Non-intrusive electric equipment monitoring method and system fusing high and low frequency data

Abstract

The invention discloses a non-intrusive electric equipment monitoring method and system fusing high and low frequency data, and relates to the technical field of building indoor electric system monitoring, and the method comprises the steps: carrying out the high-frequency sampling through a sliding window monitoring method, and collecting the current and voltage signals of a main loop of electric equipment in a plurality of monitoring periods in real time; for a time period comprising a plurality of monitoring periods, performing high-frequency trigger detection based on the current and voltage signals, and primarily judging whether an electric equipment switching event occurs or not; if not, low-frequency trigger detection is carried out again, and whether an electric equipment switching event occurs or not is judged for the second time; if the first judgment or the second judgment is yes, extracting a high-frequency sampling current and voltage signal in the time period; and extracting electrical signal characteristics of the current and voltage signals, matching and identifying typical and atypical characteristic models of the electric equipment based on the characteristics, and outputting an identification result of the electric equipment in which the switching event occurs currently. According to the invention, the accuracy of switching event occurrence and equipment identification can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of indoor power consumption system monitoring in buildings, and in particular to a non-intrusive power consumption equipment monitoring method and system integrating high-frequency and low-frequency data. Background Art

[0002] The operation monitoring and management of indoor power loads in traditional buildings has always been a difficult problem. With the gradual maturity of non-intrusive load monitoring (NILM) technology, it has broad application prospects in the fields of equipment energy consumption monitoring, management and optimization in office buildings, commercial buildings, homes and factories. Non-intrusive load monitoring technology monitors the operating electrical parameters of the line at the incoming line end of the distribution box to identify electrical equipment. NILM technology mainly consists of four processes: data acquisition, event detection, feature extraction, and load identification. Among them, data acquisition requires the installation of smart meters or electrical parameter acquisition modules to obtain power parameters. Event detection is the basis and key to device identification. It determines whether there is a device switching event by detecting the mutation point of the time series. However, this step can only determine whether there is a device switching event but cannot determine which device has been switched. For this reason, the specific switching device type is determined through subsequent feature extraction and load identification steps.

[0003] At present, the existing non-invasive monitoring and identification methods for indoor electrical equipment in buildings mostly collect single data for monitoring and identification. For example, the patent with application number CN202210779458 discloses a method for public building equipment management and energy optimization based on non-invasive identification, which only collects high-frequency data for equipment identification. However, high-frequency data can only analyze and process equipment operation curves for a shorter period of time, and is insensitive to equipment operation curves that change slowly and continuously over a long period of time, which will lead to missed detection problems for this type of equipment. If equipment identification is performed only based on low-frequency data, there will be obvious problems of missed detection, delay, low decomposition accuracy, and low detection accuracy of low-power equipment due to the large sampling interval, which will affect the subsequent equipment operation analysis and ultimately affect the accuracy of event detection and equipment identification. Summary of the invention

[0004] In order to solve the deficiencies of the above-mentioned prior art, the present invention provides a non-invasive electrical equipment monitoring method and system that integrates high- and low-frequency data. With respect to event detection, an event detection mechanism that integrates high-frequency triggering and low-frequency triggering conditions is used to compensate for the low accuracy of some equipment detection using single high-frequency data and the poor decomposition accuracy of single low-frequency data, thereby improving the accuracy of equipment start-stop event detection. With respect to electrical equipment identification, the electrical parameter operating characteristics of electrical equipment are divided into typical and atypical characteristics, thereby establishing a feature model library. By taking into account the diversity of feature models and combining the identification of the two feature data, the accuracy of the final electrical equipment identification result is ensured.

[0005] In a first aspect, the present disclosure provides a non-intrusive electrical equipment monitoring method that integrates high-frequency and low-frequency data.

