A Fault Location and Analysis Method, System and Equipment for a CLCC Converter

Abstract

The present invention relates to a method, system and device for fault location and analysis of a CLCC converter. The method includes: constructing a fault location and analysis model for the CLCC converter; offline training the fault location and analysis model for the CLCC converter; collecting real-time data of the CLCC system and performing preprocessing, inputting the preprocessed real-time data into the offline-trained fault location and analysis model for the CLCC converter for online learning, and outputting fault information; deeply analyzing the fault information to output a fault analysis report; and the system is used to implement the above method. Compared with the prior art, the present invention can shorten the model training duration and save computing resources, and can also comprehensively analyze the faults occurring in the CLCC system and still provide accurate and efficient solutions for complex working conditions.

Description

Technical Field

[0001] The present invention relates to the field of power system fault diagnosis, and particularly to a method, system and device for fault location and analysis of a CLCC converter. Background Art

[0002] The CLCC converter bears the important task of resisting commutation failure caused by AC faults and improving the stability of DC transmission power. This process involves complex electrical conversion and energy control. Therefore, once the CLCC converter fails, it will directly affect the normal transmission and distribution of electric energy, resulting in grid voltage fluctuations, frequency offsets, and even large-scale power outages, causing serious impacts on the operation of the economic society. Moreover, the types of faults generated during the operation of the CLCC converter are numerous, such as IGBT damage, thyristor damage, capacitor failure, inverter failure, etc. These faults will not only reduce the conversion efficiency of the converter but also trigger a chain reaction, leading to more serious system faults. The existing technologies for fault analysis of complex devices such as CLCC converters are not perfect, and there are the following problems: incomplete extraction of fault characteristics; inaccurate diagnosis under complex working conditions; too long training time of the fault analysis model, resulting in large consumption of computing resources, and affecting the timeliness of fault troubleshooting.

[0003] Therefore, it is a problem to be solved to provide a method that can efficiently and accurately analyze and locate the faults of the CLCC converter. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, system and device for fault location and analysis of a CLCC converter to overcome the defects of the above-mentioned existing technologies.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] According to the first aspect of the present invention, a method for fault location and analysis of a CLCC converter is provided, and the method includes the following steps:

[0007] Construct a fault location and analysis model for the CLCC converter;

[0008] Obtain a data set, and perform offline training on the fault location and analysis model of the CLCC converter based on the data set; the data set includes normal working conditions and fault working condition data;

[0009] Collect the real-time data of the CLCC system and perform preprocessing. Input the preprocessed real-time data into the CLCC converter fault location and analysis model after offline training for online learning, and output fault information. The real-time data includes current, voltage, and temperature data collected by the CLCC system sensors and is stored in a time series manner. The fault information includes the fault location and the fault type.

[0010] Conduct in-depth analysis on the fault information and output a fault analysis report.

[0011] As a preferred technical solution, the CLCC converter fault location and analysis model includes a cascaded neural network and a fault analysis module. The cascaded neural network includes a convolutional neural network, a recurrent neural network, and two Transformer layers.

[0012] As a preferred technical solution, the method for obtaining the fault information is as follows:

[0013] The input layer of the convolutional neural network receives the preprocessed real-time data and inputs it into the convolutional layer of the convolutional neural network. The preprocessed real-time data is three-dimensional data, which is reshaped into two-dimensional data and input into the convolutional layer for processing to output the first spatial information. The first spatial information is input into the pooling layer for maximum pooling operation to output the second spatial information. The second spatial information is flattened into a one-dimensional vector by the Flatten layer and input into the first Transformer layer.

[0014] The first Transformer layer receives the one-dimensional vector and uses the multi-head attention mechanism to capture the dependencies between the one-dimensional vectors and outputs local time features.

[0015] Use the local time features as the input of the recurrent neural network. The input layer receives the output of the first Transformer layer and delivers it to the recurrent layer. The long short-term memory network layer is used to capture the time features of the input data. The Dropout layer receives the output of the recurrent layer and randomly discards some neurons to output global features.

