技术领域Technical field
本申请涉及配电网技术领域，尤其涉及一种间歇性电弧故障检测方法及相关装置。The present application relates to the technical field of distribution network, and in particular to an intermittent arc fault detection method and related devices.
背景技术Background technique
在谐振接地系统的接地故障事件中，单相接地故障发生概率高，占故障总数的绝大多数。为保障配电网的安全、稳定和可靠性，有必要在单相接地故障实际发生前做到及时检测，有效排查。大部分单相接地故障的发展过程是由初期的间歇性电弧故障(Intermittent Arc Faults,IAF)逐渐过渡到稳定弧光接地故障，最后演变为单相永久性接地故障。因此，有效检测间歇性电弧故障有利于尽早发现安全隐患，从而能够在更为严重的故障发生前及时消除具有安全隐患的故障设备或影响馈线安全运行的外界环境，以达到保护配电网稳定运行的目的。Among ground fault events in resonant grounding systems, single-phase ground faults have a high probability of occurrence and account for the vast majority of the total number of faults. In order to ensure the safety, stability and reliability of the distribution network, it is necessary to promptly detect and effectively troubleshoot single-phase ground faults before they actually occur. The development process of most single-phase ground faults gradually transitions from initial intermittent arc faults (IAF) to stable arc ground faults, and finally evolves into single-phase permanent ground faults. Therefore, effective detection of intermittent arc faults is conducive to the early detection of safety hazards, so that faulty equipment with safety hazards or external environments that affect the safe operation of feeders can be promptly eliminated before more serious faults occur, so as to protect the stable operation of the distribution network. the goal of.
现有的间歇性电弧故障检测方法，通常采用时频分析法来提取故障波形特征，然后将故障波形特征输入到分类器中进行分类，该方法提取的特征与分类器的配合并非是最优的，并且现有的分类器在实测波形的检测上抗干扰性不强，使得检测精度不高。Existing intermittent arc fault detection methods usually use time-frequency analysis to extract fault waveform features, and then input the fault waveform features into a classifier for classification. The cooperation between the features extracted by this method and the classifier is not optimal. , and the existing classifiers are not strong in anti-interference in the detection of measured waveforms, resulting in low detection accuracy.
发明内容Contents of the invention
本申请提供了一种间歇性电弧故障检测方法及相关装置，用于改善现有的间歇性电弧故障检测方法存在的检测精度不高的技术问题。This application provides an intermittent arc fault detection method and related devices, which are used to improve the technical problem of low detection accuracy in existing intermittent arc fault detection methods.
有鉴于此，本申请第一方面提供了一种间歇性电弧故障检测方法，包括：In view of this, the first aspect of this application provides an intermittent arc fault detection method, including:
在获取到配电网的待检测零序电流波形后，按照时间顺序将所述待检测零序电流波形分割为若干个波形片段，并对该波形片段进行预处理，得到若干测试样本；After obtaining the zero-sequence current waveform to be detected in the distribution network, divide the zero-sequence current waveform to be detected into several waveform segments in time order, and preprocess the waveform segments to obtain several test samples;
将各所述测试样本依次输入到间歇性电弧故障检测模型进行故障检测，得到各所述测试样本的故障检测结果；Input each of the test samples into the intermittent arc fault detection model in turn for fault detection, and obtain the fault detection results of each of the test samples;
根据各所述测试样本的故障检测结果依次确定当前测试样本是否发生间歇性电弧故障，若否，则设置从因子SF＝SF+1，若是，则设置主因子PF＝PF+1，且设置SF＝0；According to the fault detection results of each test sample, it is determined whether the current test sample has an intermittent arc fault. If not, the slave factor SF=SF+1 is set. If it is, the master factor PF=PF+1 is set, and SF is set. =0;
当主因子PF大于第一预设阈值且从因子SF大于第二预设阈值时，判定所述配电网发生了短时间歇性电弧故障；When the primary factor PF is greater than the first preset threshold and the slave factor SF is greater than the second preset threshold, it is determined that a short-term intermittent arc fault has occurred in the distribution network;
当主因子PF大于第三预设阈值时，判定所述配电网发生了长时间歇性电弧故障；When the main factor PF is greater than the third preset threshold, it is determined that a long-term intermittent arc fault has occurred in the distribution network;
当主因子PF小于第四预设阈值且从因子SF大于所述第二预设阈值时，判定所述配电网未发生间歇性电弧故障。When the primary factor PF is less than the fourth preset threshold and the slave factor SF is greater than the second preset threshold, it is determined that no intermittent arc fault occurs in the distribution network.
可选的，所述间歇性电弧故障检测模型的训练过程为：Optionally, the training process of the intermittent arc fault detection model is:
获取待训练零序电流波形，所述待训练零序电流波形包括发生了间歇性电弧故障的波形和未发生间歇性电弧故障的波形；Obtaining a zero-sequence current waveform to be trained, where the zero-sequence current waveform to be trained includes a waveform in which an intermittent arc fault occurs and a waveform in which an intermittent arc fault does not occur;
将所述待训练零序电流波形分割为若干个相同大小的波形片段，对该波形片段进行预处理，得到若干训练样本；Divide the zero-sequence current waveform to be trained into several waveform segments of the same size, preprocess the waveform segments to obtain several training samples;
通过所述训练样本训练预置卷积神经网络，得到间歇性电弧故障检测模型。The preset convolutional neural network is trained through the training samples to obtain an intermittent arc fault detection model.
可选的，对波形片段进行预处理，得到若干训练样本或测试样本，包括：Optionally, preprocess the waveform segments to obtain several training samples or test samples, including:
基于双三次差值法将各波形片段映射为预设长度的数据序列；Based on the bicubic difference method, each waveform segment is mapped into a data sequence of preset length;
对各所述数据序列进行归一化处理，得到各归一化数据序列；Perform normalization processing on each of the data sequences to obtain each normalized data sequence;
按照时间顺序将各所述归一化数据序列转换为矩阵形式，得到各波形片段对应的训练样本或测试样本。Convert each normalized data sequence into matrix form in time order to obtain training samples or test samples corresponding to each waveform segment.
