Intelligent diagnosis device for edge end of electrochemical energy storage system

Abstract

The invention provides an intelligent diagnosis device for an edge end of an electrochemical energy storage system. The intelligent diagnosis device comprises an integrated hardware platform and a display platform, and the integrated hardware platform consists of a sensor array, a data acquisition unit and a communication interface, and is used for monitoring the voltage, current and temperature of the battery cell in real time and transmitting data to the display platform. The display platform comprises a data receiving unit, a data processing unit, an intelligent diagnosis algorithm module unit and an electric core level operation state display module. The intelligent diagnosis algorithm module adopts an extended Kalman method, a forgetting factor least square method, a neural network algorithm and other technologies to calculate the SOC, the internal resistance and the SOH of the battery cell and predict potential faults such as internal short circuit and thermal runaway of the battery cell, and early warning and accurate diagnosis of the faults are achieved. The operation efficiency and safety of the energy storage system can be improved, the maintenance cost is reduced, and the service life of the system is prolonged.

Description

Technical Field

[0001] The present invention relates to the field of battery cell management technology, and more specifically, to an intelligent diagnostic device at the edge of an electrochemical energy storage system. Background Art

[0002] As a key technology in the power system, the development of electrochemical energy storage system has attracted extensive attention. With the high proportion of renewable energy access, the stability and reliability of the power system are facing challenges. Electrochemical energy storage system has become an important means to solve these problems due to its fast response and flexible adjustment capabilities. However, electrochemical energy storage system has frequent inconsistent battery cells, safety issues, and difficulties in fault prediction and diagnosis during operation, which limit its wider application and commercialization process. Traditional energy storage system monitoring methods often rely on manual inspections and simple alarm systems. These methods are not only inefficient, but also difficult to achieve early warning and accurate diagnosis of faults. With the development of technologies such as big data and machine learning, the edge intelligent diagnosis device of electrochemical energy storage system integrating hardware platform and intelligent diagnosis algorithm module has become a hot research topic. These technologies can realize real-time monitoring of key parameters of energy storage system, and use machine learning technology to conduct in-depth analysis of collected data, identify abnormal patterns, and predict potential faults.

[0003] In the process of implementing the embodiments of the present invention, the inventors found that there are at least the following problems or defects in the prior art: traditional monitoring methods cannot accurately monitor the status of battery cells in real time, and it is difficult to achieve early warning and accurate diagnosis of faults, resulting in high maintenance costs, low system safety, and cannot effectively extend the service life of the energy storage system. Summary of the invention

[0004] The present invention provides an edge-end intelligent diagnostic device for an electrochemical energy storage system, comprising:

[0005] An integrated hardware platform and a display platform, wherein the integrated hardware platform is connected to the display platform;

[0006] The integrated hardware platform includes a sensor array, a data acquisition unit and a communication interface, wherein the sensor array is used to monitor the voltage, current and temperature of all cells in the electrochemical energy storage system in real time to monitor the charge and discharge state of the electrochemical energy storage system;

[0007] The data acquisition unit is connected to the sensor array and is used to collect real-time operation data monitored by the sensor array;

[0008] The communication interface is connected to the data acquisition unit and is used to transmit the acquired data to the display platform;

[0009] The display platform includes a data receiving unit, a data processing unit, an intelligent diagnosis algorithm module unit and an electrochemical energy storage system cell-level operating status display module. The data receiving unit is used to receive data transmitted from the communication interface.

[0010] Furthermore, the sensor array also includes a pressure sensor for real-time monitoring of the pressure of all battery cells in the electrochemical energy storage system.

[0011] Furthermore, the data processing unit includes a data input unit, a data category correction unit, a data standardization unit and a data cleaning unit, wherein the data input unit is used to receive an original data set to be processed, the data category correction unit is used to identify and correct errors in the data set, the data standardization unit is used to convert the data into a unified format, and the data cleaning unit is used to clean the data.

[0012] Furthermore, the data cleaning unit adopts a sliding window mechanism, and for randomly missing data, a linear regression estimation method is used to fill it. For sudden outlier data, the window size is determined according to the data characteristics and smoothing requirements. Starting from the starting position of the time series, the data points in the window are summed and the average is calculated to replace the outliers.

[0013] Furthermore, the intelligent diagnosis algorithm module unit includes a battery cell real-time SOC calculation module, a cycle internal resistance calculation module, a cycle SOH calculation module, a battery cell consistency judgment module, a battery cell internal short circuit prediction module and a thermal runaway prediction module. The battery cell real-time SOC calculation module is used to use the extended Kalman method to calculate the SOC of the battery cell based on the real-time voltage and current during the battery cell charging and discharging process. The cycle internal resistance calculation module is used to use the forgetting factor least squares method to calculate the internal resistance of the battery cell based on all the voltages and currents during the battery cell charging and discharging process. The cycle SOH calculation module is used to use a neural network algorithm to calculate the battery cell charging and discharging. The SOH of the battery cell is calculated based on all the voltages, currents and temperatures in the process. The battery cell consistency judgment module is used to calculate the difference between the real-time battery cell voltage, temperature and SOC and the median of the corresponding parameters of multiple battery cells, as well as to calculate the difference between the internal resistance and SOH of the battery cell in each cycle and the median of the corresponding parameters of multiple battery cells, so as to determine whether the battery cell is outlier. The battery cell internal short circuit prediction module and the thermal runaway prediction module are used to extract indicators of related features of self-discharge effect and abnormal temperature rise effect based on machine learning algorithm by combining electrochemical impedance spectroscopy EIS with deep neural network DNN.

[0014] Furthermore, in the battery cell short circuit prediction module and the thermal runaway prediction module, the machine learning algorithm combining the electrochemical impedance spectroscopy EIS with the deep neural network DNN first extracts features, and then predicts the self-discharge effect and the abnormal temperature rise effect based on the extracted features.

