A method, system and device for quickly locating a controllable phase-changing valve fault

Abstract

The present invention relates to a method, system and device for quickly locating a controllable phase-changing valve fault. The method comprises the following steps: step S1, constructing a multi-source heterogeneous data recognition model for a controllable phase-changing valve; step S2, learning and training the multi-source heterogeneous data recognition model; step S3, using the trained recognition model to quickly analyze and organize the multi-source heterogeneous data; step S4, constructing and training a multi-dimensional analysis model for fault source tracing and positioning; step S5, inputting the multi-source heterogeneous data of step S3 into the trained multi-dimensional analysis model to realize the rapid positioning of the controllable phase-changing valve fault. Compared with the prior art, the present invention has the advantages of realizing the rapid positioning of the operating fault of the valve, realizing the precise judgment of the operating parameters of the controllable phase-changing valve, and improving the efficiency of on-site operation and maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of controllable phase-changing and flow-changing valves, and in particular to a method, system and device for quickly locating a fault of a controllable phase-changing and flow-changing valve. Background Art

[0002] The controllable commutation converter valve (CLCC) can solve the problem of continuous locking of multi-infeed DC systems and fundamentally eliminate the long-term threat of commutation failure. The controllable commutation converter valve maintains the six-pulse bridge structure of the traditional LCC converter valve (conventional grid commutation converter valve), and reconstructs the converter valves that constitute each bridge arm based on the idea of ​​series-parallel connection of half-controlled thyristor devices and fully controlled IGBT devices to form a hybrid converter valve. The converter valve makes full use of the technical characteristics of large capacity and low loss of thyristor devices and strong IGBT shutdown capability, taking advantage of their strengths and making up for their weaknesses. It not only inherits the advantages of good economy of traditional LCC converter valves, but also fundamentally avoids the occurrence of commutation failure.

[0003] Compared with traditional LCC converter valve technology, CLCC converter valves use more components, and CLCC operating conditions are more complex than traditional LCC converter valves. Therefore, there are more types of fault events or alarm messages, and more combinations of event messages. If traditional manual fault problem analysis is used, the workload of operators will be greatly increased, and it will be difficult to achieve efficient operation and maintenance support. Therefore, it is urgent to study CLCC converter valve fault rapid location auxiliary technology, by automatically locating the converter valve fault area and fault component category, saving operators' accident location time and improving emergency response capabilities.

[0004] How to quickly locate the fault of a controllable phase-changing valve has become a technical problem that needs to be solved. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a method, system and device for quickly locating the fault of a controllable phase-changing flow control valve.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] According to one aspect of the present invention, a method for quickly locating a controllable phase-changing valve fault is provided, the method comprising the following steps:

[0008] Step S1, constructing a multi-source heterogeneous data recognition model for a controllable phase-changing valve;

[0009] Step S2, learning and training the multi-source heterogeneous data recognition model;

[0010] Step S3, using the trained recognition model to quickly analyze and organize multi-source heterogeneous data;

[0011] Step S4, constructing and training a multi-dimensional analysis model for fault tracing and location;

[0012] Step S5, inputting the multi-source heterogeneous data of step S3 into the trained multi-dimensional analysis model to achieve rapid positioning of the controllable phase-changing valve fault.

[0013] Preferably, the multi-source heterogeneous data recognition model for constructing a controllable phase-changing valve is specifically as follows: event messages of different fault types of the controllable phase-changing valve are obtained, and relevant information about the location where the fault occurs is identified and extracted using a semantic recognition model; at the same time, the structural data formed by the background recording is identified, and the multi-source and multi-type heterogeneous data are sorted and arranged according to the time axis to construct a recognition model for multi-source heterogeneous data of the controllable phase-changing valve, wherein the multi-source and multi-type heterogeneous data include time text, fault location and operating condition status.