[0016] The second Transformer layer receives the global features, fuses and optimizes the global features through the global self-attention mechanism, and inputs the result into the output layer. The output layer outputs the fault information.

[0017] As a preferred technical solution, the specific steps of the offline training include:

[0018] Obtain a dataset, label the fault condition data in the dataset according to the location where the fault occurs to obtain a fault code, and construct a fault code set. The first digit of the fault code represents the major fault category, the middle digits represent the fault location, and the last digit represents the minor fault category. The major fault category and the minor fault category constitute the fault type.

[0019] The described major fault categories include circuit faults, device damage or non-operation, and communication faults;

[0020] The described fault locations include IGBT faults, thyristor faults, smoothing reactor faults, current-limiting resistor faults, cooling system faults, control system faults, converter valve system faults, capacitor faults, measurement and monitoring equipment faults, overvoltage protection faults, transformer faults, and DC grounding electrode faults;

[0021] The described minor fault categories include circuit short circuits, circuit open circuits, poor contacts, circuit failures, device mis-triggering, device functional damage, device energy saturation, and communication faults between devices;

[0022] Divide the set composed of the fault coding set and normal operating condition data into a training set, a test set, and a validation set according to a preset rule;

[0023] Use the training set to perform offline training on the CLCC converter fault location and analysis model by means of supervised learning, and use the test set and the validation set to test and validate the trained CLCC converter fault location and analysis model.

[0024] As a preferred technical solution, the specific steps of the online learning include:

[0025] Load the parameters of the CLCC converter fault location and analysis model after offline training and initialize the model;

[0026] Real-time data acquisition and preprocessing, where the real-time data includes voltage, current, and temperature in the CLCC system;

[0027] Input the preprocessed real-time data into the CLCC converter fault location and analysis model and output fault information;

[0028] Compare the output fault information with the actual fault information to evaluate the accuracy of the location and analysis model;

[0029] Based on the evaluation results, update the model parameters of the location and analysis model in real time.

[0030] As a preferred technical solution, the specific steps of the preprocessing are:

[0031] Use a filtering algorithm to denoise the real-time data and remove noise signals;

[0032] Normalize the denoised real-time data to convert data with different dimensions into values between 0 and 1;

[0033] Store the said values in a time series manner.

[0034] As a preferred technical solution, the specific steps for obtaining the above-mentioned fault analysis report are as follows:

[0035] Obtain historical fault condition data and an expert knowledge base;

[0036] Input the fault information into the fault analysis module, and identify the fault cause in combination with the historical fault condition data and the expert knowledge base;

[0037] Generate a fault analysis report according to the fault cause, and the fault analysis report includes the fault cause, the influence scope and the solution.

[0038] According to the second aspect of the present invention, a CLCC converter fault location and analysis system is provided. The fault location and analysis system is used to implement the above method, and includes:

[0039] A data acquisition and processing module: used to acquire and process data;

[0040] A fault location module: This module includes a convolutional neural network and a recurrent neural network, receives the output data of the data acquisition and processing module, processes it and outputs fault information;

[0041] A fault analysis module: receives the fault information, conducts in-depth analysis and outputs a fault analysis report;

[0042] A display module: used to display the fault information and the fault analysis report.

[0043] According to the third aspect of the present invention, an electronic device is provided, including a memory and a processor. A computer program is stored on the memory, and when the processor executes the program, the above method is implemented.

[0044] According to the fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the above method is implemented.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] 1). The training data set of the present invention covers all fault types that occur in actual working conditions, and the CLCC converter fault location and analysis model is offline trained based on this training set so that it can accurately extract the faults that occur in the CLCC system for the existing actual working conditions; and for fault conditions that have not appeared before, the present invention is based on the offline trained model and then through an online learning model, the CLCC converter fault location and analysis model continuously accumulates new faults that appear in the online learning process, continuously improves the fault data set, and enables accurate diagnosis even in the face of complex working conditions;

[0047] 2) The present invention trains the CLCC converter fault location and analysis model using an offline training and online learning mode, effectively shortening the training duration of the CLCC converter fault location and analysis model and saving computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flowchart of the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0050] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the ordinary meanings understood by those of ordinary skill in the technical field to which this application belongs. The "one", "a", "an", "the" and other similar words involved in this application do not indicate a quantity limitation and may represent a singular or plural number. The terms "including", "comprising", "having" and any variations thereof involved in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device including a series of steps or modules (units) is not limited to the listed steps or units, but may further include unlisted steps or units, or may further include other steps or units inherent to these processes, methods, products or devices. The "connection", "coupling" and other similar words involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. The "and / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, "A and / or B" may represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0051] Embodiment 1

[0052] This embodiment provides a CLCC converter fault location and analysis method, which uses deep learning technology to automatically extract the fault features of the CLCC system, quickly locate the fault location, and deeply analyze the fault cause, thereby improving the operation reliability and maintenance efficiency of the CLCC system.

[0053] The flowchart of the fault location and analysis method for the CLCC converter is as follows Figure 1 shown, and the detailed steps are as follows:

[0054] S1. Build a model: The fault location and analysis model for the CLCC converter includes a cascaded neural network and a fault analysis module. The cascaded neural network includes a convolutional neural network, a recurrent neural network, and two Transformer layers.

[0055] S2. Offline train the fault location and analysis model for the CLCC converter:

[0056] S21. Obtain a dataset, including normal operating conditions and fault condition data, and label the fault condition data in the dataset according to the location where the fault occurs to obtain a fault code. Among them, the first digit of the fault code represents the major fault category, the middle digits represent the fault location, and the last digit represents the minor fault category. And the major fault category and the minor fault category constitute the fault type. According to the above information, build a fault code set. The detailed information of the first digit of the fault code is shown in Table 1:

[0057] Table 1 Information Table of the First Digit of the Fault Code

[0058]

[0059] The detailed information of the middle digits of the fault code is shown in Table 2:

[0060] Table 2 Information Table of the Middle Digits of the Fault Code

[0061]

[0062] The detailed information of the last digit of the fault code is shown in Table 3.

[0063] Table 3 Information of the Last Digit of the Fault Code

[0064]

[0065] S22. Divide the set composed of the fault code set and the normal operating condition data into a training set, a test set, and a validation set according to 7:2:1.

[0066] S23. Use the training set to offline train the fault location and analysis model for the CLCC converter by the method of supervised learning, and use the test set and the validation set to test and validate the trained fault location and analysis model for the CLCC converter.

[0067] S3. Online learning of the model and obtaining fault information:

[0068] S31. Real-time data acquisition and preprocessing:

[0069] S311. Collect the real-time data of the CLCC system through the sensors set in the CLCC system, where the real-time data includes the voltage, current, and temperature data collected by the sensors and is stored in a time series manner;

[0070] S312. Preprocess the real-time data. Specifically, use a filtering algorithm to denoise the real-time data and remove the noise signals; normalize the denoised real-time data, convert the data with different dimensions into values between 0 and 1, and then store the values in a time series manner;

[0071] S313. Input the preprocessed real-time data in three-dimensional form into the CLCC converter fault location and analysis model after offline training for online learning.

[0072] S32. Model online learning:

[0073] S321. Load the parameters of the CLCC converter fault location and analysis model after offline training and initialize the CLCC converter fault location and analysis model;

[0074] S322. Real-time data collection and preprocessing. The real-time data includes the voltage, current, and temperature in the CLCC system;

[0075] S323. Input the preprocessed real-time data into the CLCC converter fault location and analysis model and output the fault information;

[0076] S324. Compare the output fault information with the actual fault information to evaluate the accuracy of the location and analysis model;

[0077] S325. Based on the evaluation results, update the model parameters of the location and analysis model in real time.

[0078] S33. Output the fault information:

[0079] S331. The input layer of the convolutional neural network receives the preprocessed real-time data and inputs it into the convolutional layer of the convolutional neural network. The preprocessed real-time data is three-dimensional data. Reshape the three-dimensional data into two-dimensional data and input it into the convolutional layer with a convolutional kernel of 3×3 for processing to output the first spatial information. The first spatial information is input into the pooling layer to apply the max pooling operation to output the second spatial information. The second spatial information is flattened into a one-dimensional vector by the Flatten layer and input into the first Transformer layer;

[0080] S332. The first Transformer layer receives the one-dimensional vector and uses the multi-head attention mechanism to capture the dependencies between the one-dimensional vectors and outputs the local time features;

[0081] S333. Use the local time features as the input of the recurrent neural network. The input layer receives the output of the first Transformer layer and delivers it to the recurrent layer. The long short-term memory network layer is used to capture the time features of the input data. The Dropout layer receives the output of the recurrent layer and randomly discards the outputs of some neurons to output the global features, preventing overfitting and enhancing the generalization ability of the model.

[0082] S334. The second Transformer layer receives the global features, fuses and optimizes the global features through the global self-attention mechanism, and inputs the result into the output layer. This process enables the model to apply more weights to the key time points when the fault occurs, thereby improving the accuracy of fault detection.

[0083] S335. The output layer receives the output result of the second Transformer layer and outputs the fault information, where the fault information includes the fault location and the fault type.

[0084] S4. Conduct in-depth analysis on the fault information and output a fault analysis report:

[0085] S41. Obtain historical fault condition data and an expert knowledge base.

[0086] S42. Input the fault information into the fault analysis module, and identify the fault cause in combination with the historical fault condition data and the expert knowledge base.

[0087] S43. Generate a fault analysis report according to the fault cause, where the fault analysis report includes the fault cause, the influence scope, and the solution.

[0088] Embodiment 2

[0089] The above is the introduction of the method embodiments. The following further illustrates the solution of the present invention through system embodiments.

[0090] This embodiment provides a CLCC converter fault location and analysis system for implementing the method provided in the above embodiment. The fault location and analysis system includes:

[0091] Data acquisition and processing module: used for acquiring and processing data;

[0092] Fault location module: This module includes a convolutional neural network and a recurrent neural network, receives the output data of the data acquisition and processing module, processes it, and outputs the fault information;

[0093] Fault analysis module: receives the fault information, conducts in-depth analysis, and outputs a fault analysis report;

[0094] Display module: used for displaying the fault information and the fault analysis report.

[0095] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0096] Embodiment 3

[0097] The electronic device of the present invention includes a central processing unit (CPU), which can execute various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or computer program instructions loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0098] Multiple components in the device are connected to the I / O interface, including: an input unit, such as a keyboard, a mouse, etc.; an output unit, such as various types of displays, speakers, etc.; a storage unit, such as a magnetic disk, an optical disc, etc.; and a communication unit, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit allows the device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0099] The processing unit executes the various methods and processes described above, such as methods S1 - S4. For example, in some embodiments, methods S1 - S4 can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device via the ROM and / or the communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 - S4 described above can be executed. Alternatively, in other embodiments, the CPU can be configured to execute methods S1 - S4 by any other suitable means (e.g., by means of firmware).

[0100] The functions described above herein can be at least partially executed by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), and so on.

[0101] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program codes can be executed entirely on the machine, partially on the machine, executed partially on the machine and partially on a remote machine as an independent software package, or executed entirely on a remote machine or server.

[0102] In the context of the present invention, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0103] As described above, only the specific embodiments of the present invention are provided, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A CLCC converter fault location and analysis method, characterized in that: The method comprises the following steps: Constructing a CLCC converter fault location and analysis model, wherein the CLCC converter fault location and analysis model comprises a cascade neural network and a fault analysis module; the cascade neural network comprises a convolutional neural network, a recurrent neural network and two Transformer layers; Acquire a data set, and perform offline training on a CLCC converter fault location and analysis model based on the data set; the data set includes normal operating condition data and fault operating condition data; The real-time data of the CLCC system is collected and preprocessed, and the preprocessed real-time data is input into the CLCC converter fault location and analysis model after offline training for online learning, and the fault information is output; the method for outputting the fault information is as follows: the input layer of the convolutional neural network receives the preprocessed real-time data and inputs it into the convolutional layer of the convolutional neural network, the preprocessed real-time data is three-dimensional data, the three-dimensional data is reshaped into two-dimensional data, the convolutional layer is input for processing, and the first spatial information is output, the first spatial information is input into the pooling layer to apply the maximum pooling operation, and the second spatial information is output, and the second spatial information is flattened into a one-dimensional vector by the Flatten layer and input into the first T The first Transformer layer receives the one-dimensional vector and uses the multi-head attention mechanism to capture the dependency between the one-dimensional vectors and output the local time features. The local time features are used as the input of the recurrent neural network. The input layer receives the output of the first Transformer layer and transmits it to the recurrent layer. The long short-term memory network layer is used to capture the time features of the input data. The Dropout layer receives the output of the recurrent layer and randomly discards some neurons to output the global features. The second Transformer layer receives the global features, fuses and optimizes the global features through the global self-attention mechanism, and inputs the results into the output layer, which outputs the fault information. The real-time data includes current, voltage and temperature data collected by the CLCC system sensors and stored in a time series manner; the fault information includes the fault location and fault type; Conduct in-depth analysis on fault information and output fault analysis report.

2. A CLCC converter fault location and analysis method according to claim 1, characterized in that: The specific steps of the offline training include: Acquire a data set, and mark the fault condition data in the data set according to the location where the fault occurs to obtain a fault code, and construct a fault code set; the first digit of the fault code indicates a major fault category, the middle digit indicates a fault location, and the last digit indicates a minor fault category, and the major fault category and minor fault category constitute a fault type; The major types of faults include circuit faults, device damage or malfunction, and communication faults; The fault locations include IGBT fault, thyristor fault, smoothing reactor fault, current limiting resistor fault, cooling system fault, control system fault, converter valve system fault, capacitor fault, measurement and monitoring equipment fault, overvoltage protection fault, transformer fault and DC grounding electrode fault; The fault subcategories include circuit short circuit, circuit open circuit, poor contact, circuit failure, device mis-triggering, device functional damage, device energy saturation, and communication failure between devices; The set consisting of the fault code set and the normal operating condition data is divided into a training set, a test set and a validation set according to preset rules; The training set is used to perform offline training on the CLCC converter fault location and analysis model using a supervised learning method, and the trained CLCC converter fault location and analysis model is tested and verified using the test set and validation set.

3. A CLCC converter fault location and analysis method according to claim 1, characterized in that: The specific steps of online learning include: Load the parameters of the CLCC converter fault location and analysis model after offline training and initialize the model; Real-time data acquisition and preprocessing, the real-time data includes voltage, current and temperature in the CLCC system; Input the preprocessed real-time data into the CLCC converter fault location and analysis model, and output the fault information; Compare the output fault information with the actual fault information to evaluate the accuracy of the positioning and analysis models; Model parameters of the positioning and analysis models are updated in real time based on the evaluation results.

4. A CLCC converter fault location and analysis method according to claim 3, characterized in that: The specific steps of the pretreatment are: Use filtering algorithms to denoise real-time data and remove noise signals; Normalize the denoised real-time data and convert data of different dimensions into values ​​between 0 and 1; The numerical values ​​are stored in a time series manner.

5. A CLCC converter fault location and analysis method according to claim 1, characterized in that: The specific steps for obtaining the fault analysis report are: Obtain historical fault condition data and expert knowledge base; Inputting the fault information into the fault analysis module, and identifying the cause of the fault by combining the historical fault condition data and the expert knowledge base; A fault analysis report is generated based on the cause of the fault, and the fault analysis report includes the cause of the fault, the scope of impact, and a solution.

6. A CLCC converter fault location and analysis system, characterized in that: The fault location and analysis system is used to implement the method described in any one of claims 1 to 5, comprising: Data acquisition and processing module: used to collect and process data; Fault location module: This module includes convolutional neural network and recurrent neural network, receives the output data of data acquisition and processing module, processes and outputs fault information; Fault analysis module: receives fault information, conducts in-depth analysis and outputs fault analysis reports; Display module: used to display fault information and fault analysis reports.

7. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images