[0006] A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data, comprising:

[0007] The sliding window monitoring method is used for high-frequency sampling to collect the current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time;

[0008] For a time period including multiple monitoring cycles, a high-frequency trigger detection is performed based on the current and voltage signals to initially determine whether an electrical equipment switching event has occurred; if the initial determination is no, a low-frequency trigger detection is performed based on the current and voltage signals to secondarily determine whether an electrical equipment switching event has occurred; if the initial determination or the second determination is yes, the high-frequency sampled current and voltage signals of the time period are extracted;

[0009] Based on high-frequency sampling of current and voltage signals, electrical signal characteristics of current and voltage signals are extracted;

[0010] Based on the current electrical signal characteristics, the typical and atypical feature models of the electrical equipment are matched and identified, and the identification results of the electrical equipment that is currently switching on and off are output.

[0011] According to a further technical solution, the electrical signal characteristics include numerical characteristics and waveform characteristics; the numerical characteristics include peak-to-peak values, absolute average values, effective values, and harmonics of current and voltage.

[0012] A further technical solution is to identify the electrical equipment where the switching event occurs, including:

[0013] The current electrical signal characteristics are matched one by one with the typical characteristic models of all devices in the known device model library. If the match is successful, the electrical device identification result is output; if the match fails, the non-typical characteristic models of all devices in the known device model library are matched one by one.

[0014] If the current electrical signal feature successfully matches the atypical feature module, the recognition result will be output after adding the atypical matching degree label; if the match fails, it will be identified as an unfamiliar electrical device and stored in the library to be identified.

[0015] A further technical solution is to record the number of occurrences of the current tag when adding an atypical matching degree tag, and calculate the atypical matching degree of the current recognition result based on the number of occurrences of the tag.

[0016] A further technical solution is to determine whether the atypical matching degree of the current recognition result exceeds the set value after adding the atypical matching degree label. If so, the recognition is considered successful and the recognition result of the electrical equipment without the label is directly output; otherwise, the recognition result of the electrical equipment marked with the atypical matching degree is output.

[0017] A further technical solution is to set an update cycle to periodically classify the electrical signal features of unfamiliar electrical equipment identified in the identification library, label all electrical signal features with the type of electrical equipment after classification, and update and store them in the equipment model library.

[0018] In a second aspect, the present disclosure provides a non-intrusive electrical equipment monitoring system that integrates high-frequency and low-frequency data.

[0019] A non-intrusive electrical equipment monitoring system integrating high and low frequency data, comprising:

[0020] A data acquisition module is used to perform high-frequency sampling using a sliding window monitoring method to collect current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time;

[0021] The switching event detection module is used to perform high-frequency trigger detection based on the current and voltage signals for a time period including multiple monitoring cycles, and initially determine whether a switching event of an electric device has occurred; if the initial determination is no, then perform low-frequency trigger detection based on the current and voltage signals, and secondarily determine whether a switching event of an electric device has occurred; if the initial determination or the second determination is yes, then extract the high-frequency sampled current and voltage signals of the time period;

[0022] A feature extraction module, used for extracting electrical signal features of current and voltage signals based on high-frequency sampling of current and voltage signals;

[0023] The electrical equipment identification module is used to match and identify typical and atypical feature models of electrical equipment based on the current electrical signal characteristics, and output the identification result of the electrical equipment that is currently experiencing the switching event.

[0024] In a third aspect, the present disclosure further provides an electronic device, comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the method described in the first aspect are completed.

[0025] In a fourth aspect, the present disclosure further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the method described in the first aspect.

[0026] One or more of the above technical solutions have the following beneficial effects:

[0027] 1. The present invention provides a non-invasive electrical equipment monitoring method and system that integrates high-frequency and low-frequency data. With respect to event detection, an event detection mechanism that integrates high-frequency triggering and low-frequency triggering conditions is used to compensate for the low accuracy of some equipment detection using single high-frequency data and the poor decomposition accuracy of single low-frequency data, thereby improving the accuracy of equipment start-stop event detection. With respect to electrical equipment identification, the electrical parameter operating characteristics of electrical equipment are divided into typical and atypical characteristics, thereby establishing a feature model library. By taking into account the diversity of feature models and combining the identification of two types of feature data, the accuracy of the final electrical equipment identification result is guaranteed.