[0014] Preferably, the process of learning and training the multi-source heterogeneous data recognition model includes the following steps:

[0015] 1) The message data obtained during the valve control simulation reporting device and the converter valve real operation test is used to form a training set, and the multi-source heterogeneous data recognition model in step S1 is trained and its parameters are adjusted;

[0016] 2) Randomly select message data from the training set to form a validation set, and verify the recognition speed of the model after parameter optimization;

[0017] 3) Use the message data from the engineering site to form a test set to test the recognition model, identify the corresponding fault type from each message data and assign characteristics, automatically identify and judge the fault semiconductor level number involved in the message data, and automatically map the spatial position corresponding to the associated fault semiconductor level.

[0018] Preferably, the use of the trained recognition model to quickly analyze and organize multi-source heterogeneous data is specifically: using a hardware closed-loop simulation environment to generate multi-source heterogeneous data required by the multi-source heterogeneous data recognition model in real time, performing actual testing on the recognition model, and realizing rapid analysis and organization of the multi-source heterogeneous data.

[0019] More preferably, the hardware closed-loop simulation environment includes a real-time digital simulation experiment platform, a valve control platform and a fault simulation reporting device.

[0020] More preferably, the multi-source heterogeneous data includes messages, trigger angles, device switch states, IGBT instantaneous voltages, and system currents.

[0021] Preferably, the construction and training of a multi-dimensional analysis model for fault source tracing and location includes the following steps:

[0022] Firstly, all fault types and corresponding fault sequences of the thyristor level, IGBT level and valve control protection system in the controllable phase-commutated valve are traversed during operation, and the event combination rules, event time sorting rules and event interval time rules generated by the association of fault types and fault sequences are formed into a correlation time axis training set to obtain the analysis model of the correlation time axis.

[0023] Secondly, the fault message rules of the thyristor level, IGBT level and valve control protection system under different operating conditions are analyzed, so as to associate the fault phenomenon with the operating state and obtain an analysis model based on the operating state;

[0024] Finally, the analysis model of the associated time axis and the analysis model based on the operating status are combined to form a multi-dimensional analysis model based on time and working conditions, and training work is carried out on the multi-dimensional analysis model based on time and working conditions.

[0025] Preferably, the fault message rules include analysis of the impact of operating conditions on fault phenomena and analysis of the impact of operating conditions on event sequencing rules; the operating conditions include system lockout, LCC mode operation and CLCC mode operation.

[0026] According to one aspect of the present invention, a rapid positioning system for a controllable phase-changing valve fault is provided, the system comprising a semantic recognition module, a data parsing and extraction module, and a fault tracing and positioning analysis module;

[0027] The semantic recognition module is used to obtain event messages under different fault types of the CLCC converter valve, and identify and extract relevant information about the location where the fault occurred;

[0028] The data analysis and extraction module is used to realize rapid analysis and arrangement of multi-source heterogeneous data by using a method for automatically identifying multi-source heterogeneous data of a controllable phase-changing valve;

[0029] The fault tracing and locating analysis module is used to construct and train a multi-dimensional analysis model based on time and working conditions, and to locate and analyze the fault of the controllable phase-changing valve based on the trained multi-dimensional analysis model of time and working conditions.

[0030] According to a third aspect of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1) The present invention proposes automatic identification of multi-source and multi-type heterogeneous data, combined with a time-operating condition multi-dimensional analysis model with the ability to jointly analyze operating condition-time series multi-factors. This is a first at home and abroad, and can achieve rapid positioning of converter valve operating faults, greatly improving on-site operation and maintenance efficiency.

[0033] 2) The present invention constructs a recognition model for multi-source heterogeneous data of controllable phase-changing valves, extracts and recognizes features of the operating status data of the controllable phase-changing valves, and can automatically recognize multi-source and multi-type heterogeneous data efficiently and accurately.

[0034] 3) The present invention takes into account all operating conditions and operating sequences of the controllable phase-changing valve, and realizes real-time monitoring of the electrical parameters and operating status of key components such as thyristors and IGBTs in the controllable phase-changing valve according to the corresponding operating conditions. The time-operating condition multi-dimensional analysis model of the present invention is used to realize the precise judgment of the operating parameters of the controllable phase-changing valve and achieve the accuracy of fault location. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of the method for automatically identifying multi-source heterogeneous data of a controllable phase-changing valve in the present invention;

[0036] Figure 2 It is a flow chart of the controllable commutation fault tracing and locating method of the time-operating condition multi-dimensional analysis model in the present invention;