[0015] Furthermore, in the battery cell consistency determination module, the medians of the corresponding parameters of the battery cells calculated include the medians of the voltage, temperature, SOC, internal resistance and SOH of the battery cells.

[0016] Furthermore, the electrochemical energy storage system cell-level operating status display module includes an energy storage system cell-level operating status overview module, an energy storage system cell consistency status display module, an energy storage system cell prediction and warning module, and an energy storage system cell fault information display module. The energy storage system cell-level operating status overview module is used to dynamically display the charging and discharging status, operating voltage, current and temperature of the system level, PACK level and cell level in real time in the form of curves, and displays the calculated SOC, internal resistance and SOH at the cell level. The energy storage system cell consistency status display module is used to display the voltage, temperature, internal resistance and SOH consistency diagnosis results of all cells in the energy storage system. The energy storage system cell prediction and warning module is used to dynamically display the voltage and temperature of the charging and discharging cycle predicted by the energy storage system in real time in the form of curves, and dynamically display the probability of internal short circuit and thermal runaway occurring in each cycle in real time. The energy storage system cell fault information display module is used to display the reported cell fault information and the cell number.

[0017] Furthermore, the communication interface supports multiple communication protocols, including but not limited to CAN bus, Modbus and Ethernet protocols.

[0018] Furthermore, the display platform also includes a user interaction module, which is used to receive user operation instructions and adjust display content and diagnostic parameters according to the user instructions.

[0019] According to the above-mentioned embodiments of the present invention, at least the following beneficial effects are achieved: the intelligent diagnostic device at the edge of the electrochemical energy storage system of the present invention can realize real-time monitoring and accurate collection of key parameters such as battery cell voltage, current, and temperature by integrating the hardware platform and the intelligent diagnostic algorithm module. The intelligent diagnostic algorithm module uses advanced technologies such as the extended Kalman method, the forgetting factor least squares method, and the neural network algorithm to conduct in-depth analysis and processing of the collected data, and can accurately calculate the SOC, internal resistance, and SOH of the battery cell, and predict potential faults such as internal short circuit and thermal runaway of the battery cell, thereby achieving early warning and accurate diagnosis of faults. In addition, the device can also clearly display the operating status and fault information of the battery cell through the display platform, providing an intuitive reference for operation and maintenance personnel.

[0020] Through the above functions, the device of the present invention can significantly improve the operating efficiency and safety of the electrochemical energy storage system. Real-time monitoring and intelligent diagnosis capabilities enable the system to detect and handle potential faults in a timely manner, reduce the risk of faults, and ensure the stable operation of the energy storage system. At the same time, the early warning function can reduce downtime and maintenance costs caused by faults, extend the service life of the battery cells, and improve the economy and reliability of the energy storage system. The device is suitable for electrochemical energy storage systems of various sizes and provides strong technical support for the stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:

[0022] Figure 1 A schematic diagram of the structure of an intelligent diagnostic device at the edge of an electrochemical energy storage system provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0023] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0024] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present invention can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0025] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0026] Reference below Figure 1 , Figure 1 This is a schematic diagram of the structure of an intelligent diagnostic device for an electrochemical energy storage system edge provided by an embodiment of the present invention. Figure 1 As shown, an electrochemical energy storage system edge end intelligent diagnosis device 100 includes:

[0027] An integrated hardware platform and a display platform, wherein the integrated hardware platform is connected to the display platform;

[0028] The integrated hardware platform includes a sensor array, a data acquisition unit and a communication interface, wherein the sensor array is used to monitor the voltage, current and temperature of all cells in the electrochemical energy storage system in real time to monitor the charge and discharge state of the electrochemical energy storage system;

[0029] The data acquisition unit is connected to the sensor array and is used to collect real-time operation data monitored by the sensor array;

[0030] The communication interface is connected to the data acquisition unit and is used to transmit the acquired data to the display platform;

[0031] The display platform includes a data receiving unit, a data processing unit, an intelligent diagnosis algorithm module unit and an electrochemical energy storage system cell-level operating status display module. The data receiving unit is used to receive data transmitted from the communication interface.

[0032] It should be noted that the intelligent diagnostic device at the edge of the electrochemical energy storage system of the present invention includes an integrated hardware platform and a display platform, and the two platforms are connected to realize data collection and processing. The integrated hardware platform is composed of a sensor array, a data acquisition unit and a communication interface. The sensor array is used to monitor the voltage, current and temperature of all batteries in the electrochemical energy storage system in real time to monitor the charge and discharge status of the electrochemical energy storage system. The data acquisition unit is connected to the sensor array and is responsible for collecting real-time operating data monitored by the sensor array. The communication interface is used to transmit the collected data to the display platform for further processing and display. The display platform includes a data receiving unit, a data processing unit, an intelligent diagnostic algorithm module unit and an electrochemical energy storage system battery cell level operating status display module, wherein the data receiving unit is used to receive data transmitted from the communication interface.

[0033] Specifically, the sensor array is composed of a variety of sensors, including voltage sensors, current sensors, and temperature sensors. These sensors can monitor the voltage, current, and temperature changes of the battery cells in real time to ensure accurate grasp of the charging and discharging status of the electrochemical energy storage system. The data acquisition unit usually uses a high-precision analog-to-digital converter to convert the analog signals collected by the sensor into digital signals for subsequent data processing and analysis. The communication interface can use a variety of communication protocols, such as CAN bus, Modbus, or Ethernet protocol, to achieve efficient data transmission with the display platform. The data processing unit of the display platform includes a data input unit, a data category correction unit, a data standardization unit, and a data cleaning unit. These units work together to pre-process the collected data to ensure the accuracy and consistency of the data.

[0034] Preferably, the data cleaning unit can use a sliding window mechanism to process missing values ​​and outliers in the data. For randomly missing data, a linear regression estimation method can be used to fill it to ensure the integrity of the data. For sudden outlier data, the window size can be determined according to the data characteristics and smoothing requirements. Starting from the starting position of the time series, the data points in the window are summed and the average value is calculated to replace the outliers.

[0035] Furthermore, the real-time SOC calculation module of the battery cell in the intelligent diagnosis algorithm module can use the extended Kalman method to calculate the real-time voltage and current during the charging and discharging process of the battery cell to obtain the SOC of the battery cell. The cycle internal resistance calculation module can use the forgetting factor least squares method to calculate all the voltages and currents during the charging and discharging process of the battery cell to obtain the internal resistance of the battery cell. The combined use of these algorithms can improve the accuracy and reliability of diagnosis.

[0036] In some embodiments, the sensor array further includes a pressure sensor for real-time monitoring of the pressure of all battery cells in the electrochemical energy storage system.

[0037] It should be noted that the sensor array plays a key role in the intelligent diagnostic device at the edge of the electrochemical energy storage system. The sensor array includes not only basic sensors for monitoring the voltage, current and temperature of the battery cells, but also expanded to include pressure sensors. The pressure sensor is used to monitor the pressure changes of all battery cells in the electrochemical energy storage system in real time. This extension enables the device to monitor the operating status of the battery cells more comprehensively, thereby more accurately assessing the health status and potential safety risks of the battery cells. The addition of pressure sensors provides an additional dimension for the comprehensive diagnosis of battery cells, which helps to promptly detect abnormal conditions of the battery cells, such as expansion, thereby improving the safety and reliability of the system.

[0038] Specifically, the pressure sensor can adopt piezoelectric or piezoresistive sensors, which can convert the pressure changes of the battery cell into electrical signal output. In practical applications, the pressure sensor can be installed in the casing of the battery cell or in a position in direct contact with the battery cell to accurately monitor the pressure changes of the battery cell. The measurement range and accuracy of the sensor can be set according to the working pressure range of the battery cell and the system's requirements for pressure monitoring accuracy. For example, for common lithium-ion batteries, the measurement range of the pressure sensor can be set between 0 and 5MPa, and the accuracy can reach 0.01MPa. In addition, the data acquisition frequency of the pressure sensor should also be consistent with that of the voltage, current and temperature sensors to ensure the synchronization and integrity of the data, and it can usually be set to collect data once per second.

[0039] Preferably, in order to improve the monitoring effect of the pressure sensor and the reliability of the system, multiple pressure sensors can be used to redundantly monitor the battery cells. For example, by installing pressure sensors at multiple locations in a battery cell, or installing pressure sensors in multiple battery cells, by comparing the monitoring data of different sensors, the pressure change of the battery cell can be judged more accurately, avoiding misjudgment due to failure or error of a single sensor.

[0040] Furthermore, data from other sensors, such as temperature sensors, can be combined for comprehensive analysis to more comprehensively assess the health of the battery cell. For example, when the pressure and temperature of the battery cell increase at the same time, it may indicate that the battery cell is at risk of overheating or internal reaction, and timely measures need to be taken to deal with it.

[0041] In some embodiments, the data processing unit includes a data input unit, a data category correction unit, a data standardization unit and a data cleaning unit, wherein the data input unit is used to receive an original data set to be processed, the data category correction unit is used to identify and correct errors in the data set, the data standardization unit is used to convert the data into a unified format, and the data cleaning unit is used to clean the data.

[0042] It should be noted that the data processing unit plays a vital role in the edge intelligent diagnostic device of the electrochemical energy storage system. It includes a data input unit, a data category correction unit, a data standardization unit and a data cleaning unit. The data input unit is used to receive raw data sets from the sensor array and the data acquisition unit. These data sets contain real-time monitoring data of key parameters such as voltage, current, and temperature of the battery cell. The data category correction unit is responsible for identifying and correcting errors in the data set, such as spelling errors or inconsistent formats, to ensure the accuracy and consistency of the data. The data standardization unit converts the data into a unified format, such as a unified date format and unit, for subsequent processing and analysis. The data cleaning unit is used to clean the data, remove noise and outliers, and improve the quality and reliability of the data.

[0043] Specifically, the data input unit can use a variety of data interfaces and protocols, such as serial communication interfaces, Ethernet interfaces, etc., to achieve efficient connection and data transmission with the sensor array and the data acquisition unit. The data category correction unit can use algorithms or software tools, such as regular expressions or data verification programs, to automatically identify and correct errors in the data set. The data standardization unit can convert the data into a unified format according to preset rules and standards, such as unifying the date format to YYYY-MM-DD, unifying the voltage unit to volts, etc. The data cleaning unit can use statistical methods and machine learning algorithms, such as sliding window mechanisms and linear regression estimation methods, to identify and process noise and outliers in the data.

[0044] Preferably, the data cleaning unit can be further refined into multiple submodules to achieve a more efficient cleaning effect. For example, a missing value processing submodule can be set up to specifically process missing values ​​in the data, using interpolation, mean filling, and other methods to fill them; an outlier detection submodule can be set up to detect and process outliers in the data, using standard deviation method, box plot method, and other methods to identify and replace them.

[0045] Furthermore, data fusion technology can be introduced to fuse and integrate data from different sensors and data sources to improve data integrity and accuracy. For example, the data of voltage sensors and current sensors can be fused to calculate parameters such as power and energy consumption of the battery cell, providing more comprehensive data support for performance evaluation and fault diagnosis of the battery cell.

[0046] In some embodiments, the data cleaning unit adopts a sliding window mechanism, and uses a linear regression estimation method to fill in randomly missing data. For sudden outlier data, the window size is determined according to data characteristics and smoothing requirements. Starting from the starting position of the time series, the data points in the window are summed and the average is calculated to replace the outliers.

[0047] It should be noted that the data cleaning unit uses a sliding window mechanism to improve the accuracy and consistency of the data when processing the data collected in the electrochemical energy storage system. The sliding window mechanism is a commonly used data processing technology that processes and analyzes the data in the window by sliding a fixed-size window on the time series data. For randomly missing data, the data cleaning unit uses a linear regression estimation method to fill it in, that is, a linear regression model is established through existing data points to predict the value of the missing data points. For sudden outlier data, the data cleaning unit determines the window size based on the data characteristics and smoothing requirements, and starts from the starting position of the time series, sums the data points in the window and calculates the average value to replace the outliers, thereby smoothing data fluctuations and improving data stability.

[0048] Specifically, the size of the sliding window can be set according to the sampling frequency of the data and the system's requirements for data smoothness. For example, if the sampling frequency of the data is once per second, the size of the sliding window can be set to 10 seconds or 30 seconds to cover enough data points for smoothing. In the linear regression estimation method, the parameters of the linear regression model can be solved by the least squares method, that is, to find a straight line so that the sum of the vertical distances of all data points to the straight line is minimized, thereby achieving accurate prediction of missing data points. For the processing of outliers, a threshold can be set according to the distribution characteristics of the data and the system's requirements for data stability. When the value of a data point exceeds the threshold, it is regarded as an outlier and replaced with the average value of other data points in the window.

[0049] Preferably, the data cleaning unit can also be combined with other data processing technologies, such as filtering algorithms and anomaly detection algorithms, to further improve the accuracy and reliability of the data. For example, a low-pass filter can be used to filter the data to remove high-frequency noise and retain low-frequency valid signals; an anomaly detection algorithm based on statistics, such as the standard deviation method or the box-and-whisker method, can be used to identify and mark outliers in the data for subsequent processing and analysis.

[0050] Furthermore, the size of the sliding window and the parameters of data processing can be dynamically adjusted according to the actual application scenarios and data characteristics of the system to adapt to different data processing needs and system performance requirements. For example, in the early stage of system operation, a larger sliding window can be set to quickly smooth data fluctuations; after the system runs stably, the size of the sliding window can be appropriately reduced to improve the accuracy and efficiency of data processing.

[0051] In some embodiments, the intelligent diagnosis algorithm module unit includes a battery cell real-time SOC calculation module, a cycle internal resistance calculation module, a cycle SOH calculation module, a battery cell consistency judgment module, a battery cell internal short circuit prediction module and a thermal runaway prediction module. The battery cell real-time SOC calculation module is used to use the extended Kalman method to calculate the SOC of the battery cell for the real-time voltage and current during the battery cell charging and discharging process. The cycle internal resistance calculation module is used to use the forgetting factor least squares method to calculate the internal resistance of the battery cell for all voltages and currents during the battery cell charging and discharging process. The cycle SOH calculation module is used to use a neural network algorithm to calculate the battery cell charging and discharging process. The SOH of the battery cell is calculated using all voltages, currents and temperatures during the discharge process. The battery cell consistency determination module is used to calculate the difference between the real-time battery cell voltage, temperature and SOC and the median of the corresponding parameters of multiple battery cells, as well as to calculate the difference between the internal resistance and SOH of the battery cell in each cycle and the median of the corresponding parameters of multiple battery cells, so as to determine whether the battery cell is outlier. The battery cell internal short circuit prediction module and the thermal runaway prediction module are used to extract indicators of related features of self-discharge effect and abnormal temperature rise effect based on a machine learning algorithm by combining electrochemical impedance spectroscopy EIS with a deep neural network DNN.

[0052] It should be noted that the intelligent diagnosis algorithm module unit is the core component of the edge intelligent diagnosis device of the electrochemical energy storage system. It uses a variety of algorithm modules to perform real-time calculation and analysis on the state of the battery cell to achieve accurate assessment of the health status of the battery cell and early warning of potential faults. The real-time SOC calculation module of the battery cell adopts the extended Kalman filter algorithm to estimate the SOC (state of charge) of the battery cell, that is, the current percentage of the battery cell, based on the real-time voltage and current data during the charging and discharging process of the battery cell. The cycle internal resistance calculation module uses the forgetting factor least squares method to fit all the voltage and current data during the charging and discharging process of the battery cell to calculate the internal resistance of the battery cell, that is, the degree of resistance to the flow of current during the charging and discharging process of the battery cell. The cycle SOH calculation module uses a neural network algorithm, combined with the voltage, current and temperature data during the charging and discharging process of the battery cell, to evaluate the SOH (health state) of the battery cell, that is, the degree of degradation of the performance of the battery cell compared with the new battery cell. The cell consistency determination module determines whether the cell is outlier, that is, whether there is a consistency problem, by calculating the difference between the real-time cell voltage, temperature, SOC and other parameters and the median of the corresponding parameters of multiple cells. The cell internal short circuit prediction module and thermal runaway prediction module combine electrochemical impedance spectroscopy (EIS) and deep neural network (DNN) to extract the relevant features of self-discharge effect and abnormal temperature rise effect, and predict potential faults such as internal short circuit and thermal runaway of the cell.

[0053] Specifically, the extended Kalman filter algorithm is a recursive estimation algorithm that continuously updates the estimated value of the cell SOC by establishing a battery model of the cell and taking the real-time voltage and current data as observation values. When calculating the internal resistance, the forgetting factor is introduced in the least squares method to give a higher weight to the recent data, so that the calculation result of the internal resistance can better reflect the current state of the cell. In the SOH calculation, the neural network algorithm trains a neural network model to learn the mapping relationship between the cell voltage, current and temperature data and the SOH, thereby achieving an accurate assessment of the cell SOH. In the cell consistency judgment module, the median is the value in the middle position after sorting the parameter values ​​of multiple cells by size. By calculating the difference between the real-time cell parameters and the median, it can be determined whether the cell deviates from the normal range. In the cell internal short circuit prediction module and the thermal runaway prediction module, EIS is used to measure the impedance characteristics of the cell at different frequencies, and DNN predicts the potential faults of the cell by learning the relationship between EIS data and fault characteristics.

[0054] Preferably, the intelligent diagnosis algorithm module unit can also optimize and adjust the algorithm according to the type and application scenario of the battery cell. For example, for different types of lithium-ion batteries, their battery models and parameters may be different. For different types of batteries, the battery model in the extended Kalman filter algorithm can be established and optimized respectively to improve the accuracy of SOC calculation. When calculating the internal resistance, the size of the forgetting factor can be dynamically adjusted according to the charging and discharging characteristics of the battery cell, so that the calculation result of the internal resistance can better adapt to the actual working state of the battery cell. For SOH calculation, more influencing factors can be introduced, such as the number of cycles and service life of the battery cell, to further improve the accuracy of SOH evaluation. In the battery cell consistency judgment module, in addition to calculating the difference with the median, the difference with the mean, standard deviation and other statistical quantities can also be calculated to comprehensively judge the consistency of the battery cell. For the battery cell internal short circuit and thermal runaway prediction module, more fault characteristics and data sources can be combined, such as vibration data and acoustic emission data of the battery cell, to improve the accuracy and reliability of fault prediction.

[0055] In some embodiments, in the battery cell short circuit prediction module and the thermal runaway prediction module, the machine learning algorithm combining the electrochemical impedance spectroscopy EIS with the deep neural network DNN first performs feature extraction, and then predicts the self-discharge effect and abnormal temperature rise effect based on the extracted features.

[0056] It should be noted that the short circuit prediction module in the battery cell and the thermal runaway prediction module predict potential failures of the battery cell by combining electrochemical impedance spectroscopy (EIS) with deep neural network (DNN). Electrochemical impedance spectroscopy is a technology that measures the impedance characteristics of the battery cell at different frequencies, which can reflect the internal structure and electrochemical reaction characteristics of the battery cell. Deep neural network is a machine learning algorithm based on artificial neural network, which has powerful nonlinear mapping and feature extraction capabilities. By inputting EIS data into DNN, characteristic indicators related to short circuit and thermal runaway in the battery cell can be extracted, thereby achieving early warning and prediction of these potential failures. This method can effectively identify abnormal conditions that may occur in the battery cell during charging and discharging, and improve the safety and reliability of the system.

[0057] Specifically, the measurement of electrochemical impedance spectroscopy can be achieved by applying AC signals of different frequencies to the battery cell and measuring the impedance value of its response. The frequency range of the measurement can be selected according to the characteristics of the battery cell, usually covering the range from low frequency to high frequency, so as to capture the electrochemical reaction process at different time scales inside the battery cell. The structure of a deep neural network can include multiple hidden layers and neurons. The specific number of layers and neurons can be adjusted according to the complexity of the data and the needs of the prediction task. For example, a DNN structure containing 3 hidden layers and 128 neurons in each layer can be used to achieve full learning and feature extraction of EIS data. When training the DNN model, a large amount of historical EIS data and corresponding fault labels can be used as training samples, and the back propagation algorithm and gradient descent method can be used to optimize the parameters of the model so that it can accurately predict faults such as internal short circuit and thermal runaway of the battery cell.

[0058] Preferably, in order to improve the accuracy and robustness of the prediction, regularization techniques such as L1 regularization or L2 regularization can be introduced into the DNN model to prevent overfitting of the model. In addition, an ensemble learning method can be used to integrate multiple DNN models to combine the prediction results of multiple models to further improve the accuracy and stability of the prediction. For example, an ensemble learning algorithm such as a random forest or a gradient boosting tree can be used to combine the prediction results of multiple DNN models to achieve a comprehensive judgment and early warning of battery cell failures. In addition, other fault detection methods, such as rule-based fault detection algorithms, can be combined to complement the method of combining EIS and DNN to further improve the comprehensiveness and reliability of battery cell fault prediction.

[0059] In some embodiments, in the battery cell consistency determination module, the medians of the corresponding parameters of the battery cells calculated include the medians of the voltage, temperature, SOC, internal resistance and SOH of the battery cells.

[0060] It should be noted that the cell consistency judgment module is used in the edge intelligent diagnostic device of the electrochemical energy storage system to evaluate the performance consistency between cells. This module determines whether the cell is outlier, that is, whether there is a consistency problem, by calculating the difference between the real-time cell voltage, temperature, SOC, internal resistance and SOH parameters and the median of the corresponding parameters of multiple cells. The median is a statistic that arranges a set of data in order of size and is located in the middle. It can effectively reflect the central trend of the data and has strong robustness to outliers. By comparing with the median, cells that deviate from the normal range can be intuitively identified, thereby providing a basis for the maintenance and management of the cells. This method helps to detect deviations in cell performance in a timely manner, ensuring the stable operation of the energy storage system and the balanced use of cells.

[0061] Specifically, the parameter calculation in the cell consistency judgment module can adopt the following steps: first, collect parameter data such as voltage, temperature, SOC, internal resistance and SOH of multiple cells under the same conditions; then, sort the data of each parameter and find the median; then, calculate the difference between the parameter value of each cell and the median to obtain the outlier metric; finally, judge whether the cell is outlier according to the preset threshold. For example, for the voltage parameter, a threshold value of 0.1V can be set. When the difference between the voltage of a cell and the median exceeds 0.1V, the cell is considered to be outlier. In addition, other statistics, such as mean and standard deviation, can be combined for comprehensive judgment to improve the accuracy and reliability of consistency judgment. For example, the mean and standard deviation of the cell parameters can be calculated, and then the outlier metric can be compared with the standard deviation to further judge the consistency of the cell.

[0062] Preferably, in order to improve the efficiency and accuracy of cell consistency determination, an automated and intelligent data processing method can be used. For example, a dedicated algorithm or software tool can be developed to automatically collect and process cell parameter data, quickly calculate the median and outlier metrics, and perform consistency determination.

[0063] Furthermore, machine learning technology can be introduced to learn the relationship between battery cell parameters and consistency through training models, so as to realize intelligent prediction and judgment of battery cell consistency. For example, clustering algorithms can be used to group battery cells and identify groups of battery cells with similar performance, so as to more accurately evaluate the consistency of battery cells. At the same time, real-time monitoring and historical data analysis can be combined to dynamically adjust the threshold and parameters of consistency judgment to adapt to changes in battery cell performance and the actual situation of system operation, so as to further improve the accuracy and reliability of battery cell consistency judgment.

[0064] In some embodiments, the electrochemical energy storage system cell-level operating status display module includes an energy storage system cell-level operating status overview module, an energy storage system cell consistency status display module, an energy storage system cell prediction and warning module, and an energy storage system cell fault information display module. The energy storage system cell-level operating status overview module is used to dynamically display the charging and discharging status, operating voltage, current and temperature of the system level, PACK level and cell level in real time in the form of curves, and displays the calculated SOC, internal resistance and SOH at the cell level. The energy storage system cell consistency status display module is used to display the voltage, temperature, internal resistance and SOH consistency diagnosis results of all cells in the energy storage system. The energy storage system cell prediction and warning module is used to dynamically display the voltage and temperature of the next 14 charging and discharging cycles predicted by the energy storage system in real time in the form of curves, and dynamically display the probability of internal short circuit and thermal runaway occurring in each cycle in real time. The energy storage system cell fault information display module is used to display the reported cell fault information and the cell number.

[0065] It should be noted that the cell-level operation status display module of the electrochemical energy storage system is an important component of the edge-end intelligent diagnostic device of the electrochemical energy storage system, and is used to intuitively display the operation status and related information of the cell. The module includes an energy storage system cell-level operation status overview module, an energy storage system cell consistency status display module, an energy storage system cell prediction and warning module, and an energy storage system cell fault information display module. The energy storage system cell-level operation status overview module is used to dynamically display the charge and discharge status, operating voltage, current, and temperature of the system level, PACK level, and cell level in real time in a curve form, so that users can fully understand the real-time operation of the cell. The energy storage system cell consistency status display module is used to display the voltage, temperature, internal resistance, and SOH consistency diagnosis results of all cells in the energy storage system to determine whether there is a performance deviation between the cells. The energy storage system cell prediction and warning module is used to dynamically display the voltage and temperature of the next 14 charge and discharge cycles predicted by the energy storage system in real time in a curve form, and dynamically display the probability of internal short circuit and thermal runaway occurring in each cycle in real time to warn of potential faults in advance. The energy storage system battery cell fault information display module is used to display the reported battery cell fault information and the battery cell number, so that users can find and handle the fault in time.

[0066] Specifically, the energy storage system cell-level operation status overview module can display the real-time data of the cell through a graphical interface, such as using a line chart or a curve chart to represent the change trend of the voltage, current and temperature of the cell over time. At the cell level, the calculated parameters such as SOC, internal resistance and SOH can also be displayed to facilitate users to have a more comprehensive understanding of the status of the cell. The energy storage system cell consistency status display module can use color coding or level division to represent the consistency status of the cell, for example, the consistency status is divided into four levels of excellent, good, qualified and unqualified, and different colors are used to distinguish them, so that users can quickly identify the consistency problems of the cell. The energy storage system cell prediction and warning module can set different warning levels and colors according to the prediction results. For example, when the probability of internal short circuit or thermal runaway is high, a red warning is displayed; when the probability is low, a yellow warning is displayed. The energy storage system cell fault information display module can be displayed in a position diagram with a summary of fault information, for example, the location of the faulty cell is marked on the schematic diagram of the cell, and the specific information of the fault is displayed next to it, such as the fault type, occurrence time, etc.

[0067] Preferably, in order to improve the user experience and practicality of the cell-level operating status display module of the electrochemical energy storage system, the interface design and function settings can be further optimized. For example, an interactive interface can be provided to allow the user to select different display contents and parameters as needed, such as only displaying the voltage and current of the cell, or only displaying the consistency status of the cell, etc. A data export function can also be added to allow the user to export the displayed data into files in formats such as Excel or CSV for further analysis and recording.

[0068] Furthermore, the layout and functions of the display module can be continuously optimized based on user needs and feedback to improve its usability and flexibility. For example, the function of fault analysis and diagnosis suggestion can be added. When a battery cell fails, it can not only display the fault information, but also provide possible fault causes and handling suggestions to help users better handle faults and perform maintenance.

[0069] In some embodiments, the communication interface supports multiple communication protocols, including but not limited to CAN bus, Modbus and Ethernet protocols.

[0070] It should be noted that the communication interface plays a key role in data transmission in the intelligent diagnostic device at the edge of the electrochemical energy storage system. It supports multiple communication protocols to ensure that data can be transmitted efficiently and accurately between different devices and systems. The protocols that can be used by the communication interface include but are not limited to CAN bus, Modbus and Ethernet protocols. CAN bus is a serial communication protocol commonly used in the fields of automobiles and industrial automation. It has a high transmission rate and reliability and is suitable for data transmission scenarios with high real-time requirements. The Modbus protocol is a communication protocol widely used in the field of industrial automation. It supports multiple communication media, such as serial communication and Ethernet, and has good compatibility and scalability. The Ethernet protocol is currently the most commonly used LAN communication protocol. It has the advantages of high transmission rate and wide coverage, and is suitable for large-scale data transmission and networked systems.

[0071] Specifically, the communication interface can be configured and set by selecting the appropriate communication protocol according to the actual needs and application scenarios of the system. For example, in scenarios where real-time monitoring and control of the battery status is required, the CAN bus protocol can be selected to achieve fast data transmission and real-time communication response. In scenarios where data exchange with other industrial automation equipment is required, the Modbus protocol can be selected to achieve compatibility and communication with different devices. In scenarios where large-scale data transmission and network management are required, the Ethernet protocol can be selected to achieve efficient data transmission and the construction of a networked system.

[0072] More specifically, the communication interface can also support multiple communication media, such as wired communication and wireless communication, to meet the communication needs in different scenarios. For example, wired communication can be used to connect devices and systems through cables to ensure the stability and reliability of data transmission; wireless communication can also be used to connect devices and systems through wireless networks to improve the flexibility and convenience of the system.

[0073] Preferably, in order to improve the performance and reliability of the communication interface, some optimization measures and technologies can be adopted. For example, the communication interface can be designed to be redundant, that is, multiple communication interfaces can be set in the system. When one interface fails, other interfaces can automatically take over the data transmission task, thereby ensuring the normal operation of the system and the continuous transmission of data. Data encryption and authentication technology can also be used to encrypt and authenticate the transmitted data to improve the security and confidentiality of the data and prevent the data from being illegally intercepted or tampered with.

[0074] Furthermore, the communication interface can also be used for performance monitoring and fault diagnosis, real-time monitoring of the operating status and performance indicators of the communication interface, timely detection and processing of communication faults, and ensuring the stability and reliability of the communication interface. For example, a performance monitoring module of the communication interface can be set up to collect and analyze the transmission rate, bit error rate and other performance indicators of the communication interface in real time; a fault diagnosis module of the communication interface can also be set up to automatically diagnose and locate communication faults based on the operating status and performance indicators of the communication interface, and provide corresponding fault handling suggestions and solutions.

[0075] In some embodiments, the display platform further includes a user interaction module for receiving user operation instructions and adjusting display content and diagnostic parameters according to the user instructions.

[0076] It should be noted that the user interaction module provides an interactive interface between the user and the system in the edge intelligent diagnosis device of the electrochemical energy storage system, so that the user can adjust and set the display content and diagnostic parameters according to their own needs and operating habits. The user interaction module usually includes a graphical user interface (GUI) and input devices, such as a touch screen, keyboard, and mouse. The graphical user interface is the main interface for the user to interact with the system. It displays the information and functions of the system in a graphical way, so that the user can intuitively understand the status and operation methods of the system. The input device is used to receive the user's operating instructions, such as clicking, dragging, and inputting, so that the system can perform corresponding operations and functions according to the user's instructions. The design and implementation of the user interaction module makes the edge intelligent diagnosis device of the electrochemical energy storage system more user-friendly and easy to use, and improves the user's operating experience and the practicality of the system.

[0077] Specifically, the graphical user interface of the user interaction module can adopt a variety of layouts and design styles to meet the needs and aesthetics of different users. For example, a concise and clear layout style can be used to display the main functions and information of the system with clear icons and texts, so that users can quickly find the required functions and information; or a beautiful and generous design style can be used to enhance the aesthetics of the interface and the user's visual experience through color matching and the use of graphic elements. Input devices can be selected and configured according to the actual application scenarios of the system and the needs of users. For example, on a touch screen device, users can operate by touching the screen, such as clicking icons, dragging sliders, etc.; on a desktop device, users can operate by using the keyboard and mouse, such as entering text, clicking buttons, etc.

[0078] More specifically, the user interaction module can also provide a variety of interaction methods, such as voice interaction, gesture interaction, etc., to meet the operating habits and needs of different users. For example, users can control the operation of the system through voice commands, such as displaying the battery cell voltage, adjusting diagnostic parameters, etc.; they can also interact through gestures, such as waving to switch interfaces or perform other operations.

[0079] Preferably, in order to further improve the interactivity and flexibility of the user interaction module, some intelligent and personalized designs can be introduced. For example, artificial intelligence technology can be used to automatically adjust the interface layout and function settings according to the user's operating habits and preferences to achieve a personalized interactive experience. User customization functions can also be provided to allow users to customize the interface layout, color, font, etc., as well as the system's diagnostic parameters and operating procedures, etc. according to their own needs and preferences.

[0080] Furthermore, user feedback and learning mechanisms can be added to continuously optimize and improve the design and functions of the user interaction module by collecting user operation data and feedback information, so as to improve user satisfaction and the user experience of the system. For example, a user feedback button can be set so that users can provide feedback on their usage experience and suggestions at any time; machine learning algorithms can also be used to analyze and learn user operation data, automatically optimize interface layout and function settings, and better meet user needs and habits.

[0081] The above-mentioned embodiments of the present invention have the following beneficial effects: the present invention can realize real-time monitoring and data collection of key parameters such as voltage, current, temperature, etc. of the battery cell in the electrochemical energy storage system by integrating the hardware platform and the display platform. The diversity of the sensor array and the high efficiency of the data acquisition unit ensure the accuracy and integrity of the data. The intelligent diagnosis algorithm module adopts advanced machine learning techniques, such as the extended Kalman method, the forgetting factor least squares method, and the neural network algorithm, etc., which can conduct in-depth analysis of the collected data, accurately calculate the SOC, internal resistance and SOH of the battery cell, and predict potential internal short circuits and thermal runaway faults, thereby achieving early warning and accurate diagnosis of faults. These functions enable the device to detect abnormal conditions in time during the operation of the battery cell, thereby improving the safety and reliability of the system.

[0082] In addition, the data cleaning unit adopts a sliding window mechanism and a linear regression estimation method, which can effectively handle missing values ​​and outliers in the data, ensure the accuracy and consistency of the data, and provide a reliable data basis for subsequent intelligent diagnosis. The communication interface supports a variety of communication protocols, such as CAN bus, Modbus, and Ethernet protocols, so that the device can flexibly exchange and communicate data with other systems or devices, enhancing the compatibility and scalability of the system. The cell-level operating status display module of the display platform can clearly display the operating status and fault information of the cell, and the user interaction module can adjust the display content and diagnostic parameters according to the user's operating instructions, improving the user experience and practicality of the device. These functions enable the device to not only meet the needs of professional operation and maintenance personnel, but also facilitate operation and monitoring by ordinary users, further improving the management efficiency and maintenance convenience of the electrochemical energy storage system.

[0083] Furthermore, the storage medium of the embodiment of the present application stores program instructions that can implement all the above methods, wherein the program instructions can be stored in the above storage medium in the form of a software product, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or terminal devices such as a computer, a server, a mobile phone, and a tablet.

[0084] The above descriptions are only some preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention to form a technical solution.

Claims

1. An intelligent diagnostic device at the edge of an electrochemical energy storage system, characterized in that: It includes an integrated hardware platform and a display platform, wherein the integrated hardware platform is connected to the display platform; The integrated hardware platform includes a sensor array, a data acquisition unit and a communication interface, wherein the sensor array is used to monitor the voltage, current and temperature of all cells in the electrochemical energy storage system in real time to monitor the charge and discharge state of the electrochemical energy storage system; The data acquisition unit is connected to the sensor array and is used to collect real-time operation data monitored by the sensor array; The communication interface is connected to the data acquisition unit and is used to transmit the acquired data to the display platform; The display platform includes a data receiving unit, a data processing unit, an intelligent diagnosis algorithm module unit and an electrochemical energy storage system cell-level operating status display module. The data receiving unit is used to receive data transmitted from the communication interface.

2. The electrochemical energy storage system edge intelligent diagnosis device according to claim 1, characterized in that: The sensor array also includes a pressure sensor for real-time monitoring of the pressure of all cells in the electrochemical energy storage system.

3. The electrochemical energy storage system edge intelligent diagnosis device according to claim 1, characterized in that: The data processing unit includes a data input unit, a data category correction unit, a data standardization unit and a data cleaning unit. The data input unit is used to receive the original data set to be processed, the data category correction unit is used to identify and correct errors in the data set, the data standardization unit is used to convert the data into a unified format, and the data cleaning unit is used to clean the data.

4. The electrochemical energy storage system edge intelligent diagnosis device according to claim 3, characterized in that: The data cleaning unit adopts a sliding window mechanism. For randomly missing data, a linear regression estimation method is used to fill it. For sudden outlier data, the window size is determined according to the data characteristics and smoothing requirements. Starting from the starting position of the time series, the data points in the window are summed and the average is calculated to replace the outliers.

5. The electrochemical energy storage system edge intelligent diagnosis device according to claim 1, characterized in that: The intelligent diagnosis algorithm module unit includes a battery cell real-time SOC calculation module, a cycle internal resistance calculation module, a cycle SOH calculation module, a battery cell consistency judgment module, a battery cell internal short circuit prediction module and a thermal runaway prediction module. The battery cell real-time SOC calculation module is used to use the extended Kalman method to calculate the SOC of the battery cell based on the real-time voltage and current during the battery cell charging and discharging process. The cycle internal resistance calculation module is used to use the forgetting factor least squares method to calculate the internal resistance of the battery cell based on all the voltages and currents during the battery cell charging and discharging process. The cycle SOH calculation module is used to use a neural network algorithm to calculate the battery cell charging and discharging process. The SOH of the battery cell is calculated based on all the voltages, currents and temperatures in the battery cell. The battery cell consistency judgment module is used to calculate the difference between the real-time battery cell voltage, temperature and SOC and the median of the corresponding parameters of multiple battery cells, and calculate the difference between the internal resistance and SOH of the battery cell in each cycle and the median of the corresponding parameters of multiple battery cells, so as to determine whether the battery cell is outlier. The battery cell internal short circuit prediction module and the thermal runaway prediction module are used to extract indicators of related features of self-discharge effect and abnormal temperature rise effect based on machine learning algorithm by combining electrochemical impedance spectroscopy EIS with deep neural network DNN.

6. The electrochemical energy storage system edge intelligent diagnosis device according to claim 5, characterized in that: In the battery cell short circuit prediction module and the thermal runaway prediction module, the machine learning algorithm combining the electrochemical impedance spectroscopy EIS with the deep neural network DNN first extracts features, and then predicts the self-discharge effect and abnormal temperature rise effect based on the extracted features.

7. The electrochemical energy storage system edge intelligent diagnosis device according to claim 5, characterized in that: In the battery cell consistency determination module, the medians of the corresponding parameters of the battery cells calculated include the medians of the voltage, temperature, SOC, internal resistance and SOH of the battery cells.

8. The electrochemical energy storage system edge intelligent diagnosis device according to claim 1, characterized in that: The electrochemical energy storage system cell-level operating status display module includes an energy storage system cell-level operating status overview module, an energy storage system cell consistency status display module, an energy storage system cell prediction and warning module, and an energy storage system cell fault information display module. The energy storage system cell-level operating status overview module is used to dynamically display the charging and discharging status, operating voltage, current and temperature of the system level, PACK level and cell level in real time in a curve form, and display the calculated SOC, internal resistance and SOH at the cell level. The energy storage system cell consistency status display module is used to display the voltage, temperature, internal resistance and SOH consistency diagnosis results of all cells in the energy storage system. The energy storage system cell prediction and warning module is used to dynamically display the voltage and temperature of the charging and discharging cycle predicted by the energy storage system in real time in a curve form and dynamically display the probability of internal short circuit and thermal runaway occurring in each cycle in real time. The energy storage system cell fault information display module is used to display the reported cell fault information and the cell number.

9. The electrochemical energy storage system edge intelligent diagnosis device according to claim 1, characterized in that: The communication interface supports multiple communication protocols, including but not limited to CAN bus, Modbus and Ethernet protocols.

10. The electrochemical energy storage system edge intelligent diagnosis device according to claim 1, characterized in that: The display platform also includes a user interaction module, which is used to receive user operation instructions and adjust display content and diagnostic parameters according to the user instructions.

Images