[0028] 2. In the present invention, in the process of triggering and reporting equipment start and stop events, in addition to typical feature data, there are also atypical equipment switching feature data and unfamiliar equipment feature data. These feature data are temporarily stored in the to-be-identified library. After the atypical equipment switching feature data is identified by the recognition algorithm, the identification status and recognition result of the equipment are updated; at the same time, the feature data of the unfamiliar equipment is processed, uniformly classified and labeled, so as to facilitate the subsequent identification of the feature data of the electrical equipment; the above-mentioned identification method effectively identifies various types of feature data, which can effectively ensure the accuracy and comprehensiveness of the final identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0030] Figure 1 This is a flow chart of a non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data according to an embodiment of the present invention;

[0031] Figure 2 Schematic diagram of the slow change of current during the switching process of some electrical equipment in the embodiment of the present invention; (a) is the current change waveform when the microwave oven is turned on, and (b) is the current change waveform when the computer is turned off;

[0032] Figure 3This is a flowchart of a non-intrusive electrical equipment monitoring system that integrates high-frequency and low-frequency data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] It should be noted that the following detailed descriptions are exemplary only, are intended to describe specific embodiments, are intended to provide further explanation of the present invention, and are not intended to limit exemplary embodiments according to the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those of ordinary skill in the art to which the present invention belongs. In addition, it should also be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0034] Embodiment 1

[0035] This embodiment provides a non-intrusive electrical equipment monitoring method that integrates high-frequency and low-frequency data. Figure 1 As shown, the specific steps include:

[0036] The sliding window monitoring method is used for high-frequency sampling to collect the current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time;

[0037] For a time period including multiple monitoring cycles, a high-frequency trigger detection is performed based on the current and voltage signals to initially determine whether an electrical equipment switching event has occurred; if the initial determination is no, a low-frequency trigger detection is performed based on the current and voltage signals to secondarily determine whether an electrical equipment switching event has occurred; if the initial determination or the second determination is yes, the high-frequency sampled current and voltage signals of the time period are extracted;

[0038] Based on high-frequency sampling of current and voltage signals, electrical signal characteristics of current and voltage signals are extracted;

[0039] Based on the current electrical signal characteristics, the typical and atypical feature models of the electrical equipment are matched and identified, and the identification results of the electrical equipment that is currently switching on and off are output.

[0040] The following content introduces the non-intrusive electrical equipment monitoring method for integrating high-frequency and low-frequency data proposed in this embodiment in more detail.

[0041] Step S1, adopting the sliding window monitoring method to perform high frequency sampling, and collecting the current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time.

[0042] In this embodiment, a sliding window monitoring method is used in a high-frequency sampling manner to monitor and collect the current and voltage signals of the total circuit of the electrical equipment in real time, wherein the frequency of the high-frequency sampling is above 1kHz. The collected real-time current and voltage signals are stored in a memory, in which a high-frequency data buffer and a low-frequency data buffer are opened. The high-frequency data buffer temporarily stores the latest k cycles (k is not greater than 10 in this embodiment) of current and voltage waveforms, and the low-frequency data buffer is set up for calculating the low-frequency trigger condition, in which at least n cycles of current and voltage waveforms are temporarily stored.

[0043] Furthermore, through the sliding window monitoring method, a fixed-length window is set on the current and voltage signals. The window serves as a monitoring period and slides forward continuously over time to facilitate continuous monitoring and real-time analysis of the current and voltage signals within the window, i.e., the monitoring period.

[0044] Step S2: for a time period including multiple monitoring cycles, high-frequency trigger detection is performed based on the current and voltage signals to initially determine whether an electrical equipment switching event occurs; if the initial determination is no, low-frequency trigger detection is performed based on the current and voltage signals to secondary determine whether an electrical equipment switching event occurs; if the initial or secondary determination is yes, the high-frequency sampled current and voltage signals of the time period are extracted.

