WO2023015576A1 - 基于物联网的建筑电气火灾串联故障电弧识别方法及*** - Google Patents

基于物联网的建筑电气火灾串联故障电弧识别方法及*** Download PDF

Info

Publication number
WO2023015576A1
WO2023015576A1 PCT/CN2021/112821 CN2021112821W WO2023015576A1 WO 2023015576 A1 WO2023015576 A1 WO 2023015576A1 CN 2021112821 W CN2021112821 W CN 2021112821W WO 2023015576 A1 WO2023015576 A1 WO 2023015576A1
Authority
WO
WIPO (PCT)
Prior art keywords
current
things
internet
fault
feature extraction
Prior art date
Application number
PCT/CN2021/112821
Other languages
English (en)
French (fr)
Inventor
李成栋
张汉元
***
田晨璐
阎俏
张桂青
Original Assignee
山东建筑大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 山东建筑大学 filed Critical 山东建筑大学
Publication of WO2023015576A1 publication Critical patent/WO2023015576A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/088—Aspects of digital computing

Definitions

  • the present disclosure relates to the technical field of smart buildings of the Internet of Things, and in particular to a method and system for identifying a series fault arc of a building electrical fire based on the Internet of Things.
  • the Internet of Things technology has been widely used in a series of systems such as smart home, security, construction equipment monitoring, electronic distribution frame, card, remote meter reading, professional applications, and real-time or timing collection of electrical parameters. It has become the focus of the application of Internet of Things technology in smart buildings.
  • the present disclosure provides a method and system for identifying arc faults in series in building electrical fires based on the Internet of Things. For high-frequency electrical parameters, through first-order feature extraction and second-order feature extraction, the line series connection is realized. Accurate and efficient identification of arc faults.
  • the first aspect of the present disclosure provides a method for identifying a series fault arc in a building electrical fire based on the Internet of Things.
  • a method for identifying arc faults in series in building electrical fires based on the Internet of Things including the following processes:
  • First-order feature extraction including: when the pre-processed current RMS value is greater than the first preset threshold, determine the load type according to the pre-processed current harmonic;
  • the second-order feature extraction includes: performing complementary set empirical mode decomposition on the high-frequency periodic current of the suspected abnormal data after the first-order extraction, combining the load type, and obtaining the series fault according to the comparison between the decomposition result and the preset fault eigenmode function Arc results.
  • the preprocessing includes: respectively performing pasteurization filtering and normalization processing on the acquired current effective value, high-frequency periodic current and current harmonics.
  • the identified feature data is fitted with a Gaussian function to obtain the load type,
  • the load type is resistive or resistive or nonlinear.
  • the second-order feature extraction process after the first-order feature recognition is passed, continuously capture multiple continuous cycle electrical parameters, and perform complementary set empirical mode decomposition on the high-frequency periodic current of suspected abnormal data after the first-order feature extraction;
  • eigenmode functions are obtained according to the time scale characteristics of the data itself, and the results of series fault arcs are obtained according to the comparison between eigenmode functions and fault eigenmode functions.
  • D(x m ,y n ) is the distance cost matrix composed of the corresponding elements of S 1 and S 2 , the smaller ⁇ (S 1 ,S 2 ), the greater the difference between the fault eigenmode function and the eigenmode function to be identified The higher the similarity.
  • complementary set empirical mode decomposition includes:
  • each pair of white noise has the same amplitude but opposite polarity
  • the final IMF component is obtained by averaging the IMF+ function and the IMF- function of the two sets of integrations.
  • R k (t) R k-1 (t)-IMF k
  • E 1 and E k are operator expectations
  • ⁇ i (t) is a typical Gaussian white noise signal with zero mean
  • ⁇ k is the coefficient
  • I is the number of typical Gaussian white noise signals whose mean is zero.
  • the second aspect of the present disclosure provides an Internet of Things-based identification system for building electrical fire series fault arcs.
  • a building electrical fire series fault arc identification system based on the Internet of Things including:
  • the data acquisition module is configured to: acquire the current effective value, high-frequency periodic current and current harmonics of the line to be detected;
  • the preprocessing module is configured to: preprocess the acquired data
  • the first-order feature extraction module is configured to: determine the load type according to the preprocessed current harmonics when the preprocessed current effective value is greater than the first preset threshold;
  • the second-order feature extraction module is configured to: conduct complementary set empirical mode decomposition on the high-frequency periodic current of suspected abnormal data after the first-order extraction, and combine the load type, according to the comparison between the decomposition result and the preset fault eigenmode function Get series arc fault results.
  • the third aspect of the present disclosure provides a computer-readable storage medium, on which a program is stored, and when the program is executed by a processor, the method for identifying a series fault arc in a building electrical fire based on the Internet of Things as described in the first aspect of the present disclosure is implemented. in the steps.
  • the fourth aspect of the present disclosure provides an electronic device, including a memory, a processor, and a program stored in the memory and operable on the processor, and the processor implements the program described in the first aspect of the present disclosure when executing the program. Steps in the IoT-based identification method for series fault arcs in building electrical fires.
  • the first-order is rough extraction, and is realized by double thresholds of current effective value and harmonic Preliminary judgment (making full use of the appearance features, that is, the electrical parameters that can be obtained directly); the second-order extraction deeply excavates the high-frequency periodic current, and realizes more accurate and efficient identification of line series fault arcs.
  • the method, system, medium or electronic device described in this disclosure after passing the first-order feature recognition, continues to capture multiple continuous cycle electrical parameters, and performs complementary collection of high-frequency periodic currents of suspected abnormal data after the first-order extraction Empirical mode decomposition enables more accurate identification of series fault arcs.
