Fault early warning method and system for high-voltage cascade energy storage converter system

Abstract

The invention relates to the technical field of fault early warning, and provides a fault early warning method and system for a high-voltage cascade energy storage converter system, a monitoring module comprises a plurality of sensors distributed in each part of the high-voltage cascade energy storage converter system and used for monitoring system parameter data changes in real time, and a fault information processing module collects historical fault data and sends the historical fault data to a fault early warning module. The fault information processing module classifies different fault types, and the analysis module is in data connection with the fault information processing module and the monitoring module, performs early warning analysis on data processed by the fault information processing module, and identifies and compares specific fault types; the early warning module is in data connection with the analysis module, carries out early warning according to an early warning analysis result of the analysis module and informs a specific fault type; according to the invention, the analysis module uses two prediction algorithms to take an average number as a final prediction value, the prediction accuracy and timeliness are improved, and the fault data processing module classifies different fault types to realize accurate positioning and rapid identification of faults.

Description

Technical Field

[0001] The present invention relates to the technical field of fault warning, and in particular to a fault warning method and system for a high-voltage cascade energy storage converter system. Background Art

[0002] With the continuous development of the energy field, high-voltage cascade energy storage and conversion systems play a vital role in many fields such as new energy power generation, smart grids, and electric vehicles.

[0003] High-voltage cascade energy storage converter systems are usually composed of power modules, control systems, energy storage units, etc. They can achieve high-voltage power conversion, use modular design for flexible configuration and expansion, use advanced power electronics technology for high energy conversion efficiency, and the control system can implement a variety of control strategies to meet different application requirements. However, due to the complex structure of the system, the changeable working environment and the high-voltage operation state, the risk of failure is high.

[0004] The existing background technology still has the following shortcomings:

[0005] 1. Lack of efficient early warning systems and methods. Most existing technologies monitor system status in real time and then analyze data to issue an alarm. This method often has a lag in data processing and cannot issue an early warning in advance of a failure, resulting in insufficient preparation time for system maintenance and response measures.

[0006] 2. When the alarm sounds, it is difficult to know the specific cause of the fault. When the alarm sounds, maintenance personnel need to spend a lot of time on troubleshooting and diagnosis after receiving the alarm, which delays the maintenance time, increases system downtime and maintenance costs, and also reduces the system availability and stability. Summary of the invention

[0007] In order to solve the above technical problems, the present invention provides a fault warning method and system for a high-voltage cascade energy storage converter system, which classifies the fault type and determines the extracted characteristic value range to compare with the predicted value of the prediction algorithm to perform fault warning, so as to solve the problems in the prior art.

[0008] A fault warning system for a high-voltage cascade energy storage converter system, comprising:

[0009] A monitoring module, which includes a variety of sensors distributed in various parts of the high-voltage energy storage and conversion system to monitor changes in system parameter data in real time;

[0010] A fault information processing module, which collects historical fault data from system maintenance logs and fault reports, classifies and extracts features of different fault types, and determines the range of sensor data feature values ​​corresponding to each fault type;

[0011] An analysis module, which is connected to the fault information processing module and the monitoring module, and is used to perform prediction calculations on the data monitored by the monitoring module, and identify the fault type in combination with the fault information processing module to compare the specific fault type;

[0012] The early warning module is connected to the analysis module data, and is used to issue early warnings based on the early warning analysis results of the analysis module, and inform relevant staff of the specific fault type.

[0013] Preferably, the sensors in the monitoring module include but are not limited to temperature sensors, current sensors, voltage sensors, vibration sensors and power sensors, and the monitoring module performs simple processing on the real-time monitored data including statistical methods to remove outliers and linear interpolation methods to process missing values.

[0014] Preferably, the fault information processing module analyzes the collected historical fault data;

[0015] Classify faults according to their manifestation or location, and assign a unique identifier to each fault type; extract relevant sensor data features for each fault type, and analyze the value range of these features under different fault types;

[0016] Through statistical analysis, the typical value range of the characteristic value under a specific fault type is determined. The characteristic performance of the same fault type in all historical fault data is comprehensively considered to determine the characteristic value range of the sensor data corresponding to each fault type.