[0045] This embodiment proposes an event detection mechanism that integrates high-frequency triggering and low-frequency triggering conditions, wherein high-frequency triggering refers to judging, for the high-frequency sampled current and voltage signals of each monitoring cycle, whether the change ΔI of the effective value of the current in adjacent cycles is greater than a set threshold value. If so, it is judged that a switching event of the electrical equipment being turned on or the state being switched has occurred, and current and voltage signals of several cycles are captured at this time; otherwise, it is judged that a switching event of the electrical equipment being turned on or the state being switched has not occurred.

[0046] Specifically, the judgment conditions for high-frequency triggering are:

[0047] ΔI 1 =|I s -I s-1 |≥I th1 (1)

[0048] In the above formula, I s is the effective value of the current at the current moment, I s-1 is the effective value of the current at the previous moment, I th1 is the set threshold. The calculation formula of the effective value of current is as follows:

[0049]

[0050] In the above formula, I (i)is the instantaneous value of current, M is the number of sampling points in one cycle, and i is the serial number of the sampling point.

[0051] However, considering that for many electrical devices, due to the characteristics of the power circuit or working circuit, the change in the current of adjacent cycles when turned on or off is less than the set threshold, the current usually changes continuously and steadily, such as Figure 2 The slow current change process of some electrical equipment in a certain time period is shown, that is, the current change waveform within 5 cycles when (a) the microwave oven is turned on and (b) the computer is turned off. The switching events of this part of the electrical equipment have the problem of missed detection under the detection of high-frequency trigger conditions. Therefore, in order to improve the detection accuracy, this embodiment also sets a low-frequency trigger condition, with the high-frequency trigger condition as the primary judgment condition and the low-frequency trigger condition as the secondary judgment condition, so as to assist the accuracy of the judgment of the switching event and solve the problems of decomposition delay and missed detection based solely on low-frequency data and missed detection based solely on high-frequency data.

[0052] Low-frequency triggering means that during the monitoring process, a current effective value is calculated based on the sampling points of each cycle in the low-frequency data buffer at intervals of p cycles (in this embodiment, p=5), and q current effective values ​​are calculated continuously. These current effective values ​​are then compared to determine whether the current sequence changes continuously and stably and exceeds the set threshold. If so, it is determined that a switching event of the electrical equipment being turned on or the state being switched has occurred. At this time, the corresponding high-frequency sequence signal in the high-frequency data buffer is intercepted and uploaded.

[0053] Specifically, taking the low-frequency trigger condition for judging whether the device is “on” as an example, the j-th difference ΔI is calculated by the following formula (3): j , if equation (3) is satisfied, then the next difference ΔI of the subsequent time series is calculated j+1 , until the q consecutive differences ΔI are calculated j+q If both satisfy formula (3), it means that the current increases stably and continuously, and it is determined that the device is in the startup state, that is, a switching event occurs at this time.

[0054] ΔI j =I j -I j-1 ≥I th2 (3)

[0055] In the above formula, I j is the effective value of the current at the current moment, I j-1 is the effective value of the current before the interval p cycles, I th2 To set the threshold.

[0056] Furthermore, when the high-frequency trigger condition or the low-frequency trigger condition is met, it is determined that a switching event has occurred in the electrical equipment, and at this time, the latest current and voltage signal data stored in the current high-frequency data buffer is uploaded to the original database through any communication method.

[0057] As another implementation method, during the monitoring process, electrical data such as steady-state voltage, current, active power, reactive power, power factor, etc. are uploaded at fixed intervals (30 minutes in this embodiment). If an event occurs during the interval to trigger the upload, the steady-state data is also uploaded once and the interval timing is recalculated. The above-mentioned regularly uploaded steady-state data provides data support for subsequent equipment management.

[0058] In this embodiment, a database is provided in the cloud, and the current and voltage signal data are uploaded to the database. The database can be divided into an original database, a device model library, and a library to be identified. Among them, the original database is a real-time database that stores the uploaded high-frequency current and voltage signals; the device model library is used to store typical feature models and atypical feature models of known electrical equipment; when the data of the electrical equipment to be identified is uploaded, if the data cannot match the data feature model (including typical and atypical) in the device model library, the data is stored in the library to be identified, and similar data in the library to be identified can be uniformly processed and marked periodically.

