Battery pack state monitoring system and method based on artificial intelligence

Abstract

The invention discloses a battery pack state monitoring system and method based on artificial intelligence. The system comprises a data acquisition module, a data processing module and a state early warning module. Relates to the technical field of battery pack monitoring, and solves the technical problem of low efficiency and accuracy of state monitoring of a battery pack caused by difficulty in dynamically adjusting monitoring frequencies of different areas in the battery pack in the prior art. The method comprises the following steps: performing region division on a battery pack based on historical abnormal alarm information to obtain a plurality of monitoring regions; acquiring operation parameters of each monitoring area in a plurality of continuous periods; performing data analysis on the operation parameters to obtain operation characteristic data; calculating a safety risk coefficient of each monitoring area based on the operation characteristic data; generating an early warning signal based on the safety risk coefficient predicted value and a preset risk threshold range; the specific monitoring area is monitored in a targeted manner according to the early warning signal, so that the efficiency of monitoring the state of the battery pack is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery pack monitoring, and specifically relates to a battery pack status monitoring system and method based on artificial intelligence. Background Art

[0002] With the increasing popularity of electric vehicles, battery packs play a crucial role as energy storage devices. However, battery packs are prone to various abnormalities such as overheating, abnormal pressure, and changes in contact resistance over long periods of use and in harsh environments. These abnormalities can negatively impact the performance, lifespan, and safety of the battery pack.

[0003] Existing technologies achieve a certain degree of battery pack status monitoring by installing several sensors in the battery pack and regularly acquiring key data collected by the sensors. However, in practice, due to varying heat dissipation conditions and external impact angles in different areas of the battery pack, the probability of abnormalities occurring in each area varies. Existing solutions struggle to dynamically adjust the monitoring frequency for different areas of the battery pack, resulting in low efficiency and accuracy in battery pack status monitoring.

[0004] The present invention provides a battery pack status monitoring system and method based on artificial intelligence to solve the above technical problems. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a battery pack status monitoring system and method based on artificial intelligence, which is used to solve the technical problem that it is difficult to dynamically adjust the monitoring frequency of different areas in the battery pack, resulting in low efficiency and accuracy of battery pack status monitoring.

[0006] To achieve the above objectives, a first aspect of the present invention provides an artificial intelligence-based battery pack status monitoring system, comprising: a data processing module, and a data acquisition module and a status warning module connected thereto;

[0007] The data acquisition module is used to obtain historical abnormal alarm information of the battery pack, divide the battery pack into regions based on the historical abnormal alarm information, and obtain multiple monitoring regions; collect operating parameters of each monitoring region within several consecutive cycles;

[0008] The data processing module is used to perform data analysis on the operating parameters to obtain operating characteristic data; calculate the safety risk coefficient of each monitoring area based on the operating characteristic data; input the safety risk coefficient of each monitoring area into the safety risk prediction model to obtain a safety risk coefficient prediction value; wherein the safety risk prediction model is obtained by training an artificial intelligence model;

[0009] The status warning module is used to generate a warning signal based on the safety risk coefficient prediction value and the preset risk threshold range; and perform targeted monitoring of the battery pack based on the warning signal.

[0010] Preferably, the dividing the battery pack into regions based on historical abnormal alarm information includes:

[0011] Extract historical abnormal alarm information of the battery pack; wherein the historical abnormal alarm information includes abnormal alarm levels of multiple cells in the battery pack and the number of abnormal occurrences of each abnormal alarm level; abnormal alarm levels include level one abnormality and level two abnormality, and level one abnormality is greater than level two abnormality; the number of abnormal occurrences includes the number of level one abnormality and the number of level two abnormality;

[0012] The battery pack is divided into multiple sub-areas based on the distribution information of the battery cells inside the battery pack, and the sub-areas are numbered as i. Wherein, a single sub-area has only one battery cell; i = 1, 2, ..., n, where n is the total number of battery cells inside the battery pack;

[0013] Calculate the stability score of each battery cell based on the abnormal alarm level and the corresponding number of abnormal occurrences;

[0014] Multiple sub-areas with stability scores less than the preset lower limit of the scoring threshold are marked as monitoring area 1; multiple sub-areas with stability scores within the preset scoring threshold range are marked as monitoring area 2; and multiple sub-areas with stability scores greater than the preset upper limit of the scoring threshold are marked as monitoring area 3.