可选的，所述预置卷积神经网络由输入层、5个可分离卷积层、4个连接层、1个卷积层、1个全连接模块和输出层构成，所述全连接模块由若干个全连接层串联构成；Optionally, the preset convolutional neural network consists of an input layer, 5 separable convolutional layers, 4 connection layers, 1 convolutional layer, 1 fully connected module and an output layer. The fully connected module It is composed of several fully connected layers connected in series;
所述输入层的输出端分别与第一可分离卷积层、第二可分离卷积层、第三可分离卷积层的输入端连接；The output end of the input layer is connected to the input end of the first separable convolution layer, the second separable convolution layer, and the third separable convolution layer respectively;
第一连接层与所述第一可分离卷积层、所述第二可分离卷积层的输出端连接，第二连接层与所述第二可分离卷积层、所述第三可分离卷积层的输出端连接；The first connection layer is connected to the output terminals of the first separable convolution layer and the second separable convolution layer, and the second connection layer is connected to the second separable convolution layer and the third separable convolution layer. The output terminal of the convolutional layer is connected;
第四可分离卷积层的输入端与所述第一连接层的输出端连接，输出端通过第三连接层与所述卷积层的输入端连接；The input end of the fourth separable convolution layer is connected to the output end of the first connection layer, and the output end is connected to the input end of the convolution layer through a third connection layer;
第五可分离卷积层的输入端与所述第二连接层的输出端连接；The input terminal of the fifth separable convolution layer is connected to the output terminal of the second connection layer;
第四连接层的输入端分别与所述卷积层、所述的第五可分离卷积层的输出端连接，输出端通过所述全连接层模块与所述输出层连接。The input end of the fourth connection layer is connected to the output end of the convolution layer and the fifth separable convolution layer respectively, and the output end is connected to the output layer through the fully connected layer module.
可选的，所述训练样本包括故障样本和非故障样本，所述故障样本对应的波形片段为所述待训练零序电流波形中故障突变时刻前后各一周波的波形。Optionally, the training samples include fault samples and non-fault samples, and the waveform segments corresponding to the fault samples are waveforms of one cycle before and after the fault mutation moment in the zero-sequence current waveform to be trained.
本申请第二方面提供了一种间歇性电弧故障检测装置，包括：The second aspect of this application provides an intermittent arc fault detection device, including:
分割单元，用于在获取到配电网的待检测零序电流波形后，按照时间顺序将所述待检测零序电流波形分割为若干个波形片段，并对该波形片段进行预处理，得到若干测试样本；The segmentation unit is used to, after acquiring the zero-sequence current waveform to be detected in the distribution network, segment the zero-sequence current waveform to be detected into several waveform segments in chronological order, and preprocess the waveform segments to obtain several waveform segments. test samples;
故障检测单元，用于将各所述测试样本依次输入到间歇性电弧故障检测模型进行故障检测，得到各所述测试样本的故障检测结果；a fault detection unit, configured to sequentially input each of the test samples into the intermittent arc fault detection model for fault detection, and obtain the fault detection results of each of the test samples;
设置单元，用于根据各所述测试样本的故障检测结果依次确定当前测试样本是否发生间歇性电弧故障，若否，则设置从因子SF＝SF+1，若是，则设置主因子PF＝PF+1，且设置SF＝0；A setting unit configured to sequentially determine whether an intermittent arc fault occurs in the current test sample based on the fault detection results of each test sample. If not, set the slave factor SF=SF+1, and if so, set the master factor PF=PF+ 1, and set SF=0;
判定单元，用于当主因子PF大于第一预设阈值且从因子SF大于第二预设阈值时，判定所述配电网发生了短时间歇性电弧故障；A determination unit configured to determine that a short-term intermittent arc fault has occurred in the distribution network when the primary factor PF is greater than the first preset threshold and the slave factor SF is greater than the second preset threshold;
当主因子PF大于第三预设阈值时，判定所述配电网发生了长时间歇性电弧故障；When the main factor PF is greater than the third preset threshold, it is determined that a long-term intermittent arc fault has occurred in the distribution network;
当主因子PF小于第四预设阈值且从因子SF大于所述第二预设阈值时，判定所述配电网未发生间歇性电弧故障。When the primary factor PF is less than the fourth preset threshold and the slave factor SF is greater than the second preset threshold, it is determined that no intermittent arc fault occurs in the distribution network.
可选的，还包括：训练单元，用于：Optionally, also includes: training units for:
获取待训练零序电流波形，所述待训练零序电流波形包括发生了间歇性电弧故障的波形和未发生间歇性电弧故障的波形；Obtaining a zero-sequence current waveform to be trained, where the zero-sequence current waveform to be trained includes a waveform in which intermittent arc faults occur and a waveform in which intermittent arc faults do not occur;
将所述待训练零序电流波形分割为若干个相同大小的波形片段，对该波形片段进行预处理，得到若干训练样本；Divide the zero-sequence current waveform to be trained into several waveform segments of the same size, preprocess the waveform segments to obtain several training samples;
通过所述训练样本训练预置卷积神经网络，得到间歇性电弧故障检测模型。The preset convolutional neural network is trained through the training samples to obtain an intermittent arc fault detection model.
可选的，所述预置卷积神经网络由输入层、5个可分离卷积层、4个连接层、1个卷积层、1个全连接模块和输出层构成，所述全连接模块由若干个全连接层串联构成；Optionally, the preset convolutional neural network consists of an input layer, 5 separable convolutional layers, 4 connection layers, 1 convolutional layer, 1 fully connected module and an output layer. The fully connected module It is composed of several fully connected layers connected in series;
所述输入层的输出端分别与第一可分离卷积层、第二可分离卷积层、第三可分离卷积层的输入端连接；The output end of the input layer is connected to the input end of the first separable convolution layer, the second separable convolution layer, and the third separable convolution layer respectively;
第一连接层与所述第一可分离卷积层、所述第二可分离卷积层的输出端连接，第二连接层与所述第二可分离卷积层、所述第三可分离卷积层的输出端连接；The first connection layer is connected to the output terminals of the first separable convolution layer and the second separable convolution layer, and the second connection layer is connected to the second separable convolution layer and the third separable convolution layer. The output terminal of the convolutional layer is connected;
第四可分离卷积层的输入端与所述第一连接层的输出端连接，输出端通过第三连接层与所述卷积层的输入端连接；The input end of the fourth separable convolution layer is connected to the output end of the first connection layer, and the output end is connected to the input end of the convolution layer through a third connection layer;
第五可分离卷积层的输入端与所述第二连接层的输出端连接；The input terminal of the fifth separable convolution layer is connected to the output terminal of the second connection layer;
第四连接层的输入端分别与所述卷积层、所述的第五可分离卷积层的输出端连接，输出端通过所述全连接层模块与所述输出层连接。The input end of the fourth connection layer is connected to the output end of the convolution layer and the fifth separable convolution layer respectively, and the output end is connected to the output layer through the fully connected layer module.
本申请第三方面提供了一种间歇性电弧故障检测设备，所述设备包括处理器以及存储器；The third aspect of this application provides an intermittent arc fault detection device, which includes a processor and a memory;
所述存储器用于存储程序代码，并将所述程序代码传输给所述处理器；The memory is used to store program code and transmit the program code to the processor;
所述处理器用于根据所述程序代码中的指令执行第一方面任一种所述的间歇性电弧故障检测方法。The processor is configured to execute any one of the intermittent arc fault detection methods described in the first aspect according to instructions in the program code.
