Arc fault detection method and system, electronic equipment and computer program product

Abstract

The invention is suitable for the technical field of arc detection, and provides an arc fault detection method and system, electronic equipment and a computer program product, and the method comprises the steps: collecting an analog current signal in a target electrical circuit, constructing a data frame, obtaining a pre-training measurement matrix and an over-complete dictionary of a non-arc current signal, and carrying out compression measurement on the data frame based on the pre-training measurement matrix to obtain a low-dimensional measurement vector, carrying out sparse decomposition iteration on the low-dimensional measurement vector based on the over-complete dictionary to obtain a sparse coefficient, and carrying out arc fault detection on a target electrical circuit by using the sparse coefficient. And repeating the steps until an arc fault is detected. According to the arc fault detection method, efficient and accurate detection of the arc fault in the electrical circuit can be realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of arc detection, and particularly relates to an arc fault detection method and system, an electronic device, and a computer program product. BACKGROUND

[0002] An arc fault refers to a persistent or intermittent arc discharge phenomenon caused by insulation damage, poor contact, and other abnormal conditions in an electrical circuit. Arc discharge can produce local high temperature (up to several thousand degrees Celsius), accompanied by sparks and molten metal splashing, which is extremely easy to ignite surrounding combustible materials, and is one of the important reasons for electrical fires. Therefore, it is crucial to detect arc faults quickly and accurately to ensure electrical safety. SUMMARY

[0003] Therefore, the embodiments of the present application provide an arc fault detection method, system, electronic device, and computer program product to efficiently and accurately detect arc faults in an electrical circuit.

[0004] A first aspect of the embodiments of the present application provides an arc fault detection method, comprising:

[0005] Collecting an analog current signal in a target electrical circuit and constructing a data frame, wherein the data frame comprises a plurality of continuous current sampling signals;

[0006] Obtaining a pre-trained measurement matrix and an overcomplete dictionary of an arc-free current signal;

[0007] Performing compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector;

[0008] Performing sparse decomposition iteration on the low-dimensional measurement vector based on the overcomplete dictionary to obtain sparse coefficients;

[0009] Using the sparse coefficients to detect arc faults in the target electrical circuit;

[0010] Returning to the step of collecting an analog current signal in a target electrical circuit and constructing a data frame until an arc fault is detected.

[0011] In an implementation form of the first aspect, the method further comprises a pre-training phase as follows:

[0012] Obtaining an arc-free current signal dataset;

[0013] Performing dictionary learning based on the arc-free current signal dataset to obtain an overcomplete dictionary of an arc-free current signal;

[0014] Constructing a measurement matrix, wherein the measurement matrix and the overcomplete dictionary constitute a sensing matrix, and the sensing matrix has a restricted isometry property.

[0015] In an implementation form of the first aspect, the iteratively sparse decomposing the low-dimensional measurement vector based on the overcomplete dictionary comprises:

[0016] calculating an intermediate variable v (k) based on the following formula:

[0017]

[0018] wherein k is a current iteration number, α (k—1) is a sparse coefficient of a previous iteration, L is a Lipschitz constant, Φ is the measurement matrix, D noArc is the overcomplete dictionary, ΦD noArc is a sensing matrix, and y is the low-dimensional measurement vector.

[0019] soft-thresholding each element in the intermediate variable v (k) based on the following formula:

[0020]

[0021] wherein α is the sparse coefficient of the kth iteration, α (k) is an i-th component of the sparse coefficient of the kth iteration, v (k) is an i-th component of the intermediate variable v is a regularization parameter for controlling the sparsity strength, and sign(·) is a sign function.

[0022] if the termination iteration condition is not triggered, returning to the step of calculating the intermediate variable v (k) based on the following formula.

[0023] In an implementation form of the first aspect, the termination iteration condition comprises: the sparsity of the sparse coefficient α (k) obtained by iteration is invariant for a plurality of times continuously and the total iteration number k is greater than a minimum iteration threshold.

[0024] In an implementation form of the first aspect, the utilizing the sparse coefficient to detect the arc fault of the target electrical circuit comprises:

[0025] if the sparsity of the sparse coefficient is less than a preset sparsity threshold, determining that the target electrical circuit has no arc fault.

[0026] if the sparsity of the sparse coefficient is not less than the preset sparsity threshold, determining that the target electrical circuit has an arc fault.