[0015] Preferably, the calculation of the stability score of each battery cell based on the abnormality alarm level and the corresponding number of abnormality occurrences includes:

[0016] By formula Calculate the stability score WDPi of sub-region i; where, is the number of first-level abnormalities of the i-th battery cell, is the number of secondary abnormalities of the i-th battery cell; a is the weight coefficient corresponding to the primary abnormality, and 0 <a<1。

[0017] Preferably, the performing data analysis on the operating parameters to obtain the operating characteristic data includes:

[0018] Extract the operating parameters of each monitoring area over several consecutive cycles; the operating parameters include average internal resistance, pressure data, and battery cell temperature;

[0019] By calculating the ratio of average voltage to average current, the average internal resistance of the corresponding monitoring area is obtained;

[0020] Performing linear fitting on the operating parameters of several consecutive cycles to obtain operating parameter curves; wherein the operating parameter curves include an average internal resistance curve, a pressure data curve, and a battery cell temperature curve;

[0021] Calculate the first-order derivative function of each operating parameter curve to obtain the operating parameter change derivative function; extract the maximum function value and the minimum function value in the operating parameter change derivative function, calculate the average value of the maximum function value and the minimum function value and mark it as the operating characteristic data; wherein the operating characteristic data includes the average internal resistance change rate, the pressure data change rate and the battery cell temperature change rate.

[0022] Preferably, performing linear fitting on the operating parameters of a plurality of consecutive cycles respectively includes:

[0023] The average internal resistance, pressure data, and battery cell temperature of each monitoring area in several consecutive cycles are extracted; with time as the independent variable, and the average internal resistance, pressure data, and battery cell temperature as independent variables, the average internal resistance curve, pressure data curve, and battery cell temperature curve are obtained through linear fitting.

[0024] Preferably, the calculation of the safety risk coefficient of each monitoring area based on the operation characteristic data includes:

[0025] Extract the operating characteristic data of each monitoring area; through the formula Calculate the safety risk factor AFXj for monitoring area i; where NZBj is the average internal resistance change rate of monitoring area j, YLBj is the pressure data change rate of monitoring area j, and WDBj is the cell temperature change rate of monitoring area j; b1, b2, and b3 are all proportional coefficients greater than 0; j = 1, 2, or 3; e is a natural constant, and ln() is a logarithmic function with the natural constant as the base.

[0026] It should be noted that the values ​​of the proportional coefficients b1, b2, and b3 are related to the usage time of the battery pack. When ... is larger, the actual value setting of the corresponding proportional coefficients b1, b2, and b3 is larger.

[0027] Preferably, the security risk prediction model is obtained through artificial intelligence model training, including:

[0028] Extract historical data on the computing volume of several computing devices in each monitoring area and integrate them into several groups of training data and test data; use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a security risk prediction model whose input is the computing volume of several recent consecutive cycles and whose output is the computing volume of the prediction cycle; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model.

[0029] Preferably, the generating of the warning signal based on the safety risk coefficient prediction value and the preset risk threshold range includes:

[0030] T1: Extract the predicted value of safety risk factor of each monitoring area;

[0031] T2: Determine whether the predicted value of the safety risk factor is within the preset risk threshold; if yes, set the warning signal of the corresponding monitoring area to warning signal 1; if not, jump to T3:

[0032] T3: Determine whether the predicted value of the safety risk factor is greater than the preset risk threshold upper limit; if yes, generate an early warning signal for the corresponding monitoring area and set it as early warning signal 2; if no, continue to monitor the status of the battery pack; among them, the early warning signal includes early warning signal 1 and early warning signal 2.