[0059] Step S3: extracting electrical signal features of the current and voltage signals based on the high-frequency sampled current and voltage signals.

[0060] For the uploaded current and voltage signals, the electrical signal characteristics of the signal data are extracted, which include numerical characteristics and waveform characteristics. The numerical characteristics include the peak-to-peak value, absolute average value, effective value, harmonics, etc. of the current and voltage, as shown in Table 1 below. The calculation formulas for the proposed numerical characteristics are shown in Table 2 below.

[0061] Table 1 Symbols, names and meanings of various numerical characteristics

[0062]

[0063] Table 2 Calculation formulas for various numerical features

[0064]

[0065] In the above formula, i n is the instantaneous value of current, N is the number of sampling points in one cycle, and n is the sequence number of the sampling point.

[0066] Furthermore, the establishment of the above-mentioned device model library includes: at the initial stage of system operation, first establish a device model library for known electrical equipment. Models are established in the device model library according to device classification, and the model of each type of equipment is the electrical signal feature of the device. Since the event triggering of the device is not only once during the start-up or operation of the device, for example, when a device is turned on, it meets the event detection conditions multiple times, and multiple signal waveforms will be uploaded at this time. These waveforms are not completely consistent. A clustering algorithm is used to select the signal waveform with the most occurrences from multiple signal waveforms, and its characteristic value is calculated. The electrical signal feature composed of the characteristic value and the characteristic waveform is used as the typical characteristic model of the device, and the electrical signal features of other waveforms are used as atypical characteristic models. The device model library constructed here is the training data set for device identification using the algorithm described below.

[0067] By establishing a device model library in the above manner, a device has only one typical feature model, while there can be multiple atypical feature models. The establishment of typical and atypical feature models takes into account the fluctuation of equipment operating parameters due to complex on-site conditions, and takes into account the diversity of models, combining the recognition of the two feature data to ensure the accuracy of the final electrical equipment identification results.

[0068] Step S4: Based on the current electrical signal characteristics, typical and atypical characteristic models of the electrical equipment are matched and identified, and the identification result of the electrical equipment that currently has a switching event is output.

[0069] Specifically, first, the current electrical signal characteristics are matched one by one with the typical characteristic models of all devices in the known device model library. If the match is successful, the electrical device identification result is output and the device status is updated; otherwise, if the match fails, the non-typical characteristic models of all devices in the known device model library are continued to be matched one by one.

[0070] Secondly, if the current electrical signal feature successfully matches the atypical feature module, an atypical matching degree label is added to the recognition result, that is, a label of "atypical matching degree m" is added, and then the annotated recognition result is output, indicating that although the recognition is successfully matched, the recognition accuracy is not high. Among them, when adding the atypical matching degree label, the number of occurrences of the current label is recorded, and the atypical matching degree m of the current recognition result is calculated according to the number of occurrences of the label. The calculation formula is shown in the following formula (4):

[0071] m=in×100% (4)

[0072] In the above formula, i represents the number of times the label appears, i = 1, 2, ..., n, and n represents the total number of times. This parameter can be set manually.

[0073] Furthermore, after adding the atypical matching degree label as above, it is determined whether the atypical matching degree of the current recognition result exceeds the set value, that is, it is determined whether the number of times the label appears exceeds the set number (that is, the total number of times n set above). If it exceeds, it is considered that the recognition is successful, and the recognition result of the electrical equipment without label is directly output; otherwise, the recognition result of the electrical equipment marked with the atypical matching degree is output. In this embodiment, when the number of times the label of "atypical matching degree m" appears for the same type of equipment in the same working state is large, reaching n times, the matching degree m = 100%, and the recognition is considered successful at this time. After removing the marked label, the recognition result of the electrical equipment is output.

[0074] As another implementation method, this embodiment does not limit the specific device identification algorithm. It can use sequence matching algorithms, such as dynamic time warping (DTW) and cross-correlation algorithms for feature matching, or machine learning classification algorithms, such as K-nearest neighbor algorithm (KNN), to calculate and match feature values ​​such as peak-to-peak value, absolute average, effective value, harmonics, etc. It can also use complex pattern recognition algorithms, such as neural networks and deep learning, to extract high-level waveform features for device identification.

[0075] Finally, if the current electrical signal feature fails to match the atypical feature module, it is identified as an unfamiliar electrical device and stored in the library to be identified.

[0076] Furthermore, an update cycle is set to periodically classify and process the electrical signal features of unfamiliar electrical equipment in the identification library. After accumulating a certain amount of data, all classified electrical signal features are manually labeled with the type of electrical equipment, updated and stored in the equipment model library to facilitate subsequent classification and identification.

[0077] Through the above scheme, accurate detection of switching events of electric equipment is completed, and the electric equipment where the switching event occurs is identified.

[0078] Embodiment 2

[0079] This embodiment provides a non-intrusive electrical equipment monitoring system integrating high-frequency and low-frequency data, including:

[0080] A data acquisition module is used to perform high-frequency sampling using a sliding window monitoring method to collect current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time;

[0081] The switching event detection module is used to perform high-frequency trigger detection based on the current and voltage signals for a time period including multiple monitoring cycles, and initially determine whether a switching event of an electric device has occurred; if the initial determination is no, then perform low-frequency trigger detection based on the current and voltage signals, and secondarily determine whether a switching event of an electric device has occurred; if the initial determination or the second determination is yes, then extract the high-frequency sampled current and voltage signals of the time period;

[0082] A feature extraction module, used for extracting electrical signal features of current and voltage signals based on high-frequency sampling of current and voltage signals;

[0083] The electrical equipment identification module is used to match and identify typical and atypical feature models of electrical equipment based on the current electrical signal characteristics, and output the identification result of the electrical equipment that is currently experiencing the switching event.

[0084] like Figure 3 As shown, the architecture of the non-intrusive electrical equipment monitoring system that integrates high-frequency and low-frequency data proposed in this embodiment is a cloud-edge structure, which is mainly composed of an edge data acquisition module and a main system on the cloud.

[0085] Among them, the data acquisition module monitors and collects the real-time current and voltage of the circuit by high-frequency sampling, and the frequency of high-frequency sampling is above 1kHz; the switching event detection module has calculation and judgment functions, and judges whether the equipment has a start-up or state switching event based on high and low frequency data; when the event detection conditions are met, the high-frequency data is uploaded to the database of the cloud system through the communication module.

[0086] The system database is divided into the original database, the device model library and the library to be identified. The original database stores the high-frequency current and voltage waveforms uploaded by the data acquisition module after executing the event detection algorithm, and the device model library stores the typical feature models and atypical feature models for known devices. When the device data to be identified cannot be matched with the device in the device model library after being uploaded, it will be stored in the library to be identified. The system can then periodically process and mark similar data in the library to be identified.

[0087] The feature extraction module processes the feature data in the original database, the device feature library and the library to be identified. The electrical equipment identification module uses the typical equipment startup and state switching waveforms and features as the identification results. Other waveforms uploaded by the device are used to detect and update the identification results. Similar data in the library to be identified are classified and analyzed to determine whether it is a new electrical equipment.

[0088] Embodiment 3

[0089] This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are executed by the processor, the steps in the non-invasive electrical equipment monitoring method for fusing high and low frequency data as described above are completed.

[0090] Embodiment 4

[0091] This embodiment further provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps in the non-intrusive electrical equipment monitoring method for fusing high-frequency and low-frequency data as described above are completed.

[0092] The steps involved in the above embodiments 2 to 4 correspond to the method embodiment 1, and the specific implementation methods can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0093] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0094] The above description is only a preferred embodiment of the present invention. Although the specific implementation mode of the present invention is described in conjunction with the accompanying drawings, it is not a limitation of the protection scope of the present invention. Those skilled in the art should understand that on the basis of the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without creative work are still within the protection scope of the present invention.

