Equipment security registration method and system, equipment and storage medium

Abstract

The invention provides a device security registration method and system, a device and a storage medium. Relates to the technical field of electric equipment identification. The method is applied to an equipment security registration system, and comprises the following steps: monitoring voltage and current data at a main incoming line of a household distribution box; in response to a switching event of the access equipment, target features are determined according to the voltage and current data, the target features comprise steady-state features and transient-state features of the access equipment, and the switching event is used for indicating that the access equipment is connected or started; according to the target features, security detection is performed on the access device, and the security detection is used for identifying and filtering abnormal devices; when the access device passes the security detection, determining the device type of the access device according to the target feature and a preset device fingerprint database; and automatically registering the access equipment according to the equipment type. According to the invention, non-intrusive automatic discovery and safe registration of intelligent electric equipment and non-intelligent electric equipment in a family can be realized.

Description

Technical Field

[0001] This application relates to the field of electrical equipment identification technology, and in particular to a method, system, device and storage medium for secure registration of equipment. Background Technology

[0002] With the rapid development of smart home technology, home device management systems have gradually become an important part of modern families. In practical applications, the diversity of home devices presents significant challenges to device access and management.

[0003] In related technologies, the access and registration of electrical devices are accomplished through wireless protocol-based discovery mechanisms or manual input methods. However, wireless protocol-based discovery mechanisms are not applicable to non-smart home appliances, and manual input methods are cumbersome. Furthermore, household circuits may contain safety hazards such as substandard appliances and equipment with aging insulation. Therefore, there is an urgent need for a method that can automatically discover unknown new devices and complete their secure registration. Summary of the Invention

[0004] This application provides a device security registration method, system, device, and storage medium to solve the technical problems in related technologies that cannot automatically identify non-smart home appliances and lack active security protection measures.

[0005] In a first aspect, this application provides a device security registration method, applied to a device security registration system, the system including a data acquisition unit, an edge computing unit, and a cloud collaboration unit, the method including:

[0006] Monitor the voltage and current data at the main incoming line of the household electrical distribution box;

[0007] In response to the switching event of the access device, the target characteristics are determined based on voltage and current data. The target characteristics include the steady-state and transient characteristics of the access device. The switching event is used to indicate whether the access device is connected or powered on.

[0008] Based on the target characteristics, security detection is performed on the access devices to identify and filter abnormal devices.

[0009] When an access device passes a security test, the device type of the access device is determined based on the target characteristics and a pre-set device fingerprint database.

[0010] Automatic registration is performed on the access devices based on their type.

[0011] In one possible implementation, in response to a switching event of an access device, target characteristics are determined based on voltage and current data, including:

[0012] Based on voltage and current data, switching events are detected. A switching event is triggered when the current jumps above a first threshold.

[0013] In response to a switching event, steady-state characteristics are extracted based on voltage and current data. The steady-state characteristics include at least one of the following: active power, reactive power, total harmonic distortion of current, and power factor after switching.

[0014] Based on voltage and current data, transient features are extracted. The transient features include at least one of the following: the area of ​​the loop in the voltage and current trajectory, the flatness of the voltage and current trajectory, the peak value of the high-frequency energy distribution, and the local singularity features of the current waveform.

[0015] The target feature is obtained by combining steady-state and transient features.

[0016] In one possible implementation, transient features are extracted based on voltage and current data, including:

[0017] Based on the voltage and current data, extract current waveform segments before and after the switching.

[0018] Based on the current waveform segment, the energy distribution from low frequency to high frequency is analyzed by fast Fourier transform to obtain the peak value of the high frequency energy distribution;

[0019] Draw the voltage and current trajectory diagram based on the current waveform segment;

[0020] Extract the shape features of the voltage and current trajectory graph to obtain the annular area and flatness of the voltage and current trajectory graph;

[0021] Wavelet transform is applied to current waveform segments to extract local singularity features of the current waveform.

[0022] In one possible implementation, security checks are performed on the access device based on target characteristics, including:

[0023] Determine whether the target features meet the preset detection conditions, wherein the detection conditions include at least one of the following: the total harmonic distortion rate of the current is greater than the second threshold, the power factor is less than the third threshold, and the active power fluctuation exceeds the fourth threshold.

[0024] When the target features meet the detection conditions, it is determined that the access device has failed the security test and an abnormal device access alarm is generated.

[0025] If the target features do not meet the detection conditions, the access device is determined to have passed the security test.

[0026] In one possible implementation, the device type of the access device is determined based on the target characteristics and a pre-defined device fingerprint database, including:

[0027] Based on the dynamic time warping algorithm and / or deep neural network model, calculate the similarity between the target feature and the fingerprints of various devices in the device fingerprint database, and determine the highest similarity.

[0028] Determine the first device type corresponding to the highest similarity;

[0029] The device type of the access device is determined based on the first device type and the preset determination rules.

[0030] In one possible implementation, the device type of the access device is determined based on a first device type and a preset determination rule, including:

[0031] When the highest similarity is greater than the fifth threshold and the difference between the highest similarity and other similarities is greater than the sixth threshold, the first device type is determined as the device type of the access device.

[0032] When the highest similarity is between the seventh threshold and the fifth threshold, or when the highest similarity is greater than the fifth threshold but the difference between the highest similarity and other similarities is less than the sixth threshold, type confirmation information is pushed to the user device, and the first type information returned by the user device is received. The device type of the access device is determined according to the first type information, wherein the seventh threshold is less than the fifth threshold.

[0033] When the highest similarity is less than the seventh threshold, the access device is determined to be an unknown device. It is then classified into the target unknown class using a clustering algorithm, and its fingerprint is recorded. When the number of records in the target unknown class accumulates to the eighth threshold, type definition information is pushed to the user device, and the second type information returned by the user device is received. The device type of the access device is determined based on the second type information.

[0034] In one possible implementation, the access device is automatically registered based on the device type, including: creating a new device ID in the user device list and binding the device type and default icon.

[0035] In one possible implementation, the method also includes:

[0036] Receive correction information returned by the user equipment;

[0037] The target features and device types corresponding to the corrected information are identified as new sample data;

[0038] Add the new sample data to the device fingerprint database;

[0039] The deep neural network model is updated based on the new sample data.

[0040] In one possible implementation, the method also includes:

[0041] Obtain environmental information, which includes at least one of the following: time information, season information, and temperature information;

[0042] Adjust the second and / or third and / or fourth thresholds based on environmental information.

[0043] Secondly, this application provides a device security registration system, which includes a data acquisition unit, an edge computing unit, and a cloud collaboration unit, wherein...

[0044] The data acquisition unit is deployed at the main incoming line of the household electrical distribution box to monitor the voltage and current data at the main incoming line of the household electrical distribution box;

[0045] The edge computing unit is communicatively connected to the data acquisition unit and is used to respond to the switching events of the access device. Based on voltage and current data, it determines the target characteristics, including the steady-state and transient characteristics of the access device. Based on the target characteristics, it performs security detection on the access device to identify and filter abnormal devices.

[0046] The cloud collaboration unit communicates with the edge computing unit to determine the device type of the access device based on the target characteristics and the preset device fingerprint database when the access device passes the security test, and automatically registers the access device according to the device type.

[0047] Thirdly, this application provides an electronic device, including: a memory and a processor;

[0048] The memory stores the instructions that the computer executes;

[0049] The processor executes computer execution instructions stored in memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0050] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible embodiments of the first aspect.

[0051] The device security registration method, system, device, and storage medium provided in this application include: monitoring voltage and current data at the main incoming line of a household distribution box; responding to a switching event of an access device, determining target characteristics based on the voltage and current data, the target characteristics including steady-state and transient characteristics of the access device, the switching event indicating that the access device is connected or powered on; performing security detection on the access device based on the target characteristics, the security detection being used to identify and filter abnormal devices; when the access device passes the security detection, determining the device type of the access device based on the target characteristics and a preset device fingerprint database; and automatically registering the access device based on the device type. By monitoring the voltage and current at the main incoming line to construct target characteristics including steady-state and transient characteristics, first performing security detection on the access device to filter abnormal devices, and then identifying the device type of the access device through the device fingerprint database and completing automatic registration, non-intrusive automatic discovery and secure registration of smart and non-smart electrical devices in the home can be achieved, improving the intelligence of smart home device management and electrical safety. Attached Figure Description

[0052] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0053] Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of this application;

[0054] Figure 2 A flowchart illustrating a device security registration method provided in this application embodiment. Figure 1 ;

[0055] Figure 3 for Figure 2 The flowchart of the specific implementation method of step S202 in the embodiment shown is as follows;

[0056] Figure 4 for Figure 2 The flowchart illustrating the specific implementation of step S204 in the illustrated embodiment is shown.

[0057] Figure 5 A flowchart illustrating a device security registration method provided in this application embodiment. Figure 2 ;

[0058] Figure 6 This is a schematic diagram of the structure of the device security registration system provided in the embodiments of this application;

[0059] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0060] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0061] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0062] With the rapid development of smart home technology, home device management systems have gradually become an important part of modern families, with core functions including device management, energy consumption monitoring, and smart linkage. However, in practical applications, the diversity and heterogeneity of home devices (such as traditional non-smart appliances, old devices, and devices from different brands) pose significant challenges to device access and management.

[0063] In related technologies, the access and management of home devices mainly rely on the following solutions: wireless protocol-based discovery mechanisms, manual input and QR code binding, smart socket solutions, and Non-Intrusive Load Monitoring (NILM) technology. Among these, wireless protocol-based discovery mechanisms automatically discover smart devices within the network using protocols such as Multicast Domain Name System (mDNS) and Universal Plug and Play (UPnP). This solution requires devices to have network connectivity but cannot cover traditional non-smart home appliances and is susceptible to protocol compatibility issues. Manual input and QR code binding requires users to scan the device's QR code via a mobile app or manually enter the model number for binding. This method is cumbersome, error-prone, and unfriendly to elderly users. Smart socket solutions achieve access monitoring by installing smart sockets at the back end of the device, but require modification of the home circuit, resulting in high costs and inability to cover existing devices. NILM technology identifies the operating status of known devices by analyzing the total incoming current, such as whether the air conditioner is on, but lacks the ability to actively identify new devices and does not involve real-time detection of electrical safety hazards. Therefore, there is an urgent need for a method that can automatically discover unknown new devices and complete secure registration.

[0064] Based on this, embodiments of this application provide a device security registration method and system. By monitoring the voltage and current at the main incoming line, and constructing target features including steady-state and transient characteristics based on the switching events of the access devices, the access devices are first subjected to security detection to filter abnormal devices. Then, the device type of the access devices is identified through the device fingerprint database and automatic registration is completed. This enables non-intrusive automatic discovery and secure registration of smart and non-smart electrical devices in the home, improving the intelligence of smart home device management and power security.

[0065] Figure 1 This is a schematic diagram of an application scenario provided in an embodiment of this application, such as... Figure 1 As shown, it includes the device security registration system provided in this application and multiple electrical devices.

[0066] The device security registration system is deployed at the main incoming line of the household distribution box to manage multiple electrical devices in the home, including registration management when the device is first connected and status management when it is not first connected. Household electrical devices may include smart devices such as smart air conditioners and smart refrigerators, as well as non-smart devices such as older washing machines and older air conditioners.

[0067] The device security registration system comprises a data acquisition unit, an edge computing unit, and a cloud collaboration unit. The data acquisition unit collects total voltage and current waveforms in real time at a sampling rate of at least 10kHz and transmits them to the edge computing unit. Since the transient process of switching between home devices typically lasts 1-10ms, a 10kHz sampling rate can fully capture waveform details, ensuring no transient features are lost. The edge computing unit runs lightweight transient event detection algorithms, feature extraction algorithms, and preliminary security prediction algorithms. It captures device switching events in real time, extracts core features, and completes preliminary security screening. This enables local real-time response, avoiding cloud transmission delays and ensuring rapid interception of abnormal devices. The cloud collaboration unit performs deep matching in a feature library using a device type recognition model to achieve accurate device type identification. It also collects user feedback data to iteratively update the device type recognition model and the feature library.

[0068] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0069] Figure 2 A flowchart illustrating a device security registration method provided in this application embodiment. Figure 1 This method is applied to device security registration systems. For example... Figure 2 As shown, the method includes:

[0070] S201. Monitor the voltage and current data at the main incoming line of the household distribution box.

[0071] Voltage and current data are monitored by the data acquisition unit in the device safety registration system. This data includes the total voltage and current waveforms at the main incoming line of the household distribution box. For example, the data acquisition unit acquires the total voltage and current waveforms in real time at a sampling rate of at least 10 kHz and transmits them to the edge computing unit.

[0072] S202. In response to the switching event of the access device, determine the target characteristics based on voltage and current data. The target characteristics include the steady-state characteristics and transient characteristics of the access device.

[0073] Connected devices are electrical appliances in a household, such as washing machines and air conditioners. A switching event is when a connected device is connected or turned on. The target features are the steady-state and transient features of the connected device extracted from the current and voltage data before and after the switching event. Steady-state features include active power, reactive power, and total harmonic distortion of the current, while transient features include high-frequency energy distribution peaks, shape characteristics of the voltage and current (VI) trajectory (such as annular area and flatness), and local singularity characteristics of the current waveform.

[0074] The edge computing unit receives voltage and current data from the data acquisition unit and detects whether the total current has changed by a factor exceeding a preset threshold. When a change in the total current exceeding the preset threshold is detected, it indicates that a switching event of the access device has been triggered. After the switching event is triggered, the voltage and current data are analyzed and processed to obtain target features. The target features may include at least one of the steady-state features and at least one of the transient features.

[0075] S203. Based on the target characteristics, perform security detection on the access devices. The security detection is used to identify and filter abnormal devices.

[0076] Abnormal equipment refers to electrical appliances in the home that pose safety hazards, such as substandard chargers, faulty appliances, and aging appliances.

[0077] Security testing of access devices can be performed by comparing the values ​​of the target characteristics with those of the characteristics during normal operation to determine whether the access device is abnormal. For example, when a normal device is connected, the total harmonic distortion (THD) of the current should generally be less than 40%. If the THD of the current in the target characteristics of the access device is greater than 40%, the access device is determined to be abnormal and fails the security test.

[0078] For access devices that fail the security test, the registration process is not executed. Instead, an abnormal device access alarm is generated and sent to the user's device to remind the user to check the access device.

[0079] For access devices that pass the security test, continue with steps S204 to S205.

[0080] S204. When the access device passes the security test, determine the device type of the access device based on the target characteristics and the preset device fingerprint database.

[0081] The device fingerprint database is pre-built and includes device fingerprints for various known device types and models. Device fingerprints refer to the unique current and voltage waveform characteristics of electrical equipment during startup, operation, and shutdown, including steady-state power, harmonic components, and time-frequency domain characteristics of transient waveforms, which can uniquely identify the device type and status. For example, the device fingerprint database includes device fingerprints for various manufacturers and models of electrical equipment such as air conditioners, washing machines, and hair dryers.

[0082] The device type of the access device can be determined by searching the device fingerprint in the device fingerprint database that has the highest similarity to the target feature.

[0083] S205. Automatically register the access devices according to their type.

[0084] Automatic registration of access devices involves creating a new device ID in the user device list and binding it to the device type and default icon. After automatic registration is complete, the access device can be continuously monitored and managed through the secure registration system.

[0085] The device security registration method provided in this application monitors the voltage and current at the main incoming line, constructs target features including steady-state and transient characteristics based on the switching events of the access devices, first performs security detection on the access devices to filter out abnormal devices, and then identifies the device type of the access devices through the device fingerprint database and completes automatic registration. This enables non-intrusive automatic discovery and security registration of smart and non-smart electrical devices in the home, improving the intelligence of smart home device management and power security.

[0086] Optional, see Figure 3 , Figure 3 for Figure 2 The flowchart of the specific implementation of step S202 in the illustrated embodiment shows that, based on voltage and current data, the target features are determined, including:

[0087] S2021. Based on voltage and current data, detect switching events, wherein a switching event is triggered when the current jumps above a first threshold.

[0088] The first threshold is a preset dynamic judgment threshold used to determine whether a switching event has occurred. The edge computing unit receives the total incoming current data from the data acquisition unit in real time, continuously calculates the difference between the instantaneous current and the baseline steady-state current, and determines that a switching event has occurred when the absolute value of the difference exceeds the preset first threshold ΔI. The first threshold is, for example, 0.2A.

[0089] S2022. In response to the switching event, extract steady-state features based on voltage and current data, wherein the steady-state features include at least one of the following: active power, reactive power, total harmonic distortion of current, and power factor after switching.

[0090] Wherein, active power is represented by P, and the unit is watt (W) or kilowatt (KW); reactive power is represented by Q, and the unit is var (Var); total harmonic distortion (THD) is represented by THD as a percentage; and power factor is in ratio form, with a value range of [0,1].

[0091] The edge computing unit calculates active power, reactive power, and power factor based on voltage and current data using power parameter formulas, and calculates THD by analyzing fundamental and harmonic components using fast Fourier transform.

[0092] S2023. Based on the voltage and current data, extract transient features, which include at least one of the following: the ring area of ​​the voltage and current trajectory, the flatness of the voltage and current trajectory, the peak value of the high-frequency energy distribution, and the local singularity features of the current waveform.

[0093] For example, based on voltage and current data, transient features are extracted, including: extracting current waveform segments before and after switching based on voltage and current data; analyzing the energy distribution from low frequency to high frequency using fast Fourier transform based on the current waveform segments to obtain the high-frequency energy distribution peak; drawing voltage and current trajectory diagrams based on the current waveform segments; extracting the shape features of the voltage and current trajectory diagrams to obtain the annular area and flatness of the voltage and current trajectory diagrams; and performing wavelet transform on the current waveform segments to extract the local singularity features of the current waveform.

[0094] Among these features, the uniqueness of the VI trajectory diagrams for different types of equipment constitutes the core visual characteristic of equipment fingerprinting. The significant differences in the annular area of ​​the VI trajectory diagrams for different load types (resistive, inductive, capacitive, and nonlinear) are used to accurately distinguish equipment types. The flatness of the VI trajectory diagram is used to further refine the equipment load characteristics and reduce the identification error of similar equipment. The high-frequency energy distribution peak reflects the energy intensity of high-frequency harmonics at the moment of switching of the connected equipment; the high-frequency energy peak of nonlinear equipment (variable frequency appliances, switching power supply equipment) is much higher than that of ordinary resistive equipment. The local singularity characteristics of the current waveform can reflect the non-smoothness and discontinuity of the waveform; through wavelet transform extraction, subtle equipment differences that cannot be identified by conventional frequency domain analysis can be captured.

[0095] S2024. Combine the steady-state features and transient features to obtain the target features.

[0096] The edge computing unit extracts the steady-state and transient features, concatenates them in a fixed-dimensional order, and performs normalization and standardization processing to unify the numerical range to 0 to 1, eliminating numerical differences between parameters of different dimensions, and forming a standardized multidimensional target feature F. The target feature may include any one or more of the steady-state features, and may also include any one or more of the transient features.

[0097] In the above method, switching events are detected by current jump threshold, and steady-state features such as active power, reactive power, total harmonic distortion rate, and power factor are extracted, as well as transient features such as VI trajectory shape, high-frequency energy, and waveform local singularity, to form multi-dimensional target features. This can comprehensively characterize the unique electrical fingerprint of the access device, improve the distinguishability and recognition reliability of unknown devices, and provide an accurate data foundation for automatic discovery and secure registration.

[0098] Optionally, based on the target characteristics, security testing is performed on the access device, including: determining whether the target characteristics meet preset testing conditions, wherein the testing conditions include at least one of the following: the total harmonic distortion rate of the current is greater than a second threshold, the power factor is less than a third threshold, and the active power fluctuation exceeds a fourth threshold; when the target characteristics meet the testing conditions, it is determined that the access device has failed the security test and an abnormal device access alarm is generated; when the target characteristics do not meet the testing conditions, it is determined that the access device has passed the security test.

[0099] The detection conditions are preset quantitative judgment rules used to determine whether the access device has safety hazards or quality defects. These conditions may include multiple conditions. For example, the first condition is that the total harmonic distortion rate of the current is greater than a second threshold; the second condition is that the power factor is less than a third threshold; and the third condition is that the active power fluctuation exceeds a fourth threshold. When the access device meets any one of these conditions, it indicates that the access device has safety hazards or quality defects.

[0100] The upper limit of the second threshold THD is, for example, 40%. That is, when the THD of the access device is greater than 40%, it means that the operating current waveform of the access device deviates from the standard sine waveform, the harmonic content is too high, and there may be a large number of nonlinear components inside the access device, such as inferior rectifier circuits, switching power supplies, etc. The access device is judged to have failed the safety test.

[0101] The third threshold is the lower limit of the power factor (PF) threshold. For example, the third threshold is 0.5. That is, when the PF of the access device is less than 0.5, it means that the grid energy utilization efficiency of the access device is poor. It may be due to inferior electrical appliances, old and faulty equipment, non-standard modified loads, etc., and the access device is judged to have failed the safety test.

[0102] The fourth threshold is the deviation threshold of the active power fluctuation range. For example, the fourth threshold is ±20%. That is, when the active power fluctuation exceeds ±20%, it indicates that the access device is unstable and the internal components may fail. The access device is then judged to have failed the safety test.

[0103] For access devices that fail the security test, an abnormal device access alarm is generated and pushed to the user's device, reminding the user to conduct further inspection and handling of the access device.

[0104] For access devices that pass the security test, the subsequent registration process will continue.

[0105] The above method constructs multi-level safety detection conditions based on total harmonic distortion, power factor, and active power fluctuations, directly identifies and alarms abnormal devices, and can achieve pre-connection safety screening of devices, intercepting inferior, faulty, and interfering devices before registration, ensuring that all registered devices are safe devices, and strengthening the safety of household electricity use.

[0106] Optionally, the method further includes: acquiring environmental information, which includes at least one of the following: time information, season information, and temperature information; and adjusting the second threshold and / or the third threshold and / or the fourth threshold based on the environmental information.

[0107] Time information is used to indicate whether the current electricity consumption period is peak (e.g., 7-11 PM) or off-peak (e.g., 2-5 AM). During peak hours, high-power motor loads are concentrated, so the third and fourth thresholds can be appropriately increased, and the second threshold can be appropriately decreased. During off-peak hours, the third threshold can be appropriately decreased, the fourth threshold can remain unchanged, and the second threshold can be appropriately increased.

[0108] Seasonal information is used to indicate whether it is currently summer, winter, or spring / autumn. Summer sees more nonlinear loads due to the use of cooling equipment, while winter sees more high-power inductive / resistive devices due to the use of heating equipment. For winter, the second threshold can be appropriately lowered or kept unchanged, the third threshold can be appropriately lowered, and the fourth threshold can be appropriately increased. For summer, the second threshold can be appropriately increased, while the third and fourth thresholds remain unchanged.

[0109] Temperature information is used to indicate the current ambient temperature. For example, a temperature above 35°C is defined as a high-temperature environment, and a temperature below 5°C is defined as a low-temperature environment. The threshold adjustment rules can refer to the logic when adjusting the threshold based on seasonal information.

[0110] The above method, by adaptively adjusting the security detection threshold based on environmental information such as time, season, and temperature, can improve the environmental adaptability and robustness of abnormal device identification, achieve security screening in different application scenarios, and ensure the reliability of the registration process.

[0111] Optional, see Figure 4 , Figure 4 for Figure 2 The flowchart of the specific implementation of step S204 in the illustrated embodiment shows that, based on the target characteristics and a preset device fingerprint database, the device type of the access device is determined, including:

[0112] S2041. Based on the dynamic time warping algorithm and / or deep neural network model, calculate the similarity between the target feature and the fingerprints of various devices in the device fingerprint database, and determine the highest similarity.

[0113] Optionally, the target features can be temporally aligned and similarity matched with the features in the device fingerprint database using the Dynamic Time Warping (DTW) algorithm. This will obtain the similarity between the access device and various types of devices in the device fingerprint database, and then determine the highest similarity among them. The highest similarity is denoted as Smax.

[0114] Optionally, a deep neural network (DNN) model can be used to analyze the target features and device fingerprints in the device fingerprint database to obtain similarity and determine the highest similarity. Here, the DNN is a deep learning model based on multi-layer nonlinear transformations, which is pre-trained based on known device types and their corresponding features.

[0115] Optionally, the highest similarity can be determined jointly by DTW and DNN. For example, first use DNN to quickly match the top N types of devices with the highest similarity, and then use DTW to accurately calculate the highest similarity among the N types of devices.

[0116] S2042. Determine the first device type corresponding to the highest similarity.

[0117] Based on the highest similarity value, the device category label in the device fingerprint database is matched in reverse to obtain the device type corresponding to the label, and it is determined as the first device type.

[0118] S2043. Determine the device type of the access device according to the first device type and the preset determination rules.

[0119] The determination rule is a pre-set rule used to determine the device type of the access device, which may include multiple restrictions.

[0120] For example, determining the device type of the access device according to the determination rule includes: when the highest similarity is greater than a fifth threshold and the difference between the highest similarity and other similarities is greater than a sixth threshold, determining the first device type as the device type of the access device; when the highest similarity is between a seventh threshold and a fifth threshold, or when the highest similarity is greater than a fifth threshold but the difference between the highest similarity and other similarities is less than a sixth threshold, pushing type confirmation information to the user device and receiving the first type information returned by the user device, determining the device type of the access device according to the first type information, wherein the seventh threshold is less than the fifth threshold; when the highest similarity is less than the seventh threshold, determining the access device as an unknown device, classifying it into a target unknown class through a clustering algorithm, recording its fingerprint, and when the number of records of the target unknown class accumulates to an eighth threshold, pushing type definition information to the user device and receiving the second type information returned by the user device, determining the device type of the access device according to the second type information.

[0121] The fifth threshold is 90%, and the sixth threshold is 20%. That is, when Smax > 90% and the difference between it and other similarities is greater than 20%, the first device type is determined as the device type of the access device, and the registration process is further initiated.

[0122] The seventh threshold is, for example, 70%. That is, when Smax is between 70% and 90%, the fuzzy recognition process begins. The fuzzy recognition process includes: pushing type confirmation information to the user device, pushing one or more possible device types to the user, the user manually selecting to determine the device type of the access device, generating first type information, returning the first type information to the cloud collaboration unit, the cloud collaboration unit receiving the first type information, and determining the device type indicated in the first type information as the device type of the access device.

[0123] Furthermore, when Smax > 90%, but the difference between it and other similarities is less than 20%, it enters the fuzzy recognition process.

[0124] When Smax < 70%, the device is identified as an unknown device and enters the cold start identification process. The cold start identification process includes: classifying the access device into the target unknown class using a clustering algorithm, recording its fingerprint, and pushing type definition information to the user device when the number of records for the target unknown class reaches an eighth threshold. The user then manually inputs the device type of the access device, generating second type information, which is returned to the cloud collaboration unit. The cloud collaboration unit receives the second type information and determines the device type indicated in the second type information as the access device's device type. If the eighth threshold is not reached, only the fingerprint of the access device is recorded in this instance. For example, the eighth threshold is 5 times; that is, when the same access device accesses for the 5th time, type definition information is pushed to the user device and subsequent steps are executed; otherwise, only its fingerprint is recorded, and no further operations are performed in this instance.

[0125] In the above method, feature similarity matching is performed by using dynamic time warping algorithm and / or deep neural network model, and device type is determined by combining preset judgment rules. Features and device fingerprints are matched efficiently and accurately to achieve automatic identification of unknown devices. Through multi-level similarity classification rules, accurate confirmation, user-assisted confirmation and cold start of unknown device clustering are achieved in stages. This can ensure the rapid registration of known devices and the continuous learning and expansion of unknown devices, thus realizing the applicability of automatic device registration.

[0126] Optionally, after automatically registering the access device, the method also includes a user feedback process and a model update process, specifically including: receiving correction information returned by the user device; determining the target features and device type corresponding to the correction information as new sample data; adding the new sample data to the device fingerprint database; and updating the deep neural network model based on the new sample data.

[0127] The user feedback process involves the system identifying the device type indicated in the user's correction information as a label when the user provides feedback on the first or second type of information through their device, or when the user manually corrects the identification result (e.g., correcting an electric kettle to a tea maker). The system then matches this label with the target features, combines them into new sample data, and adds the new sample data to the device fingerprint database to expand the database and improve its accuracy.

[0128] Optionally, when similarity is determined through a deep neural network model, the method also includes using new sample data to update the deep neural network model through a transfer learning algorithm, so as to achieve increasing accuracy with repeated use.

[0129] In the above method, the device fingerprint database and deep neural network model are continuously optimized by the correction information provided by the user, so that the system can continuously improve the accuracy of unknown device identification during use, forming a closed-loop automatic registration mechanism that becomes more and more accurate with use.

[0130] Figure 5 A flowchart illustrating a device security registration method provided in this application embodiment. Figure 2 ,like Figure 5 As shown, the method includes:

[0131] S301. Monitor the voltage and current data at the main incoming line of the household distribution box.

[0132] S302. Determine whether a switching event has occurred.

[0133] If yes, proceed to step S303; otherwise, return to step S301.

[0134] S303. Determine the target characteristics based on voltage and current data.

[0135] S304. Determine whether the access device has passed the security test.

[0136] If the access device passes the security test, proceed to step S305; if the access device fails the security test, proceed to step S313, generate an abnormal device access alarm, and end the method.

[0137] S305. Determine the similarity between the target features and the fingerprints of various devices in the device fingerprint database.

[0138] S306. Determine whether the highest similarity is greater than the fifth threshold and whether the difference between it and other similarities is greater than the sixth threshold.

[0139] If not, proceed to step S307; if yes, proceed to step S308.

[0140] S307. Determine whether the highest similarity is greater than the seventh threshold.

[0141] If not, proceed to step S309; ​​if yes, proceed to step S310.

[0142] S308. Determine the device type of the access device as the device corresponding to the highest similarity.

[0143] After step S308, step S311 is executed to automatically register the access device.

[0144] S309. Enter the fuzzy recognition process.

[0145] The fuzzy recognition process includes: pushing type confirmation information to the user device, pushing one or more possible device types to the user, the user manually selecting to determine the device type of the access device, generating first type information, returning the first type information to the cloud collaboration unit, the cloud collaboration unit receiving the first type information, and determining the device type indicated in the first type information as the device type of the access device.

[0146] After step S309, step S311 is executed to automatically register the access device, and step S312 is executed to determine the first type of information as correction information and update the device fingerprint database and / or deep neural network model.

[0147] S310, Entering the cold start identification process.

[0148] The cold start identification process includes: classifying the access device into the target unknown class using a clustering algorithm, recording its fingerprint, and pushing type definition information to the user device when the number of records for the target unknown class reaches the eighth threshold. The user then manually inputs the device type of the access device to generate second type information, which is then returned to the cloud collaboration unit. The cloud collaboration unit receives the second type information and determines the device type indicated in the second type information as the device type of the access device.

[0149] After receiving the second type of information, step S311 is executed to automatically register the access device, and step S312 is further executed to use the second type of information as correction information to update the device fingerprint database and / or deep neural network model.

[0150] S311. Automatically register the access devices according to their type.

[0151] S312. Receive correction information returned by the user equipment and update the device fingerprint database and / or deep neural network model.

[0152] S313, Generate an alarm for abnormal device access.

[0153] The device security registration method provided in this application monitors the voltage and current at the main incoming line, constructs target features including steady-state and transient characteristics based on the switching events of the access devices, first performs security detection on the access devices to filter out abnormal devices, and then identifies the device type of the access devices through the device fingerprint database and completes automatic registration. This enables non-intrusive automatic discovery and security registration of smart and non-smart electrical devices in the home, improving the intelligence of smart home device management and power security.

[0154] Figure 6 This is a schematic diagram of the device security registration system provided in the embodiments of this application, such as... Figure 6 As shown, the device security registration system 60 provided in this embodiment includes a data acquisition unit 61, an edge computing unit 62, and a cloud collaboration unit 63, wherein...

[0155] The data acquisition unit 61 is deployed at the main incoming line of the household distribution box to monitor the voltage and current data at the main incoming line of the household distribution box.

[0156] The edge computing unit 62 is communicatively connected to the data acquisition unit 61 and is used to respond to the switching event of the access device, determine the target characteristics based on voltage and current data, the target characteristics include the steady-state characteristics and transient characteristics of the access device, and perform security detection on the access device based on the target characteristics. The security detection is used to identify and filter abnormal devices.

[0157] The cloud collaboration unit 63 is communicatively connected to the edge computing unit 62. When the access device passes the security test, it determines the device type of the access device based on the target characteristics and the preset device fingerprint database, and automatically registers the access device according to the device type.

[0158] In one possible implementation, the edge computing unit 62 is specifically used for:

[0159] Based on voltage and current data, switching events are detected. A switching event is triggered when the current jumps above a first threshold.

[0160] In response to a switching event, steady-state characteristics are extracted based on voltage and current data. The steady-state characteristics include at least one of the following: active power, reactive power, total harmonic distortion of current, and power factor after switching.

[0161] Based on voltage and current data, transient features are extracted. The transient features include at least one of the following: the area of ​​the loop in the voltage and current trajectory, the flatness of the voltage and current trajectory, the peak value of the high-frequency energy distribution, and the local singularity features of the current waveform.

[0162] The target feature is obtained by combining steady-state and transient features.

[0163] In one possible implementation, the edge computing unit 62 is also used for:

[0164] Based on the voltage and current data, extract current waveform segments before and after the switching.

[0165] Based on the current waveform segment, the energy distribution from low frequency to high frequency is analyzed by fast Fourier transform to obtain the peak value of the high frequency energy distribution;

[0166] Draw the voltage and current trajectory diagram based on the current waveform segment;

[0167] Extract the shape features of the voltage and current trajectory graph to obtain the annular area and flatness of the voltage and current trajectory graph;

[0168] Wavelet transform is applied to current waveform segments to extract local singularity features of the current waveform.

[0169] In one possible implementation, the edge computing unit 62 is also used for:

[0170] Determine whether the target features meet the preset detection conditions, wherein the detection conditions include at least one of the following: the total harmonic distortion rate of the current is greater than the second threshold, the power factor is less than the third threshold, and the active power fluctuation exceeds the fourth threshold.

[0171] When the target features meet the detection conditions, it is determined that the access device has failed the security test and an abnormal device access alarm is generated.

[0172] If the target features do not meet the detection conditions, the access device is determined to have passed the security test.

[0173] In one possible implementation, the cloud collaboration unit 63 is specifically used for:

[0174] Based on the dynamic time warping algorithm and / or deep neural network model, calculate the similarity between the target feature and the fingerprints of various devices in the device fingerprint database, and determine the highest similarity.

[0175] Determine the first device type corresponding to the highest similarity;

[0176] The device type of the access device is determined based on the first device type and the preset determination rules.

[0177] In one possible implementation, the cloud collaboration unit 63 is also used for:

[0178] When the highest similarity is greater than the fifth threshold and the difference between the highest similarity and other similarities is greater than the sixth threshold, the first device type is determined as the device type of the access device.

[0179] When the highest similarity is between the seventh threshold and the fifth threshold, or when the highest similarity is greater than the fifth threshold but the difference between the highest similarity and other similarities is less than the sixth threshold, type confirmation information is pushed to the user device, and the first type information returned by the user device is received. The device type of the access device is determined according to the first type information, wherein the seventh threshold is less than the fifth threshold.

[0180] When the highest similarity is less than the seventh threshold, the access device is determined to be an unknown device. It is then classified into the target unknown class using a clustering algorithm, and its fingerprint is recorded. When the number of records in the target unknown class accumulates to the eighth threshold, type definition information is pushed to the user device, and the second type information returned by the user device is received. The device type of the access device is determined based on the second type information.

[0181] In one possible implementation, the cloud collaboration unit 63 is also used to: create a new device ID in the user device list and bind the device type and default icon.

[0182] In one possible implementation, the cloud collaboration unit 63 is also used for:

[0183] Receive correction information returned by the user equipment;

[0184] The target features and device types corresponding to the corrected information are identified as new sample data;

[0185] Add the new sample data to the device fingerprint database;

[0186] The deep neural network model is updated based on the new sample data.

[0187] In one possible implementation, the edge computing unit 62 is also used for:

[0188] Obtain environmental information, which includes at least one of the following: time information, season information, and temperature information;

[0189] Adjust the second and / or third and / or fourth thresholds based on environmental information.

[0190] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 7 As shown, the electronic device 70 provided in this embodiment includes at least one processor 701 and a memory 702. Optionally, the electronic device 70 further includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus.

[0191] In a specific implementation, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to execute the above-described device security registration method embodiment.

[0192] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar, so it will not be repeated here. In the above embodiments, it should be understood that the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the device security registration method disclosed in this invention can be directly reflected as being executed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0193] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0194] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0195] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the aforementioned device security registration method.

[0196] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory, electrically erasable programmable read-only memory, erasable programmable read-only memory, programmable read-only memory, read-only memory, magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0197] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0198] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0199] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0200] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0201] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0202] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0203] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for secure device registration, characterized in that, The method, applied to a device security registration system, includes: Monitor the voltage and current data at the main incoming line of the household electrical distribution box; In response to a switching event of an access device, a target feature is determined based on the voltage and current data. The target feature includes the steady-state and transient features of the access device. The switching event is used to indicate that the access device is connected or powered on. Based on the target characteristics, the access device is subjected to security detection, which is used to identify and filter abnormal devices; When the access device passes the security detection, the device type of the access device is determined based on the target characteristics and a preset device fingerprint database; The access device is automatically registered according to the device type. The step of performing security detection on the access device based on the target characteristics includes: determining whether the target characteristics meet preset detection conditions, wherein the detection conditions include at least one of the following: the total harmonic distortion rate of the current is greater than a second threshold, the power factor is less than a third threshold, and the active power fluctuation exceeds a fourth threshold; when the target characteristics meet the detection conditions, it is determined that the access device has failed the security detection and an abnormal device access alarm is generated; when the target characteristics do not meet the detection conditions, it is determined that the access device has passed the security detection.

2. The method according to claim 1, characterized in that, The method of responding to the switching event of the access device and determining the target characteristics based on the voltage and current data includes: Based on the voltage and current data, the switching event is detected, wherein the switching event is triggered when the current changes by more than a first threshold. In response to the switching event, the steady-state characteristics are extracted based on the voltage and current data, wherein the steady-state characteristics include at least one of the following: active power, reactive power, total harmonic distortion of current, and power factor after switching; Based on the voltage and current data, the transient features are extracted, and the transient features include at least one of the following: the ring area of ​​the voltage and current trajectory, the flatness of the voltage and current trajectory, the peak value of the high-frequency energy distribution, and the local singularity features of the current waveform; The target feature is obtained by combining the steady-state feature and the transient feature.

3. The method according to claim 2, characterized in that, The step of extracting the transient features based on the voltage and current data includes: Based on the voltage and current data, extract current waveform segments before and after the switching. Based on the current waveform segment, the energy distribution from low frequency to high frequency is analyzed by fast Fourier transform to obtain the peak value of the high frequency energy distribution; Based on the current waveform segment, draw the voltage and current trajectory diagram; Extract the shape features of the voltage and current trajectory diagram to obtain the annular area and flatness of the voltage and current trajectory diagram; Wavelet transform is performed on the current waveform segment to extract the local singularity features of the current waveform.

4. The method according to claim 1, characterized in that, The step of determining the device type of the access device based on the target features and a preset device fingerprint database includes: Based on the dynamic time warping algorithm and / or deep neural network model, calculate the similarity between the target feature and the fingerprints of various devices in the device fingerprint database, and determine the highest similarity. Determine the first device type corresponding to the highest similarity; The device type of the access device is determined based on the first device type and the preset determination rules.

5. The method according to claim 4, characterized in that, The step of determining the device type of the access device based on the first device type and a preset determination rule includes: When the highest similarity is greater than the fifth threshold and the difference between the highest similarity and other similarities is greater than the sixth threshold, the first device type is determined as the device type of the access device. When the highest similarity is between the seventh threshold and the fifth threshold, or when the highest similarity is greater than the fifth threshold, but the difference between the highest similarity and other similarities is less than the sixth threshold, type confirmation information is pushed to the user equipment, and first type information returned by the user equipment is received. The device type of the access device is determined according to the first type information, wherein the seventh threshold is less than the fifth threshold. When the highest similarity is less than the seventh threshold, the access device is determined to be an unknown device. It is then classified into the target unknown class using a clustering algorithm, and its fingerprint is recorded. When the number of records in the target unknown class accumulates to the eighth threshold, type definition information is pushed to the user device, and the second type information returned by the user device is received. The device type of the access device is determined based on the second type information.

6. The method according to claim 1, characterized in that, The step of automatically registering the access device according to the device type includes: creating a new device ID in the user device list and binding the device type and default icon.

7. The method according to claim 4, characterized in that, The method further includes: Receive correction information returned by the user equipment; The target features and device types corresponding to the corrected information are determined as new sample data; Add the new sample data to the device fingerprint database; The deep neural network model is updated based on the new sample data.

8. The method according to claim 1, characterized in that, The method further includes: Obtain environmental information, which includes at least one of the following: time information, season information, and temperature information; Based on the environmental information, the second threshold and / or the third threshold and / or the fourth threshold are adjusted.

9. A device security registration system, characterized in that, The device security registration system includes a data acquisition unit, an edge computing unit, and a cloud collaboration unit, wherein... The data acquisition unit is deployed at the main incoming line of the household electrical distribution box to monitor the voltage and current data at the main incoming line of the household electrical distribution box. The edge computing unit is communicatively connected to the data acquisition unit and is used to respond to the switching event of the access device, determine target characteristics based on the voltage and current data, the target characteristics including the steady-state and transient characteristics of the access device, and perform security detection on the access device based on the target characteristics. The security detection is used to identify and filter abnormal devices. The security detection includes: determining whether the target characteristics meet preset detection conditions, wherein the detection conditions include at least one of the following: the total harmonic distortion rate of the current is greater than a second threshold, the power factor is less than a third threshold, and the active power fluctuation exceeds a fourth threshold; when the target characteristics meet the detection conditions, it is determined that the access device has failed the security detection and an abnormal device access alarm is generated; when the target characteristics do not meet the detection conditions, it is determined that the access device has passed the security detection. The cloud collaboration unit is communicatively connected to the edge computing unit and is used to determine the device type of the access device based on the target characteristics and a preset device fingerprint database when the access device passes the security detection, and to automatically register the access device based on the device type.

10. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 8.

Images