[0017] Preferably, the analysis module includes an analysis unit and a fault identification unit;

[0018] The analysis unit performs analysis and prediction based on the real-time monitoring data of the monitoring module to obtain a predicted value;

[0019] The fault identification unit determines whether the predicted value falls within a characteristic value range of a certain fault type, and if so, issues a warning for this type of fault.

[0020] Preferably, the analysis unit uses an exponential smoothing method combined with a moving average algorithm to perform fault prediction;

[0021] First, arrange the monitored data in time series as (y1, y2, ..., y n,), where y is the value of the data point monitored in the time series, n is the number of monitored data points, and the prediction formula is;

[0022] x t+1 = αy t +(1 - α)x t

[0023] Among them, t represents the time point of each collection of monitoring data in the time series, and x t+1 represents the predicted value of the next time point t + 1 calculated by the exponential smoothing method algorithm, y t represents the actual monitored value of the current time point t, and x t is the predicted value of the previous time point. α is the smoothing coefficient, and its value range is 0 < α < 1. The larger the α value, the more the prediction result tends to the recent data, and the smaller the value, the more the prediction result tends to the historical data;

[0024] Then, the data is used to calculate the predicted value using the moving average formula. The simple moving average formula is:

[0025]

[0026] Among them, z t+1 represents the predicted value of the next time point t + 1 calculated by the moving average formula, y t-1 represents the actual monitored value of the previous time point t - 1, and k is the window size. The value range is 1 < k < n. The smaller the k value, the more the prediction result tends to the recent data, and the larger the value, the more the prediction result tends to the historical data;

[0027] Take the average of the predicted values of the next time point t + 1 predicted by the two algorithms;

[0028]

[0029] Among them, F t+1 is the predicted value finally analyzed by the fault analysis unit.

[0030] Preferably, the fault identification unit traverses all known fault types and their corresponding eigenvalue ranges;

[0031] For each fault type, compare the predicted value with the eigenvalue range of this fault type;

[0032] If the predicted value falls within the eigenvalue range of this fault type, it is considered that this type of fault has occurred;

[0033] And send a warning message of this type of fault to the warning module, and specific fault type description information can be attached at the same time; if the predicted value is not within the eigenvalue range of any known fault type, no fault warning is issued, but this predicted value is recorded.

[0034] A fault warning method for a high-voltage cascade energy storage converter system comprises the following steps:

[0035] Step 1: Data collection and preprocessing, using various sensors in the monitoring module including temperature sensors, current sensors, voltage sensors, vibration sensors and power sensors to monitor various parts of the high-voltage cascade energy storage converter system in real time, and collect various parameter data including voltage, current, temperature and power;

[0036] Step 2: Perform simple processing on the real-time monitored data, use statistical methods to remove outliers, and use linear interpolation to handle missing values ​​to ensure data integrity;

[0037] Step 3: Processing historical fault data. The fault information processing module collects historical fault data of the high-voltage cascade energy storage converter system from the system maintenance log and fault report, classifies the faults according to the manifestation or location of the fault, and assigns a unique identifier to each fault type. For each fault type, extract the sensor data features related to it, analyze the value range of these features under different fault types, and use statistical analysis methods to comprehensively consider the feature performance of the same fault type in all historical fault data to determine the sensor data feature value range corresponding to each fault type.

[0038] Step 4: Data analysis and prediction. The analysis unit combines the exponential smoothing method and the moving average algorithm to predict faults. First, the monitored data is arranged in time series, and the predicted value of the next time point is calculated using the exponential smoothing method. Then, the predicted value is calculated using the moving average. The predicted values ​​of the next time point predicted by the two algorithms are averaged to obtain the final predicted value.

[0039] Step 5: Fault identification. The fault identification unit traverses all known fault types and their corresponding characteristic value ranges. For each fault type, the predicted value is compared with the characteristic value range of the fault type. If the predicted value falls within the characteristic value range of the fault type, it is considered that this type of fault has occurred, and a warning message of this type of fault is sent to the warning module, along with specific fault type description information. If the predicted value is not within the characteristic value range of any known fault type, no fault warning is issued, but the predicted value is recorded.

