A Fault Prediction Method for Energy Storage Converter
By collecting and analyzing fault data of energy storage converters in real time, using artificial intelligence and statistical technology, combined with cloud service modules, the problem of difficult to predict multiple energy storage converters in the existing technology is solved, and more accurate Improved fault prediction and maintenance efficiency.
Patent Information
- Application Number
- CN202210960694.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-11
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-08-11
Smart Images

Figure CN115347670B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of fault prediction, and more specifically, to a method for predicting faults of an energy storage converter. Background Art
[0002] With the continuous development and growth of new energy, energy storage inverters adjust the randomness, volatility and indirectness of wind and photovoltaic power generation. By controlling the charging and discharging process of the battery, AC and DC conversion is performed to supply power to the load, achieving complementary output of wind energy, solar energy and energy storage, improving the predictability and controllability of energy generation, solving the problem of safe and stable operation of new energy, and enabling the power system to achieve a dynamic and stable balance.
[0003] With the rapid development of new energy and microgrids, the application scale of energy storage inverters is also getting larger and larger. The performance status of energy storage inverters affects the entire power system. In order to better monitor the operating status of energy storage inverters, timely and accurate prediction of energy storage inverter failures is conducive to ensuring the safe and stable operation of the entire power system, while reducing the power outage time due to post-processing and maintenance, and improving maintenance efficiency.
[0004] At present, the continuous change of workload of the energy storage inverter operation system causes various operation failures of the system. The existing fault prediction uses historical labeled data for learning prediction, but in practice it is difficult to obtain labeled historical data of system operation. In addition, the system fault prediction can only predict faults for a single energy storage inverter. When multiple energy storage inverters have the same fault, it is impossible to analyze the cause of the fault based on the same output generated by the fault data node, which causes lag in the maintenance and production of energy storage inverters. Summary of the invention
[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for predicting faults of an energy storage converter. A real-time acquisition module is used to connect a fault recorder with a voltage sensor and a current sensor to collect fault data of the storage converter to obtain accurate fault data. A fault data management module uses fault prediction technology based on data-driven artificial intelligence technology and statistical technology to analyze and process various data to mine implicit information and perform prediction operations. A cloud service module generates the same output according to the fault data node to analyze the cause of the fault, thereby providing favorable fault prediction for the energy storage converter during later production and installation, and playing a positive role in avoiding the occurrence of faults, so as to solve the problems raised in the above-mentioned background technology.
[0006] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a method for predicting faults of energy storage converters, comprising a plurality of PCS modules, a distributed power supply module, a real-time acquisition module, a fault data management module, a PCS control module, a communication module, an operation and maintenance mobile terminal, a visualization module and a cloud service module, wherein the plurality of PCS modules are connected and used for energy storage of new energy electric energy and DC-AC conversion to supply power to the load area, the real-time acquisition module is used for real-time acquisition of current, voltage and temperature of the plurality of PCS modules and transmits them to the fault data management module, and the fault data management module is used for data analysis and data in the fault prediction system Storage, transmit the data analysis results to the PCS control module, and store the data locally or in a cloud service module. The PCS control module is used for the control and scheduling of several PCS modules. The PCS control module is connected to the communication module to transmit the fault information to the cloud service module or the operation and maintenance mobile terminal module. The communication module includes wifi wireless network technology, mobile data technology or Bluetooth technology to transmit data. The operation and maintenance mobile terminal receives the fault warning message and displays the fault code and cause according to the fault characteristics through the visualization module. The cloud service module receives the data storage and calculates the data fault type and transmits it to the operation and maintenance mobile terminal.
[0007] In a preferred embodiment, the specific control steps are as follows:
[0008] S1, the plurality of PCS modules are connected to supply power to the load area through DC or AC conversion, and the real-time acquisition module collects the uninterrupted wavelength emitted when a fault occurs in the energy storage converter in real time;
[0009] S2, the fault data management module performs data analysis and data storage, transmits the data analysis results to the PCS control module, and stores the data locally or in a cloud service module;
[0010] S3, the PCS control module controls and dispatches several PCS modules according to the load conditions, and the PCS control module is connected to the communication module to transmit the fault information to the cloud service module or the operation and maintenance mobile terminal module;
[0011] S4. The operation and maintenance mobile terminal receives the fault warning message, and displays the fault code and cause through the visualization module according to the fault characteristics. The operation and maintenance personnel view it through the operation and maintenance mobile terminal and go to the site for repair. The cloud service module receives and stores the data and calculates the data fault type and transmits it to the operation and maintenance mobile terminal.