[0037] Figure 3 It is a schematic diagram of the structure of a rapid positioning system for a controllable phase-changing flow control valve fault in the present invention;

[0038] Figure 4 The figure is a flow chart of a method for quickly locating a fault of a controllable phase-changing flow control valve in the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] Example 1

[0041] This embodiment relates to a method for quickly locating a controllable phase-changing valve fault, such as Figure 4 , the method comprising:

[0042] Step S1, constructing a multi-source heterogeneous data recognition model for a controllable phase-changing valve;

[0043] The various fault level information in the controllable phase commutation (CLCC) converter valve event message is obtained, and the semantic recognition model is used to identify and extract the relevant information of the fault location.

[0044] Step S2, learning and training the multi-source heterogeneous data recognition model;

[0045] Step S3, using the trained recognition model to quickly analyze and organize multi-source heterogeneous data;

[0046] Construct a correspondence database between the time series of controllable phase-changing valve fault events and the fault causes, and realize the analysis and location of different components and various fault causes of the controllable phase-changing valve under different time sequencing and different operating conditions;

[0047] Step S4, constructing and training a multi-dimensional analysis model for fault tracing and location;

[0048] Based on the operation status extraction data, the event time sequence arrangement and time interval rules caused by different fault causes of each component of the controllable phase-changing valve under different operating conditions are compared and analyzed. Combined with the time axis correlation method, a controllable phase-changing valve fault tracing and positioning method based on the time-operating condition multi-dimensional analysis model is formed. Among them, the operating conditions of the controllable phase-changing valve include but are not limited to valve locking, valve LCC mode operation, and valve CLCC mode operation;

[0049] Step S5, inputting the multi-source heterogeneous data of step S3 into the trained multi-dimensional analysis model to achieve rapid positioning of the controllable phase-changing valve fault.

[0050] For unstructured data such as background event message text and time, as well as structured data such as recorded data, a unified analysis based on the time axis is carried out, and a controllable phase-changing valve fault tracing and positioning method based on the time-operating condition multi-dimensional analysis model is proposed to achieve rapid positioning of controllable phase-changing valve faults.

[0051] like Figure 1 As shown, the method for automatically identifying multi-source heterogeneous data of a controllable phase-changing valve includes constructing a multi-source heterogeneous data identification model, learning and training the multi-source heterogeneous data identification model, and testing and verifying the multi-source heterogeneous data identification model, so as to realize rapid analysis and organization of multi-source heterogeneous data.

[0052] Firstly, constructing a multi-source heterogeneous data recognition model specifically includes: obtaining event messages of different fault types of controllable phase commutation (CLCC) valves, and using the semantic recognition model to identify and extract relevant information about the fault location; at the same time, identifying the structured data formed by the background recording, and arranging and sorting multi-source and multi-type heterogeneous data such as time text, fault location, and operating condition according to the time axis to obtain the recognition model of multi-source heterogeneous data of controllable phase commutation valves.

[0053] Then, the multi-source heterogeneous data recognition model is learned and trained:

[0054] 1) The message data obtained during the valve control simulation reporting device and the converter valve real operation test is used to form a multi-source heterogeneous data recognition model training set, and the model is trained and parameter adjusted;

[0055] 2) Randomly select message data from the multi-source heterogeneous data recognition model training set to form a multi-source heterogeneous data recognition model verification set, and verify the recognition speed of the model after parameter optimization;

[0056] 3) Use the event message data from the engineering site to form a multi-source heterogeneous data recognition model test set, and use the trained multi-source heterogeneous data recognition model for testing, and finally realize the identification of the corresponding fault type from each text and the assignment of features, as well as the automatic identification and judgment of the fault semiconductor level number involved in the text, and automatically map the spatial position corresponding to the fault level.

[0057] Finally, the multi-source heterogeneous data recognition model was tested and verified: RT-LAB (Real Time Laboratory, a real-time digital simulation experimental platform, a scientific instrument used in the field of power and electrical engineering), valve control platform and fault simulation reporting device were used to build a hardware closed-loop simulation platform to generate multi-source heterogeneous data such as messages, trigger angles, device switch states, IGBT instantaneous voltage, system current, etc. required for the recognition model in real time, and the model was actually tested to achieve rapid analysis and organization of multi-source heterogeneous data. The hardware closed-loop simulation platform is equipped with a multi-source heterogeneous data recognition model.

[0058] Example 3

[0059] like Figure 2 As shown, the controllable phase-changing valve fault tracing and positioning based on multi-dimensional information includes the following steps:

[0060] Firstly, all fault types (including semiconductor faults, board faults and other primary equipment faults) and the corresponding fault sequence that occur in the thyristor level, IGBT level and valve control protection system of the CLCC converter valve during operation are traversed. The event combination rules, event time sorting rules and event interval time rules generated by associating the fault types and fault sequences are formed into a correlation time axis training set to obtain the analysis model of the correlation time axis.

[0061] Secondly, the fault message rules of the thyristor level, IGBT level and valve control protection system under different operating conditions (system lockout, LCC mode operation, CLCC mode operation) are analyzed, including the analysis of the impact of operating conditions on fault phenomena and the analysis of the impact of operating conditions on event sequencing rules, so as to associate the fault phenomenon with the operating status and obtain an analysis model based on the operating status.

[0062] Finally, the analysis model of the associated time axis is combined with the analysis model based on the operating status to form a time-operating condition multi-dimensional analysis model, and the time-operating condition multi-dimensional analysis model is trained;

[0063] The fault location analysis of the controllable phase-changing valve based on the trained time-operating condition multi-dimensional analysis model can fully improve the accuracy of the fault.

[0064] Example 2

[0065] This embodiment relates to a fast positioning system for a controllable phase-changing valve fault, such as Figure 3 As shown, the system includes a semantic recognition module, a data parsing and extraction module, and a fault tracing and positioning analysis module.

[0066] The semantic recognition module is used to obtain event messages under different fault types of the CLCC converter valve, and use the semantic recognition model to identify and extract relevant information about the fault location.

[0067] The data analysis and extraction module is used to adopt the automatic identification method of multi-source heterogeneous data of controllable phase-changing valves to realize the rapid analysis and organization of multi-source heterogeneous data such as messages, trigger angles, device switching states, IGBT instantaneous voltage, system current, etc.

[0068] The fault tracing and positioning analysis module is used to construct and train a multi-dimensional analysis model based on time and working conditions, and to locate and analyze the faults of the controllable phase-changing valve based on the trained multi-dimensional analysis model of time and working conditions.

[0069] Example 3

[0070] This embodiment relates to an electronic device.

[0071] The electronic device is a physical medium for CLCC fault location that realizes automatic identification of multi-source heterogeneous data of controllable phase-changing valves and joint analysis of working conditions and timing multi-factors, including a memory and a processor. The memory stores a computer program, and the processor implements the method described in the above section when executing the computer program. When the computer program is executed by the processor, the above-mentioned method for automatic identification of multi-source heterogeneous data of controllable phase-changing valves and the method for tracing and locating faults of controllable phase-changing valves based on multi-dimensional information are implemented.

[0072] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0073] In some embodiments, part or all of the computer program can be loaded and / or installed on 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 the method for automatically identifying multi-source heterogeneous data of a controllable phase-changing valve and the method for tracing the source of a fault of a controllable phase-changing valve based on multi-dimensional information described above can be executed. Alternatively, in other embodiments, the CPU can be configured to execute the method for automatically identifying multi-source heterogeneous data of a controllable phase-changing valve and the method for tracing the source of a fault of a controllable phase-changing valve based on multi-dimensional information in any other appropriate manner (for example, by means of firmware).

[0074] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0075] 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 device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0076] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for quickly locating a controllable phase-changing valve fault, characterized in that: The method comprises the following steps: Step S1, constructing a multi-source heterogeneous data recognition model for a controllable phase-changing valve; Step S2, learning and training the multi-source heterogeneous data recognition model; Step S3, using the trained recognition model to quickly analyze and organize multi-source heterogeneous data; Step S4, constructing and training a multi-dimensional analysis model for fault tracing and location; Step S5, inputting the multi-source heterogeneous data of step S3 into the trained multi-dimensional analysis model to achieve rapid positioning of the controllable phase-changing valve fault; The construction and training of a multi-dimensional analysis model for fault source tracing and location includes the following steps: Firstly, all fault types and corresponding fault sequences of the thyristor level, IGBT level and valve control protection system in the controllable phase-commutated valve are traversed during operation, and the event combination rules, event time sorting rules and event interval time rules generated by the association of fault types and fault sequences are formed into a correlation time axis training set to obtain the analysis model of the correlation time axis. Secondly, the fault message rules of the thyristor level, IGBT level and valve control protection system under different operating conditions are analyzed, so as to associate the fault phenomenon with the operating state and obtain an analysis model based on the operating state; Finally, the analysis model of the associated time axis and the analysis model based on the operating status are combined to form a multi-dimensional analysis model based on time and working conditions, and training work is carried out on the multi-dimensional analysis model based on time and working conditions.

2. A method for quickly locating a controllable phase-changing valve fault according to claim 1, characterized in that: The multi-source heterogeneous data recognition model for building a controllable phase-changing valve is specifically as follows: event messages under different fault types of the controllable phase-changing valve are obtained, and relevant information about the fault location is identified and extracted using a semantic recognition model; at the same time, the structural data formed by the background recording is identified, and the multi-source and multi-type heterogeneous data are sorted and arranged according to the time axis to build a recognition model for multi-source heterogeneous data of the controllable phase-changing valve, wherein the multi-source and multi-type heterogeneous data include time text, fault location and operating condition status.

3. A method for quickly locating a controllable phase-changing valve fault according to claim 1, characterized in that: The process of learning and training the multi-source heterogeneous data recognition model includes the following steps: 1) The message data obtained during the valve control simulation reporting device and the converter valve real operation test is used to form a training set, and the multi-source heterogeneous data recognition model in step S1 is trained and its parameters are adjusted; 2) Randomly select message data from the training set to form a validation set, and verify the recognition speed of the model after parameter optimization; 3) Use the message data from the engineering site to form a test set to test the recognition model, identify the corresponding fault type from each message data and assign characteristics, automatically identify and judge the fault semiconductor level number involved in the message data, and automatically map the spatial position corresponding to the associated fault semiconductor level.

4. A method for quickly locating a controllable phase-changing valve fault according to claim 1, characterized in that: The rapid analysis and organization of multi-source heterogeneous data using the trained recognition model specifically includes: using a hardware closed-loop simulation environment to generate multi-source heterogeneous data required by the multi-source heterogeneous data recognition model in real time, performing actual testing on the recognition model, and realizing rapid analysis and organization of multi-source heterogeneous data.

5. A method for quickly locating a controllable phase-changing valve fault according to claim 4, characterized in that: The hardware closed-loop simulation environment includes a real-time digital simulation experiment platform, a valve control platform and a fault simulation reporting device.

6. A method for quickly locating a controllable phase-changing valve fault according to claim 4, characterized in that: The multi-source heterogeneous data includes messages, trigger angles, device switch states, IGBT instantaneous voltages, and system currents.

7. A method for quickly locating a controllable phase-changing valve fault according to claim 1, characterized in that: The fault message rules include analysis of the impact of operating conditions on fault phenomena and analysis of the impact of operating conditions on event sequencing rules; the operating conditions include system locking, LCC mode operation and CLCC mode operation.

8. A system using the method for quickly locating a controllable phase-changing valve fault according to claim 1, the system comprising a semantic recognition module, a data parsing and extraction module, and a fault tracing and locating analysis module; The semantic recognition module is used to obtain event messages under different fault types of the CLCC converter valve, and identify and extract relevant information about the location where the fault occurred; The data analysis and extraction module is used to realize rapid analysis and arrangement of multi-source heterogeneous data by using a method for automatically identifying multi-source heterogeneous data of a controllable phase-changing valve; The fault tracing and locating analysis module is used to construct and train a multi-dimensional analysis model based on time and working conditions, and to locate and analyze the fault of the controllable phase-changing valve based on the trained multi-dimensional analysis model of time and working conditions.

9. 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 7 is implemented.

Images