本申请第四方面提供了一种计算机可读存储介质，所述计算机可读存储介质用于存储程序代码，所述程序代码用于执行第一方面任一种所述的间歇性电弧故障检测方法。The fourth aspect of the present application provides a computer-readable storage medium, the computer-readable storage medium is used to store program code, and the program code is used to execute any one of the intermittent arc fault detection methods described in the first aspect. .
从以上技术方案可以看出，本申请具有以下优点：It can be seen from the above technical solutions that this application has the following advantages:
本申请提供了一种间歇性电弧故障检测方法，包括：在获取到配电网的待检测零序电流波形后，按照时间顺序将待检测零序电流波形分割为若干个波形片段，并对该波形片段进行预处理，得到若干测试样本；将各测试样本依次输入到间歇性电弧故障检测模型进行故障检测，得到各测试样本的故障检测结果；根据各测试样本的故障检测结果依次确定当前测试样本是否发生间歇性电弧故障，若否，则设置从因子SF＝SF+1，若是，则设置主因子PF＝PF+1，且设置SF＝0；当主因子PF大于第一预设阈值且从因子SF大于第二预设阈值时，判定配电网发生了短时间歇性电弧故障；当主因子PF大于第三预设阈值时，判定配电网发生了长时间歇性电弧故障；当主因子PF小于第四预设阈值且从因子SF大于第二预设阈值时，判定配电网未发生间歇性电弧故障。This application provides an intermittent arc fault detection method, which includes: after obtaining the zero-sequence current waveform to be detected in the distribution network, dividing the zero-sequence current waveform to be detected into several waveform segments in chronological order, and The waveform fragments are preprocessed to obtain several test samples; each test sample is sequentially input into the intermittent arc fault detection model for fault detection, and the fault detection results of each test sample are obtained; the current test sample is determined in turn based on the fault detection results of each test sample. Whether an intermittent arc fault occurs, if not, set the slave factor SF=SF+1, if so, set the master factor PF=PF+1, and set SF=0; when the master factor PF is greater than the first preset threshold and the slave factor When SF is greater than the second preset threshold, it is determined that a short-term intermittent arc fault has occurred in the distribution network; when the main factor PF is greater than the third preset threshold, it is determined that a long-term intermittent arc fault has occurred in the distribution network; when the main factor PF is less than When the fourth preset threshold and slave factor SF is greater than the second preset threshold, it is determined that no intermittent arc fault occurs in the distribution network.
本申请中，通过间歇性电弧故障检测模型进行故障检测，由于卷积神经网络具有较强的自学习能力，可以自适应提取最佳特征和故障检测，有助于提高故障检测结果；并且考虑到实际的间歇性电弧故障具有随机、间歇性，不能被仿真模型完美地模拟，基于主从因子法对各测试样本进一步处理，以提高检测结果的抗干扰性和准确性，从而改善了现有的间歇性电弧故障检测方法存在的检测精度不高的技术问题。In this application, fault detection is carried out through the intermittent arc fault detection model. Since the convolutional neural network has strong self-learning ability, it can adaptively extract the best features and fault detection, which helps to improve the fault detection results; and considering Actual intermittent arc faults are random and intermittent and cannot be perfectly simulated by the simulation model. Each test sample is further processed based on the master-slave factor method to improve the anti-interference and accuracy of the detection results, thereby improving the existing The intermittent arc fault detection method has the technical problem of low detection accuracy.
附图说明Description of the drawings
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其它的附图。In order to explain the embodiments of the present application or the technical solutions in the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are only These are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting any creative effort.
图1为本申请实施例提供的一种间歇性电弧故障检测方法的一个流程示意图；Figure 1 is a schematic flow chart of an intermittent arc fault detection method provided by an embodiment of the present application;
图2为本申请实施例提供的一种间歇性电弧故障检测模型的一个结构示意图；Figure 2 is a schematic structural diagram of an intermittent arc fault detection model provided by an embodiment of the present application;
图3为本申请实施例提供的一种基于主从因子的故障决策过程的一个示意图；Figure 3 is a schematic diagram of a fault decision-making process based on master-slave factors provided by the embodiment of the present application;
图4为本申请实施例提供的一种间歇性电弧故障检测装置的一个结构示意图。Figure 4 is a schematic structural diagram of an intermittent arc fault detection device provided by an embodiment of the present application.
具体实施方式Detailed ways
为了使本技术领域的人员更好地理解本申请方案，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。In order to enable those in the technical field to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only These are part of the embodiments of this application, but not all of them. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.
为了便于理解，请参阅图1，本申请提供的一种间歇性电弧故障检测方法的一个实施例，包括：For ease of understanding, please refer to Figure 1. This application provides an embodiment of an intermittent arc fault detection method, including:
步骤101、在获取到配电网的待检测零序电流波形后，按照时间顺序将待检测零序电流波形分割为若干个波形片段，并对该波形片段进行预处理，得到若干测试样本。Step 101. After obtaining the zero-sequence current waveform to be detected of the distribution network, divide the zero-sequence current waveform to be detected into several waveform segments in time order, and preprocess the waveform segments to obtain several test samples.
可以通过录波装置获取配电网的完整的待检测零序电流波形，基于采样步长fs，按照时间顺序将待检测零序电流波形分割为若干个波形片段，一个波形片段对应一个检测窗口，分割后的待检测零序电流波形由若干个连续的检测窗口构成。The complete zero-sequence current waveform to be detected in the distribution network can be obtained through the wave recording device. Based on the sampling step fs , the zero-sequence current waveform to be detected is divided into several waveform segments in time order. One waveform segment corresponds to one detection window. , the divided zero-sequence current waveform to be detected consists of several continuous detection windows.
然后对各波形片段进行预处理，得到若干测试样本。具体的，基于双三次差值法将各波形片段映射为预设长度的数据序列；对各数据序列进行归一化处理，得到各归一化数据序列；按照时间顺序将各归一化数据序列转换为矩阵形式，得到各波形片段对应的测试样本。Then each waveform segment is preprocessed to obtain several test samples. Specifically, each waveform segment is mapped to a data sequence of a preset length based on the bicubic difference method; each data sequence is normalized to obtain each normalized data sequence; each normalized data sequence is obtained in time order. Convert to matrix form to obtain the test samples corresponding to each waveform segment.
本申请实施例中，优选通过双三次差值法将各波形片段映射为长度为1024的数据序列，纵坐标数值范围不变；然后通过对各数据序列进行归一化处理，将数据序列的纵坐标数值标准化为0～1之间的标量，归一化公式为：In the embodiment of the present application, it is preferable to map each waveform segment into a data sequence with a length of 1024 by using the bicubic difference method, and the ordinate value range remains unchanged; and then normalize each data sequence to convert the ordinate of the data sequence. The coordinate values are normalized to a scalar between 0 and 1. The normalization formula is:
式中，x′i为第k个波形片段Win(k)中第i个采样点的归一化后的值，xi为第k个波形片段Win(k)中第i个采样点的值，max、min分别为取最大值运算符、取最小值运算符。In the formula, x′i is the normalized value of the i-th sampling point in the k-th waveform segment Win(k), and xi is the value of the i-th sampling point in the k-th waveform segment Win(k). , max and min are the maximum value operator and the minimum value operator respectively.