[0012] In a preferred embodiment, the real-time acquisition module is connected to the voltage sensor and the current sensor through a fault recorder to collect the fault data of the storage converter. The fault recorder continuously and uninterruptedly performs fast sampling of the collected AC or DC power, scans the switch and relay protection switch quantity, and performs simple calculation and analysis of the fault data through the harmonic analysis method during the sampling and scanning. When the energy storage converter fails, the fault filter stores the fault data before, during and after the fault in the data storage unit of the fault data management module. During the operation of the energy storage converter, the AC voltage and current are obtained to obtain a sinusoidal waveform. The harmonic source in the memory of other systems changes periodically and satisfies the Dirichlet principle. The harmonics can be decomposed into the following formula:
[0013] In a preferred embodiment, the harmonic decomposition formula is as follows: ω1 is the power frequency angular frequency, n is the harmonic order, and U n ,I n The root mean square value of the nth harmonic voltage and current is obtained respectively, α n , β n are the initial phase angles of the nth harmonic voltage and current respectively, M is the highest harmonic order, and it is determined by the degree of waveform distortion and the accuracy requirement of analysis, and M≤50.
[0014] In a preferred embodiment, the fault data management module includes a data analysis unit and a data storage unit, wherein the data analysis unit uses fault prediction technology to perform further analysis after obtaining the fault data. The fault prediction technology is based on data-driven use of artificial intelligence technology and statistical technology to analyze and process various data to mine implicit information and perform predictive operations. Artificial intelligence technology uses a decision tree method for analysis, and statistical technology uses a canonical variable analysis method to analyze the fault collection data, and the analysis results are transmitted to the PCS control module.
[0015] In a preferred embodiment, the PCS control module includes an early warning unit and a control unit. The control unit is responsible for issuing control instructions to control the entire fault prediction system, including the coordinated and correct operation of various components. The early warning unit obtains control limits based on the probability empirical distribution of historical contribution data to determine the change in the new contribution data rate. If the variance of the historical contribution is small and the variance of the new sample is large when there is no fault, this indicates that the energy storage inverter has a certain specific fault. By analyzing all process variables through control limits, a small number of process variables with abnormal contribution changes are found as fault points. For this purpose, the early warning unit transmits the fault warning signal to the communication module and transmits it to the operation and maintenance mobile terminal module.
[0016] In a preferred embodiment, the fault code generation process is divided into three levels based on the specific process on MATLAB. The first level is the control prototype development stage, that is, to minimize errors in subsequent links through control algorithm modeling and array error shooting. The second level is connected to the first level through a code generation tool, and the third stage is carried out after code implantation or code modification and compilation optimization. The third level is data verification and parameter calibration of the control algorithm. When the control algorithm adjusts the abnormal data, it returns to MATLAB for modification.
[0017] In a preferred embodiment, when one or several energy storage converters fail while several PCS modules are operating, the cloud service module collects fault data through the real-time acquisition module, analyzes the fault data through the fault data management module and stores it in the cloud. After the data is stored in the cloud, the data is further analyzed, and frequently occurring fault data information is selected for processing, and the cause of the fault is analyzed based on the same output generated by the fault data node.
[0018] Technical effects and advantages of the present invention:
[0019] The present invention specifically utilizes a fault recorder to connect with a voltage sensor and a current sensor through a real-time acquisition module to collect storage converter fault data to obtain accurate fault data; a fault data management module utilizes fault prediction technology based on data-driven use of artificial intelligence technology and statistical technology to analyze and process various data to mine implicit information and perform predictive operations; a cloud service module generates the same output according to the fault data node to analyze the cause of the fault, thereby providing favorable fault prediction for the energy storage converter during later production and installation, and playing a positive role in avoiding the occurrence of faults. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 It is a system structure block diagram of the present invention.
[0021] Figure 2 It is the operation flow chart of the fault recorder of the present invention. DETAILED DESCRIPTION
[0022] 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 only 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 are within the scope of protection of the present invention.
[0023] Example
[0024] This embodiment provides Figure 1A method for predicting faults of an energy storage converter is shown, comprising a plurality of PCS modules, a distributed power supply module, a real-time acquisition module, a fault data management module, a PCS control module, a communication module, an operation and maintenance mobile terminal, a visualization module and a cloud service module. The plurality of PCS modules are connected and used for energy storage of new energy electric energy and DC-AC conversion to power the load area. The real-time acquisition module is used for real-time acquisition of current, voltage and temperature of the plurality of PCS modules and transmits them to the fault data management module. The fault data management module is used for data analysis and data storage in the fault prediction system, transmits the data analysis results to the PCS control module, and stores the data locally or in a cloud service module. The PCS control module is used for control and scheduling of the plurality of PCS modules. The PCS control module is connected to the communication module, and transmits the fault information to the cloud service module or the operation and maintenance mobile terminal module. The communication module comprises wifi wireless network technology, mobile data technology or Bluetooth technology to transmit data. The operation and maintenance mobile terminal receives the fault warning message, and displays the fault code and cause according to the fault characteristics through the visualization module. The cloud service module receives the data storage and calculates the data fault type and transmits it to the operation and maintenance mobile terminal.
[0025] S1, the plurality of PCS modules are connected to supply power to the load area through DC or AC conversion, and the real-time acquisition module collects the uninterrupted wavelength emitted when a fault occurs in the energy storage converter in real time;
[0026] like Figure 1 In this embodiment, it is necessary to specifically explain that the real-time acquisition module is connected to the voltage sensor and the current sensor through a fault recorder to collect the fault data of the storage converter. The fault recorder continuously and uninterruptedly collects AC or DC power for fast sampling, scans the switch and relay protection switch quantity, and performs simple calculation and analysis on the fault data through the harmonic analysis method during the sampling and scanning. When the energy storage converter fails, the fault filter stores the fault data before, during and after the fault in the data storage unit of the fault data management module. During the operation of the energy storage converter, the AC voltage and current are obtained to obtain a sinusoidal waveform. The harmonic source in the memory of other systems is periodically changing and satisfies the Dirichlet principle. The harmonics can be decomposed into the following formula:
[0027] Among them, ω1 is the angular frequency of the power frequency, n is the harmonic number, U n ,I n The root mean square value of the nth harmonic voltage and current is obtained respectively, α n , β n are the initial phase angles of the nth harmonic voltage and current respectively, M is the highest harmonic order, and it is determined by the degree of waveform distortion and the accuracy requirement of analysis, and M≤50.
[0028] S2, the fault data management module performs data analysis and data storage, transmits the data analysis results to the PCS control module, and stores the data locally or in a cloud service module;
[0029] What needs to be specifically explained in this embodiment is that the fault data management module includes a data analysis unit and a data storage unit, wherein the data analysis unit uses fault prediction technology to perform further analysis after obtaining the fault data, and the fault prediction technology is based on data-driven use of artificial intelligence technology and statistical technology to analyze and process various data to mine implicit information and perform predictive operations, artificial intelligence technology uses decision trees for analysis, and statistical technology uses canonical variable analysis methods to analyze the fault collection data, and the analysis results are transmitted to the PCS control module.
[0030] S3, the PCS control module controls and dispatches several PCS modules according to the load conditions, and the PCS control module is connected to the communication module to transmit the fault information to the cloud service module or the operation and maintenance mobile terminal module;
[0031] The PCS control module of this embodiment includes an early warning unit and a control unit. The control unit is responsible for issuing control instructions to control the entire fault prediction system, including the coordinated and correct operation of various components. It should be specifically explained that the early warning unit obtains control limits based on the probability empirical distribution of historical contribution data to determine the change in the new contribution data rate. If the variance of the historical contribution is small and the variance of the new sample is large when there is no fault, this indicates that the energy storage inverter has a certain specific fault. Through the control limit analysis of all process variables, a small number of process variables with abnormal contribution changes are found to be the fault points. For this purpose, the early warning unit transmits the fault warning signal to the communication module and transmits it to the operation and maintenance mobile terminal module.
[0032] S4. The operation and maintenance mobile terminal receives the fault warning message, and displays the fault code and cause through the visualization module according to the fault characteristics. The operation and maintenance personnel view it through the operation and maintenance mobile terminal and go to the site for repair. The cloud service module receives and stores the data and calculates the data fault type and transmits it to the operation and maintenance mobile terminal.
[0033] What needs to be specifically explained in this embodiment is that the fault code generation process is divided into three levels based on the specific process on MATLAB. The first level is the control prototype development stage, that is, the control algorithm modeling and array error shooting are used to minimize errors in subsequent links. The second level is connected to the first level through the code generation tool. The third stage is carried out after the code is implanted or the code is modified through compilation and optimization. The third level is the control algorithm for data verification and parameter calibration. The control algorithm adjusts the data abnormal data and then returns to MATLAB for modification. This method reduces the repeated iterations of the code generation link and improves the code generation efficiency.
[0034] What needs to be specifically explained in this embodiment is that when the cloud service module is under the operating conditions of several PCS modules, when one or several energy storage converters fail, the real-time acquisition module is used to collect fault data, and the fault data management module is used to analyze the fault data and store it in the cloud. After the data is stored in the cloud, the data is further analyzed, and frequently occurring fault data information is selected for processing. The cause of the fault is analyzed based on the same output generated by the fault data node, so as to provide favorable fault prediction for the energy storage converter during the later production and installation, and play a positive role in avoiding the occurrence of faults.
[0035] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for predicting faults of energy storage converters. Features: It includes several PCS modules, distributed power supply modules, real-time acquisition modules, fault data management modules, PCS control modules, communication modules, operation and maintenance mobile terminals, visualization modules and cloud service modules. The several PCS modules are connected and used for energy storage of new energy and DC-AC conversion to power the load area. The real-time acquisition module is used for real-time acquisition of current, voltage and temperature of several PCS modules and transmits them to the fault data management module. The fault data management module is used for data analysis and data storage in the fault prediction system, transmits the data analysis results to the PCS control module, and stores the data locally or in the cloud service module. The PCS control module is used for control and scheduling of several PCS modules. The PCS control module is connected to the communication module to transmit the fault information to the cloud service module or the operation and maintenance mobile terminal module. The communication module includes wifi wireless network technology, mobile data technology or Bluetooth technology to transmit data. The operation and maintenance mobile terminal receives the fault warning message, and displays the fault code and cause according to the fault characteristics through the visualization module. The cloud service module receives the data storage and calculates the data fault type and transmits it to the operation and maintenance mobile terminal. The specific control steps are as follows: S1, the plurality of PCS modules are connected to supply power to the load area through DC or AC conversion, and the real-time acquisition module collects the uninterrupted wavelength emitted when a fault occurs in the energy storage converter in real time; S2, the fault data management module performs data analysis and data storage, transmits the data analysis results to the PCS control module, and stores the data locally or in a cloud service module; S3, the PCS control module controls and dispatches several PCS modules according to the load conditions, and the PCS control module is connected to the communication module to transmit the fault information to the cloud service module or the operation and maintenance mobile terminal module; S4, the operation and maintenance mobile terminal receives the fault warning message, and displays the fault code and cause through the visualization module according to the fault characteristics. The operation and maintenance personnel view it through the operation and maintenance mobile terminal and go to the site for repair. The cloud service module receives and stores the data and calculates the data fault type and transmits it to the operation and maintenance mobile terminal; The real-time acquisition module is connected to the voltage sensor and the current sensor through a fault recorder to collect the fault data of the storage converter. The fault recorder continuously and uninterruptedly performs fast sampling of the collected AC or DC power, scans the switch and relay protection switch quantity, and performs simple calculation and analysis of the fault data through the harmonic analysis method during the sampling and scanning. When the energy storage converter fails, the fault filter stores the fault data before, during and after the fault in the data storage unit of the fault data management module. During the operation of the energy storage converter, the AC voltage and current are obtained to obtain a sinusoidal waveform. The harmonic source in the memory of other systems is periodically changing and satisfies the Dirichlet principle. The harmonics can be decomposed into the following formula: ; ; The fault data management module includes a data analysis unit and a data storage unit, wherein the data analysis unit uses the fault prediction technology to further analyze the fault data after obtaining it. The fault prediction technology is based on data-driven use of artificial intelligence technology and statistical technology to analyze and process various data to mine implicit information and perform prediction operations. Artificial intelligence technology uses a decision tree method to analyze, and statistical technology uses a canonical variable analysis method to analyze the fault collection data, and the analysis results are transmitted to the PCS control module; The PCS control module includes an early warning unit and a control unit. The control unit is responsible for issuing control instructions to control the entire fault prediction system, including the coordinated and correct operation of various components. The early warning unit obtains control limits based on the probability empirical distribution of historical contribution data to determine the change in the new contribution data rate. If the variance of the historical contribution is small and the variance of the new sample is large when there is no fault, this indicates that the energy storage inverter has a specific fault. Through the control limit analysis of all process variables, a small number of process variables with abnormal contribution changes are found to be the fault points. For this purpose, the early warning unit transmits the fault warning signal to the communication module and transmits it to the operation and maintenance mobile terminal module.
2. A method for predicting faults of an energy storage converter according to claim 1, Features: The harmonic decomposition formula is is the angular frequency of the power frequency, n is the harmonic number, , Get the RMS value for the nth harmonic voltage and current respectively, , are the initial phase angles of the nth harmonic voltage and current respectively, M is the highest harmonic order, and it is determined by the degree of waveform distortion and the accuracy requirement of analysis, and M≤50.
3. A method for predicting faults of an energy storage converter according to claim 2, Features: The fault code generation process is divided into three levels based on the specific process on MATLAB. The first level is the control prototype development stage, which is to minimize errors in subsequent links through control algorithm modeling and array error shooting. The second level is connected to the first level through a code generation tool. The third stage is carried out after code implantation or code modification and compilation optimization. The third level is data verification and parameter calibration of the control algorithm. The control algorithm adjusts the abnormal data and then returns to MATLAB for modification.
4. A method for predicting faults of an energy storage converter according to claim 3, Features: When a certain energy storage converter fails or several energy storage converters fail under the operating conditions of several PCS modules, the cloud service module collects fault data through the real-time acquisition module, analyzes fault data through the fault data management module and stores it in the cloud. After the data is stored in the cloud, the data is further analyzed, and frequently occurring fault data information is selected for processing. The cause of the fault is analyzed based on the same output generated by the fault data node.