[0033] Preferably, the targeted monitoring of the battery pack based on the early warning signal includes:

[0034] Extract early warning signals from each monitoring area;

[0035] Determine whether the warning signal of monitoring area i is warning signal 1; if yes, set the data collection frequency of monitoring area i to 2F; if not, set the data collection frequency of monitoring area i to 1.5F; where F is the default value of the data collection frequency.

[0036] A second aspect of the present invention provides a battery pack status monitoring method based on artificial intelligence, comprising:

[0037] Obtain historical abnormal alarm information of the battery pack, divide the battery pack into regions based on the historical abnormal alarm information, and obtain multiple monitoring areas; collect operating parameters of each monitoring area over several consecutive cycles;

[0038] Performing data analysis on the operating parameters to obtain operating characteristic data;

[0039] Calculate the safety risk factor of each monitoring area based on the operating characteristic data;

[0040] Input the safety risk coefficient of each monitoring area into the safety risk prediction model to obtain the safety risk coefficient prediction value;

[0041] Generate early warning signals based on the predicted value of the safety risk factor and the preset risk threshold range;

[0042] Conduct targeted monitoring of battery packs based on early warning signals.

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

[0044] The present invention first uses historical abnormal data to accurately divide the battery pack into different areas to form multiple monitoring areas, so as to implement refined management for different risk points. Subsequently, by collecting the operating parameters of each monitoring area within a continuous period and conducting in-depth data analysis to extract characteristic data, the actual operating status and potential risk trends of each area can be accurately reflected. The safety risk coefficient is calculated based on the obtained operating characteristic data, and the risk level of each area is quantified to provide a basis for subsequent early warning. These risk coefficients are input into a pre-trained safety risk prediction model to obtain possible future risk coefficient prediction values, so that high-risk areas can be identified in advance and preventive measures can be taken. According to the prediction results and the preset risk threshold range, the corresponding early warning signal is generated, realizing the transition from passive response to active prevention. Finally, targeted monitoring is performed on specific monitoring areas based on the early warning signal, which is conducive to improving the efficiency of battery pack status monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 This is an overall flow chart of the battery pack status monitoring method based on artificial intelligence of the present invention;

[0047] Figure 2 Schematic diagram of the principle of the artificial intelligence-based battery pack status monitoring system of the present invention. DETAILED DESCRIPTION

[0048] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0049] See also Figure 1-Figure 2 , a first embodiment of the present invention provides a battery pack status monitoring system based on artificial intelligence, comprising: a data processing module, and a data acquisition module and a status warning module connected thereto;

[0050] Data acquisition module: used to obtain historical abnormal alarm information of the battery pack, divide the battery pack into regions based on the historical abnormal alarm information, and obtain multiple monitoring areas; collect the operating parameters of each monitoring area within several consecutive cycles;

[0051] Data processing module: used to analyze operating parameters to obtain operating characteristic data; calculate the safety risk coefficient of each monitoring area based on the operating characteristic data; input the safety risk coefficient of each monitoring area into the safety risk prediction model to obtain the safety risk coefficient prediction value; wherein, the safety risk prediction model is obtained through artificial intelligence model training;

[0052] Status warning module: used to generate warning signals based on the predicted value of the safety risk factor and the preset risk threshold range; and conduct targeted monitoring of the battery pack based on the warning signals.