改变原数据波形，将每个检测窗口波形等分为32段，按时间的方向顺序排成32行，每行有32个数据点，因此，测试样本格式是32×32的方形矩阵(1024＝32×32)。Change the original data waveform, divide each detection window waveform into 32 equal segments, and arrange them into 32 rows in the direction of time. Each row has 32 data points. Therefore, the test sample format is a 32×32 square matrix (1024= 32×32).
步骤102、将各测试样本依次输入到间歇性电弧故障检测模型进行故障检测，得到各测试样本的故障检测结果。Step 102: Input each test sample into the intermittent arc fault detection model in turn for fault detection, and obtain the fault detection results of each test sample.
进一步，间歇性电弧故障检测模型的训练过程为：Further, the training process of the intermittent arc fault detection model is:
获取待训练零序电流波形，待训练零序电流波形包括发生了间歇性电弧故障的波形和未发生间歇性电弧故障的波形；将待训练零序电流波形分割为若干个相同大小的波形片段，对该波形片段进行预处理，得到若干训练样本；通过训练样本训练预置卷积神经网络，得到间歇性电弧故障检测模型。Obtain the zero-sequence current waveform to be trained. The zero-sequence current waveform to be trained includes a waveform with intermittent arc faults and a waveform without intermittent arc faults. Divide the zero-sequence current waveform to be trained into several waveform segments of the same size. The waveform fragment is preprocessed to obtain several training samples; the preset convolutional neural network is trained through the training samples to obtain an intermittent arc fault detection model.
进一步，对波形片段进行预处理，得到若干训练样本或测试样本，包括：Further, preprocess the waveform fragments to obtain several training samples or test samples, including:
基于双三次差值法将各波形片段映射为预设长度的数据序列；对各数据序列进行归一化处理，得到各归一化数据序列；按照时间顺序将各归一化数据序列转换为矩阵形式，得到各波形片段对应的训练样本。Based on the bicubic difference method, each waveform segment is mapped into a data sequence of preset length; each data sequence is normalized to obtain each normalized data sequence; each normalized data sequence is converted into a matrix in time order In the form, the training samples corresponding to each waveform segment are obtained.
本申请实施例中，训练样本包括故障样本和非故障样本，故障样本对应的波形片段为待训练零序电流波形中故障突变时刻前后各一周波的波形，即，本申请实施例中，每个波形片段的长度单位为两个工频周期，采样频率可以为10kHz。将间歇性电弧故障完整的波形数据流分割成若干个以两个周波为单位的波形片段，一个波形片段对应一个检测窗口，利用双三次插值法将检测窗口的序列映射成长度为1024的数据序列，纵坐标数值范围不变。然后最小最大归一化方法，将纵坐标数据值标准化成0～1之间的标量。最后，改变原数据波形，将每个检测窗口波形等分为32段，按时间的方向顺序排成32行，每行有32个数据点，因此，训练样本格式是32×32的方形矩阵(1024＝32×32)。In the embodiment of the present application, the training samples include fault samples and non-fault samples, and the waveform segments corresponding to the fault samples are the waveforms of one cycle before and after the fault mutation moment in the zero-sequence current waveform to be trained, that is, in the embodiment of the present application, each The length unit of the waveform segment is two power frequency cycles, and the sampling frequency can be 10kHz. Divide the complete waveform data stream of the intermittent arc fault into several waveform segments with two cycles as the unit. Each waveform segment corresponds to a detection window. The sequence of the detection window is mapped into a data sequence with a length of 1024 using the bicubic interpolation method. , the numerical range of the ordinate remains unchanged. Then the minimum-maximum normalization method normalizes the ordinate data value into a scalar between 0 and 1. Finally, change the original data waveform, divide each detection window waveform into 32 equal segments, and arrange them into 32 rows in the direction of time. Each row has 32 data points. Therefore, the training sample format is a 32×32 square matrix ( 1024=32×32).
进一步，本申请实施例中的预置卷积神经网络由输入层(Input)、5个可分离卷积层(Sep_conv)、4个连接层(Concatenate)、1个卷积层(Conv)、1个全连接模块和输出层(Output)构成，全连接模块由若干个全连接层(Dense)串联构成，具体网络结构和各层输入输出特征图尺寸参数可以参考图2；Furthermore, the preset convolutional neural network in the embodiment of this application consists of an input layer (Input), 5 separable convolution layers (Sep_conv), 4 connection layers (Concatenate), 1 convolution layer (Conv), 1 It consists of a fully connected module and an output layer (Output). The fully connected module is composed of several fully connected layers (Dense) connected in series. The specific network structure and input and output feature map size parameters of each layer can be referred to Figure 2;
输入层的输出端分别与第一可分离卷积层、第二可分离卷积层、第三可分离卷积层的输入端连接；The output end of the input layer is connected to the input end of the first separable convolution layer, the second separable convolution layer, and the third separable convolution layer respectively;
第一连接层与第一可分离卷积层、第二可分离卷积层的输出端连接，第二连接层与第二可分离卷积层、第三可分离卷积层的输出端连接；The first connection layer is connected to the output terminals of the first separable convolution layer and the second separable convolution layer, and the second connection layer is connected to the output terminals of the second separable convolution layer and the third separable convolution layer;
第四可分离卷积层的输入端与第一连接层的输出端连接，输出端通过第三连接层与卷积层的输入端连接；The input end of the fourth separable convolution layer is connected to the output end of the first connection layer, and the output end is connected to the input end of the convolution layer through the third connection layer;
第五可分离卷积层的输入端与第二连接层的输出端连接；The input terminal of the fifth separable convolution layer is connected to the output terminal of the second connection layer;
第四连接层的输入端分别与卷积层、的第五可分离卷积层的输出端连接，输出端通过全连接层模块与输出层连接。The input end of the fourth connection layer is connected to the output end of the convolution layer and the fifth separable convolution layer respectively, and the output end is connected to the output layer through the fully connected layer module.
本申请中的间歇性电弧故障检测模型具有10层网络结构，第2层至第11层为inception架构与Dense block相结合的创新结构，该结构用于代替人工设计的特征提取方法，第12层至第15层是全连接层，主要是起到分类器的作用。The intermittent arc fault detection model in this application has a 10-layer network structure. The 2nd to 11th layers are an innovative structure combining the inception architecture and Dense block. This structure is used to replace the manually designed feature extraction method. The 12th layer The 15th layer is the fully connected layer, which mainly functions as a classifier.
