before and after the event is aligned with the transient voltage waveform to remove the influence of the background electrical appliance and the steady-state part. Then, from the aligned new current waveform, the first point that deviates from the zero axis by a certain proportion is found, which is used as the transient starting point and end point respectively, so that the transient interval to which the electrical appliance is connected can be accurately determined.

[0073] The embodiment of the present application determines the event point twice using three panes and aligns the steady-state waveforms before and after the transient state of the event point to efficiently and accurately determine the transient interval of the appliance connection, thereby better helping the non-invasive identification algorithm to improve the load recognition accuracy, effectively removing the impact of background appliances in the non-invasive circuit on newly connected appliances, and effectively eliminating noise interference, ultimately obtaining a precise transient interval for appliance connection startup, solving the problems of noise interference and inaccurate transient intervals in the prior art. Furthermore, the embodiment of the present application is suitable for non-invasive appliance monitoring in various complex environments, and is particularly suitable for smart homes and energy management systems, providing higher-quality data input for subsequent load identification algorithms, helping the system better identify appliance types, and providing higher reliability and accuracy for non-invasive appliance monitoring technology.

[0074] Based on this, Figure 8 As shown, the embodiment of the present application further provides a load event identification method, which implements non-intrusive load identification through the transient interval determined in the above embodiment; for reference, the load event identification method includes the following steps: S810: Extract transient features according to the transient interval.

[0075] S820: Determine the start / stop status of each electrical appliance in the target circuit according to the transient characteristics.

[0076] The transient characteristics of the current signal are extracted from the transient interval, and the steady-state characteristics are extracted from the continuous current waveform. These characteristics are combined to identify non-intrusive loads when electrical appliances are connected, and thus determine the start and stop status of each appliance in the circuit. Transient characteristics are used to identify the start or operation status of an appliance, while steady-state characteristics are used to identify the dormant or off state of an appliance.

[0077] In one example, a corresponding feature processing model is obtained by training using machine learning or signal processing algorithms, and transient features and steady-state features are input into the feature processing model to output the start and stop status data of each electrical appliance, thereby obtaining the behavior of each electrical appliance.

[0078] The embodiment of the present application identifies non-intrusive loads through accurate transient intervals, effectively removing the impact of background appliances in the non-intrusive circuit on newly connected appliances, improving the recognition accuracy, and further effectively identifying the start and stop status of appliances, facilitating subsequent efficient and convenient appliance monitoring and management.

[0079] As a feasible implementation method, the embodiment of the present application further provides an event point determination device, which may include: an event occurrence monitoring module, configured to use a plurality of first groups of panes to cache data of a target cycle length in a current connection waveform of a target circuit, and preliminarily determine that an electrical appliance connection event has occurred in the target circuit if the current data in any one or more of the first groups of panes meets a target condition; a pseudo event point determination module, configured to use a plurality of second groups of panes to cache data of a target cycle length in the current connection waveform, and to filter out pseudo event points corresponding to the appliance connection event from a plurality of sampling points contained in each of the second groups of panes; The real event point determination module is used to obtain, in the current continuous waveform, current data of a complete cycle length at positions separated by a second target cycle length before and after each of the pseudo event points, based on the pseudo event points, and to select a real event point corresponding to the appliance connection event from each of the pseudo event points based on the current data of each complete cycle length.

[0080] It can be understood that the event point determination device of this embodiment corresponds to the event point determination method of the above embodiment, and the optional items in the above embodiment are also applicable to this embodiment, so they will not be described again here.

[0081] As an optional implementation manner, the embodiment of the present application further provides a transient interval determination device, which may include: a steady-state waveform acquisition module configured to capture a steady-state current waveform of a full cycle length at a second target distance before and after a true event point in the current connection waveform of the target circuit; the true event point being determined based on the event point determination method of the aforementioned embodiment; The transient interval determination module is configured to perform phase alignment based on the steady-state current waveforms, and determine a start point and an end point of the transient interval according to the aligned steady-state current waveforms.

[0082] It can be understood that the transient interval determination device of this embodiment corresponds to the transient interval determination method of the above embodiment, and the options in the above embodiment are also applicable to this embodiment, so they will not be described again here.

[0083] As an optional implementation, the embodiment of the present application further provides a load event identification device, which may include: A transient feature extraction module, configured to extract transient features based on a transient interval; the transient interval is determined based on the transient interval determination method in the aforementioned embodiment; The electrical appliance state determination module is used to determine the start and stop state of each electrical appliance in the target circuit according to the transient characteristics.

[0084] It can be understood that the load event identification device of this embodiment corresponds to the load event identification method of the above embodiment, and the optional items in the above embodiment are also applicable to this embodiment, so they will not be described again here.

[0085] The present application also provides a computer device. Exemplarily, the computer device includes a processor and a memory, wherein the memory stores a computer program, and the processor runs the computer program to enable the computer device to execute the above-mentioned event point determination method, or the above-mentioned transient interval determination method, or the above-mentioned load event identification method, or the functions of the various modules of the above-mentioned event point determination device, or the functions of the various modules of the above-mentioned transient interval determination device, or the functions of the various modules of the above-mentioned load event identification device.

[0086] The processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including at least one of a central processing unit (CPU), a graphics processing unit (GPU), a network processor (NP), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor, etc., and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0087] The memory may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The memory is used to store a computer program, and the processor may execute the computer program accordingly after receiving an execution instruction.

[0088] The present application also provides a computer storage medium for storing the computer program used in the above-mentioned computer device. The computer storage medium may be a readable storage medium, a non-volatile storage medium, or a volatile storage medium. For example, the computer storage medium may include, but is not limited to, various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0089] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and structure diagrams in the accompanying drawings show the possible architectures, functions and operations of the devices, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in an alternative implementation, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the structure diagram and / or flowchart, and the combination of boxes in the structure diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.

[0090] In addition, the functional modules or units in the various embodiments of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0091] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a smart phone, personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application.

[0092] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A method for determining an event point, characterized in that: include: Using a plurality of first groups of panes to cache data of a target cycle length in a current connection waveform of a target circuit, if the current data in any one or more of the first groups of panes meets a target condition, it is preliminarily determined that an electrical appliance connection event has occurred in the target circuit; Using a plurality of second groups of panes to cache data of a target cycle length in the current connection waveform, and filtering out pseudo event points corresponding to the electrical appliance connection event from a plurality of sampling points contained in each of the second groups of panes; Taking the pseudo event points as a reference, in the current continuous waveform, current data of a complete cycle length at positions separated by a second target cycle length before and after each of the pseudo event points is obtained. Based on the current data of each complete cycle length, a real event point corresponding to the electrical appliance connection event is screened out from each of the pseudo event points.

2. The event point determination method according to claim 1, characterized in that: Each of the first group of panes includes first to third panes of the same size; The starting position and the ending position of each of the second group of panes are the starting position and the ending position corresponding to the first group of panes after the first group of panes is moved backward by a unit distance of one step in the current connection waveform; each of the second group of panes includes fourth to sixth panes of the same size; Each of the window panes contains current data of a plurality of sampling points.

3. The event point determination method according to claim 2, characterized in that: The target condition is: if the average current value of the third pane is the maximum value among the average current values ​​corresponding to the panes in the first group of panes; The average current value of each window pane is a value obtained by dividing the sum of the absolute values ​​of the current data of one cycle length in the corresponding window pane by the wavelength.

4. The event point determination method according to claim 2 or 3, characterized in that: For any of the second group of panes, filtering out pseudo event points from a plurality of sampling points included in the second group of panes includes: calculating an adjustment coefficient based on a ratio between an average current value of the sixth window pane and an average current value of the fourth window pane; Starting from the first sampling point in the fifth window pane, sequentially calculating the cumulative sum of the absolute values ​​of the currents of all sampling points between the first sampling point and the current sampling point, and then calculating the average value of the cumulative sums; If the average value of the accumulated sum is greater than the product of the adjustment coefficient and the average current value of the fourth window pane, the current sampling point is determined to be a pseudo event point.

5. The event point determination method according to claim 1, characterized in that: The step of selecting a real event point from the pseudo event points based on the current data of each complete cycle length includes: In the current connection waveform, current data of a complete cycle length is intercepted at a first target distance before and after the pseudo event point to obtain two current cycle data; Obtaining the maximum absolute value of the current value at each sampling point from the two current cycle data to obtain two maximum current absolute values; If the difference between the two maximum current absolute values ​​is greater than a preset threshold, the pseudo event point is determined to be a true event point.

6. A method for determining a transient interval, characterized in that: include: In the current connection waveform of the target circuit, a steady-state current waveform of a full cycle length is intercepted at a second target distance before and after a real event point; the real event point is determined based on the event point determination method according to any one of claims 1 to 5; Phase alignment is performed based on each of the steady-state current waveforms, and a start point and an end point of a transient interval are determined according to the aligned steady-state current waveforms.

7. The method for determining a transient interval according to claim 6, wherein: If the steady-state current waveform is a previous steady-state current waveform before the actual event point in the current continuous waveform; The performing phase alignment based on the steady-state current waveforms and determining the starting point of the transient interval according to the aligned steady-state current waveforms includes: In the voltage continuous waveform of the target circuit, taking the actual event point as a reference point, obtaining a first transient voltage waveform before the actual event point and including different voltage phases, and obtaining a first steady-state voltage waveform at a corresponding moment of the previous steady-state current waveform; aligning the phases of the first steady-state voltage waveform and the first transient voltage waveform to obtain a first identical voltage phase; Obtaining a first current waveform at a time corresponding to the first same voltage phase from the current continuous waveform; subtracting the previous steady-state current waveform from the first current waveform to obtain a first new current waveform; The first point in the first new current waveform that deviates from the zero axis by a target proportional value is obtained from left to right as the starting point of the transient interval.

8. The method for determining a transient interval according to claim 6, wherein: If the steady-state current waveform is a post-steady-state current waveform after the real event point in the current continuous waveform; The performing phase alignment based on the steady-state current waveforms and determining the end point of the transient interval according to the aligned steady-state current waveforms includes: In the voltage continuous waveform of the target circuit, taking the actual event point as a reference point, obtaining a second transient voltage waveform before the actual event point and including a different voltage phase, and obtaining a second steady-state voltage waveform at a corresponding moment of the post-steady-state current waveform; aligning the phases of the second steady-state voltage waveform and the second transient voltage waveform to obtain a second identical voltage phase; Obtaining a second current waveform at a time corresponding to the second same voltage phase from the continuous current waveform; subtracting the post-steady-state current waveform from the second current waveform to obtain a second new current waveform; The first point in the second new current waveform that deviates from the zero axis by a target proportional value is obtained from right to left as the end point of the transient interval.

9. A load event identification method, characterized in that: include: Extract transient features based on transient intervals; The transient interval is determined based on the transient interval determination method according to any one of claims 6 to 8; The start and stop states of each electrical appliance in the target circuit are determined according to the transient characteristics.

10. A computer device, characterized in that: The computer device includes a processor and a memory, the memory stores a computer program, and the processor is used to execute the computer program to implement the event point determination method described in any one of claims 1 to 5, or the transient interval determination method described in any one of claims 6 to 8, or the load event identification method described in claim 9.

Images