[0053] In this embodiment, the battery pack is divided into regions based on historical abnormal alarm information, including:

[0054] Extract historical abnormal alarm information of the battery pack; wherein the historical abnormal alarm information includes abnormal alarm levels of multiple cells in the battery pack and the number of abnormal occurrences of each abnormal alarm level; abnormal alarm levels include level one abnormality and level two abnormality, and level one abnormality is greater than level two abnormality; the number of abnormal occurrences includes the number of level one abnormality and the number of level two abnormality;

[0055] The battery pack is divided into multiple sub-areas based on the distribution information of the battery cells inside the battery pack, and the sub-areas are numbered as i. Wherein, a single sub-area has only one battery cell; i = 1, 2, ..., n, where n is the total number of battery cells inside the battery pack;

[0056] Calculate the stability score of each battery cell based on the abnormal alarm level and the corresponding number of abnormal occurrences;

[0057] Multiple sub-areas with stability scores less than the preset lower limit of the scoring threshold are marked as monitoring area 1; multiple sub-areas with stability scores within the preset scoring threshold range are marked as monitoring area 2; and multiple sub-areas with stability scores greater than the preset upper limit of the scoring threshold are marked as monitoring area 3.

[0058] The present invention extracts the historical abnormal alarm information of each battery cell in the battery pack, including the first-level abnormality and the second-level abnormality and the number of occurrences, and divides the battery pack into single-cell sub-areas according to the distribution of the battery cells. Based on this, the stability score of each battery cell is calculated, and then multiple monitoring areas are divided to achieve refined management. First, a stability scoring system is established by quantifying the abnormality level and frequency to accurately identify the risk status of each battery cell. Secondly, the regional division is carried out based on a single battery cell, which avoids the problem of inaccurate fault location caused by larger monitoring units in traditional methods and improves the accuracy and efficiency of fault detection. Then, the battery pack is divided into regions according to the preset scoring threshold, which facilitates the subsequent targeted adjustment of the data acquisition frequency of each monitoring area, thereby helping to improve the efficiency and accuracy of battery pack status monitoring.

[0059] In this embodiment, the stability score of each battery cell is calculated based on the abnormal alarm level and the corresponding number of abnormal occurrences, including:

[0060] By formula Calculate the stability score WDPi of sub-region i; where, is the number of first-level abnormalities of the i-th battery cell, is the number of secondary abnormalities of the i-th battery cell; a is the weight coefficient corresponding to the primary abnormality, and 0 <a<1。

[0061] For example, the weight coefficient a corresponding to the first-level abnormality is set to 0.6; the number of first-level abnormalities of the first cell is Number of secondary abnormalities of the first battery cell The stability score of the first battery cell is calculated by the formula WDP1=2.2.

[0062] In this embodiment, the operation characteristic data is obtained by performing data analysis on the operation parameters, including:

[0063] Extract the operating parameters of each monitoring area over several consecutive cycles; the operating parameters include average internal resistance, pressure data, and battery cell temperature;

[0064] By calculating the ratio of average voltage to average current, the average internal resistance of the corresponding monitoring area is obtained;

[0065] Performing linear fitting on the operating parameters of several consecutive cycles to obtain operating parameter curves; wherein the operating parameter curves include an average internal resistance curve, a pressure data curve, and a battery cell temperature curve;

[0066] Calculate the first-order derivative function of each operating parameter curve to obtain the operating parameter change derivative function; extract the maximum function value and the minimum function value in the operating parameter change derivative function, calculate the average value of the maximum function value and the minimum function value and mark it as the operating characteristic data; wherein the operating characteristic data includes the average internal resistance change rate, the pressure data change rate and the battery cell temperature change rate.

[0067] In this embodiment, linear fitting is performed on the operating parameters of several consecutive cycles, including:

[0068] The average internal resistance, pressure data, and battery cell temperature of each monitoring area in several consecutive cycles are extracted; with time as the independent variable, and the average internal resistance, pressure data, and battery cell temperature as independent variables, the average internal resistance curve, pressure data curve, and battery cell temperature curve are obtained through linear fitting.

[0069] In this embodiment, the safety risk coefficient of each monitoring area is calculated based on the operating characteristic data, including:

[0070] Extract the operating characteristic data of each monitoring area; through the formula Calculate the safety risk factor AFXj for monitoring area i; where NZBj is the average internal resistance change rate of monitoring area j, YLBj is the pressure data change rate of monitoring area j, and WDBj is the cell temperature change rate of monitoring area j; b1, b2, and b3 are all proportional coefficients greater than 0; j = 1, 2, or 3; e is a natural constant, and ln() is a logarithmic function with the natural constant as the base.

[0071] For example, the proportional coefficients b1=10, b2=5, and b3=4 are set; the average internal resistance change rate NZB1=0.2mΩ / min in the monitoring area 1, the pressure data change rate YLB1=15kPa / min in the monitoring area 1, and the battery cell temperature change rate WDB1=5℃ / min in the monitoring area 1; the safety risk coefficient AFX1≈32.67 of the monitoring area 1 is calculated by the formula.

[0072] In this embodiment, the security risk prediction model is obtained through artificial intelligence model training, including:

[0073] The safety risk coefficients of each monitoring area in several consecutive periods are extracted and integrated into several groups of original data, 70% of the original data are used as training data and 30% as test data; the artificial intelligence model is trained using the training data; the trained artificial intelligence model is tested using the test data, and the artificial intelligence model is adjusted according to the test results; and finally a safety risk prediction model is obtained, whose input is the safety risk coefficient of several recent consecutive periods and whose output is the predicted value of the safety risk coefficient of the prediction period; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model.

[0074] In this embodiment, the early warning signal is generated based on the predicted value of the safety risk factor and the preset risk threshold range, including:

[0075] T1: Extract the predicted value of safety risk factor of each monitoring area;

[0076] T2: Determine whether the predicted value of the safety risk factor is within the preset risk threshold; if yes, set the warning signal of the corresponding monitoring area to warning signal 1; if not, jump to T3:

[0077] T3: Determine whether the predicted value of the safety risk factor is greater than the preset risk threshold upper limit; if yes, generate an early warning signal for the corresponding monitoring area and set it as early warning signal 2; if no, continue to monitor the status of the battery pack;

[0078] Among them, the early warning signals include early warning signal one and early warning signal two.

[0079] For example, the safety risk coefficient prediction value of monitoring area 1 is set to 35, and the preset risk threshold range is [30, 40]; since the safety risk coefficient prediction value of monitoring area 1 is within the preset risk threshold range, the warning signal of monitoring area 1 is set to warning signal one.

[0080] In this embodiment, targeted monitoring of the battery pack based on the early warning signal includes:

[0081] Extract early warning signals from each monitoring area;

[0082] Determine whether the warning signal of monitoring area i is warning signal 1; if yes, set the data collection frequency of monitoring area i to 2F; if not, set the data collection frequency of monitoring area i to 1.5F; where F is the default value of the data collection frequency.

[0083] This invention extracts warning signals from each monitoring area and responds in a graded manner. When a monitoring area triggers warning signal one, the data collection frequency is increased to 2F; when a monitoring area triggers warning signal two, the data collection frequency is increased to 1.5F. This dynamically adjusts the monitoring frequency of each monitoring area in the battery pack, allowing for faster acquisition of operating parameters in abnormal areas during subsequent battery pack adjustments, thereby improving the efficiency and accuracy of battery pack status monitoring.

[0084] Exemplarily, the early warning signal of monitoring area 1 is set to early warning signal 1, and the data collection frequency of monitoring area 1 is set to twice the default value.

[0085] A second embodiment of the present invention provides a battery pack status monitoring method based on artificial intelligence, comprising:

[0086] S1: Obtain historical abnormal alarm information of the battery pack, divide the battery pack into regions based on the historical abnormal alarm information, and obtain multiple monitoring areas; collect operating parameters of each monitoring area over several consecutive cycles;

[0087] S2: Analyze the operating parameters to obtain operating characteristic data;

[0088] S3: Calculate the safety risk factor of each monitoring area based on the operating characteristic data;

[0089] S4: Input the safety risk coefficient of each monitoring area into the safety risk prediction model to obtain a safety risk coefficient prediction value;

[0090] S5: Generate an early warning signal based on the predicted value of the safety risk factor and the preset risk threshold range;

[0091] S6: Conduct targeted monitoring of the battery pack based on early warning signals.