  • the starting moment of the arc occurrence point is random, and there may be up and down half-wave or front and back half-cycle interference.
  • the method, system, medium or electronic device described in this disclosure, through the CEEMD method, can maximize The impact of the randomness of the arc occurrence point on the recognition results is reduced to a certain extent.
  • FIG. 1 is a schematic flowchart of a method for identifying a series fault arc in a building electrical fire based on the Internet of Things provided by Embodiment 1 of the present disclosure.
  • FIG. 2 is a schematic flow chart of performing EMD decomposition on p positive-noise mixed signals provided by Embodiment 1 of the present disclosure.
  • FIG. 3 is a schematic flow chart of performing EMD decomposition on p negative noise mixed signals provided by Embodiment 1 of the present disclosure.
  • FIG. 4 is a radar distribution diagram of a recognition result provided by Embodiment 1 of the present disclosure.
  • FIG. 5 is a schematic diagram of the accuracy of recognition results provided by Embodiment 1 of the present disclosure.
  • FIG. 6 is a schematic diagram of connection of an experimental circuit provided by Embodiment 1 of the present disclosure.
  • FIG. 7 is a schematic diagram of a resistive load series arc current waveform and a normal current waveform provided by Embodiment 1 of the present disclosure.
  • FIG. 8 is a schematic diagram of a current waveform and a normal current waveform of a resistive-inductive load series arc provided by Embodiment 1 of the present disclosure.
  • FIG. 9 is a schematic diagram of a current waveform and a normal current waveform of a nonlinear load series arc current waveform provided by Embodiment 1 of the present disclosure.
  • FIG. 10 is an exploded view of the resistive load circuit in normal operation and the series fault arc CEEMD provided by Embodiment 1 of the present disclosure.
  • FIG. 11 is an exploded view of the resistive-inductive load circuit in normal operation and the series fault arc CEEMD provided by Embodiment 1 of the present disclosure.
  • Fig. 12 is a CEEMD exploded view of the non-linear load circuit in normal operation and the series fault arc provided by Embodiment 1 of the present disclosure.
  • Fig. 13 is a schematic structural diagram of a building electrical fire series arc fault identification system based on the Internet of Things provided by Embodiment 5 of the present disclosure.
  • Fig. 14 is a schematic workflow diagram of a system for identifying series fault arcs in building electrical fires based on the Internet of Things provided by Embodiment 5 of the present disclosure.
  • Embodiment 1 of the present disclosure provides a method for identifying a series fault arc in a building electrical fire based on the Internet of Things, including the following process:
  • First-order feature extraction including: when the pre-processed current RMS value is greater than the first preset threshold, determine the load type according to the pre-processed current harmonic;
  • the second-order feature extraction includes: performing complementary set empirical mode decomposition on the high-frequency periodic current of the suspected abnormal data after the first-order extraction, combining the load type, and obtaining the series fault according to the comparison between the decomposition result and the preset fault eigenmode function Arc results.
  • the high-frequency electrical parameter monitor collects the electrical parameters of the tested line in real time, and performs normalization processing after Pasteur filtering to eliminate the interference of interference signals on the identification of series arcs and abnormal contact resistance, realize standardization, and eliminate differences under the same type of load. The effect of power improves the accuracy of recognition.
  • S2 First-order feature extraction and recognition.
  • the first-order extraction is rough extraction, and the possible series fault arc is included in the second-order extraction range with a relatively wide threshold, and the load type is determined at the same time.
  • the high-frequency periodic current data of different types of loads operating under normal conditions are obtained.
  • the threshold 1 of the first-order extraction and identification mechanism is triggered, the first cycle data will be reported according to the identification characteristics.
  • Data fitting is carried out through the Gaussian function, and the fitting produces 3 types. Output results: resistive, resistive, nonlinear, determine the load type.
  • the first-order extraction is a double-threshold judgment.
  • Threshold 1 is the current RMS change ⁇ 0.15A after preprocessing
  • threshold 2 is the proportion of each harmonic in the current harmonics. See Table 1 for details.
  • threshold 1 is taken as 0.15A
  • threshold 2 harmonic proportion value is obtained from simulation experiments.
  • P n is the proportion of the nth harmonic
  • N n is the value of the nth harmonic
  • It is the sum of the values of the 2nd to 13th harmonics except the fundamental wave (1Exclude the fundamental wave: the fundamental wave value is high, which affects the proportion of high-order harmonics; 2The harmonics above the 13th order basically exist in the form of 0, which is related to the overall identification little sex).
  • the second-order extraction continuously extracts multi-cycle electrical parameter information for specific fault types, further confirms and verifies, and at the same time integrates the concept with the data in step 4, and multi-party data is mutually verified.
  • the first-order feature recognition After the first-order feature recognition is passed, continue to capture the continuous electrical parameters of multiple cycles thereafter, and perform complementary set empirical mode decomposition (CEEMD, the specific decomposition process is attached) for the high-frequency periodic current of the suspected abnormal data after the first-order extraction.
  • CEEMD complementary set empirical mode decomposition
  • IMFx Multiple intrinsic mode functions IMFx are obtained from its own time scale characteristics. After simulation experiments, the intrinsic mode function IMF3 determined by the Pearson correlation coefficient has the strongest correlation with series arc faults.
  • Complementary set empirical mode decomposition including the following process:
  • IMF 2 is calculated as:
  • R k (t) R k-1 (t)-IMF k
  • E k is the operator expectation
  • ⁇ i (t) is a typical Gaussian white noise signal with zero mean
  • ⁇ k is a coefficient that allows the selection of SNR at each stage
  • x(t) is the original high frequency Periodic current sequence.
  • test set For the test set: perform second-order feature extraction, obtain the intrinsic mode function, and input DTW (Dynamic Time Warping) together with the standard data of the arc feature training set corresponding to the load type.
  • DTW Dynamic Time Warping
  • DTW is not constrained by time series length, frequency and other conditions.
  • the distance is used as a measure to characterize the similarity of any two time signals.
  • the distance cost matrix D is composed of the corresponding elements of the above two time series:
  • d(x m ,y n ) is the Euclidean distance between the two sequences:
  • Pearson correlation coefficient between two variables is defined as the quotient of the covariance and standard deviation between the two variables:
  • the radar chart of the recognition result is shown in Figure 4, and the accuracy of the recognition result is shown in Figure 5.
  • S5 Data fusion verification.
  • a series fault arc or abnormal contact resistance will cause the temperature of the line to rise, flashes near the fault point, and abnormal air pressure.
  • the cost of accurately locating the fault point is too high, and it is not necessary at the same time. Therefore, the temperature rise of the line is selected as the auxiliary identification feature, but because the temperature extension is affected by the time scale Obviously, and the impact of external interference is great, so the confidence factor is not set for the auxiliary identification feature in this identification method.
  • Figure 6 it is a schematic diagram of the specific experimental circuit
  • Figure 7, Figure 8 and Figure 9 are schematic diagrams of the series arc current waveform and normal current waveform of resistive load, resistive inductive load and nonlinear load, respectively.
  • Figure 10, Figure 11 and Figure 12 are the CEEMD decomposition diagrams of resistive load circuit, resistive inductive load and nonlinear load in normal operation and series fault arc respectively. According to the obtained load type, combined with the CEEMD decomposition results of normal and fault states corresponding to the load , the final fault identification result can be obtained.
  • Embodiment 2 of the present disclosure provides a building electrical fire series fault arc identification system based on the Internet of Things, including:
  • the data acquisition module is configured to: acquire the current effective value, high-frequency periodic current and current harmonics of the line to be detected;
  • the preprocessing module is configured to: preprocess the acquired data
  • the first-order feature extraction module is configured to: determine the load type according to the preprocessed current harmonics when the preprocessed current effective value is greater than the first preset threshold;
  • the second-order feature extraction module is configured to: conduct complementary set empirical mode decomposition on the high-frequency periodic current of suspected abnormal data after the first-order extraction, and combine the load type, according to the comparison between the decomposition result and the preset fault eigenmode function Get series arc fault results.
  • the working method of the system is the same as the IoT-based identification method for building electrical fire series fault arc identification provided in Embodiment 1, and will not be repeated here.
  • Embodiment 3 of the present disclosure provides a computer-readable storage medium, on which a program is stored.
  • the program is executed by a processor, the method for identifying arc faults in series in electrical fires in buildings based on the Internet of Things as described in Embodiment 1 of the present disclosure is implemented. in the steps.
  • Embodiment 4 of the present disclosure provides an electronic device, including a memory, a processor, and a program stored in the memory and operable on the processor.
  • the processor executes the program, the implementation as described in Embodiment 1 of the present disclosure Steps in the IoT-based identification method for series fault arcs in building electrical fires.
  • Embodiment 5 of the present disclosure provides an Internet of Things-based building electrical fire series fault arc identification system, including: a high-frequency electrical parameter collector, a host computer and an intelligent circuit breaker.
  • the high-frequency electrical parameter collector is a new type of terminal equipment based on the Internet of Things technology. It has functions such as data collection and data communication. The collector is easy to install and use.
  • the current RMS value, high-frequency cycle current and current harmonics of the hourly detection line are sampled at 128 points per cycle, and the sampling frequency is 6400Hz.
  • the data is reported to the host computer based on the event trigger mechanism.
  • the host computer executes the steps in the method for identifying arc faults in series of electrical fires in buildings based on the Internet of Things described in Embodiment 1 according to the acquired data, which will not be repeated here.
  • each room or every few rooms will be equipped with a distribution box, the floor will be equipped with a separate floor distribution box, and the building's main incoming line will be installed with a building distribution box.
  • Electrical fire monitoring devices are installed at multiple key distribution boxes in multiple buildings, each device is directly connected to the cloud server, and the host computer program is deployed on the cloud, which can realize real-time monitoring and fire hazards of electrical fires in multiple buildings position.
  • the detection target of the system is the whole building, and the monitoring module used is small in size and can be installed in the distribution box without changing the existing lines.
  • Using the identification algorithm described in Embodiment 1 can effectively realize the identification of line series fault arcs, and provide real and effective technical support for fire safety personnel, building operation and maintenance personnel, users, etc. to discover fire hazards in time.
  • the first-order feature extraction and identification are implemented in local embedded devices, and the second-order feature extraction and identification are implemented on cloud servers.
  • the first-order feature extraction and identification are implemented in local embedded devices, and the second-order feature extraction and identification are implemented on cloud servers.
  • the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Accordingly, the present disclosure may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) having computer-usable program code embodied therein.
  • a computer-usable storage media including but not limited to disk storage, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions
  • the device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random AccessMemory, RAM), etc.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Testing Relating To Insulation (AREA)

Abstract

本公开提供了一种基于物联网的建筑电气火灾串联故障电弧识别方法及***，获取待检测线路的电流有效值、高频周期电流和电流谐波；对获取的数据进行预处理；一阶特征提取，包括：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；二阶特征提取，包括：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果；本公开针对高频电气参数，通过一阶特征提取和二阶特征提取，实现了线路串联故障电弧的准确和高效识别。

Description

基于物联网的建筑电气火灾串联故障电弧识别方法及*** 技术领域
本公开涉及物联网智慧建筑技术领域，特别涉及一种基于物联网的建筑电气火灾串联故障电弧识别方法及***。
背景技术
本部分的陈述仅仅是提供了与本公开相关的背景技术，并不必然构成现有技术。
随着工业电气设备和家用电器种类不断增多，功能不断强大，随之而来的建筑电气火灾事故发生率也始终占据高位，多项标准中明确规定了应设置电气火灾监控***的场所，相比其他火灾诱因，串联故障电弧的隐蔽性极强，故障状态下线路电流不会发生剧烈变化，断路器、熔断器等保护设备无法发挥保护作用，故障持续发生时，集聚大量热量，从而引发电气火灾，造成重大事故。
目前，物联技术已经在智能家居、安防、建筑设备监控、电子配线架、一卡通、远传抄表、专业应用等一系列***中实现了广泛的应用，而针对电气参数的实时或者定时采集也已成为物联网技术在智慧建筑中应用的重点。
发明人发现，传统的电气火灾监测主要是以单一的表观信息为特征量，通过线缆温度、烟雾浓度等进行识别，往往火灾已经发生或最佳扑救机会失去后，相应***才会后知后觉，而且随着用电设备种类的增加，导致传统的检测方法漏报、误报概率大幅增加。
发明内容
为了解决现有技术的不足，本公开提供了一种基于物联网的建筑电气火灾串联故障电弧识别方法及***，针对高频电气参数，通过一阶特征提取和二阶特征提取，实现了线路串联故障电弧的准确和高效识别。
为了实现上述目的，本公开采用如下技术方案：
本公开第一方面提供了一种基于物联网的建筑电气火灾串联故障电弧识别方法。
一种基于物联网的建筑电气火灾串联故障电弧识别方法，包括以下过程：
获取待检测线路的电流有效值、高频周期电流和电流谐波；
对获取的数据进行预处理；
一阶特征提取，包括：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；
二阶特征提取，包括：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果。
进一步的，预处理包括：对获取的电流有效值、高频周期电流和电流谐波分别进行巴氏滤波和归一化处理。
进一步的，一阶特征提取过程中：当预处理后的电流有效值大于第一预设阈值时，根据第一周波数据，将识别到的特征数据通过高斯函数进行数据拟合，得到负载类型，所述负载类型为阻性或者阻感或者非线性。
进一步的，二阶特征提取过程中：一阶特征识别通过后，持续捕捉此后连续多个周波电气参数，对一阶特征提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解；
根据数据自身时间尺度特征获取多个本征模态函数，根据本征模态函数与故障本征模态函数的对比，得到串联故障电弧结果。
更进一步的，故障本征模态函数为S 1＝[x 1,x 2,x 3…x m]，待识别本征模态函数为S 2＝[y 1,y 2,y 3…y n]，两序列间的最优路径为：
ξ(S 1,S 2)＝D(x m,y n)+min[ξ(x m-1,x n),ξ(x m-1,x n-1),ξ(x m,x n-1)
其中，D(x m,y n)为S 1和S 2对应元素组成的距离代价矩阵，ξ(S 1,S 2)越小，故障本征模态函数与待识别本征模态函数的相似度越高。
进一步的，互补集合经验模态分解，包括：
在原始信号中加入p对的高斯白噪声信号，每对白噪声幅值相同但极性相反；
对p个正噪声混合信号进行经验模态分解，并进行集成平均，得到一组IMF+函数；
对p个负噪声混合信号进行经验模态分解，并进行集成平均，得到一组IMF-函数；
对两组集成的IMF+函数和IMF-函数求平均值，得到最终的IMF分量。
更进一步的，第k+1个分解结果IMF k+1，具体为：
Figure PCTCN2021112821-appb-000001
其中，R k(t)＝R k-1(t)-IMF k，E 1和E k均为算子期望，ω i(t)为均值是零的典型高斯白噪声信号，ε k为系数，I为均值是零的典型高斯白噪声信号数量。
本公开第二方面提供了一种基于物联网的建筑电气火灾串联故障电弧识别***。
一种基于物联网的建筑电气火灾串联故障电弧识别***，包括：
数据获取模块，被配置为：获取待检测线路的电流有效值、高频周期电流和电流谐波；
预处理模块，被配置为：对获取的数据进行预处理；
一阶特征提取模块，被配置为：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；
二阶特征提取模块，被配置为：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果。
本公开第三方面提供了一种计算机可读存储介质，其上存储有程序，该程序被处理器执行时实现如本公开第一方面所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤。
本公开第四方面提供了一种电子设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的程序，所述处理器执行所述程序时实现如本公开第一方面所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤。
与现有技术相比，本公开的有益效果是：
1、本公开所述的方法、***、介质或电子设备，针对高频电气参数，通过一阶特征提取和二阶特征提取，一阶为粗提取，通过电流有效值和谐波的双重阈值实现初步判定(较为充分的利用了表象特征，即直接能够获取的电气参数)；二阶提取则对高频周期电流进行了进行深入挖掘，实现了线路串联故障电弧的更准确和高效识别。
2、本公开所述的方法、***、介质或电子设备，一阶特征识别通过后，持续捕捉此后连续多个周波电气参数，对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，实现了串联故障电弧的更准确识别。
3、串联故障电弧发生时，电弧发生点的起始时刻是随机的，可能会出现上下半波或前后半周期干扰，本公开所述的方法、***、介质或电子设备，通过CEEMD方法，最大程度降低了电弧发生点随机性对识别结果的影响。
4、本公开所述的方法、***、介质或电子设备，在一阶特征提取后巧妙利用第一周波数据实现了负载类型的区分，在负载类型分类的基础上进一步进行电弧识别，进一步的提高了电弧识别的准确性。
附图说明
构成本公开的一部分的说明书附图用来提供对本公开的进一步理解，本公开的示意性实施例及其说明用于解释本公开，并不构成对本公开的不当限定。
图1为本公开实施例1提供的基于物联网的建筑电气火灾串联故障电弧识别方法的流程示意图。
图2为本公开实施例1提供的对p个正噪声混合信号进行EMD分解的流程示意图。
图3为本公开实施例1提供的对p个负噪声混合信号进行EMD分解的流程示意图。
图4为本公开实施例1提供的识别结果雷达分布图。
图5为本公开实施例1提供的识别结果准确度示意图。
图6为本公开实施例1提供的实验电路连接示意图。
图7为本公开实施例1提供的阻性负载串联电弧电流波形和正常电流波形示意图。
图8为本公开实施例1提供的阻感负载串联电弧电流波形和正常电流波形示意图。
图9为本公开实施例1提供的非线性负载串联电弧电流波形和正常电流波形示意图。
图10为本公开实施例1提供的阻性负载电路正常工作与串联故障电弧CEEMD分解图。
图11为本公开实施例1提供的阻感负载电路正常工作与串联故障电弧CEEMD分解图。
图12为本公开实施例1提供的非线性负载电路正常工作与串联故障电弧CEEMD分解图。
图13为本公开实施例5提供的基于物联网的建筑电气火灾串联故障电弧识别***的结构示意图。
图14为本公开实施例5提供的基于物联网的建筑电气火灾串联故障电弧识别***的工作流程示意图。
具体实施方式
下面结合附图与实施例对本公开作进一步说明。
应该指出，以下详细说明都是例示性的，旨在对本公开提供进一步的说明。除非另有指明，本文使用的所有技术和科学术语具有与本公开所属技术领域的普通技术人员通常理解的相同含义。
需要注意的是，这里所使用的术语仅是为了描述具体实施方式，而非意图限制根据本公开的示例性实施方式。如在这里所使用的，除非上下文另外明确指出，否则单数形式也意图包括复数形式，此外，还应当理解的是，当在本说明书中使用术语“包含”和/或“包括”时，其指明存在特征、步骤、操作、器件、组件和/或它们的组合。
在不冲突的情况下，本公开中的实施例及实施例中的特征可以相互组合。
实施例1：
如图1所示，本公开实施例1提供了一种基于物联网的建筑电气火灾串联故障电弧识别方法，包括以下过程：
获取待检测线路的电流有效值、高频周期电流和电流谐波；
对获取的数据进行预处理；
一阶特征提取，包括：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；
二阶特征提取，包括：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果。
具体的，包括以下内容：
S1：数据预处理。高频电气参数监测器实时采集被测线路电气参数，经巴氏滤波后，进行归一化处理，排除干扰信号对串联电弧和接触电阻异常识别的干扰，实现标准化，消除同种类型负载下不同功率的影响，提高识别的准确度。
S2：一阶特征提取识别。一阶提取为粗提取，以相对宽泛的阈值将可能的串联故障电弧纳入二阶提取范围，同时确定负载类型。
根据模拟实验获取不同类型负载正常状态下运行的高频周期电流数据，一阶提取识别机制阈值1触发后，将上报第一周波数据按识别特征通过高斯函数进行数据拟合，拟合产生3种输出结果：阻性、阻感、非线性，确定负载类型。
其中，一阶提取为双阈值判定，阈值1为预处理后电流有效值变化≥0.15A，阈值2为电流谐波中各次谐波占比，详见表1。
表1：
Figure PCTCN2021112821-appb-000002
由于一般建筑常规用电器功率为几十到几百瓦不等，正常工作电流有效值一般在0.2A以上，故阈值1取为0.15A，阈值2谐波占比数值由模拟实验得出。
其中，n次谐波占比计算方法：
Figure PCTCN2021112821-appb-000003
其中，P n为n次谐波占比，N n为n次谐波数值，
Figure PCTCN2021112821-appb-000004
为除基波外2～13次谐波数值的总和(①排除基波：基波数值较高，影响高次谐波占比数值；②13次以上谐波基本以0形式存在，与整体识别关联性不大)。
S3：二阶特征提取识别。
二阶提取针对具体故障类型连续提取多周波电气参数信息，进一步确认、核实，同时与步骤4数据融合理念，多方数据互为验证。
一阶特征识别通过后，持续捕捉此后连续多个周波电气参数，对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解(CEEMD，具体分解流程附后)，根据数据自身时间尺度特征获取多个本征模态函数IMFx，经模拟实验，由皮尔逊相关系数确定本征模态函数IMF3与串联电弧故障关联性最强。
互补集合经验模态分解，包括以下过程：
S3.1：在原始信号中加入p对的高斯白噪声信号，每对白噪声幅值相同但极性相反；
S3.2：对p个正噪声混合信号进行EMD分解(如图2所示)，并进行集成平均，得到一组IMF+函数；
S3.3：对p个负噪声混合信号进行EMD分解(如图3所示)，并进行集成平均，得到一组IMF-函数；
S3.4：对两组集成的IMF(±)求平均值，得到最终的IMF分量。
设x(t)为原周期电流序列，互补集合经验模态分量IMF 1的计算方法为：
Figure PCTCN2021112821-appb-000005
求第一重R 1(t)：
R 1(t)＝x(t)-IMF 1
IMF 2的计算方法为：
Figure PCTCN2021112821-appb-000006
求第k重R k(t)：
R k(t)＝R k-1(t)-IMF k
IMF k+1的计算方法为：
Figure PCTCN2021112821-appb-000007
重复直至达到人为限定的分解次数终止条件。
上述算式中，E k即为算子期望，ω i(t)为均值是零的典型高斯白噪声信号，ε k为系数允许在每个阶段选择信噪比，x(t)为原高频周期电流序列。
S4：匹配识别过程：
S4.1：针对训练集：对三种不同负载类型下的串联电弧故障数据分别进行一阶、二阶特征提取，获取本征模态函数，并作为电弧特征训练集标准数据，用于实际识别时做计算的对比量。
S4.2：针对测试集：进行二阶特征提取，获取本征模态函数后，与对应负载类型的电弧特征训练集标准数据一同输入DTW(Dynamic Time Warping，动态时间规整)。
DTW不受时间序列长度、频率等条件的约束，通过动态匹配最优路径，以距离作为衡量标准表征任意两个时许信号的相似性。
将电弧特征训练集标准数据作为S 1：
S 1＝[x 1,x 2,x 3…x m]
将待识别的本征模态函数作为S 2：
S 2＝[y 1,y 2,y 3…y n]
以上述两时间序列对应元素组成距离代价矩阵D：
Figure PCTCN2021112821-appb-000008
其中，d(x m,y n)为此两序列的欧氏距离：
Figure PCTCN2021112821-appb-000009
两序列间的最优路径ξ(S 1,S 2)：
ξ(S 1,S 2)＝D(x m,y n)+min[ξ(x m-1,x n),ξ(x m-1,x n-1),ξ(x m,x n-1)
其中，
Figure PCTCN2021112821-appb-000010
Figure PCTCN2021112821-appb-000011
两时间序列“相似度”越高，ξ(S 1,S 2)的数值越接近零，反之越大。
两个变量之间的皮尔逊相关系数定义为两个变量之间的协方差和标准差的商：
Figure PCTCN2021112821-appb-000012
S4.3：相似度判别：DTW的输出结果决定了电弧识别的结果，相似度判别条件见下表：
Figure PCTCN2021112821-appb-000013
识别结果的雷达图如图4所示，识别结果的准确度如图5所示。
S5：数据融合核实。串联故障电弧或接触电阻异常都会出现线路温度升高，故障点附近闪光、气压异常等表现。为进一步验证S4的识别结果，考虑非侵入式识别的特性，精确实现故障点定位投入成本过高，同时也非必要，因此选择线路温升作为辅助识别特征，但由于温度的延展受时间尺度影响明显，且外界干扰影响大，故辅助识别特征在本识别方法中未设置置信因数。
如图6所示，为具体的实验电路示意图，图7、图8和图9分别为阻性负载、阻感负载和非线性负载串联电弧电流波形和正常电流波形示意图。
图10、图11和图12分别为阻性负载电路、阻感负载和非线性负载正常工作与串联故障电弧CEEMD分解图，根据得到的负载类型，结合负载对应的正常和故障状态的CEEMD分解结果，可以得到最终的故障识别结果。
实施例2：
本公开实施例2提供了一种基于物联网的建筑电气火灾串联故障电弧识别***，包括：
数据获取模块，被配置为：获取待检测线路的电流有效值、高频周期电流和电流谐波；
预处理模块，被配置为：对获取的数据进行预处理；
一阶特征提取模块，被配置为：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；
二阶特征提取模块，被配置为：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果。
所述***的工作方法与实施例1提供的基于物联网的建筑电气火灾串联故障电弧识别方法相同，这里不再赘述。
实施例3：
本公开实施例3提供了一种计算机可读存储介质，其上存储有程序，该程序被处理器执行时实现如本公开实施例1所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤。
实施例4：
本公开实施例4提供了一种电子设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的程序，所述处理器执行所述程序时实现如本公开实施例1所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤。
实施例5：
如图13和14所示，本公开实施例5提供了一种基于物联网的建筑电气火灾串联故障电弧识别***，包括：高频电气参数采集器、上位机和智能断路器。
高频电气参数采集器是基于物联网技术的新型末端设备，具备数据采集、数据通信等功能，采集器安装使用方便，以物理方式安装于配电箱内，以二次侧线圈感应的方式24小时检测线路的电流有效值、高频周期电流和电流谐波，每个周波采样128个点，采样频率6400Hz，基于事件触发机制实现数据向上位机的上报。
所述上位机根据获取到的数据执行实施例1所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤，这里不再赘述。
对于一栋完整建筑而言，每个房间或每几个房间都会设置配电箱，楼层会单独设置楼层配电箱，楼宇总进线处会设置楼宇配电箱。在多栋建筑的多个关键配电箱位置都设立电气火灾监测装置，每个装置直接与云端服务器连接，上位机程序部署在云端，就可以实现对多栋建筑电气火灾的实时监测与火灾隐患定位。
***的检测目标为整栋建筑，采用的监测模块体积小，可安装在配电箱内，无需改动已有线路。采用实施例1所述的识别算法可有效的实现线路串联故障电弧的识别，为消防安全人 员、楼宇运维人员、用户等及时发现火灾隐患提供真实有效的技术支撑。
作为可选的实施方式，一阶特征提取识别在本地嵌入式设备中实现，二阶特征提取识别在云端服务器实现，本领域技术人员可以根据具体工况进行一阶和二阶提取的设备选择，这里不再赘述。
本领域内的技术人员应明白，本公开的实施例可提供为方法、***、或计算机程序产品。因此，本公开可采用硬件实施例、软件实施例、或结合软件和硬件方面的实施例的形式。而且，本公开可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器和光学存储器等)上实施的计算机程序产品的形式。
本公开是参照根据本公开实施例的方法、设备(***)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的程序可存储于一计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)或随机存储记忆体(Random AccessMemory，RAM)等。
以上所述仅为本公开的优选实施例而已，并不用于限制本公开，对于本领域的技术人员来说，本公开可以有各种更改和变化。凡在本公开的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本公开的保护范围之内。

Claims (10)

  1. 一种基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：包括以下过程：
    获取待检测线路的电流有效值、高频周期电流和电流谐波；
    对获取的数据进行预处理；
    一阶特征提取，包括：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；
    二阶特征提取，包括：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果。
  2. 如权利要求1所述的基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：
    预处理包括：对获取的电流有效值、高频周期电流和电流谐波分别进行巴氏滤波和归一化处理。
  3. 如权利要求1所述的基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：
    一阶特征提取过程中：当预处理后的电流有效值大于第一预设阈值时，根据第一周波数据，将识别到的特征数据通过高斯函数进行数据拟合，得到负载类型，所述负载类型为阻性或者阻感或者非线性。
  4. 如权利要求1所述的基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：
    二阶特征提取过程中：一阶特征识别通过后，持续捕捉此后连续多个周波电气参数，对一阶特征提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解；
    根据数据自身时间尺度特征获取多个本征模态函数，根据本征模态函数与故障本征模态函数的对比，得到串联故障电弧结果。
  5. 如权利要求4所述的基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：
    故障本征模态函数为S 1＝[x 1,x 2,x 3…x m]，待识别本征模态函数为S 2＝[y 1,y 2,y 3…y n]，两序列间的最优路径为：
    ξ(S 1,S 2)＝D(x m,y n)+min[ξ(x m-1,x n),ξ(x m-1,x n-1),ξ(x m,x n-1)]
    其中，D(x m,y n)为S 1和S 2对应元素组成的距离代价矩阵，ξ(S 1,S 2)越小，故障本征模态函数与待识别本征模态函数的相似度越高。
  6. 如权利要求1所述的基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：
    互补集合经验模态分解，包括：
    在原始信号中加入p对的高斯白噪声信号，每对白噪声幅值相同但极性相反；
    对p个正噪声混合信号进行经验模态分解，并进行集成平均，得到一组IMF+函数；
    对p个负噪声混合信号进行经验模态分解，并进行集成平均，得到一组IMF-函数；
    对两组集成的IMF+函数和IMF-函数求平均值，得到最终的IMF分量。
  7. 如权利要求6所述的基于物联网的建筑电气火灾串联故障电弧识别方法，其特征在于：
    第k+1个分解结果IMF k+1，具体为：
    Figure PCTCN2021112821-appb-100001
    其中，R k(t)＝R k-1(t)-IMF k，E 1和E k均为算子期望，ω i(t)为均值是零的典型高斯白噪声信号，ε k为系数，I为均值是零的典型高斯白噪声信号数量。
  8. 一种基于物联网的建筑电气火灾串联故障电弧识别***，其特征在于：包括：
    数据获取模块，被配置为：获取待检测线路的电流有效值、高频周期电流和电流谐波；
    预处理模块，被配置为：对获取的数据进行预处理；
    一阶特征提取模块，被配置为：当预处理后的电流有效值大于第一预设阈值时，根据预处理后的电流谐波确定负载类型；
    二阶特征提取模块，被配置为：对一阶提取后的疑似异常数据的高频周期电流进行互补集合经验模态分解，结合负载类型，根据分解结果与预设故障本征模态函数的对比得到串联故障电弧结果。
  9. 一种计算机可读存储介质，其上存储有程序，其特征在于，该程序被处理器执行时实现如权利要求1-7任一项所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤。
  10. 一种电子设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的程序，其特征在于，所述处理器执行所述程序时实现如权利要求1-7任一项所述的基于物联网的建筑电气火灾串联故障电弧识别方法中的步骤。
PCT/CN2021/112821 2021-08-09 2021-08-16 基于物联网的建筑电气火灾串联故障电弧识别方法及*** WO2023015576A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110909775.0A CN113589105B (zh) 2021-08-09 2021-08-09 基于物联网的建筑电气火灾串联故障电弧识别方法及***
CN202110909775.0 2021-08-09

Publications (1)

Publication Number Publication Date
WO2023015576A1 true WO2023015576A1 (zh) 2023-02-16

Family

ID=78256447

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/112821 WO2023015576A1 (zh) 2021-08-09 2021-08-16 基于物联网的建筑电气火灾串联故障电弧识别方法及***

Country Status (2)

Country Link
CN (1) CN113589105B (zh)
WO (1) WO2023015576A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117092980A (zh) * 2023-08-05 2023-11-21 淮阴师范学院 一种基于大数据的电气故障检测控制***
CN117239942A (zh) * 2023-11-16 2023-12-15 天津瑞芯源智能科技有限责任公司 一种具有监控功能的电表
CN117434406A (zh) * 2023-12-20 2024-01-23 天津航空机电有限公司 一种基于互补集合经验模态分解的电弧故障检测方法
CN117473445A (zh) * 2023-12-27 2024-01-30 深圳市明心数智科技有限公司 基于极限学习机的设备异常分析方法及装置
CN117874434A (zh) * 2024-03-11 2024-04-12 湖南工程学院 一种环境温湿度智能检测过程中数据清洗方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114999095B (zh) * 2022-05-23 2023-11-14 山东建筑大学 基于时间和空间融合的建筑电气火灾监测方法及***

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109490701A (zh) * 2018-09-17 2019-03-19 南京航空航天大学 一种工频串联电弧故障检测方法
CN109521301A (zh) * 2018-11-30 2019-03-26 北京航空航天大学 一种故障电弧产生装置及其检测方法
CN110568331A (zh) * 2019-09-17 2019-12-13 珠海格力电器股份有限公司 一种电弧故障检测方法及保护装置
US20200099334A1 (en) * 2017-06-02 2020-03-26 Sma Solar Technology Ag Method for detecting a contact fault in a photovoltaic system

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102135555B (zh) * 2010-12-29 2013-05-01 重庆大学 低压***串联电弧故障识别方法
CN105067966B (zh) * 2015-07-08 2017-10-10 上海交通大学 基于特征模态分量能量分析的低压交流故障电弧检测方法
CN107064752B (zh) * 2017-03-22 2019-09-27 北京航空航天大学 一种航空故障电弧检测的判别算法
CN112505511B (zh) * 2020-12-14 2022-06-07 天津求实智源科技有限公司 一种非侵入式低压故障电弧检测与定位方法及***

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200099334A1 (en) * 2017-06-02 2020-03-26 Sma Solar Technology Ag Method for detecting a contact fault in a photovoltaic system
CN109490701A (zh) * 2018-09-17 2019-03-19 南京航空航天大学 一种工频串联电弧故障检测方法
CN109521301A (zh) * 2018-11-30 2019-03-26 北京航空航天大学 一种故障电弧产生装置及其检测方法
CN110568331A (zh) * 2019-09-17 2019-12-13 珠海格力电器股份有限公司 一种电弧故障检测方法及保护装置

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117092980A (zh) * 2023-08-05 2023-11-21 淮阴师范学院 一种基于大数据的电气故障检测控制***
CN117092980B (zh) * 2023-08-05 2024-02-06 淮阴师范学院 一种基于大数据的电气故障检测控制***
CN117239942A (zh) * 2023-11-16 2023-12-15 天津瑞芯源智能科技有限责任公司 一种具有监控功能的电表
CN117239942B (zh) * 2023-11-16 2024-01-19 天津瑞芯源智能科技有限责任公司 一种具有监控功能的电表
CN117434406A (zh) * 2023-12-20 2024-01-23 天津航空机电有限公司 一种基于互补集合经验模态分解的电弧故障检测方法
CN117434406B (zh) * 2023-12-20 2024-04-09 天津航空机电有限公司 一种基于互补集合经验模态分解的电弧故障检测方法
CN117473445A (zh) * 2023-12-27 2024-01-30 深圳市明心数智科技有限公司 基于极限学习机的设备异常分析方法及装置
CN117473445B (zh) * 2023-12-27 2024-04-16 深圳市明心数智科技有限公司 基于极限学习机的设备异常分析方法及装置
CN117874434A (zh) * 2024-03-11 2024-04-12 湖南工程学院 一种环境温湿度智能检测过程中数据清洗方法
CN117874434B (zh) * 2024-03-11 2024-05-24 湖南工程学院 一种环境温湿度智能检测过程中数据清洗方法

Also Published As

Publication number Publication date
CN113589105A (zh) 2021-11-02
CN113589105B (zh) 2022-07-26

Similar Documents

Publication Publication Date Title
WO2023015576A1 (zh) 基于物联网的建筑电气火灾串联故障电弧识别方法及***
CN112505511B (zh) 一种非侵入式低压故障电弧检测与定位方法及***
CN104678265A (zh) 一种串联故障电弧检测装置及检测方法
Chen et al. Intelligent home-appliance recognition over IoT cloud network
CN102135555B (zh) 低压***串联电弧故障识别方法
CN110007197B (zh) 串联电弧故障检测***及检测方法
CN103576059B (zh) 一种变压器匝间放电综合故障诊断方法及***
Guo et al. Detection and line selection of series arc fault in multi-load circuit
CN109490701B (zh) 一种工频串联电弧故障检测方法
CN107370123A (zh) 电弧故障检测装置
CN110543921A (zh) 一种基于波形学习的电缆早期故障识别方法
TWI492182B (zh) 電器負載辨識方法
CN112183628B (zh) 基于多种线性时频变换的交流电弧故障检测方法及***
CN113567794A (zh) 基于动态时间规整的电动自行车户内充电辨识方法及***
CN110376489B (zh) 一种智能配电网故障电弧检测***
CN109920193A (zh) 一种电气火灾隐患智能检测方法、***、设备及介质
Gong et al. Series arc fault identification method based on wavelet transform and feature values decomposition fusion DNN
CN113281625B (zh) 一种低压用户非侵入式电弧故障精准检测方法及***
CN115144703B (zh) 基于零序差分电流和能量矩指标的高阻接地故障识别方法
Liu et al. Classification and identification of electric shock current for safety operation in power distribution network
CN114755533A (zh) 一种基于电压变化沿特征识别的电弧故障检测方法及***
CN109581160B (zh) 故障电弧检测方法
Zeng et al. High impedance fault detection in distribution network based on S-transform and average singular entropy
US20230341477A1 (en) Series arc detection firmware
Wang et al. Research on AC Arc Fault Detection Method Based on Wigner-Hough Transform and NG-Boost

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21953189

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE