Fault prediction method and system for lithium battery

Abstract

The invention relates to the technical field of lithium battery fault prediction, in particular to a lithium battery fault prediction method and system. The method comprises the steps that input current excitation is carried out when a lithium battery is charged online; performing prediction based on current excitation by using a battery electrochemical principle equation to form AI model prediction values of the voltage U and the temperature T; comparing a predicted value of the AI model with an actual detection value of a battery sensor, and if an error exceeds a normal range after comparison, performing boundary judgment on the internal SOC and internal resistance characteristic parameters of the battery calculated by the battery electrochemical principle equation; if the characteristic parameters meet the boundary judgment conditions, determining that the fault is an internal short circuit fault; therefore, the energy storage system has a rapid thermal runaway early warning capability and a lithium ion battery safety and service life prediction function with high prediction accuracy, and the overall safety of the energy storage system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium battery fault prediction, and particularly to a method and system for predicting faults of lithium batteries. Background Art

[0002] With the continuous development of the "dual-high" power system, the demand for electrochemical energy storage in the new power system is also increasing. The electrochemical energy storage system is an indispensable part of the six links of "power generation - power transmission - power distribution - power consumption - power storage", especially as a power storage module for renewable energy such as solar energy, load peak shaving and valley filling of the power grid, and an emergency backup energy storage system for the power grid, etc., which has a wide range of application prospects. In order to meet the voltage and power requirements of the energy storage system, the lithium battery energy storage system usually consists of hundreds or thousands of battery cells connected in series and parallel. For large-scale battery systems, the system may consist of different types, different magnitudes, and different echelon stacks, and the inconsistency of system parameters and performance will be more obvious, resulting in a more serious difference in the attenuation rate of the core life.

[0003] In the aforementioned prior art, during the use stage of lithium batteries, there is no function to predict lithium battery faults, especially it is impossible to quickly warn of the thermal runaway of lithium batteries, and thus it is impossible to improve the overall safety of the energy storage system. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for predicting faults of lithium batteries, to solve the problem that in the prior art, during the use stage of lithium batteries, there is no function to predict lithium battery faults, especially it is impossible to quickly warn of the thermal runaway of lithium batteries, and thus it is impossible to improve the overall safety of the energy storage system.

[0005] To achieve the above purpose, the present invention provides a method for predicting faults of lithium batteries: Apply an input current excitation during the online charging of lithium batteries; Use the battery electrochemical principle equation to perform prediction based on the current excitation to obtain the AI model prediction values of voltage U and temperature T; Compare the model prediction values with the actual detection values of the battery sensor. If the error after comparison exceeds the normal range, then perform boundary judgment on the internal SOC (battery charge) and internal resistance characteristic parameters of the battery calculated by the battery electrochemical principle equation; If the characteristic parameters meet the boundary judgment conditions, it is determined as an internal short circuit fault.

[0006] Among them, in the step of using the battery electrochemical principle equation to perform prediction based on the current excitation to form the model prediction values of voltage U and temperature T: Input a current with a length of 1000 s and a maximum charging current of 1.6 C; Based on the current, charging is carried out using a multi-stage pulsed current charging curve to obtain voltage U and temperature T response values; Perform sensitivity analysis on the battery aging mechanism parameters of the battery based on the multi-stage pulsed current charging curve, where the battery aging mechanism parameters include the maximum lithium-ion concentration at the negative electrode of the battery, the maximum lithium-ion concentration at the positive electrode of the battery, the percentage of active material at the negative electrode of the battery, and the percentage of active material at the positive electrode of the battery.

[0007] Among them, in the step of performing sensitivity analysis on the battery aging mechanism parameters of the battery based on the multi-stage pulsed current charging curve: Based on the battery electrochemical principle equation, calculate various battery aging mechanism parameters to obtain that the maximum lithium-ion concentration at the negative electrode of the battery has the greatest sensitivity; Based on the calculated battery aging mechanism parameters and the capacity estimation formula, obtain the predicted value of the maximum available capacity of the battery.

[0008] Among them, in the step of comparing the model predicted value with the actual detection value of the battery sensor, if the error after comparison exceeds the normal range, then in the step of performing boundary judgment on the internal SOC and internal resistance characteristic parameters of the battery calculated by the battery electrochemical principle equation: Actually detect the voltage U and temperature T parameters of the battery through the sensor; Compare the actual detection value with the predicted voltage and temperature to obtain an error value; Based on the internal short circuit model, set the normal range of the error, and input the error value into the internal short circuit model; When the error value exceeds the normal range determined by the internal short circuit model, initially judge that there is an internal short circuit; Calculate the predicted values of the characteristic parameters of the SOC and internal resistance of the single battery through the battery electrochemical principle equation, and record the changes in the predicted values of the characteristic parameters.

[0009] Among them, in the step of determining an internal short circuit fault if the characteristic parameters meet the boundary judgment conditions: Based on the internal short circuit model, set the characteristic parameter boundary, and input the change in the predicted value of the characteristic parameter into the internal short circuit model for judgment; When the change in the predicted value of the characteristic parameter meets the characteristic parameter boundary, determine that there is an internal short circuit fault; If the change in the predicted value of the characteristic parameter does not meet the characteristic parameter boundary, determine that other faults have occurred.

[0010] The present invention also provides a fault prediction system for a lithium battery, which applies the above-mentioned fault prediction method for a lithium battery and includes a system login module, a station overview module, an operation monitoring module, an alarm setting module, an alarm query module, a model management module, an intelligent operation and maintenance module, and an analysis report module. The system login module, the station overview module, the operation monitoring module, the alarm setting module, the alarm query module, the model management module, the intelligent operation and maintenance module, and the analysis report module are connected in sequence; The system login module is used for staff to log in to the energy storage battery safety diagnosis system; The station overview module is used to display the key indicators of the energy storage station; The operation monitoring module is used to display the information of battery equipment; The alarm setting module is used to set the alarm rules and alarm thresholds of the energy storage station; The alarm query module is used to query alarm information; The model management module is used to configure the energy storage algorithm model; The intelligent operation and maintenance module is used to give analysis and suggestions based on detected anomalies; The analysis report module is used to analyze the battery cluster and generate reports.

[0011] Among them, the operation monitoring module includes a device monitoring unit and a device matrix unit, and the device monitoring unit is connected to the device matrix unit; The device monitoring unit is used to display the device information in the energy storage station in a device tree structure; The device matrix unit is used to visually display the operation status of each layer of device units in the current energy storage station in a card-by-layer manner.

[0012] A fault prediction method and system for a lithium battery according to the present invention: an input current excitation is performed during online charging of the lithium battery; based on the current excitation, predictions are made using the battery electrochemistry principle equation to form AI model prediction values of voltage U and temperature T; the AI model prediction values are compared with the actual detection values of the battery sensor. If the error after comparison exceeds the normal range, boundary judgment is performed on the internal SOC and internal resistance characteristic parameters of the battery calculated by the battery electrochemistry principle equation; if the characteristic parameters meet the boundary judgment conditions, it is determined as an internal short circuit fault; This enables the energy storage system to have the functions of rapid thermal runaway early warning and high-precision prediction of the safety and life of lithium-ion batteries, improving the overall safety and operation and maintenance efficiency of the energy storage system, thus laying a foundation for its application in large-scale energy storage systems such as energy storage power stations; predicting the life of lithium-ion batteries through the battery electrochemical principle equation to ensure that batteries reaching the end of life can be replaced in a timely manner, and establishing a mathematical model between the external stress factors, internal aging mechanism and external health behavior of lithium-ion batteries based on the remaining life prediction theory, providing theoretical support for improving the safe, healthy and efficient operation of the lithium battery energy storage system and ensuring the healthy, long-life and efficient operation of the lithium battery energy storage system. Description of the Drawings

[0013] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art.

[0014] Figure 1 It is the flowchart of the steps for internal short circuit detection of the present invention.

[0015] Figure 2 It is the curve graph of multi-stage pulse current charging of the present invention.

[0016] Figure 3 It is the curve graph of the battery voltage response of the present invention.

[0017] Figure 4 It is the calculation parameter graph of the sensitivity analysis of battery electrochemical aging parameters of the present invention.

[0018] Figure 5 It is the flowchart of the steps for the fault prediction method of the lithium battery of the present invention.

[0019] Figure 6 It is the flowchart of the steps for predicting based on current excitation using the battery electrochemical principle equation to form predicted values of voltage U and temperature T of the present invention.

[0020] Figure 7 It is the flowchart of the steps for performing sensitivity analysis of parameters of the battery aging mechanism parameters of the battery based on the multi-stage pulse current charging curve of the present invention.

[0021] Figure 8 It is the flowchart of the steps for comparing the predicted values of the model with the actual detection values of the battery sensor of the present invention.

[0022] Figure 9 It is the schematic diagram of the fault prediction system of the lithium battery of the present invention.

[0023] 1 - System Login Module, 2 - Yard Overview Module, 3 - Operation Monitoring Module, 301 - Equipment Monitoring Unit, 302 - Equipment Matrix Unit, 4 - Alarm Setting Module, 5 - Alarm Query Module, 6 - Model Management Module, 7 - Intelligent Operation and Maintenance Module, 8 - Analysis Report Module. Detailed Implementation Manner

[0024] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention and should not be construed as a limitation to the present invention.

[0025] Please refer to Figures 1 to 8 , the present invention provides a method for fault prediction of a lithium battery, specifically including: S101: Apply an input current excitation during the online charging of the lithium battery; By applying a current excitation to the lithium battery and based on the battery data collected in real time by the BMS (Battery Management System), combined with the battery electrochemistry principle equation, a parameter sensitivity analysis can be performed on the crucial aging parameters that affect the battery performance; Since the aging parameters are the fundamental characteristics of the battery capacity and health state, therefore, information such as the battery capacity and health state (SOH) can be directly obtained through these electrochemistry aging parameters.

[0026] S102: Use the battery electrochemistry principle equation to perform prediction based on the current excitation to form predicted values of voltage U and temperature T; Specifically including: S1021: Input a current with a length of 1000 s and a maximum charging current of 1.6 C; S1022: Based on the current, charge using a multi-stage pulsed current charging curve to obtain the voltage U and temperature T responses; Use the Figure 2 attached multi-stage pulsed current charging curve to charge and obtain the Figure 3 attached voltage response; S1023: Perform a sensitivity analysis of the parameters of the battery aging mechanism parameters of the battery based on the multi-stage pulsed current charging curve. The battery aging mechanism parameters include the maximum lithium ion concentration of the battery negative electrode, the maximum lithium ion concentration of the battery positive electrode, the percentage of active material of the battery negative electrode, and the percentage of active material of the battery positive electrode.

[0027] Associate the selected parameters with the voltage response of the battery and evaluate the sensitivity of each parameter under the given working conditions.

[0028] Specifically including: S10231: Based on the battery electrochemical principle equation, calculate various battery aging mechanism parameters to obtain the maximum sensitivity of the maximum lithium ion concentration at the negative electrode of the battery; As shown in the appendix Figure 4 , use the known voltage response data and sensitivity analysis results to directly calculate the maximum lithium ion concentration at the negative electrode of the battery.

[0029] S10232: Based on the calculated battery aging mechanism parameters and capacity estimation formula, obtain the predicted value of the maximum available capacity of the battery; After calculating the maximum lithium ion concentration, based on the battery electrochemical principle equation, combined with known conditions such as voltage, current, and temperature, calculate other sensitive parameters in sequence. The parameter calculation objective function is as follows, that is, by finding a set of optimal parameter solutions to minimize the error between the measured terminal voltage and the model terminal voltage: where, θ * — Optimal parameter set, obtained by minimizing the experimental voltage and the simulated voltage The goal is to find a set of parameters to minimize the sum of the squares of the errors between the experimental data and the model prediction values; — Maximum lithium ion concentration of the negative electrode material; — Effective volume fraction of the negative electrode, representing the proportion of the active material in the negative electrode; — Maximum concentration of the positive electrode material; — Effective volume fraction of the positive electrode, representing the proportion of the active material in the positive electrode; The capacity calculation formula is as follows: Formula for calculating the negative electrode capacity: where, Q neg — Is the capacity of the negative electrode; A— Plate area; F— Faraday constant; L neg — Negative electrode plate thickness; — Electrochemical equivalent change value of the negative electrode; Formula for calculating the positive electrode capacity: where, Q pos — Is the capacity of the positive electrode; A— Plate area; F— Faraday constant; L pos — Thickness of the positive electrode plate; — Change value of the electrochemical equivalent of the positive electrode; The maximum actually available capacity of the battery is the smaller value of the two: ; Combined with four electrochemistry calculation parameters ( ), the available capacity of the battery can be calculated in real time.

[0030] S103: Compare the model prediction value with the actual detection value of the battery sensor. If the error after comparison exceeds the normal range, perform boundary judgment on the internal SOC and internal resistance characteristic parameters of the battery calculated by the battery electrochemistry principle equation; Specifically include: S1031: Actually detect the voltage U and temperature T parameters of the battery through the sensor; Detect and record the voltage U and temperature T parameters of the battery through the sensor for subsequent comparison with the predicted parameters.

[0031] S1032: Compare the actual detection value with the predicted voltage U and temperature T to obtain an error value; Ensure that the actual detection value and the predicted value are time-aligned, that is, they are measured at the same time point or within the same time period. If the timestamps do not match, interpolation or time synchronization processing may be required, and the change trend of the error value over time can also be observed to identify any potential abnormal patterns or trends.

[0032] S1033: Set the normal range of the error based on the internal short circuit model, and input the error value into the internal short circuit model; Compare the error value with the normal range set in the internal short circuit model to determine whether it exceeds the range; the internal short circuit model is an ECM-T-ISC coupling model.

[0033] S1034: When the error value exceeds the normal range determined by the internal short circuit model, preliminarily judge that there is an internal short circuit; Determine the normal range or threshold of the error value. If the error exceeds this range, it may indicate that the battery is abnormal, so that subsequent staff can handle it and perform timely maintenance, thus achieving the effect of fault prediction for the battery and improving the safety of the energy storage device.

[0034] S1035: Calculate the predicted values of the characteristic parameters of the single-cell battery SOC and internal resistance through the battery electrochemistry principle equation, and record the changes of the characteristic parameter predicted values.

[0035] The State of Charge (SOC) is an important parameter describing the remaining battery charge. Its calculation is usually based on the battery charging and discharging current, time, and the initial SOC value of the battery. The battery internal resistance is an important parameter describing the battery's resistance to current, and its magnitude is related to factors such as the battery state, temperature, and charging and discharging history. Thus, by recording the changes in the predicted values of these two characteristic parameters, it is possible to judge battery faults at a later stage.

[0036] S104: If the characteristic parameter meets the boundary judgment condition, it is determined as an internal short - circuit fault.

[0037] Specifically, it includes: S1041: Set the characteristic parameter boundary based on the internal short - circuit model, and input the change in the predicted value of the characteristic parameter into the internal short - circuit model for judgment; Set appropriate characteristic parameter boundaries within the internal short - circuit model, and then make a judgment with the change in the predicted value of the characteristic parameter.

[0038] S1042: When the change in the predicted value of the characteristic parameter meets the characteristic parameter boundary, it is determined as an internal short - circuit fault; Compare the predicted value of SOC with the preset SOC boundaries (such as the lowest SOC and the highest SOC). If the predicted value is lower than the lowest SOC boundary, it may mean that the battery charge is too low and needs to be charged. If the predicted value is higher than the highest SOC boundary, it may mean that the battery is overcharged and charging needs to be stopped and measures should be taken to prevent battery damage; After the first judgment of the error value, if the battery may be abnormal, then a further analysis of the second characteristic parameter boundary is carried out, thereby avoiding the second - level characteristic parameter boundary analysis when the battery is not abnormal, greatly improving the efficiency, and at the same time improving the accuracy of fault judgment.

[0039] S1043: If the change in the predicted value of the characteristic parameter does not meet the characteristic parameter boundary, it is determined that other faults have occurred.

[0040] Please refer to Figure 9 , the present invention also provides a fault prediction system for a lithium - ion battery. The system login module 1, the station overview module 2, the operation monitoring module 3, the alarm setting module 4, the alarm query module 5, the model management module 6, the intelligent operation and maintenance module 7, and the analysis report module 8 are connected in sequence; The system login module 1 is used for staff to log in to the energy storage battery safety diagnosis system; By opening a browser, entering the address of the system, entering the system login interface, and then logging in to the system to manage battery safety.

[0041] The station overview module 2 is used to display the key indicators of the energy storage station; After logging in to the system, it will default to enter the above-mentioned substation overview module 2, which displays the key indicators of the energy storage substation, such as the alarm overview, alarm list, alarm ranking, and statistics of the charge and discharge data information of the energy storage substation; Key indicators of the energy storage substation: display the total power and capacity of the current energy storage substation, the daily charge and discharge volume, the cumulative charge and discharge volume, and the SOC of the substation respectively; Alarm overview: It is possible to query the alarm statistics of each level in the substation within a specified date range by selecting three time dimensions: year, month, and day, and display them in a pie chart; Alarm list: Display the latest 6 alarm record information in the current substation in sequence, showing the device name, alarm content, alarm level, and alarm trigger time of the alarm. There is a "More" button, and clicking this button can view all the alarm list information; Alarm ranking: It is possible to query the top 6 devices with the most alarms and the top 6 alarm contents in the substation within a specified date range by selecting three time dimensions: year, month, and day, and it is also possible to switch and view by clicking the tab page; Energy storage charge and discharge: It is possible to query the statistical information of the charging and discharging volume and the power change curve of the substation within a specified date range by selecting three time dimensions: year, month, and day, and it is also possible to switch and view by clicking the tab page.

[0042] The above-mentioned operation monitoring module 3 is used to display the information of the battery device; Specifically including: The device monitoring unit 301 is used to display the device information in the energy storage substation in a device tree structure; Display the device information in the energy storage substation in a device tree structure. It can be clicked to expand and collapse, and a certain device can also be selected to view the detailed information of the device. The selection status of the device number can be viewed on the left side of the operation interface, and the operation information and statistical chart of the selected device are displayed on the right side; The interface is set with a battery compartment: when the device at the battery compartment level is selected, the real-time operation data of the battery compartment is displayed on the right side, including SOC, SOH, the highest monomer temperature, the maximum temperature difference, the theoretical SOH, the current battery efficiency, the highest monomer voltage, the maximum voltage difference, the daily charging volume, the daily discharging volume, the cumulative charging volume, the cumulative discharging volume, the cumulative cycle times, and the remaining cycle times; Multiple tab pages are respectively displayed below for capacity prediction, revenue prediction, battery efficiency, and alarm records; Display Capacity Prediction: Display the actual values of the actual SOH and theoretical SOH of the current energy storage device unit in two line charts, and use a dotted line to display the predicted value trends of the actual SOH and theoretical SOH at future time points according to the prediction model; Display Energy Storage Temperature Difference: Display the line chart of the energy storage temperature difference of the current energy storage device unit in a line chart. The analysis period supports up to 7 days, with an analysis dimension of 0 - 3 days and a data frequency of 30 seconds, and a data frequency of 1 minute for the 4 - 7 day dimension; Display Alarm Records: Display the alarm records of the current energy storage device unit within the selected time period in a list form, and support customizing the time range for filtering; Display Battery Stack: When a device at the battery stack level is selected, display the real-time operating data of the battery stack on the right, including total voltage, total current, maximum charging current, maximum discharge current, SOC, SOH, theoretical SOH, SCR, daily charging amount, daily discharge amount, cumulative charging amount, cumulative discharge amount; Below, display battery capacity, trend data, and power statistics in multiple tabs respectively; Display Battery Capacity: Display the actual values of the actual SOH and theoretical SOH of the current energy storage device unit in two line charts, and use a dotted line to display the predicted value trends of the actual SOH and theoretical SOH at future time points according to the prediction model; Display Trend Data: According to the selected data indicators and two dates, display the change trend chart of the indicator values of the current energy storage device unit in a curve chart. The data indicators that can be selected at the battery stack level include total voltage, total current, SOC, SOH, and support synchronous comparison of multiple indicators, with a maximum of 4 supported; Display Power Statistics: Display the charge and discharge amount statistics information of the current energy storage device unit within the selected date range in a bar chart; Display the alarm information of the currently selected energy storage device unit in a list form, and support customizing the time range for filtering; It can also display the battery cluster separately: When a device at the battery cluster level is selected, display the real-time operating data of the battery cluster on the right, including total voltage, highest single-cell voltage, lowest single-cell voltage, total current, maximum charging current, maximum discharge current, highest temperature, lowest temperature, SOC, highest single-cell SOC, highest single-cell SOC number, lowest single-cell SOC, lowest single-cell SOC number, SOH, theoretical SOH, SCR, daily charging amount, daily discharge amount, cumulative charging amount, cumulative discharge amount; Below, display trend data, power statistics, temperature monitoring, and alarm information in multiple tabs respectively; Display Trend Data: According to the selected data indicators and two dates, display the change trend chart of the indicator values of the current energy storage device unit in a curve chart. The data indicators that can be selected at the battery cluster level include total voltage, total current, SOC, SOH, highest temperature, lowest temperature, highest single-cell voltage, lowest single-cell voltage, and support synchronous comparison of multiple indicators, with a maximum of 4 supported; Display power statistics: Show the charge and discharge volume statistics of the current energy storage device unit within the selected date range in the form of a bar chart; Display temperature monitoring: List in detail the real-time temperature data of each module under the current battery cluster in tabular form; Display battery cells: When the device at the battery cell level is selected, display the real-time operation data of the battery cell on the right; The device matrix unit 302 is used to visually display the operation status of each level of device units in the current energy storage station in a card-by-layer manner.

[0043] Visually display the operation status of each level of device units in the current energy storage station in a card-by-layer manner; First, from the perspective of the battery compartment level, display the operation data of each battery stack in the current energy storage station; Display the alarm status of the device unit in different icon colors. Green indicates that the device and all subordinate devices have no real-time alarms or only low-risk real-time alarms. Orange indicates that the device and all subordinate devices have at most medium-risk real-time alarms. Red indicates that the device and subordinate devices have high-risk real-time alarms; Set a switch button "Only view alarm status" in the upper right corner of the system. After clicking, switch between displaying all devices and only displaying devices in the alarm status; The content displayed on the battery stack level card includes: total voltage, total current, SOC, SOH, highest SCR, lowest SCR, highest theoretical SOH, lowest theoretical SOH; Set the "battery stack card" interface to enter the perspective of this battery stack level; At the same time, by default, display the operation status of all battery cluster device units under the battery stack of the selected battery compartment; The content displayed on the battery cluster level card includes: total voltage, total current, SOC, SOH, highest SCR, lowest SCR, highest theoretical SOH, lowest theoretical SOH, highest single cell temperature, lowest single cell temperature; Click on the "temperature heat map" set on the system interface to view the real-time temperature 3D model diagram of each battery cluster and single battery cell under the battery stack device; Set the "battery cluster card" to display the operation information of the devices under the battery cluster, including: number of battery cells, highest SCR, lowest SCR, highest theoretical SOH, lowest theoretical SOH, highest battery cell SOC, lowest battery cell SOC, highest measurement point temperature, lowest measurement point temperature, highest single cell voltage, lowest single cell voltage; At the same time, display the operation data of each battery cell under the battery cluster in the form of a card below, and the display content includes SOC, SOH, SCR, theoretical SOH, voltage U, temperature T.

[0044] The alarm setting module 4 is used to set the alarm rules and alarm thresholds of the energy storage station; Manual configuration operations can be performed on the alarm rules and alarm thresholds of the current station. The list shows the maintained alarm rule information separately for collection and calculation type metrics, and filtering operations can be performed through query conditions; When setting the alarm rules, fill in the alarm name according to the actual business, select the corresponding device type, which is divided into battery compartments, battery stacks, battery clusters, and battery cells, and then select the data items to be monitored under the device type, such as battery stack voltage, battery stack current, battery stack SOC, battery stack SOH, battery cluster voltage, battery cluster current, etc.; The alarm query module 5 is used to query alarm information; It is used to query different classification and different level alarm information generated since the operation of the current energy storage station system, as well as the alarm push records for docking real-time data; The list shows the generated alarm record information, and filtering operations can be performed through query conditions; the current alarm list data can also be exported in a specified format file; at the same time, in the trigger rule column, you can click to view the alarm rule basis of the selected alarm record and view the alarm rule pop-up window; In addition, the data point values of the device 3 minutes before and after the alarm trigger moment can be downloaded for analysis and viewing.

[0045] The model management module 6 is used to configure the energy storage algorithm model; It is used to configure the energy storage algorithm model to facilitate the maintenance and parameter configuration of the corresponding algorithm program in the system background; the maintained algorithm model information is displayed in the system list, and filtering operations can be performed through query conditions, and algorithm model configurations can also be added manually; for the maintained algorithm models, editing and modification operations can be performed, and invalid algorithm models can also be selected and deleted.

[0046] The intelligent operation and maintenance module 7 is used to give analysis and suggestions based on the detected anomalies; Set up an "expert knowledge base" to provide a fault diagnosis function based on the expert knowledge base. When the system detects an anomaly, it can give an analysis of the cause of the fault and suggested handling measures; the maintained expert knowledge base information is displayed, and filtering operations can be performed through query conditions; possible causes and handling suggestions can be maintained according to the alarm type, alarm level, etc.; in addition, the maintained expert knowledge can be edited and modified; corresponding information can also be selected and deleted.

[0047] The analysis report module 8 is used to analyze the battery cluster and generate reports.

[0048] Automatically generate reports for the operation data of the energy storage station according to the selected date range for business analysis; the statistical dimensions (battery stack, battery cluster) can be selected, then the statistical date range can be selected, and finally the "Query" button on the motor interface can be used to generate the operation data of the corresponding energy storage device unit. Currently, only the statistical analysis of surface data is supported, including the cumulative charge, cumulative discharge, maximum SCR value, minimum theoretical SOH, total number of alarms, high-risk number, medium-risk number, and low-risk number. It can also perform temperature consistency analysis and voltage consistency analysis on the battery clusters in the energy storage station within a specified time period; the analysis content (voltage U, temperature T) and merging method (energy storage unit merging, battery cluster merging) can be selected, and then the analysis time range can be selected to generate the operation data of the corresponding energy storage device. It can also provide parameter records for each charge / discharge stage of each battery compartment in the energy storage station on a single day, including the voltage range difference and temperature range difference of the charging monomers and discharging monomers of each battery cluster, and determine whether it exceeds the limit according to the threshold configuration. According to the minimum / maximum value range of the charge / discharge type, voltage range difference, and temperature range difference, and then selecting the analysis date range, the operation data of the corresponding energy storage device can be generated. The detailed data information of the current round of charge / discharge of the device on the same day can be viewed, and the current query analysis report can also be exported as a file in a specified format.

[0049] The above disclosure is only one or more preferred embodiments of the present application, and cannot be used to limit the scope of rights of the present application. Those of ordinary skill in the art can understand all or part of the processes of implementing the above embodiments, and the equivalent changes made according to the claims of the present application still fall within the scope covered by the present application.

Claims

1. A method for predicting a lithium battery failure, characterized in that: The steps include: Input current excitation when lithium battery is charged online; Use the battery electrochemical principle equation to make predictions based on current excitation to form AI model prediction values ​​of voltage U and temperature T; Compare the AI ​​model prediction value with the actual detection value of the battery sensor. If the error after comparison exceeds the normal range, perform boundary judgment on the battery internal SOC and internal resistance characteristic parameters calculated by the battery electrochemical principle equation; If the characteristic parameters meet the boundary judgment conditions, it is determined to be an internal short circuit fault.

2. The lithium battery failure prediction method according to claim 1, characterized in that: In the step of using the battery electrochemical principle equation to make predictions based on current excitation to form AI model prediction values ​​of voltage U and temperature T: Input current length of 1000 s, maximum charging current of 1.6 C; Based on the current, a multi-stage pulse current charging curve is used for charging to obtain voltage U and temperature T response values; The battery aging mechanism parameters of the battery are subjected to parameter sensitivity analysis based on the multi-stage pulse current charging curve, and the battery aging mechanism parameters include the maximum lithium ion concentration of the battery negative electrode, the maximum lithium ion concentration of the battery positive electrode, the percentage of active materials in the battery negative electrode and the percentage of active materials in the battery positive electrode.

3. The lithium battery failure prediction method according to claim 2, characterized in that: In the step of performing sensitivity analysis on the battery aging mechanism parameters of the battery based on the multi-stage pulse current charging curve: Based on the battery electrochemical principle equation, a variety of battery aging mechanism parameters are calculated to obtain the maximum lithium ion concentration of the battery negative electrode with the maximum sensitivity; Based on the calculated battery aging mechanism parameters and capacity estimation formula, a predicted value of the actual maximum available capacity of the battery is obtained.

4. The lithium battery failure prediction method according to claim 3, characterized in that: When comparing the model prediction value with the actual detection value of the battery sensor, if the error after comparison exceeds the normal range, the battery internal SOC and internal resistance characteristic parameters calculated by the battery electrochemical principle equation are subjected to boundary judgment: The voltage U and temperature T parameters of the battery are actually detected by the sensor; Compare the actual detected value with the predicted voltage temperature to obtain the error value; Setting a normal range of error based on an internal short-circuit model, and inputting the error value into the internal short-circuit model; When the error value exceeds the normal range determined by the internal short circuit model, it is preliminarily determined that an internal short circuit exists; The characteristic parameter prediction values ​​of the SOC and internal resistance of the single cell are calculated by the battery electrochemical principle equation, and the changes of the characteristic parameter prediction values ​​are recorded.

5. The lithium battery failure prediction method according to claim 4, characterized in that: In the step of determining that an internal short circuit fault occurs if the characteristic parameter meets the boundary judgment condition: Setting a characteristic parameter boundary based on the internal short-circuit model, and inputting a change in the characteristic parameter prediction value into the internal short-circuit model for judgment; When the change of the predicted value of the characteristic parameter meets the characteristic parameter boundary, it is determined to be an internal short circuit fault; If the change in the predicted value of the characteristic parameter does not satisfy the characteristic parameter boundary, it is determined that other faults have occurred.

6. A lithium battery fault prediction system, using the lithium battery fault prediction method as claimed in claim 5, characterized in that: It includes a system login module, a station overview module, an operation monitoring module, an alarm setting module, an alarm query module, a model management module, an intelligent operation and maintenance module, and an analysis report module, wherein the system login module, the station overview module, the operation monitoring module, the alarm setting module, the alarm query module, the model management module, the intelligent operation and maintenance module, and the analysis report module are connected in sequence; The system login module is used by staff to log in to the energy storage battery safety diagnosis system; The station overview module is used to display key indicators of the energy storage station; The operation monitoring module is used to display the information of the battery equipment; The alarm setting module is used to set the alarm rules and alarm thresholds of the energy storage station; The alarm query module is used to query the alarm information; The model management module is used to configure the energy storage algorithm model; The intelligent operation and maintenance module is used to provide analysis and suggestions based on detected anomalies; The analysis report module is used to analyze the battery cluster and generate reports.

7. The lithium battery failure prediction system according to claim 6, characterized in that: The operation monitoring module includes a device monitoring unit and a device matrix unit, and the device monitoring unit is connected to the device matrix unit; The equipment monitoring unit is used to display the equipment information in the energy storage station in an equipment tree structure; The equipment matrix unit is used to display the operation status of each level of equipment units in the current energy storage station layer by layer in a visual card format.

Images