[0092] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0093] Working principle of the present invention:

[0094] The present invention obtains historical abnormal alarm information of the battery pack, divides the battery pack into regions based on the historical abnormal alarm information, and obtains multiple monitoring areas; collects operating parameters of each monitoring area within several consecutive cycles; performs data analysis on the operating parameters to obtain operating characteristic data; calculates the safety risk coefficient of each monitoring area based on the operating characteristic data; inputs the safety risk coefficient of each monitoring area into a safety risk prediction model to obtain a safety risk coefficient prediction value; generates an early warning signal based on the safety risk coefficient prediction value and a preset risk threshold range; and performs targeted monitoring of the battery pack based on the early warning signal.

[0095] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A battery pack status monitoring system based on artificial intelligence, comprising: The data processing module, and the data acquisition module and status warning module connected thereto are characterized in that: The data acquisition module is used to obtain historical abnormal alarm information of the battery pack, divide the battery pack into regions based on the historical abnormal alarm information, and obtain multiple monitoring regions; collect operating parameters of each monitoring region within several consecutive cycles; The data processing module is used to perform data analysis on the operating parameters to obtain operating characteristic data; calculate the safety risk coefficient of each monitoring area based on the operating characteristic data; input the safety risk coefficient of each monitoring area into the safety risk prediction model to obtain a safety risk coefficient prediction value; wherein the safety risk prediction model is obtained by training an artificial intelligence model; The status warning module is used to generate a warning signal based on the safety risk coefficient prediction value and the preset risk threshold range; and perform targeted monitoring of the battery pack based on the warning signal.

2. The artificial intelligence-based battery pack status monitoring system according to claim 1, characterized in that: The dividing of battery packs into regions based on historical abnormal alarm information includes: Extract historical abnormal alarm information of the battery pack; wherein the historical abnormal alarm information includes abnormal alarm levels of multiple cells in the battery pack and the number of abnormal occurrences of each abnormal alarm level; abnormal alarm levels include level one abnormality and level two abnormality, and level one abnormality is greater than level two abnormality; the number of abnormal occurrences includes the number of level one abnormality and the number of level two abnormality; The battery pack is divided into multiple sub-areas based on the distribution information of the battery cells inside the battery pack, and the sub-areas are numbered as i. Wherein, a single sub-area has only one battery cell; i = 1, 2, ..., n, where n is the total number of battery cells inside the battery pack; Calculate the stability score of each battery cell based on the abnormal alarm level and the corresponding number of abnormal occurrences; Multiple sub-areas with stability scores less than the preset lower limit of the scoring threshold are marked as monitoring area 1; multiple sub-areas with stability scores within the preset scoring threshold range are marked as monitoring area 2; and multiple sub-areas with stability scores greater than the preset upper limit of the scoring threshold are marked as monitoring area 3.

3. The artificial intelligence-based battery pack status monitoring system according to claim 2, characterized in that: The calculation of the stability score of each battery cell based on the abnormal alarm level and the corresponding number of abnormal occurrences includes: By formula Calculate the stability score WDPi of sub-region i; where, is the number of first-level abnormalities of the i-th battery cell, is the number of secondary abnormalities of the i-th battery cell; a is the weight coefficient corresponding to the primary abnormality, and 0 <a<1。 4. The artificial intelligence-based battery pack status monitoring system according to claim 1, characterized in that: The step of performing data analysis on the operating parameters to obtain the operating characteristic data includes: Extract the operating parameters of each monitoring area over several consecutive cycles; the operating parameters include average internal resistance, pressure data, and battery cell temperature; By calculating the ratio of average voltage to average current, the average internal resistance of the corresponding monitoring area is obtained; Performing linear fitting on the average internal resistance, pressure data, and battery cell temperature of several consecutive cycles to obtain an operating parameter curve; wherein the operating parameter curve includes an average internal resistance curve, a pressure data curve, and a battery cell temperature curve; Calculate the first-order derivative function of each operating parameter curve to obtain the operating parameter change derivative function; extract the maximum function value and the minimum function value in the operating parameter change derivative function, calculate the average value of the maximum function value and the minimum function value and mark it as the operating characteristic data; wherein the operating characteristic data includes the average internal resistance change rate, the pressure data change rate and the battery cell temperature change rate.

5. The artificial intelligence-based battery pack status monitoring system according to claim 4, characterized in that: The linear fitting of the operating parameters of a plurality of consecutive cycles respectively includes: The average internal resistance, pressure data, and battery cell temperature of each monitoring area in several consecutive cycles are extracted; with time as the independent variable, and the average internal resistance, pressure data, and battery cell temperature as independent variables, the average internal resistance curve, pressure data curve, and battery cell temperature curve are obtained through linear fitting.

6. The artificial intelligence-based battery pack status monitoring system according to claim 4, characterized in that: The calculation of the safety risk coefficient of each monitoring area based on the operation characteristic data includes: Extract the operating characteristic data of each monitoring area; through the formula Calculate the safety risk factor AFXj for monitoring area i; where NZBj is the average internal resistance change rate of monitoring area j, YLBj is the pressure data change rate of monitoring area j, and WDBj is the cell temperature change rate of monitoring area j; b1, b2, and b3 are all proportional coefficients greater than 0; j = 1, 2, or 3; e is a natural constant, and ln() is a logarithmic function with the natural constant as the base.

7. The artificial intelligence-based battery pack status monitoring system according to claim 1, characterized in that: The security risk prediction model is obtained through artificial intelligence model training, including: The safety risk coefficients of each monitoring area in several consecutive periods are extracted and integrated into several groups of training data and test data; the artificial intelligence model is trained using the training data; the trained artificial intelligence model is tested using the test data, and the artificial intelligence model is adjusted according to the test results; and finally a safety risk prediction model is obtained, whose input is the safety risk coefficient of the most recent several consecutive periods and whose output is the predicted value of the safety risk coefficient of the prediction period; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model.

8. The artificial intelligence-based battery pack status monitoring system according to claim 1, characterized in that: The generating of the warning signal based on the safety risk coefficient prediction value and the preset risk threshold range includes: T1: Extract the predicted value of safety risk factor of each monitoring area; T2: Determine whether the predicted value of the safety risk factor is within the preset risk threshold; if yes, set the warning signal of the corresponding monitoring area to warning signal 1; if not, jump to T3: T3: Determine whether the predicted value of the safety risk factor is greater than the preset risk threshold upper limit; if yes, generate an early warning signal for the corresponding monitoring area and set it as early warning signal 2; if no, continue to monitor the status of the battery pack; among them, the early warning signal includes early warning signal 1 and early warning signal 2.

9. The artificial intelligence-based battery pack status monitoring system according to claim 8, characterized in that: The targeted monitoring of the battery pack based on the early warning signal includes: Extract early warning signals from each monitoring area; Determine whether the warning signal of monitoring area i is warning signal 1; if yes, set the data collection frequency of monitoring area i to 2F; if not, set the data collection frequency of monitoring area i to 1.5F; where F is the default value of the data collection frequency.

10. A battery pack status monitoring method based on artificial intelligence, operating based on an artificial intelligence-based battery pack status monitoring system according to any one of claims 1 to 9, characterized in that: include: Obtain historical abnormal alarm information of the battery pack, divide the battery pack into regions based on the historical abnormal alarm information, and obtain multiple monitoring areas; Collect the operating parameters of each monitoring area in several consecutive cycles; Performing data analysis on the operating parameters to obtain operating characteristic data; Calculate the safety risk factor of each monitoring area based on the operating characteristic data; Input the safety risk coefficient of each monitoring area into the safety risk prediction model to obtain the safety risk coefficient prediction value; Generate early warning signals based on the predicted value of the safety risk factor and the preset risk threshold range; Conduct targeted monitoring of battery packs based on early warning signals.

Images