[0040] Step 6: Early warning and response. The early warning module issues early warnings based on the early warning analysis results of the analysis module, through sound and light alarms and SMS notifications. The SMS notification also informs relevant staff of the specific fault type.

[0041] Step 7: After receiving the warning information, the staff will inspect and repair the high-voltage cascade energy storage converter system according to the fault cause provided;

[0042] Step 8: System maintenance and update. Regularly maintain and update the fault warning system, including calibrating sensors, optimizing data analysis algorithms, and upgrading warning module warning methods.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention adopts a fault warning system consisting of a monitoring module, an analysis module and a warning module. The analysis unit takes the average of the values ​​predicted by two prediction algorithms as the final prediction value, which greatly improves the accuracy and reliability of the prediction and realizes comprehensive monitoring and accurate warning of the high-voltage cascade energy storage converter system. The advance prediction value effectively improves the problem of delayed or untimely warning in the prior art, and provides sufficient preparation time for system maintenance and response measures.

[0045] 2. The present invention classifies different fault types through a fault data processing module, quickly identifies fault categories through an analysis module, and attaches specific fault type description information when issuing an early warning, thereby achieving accurate positioning and rapid identification of faults, solving the problem that it is difficult to know the specific cause of a fault when an early warning sounds, reducing the troubleshooting and diagnosis time of maintenance personnel, reducing system downtime and maintenance costs, and improving system availability and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a working flow diagram of the system of the present invention;

[0047] Figure 2 is a flow chart of the steps of the method of the present invention;

[0048] Figure 3 It is a schematic diagram of the calculation flow of the analysis unit of the present invention. DETAILED DESCRIPTION

[0049] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0050] The present invention provides a fault warning system for a high-voltage cascade energy storage converter system, comprising:

[0051] A monitoring module, which includes a variety of sensors distributed in various parts of the high-voltage energy storage and conversion system to monitor changes in system parameter data in real time;

[0052] A fault information processing module, which collects historical fault data from system maintenance logs and fault reports, classifies and extracts features of different fault types, and determines the range of sensor data feature values ​​corresponding to each fault type;

[0053] An analysis module, which is connected to the fault information processing module and the monitoring module, and is used to perform prediction calculations on the data monitored by the monitoring module, and identify the fault type in combination with the fault information processing module to compare the specific fault type;

[0054] The early warning module is connected to the analysis module data, and is used to issue early warnings based on the early warning analysis results of the analysis module, and inform relevant staff of the specific fault type.

[0055] Embodiment 1: Figure 1-3 As shown, in this embodiment, a large energy storage station adopts a high-voltage cascade energy storage and conversion system. In order to ensure the stable operation of the system, the fault warning system of the present invention is introduced.

[0056] The temperature sensor, current sensor, voltage sensor, vibration sensor and power sensor in the monitoring module implement monitoring data, and the monitoring module performs simple processing on the real-time monitored data including statistical method to remove outliers and linear interpolation method to process missing values.

[0057] The fault information processing module analyzes the collected historical fault data;

[0058] Classify faults according to their manifestation or location, and assign a unique identifier to each fault type; extract relevant sensor data features for each fault type, and analyze the value range of these features under different fault types;

[0059] Through statistical analysis, the typical value range of the characteristic value under a specific fault type is determined. The characteristic performance of the same fault type in all historical fault data is comprehensively considered to determine the characteristic value range of the sensor data corresponding to each fault type.

[0060] The analysis module includes an analysis unit and a fault identification unit;

[0061] The analysis unit analyzes and predicts the real-time monitoring data of the monitoring module to obtain a predicted value;

[0062] The fault identification unit determines whether the predicted value falls within the characteristic value range of a certain fault type. If so, an early warning of this type of fault is issued.

[0063] The fault data processing module classifies different fault types, and the analysis module quickly identifies the fault category, achieving accurate positioning and rapid identification of the fault, solving the problem of difficulty in knowing the specific cause of the fault when the alarm sounds, reducing the troubleshooting and diagnosis time of maintenance personnel, reducing system downtime and maintenance costs, and improving the availability and stability of the high-voltage cascade energy storage converter system.

[0064] The analysis unit uses the exponential smoothing method combined with the moving average algorithm for fault prediction;

[0065] First, arrange the monitored data in a time series and set it as (y1, y2,..., y n ,), where y is the value of the data point monitored on the time series, n is the number of data points monitored, and the prediction formula is;

[0066] x t+1 = αy t +(1 - α)x t

[0067] Among them, t represents the time point of each collection of monitored data on the time series, x t+1 represents the predicted value of the next time point t + 1 calculated by the exponential smoothing method algorithm, y t represents the actual monitored value of the current time point t, x t is the predicted value of the previous time point, α is the smoothing coefficient, and its value range is 0 < α < 1. The larger the value of α, the more the prediction result tends to recent data, and the smaller the value, the more the prediction result tends to historical data;

[0068] Then, calculate the predicted value of the data using the moving average formula. The simple moving average formula is:

[0069]

[0070] Among them, z t+1 represents the predicted value of the next time point t + 1 calculated by the moving average formula, y t-1 represents the actual monitored value of the previous time point t - 1, k is the window size, and its value range is 1 < k < n. The smaller the value of k, the more the prediction result tends to recent data, and the larger the value, the more the prediction result tends to historical data;

[0071] Take the average of the predicted values of the next time point t + 1 predicted by the two algorithms;

[0072]

[0073] Among them, F t+1 is the predicted value finally analyzed by the fault analysis unit.

[0074] The analysis unit combines two algorithms to more accurately predict the state changes of the system and improve the accuracy and reliability of the prediction. The exponential smoothing method can quickly respond to the changing trend of the data, while the moving average algorithm can smooth the noise. The combination of the two can give full play to their respective advantages and improve the accuracy of the prediction. In addition, by adjusting the smoothing coefficient and window size, it can flexibly adapt to different data characteristics and system states, improving the adaptability of the system.

[0075] The fault identification unit traverses all known fault types and their corresponding characteristic value ranges;

[0076] For each fault type, the predicted value is compared with the characteristic value range of the fault type;

[0077] If the predicted value falls within the characteristic value range of the fault type, it is considered that this type of fault has occurred;

[0078] And send warning information of this type of fault to the warning module, and can also attach specific fault type description information; if the predicted value is not within the characteristic value range of any known fault type, no fault warning is issued, but the predicted value is recorded;

[0079] Through the fault warning system of the monitoring module, analysis module and warning module, the analysis unit takes the average of the values ​​predicted by the two prediction algorithms as the final prediction value, which greatly improves the accuracy and reliability of the prediction, and realizes the comprehensive monitoring and accurate warning of the high-voltage cascade energy storage conversion system. The advance prediction value effectively improves the problem of delayed or untimely warning in the existing technology, and provides sufficient preparation time for system maintenance and response measures.

[0080] Embodiment 2: Figure 1-3 As shown, in this embodiment, a fault warning method for a fault warning system for a high-voltage cascade energy storage converter system applicable to Embodiment 1 includes the following steps:

[0081] Step 1: Data collection and preprocessing. Use the various sensors in the monitoring module, including temperature sensors, current sensors, voltage sensors, vibration sensors, and power sensors, to monitor the various parts of the high-voltage cascade energy storage converter system in real time, and collect multiple parameter data including voltage, current, temperature, and power; providing a rich data basis for subsequent analysis and early warning.

[0082] Step 2: Perform simple processing on the real-time monitored data, use statistical methods to remove outliers, and use linear interpolation to handle missing values ​​to ensure data integrity; avoid interference with subsequent analysis, and handle missing values ​​to ensure data continuity and integrity.

[0083] Step 3: Processing historical fault data. The fault information processing module collects historical fault data of the high-voltage cascade energy storage converter system from the system maintenance log and fault report, classifies the faults according to the manifestation or location of the fault, and assigns a unique identifier to each fault type. For each fault type, extract the sensor data features related to it, analyze the value range of these features under different fault types, and use statistical analysis methods to comprehensively consider the feature performance of the same fault type in all historical fault data to determine the sensor data feature value range corresponding to each fault type.

[0084] Step 4: Data analysis and prediction. The analysis unit combines the exponential smoothing method and the moving average algorithm to predict faults. First, the monitored data is arranged in time series, and the predicted value of the next time point is calculated using the exponential smoothing method. Then, the predicted value is calculated using the moving average. The predicted values ​​of the next time point predicted by the two algorithms are averaged to obtain the final predicted value.

[0085] Step 5: Fault identification. The fault identification unit traverses all known fault types and their corresponding characteristic value ranges. For each fault type, the predicted value is compared with the characteristic value range of the fault type. If the predicted value falls within the characteristic value range of the fault type, it is considered that this type of fault has occurred, and a warning message of this type of fault is sent to the warning module, along with specific fault type description information. If the predicted value is not within the characteristic value range of any known fault type, no fault warning is issued, but the predicted value is recorded.

[0086] Step 6: Early warning and response. The early warning module issues early warnings based on the early warning analysis results of the analysis module, through sound and light alarms and SMS notifications. The SMS notification also informs relevant staff of the specific fault type.

[0087] Step 7: After receiving the warning information, the staff will inspect and repair the high-voltage cascade energy storage converter system according to the fault cause provided;

[0088] Step 8: System maintenance and update: Regularly maintain and update the fault warning system, including calibrating sensors, optimizing data analysis algorithms, and upgrading warning module warning methods;

[0089] Regular maintenance and updates adapt to changes and developments in the system, improve the system's adaptability and reliability, ensure that the fault warning system is always in the best condition, and provide continuous guarantee for the stable operation of the system.

[0090] Through the fault warning system and fault warning method of the present invention, the energy storage station successfully warned of multiple potential faults in advance, providing a strong guarantee for the stable operation of the system. Compared with the prior art, the present invention reduces system downtime and maintenance costs, and improves system availability and stability.

[0091] The embodiments of the present invention are provided for the purpose of illustration and description. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present invention. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A fault warning system for a high-voltage cascade energy storage converter system, characterized in that: include: A monitoring module, which includes a variety of sensors distributed in various parts of the high-voltage energy storage and conversion system to monitor changes in system parameter data in real time; A fault information processing module, which collects historical fault data from system maintenance logs and fault reports, classifies and extracts features of different fault types, and determines the range of sensor data feature values ​​corresponding to each fault type; An analysis module, which is connected to the fault information processing module and the monitoring module, and is used to perform prediction calculations on the data monitored by the monitoring module, and identify the fault type in combination with the fault information processing module to compare the specific fault type; The early warning module is connected to the analysis module data, and is used to issue early warnings based on the early warning analysis results of the analysis module, and inform relevant staff of the specific fault type.

2. A fault warning system for a high voltage cascade energy storage converter system as claimed in claim 1, characterized in that: The sensors in the monitoring module include but are not limited to temperature sensors, current sensors, voltage sensors, vibration sensors and power sensors, and the monitoring module performs simple processing on the real-time monitored data including statistical methods to remove outliers and linear interpolation methods to process missing values.

3. A fault warning system for a high voltage cascade energy storage converter system as claimed in claim 1, characterized in that: The fault information processing module analyzes the collected historical fault data; Classify faults according to their manifestation or location, and assign a unique identifier to each fault type; extract relevant sensor data features for each fault type, and analyze the value range of these features under different fault types; Through statistical analysis, the typical value range of the characteristic value under a specific fault type is determined. The characteristic performance of the same fault type in all historical fault data is comprehensively considered to determine the characteristic value range of the sensor data corresponding to each fault type.

4. A fault warning system for a high voltage cascade energy storage converter system as claimed in claim 3, characterized in that: The analysis module includes an analysis unit and a fault identification unit; The analysis unit performs analysis and prediction based on the real-time monitoring data of the monitoring module to obtain a predicted value; The fault identification unit determines whether the predicted value falls within a characteristic value range of a certain fault type, and if so, issues a warning for this type of fault.

5. A fault warning system for a high voltage cascade energy storage converter system as claimed in claim 4, characterized in that: The analysis unit uses exponential smoothing combined with moving average algorithm to predict faults; First, arrange the monitored data in time series as (y1, y2, ..., y n ,), where y is the value of the data point monitored in the time series, n is the number of monitored data points, and the prediction formula is; x t+1 ay t +(1-α)x t Among them, t represents the time point of each monitoring data collection in the time series, x t+1 represents the predicted value of the next time point t+1 calculated by the exponential smoothing algorithm, y t represents the actual monitoring value at the current time point t, x t is the predicted value at the previous time point, α is the smoothing coefficient, and its value range is 0<α<1. The larger the α value, the more the prediction result tends to the recent data, and the smaller the value, the more the prediction result tends to the historical data. Then use the moving average formula to calculate the predicted value. The simple moving average formula is: Among them, z t+1 represents the predicted value at the next time point t + 1 calculated by the moving average formula, y t-1 represents the actual monitored value at the previous time point t - 1. k is the window size, and its value range is 1 < k < n. The smaller the value of k, the more the prediction result tends to recent data, and the larger the value, the more the prediction result tends to historical data; Take the average of the predicted values ​​of the next time point t+1 predicted by the two algorithms; Among them, F t+1 As the predicted value finally analyzed by the fault analysis unit.

6. A fault warning system for a high voltage cascade energy storage converter system as claimed in claim 4, characterized in that: The fault identification unit traverses all known fault types and their corresponding characteristic value ranges; For each fault type, the predicted value is compared with the characteristic value range of the fault type; If the predicted value falls within the characteristic value range of the fault type, it is considered that this type of fault has occurred; And send warning information of this type of fault to the warning module, and can also attach specific fault type description information; if the predicted value is not within the characteristic value range of any known fault type, no fault warning is issued, but the predicted value is recorded.

7. A fault warning method for a high voltage cascade energy storage converter system applicable to any one of claims 1 to 6, characterized in that: The following steps are involved: Step 1: Data collection: Use various sensors in the monitoring module, including temperature sensors, current sensors, voltage sensors, vibration sensors, and power sensors, to monitor various parts of the high-voltage cascade energy storage converter system in real time, and collect various parameter data including voltage, current, temperature, and power; Step 2: Perform simple processing on the real-time monitored data, use statistical methods to remove outliers, and use linear interpolation to handle missing values ​​to ensure data integrity; Step 3: Processing historical fault data. The fault information processing module collects historical fault data of the high-voltage cascade energy storage converter system from the system maintenance log and fault report, classifies the faults according to the manifestation or location of the fault, and assigns a unique identifier to each fault type. For each fault type, extract the sensor data features related to it, analyze the value range of these features under different fault types, and use statistical analysis methods to comprehensively consider the feature performance of the same fault type in all historical fault data to determine the sensor data feature value range corresponding to each fault type. Step 4: Data analysis and prediction. The analysis unit combines the exponential smoothing method and the moving average algorithm to predict faults. First, the monitored data is arranged in time series, and the predicted value of the next time point is calculated using the exponential smoothing method. Then, the predicted value is calculated using the moving average. The predicted values ​​of the next time point predicted by the two algorithms are averaged to obtain the final predicted value. Step 5: Fault identification. The fault identification unit traverses all known fault types and their corresponding characteristic value ranges. For each fault type, the predicted value is compared with the characteristic value range of the fault type. If the predicted value falls within the characteristic value range of the fault type, it is considered that this type of fault has occurred, and a warning message of this type of fault is sent to the warning module, along with specific fault type description information. If the predicted value is not within the characteristic value range of any known fault type, no fault warning is issued, but the predicted value is recorded. Step 6: Early warning and response. The early warning module issues early warnings based on the early warning analysis results of the analysis module, through sound and light alarms and SMS notifications. The SMS notification also informs relevant staff of the specific fault type. Step 7: After receiving the warning information, the staff will inspect and repair the high-voltage cascade energy storage converter system according to the fault cause provided; Step 8: System maintenance and update. Regularly maintain and update the fault warning system, including calibrating sensors, optimizing data analysis algorithms, and upgrading warning module warning methods.

Images