第1层为输入层，输入的测试的样本格式为32×32的方形矩阵；第2-4层为可分离卷积层，分为两个阶段：首先，使用填充技术保持输出特征映射矩阵大小不变的同时，使用单个的3×3卷积核对输入矩阵进行特征提取。其次，利用36个一维卷积核生成36个特征映射矩阵；第5-6层均为连接层，第5层将第2-4层输出的特征映射矩阵进行串联组合成108个特征映射矩阵集合，第6层则将第3层与第4层输出的特征映射矩阵进行串联组合成72个特征映射矩阵集合；第7-8层为可分离卷积层，生成36个输出特征保持不变的特征映射矩阵；同第6层，第9层将第3层与第7层输出的特征映射矩阵进行串联组合；第10层为常规的卷积层；第11层为连接层，将第8层和第10层输出的特征映射矩阵进行串联组合成72个特征映射矩阵集合；第12至15层为全连接层，第12层将所有特征映射矩阵展开成1维张量，第13层的激活函数为“ReLU”，第14层全连接层采用dropout，弃权率优选设置为25％，其目的是为了降低过拟合的可能，第15层的激活函数为“Softmax”且与之对应的损失函数为交叉熵函数。The first layer is the input layer, and the input test sample format is a 32×32 square matrix; the second to fourth layers are separable convolution layers, which are divided into two stages: first, use padding technology to maintain the size of the output feature mapping matrix While remaining unchanged, a single 3×3 convolution kernel is used to extract features from the input matrix. Secondly, 36 one-dimensional convolution kernels are used to generate 36 feature mapping matrices; layers 5-6 are all connection layers, and layer 5 concatenates the feature mapping matrices output from layers 2-4 into 108 feature mapping matrices. The 6th layer concatenates the feature mapping matrices output from the 3rd and 4th layers into a set of 72 feature mapping matrices; the 7th-8th layer is a separable convolution layer, generating 36 output features that remain unchanged. The feature mapping matrix of layer and the feature mapping matrices output by the 10th layer are concatenated and combined into a set of 72 feature mapping matrices; the 12th to 15th layers are fully connected layers, and the 12th layer expands all feature mapping matrices into 1-dimensional tensors. The 13th layer The activation function is "ReLU". The 14th fully connected layer uses dropout. The abstention rate is preferably set to 25%. The purpose is to reduce the possibility of overfitting. The activation function of the 15th layer is "Softmax" and the corresponding The loss function is the cross entropy function.
将各测试样本按时间方向顺序依次输入到间歇性电弧故障检测模型进行故障检测，得到各测试样本的故障检测结果，其中，间歇性电弧故障检测模型输出“0”，表示检测当前测试样本未发生间歇性电弧故障，间歇性电弧故障检测模型输出“1”，表示检测当前测试样本发生了间歇性电弧故障。Each test sample is input into the intermittent arc fault detection model in order in the time direction for fault detection, and the fault detection results of each test sample are obtained. Among them, the intermittent arc fault detection model outputs "0", indicating that the detection of the current test sample has not occurred. Intermittent arc fault, the intermittent arc fault detection model outputs "1", indicating that an intermittent arc fault has occurred in the current test sample.
本申请实施例所提出的神经网络架构具有较强的自学习和联想储存能力，能迅速寻到优化解，鲁棒性更强，适应性更好，无论是在仿真还是在实测样本上都有着很高的准确率；本申请实施例将Inception结构与Dense block的优点相结合，增加了网络的宽度的同时使网络层前后相关联，形成了既能避免过拟合也能避免梯度消失的网络结构；并且，本申请实施例无需故障触发算法就能实时检测间歇性电弧故障。The neural network architecture proposed in the embodiments of this application has strong self-learning and associative storage capabilities, can quickly find optimal solutions, is more robust, and has better adaptability, both in simulation and measured samples. Very high accuracy; the embodiment of this application combines the advantages of the Inception structure and the Dense block to increase the width of the network and correlate the network layers before and after, forming a network that can avoid overfitting and gradient disappearance. structure; and, the embodiment of the present application can detect intermittent arc faults in real time without a fault triggering algorithm.
步骤103、根据各测试样本的故障检测结果依次确定当前测试样本是否发生间歇性电弧故障，若否，则设置从因子SF＝SF+1，若是，则设置主因子PF＝PF+1，且设置SF＝0。Step 103: Determine whether an intermittent arc fault occurs in the current test sample according to the fault detection results of each test sample. If not, set the slave factor SF=SF+1. If so, set the master factor PF=PF+1, and set SF=0.
本申请实施例考虑到，实际的间歇性电弧故障具有随机、间歇性，不能被仿真模型完美地模拟，为了能够提高检测结果的抗干扰性和准确性，实现单节点间歇性电弧故障的检测，提出了主从因子法。请参考图3，图3中的图(a)为间歇性电弧故障的间歇过程示意图，图(b)为理想主因子变化曲线图，图(c)为理想从因子变化曲线图。将完整的零序电流波形(数据流)根据采样步长fs按时间的方向顺序输入到检测窗口中，一个检测参考对应一个测试样本，通过上述步骤可以得到各测试样本的故障检测结果(“0”或“1”)，根据各测试样本的故障检测结果依次确定当前测试样本是否发生间歇性电弧故障，若否，则设置从因子SF＝SF+1，若是，则设置主因子PF＝PF+1，且设置SF＝0。其中，主因子PF和从因子SF初始值为0。The embodiments of this application take into account that actual intermittent arc faults are random and intermittent and cannot be perfectly simulated by the simulation model. In order to improve the anti-interference and accuracy of the detection results and realize the detection of single-node intermittent arc faults, The master-slave factor method was proposed. Please refer to Figure 3. Figure 3 (a) is a schematic diagram of the intermittent process of an intermittent arc fault, Figure (b) is an ideal primary factor change curve, and Figure (c) is an ideal slave factor change curve. The complete zero-sequence current waveform (data flow) is sequentially input into the detection window in the time direction according to the sampling step fs . One detection reference corresponds to one test sample. Through the above steps, the fault detection results of each test sample can be obtained ("0" or "1"), determine whether an intermittent arc fault occurs in the current test sample based on the fault detection results of each test sample. If not, set the slave factor SF = SF + 1. If so, set the master factor PF = PF +1, and set SF=0. Among them, the initial values of the primary factor PF and the secondary factor SF are 0.
步骤104、当主因子PF大于第一预设阈值且从因子SF大于第二预设阈值时，判定配电网发生了短时间歇性电弧故障；当主因子PF大于第三预设阈值时，判定配电网发生了长时间歇性电弧故障；当主因子PF小于第四预设阈值且从因子SF大于第二预设阈值时，判定配电网未发生间歇性电弧故障。Step 104: When the primary factor PF is greater than the first preset threshold and the slave factor SF is greater than the second preset threshold, it is determined that a short-term intermittent arc fault has occurred in the distribution network; when the primary factor PF is greater than the third preset threshold, it is determined that the distribution network has a short-term intermittent arc fault. A long-term intermittent arc fault has occurred in the power grid; when the primary factor PF is less than the fourth preset threshold and the slave factor SF is greater than the second preset threshold, it is determined that an intermittent arc fault has not occurred in the distribution network.
本申请实施例中的第一预设阈值优选为η1×fs，第二预设阈值优选为β×fs，第三预设阈值优选为η2×fs，第四预设阈值优选为η3×fs。η1表示主因子判断是否为短时间歇性故障的阈值系数，可以取值0.2～0.3；η2表示主因子判断是否为长时间歇性故障的阈值系数，可以取值1～2；η3表示主因子判断是否发生间歇性故障的阈值系数，可以取值0.05；β表示从因子判断是否发生间歇性电弧故障的阈值系数，可以取值0.1。In the embodiment of the present application, the first preset threshold is preferably η1 ×fs , the second preset threshold is preferably β×fs , the third preset threshold is preferably η2 ×fs , and the fourth preset threshold is preferably is η3 ×fs . η1 represents the threshold coefficient for the main factor to judge whether it is a short-term intermittent fault, and can take the value 0.2 to 0.3; η2 represents the threshold coefficient for the main factor to judge whether it is a long-term intermittent fault, and can take the value 1 to 2; η3 Represents the threshold coefficient of the main factor to determine whether an intermittent fault occurs, which can take a value of 0.05; β represents the threshold coefficient of the secondary factor to determine whether an intermittent arc fault occurs, which can take a value of 0.1.
当主因子PF>η1×fs且从因子SF>β×fs时，判定配电网发生了短时间歇性电弧故障；When the primary factor PF>η1 ×fs and the slave factor SF>β×fs , it is determined that a short-term intermittent arc fault has occurred in the distribution network;
当主因子PF>η2×fs时，判定配电网发生了长时间歇性电弧故障；When the main factor PF>η2 ×fs , it is determined that a long-term intermittent arc fault has occurred in the distribution network;
当主因子PF<η3×fs且从因子SF>β×fs时，判定配电网未发生间歇性电弧故障，并将主因子PF和从因子SF初始化为0，回到初始状态。When the main factor PF < η3 × fs and the slave factor SF > β × fs , it is determined that intermittent arc faults have not occurred in the distribution network, and the master factor PF and the slave factor SF are initialized to 0 and returned to the initial state.
本申请采用间歇性电弧故障检测模型结合主从因子方法能够分析同一个故障波形下的多个片段信息，二者的结合使用使得判断错误的偶然性大大降低，提高了对间歇性电弧故障的检测准确率。This application uses the intermittent arc fault detection model combined with the master-slave factor method to analyze multiple fragments of information under the same fault waveform. The combined use of the two greatly reduces the chance of judgment errors and improves the accuracy of the detection of intermittent arc faults. Rate.
本申请实施例中，通过间歇性电弧故障检测模型进行故障检测，由于卷积神经网络具有较强的自学习能力，可以自适应提取最佳特征和故障检测，有助于提高故障检测结果；并且考虑到实际的间歇性电弧故障具有随机、间歇性，不能被仿真模型完美地模拟，基于主从因子法对各测试样本进一步处理，以提高检测结果的抗干扰性和准确性，从而改善了现有的间歇性电弧故障检测方法存在的检测精度不高的技术问题。In the embodiment of this application, fault detection is performed through the intermittent arc fault detection model. Since the convolutional neural network has strong self-learning ability, it can adaptively extract the best features and fault detection, which helps to improve the fault detection results; and Considering that actual intermittent arc faults are random and intermittent and cannot be perfectly simulated by the simulation model, each test sample is further processed based on the master-slave factor method to improve the anti-interference and accuracy of the detection results, thus improving the current Some intermittent arc fault detection methods have technical problems such as low detection accuracy.
以上为本申请提供的一种间歇性电弧故障检测方法的一个实施例，以下为本申请提供的一种间歇性电弧故障检测装置的一个实施例。The above is an embodiment of an intermittent arc fault detection method provided by this application, and the following is an embodiment of an intermittent arc fault detection device provided by this application.
请参考图4，本申请实施例提供的一种间歇性电弧故障检测装置，包括：Please refer to Figure 4. An intermittent arc fault detection device provided by an embodiment of the present application includes:
分割单元，用于在获取到配电网的待检测零序电流波形后，按照时间顺序将待检测零序电流波形分割为若干个波形片段，并对该波形片段进行预处理，得到若干测试样本；The segmentation unit is used to divide the zero-sequence current waveform to be detected into several waveform segments in time sequence after acquiring the zero-sequence current waveform to be detected in the distribution network, and to preprocess the waveform segments to obtain several test samples. ;
故障检测单元，用于将各测试样本依次输入到间歇性电弧故障检测模型进行故障检测，得到各测试样本的故障检测结果；A fault detection unit is used to input each test sample into the intermittent arc fault detection model in sequence for fault detection, and obtain the fault detection results of each test sample;
设置单元，用于根据各测试样本的故障检测结果依次确定当前测试样本是否发生间歇性电弧故障，若否，则设置从因子SF＝SF+1，若是，则设置主因子PF＝PF+1，且设置SF＝0；The setting unit is used to determine whether an intermittent arc fault occurs in the current test sample according to the fault detection results of each test sample. If not, set the slave factor SF=SF+1. If so, set the master factor PF=PF+1. And set SF=0;
判定单元，用于当主因子PF大于第一预设阈值且从因子SF大于第二预设阈值时，判定配电网发生了短时间歇性电弧故障；a determination unit configured to determine that a short-term intermittent arc fault has occurred in the distribution network when the primary factor PF is greater than the first preset threshold and the slave factor SF is greater than the second preset threshold;
当主因子PF大于第三预设阈值时，判定配电网发生了长时间歇性电弧故障；When the main factor PF is greater than the third preset threshold, it is determined that a long-term intermittent arc fault has occurred in the distribution network;
当主因子PF小于第四预设阈值且从因子SF大于第二预设阈值时，判定配电网未发生间歇性电弧故障。When the primary factor PF is less than the fourth preset threshold and the slave factor SF is greater than the second preset threshold, it is determined that no intermittent arc fault occurs in the distribution network.
作为进一步地改进，还包括：训练单元，用于：As a further improvement, it also includes: training unit for:
获取待训练零序电流波形，待训练零序电流波形包括发生了间歇性电弧故障的波形和未发生间歇性电弧故障的波形；Obtain the zero-sequence current waveform to be trained. The zero-sequence current waveform to be trained includes a waveform with intermittent arc faults and a waveform without intermittent arc faults;
将待训练零序电流波形分割为若干个相同大小的波形片段，对该波形片段进行预处理，得到若干训练样本；Divide the zero-sequence current waveform to be trained into several waveform segments of the same size, preprocess the waveform segments to obtain several training samples;
通过训练样本训练预置卷积神经网络，得到间歇性电弧故障检测模型。The preset convolutional neural network is trained through training samples to obtain an intermittent arc fault detection model.
作为进一步地改进，预置卷积神经网络由输入层、5个可分离卷积层、4个连接层、1个卷积层、1个全连接模块和输出层构成，全连接模块由若干个全连接层串联构成；As a further improvement, the preset convolutional neural network consists of an input layer, 5 separable convolutional layers, 4 connection layers, 1 convolutional layer, 1 fully connected module and an output layer. The fully connected module consists of several It is composed of fully connected layers in series;
输入层的输出端分别与第一可分离卷积层、第二可分离卷积层、第三可分离卷积层的输入端连接；The output end of the input layer is connected to the input end of the first separable convolution layer, the second separable convolution layer, and the third separable convolution layer respectively;
第一连接层与第一可分离卷积层、第二可分离卷积层的输出端连接，第二连接层与第二可分离卷积层、第三可分离卷积层的输出端连接；The first connection layer is connected to the output terminals of the first separable convolution layer and the second separable convolution layer, and the second connection layer is connected to the output terminals of the second separable convolution layer and the third separable convolution layer;
第四可分离卷积层的输入端与第一连接层的输出端连接，输出端通过第三连接层与卷积层的输入端连接；The input end of the fourth separable convolution layer is connected to the output end of the first connection layer, and the output end is connected to the input end of the convolution layer through the third connection layer;
第五可分离卷积层的输入端与第二连接层的输出端连接；The input terminal of the fifth separable convolution layer is connected to the output terminal of the second connection layer;
第四连接层的输入端分别与卷积层、的第五可分离卷积层的输出端连接，输出端通过全连接层模块与输出层连接。The input end of the fourth connection layer is connected to the output end of the convolution layer and the fifth separable convolution layer respectively, and the output end is connected to the output layer through the fully connected layer module.
本申请实施例中，通过间歇性电弧故障检测模型进行故障检测，由于卷积神经网络具有较强的自学习能力，可以自适应提取最佳特征和故障检测，有助于提高故障检测结果；并且考虑到实际的间歇性电弧故障具有随机、间歇性，不能被仿真模型完美地模拟，基于主从因子法对各测试样本进一步处理，以提高检测结果的抗干扰性和准确性，从而改善了现有的间歇性电弧故障检测方法存在的检测精度不高的技术问题。In the embodiment of this application, fault detection is performed through the intermittent arc fault detection model. Since the convolutional neural network has strong self-learning ability, it can adaptively extract the best features and fault detection, which helps to improve the fault detection results; and Considering that actual intermittent arc faults are random and intermittent and cannot be perfectly simulated by the simulation model, each test sample is further processed based on the master-slave factor method to improve the anti-interference and accuracy of the detection results, thus improving the current Some intermittent arc fault detection methods have technical problems such as low detection accuracy.
本申请实施例还提供了一种间歇性电弧故障检测设备，设备包括处理器以及存储器；An embodiment of the present application also provides an intermittent arc fault detection device, which includes a processor and a memory;
存储器用于存储程序代码，并将程序代码传输给处理器；Memory is used to store program code and transmit the program code to the processor;
处理器用于根据程序代码中的指令执行前述方法实施例中的间歇性电弧故障检测方法。The processor is configured to execute the intermittent arc fault detection method in the foregoing method embodiment according to instructions in the program code.
本申请实施例还提供了一种计算机可读存储介质，计算机可读存储介质用于存储程序代码，程序代码用于执行前述方法实施例中的间歇性电弧故障检测方法。Embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium is used to store program codes. The program codes are used to execute the intermittent arc fault detection method in the foregoing method embodiments.
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the devices and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be described again here.
本申请的说明书及上述附图中的术语“第一”、“第二”、“第三”、“第四”等(如果存在)是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本申请的实施例例如能够以除了在这里图示或描述的那些以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。The terms "first", "second", "third", "fourth", etc. (if present) in the description of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe specific objects. Sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances so that the embodiments of the application described herein can, for example, be practiced in sequences other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, e.g., a process, method, system, product, or apparatus that encompasses a series of steps or units and need not be limited to those explicitly listed. Those steps or elements may instead include other steps or elements not expressly listed or inherent to the process, method, product or apparatus.
应当理解，在本申请中，“至少一个(项)”是指一个或者多个，“多个”是指两个或两个以上。“和/或”，用于描述关联对象的关联关系，表示可以存在三种关系，例如，“A和/或B”可以表示：只存在A，只存在B以及同时存在A和B三种情况，其中A，B可以是单数或者复数。字符“/”一般表示前后关联对象是一种“或”的关系。“以下至少一项(个)”或其类似表达，是指这些项中的任意组合，包括单项(个)或复数项(个)的任意组合。例如，a，b或c中的至少一项(个)，可以表示：a，b，c，“a和b”，“a和c”，“b和c”，或“a和b和c”，其中a，b，c可以是单个，也可以是多个。It should be understood that in this application, "at least one (item)" refers to one or more, and "plurality" refers to two or more. "And/or" is used to describe the relationship between associated objects, indicating that there can be three relationships. For example, "A and/or B" can mean: only A exists, only B exists, and A and B exist simultaneously. , where A and B can be singular or plural. The character "/" generally indicates that the related objects are in an "or" relationship. “At least one of the following” or similar expressions thereof refers to any combination of these items, including any combination of a single item (items) or a plurality of items (items). For example, at least one item (item) of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c" ”, where a, b, c can be single or multiple.
在本申请所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be in electrical, mechanical or other forms.
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or software functional units.
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以通过一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(英文全称：Read-OnlyMemory，英文缩写：ROM)、随机存取存储器(英文全称：Random Access Memory，英文缩写：RAM)、磁碟或者光盘等各种可以存储程序代码的介质。If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions for executing all or part of the steps of the methods described in various embodiments of the application through a computer device (which can be a personal computer, a server, or a network device, etc.). The aforementioned storage media include: U disk, mobile hard disk, read-only memory (English full name: Read-OnlyMemory, English abbreviation: ROM), random access memory (English full name: Random Access Memory, English abbreviation: RAM), magnetic disk Or various media such as CDs that can store program code.
以上所述，以上实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。As mentioned above, the above embodiments are only used to illustrate the technical solution of the present application, but not to limit it. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still make the foregoing technical solutions. The technical solutions described in each embodiment may be modified, or some of the technical features may be equivalently substituted; however, these modifications or substitutions shall not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110838493.6ACN113468704B (en) | 2021-07-23 | 2021-07-23 | An intermittent arc fault detection method and related devices |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110838493.6ACN113468704B (en) | 2021-07-23 | 2021-07-23 | An intermittent arc fault detection method and related devices |
| Publication Number | Publication Date |
|---|---|
| CN113468704A CN113468704A (en) | 2021-10-01 |
| CN113468704Btrue CN113468704B (en) | 2024-02-02 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110838493.6AActiveCN113468704B (en) | 2021-07-23 | 2021-07-23 | An intermittent arc fault detection method and related devices |
| Country | Link |
|---|---|
| CN (1) | CN113468704B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113850025B (en)* | 2021-10-15 | 2025-01-21 | 广东电网有限责任公司东莞供电局 | A distribution network fault sample expansion method, device, terminal and storage medium |
| CN114913142A (en)* | 2022-05-06 | 2022-08-16 | 国网河北省电力有限公司石家庄供电分公司 | A crack detection method for power equipment based on fusion of DCGAN and improved GoogLeNet |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103529316A (en)* | 2013-08-15 | 2014-01-22 | 国家电网公司 | Comprehensive detection method for high resistance ground faults of electric power system |
| CN104375061A (en)* | 2014-12-02 | 2015-02-25 | 国网上海市电力公司 | Intermittent grounding fault detection system of power distribution network |
| CN104597344A (en)* | 2015-01-08 | 2015-05-06 | 上海交通大学 | Fault arc online detecting method based on wavelet first-layer high-frequency component correlation |
| CN107871044A (en)* | 2017-11-13 | 2018-04-03 | 杭州电魂网络科技股份有限公司 | Course method to set up and device |
| CN110398663A (en)* | 2019-07-03 | 2019-11-01 | 东南大学 | A Fault Identification Method for Flexible DC Power Grid Based on Convolutional Neural Network |
| CN110501631A (en)* | 2019-08-19 | 2019-11-26 | 重庆大学 | An online intermittent fault detection and diagnosis method |
| CN110717144A (en)* | 2019-08-19 | 2020-01-21 | 珠海格力电器股份有限公司 | Fault arc detection method, device, terminal and storage medium |
| CN111123048A (en)* | 2019-12-23 | 2020-05-08 | 温州大学 | A serial fault arc detection device and method based on convolutional neural network |
| CN111707908A (en)* | 2020-07-29 | 2020-09-25 | 中国科学技术大学先进技术研究院 | Method, device and storage medium for detecting arc fault in series with multiple load circuits |
| CN113049922A (en)* | 2020-04-22 | 2021-06-29 | 青岛鼎信通讯股份有限公司 | Fault arc signal detection method adopting convolutional neural network |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10503998B2 (en)* | 2016-11-07 | 2019-12-10 | Gracenote, Inc. | Recurrent deep neural network system for detecting overlays in images |
| WO2019216878A1 (en)* | 2018-05-07 | 2019-11-14 | Inhand Networks Inc. | System for locating fault in power distribution network based on mixed mode wave recording |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103529316A (en)* | 2013-08-15 | 2014-01-22 | 国家电网公司 | Comprehensive detection method for high resistance ground faults of electric power system |
| CN104375061A (en)* | 2014-12-02 | 2015-02-25 | 国网上海市电力公司 | Intermittent grounding fault detection system of power distribution network |
| CN104597344A (en)* | 2015-01-08 | 2015-05-06 | 上海交通大学 | Fault arc online detecting method based on wavelet first-layer high-frequency component correlation |
| CN107871044A (en)* | 2017-11-13 | 2018-04-03 | 杭州电魂网络科技股份有限公司 | Course method to set up and device |
| CN110398663A (en)* | 2019-07-03 | 2019-11-01 | 东南大学 | A Fault Identification Method for Flexible DC Power Grid Based on Convolutional Neural Network |
| CN110501631A (en)* | 2019-08-19 | 2019-11-26 | 重庆大学 | An online intermittent fault detection and diagnosis method |
| CN110717144A (en)* | 2019-08-19 | 2020-01-21 | 珠海格力电器股份有限公司 | Fault arc detection method, device, terminal and storage medium |
| CN111123048A (en)* | 2019-12-23 | 2020-05-08 | 温州大学 | A serial fault arc detection device and method based on convolutional neural network |
| CN113049922A (en)* | 2020-04-22 | 2021-06-29 | 青岛鼎信通讯股份有限公司 | Fault arc signal detection method adopting convolutional neural network |
| CN111707908A (en)* | 2020-07-29 | 2020-09-25 | 中国科学技术大学先进技术研究院 | Method, device and storage medium for detecting arc fault in series with multiple load circuits |
| Publication number | Publication date |
|---|---|
| CN113468704A (en) | 2021-10-01 |
| Publication | Publication Date | Title |
|---|---|---|
| CN113049922B (en) | Fault arc signal detection method adopting convolutional neural network | |
| Xiong et al. | Incipient fault identification in power distribution systems via human-level concept learning | |
| CN112200032B (en) | An Attention Mechanism-Based On-Line Monitoring Method for the Mechanical Status of High-Voltage Circuit Breakers | |
| CN105974265B (en) | A kind of electric network fault cause diagnosis method based on svm classifier technology | |
| CN111060813B (en) | Fault diagnosis method and device for operating mechanism of high-voltage circuit breaker and electronic equipment | |
| CN106017879B (en) | Omnipotent breaker mechanical failure diagnostic method based on acoustic signal Fusion Features | |
| CN113468704B (en) | An intermittent arc fault detection method and related devices | |
| CN108680348A (en) | A kind of breaker mechanical fault diagnosis method and system based on random forest | |
| CN109828181A (en) | A kind of transformer winding minor failure detection method based on MODWT | |
| CN116524200A (en) | High-voltage circuit breaker fault diagnosis method based on image recognition | |
| CN111537853A (en) | Intelligent detection method of partial discharge in switchgear based on multi-source heterogeneous data analysis | |
| CN113533904B (en) | A method, device, equipment and medium for detecting a high-resistance ground fault in a distribution network | |
| CN110646708B (en) | 10kV single-core cable early state identification method based on double-layer long-and-short-term memory network | |
| CN116243155A (en) | Breaker fault diagnosis method, device, equipment and storage medium | |
| CN117272143A (en) | Transmission line fault identification method and device based on Gram's angle field and residual network | |
| CN117559373A (en) | Deep flyback ESD protection device and method thereof | |
| CN111898446A (en) | Single-phase-to-ground fault judgment method based on multi-algorithm normalized analysis | |
| CN103558484B (en) | A kind of substation cable is by thunderbolt interference detecting apparatus and method | |
| CN114924220A (en) | Abnormity evaluation method, device and system for electronic transformer digitization device | |
| Zhang | Cable Fault Detection and Diagnosis Method Based on Convolutional Neural Network | |
| CN113945862A (en) | Method, device and equipment for identifying high-resistance grounding fault of power distribution network | |
| CN116796187A (en) | Power transmission line partial discharge detection method | |
| CN112906489A (en) | Transformer machine winding looseness identification method based on LMD and permutation entropy | |
| CN110244127A (en) | A kind of the frequency sweep impedance test device and method of transformer | |
| CN113607992B (en) | A detection wave generation method and related devices for a DC power distribution protection device |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |