CN113240057B - High-precision error detection method and system based on electric power data acquisition - Google Patents
High-precision error detection method and system based on electric power data acquisition Download PDFInfo
- Publication number
- CN113240057B CN113240057B CN202110785930.2A CN202110785930A CN113240057B CN 113240057 B CN113240057 B CN 113240057B CN 202110785930 A CN202110785930 A CN 202110785930A CN 113240057 B CN113240057 B CN 113240057B
- Authority
- CN
- China
- Prior art keywords
- waveform
- power data
- differential
- differential waveform
- sequence
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/15—Correlation function computation including computation of convolution operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Power Engineering (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention relates to a high-precision error detection method and a high-precision error detection system based on electric power data acquisition, wherein the method comprises the following steps: acquiring a waveform diagram of the power data to be detected, and comparing the waveform diagram of each sampling period with the waveform diagrams of one or more adjacent periods to obtain one or more differential waveforms of the waveform diagram of each sampling period; respectively identifying the differential waveform of the power data to be tested and the corresponding compression sequence thereof by utilizing a target identification network and a trained bidirectional GRU based on an attention mechanism; and judging whether high-precision errors exist in the power data to be detected in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform. The method combines the target identification network and the GRU to respectively identify the differential waveform and the sensing compression sequence of the power data, and improves the accuracy and the universality of high-precision error detection of the power.
Description
Technical Field
The invention belongs to the field of electric power data processing, and particularly relates to a high-precision error detection method and system based on electric power data acquisition.
Background
The power data generally comprises voltage values, current values and variation values of various power equipment in operation, and power disturbance reflects various disturbances at a source-grid-load side, for example, power disturbance is easily caused by intermittent new energy grid connection at a power supply side or the on-off characteristic of power electronic equipment, and meteorological factors such as lightning stroke are common reasons for power disturbance; various power grid faults occurring at the power grid side and abnormal operation of equipment such as a transformer, a capacitor and the like can also cause corresponding waveform distortion in monitoring data; power disturbances are also generated by user side sensitive equipment failures and dynamic power usage behavior of the load. These disturbances, in their nature, are phenomena or events that distort the voltage, current, or deviate from desired values. Eliminating or detecting the disturbances has important significance for improving the accuracy of power data acquisition.
Research shows that a series of weak disturbances with unobvious disturbance characteristics exist in the operation process of the power system, the disturbances are difficult to extract characteristics through a traditional signal processing method, and then the disturbances are accurately detected based on deterministic criteria. The difficulty of detection is that the forms of disturbance waveforms are various, and hidden features which can be used for detection need to be extracted from slightly disturbed waveform data. In particular, some weak, continuous perturbations having similar waveform characteristics for multiple cycles are easily detected as multiple perturbation events.
Abundant equipment running state information is contained in the electric power disturbance data, the electric power disturbance is accurately detected and identified, the electric energy quality can be improved, decision support can be provided for equipment maintenance, and the power supply reliability and economy are improved. Therefore, the accurate detection of the power disturbance data has important significance. At present, a detection method of power disturbance mainly extracts characteristic signal features based on a signal processing method (wavelet transform, fourier transform), and then obtains corresponding detection criteria based on the extracted characteristic quantities.
In the deep learning field, each source or form of information may be referred to as a modality. For example, humans have touch, hearing, vision, smell; information media such as voice, video, text and the like; a wide variety of sensors such as radar, infrared, accelerometer, etc. Each of the above may be referred to as a modality. With the rise and development of multi-modal Machine Learning (MMML), the intrinsic characteristics of information (signals) are understood or learned using various angles or ways, but it is rare to apply them to the field of signal detection or processing.
Attention models have been widely used in recent years in various fields of deep learning, whether in various types of tasks for image processing, speech recognition, or natural language processing. The attention mechanism in deep learning is similar to the selective visual attention mechanism of human beings in nature, and the core target is to select information which is more critical to the current task target from a plurality of information.
Disclosure of Invention
In order to improve the detection accuracy and the universality of the power disturbance, the invention provides a high-precision error detection method based on power data acquisition in a first aspect, which comprises the following steps: acquiring a waveform diagram of the power data to be detected, and comparing the waveform diagram of each sampling period with the waveform diagrams of one or more adjacent periods to obtain one or more differential waveforms of the waveform diagram of each sampling period; respectively identifying the differential waveform of the power data to be tested and the corresponding compression sequence thereof by utilizing a target identification network and a trained bidirectional GRU based on an attention mechanism; and judging whether high-precision errors exist in the power data to be detected in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform.
In some embodiments of the present invention, the obtaining the waveform diagram of the power data to be measured, and comparing the waveform diagram of each sampling period with the waveform diagrams of one or more adjacent periods to the waveform diagram of each sampling period to obtain one or more differential waveforms of the waveform diagram of each sampling period includes the following steps: uniformly dividing the differential waveform into a plurality of subframes by using a window function, and calculating the effective value of the differential waveform in each subframe: if the effective value of the differential waveform is lower than a first threshold value, ignoring the differential waveform; and if the effective value of the differential waveform is higher than or equal to a first threshold value, performing Fourier transform on the differential waveform, and recording and storing the corresponding fundamental frequency and harmonic frequency.
In some embodiments of the present invention, the preset measurement matrix is obtained by: calculating the effective value of each differential waveform, comparing the effective values according to a preset second threshold value to obtain a symbol function sequence formed by the effective values of each differential waveforma n }; and determining each element in the preset measurement matrix according to the symbolic function sequence and the logistic mapping.
Further, determining each element in the measurement matrix according to the symbol function sequence and the Logistic mapping by determining an initial value and a parameter value of the Logistic mappingμGenerating random sequenceb n }; randomly selecting a sampling value of an effective value of a differential waveform as a symbol function sequencea n The initial value of } is set; according to said symbol function sequencea n Fill with said random sequenceb n Determining each element of the measurement matrix:Φ=a n ×b n whereinΦThe elements of the measurement matrix are represented by,a n andb n respectively representing the nth element in the sequence of the symbol function and the random sequence.
In some embodiments of the invention, the trained attention-based bidirectional GRU is trained by:
extracting the characteristics of the compressed sequence of each differential waveform to obtain a plurality of characteristic vectors;
taking the plurality of feature vectors as the input of a bidirectional GRU, taking the state parameters of a hidden layer of the bidirectional GRU as the input of an attention layer, and constructing the bidirectional GRU based on an attention mechanism;
updating the weight matrix of the attention-based bidirectional GRU until the error of the attention-based bidirectional GRU is lower than a threshold value, and obtaining a trained attention-based bidirectional GRU.
Further, before the extracting features of the compressed sequence of each differential waveform to obtain a plurality of feature vectors, the method further includes: frequency offset correction is performed on the compressed sequence of each differential waveform.
The invention provides a high-precision error detection system based on power data acquisition, which comprises an acquisition module, a sampling module, an identification module and a judgment module, wherein the acquisition module is used for acquiring a oscillogram of power data to be detected, and comparing the oscillogram of each sampling period with the oscillograms of one or more adjacent periods to obtain one or more differential waveforms of the oscillogram of each sampling period; the sampling module is used for carrying out compression sampling on each differential waveform according to a preset measurement matrix to obtain a compression sequence of each differential waveform; the identification module is used for respectively identifying the differential waveforms and the corresponding compression sequences thereof by utilizing a target identification network and a trained bidirectional GRU model based on an attention mechanism; and the judging module is used for judging whether high-precision errors exist in the power data to be detected in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform. In a third aspect of the present invention, there is provided an electronic device comprising: one or more processors; storage means for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to carry out the method provided by the first aspect of the invention.
Further, the identification module comprises a target identification network and a bidirectional GRU based on an attention mechanism, wherein the target identification network is used for identifying a differential waveform; the attention-based bidirectional GRU is used for identifying a compression sequence corresponding to a differential waveform.
In a third aspect of the present invention, there is provided an electronic device comprising: one or more processors; a storage device, configured to store one or more programs, which when executed by the one or more processors, cause the one or more processors to implement the method for detecting errors with high accuracy based on power data acquisition provided by the first aspect of the present invention.
In a fourth aspect of the present invention, a computer readable medium is provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the method of the first aspect of the present invention based on a high accuracy error detection method of power data acquisition.
The invention has the beneficial effects that:
1. the differential waveform and the effective value of the power data are used as the characteristics of the power disturbance, the waveform information and the sequence information contained in the power disturbance are respectively learned by utilizing deep learning, and the model integrating the waveform information and the sequence information has better universality and accuracy;
2. the compressed power data is sensed and compressed by using compression, so that the real-time performance of the inspection is improved on the premise of slight data integrity loss;
3. the target recognition network recognizes the waveform, recognizes sequence data based on attention bidirectional GRU, fuses characteristic information of the two, and can further improve accuracy of high-precision errors and robustness of models.
Drawings
FIG. 1 is a schematic diagram of a power disturbance detection process based on compressed sensing and deep learning according to some embodiments of the present invention;
FIG. 2 is a schematic diagram of differential waveforms collected in some embodiments of the present invention;
FIG. 3 is a schematic illustration of an attention mechanism based bidirectional GRU structure and training process in some embodiments of the invention;
FIG. 4 is a schematic diagram of a high-accuracy error detection system based on power data acquisition according to some embodiments of the present invention;
FIG. 5 is a basic block diagram of an electronic device in some embodiments of the invention.
Detailed Description
The principles and features of this invention are described below in conjunction with the following drawings, which are set forth by way of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, in a first aspect of the present invention, a high-precision error detection method based on power data acquisition is provided, including the following steps: s100, acquiring a waveform diagram of power data to be detected, and comparing the waveform diagram of each sampling period with waveform diagrams of one or more adjacent periods to obtain one or more differential waveforms of the waveform diagram of each sampling period; s300, respectively identifying the differential waveform of the power data to be detected and a corresponding compression sequence thereof by utilizing a target identification network and a trained bidirectional GRU based on an attention mechanism; and S400, judging whether high-precision errors exist in the power data to be detected in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform. It can be understood that the collected errors generally come from the aging of the equipment itself or the abnormality of the equipment caused by disturbance of the power grid, and thus errors are formed or accumulated, and the errors are not easy to be perceived by using the traditional detection or measurement method.
Referring to fig. 2, in order to improve the real-time performance of high-precision error detection, reduce transmission delay, and compress power data, in step S100 of some embodiments of the present invention, the obtaining a waveform diagram of power data to be detected, and comparing the waveform diagram of each sampling period with waveform diagrams of one or more adjacent periods thereof to obtain one or more differential waveforms of the waveform diagram of each sampling period includes the following steps: uniformly dividing the differential waveform into a plurality of subframes by using a window function, and calculating the effective value of the differential waveform in each subframe: if the effective value of the differential waveform is lower than a first threshold value, ignoring the differential waveform; and if the effective value of the differential waveform is higher than or equal to a first threshold value, performing Fourier transform on the differential waveform, and recording and storing the corresponding fundamental frequency and harmonic frequency. The first threshold is 5% -20% of the root Mean Square (MSE) of the effective values of the difference waveforms of the sub-frames in the plurality of periods. In fig. 2, parts (a), (b), and (c) show waveform diagrams of an original waveform of power data, an original waveform (adjacent waveform) delayed by one cycle, and a differential waveform, respectively.
In step S200 of some embodiments of the present invention, the preset measurement matrix is obtained by: calculating the effective value of each differential waveform, comparing the effective values according to a preset second threshold value to obtain a symbol function sequence formed by the effective values of each differential waveforma n And i.e.: the effective value of each differential waveform is compared to a second threshold: if the value is larger than the second threshold value, the value is 1; if the threshold value is smaller than the second threshold value, taking the value as-1; if the value is equal to the second threshold value, the value is 0; each element in the predetermined measurement matrix is thus determined from the sequence of sign functions and the logistic map. Wherein the second threshold value is 5% -20% of the root Mean Square (MSE) of the effective values of the plurality of differential waveforms.
Further, the variation trend of the actual differential waveform is better fitted, each element in the measurement matrix is determined according to the symbolic function sequence and the Logistic mapping, and the initial value and the parameter value of the Logistic mapping are determinedμGenerating random sequenceb n }; randomly selecting a sampling value of an effective value of a differential waveform as a symbol function sequencea n The initial value of } is set; according to said symbol function sequencea n Fill with said random sequenceb n Determining each element of the measurement matrix:Φ=a n ×b n whereinΦThe elements of the measurement matrix are shown (the row and column numbers of the elements of the matrix are omitted),a n andb n respectively representing the nth element in the sequence of the symbol function and the random sequence. In a specific embodiment, the parameter values of the Logistic mappingμ∈[1.872,2.0]Initial value ofa 0 =0.23,0.37 or 0.7; preferably, the first and second liquid crystal materials are,μ=2.0, random sequenceb n Satisfy bernoulli distribution while satisfying RIP (Restricted Isometry Property).
It can be understood that the coarse-grained variation trend of the compressed sequence of the differential waveform is fitted through the symbolic function sequence, the randomness of the fine granularity of the differential waveform is fitted through the chaos characteristic of Logistic mapping, and the accuracy of high-precision error measurement can be further improved by taking the product of the coarse-grained variation trend and the chaos characteristic as an element of a measurement matrix.
The power data has certain chaotic characteristics in a long period without loss of generality, so that the power data can be effectively compressed by using a measurement matrix generated by chaotic mapping, the Logistic mapping can be replaced by Tent mapping, Chebyshev mapping and other one-dimensional chaotic mappings, and the multi-dimensional chaotic mapping is not beneficial to detection real-time due to high calculation complexity; on the basis, in order to adapt to the electric power data acquired under different signal-to-noise ratio environments, the measurement matrix can be replaced by a random sampling value of an effective value of a differential waveform to initially determine elements of one or more measurement matrixes of a Gaussian random measurement matrix, a Bernoulli measurement matrix, a sparse measurement matrix, a Toeplitz measurement matrix, a cyclic measurement matrix and a Hadamard measurement matrix, so that the detection universality of high-precision errors is improved.
Referring to FIG. 3, in step S300 of some embodiments of the present invention, the trained attention-based bidirectional GRU is trained by S301. performing feature extraction on the compressed sequence of each differential waveform to obtain a plurality of feature vectors; s302, taking the plurality of feature vectors as input of a bidirectional GRU, taking state parameters of a hidden layer of the bidirectional GRU as input of an attention layer, and constructing the bidirectional GRU based on an attention mechanism; and S303, updating the weight matrix of the attention mechanism-based bidirectional GRU until the error of the attention mechanism-based bidirectional GRU is lower than a threshold value, and obtaining the trained attention mechanism-based bidirectional GRU.
Schematically, fig. 3 shows the basic structure and associated training process of an attention-based bidirectional GRU: input sequence (S 1 -S t ) Obtaining hidden state parameters through bidirectional GRU networkh 1 -h t ) The hidden state parameter obtains attention distribution through the learning process of attention module information selection
。
Obtaining new information quantity after the hidden state parameter is selected by the attention distribution information weighting (m 1-m t ) And finally, outputting a differential sequence result through a full-link layer: i.e. whether the difference sequence indicates the presence of a high accuracy error. It will be appreciated that the attention mechanism is represented by the hidden state parameter(s) ((h 1 -h t ) Attention distribution through attention module information selection
The full connection layer usually includes an activation function or a classification function, and the classification process of the activation function or the classification function can be implemented by the attention moduleNow.
Further, before the extracting features of the compressed sequence of each differential waveform to obtain a plurality of feature vectors, the method further includes: frequency offset correction is performed on the compressed sequence of each differential waveform. Specifically, the zero-crossing point of the reference period is determined first, and the precise period of the sampling waveform is determined by linear interpolation for a plurality of times, so that the frequency correction is performed on the sampling point data.
In step S400 of some embodiments of the present invention, whether there is a high-precision error in the power data to be measured is determined in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform. Specifically, a bidirectional GRU based on an attention mechanism and a target identification network are fused with a differential waveform prediction result or a sequence prediction result obtained by the bidirectional GRU and the target identification network through a decision tree classifier, and finally whether disturbance exists in power data is judged. The target recognition network at least comprises one of R-CNN, Fast R-CNN, Detectron, YOLO series network, SSD, FCN or DeepLab.
Example 2
Referring to fig. 4, in a second aspect of the present invention, a high-precision error detection system 1 based on power data acquisition is provided, including an obtaining module 11, a sampling module 12, an identifying module 13, and a judging module 14, where the obtaining module 11 is configured to obtain a waveform diagram of power data to be detected, and compare the waveform diagram of each sampling period with waveform diagrams of one or more adjacent periods thereof to obtain one or more differential waveforms of the waveform diagram of each sampling period; the sampling module 12 is configured to perform compression sampling on each differential waveform according to a preset measurement matrix to obtain a compression sequence of each differential waveform; the identification module 13 is configured to respectively identify the differential waveforms and the corresponding compression sequences thereof by using a target identification network and a trained attention-based bidirectional GRU model; and the judging module 14 is configured to judge whether a high-precision error exists in the power data to be detected in real time according to the identified differential waveform and the type of the compression sequence corresponding to the differential waveform.
Further, the identification module 13 includes a target identification network for identifying the differential waveform and a bidirectional GRU based on attention mechanism; the attention-based bidirectional GRU is used for identifying a compression sequence corresponding to a differential waveform.
It can be understood that after the high-precision error is detected, an interference source is determined or the collector is replaced or overhauled according to the error analysis of the high-precision error, so that the measurement precision of the collector is improved.
Example 3
In a third aspect of the present invention, there is provided an electronic device comprising: one or more processors; a storage device, configured to store one or more programs, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the method for detecting errors with high accuracy based on power data acquisition according to the first aspect of the present invention.
Referring to fig. 5, an electronic device 500 may include a processing means (e.g., central processing unit, graphics processor, etc.) 501 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)502 or a program loaded from a storage means 508 into a Random Access Memory (RAM) 503. In the RAM 503, various programs and data necessary for the operation of the electronic apparatus 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input/output (I/O) interface 505 is also connected to bus 504.
The following devices may be connected to the I/O interface 505 in general: input devices 506 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 507 including, for example, a Liquid Crystal Display (LCD), speakers, vibrators, and the like; a storage device 508 including, for example, a hard disk; and a communication device 509. The communication means 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. While fig. 5 illustrates an electronic device 500 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided. Each block shown in fig. 5 may represent one device or may represent multiple devices as desired.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 509, or installed from the storage means 508, or installed from the ROM 502. The computer program, when executed by the processing device 501, performs the above-described functions defined in the methods of embodiments of the present disclosure. It should be noted that the computer readable medium described in the embodiments of the present disclosure may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In embodiments of the disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In embodiments of the present disclosure, however, a computer readable signal medium may comprise a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.
The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device. The computer readable medium carries one or more computer programs which, when executed by the electronic device, cause the electronic device to:
computer program code for carrying out operations for embodiments of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, Python, Go, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A high-precision error detection method based on power data acquisition is characterized by comprising the following steps:
acquiring a waveform diagram of the power data to be detected, and comparing the waveform diagram of each sampling period with the waveform diagrams of one or more adjacent periods to obtain one or more differential waveforms of the waveform diagram of each sampling period;
carrying out compression sampling on each differential waveform according to a preset measurement matrix to obtain a compression sequence of each differential waveform;
respectively identifying a differential waveform of the power data to be tested and a corresponding compression sequence thereof by utilizing a target identification network and a trained bidirectional GRU based on an attention mechanism;
and judging whether high-precision errors exist in the power data to be detected in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform.
2. The method for detecting the high-precision error based on the power data acquisition as claimed in claim 1, wherein the step of obtaining the waveform diagram of the power data to be detected, and comparing the waveform diagram of each sampling period with the waveform diagrams of one or more adjacent periods to obtain one or more differential waveforms of the waveform diagram of each sampling period comprises the following steps:
uniformly dividing the differential waveform into a plurality of subframes by using a window function, and calculating the effective value of the differential waveform in each subframe:
if the effective value of the differential waveform in each subframe is lower than a first threshold value, ignoring the differential waveform in the subframe;
and if the effective value of the differential waveform in each subframe is higher than or equal to a first threshold value, performing Fourier transform on the differential waveform in the subframe, and recording and storing the corresponding fundamental frequency and harmonic frequency.
3. The method for detecting the high-precision error based on the power data acquisition as claimed in claim 1, wherein the preset measurement matrix is obtained by the following steps:
calculating the effective value of each differential waveform, comparing the effective values according to a preset second threshold value to obtain a symbol function sequence formed by the effective values of each differential waveforma n }；
And determining each element in the preset measurement matrix according to the symbolic function sequence and the logistic mapping.
4. A high accuracy error detection method based on power data collection according to claim 3, wherein each element in the measurement matrix is determined according to a sequence of symbol functions and Logistic mapping:
determining initial values and parameter values of Logistic mappingsμGenerating random sequenceb n }；
Randomly selecting a sampling value of an effective value of a differential waveform as a symbol function sequencea n The initial value of } is set;
according to said symbol function sequencea n Fill with said random sequenceb n Determining each element of the measurement matrix:Φ=a n ×b n whereinΦThe elements of the measurement matrix are represented by,a n andb n respectively representing sequences of symbol functions andthe nth element in the machine sequence.
5. A power data acquisition-based high accuracy error detection method as claimed in claim 1, wherein the trained attention mechanism-based bidirectional GRU is trained by:
extracting the characteristics of the compressed sequence of each differential waveform to obtain a plurality of characteristic vectors;
taking the plurality of feature vectors as the input of a bidirectional GRU, taking the state parameters of a hidden layer of the bidirectional GRU as the input of an attention layer, and constructing the bidirectional GRU based on an attention mechanism;
updating the weight matrix of the attention-based bidirectional GRU until the error of the attention-based bidirectional GRU is lower than a threshold value, and obtaining a trained attention-based bidirectional GRU.
6. The method according to claim 5, wherein before performing feature extraction on the compressed sequence of each differential waveform to obtain a plurality of feature vectors, the method further comprises:
frequency offset correction is performed on the compressed sequence of each differential waveform.
7. A high-precision error detection system based on electric power data acquisition is characterized by comprising an acquisition module, a sampling module, an identification module and a judgment module,
the acquisition module is used for acquiring a waveform diagram of the power data to be detected, and comparing the waveform diagram of each sampling period with the waveform diagrams of one or more adjacent periods to obtain one or more differential waveforms of the waveform diagram of each sampling period;
the sampling module is used for carrying out compression sampling on each differential waveform according to a preset measurement matrix to obtain a compression sequence of each differential waveform;
the identification module is used for respectively identifying the differential waveforms and the corresponding compression sequences thereof by utilizing a target identification network and a trained bidirectional GRU model based on an attention mechanism;
and the judging module is used for judging whether high-precision errors exist in the power data to be detected in real time according to the identified differential waveform and the class of the compression sequence corresponding to the differential waveform.
8. The power data collection-based high-precision error detection system of claim 7, wherein the identification module comprises a target identification network and an attention-based bidirectional GRU,
the target identification network is used for identifying differential waveforms;
the attention-based bidirectional GRU is used for identifying a compression sequence corresponding to a differential waveform.
9. An electronic device, comprising: one or more processors; a storage device for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the power data acquisition-based high-precision error detection method of any one of claims 1-6.
10. A computer-readable medium, on which a computer program is stored, wherein the computer program, when being executed by a processor, implements the power data acquisition based high precision error detection method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110785930.2A CN113240057B (en) | 2021-07-12 | 2021-07-12 | High-precision error detection method and system based on electric power data acquisition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110785930.2A CN113240057B (en) | 2021-07-12 | 2021-07-12 | High-precision error detection method and system based on electric power data acquisition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113240057A CN113240057A (en) | 2021-08-10 |
| CN113240057B true CN113240057B (en) | 2021-09-28 |
Family
ID=77135440
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110785930.2A Active CN113240057B (en) | 2021-07-12 | 2021-07-12 | High-precision error detection method and system based on electric power data acquisition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113240057B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112183280A (en) * | 2020-09-21 | 2021-01-05 | 西安交通大学 | Underwater sound target radiation noise classification method and system based on EMD and compressed sensing |
| CN112564280A (en) * | 2020-11-30 | 2021-03-26 | 海南电网有限责任公司电力科学研究院 | Electric energy quality detection and protection system for high-speed rail station |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10713341B2 (en) * | 2011-07-13 | 2020-07-14 | Scott F. McNulty | System, method and apparatus for generating acoustic signals based on biometric information |
| US10193396B1 (en) * | 2014-05-07 | 2019-01-29 | Energous Corporation | Cluster management of transmitters in a wireless power transmission system |
| JP5890816B2 (en) * | 2013-09-30 | 2016-03-22 | 富士重工業株式会社 | Filtering device and environment recognition system |
| CN109685314B (en) * | 2018-11-20 | 2021-10-29 | 中国电力科学研究院有限公司 | Non-intrusive load decomposition method and system based on long-term and short-term memory network |
| CN110414412B (en) * | 2019-07-25 | 2023-03-31 | 长沙理工大学 | Wide-area power grid multiple disturbance accurate identification method and device based on big data analysis |
| CN110929563A (en) * | 2019-10-14 | 2020-03-27 | 广东浩迪创新科技有限公司 | Electrical appliance identification method and device |
| CN111965409B (en) * | 2020-07-17 | 2023-01-03 | 浙江理工大学 | Voltage transient disturbance detection method based on segmented differential waveform effective value |
| CN111965541B (en) * | 2020-08-18 | 2021-06-18 | 杭州核诺瓦科技有限公司 | Device and method for detecting states of slot wedge, air gap and rotor broken bar of motor magnetic circuit |
| CN112307969B (en) * | 2020-10-30 | 2024-03-12 | 全球能源互联网欧洲研究院 | Pulse signal classification identification method and device and computer equipment |
| CN112487880B (en) * | 2020-11-13 | 2021-09-24 | 河北省科学院应用数学研究所 | Median filter, device and system based on maximum error criterion |
-
2021
- 2021-07-12 CN CN202110785930.2A patent/CN113240057B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112183280A (en) * | 2020-09-21 | 2021-01-05 | 西安交通大学 | Underwater sound target radiation noise classification method and system based on EMD and compressed sensing |
| CN112564280A (en) * | 2020-11-30 | 2021-03-26 | 海南电网有限责任公司电力科学研究院 | Electric energy quality detection and protection system for high-speed rail station |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113240057A (en) | 2021-08-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108593260B (en) | Optical cable line fault positioning and detecting method and terminal equipment | |
| US10228994B2 (en) | Information processing system, information processing method, and program | |
| CN107784276B (en) | Microseismic event identification method and device | |
| CN113884961B (en) | SOC calibration method, modeling device, computer equipment and medium | |
| US20210034918A1 (en) | Method and apparatus for detecting signal features | |
| CN105637331A (en) | Abnormality detection device, abnormality detection method, and computer-readable storage medium | |
| CN113837596B (en) | Fault determination method and device, electronic equipment and storage medium | |
| CN116416884B (en) | Testing device and testing method for display module | |
| CN113948159B (en) | Fault detection method, device and equipment for transformer | |
| CN116257663A (en) | Abnormality detection and association analysis method and related equipment for unmanned ground vehicle | |
| CN113240057B (en) | High-precision error detection method and system based on electric power data acquisition | |
| CN116030370A (en) | Behavior recognition method and device based on multi-target tracking and electronic equipment | |
| CN107894837A (en) | Dynamic sentiment analysis model sample processing method and processing device | |
| CN116702005A (en) | Neural network-based data anomaly diagnosis method and electronic equipment | |
| CN115577249A (en) | Transformer acoustic signal identification method, system and medium with multi-view feature fusion | |
| CN114500075A (en) | User abnormal behavior detection method and device, electronic equipment and storage medium | |
| CN114187552A (en) | Method and system for monitoring power environment of machine room | |
| CN115238805B (en) | Training method of abnormal data recognition model and related equipment | |
| CN111695501A (en) | Equipment soft fault detection method based on operating system kernel calling data | |
| CN113011096B (en) | Current transformer saturation waveform recovery method based on model and data hybrid driving | |
| CN114001939B (en) | Line connection device monitoring method, device, equipment and storage medium | |
| Yu et al. | Similarity study of hydrological time series based on data mining | |
| CN115237739B (en) | Analysis method, device and equipment for board card running environment and readable storage medium | |
| CN117074840B (en) | DC charging pile on-line testing system | |
| CN115617023B (en) | Spacecraft attitude control system abnormity positioning method and device |
Legal Events
| 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 |