[0027] In an implementation form of the first aspect, the step of returning the analog current signal in the target electrical circuit and constructing the data frame until an arc fault is detected comprises:

[0028] The step of returning the analog current signal in the target electrical circuit and constructing the original frame until the arc fault is detected to exist in the continuous multiple data frames.

[0029] In an implementation form of the first aspect, the method further comprises:

[0030] In the case that the arc fault is detected to exist in the continuous multiple data frames, triggering an arc fault alarm and performing a preset action sequence.

[0031] The second aspect of the embodiments of the present application provides an arc fault detection system, comprising:

[0032] A real-time sampling module configured to collect an analog current signal in a target electrical circuit and construct a data frame, the data frame comprising continuous multiple current sampling signals;

[0033] A loading module configured to obtain a pre-trained measurement matrix and an overcomplete dictionary, the overcomplete dictionary being an overcomplete dictionary of an arc-free current signal;

[0034] A compressed sensing module configured to perform compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector;

[0035] A sparse decomposition module configured to perform sparse decomposition iteration on the low-dimensional measurement vector based on the overcomplete dictionary to obtain sparse coefficients;

[0036] A fault discrimination module configured to perform arc fault detection on the target electrical circuit by using the sparse coefficients;

[0037] An iteration module configured to return and call the real-time sampling module until an arc fault is detected.

[0038] The third aspect of the embodiments of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the arc fault detection method according to the first aspect when executing the computer program.

[0039] The fourth aspect of the embodiments of the present application provides a computer program product comprising a computer program, which, when executed, causes the arc fault detection method according to the first aspect to be performed.

[0040] The beneficial effect of the first aspect of the embodiments of this application is as follows: by collecting analog current signals in the target electrical circuit and constructing a data frame, the data frame includes multiple consecutive current sampling signals, obtaining a pre-trained measurement matrix and an overcomplete dictionary of arc-free current signals, then performing compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector, and performing sparse decomposition iteration on the low-dimensional measurement vector based on the overcomplete dictionary to obtain sparse coefficients, and using the sparse coefficients to detect arc faults in the target electrical circuit, repeating the above process, the efficient and accurate detection of arc faults in the electrical circuit is achieved.

[0041] It is understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram illustrating the implementation process of the arc fault detection method provided in the embodiments of this application;

[0044] Figure 2 This is a schematic diagram of the pre-training process of the arc fault detection method provided in the embodiments of this application;

[0045] Figure 3 This is a schematic diagram of the arc fault detection system provided in the embodiments of this application;

[0046] Figure 4 This is a schematic diagram of the electronic device provided in the embodiments of this application;

[0047] Figure 5 This is a schematic diagram of a computer program product provided in an embodiment of this application. Detailed Implementation

[0048] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0049] It will be understood that the terms “comprises”, “comprising”, “includes”, “including”, “has”, “having” and variants thereof when used in this specification and in the following claims are not to be interpreted in an excluding sense but in an inclusive sense. That is, they will be understood to allow for elements, features, steps, or acts that would otherwise be excluded.

[0050] It will be further understood that the terms “comprises”, “comprising”, “includes”, “including”, “has”, “having” and variants thereof when used in this specification and in the following claims are not to be interpreted in an excluding sense but in an inclusive sense. That is, they will be understood to allow for elements, features, steps, or acts that would otherwise be excluded.

[0051] As used in this specification and claims, the terms “if’ and “when” can be interpreted to mean “upon” or “in response to determining,” or “in response to detecting,” depending on the context. Similarly, the phrase “if it is determined” or “if [a described condition or event] is detected” can be interpreted to mean “upon determining” or “in response to determining,” or “upon detecting [the described condition or event]” or “in response to detecting [the described condition or event],” depending on the context.

[0052] In addition, the description in the specification and claims of this application uses the term “comprising”, “including”, “containing”, “having” and variants thereof to mean that the features, integers, steps, or components so described are present, but not excluding the presence of one or more additional features, integers, steps, components, or groups thereof. “Comprising” is to be interpreted as including the more restrictive terms “consisting of and “consisting essentially of”.

[0053] Reference throughout this specification to “one embodiment” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. Thus, appearances of the phrases “in one embodiment”, “in some embodiments”, “in other embodiments”, “in additional embodiments”, and so on, in various places throughout this specification are not necessarily all referring to the same embodiment, unless otherwise specified. The terms “including”, “containing”, “comprising”, “having” and variations thereof in this specification are meant to be interpreted in an inclusive sense, i.e., “including, but not limited to”, unless otherwise specifically noted.

[0054] The embodiment of the present application provides an arc fault detection method, which is used for efficiently and accurately detecting an arc fault in an electrical circuit. The method provided by the present application comprises the following steps: collecting an analog current signal in a target electrical circuit and constructing a data frame, the data frame comprising a plurality of continuous current sampling signals; obtaining a pre-trained measurement matrix and an over-complete dictionary of an arc-free current signal; performing compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector; performing sparse decomposition iteration on the low-dimensional measurement vector based on the over-complete dictionary to obtain sparse coefficients; and detecting an arc fault in the target electrical circuit by using the sparse coefficients. The above process is repeated to efficiently and accurately detect the arc fault in the electrical circuit.

[0055] The arc fault detection method provided by the embodiment of the present application can be applied to electronic devices with high hardware configuration and abundant resources, such as mobile phones, tablet computers, wearable devices, augmented reality (AR) / virtual reality (VR) devices, notebook computers, ultra-mobile personal computers (UMPC), netbooks, personal digital assistants (PDA), and the like. The arc fault detection method can also be applied to resource-constrained embedded electronic devices, such as microprocessors. The embodiment of the present application does not limit the specific type of electronic device.

[0056] As shown in FIG. 1, Figure 1 The embodiment of the present application provides an arc fault detection method, which comprises the following steps:

[0057] In step S11, an analog current signal in a target electrical circuit is collected and a data frame is constructed, and the data frame comprises a plurality of continuous current sampling signals.

[0058] In application, an analog current signal from a current transformer or a current sensor is continuously collected by an analog-to-digital converter at a high sampling rate (for example, 2MSps, that is, 200 million samples per second). In application, N analog current signals (current sampling signals) sampled in a preset frame time are grouped into a data frame x∈R N , for example, 20000 analog current signals collected every 10 milliseconds are grouped into a data frame at a sampling rate of 2MSps.

[0059] In application, after collecting a first preset number of analog current signals (current sampling signals), the first preset number (N) of collected analog current signals (current sampling signals) are grouped into a data frame, for example, 1024 signals are collected to form a data frame x∈R N .

[0060] Step S12, obtaining a pre-trained measurement matrix and an overcomplete dictionary of arcless current signals.

[0061] In application, the pre-trained measurement matrix and the overcomplete dictionary of arcless current signals are obtained in advance in the pre-training phase and stored in a specific storage space or location, such as a flash memory of a microprocessor, a read-only memory (ROM), a random access memory (RAM), etc.

[0062] Step S13, compressing measurement of the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector.

[0063] In application, the pre-trained measurement matrix Φ is an M × N matrix, where M << N (for example, N = 1024, M = 100-200). The pre-trained measurement matrix is used to compress high-dimensional signals into low-dimensional measurement vectors. Based on the pre-trained measurement matrix Φ (or its generation rule), the data frame x is compressed and measured to obtain a low-dimensional measurement vector y ∈ R M , y = Φx. In application, the pre-trained measurement matrix is optimized and designed (sparse or structured) so that the matrix multiplication operation in the above compression measurement can be efficiently completed, which can be efficiently performed on a microcontroller unit (MCU) and can significantly reduce the data amount of subsequent processing.

[0064] In application, the data frame is simply pre-processed (such as band-pass filtering) after signal acquisition, and then compressed and measured.

[0065] Step S14, sparse decomposition iteration of the low-dimensional measurement vector based on the overcomplete dictionary to obtain sparse coefficients.

[0066] In application, the overcomplete dictionary of arcless current signals obtained by pre-training and the simplified FISTA (Fast Iterative Shrinkage-Thresholding Algorithm) algorithm are called to perform sparse decomposition iteration on the low-dimensional measurement vector y to obtain sparse coefficients α.

[0067] Step S15, arc fault detection of the target electrical circuit using the sparse coefficients.

[0068] In application, the sparsity of the sparse coefficients can be used to realize arc fault detection of the target electrical circuit.

[0069] Step S16, returning to the step of collecting the analog current signal in the target electrical circuit and constructing the data frame until the arc fault is detected.

[0070] In the application, if no arc fault is detected, steps S11 to S15 are continuously repeated to achieve continuous arc fault detection of the target electrical circuit. Once an arc fault is detected, execution exits and an arc fault alarm is triggered. After the arc fault is resolved, the continuous arc fault detection process can be restarted, achieving accurate and efficient detection of arc faults and ensuring the safety of the electrical circuit.

[0071] like Figure 2 As shown, in one embodiment, the method further includes the following pre-training phase:

[0072] Step S101: Obtain the arc-free current signal dataset.

[0073] In applications, the arc-free current signal dataset X noArc The dataset includes a large number of high-quality arc-free current signal samples, covering various normal load types, such as resistive loads, inductive loads, capacitive loads, motor starting, and dimmer lamp types, to ensure that subsequent dictionary learning can capture the complete features of normal current signals. The sampling rate of the arc-free current signal samples in the dataset is consistent with the sampling rate in the actual detection stage, for example, both are 2 MSps. The arc-free current signal dataset is organized into a matrix form X. noArc = [x1, x2, x i , ..., x T ]∈R N×T , where x i ∈R N For an N-dimensional signal sample, x i ∈R N For a signal segment (e.g., N = 1024 or 20000).

[0074] Step S102: Based on the arc-free current signal dataset, perform dictionary learning to obtain an overcomplete dictionary of arc-free current signals.

[0075] In applications, for the arc-free current signal dataset X noArc By applying dictionary learning algorithms, such as K-SVD (K-Singular Value Decomposition) or MOD (Method of Optimal Directions), an overcomplete dictionary D for the arc-free current signal is learned. noArc , through a complete dictionary D noArc = [d1, d2, d i , ...,d K ]∈R N×K (where the number of atoms K>N), and a set of sparse coefficient matrices α=[α1, α2, α3, α4, α5, α6, α7, α8, α9, α1 ...1, α2, α3i ,..., a T ] ∈ R K×T such that any arc-free current signal x T ∈ X noArc can be represented sparsely under this overcomplete dictionary, i.e., x t ≈ D a i , and a i is as sparse as possible.

[0076] In application, the objective function of dictionary learning is:

[0077]

[0078] where x i is the i-th arc-free signal, a i is the sparse coefficient corresponding to x i , is the square of Euclidean norm, || a i ||0represents the number of non-zero elements in a i (0-norm), and λ is the regularization coefficient (sparsity penalty coefficient).

[0079] In application, the sparse features of arc-free signals are automatically extracted by dictionary learning, so that the system can better identify abnormal arc signals (because arc signals are not sparse under the arc-free dictionary), and reduce the dependence on artificial features and threshold setting.

[0080] Step S103, construct a measurement matrix, and the measurement matrix and the overcomplete dictionary constitute a sensing matrix, and the sensing matrix has a restricted isometry property.

[0081] In application, an MxN measurement matrix Φ is constructed, where M << N (for example, N = 1024, M = 100-200). To ensure efficient operation on the MCU and the performance of compressive sensing, Φ has sparsity or structure, that is, most elements are zero, or can be generated by a small number of parameters to reduce storage and computational complexity, for example: a sparse random matrix: such as a Bernoulli random matrix, which has very few non-zero elements and low multiplication calculation amount, or a DCT / FFT matrix based on partial discrete cosine transform (DCT) / fast Fourier transform (FFT) matrix, the M rows of the DCT / FFT matrix are selected as the measurement matrix, and the fast transformation algorithm can be efficiently used to implement matrix multiplication. In addition, the measurement matrix Φ and the dictionary D noArc are jointly optimized to ensure that the sensing matrix A = ΦD noArc formed by the measurement matrix Φ and the dictionary D noArc satisfies the restricted isometry property (RIP) in the compressive sensing theory, thereby ensuring the accuracy of sparse recovery.

[0082] In practice, the aforementioned pre-training phase is completed before acquiring the pre-trained measurement matrix and the overcomplete dictionary of the arc-free current signal.

[0083] In applications, taking embedded platform microprocessors as an example, the measurement matrix and the overcomplete dictionary of arc-free current signals obtained in the pre-training stage are subjected to necessary model transformation and model quantization, and then solidified into the microprocessor. It can be solidified in a specific storage area. For example, taking an MCU with an NPU as an example, the measurement matrix and the overcomplete dictionary of arc-free current signals obtained in the pre-training stage are stored in the storage area accessible by the NPU.

[0084] In one embodiment, step S14, which involves iteratively decomposing the low-dimensional measurement vector based on the overcomplete dictionary to obtain sparse coefficients, includes:

[0085] Step S141: Calculate the intermediate variable v based on the following formula. (k) :

[0086]

[0087] Where k is the current iteration number, α (k—1) Let Φ be the sparse coefficients of the previous iteration, L be the Lipschitz constant, Φ be the measurement matrix, and D be the sparse coefficients of the previous iteration. noArc For the overcomplete dictionary, ΦD noArc Let y be the perception matrix and y be the low-dimensional measurement vector.

[0088] Step S142, based on the following formula, the intermediate variable v (k) Each element in the data undergoes soft thresholding:

[0089]

[0090] in, The sparse coefficients α in the k-th iteration (k) The i-th component, intermediate variable v (k) The i-th component, λ is the regularization parameter used to control the sparsity strength, sign(·) is the sign function, and max(·) is the maximum value function.

[0091] Step S143: If the termination iteration condition is not triggered, return to step S141 and calculate the intermediate variable v based on the following formula. (k) .

[0092] In practice, the termination condition is reaching the preset maximum number of iterations to ensure thorough sparse decomposition.

[0093] In application, the present application provides a simplified FISTA algorithm, which does not need to solve the sparse coefficient accurately as the traditional FISTA algorithm, but only needs to obtain the approximate sparsity of the sparse coefficient, so that the number of iterations can be reduced, the coefficient decomposition can be more efficient, and the overall arc fault detection efficiency can be improved. It can be understood that the above-mentioned sparse decomposition method provided by the present application further enables the overall arc fault detection method to have a millisecond-level real-time response, which can meet the real-time arc monitoring requirements, and greatly improves the real-time performance on a resource-limited embedded platform, such as a microprocessor.

[0094] In application, the sparse decomposition iteration process of the simplified FISTA algorithm includes the gradient descent step described in step S141 and the soft threshold shrinkage step described in step S142, wherein the matrix multiplication in step S141 can be efficiently performed by NPU (Neural Processing Unit), which greatly reduces the CPU (Central Processing Unit) burden and calculation time, and step S142 can also be efficiently performed on CPU and NPU.

[0095] The core calculation in the FISTA algorithm provided by the present application is the multiplication of the sensing matrix A = ΦD noArc and the transpose of the vector. Some MCUs have NPU specially optimized for such matrix operations, which can complete these calculations with extremely high efficiency, significantly shorten the single iteration time, and enable the entire sparse decomposition process to be completed within milliseconds.

[0096] In one embodiment, the termination iteration condition in step S143 includes that the sparsity of the sparse coefficient α (k) obtained by iteration is unchanged for a plurality of times in succession, and the total number of iterations k is greater than a minimum iteration threshold.

[0097] In application, in order to further improve efficiency and reduce power consumption, the sparse decomposition iteration provided by the embodiments of the present application introduces an early termination mechanism. In the iteration process of sparse decomposition, the number of non-zero items (sparsity) of the sparse coefficient α (k) obtained by current iteration is counted in real time, and the sparsity of the sparse coefficient in the last plurality of times is temporarily stored. If the sparsity remains unchanged for a plurality of times (for example, 5 times) in succession, and the total number of iterations has exceeded a preset minimum iteration threshold (to avoid premature termination), early termination is performed, and the sparse decomposition iteration is stopped. Compared with taking the preset maximum iteration number as the termination iteration condition, early termination according to the sparsity reduces unnecessary calculation, saves resource occupation and time consumption, reduces the power consumption and real-time delay of CPU / NPU, and improves the overall efficiency of the method.

[0098] In one embodiment, the step S15 of detecting arc fault of the target electrical circuit using the sparse coefficient comprises:

[0099] The step S151 of determining whether the target electrical circuit has arc fault comprises:

[0100] The step S152 of determining whether the target electrical circuit has arc fault comprises:

[0101] In application, when arc fault occurs, the current waveform usually presents abnormal characteristics different from normal working current, mainly including: wideband noise, the randomness and nonlinearity of arc will superimpose low frequency to megahertz high frequency wideband noise in current signal; irregular pulse and mutation, the extinction and reignition of arc will cause the current waveform to present sharp and random mutation and peak; harmonic distortion, the nonlinear volt-ampere characteristic of arc will introduce a large number of odd and even harmonics; zero current gap or distortion, in some types of arc (such as series arc), the current waveform near zero crossing point may appear obvious distortion, clipping or even short-term interruption. Therefore, the data frame when arc fault occurs cannot be sparsely represented by the overcomplete dictionary of non-arc current signal.

[0102] In application, the sparsity of the sparse coefficient α is ||α||0, and the preset sparsity threshold is T sparsity The sparsity ||α||0 is compared with the preset sparsity threshold T sparsity If ||α||0>T sparsity , it indicates that the current data frame can be sparsely represented under the overcomplete dictionary of non-arc current signal, and it is determined that the target electrical circuit currently has no arc fault; when arc fault occurs, the current signal will contain a large number of high frequency and random components different from normal signal, and these components lack effective sparse representation atoms in D noArc , resulting in that more dictionary atoms are needed to approximate during sparse decomposition (that is, the sparsity ||α||0 will increase significantly), so if ||α||0≤T sparsity , it indicates that the current data frame cannot be sparsely represented under the overcomplete dictionary of non-arc current signal, that is, it is no longer sparse, and it is determined that the target electrical circuit currently has arc fault.

[0103] In application, in addition to the non-zero term, the energy, mean, variance and other statistical characteristics of the sparse coefficient can also be calculated as a basis for discrimination.

[0104] In one embodiment, the step S16 of returning to the step S11 of collecting analog current signal in the target electrical circuit and constructing data frame until arc fault is detected comprises:

[0105] Returning to the step S11 of collecting the analog current signal in the target electrical circuit and constructing the original frame, until the arc fault is detected in the continuous multiple data frames.

[0106] In application, in order to improve the robustness of detection, the multiple data frame detection results are accumulated by sliding window decision, and only when the continuous multiple data frames are determined as arc fault, the arc fault is confirmed to exist, so as to avoid misjudgment caused by single transient interference.

[0107] In one embodiment, the method further comprises:

[0108] Step S17, in the case that the arc fault is detected in the continuous multiple data frames, triggering the arc fault alarm and executing the preset action sequence.

[0109] In application, in the case that the arc fault is detected in the continuous multiple data frames, triggering the arc fault alarm, such as audible and visual alarm, and executing the preset action sequence, such as relay or circuit breaker to cut off the circuit, realizing the circuit protection. The related fault information can also be sent to the upper computer or cloud platform for notification or record.

[0110] In application, taking the microprocessor as an example, the microprocessor can send the alarm signal to the audible and visual alarm or the device related to the execution of the preset action sequence through the communication module in the wireless local area network (WLAN) (such as Wi-Fi network), Bluetooth, Zigbee, mobile communication network, global navigation satellite system (GNSS), frequency modulation (FM), near field communication technology (NFC), infrared technology (IR) and other communication technologies, and send the corresponding fault information to the upper computer or cloud platform.

[0111] The related algorithm principles of dictionary learning, compressed sensing, FISTA and the like can be simulated and verified in detail in Matlab / Python and the like tools to determine the optimal dictionary size, measurement matrix type, regularization parameter λ and sparsity threshold T sparsity .

[0112] Compared with the prior art, for example, the traditional time domain / frequency domain AFD (Adaptive Fourier Decomposition), the traditional CS (Compressed Sensing) + reconstruction AFD, and the complex DL (Deep Learning) model AFD, the scheme provided by the embodiments of the present application has higher real-time performance, lower computational complexity, lower storage requirement, higher detection accuracy, lower power consumption, higher robustness, and better adaptability to embedded devices.

[0113] In applications, the arc fault detection method provided by the present application can be applied to arc fault detection occasions with strict requirements on cost, size, and response speed, such as smart sockets / switches in smart homes, miniature circuit breakers (MCB / AFDD) in electrical distribution boxes, industrial control cabinets, new energy vehicle charging piles, and line monitoring of aerospace equipment.

[0114] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0115] The embodiments of the present application also provide an arc fault detection system for executing the steps in the arc fault detection method embodiments described above. The arc fault detection system can be a virtual appliance (Virtual Appliance) in an electronic device, run by a processor of the electronic device, or can be the electronic device itself.

[0116] In one embodiment, the arc fault detection system is implemented based on a microprocessor.

[0117] In one embodiment, the arc fault detection system is implemented based on a microprocessor with an NPU, such as a microprocessor STM32N6.

[0118] As shown in Figure 3 The arc fault detection system 30 provided by the embodiments of the present application includes:

[0119] The real-time sampling module 301 is configured to collect an analog current signal in a target electrical line and construct a data frame, wherein the data frame includes a plurality of continuous current sampling signals.

[0120] The loading module 303 is configured to obtain a pre-trained measurement matrix and an overcomplete dictionary, wherein the overcomplete dictionary is an overcomplete dictionary of an arc-free current signal.

[0121] The compressed sensing module 303 is configured to perform compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector.

[0122] The sparse decomposition module 304 is used to perform sparse decomposition iteration on the low-dimensional measurement vector based on the overcomplete dictionary to obtain sparse coefficients.

[0123] The fault discrimination module 305 is used to detect arc faults in the target electrical circuit using the sparse coefficient.

[0124] Iteration module 306 is used to return and call the real-time sampling module until an arc fault is detected.

[0125] In one embodiment, the arc fault detection system 30 further includes a pre-training module 300, the pre-training module 300 being used for:

[0126] A dataset of arc-free current signals is acquired. Dictionary learning is performed based on the dataset to obtain an overcomplete dictionary of arc-free current signals. A measurement matrix is ​​constructed, and the measurement matrix and the overcomplete dictionary constitute a sensing matrix. The sensing matrix has the restricted isometry property.

[0127] In one embodiment, the sparse decomposition module 304 is used for:

[0128] Calculate the intermediate variable v based on the following formula. (k) :

[0129]

[0130] Where k is the current iteration number, α (k—1) Let Φ be the sparse coefficients of the previous iteration, L be the Lipschitz constant, Φ be the measurement matrix, and D be the sparse coefficients of the previous iteration. noArc For the overcomplete dictionary, ΦD noArc Let y be the perception matrix, and y be the low-dimensional measurement vector;

[0131] Based on the following formula, the intermediate variable v (k) Each element in the data undergoes soft thresholding:

[0132]

[0133] in, The sparse coefficients α in the k-th iteration (k) The i-th component, intermediate variable v (k) The i-th component, λ is the regularization parameter used to control the sparsity strength, sign(·) is the sign function, and max(·) is the maximum value function;

[0134] If the termination iteration condition is not triggered, return to the calculation of the intermediate variable v based on the following formula. (k) The steps.

[0135] In one embodiment, the termination condition includes the sparse coefficient α obtained from the iteration. (k) The sparsity remains constant over multiple iterations and the total number of iterations k is greater than the minimum iteration threshold.

[0136] In one embodiment, the fault detection module 305 is used to:

[0137] If the sparsity of the sparsity coefficient is less than the preset sparsity threshold, then it is determined that the target electrical circuit has no arc fault.

[0138] If the sparsity of the sparse coefficient is not less than a preset sparsity threshold, then it is determined that the target electrical circuit has an arc fault.

[0139] In one embodiment, the iteration module 306 is used to return to call the real-time sampling module until multiple consecutive data frames detect the presence of an arc fault.

[0140] In one embodiment, the arc fault detection system 30 further includes an alarm module 307, which is used to trigger an arc fault alarm and execute a preset action sequence when an arc fault is detected in multiple consecutive data frames.

[0141] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 4 of this embodiment includes: at least one processor 40 ( Figure 4 (Only one is shown in the diagram) a processor, a memory 41, and a computer program 42 stored in the memory 41 and executable on the at least one processor 40, wherein the processor 40 executes the computer program 42 to implement the steps in any of the above-described embodiments of the arc fault detection method.

[0142] The electronic device may include, but is not limited to, a processor 40 and a memory 41. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 4 and does not constitute a limitation on electronic device 4. It may include more or fewer components than shown, or combine certain components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0143] The processor 40 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0144] The memory 41 can be an internal storage unit of the electronic device 4 in some embodiments, for example, a hard disk or a memory of the electronic device 4. The memory 41 can also be an external storage device of the electronic device 4 in other embodiments, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 41 can include both the internal storage unit and the external storage device of the electronic device 4. The memory 41 is used to store an operating system, an application program, a boot loader, data and other programs, for example, program codes of the computer program, etc. The memory 41 can also be used to temporarily store data that has been output or is to be output.

[0145] It should be noted that the information interaction, execution process, etc. between the above apparatuses / units, since based on the same concept as the method embodiments of the present application, the specific functions and the brought technical effects can be referred to the method embodiments part, and will not be described here in detail.

[0146] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or software functional unit. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the unit and module in the above system can refer to the corresponding process in the foregoing method embodiment, which will not be described here.

[0147] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps in each method embodiment described above.

[0148] As shown in Figure 5 The embodiment of the present application provides a computer program product 5, which includes a computer program 42, and the computer program 42 is executed to make the arc fault detection method described above be executed.

[0149] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the present application realizes all or part of the processes in the above-mentioned embodiment methods, which can be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can realize the steps in each method embodiment described above when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable file or some intermediate form. The computer readable medium at least includes any entity or device capable of carrying the computer program code to the device / terminal equipment, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0150] In the above embodiments, the description of each embodiment focuses on different aspects, and the parts not described in detail or recorded in a certain embodiment can be referred to the relevant description of other embodiments.

[0151] Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions. The skilled person can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0152] In the embodiments provided in the present application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the above-described electronic device embodiments are only schematic, for example, the division of the modules or units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0153] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0154] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. An arc fault detection method characterized by, The method comprises the following steps: collecting an analog current signal in a target electrical circuit and constructing a data frame, the data frame comprising a plurality of continuous current sampling signals; obtaining a pre-trained measurement matrix and an over-complete dictionary of non-arc current signals; performing compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector; performing sparse decomposition iteration on the low-dimensional measurement vector based on the over-complete dictionary to obtain sparse coefficients; detecting arc fault of the target electrical circuit by using the sparse coefficients; returning to the step of collecting the analog current signal in the target electrical circuit and constructing the data frame until arc fault is detected.

2. The arc fault detection method of claim 1, wherein, The method further comprises a pre-training phase as follows: obtaining a non-arc current signal data set; performing dictionary learning based on the non-arc current signal data set to obtain an over-complete dictionary of non-arc current signals; constructing a measurement matrix, the measurement matrix and the over-complete dictionary constituting a sensing matrix, the sensing matrix having a limited isometry property.

3. The arc fault detection method of claim 1, wherein, The step of performing sparse decomposition iteration on the low-dimensional measurement vector based on the over-complete dictionary to obtain sparse coefficients comprises: The intermediate variable v is calculated based on the following equation (k) : where k is the current iteration number, a (k—1) is the sparse coefficient of the last iteration, L is the Lipschitz constant, F is the measurement matrix, D noArc is the overcomplete dictionary, F noArc is the sensing matrix, and y is the low-dimensional measurement vector. Each element in the intermediate variable v (k) is soft-thresholded based on the following equation: wherein, is the sparse coefficient a for the kth iteration (k) is the ith component of is an intermediate variable v (k) is the ith component of, and λ is a regularization parameter for controlling the sparsity strength, and sign(·) is a sign function and max(·) is a maximum function; In case the termination iteration condition is not triggered, the step of calculating an intermediate variable v (k) based on the formula v = v + (a - v) / b is returned.

4. The arc fault detection method of claim 3, wherein, The termination iteration condition comprises: the sparsity of the sparse coefficient α (k) is unchanged for continuous times and the total iteration number k is greater than a minimum iteration threshold.

5. The arc fault detection method of claim 1, wherein, The step of detecting arc fault of the target electrical circuit by using the sparse coefficients comprises: if the sparsity of the sparse coefficients is less than a preset sparsity threshold, determining that the target electrical circuit has no arc fault; if the sparsity of the sparse coefficients is not less than the preset sparsity threshold, determining that the target electrical circuit has arc fault.

6. The arc fault detection method of any one of claims 1 to 5, wherein, The step of returning to the step of collecting the analog current signal in the target electrical circuit and constructing the data frame until arc fault is detected comprises: returning to the step of collecting the analog current signal in the target electrical circuit and constructing an original frame until a plurality of continuous data frames detect that there is arc fault.

7. The arc fault detection method of claim 6, wherein, The method further comprises: in the case that a plurality of continuous data frames detect that there is arc fault, triggering arc fault alarm and executing a preset action sequence.

8. An arc fault detection system characterized by, The method comprises the following steps: a real-time sampling module for collecting an analog current signal in a target electrical circuit and constructing a data frame, the data frame comprising a plurality of continuous current sampling signals; a loading module for obtaining a pre-trained measurement matrix and an over-complete dictionary, the over-complete dictionary being an over-complete dictionary of non-arc current signals; a compressed sensing module for performing compressed measurement on the data frame based on the pre-trained measurement matrix to obtain a low-dimensional measurement vector; a sparse decomposition module for performing sparse decomposition iteration on the low-dimensional measurement vector based on the over-complete dictionary to obtain sparse coefficients; a fault discrimination module for detecting arc fault of the target electrical circuit by using the sparse coefficients; an iteration module for returning to call the real-time sampling module until arc fault is detected.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the steps of the arc fault detection method according to any one of claims 1 to 7.

10. A computer program product, characterised in that, The computer program is run to execute the arc fault detection method according to any one of claims 1 to 7.

Images