Claims

1. A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data, characterized in that: include: The sliding window monitoring method is used for high-frequency sampling to collect the current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time; For a time period including multiple monitoring cycles, high-frequency trigger detection is performed based on current and voltage signals to initially determine whether an electrical equipment switching event occurs; If the initial judgment is no, a low-frequency trigger detection is performed based on the current and voltage signals, and a secondary judgment is made as to whether an electrical equipment switching event occurs; If the initial judgment or the secondary judgment is yes, then extract the high-frequency sampling current and voltage signals of the time period; Based on high-frequency sampling of current and voltage signals, electrical signal characteristics of current and voltage signals are extracted; Based on the current electrical signal characteristics, the typical and atypical feature models of the electrical equipment are matched and identified, and the identification results of the electrical equipment that is currently switching on and off are output.

2. A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data as claimed in claim 1, characterized in that: The electrical signal characteristics include numerical characteristics and waveform characteristics; the numerical characteristics include peak-to-peak values, absolute average values, effective values, and harmonics of current and voltage.

3. A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data as claimed in claim 1, characterized in that: Identify the electrical equipment where the switching event occurred, including: The current electrical signal characteristics are matched one by one with the typical characteristic models of all devices in the known device model library. If the match is successful, the electrical device identification result is output; if the match fails, the non-typical characteristic models of all devices in the known device model library are matched one by one. If the current electrical signal feature successfully matches the atypical feature module, the recognition result will be output after adding the atypical matching degree label; if the match fails, it will be identified as an unfamiliar electrical device and stored in the library to be identified.

4. A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data as claimed in claim 3, characterized in that: When adding an atypical matching degree label, the number of occurrences of the current label is recorded, and the atypical matching degree of the current recognition result is calculated based on the number of occurrences of the label.

5. A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data as claimed in claim 3, characterized in that: After adding the atypical matching degree label, determine whether the atypical matching degree of the current recognition result exceeds the set value. If so, the recognition is considered successful and the recognition result of the electrical equipment without the label is directly output; otherwise, the recognition result of the electrical equipment marked with the atypical matching degree is output.

6. A non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data as claimed in claim 3, characterized in that: Set an update cycle to periodically classify the electrical signal features of unfamiliar electrical equipment in the identification library. After classification, label all electrical signal features with the type of electrical equipment, update and store them in the equipment model library.

7. A non-intrusive electrical equipment monitoring system integrating high and low frequency data, characterized in that: include: A data acquisition module is used to perform high-frequency sampling using a sliding window monitoring method to collect current and voltage signals of multiple monitoring cycles of the total circuit of the electrical equipment in real time; A switching event detection module is used to perform high-frequency trigger detection based on current and voltage signals for a time period including multiple monitoring cycles, and initially determine whether a switching event of an electric device occurs; If the initial judgment is no, a low-frequency trigger detection is performed based on the current and voltage signals, and a secondary judgment is made as to whether an electrical equipment switching event occurs; If the initial judgment or the secondary judgment is yes, then extract the high-frequency sampling current and voltage signals of the time period; A feature extraction module, used for extracting electrical signal features of current and voltage signals based on high-frequency sampling of current and voltage signals; The electrical equipment identification module is used to match and identify typical and atypical feature models of electrical equipment based on the current electrical signal characteristics, and output the identification result of the electrical equipment that is currently experiencing the switching event.

8. A non-intrusive electrical equipment monitoring system integrating high-frequency and low-frequency data as claimed in claim 7, characterized in that: Identify the electrical equipment where the switching event occurred, including: The current electrical signal characteristics are matched one by one with the typical characteristic models of all devices in the known device model library. If the match is successful, the electrical device identification result is output; if the match fails, the non-typical characteristic models of all devices in the known device model library are matched one by one. If the current electrical signal feature successfully matches the atypical feature module, the recognition result will be output after adding the atypical matching degree label; if the match fails, it will be identified as an unfamiliar electrical device and stored in the library to be identified.

9. An electronic device, characterized in that: The invention comprises a memory and a processor and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of a non-intrusive electrical equipment monitoring method integrating high-frequency and low-frequency data as described in any one of claims 1 to 6 are completed.

10. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of a non-intrusive electrical equipment monitoring method that integrates high- and low-frequency data as described in any one of claims 1-6.

Images