Lithium battery health state monitoring system based on multi-modal data fusion

Abstract

The invention relates to the technical field of lithium batteries, in particular to a lithium battery health state monitoring system based on multi-modal data fusion. The environment and working condition data corresponding to the lithium battery are extracted based on the database, the environment and working condition data of the lithium battery are input into the pre-deployed monitoring state determining unit, the monitoring state level is output through the monitoring state determining unit, monitoring adjustment is performed according to the monitoring state level, and the monitoring level is adjusted. The output of monitoring resources is optimized; specific grade information is obtained by identifying the monitoring state grade of the lithium battery, modal analysis data corresponding to each modal data is obtained based on a database, the specific grade information and the modal analysis data are input into an importance screening unit, and a modal fusion instruction is output through the importance screening unit; and various modal data are screened, so that the processing burden of the system is reduced, and the calculation load is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium batteries, and in particular to a lithium battery health status monitoring system based on multimodal data fusion. Background Art

[0002] As a highly efficient energy storage medium, lithium batteries are a key link in the large-scale application of renewable energy. Their high energy density and long cycle life effectively address the spatial and temporal mismatch between energy supply and demand, driving the transition to a low-carbon energy structure. Compared to traditional lead-acid and nickel-metal hydride batteries, lithium batteries offer advantages such as high energy density, low self-discharge, and long cycle life, significantly improving the endurance of devices. Therefore, monitoring the health of lithium batteries is particularly important to identify risks such as battery aging, overcharging, and short circuits, and to avoid safety incidents such as thermal runaway and fire.

[0003] Traditional battery health status monitoring systems must collect multi-source modal data from batteries and perform fusion analysis to more comprehensively characterize the battery status. However, data from more sources means that the monitoring system needs to analyze a variety of modal data separately and finally fuse and summarize them, which increases the system's burden on data processing and reduces the system's response speed. In addition, some modal data has a low correlation with battery health, resulting in a large amount of redundant analysis. Summary of the Invention

[0004] The present invention provides a lithium battery health status monitoring system based on multimodal data fusion, which is used to solve the technical problems mentioned in the above background technology.

[0005] The present invention provides a lithium battery health status monitoring system based on multimodal data fusion, which includes a multimodal data acquisition module, a monitoring status analysis module, a modal importance screening module, a multimodal data fusion module and a data storage and output module.

[0006] The multimodal data acquisition module obtains the sensor group deployed corresponding to each lithium battery, obtains the data timestamp of each sensor corresponding to the sensor group, unifies the data timestamp corresponding to each sensor through a sliding window interpolation algorithm to obtain the baseline data timestamp of each sensor corresponding to the sensor group, obtains multimodal data based on the baseline data timestamp using each sensor in the sensor group, and inputs the multimodal data corresponding to the lithium battery into the data storage and output module; the types of multimodal data include electrochemical data, thermodynamic data, and environmental and working condition data.

[0007] The monitoring status analysis module extracts the environmental and operating condition data corresponding to the lithium battery based on the database, inputs the environmental and operating condition data of the lithium battery into the pre-deployed monitoring status determination unit, outputs the monitoring status level through the monitoring status determination unit, and performs monitoring adjustments based on the monitoring status level.

[0008] As a further improvement of the present invention, the monitoring state determination unit is specifically: Obtain the pre-set standard battery health index corresponding to the lithium battery and record it as α. At the same time, obtain the pre-set two health index separation indices and record them as α1 and α2 respectively. According to the health index separation indices, the standard battery health index is divided into three health status ranges. The three health status ranges correspond to α≥α1, α1>α>α2, and α≤α2 respectively. Analyze the input environment and working condition data to obtain the real-time battery health index α corresponding to the lithium battery. 实时 , change α 实时 Matching with the three health status ranges to obtain the corresponding monitoring status level; To α 实时 For identification, when α 实时 When α≥α1, the corresponding monitoring state level is generated as low-intensity monitoring state; When α 实时 When α1>α>α2, the corresponding monitoring state level is medium intensity monitoring state; When α 实时 When α≤α2, the corresponding monitoring status level is generated as a high-light monitoring status.

[0009] As a further improvement of the present invention, the input environment and working condition data are analyzed, specifically: The environmental and working condition data are identified to obtain the battery usage time, charge and discharge cycle data and usage environment data; the battery usage time is identified to obtain the operating time and self-discharge time, the length of one day is set as a unit time, the operating days and self-discharge days corresponding to the operating time are counted, and the operating attenuation capacity and self-discharge attenuation capacity are obtained based on the capacity attenuation model of the lithium battery. The operating attenuation capacity and the self-discharge attenuation capacity are calculated and summed to obtain the total capacity attenuation, and the total capacity attenuation is calculated by ratioing the total capacity attenuation with the preset standard capacity to obtain the capacity attenuation rate.

[0010] Based on the charge and discharge cycle data, the single charge and discharge data corresponding to the lithium battery is obtained. According to the single charge and discharge data, the corresponding discharge capacity is obtained as a percentage of the rated capacity. According to the percentage of the rated discharge capacity corresponding to the single charge and discharge, the corresponding single charge and discharge depth is obtained. The single charge and discharge depth corresponding to each single charge and discharge is substituted into the calculation formula of the equivalent full cycle number. Calculate the equivalent full cycle number EFC; where DOD i Indicates the depth of single charge and discharge.

[0011] Obtaining environmental vibration data corresponding to the lithium battery according to the usage environment data, obtaining the amplitude and vibration duration of the environmental vibration based on the environmental vibration data of the lithium battery, obtaining a pre-designed amplitude threshold, marking the vibration duration corresponding to the amplitude of the environmental vibration exceeding the amplitude threshold as abnormal vibration duration, dividing the abnormal vibration duration into multiple abnormal vibration duration intervals according to preset duration intervals, and setting a vibration duration impact value for each abnormal vibration duration interval; Get the real-time vibration acceleration based on the environmental vibration data, and obtain the vibration tolerance threshold and material sensitivity coefficient corresponding to the lithium battery stored in the database, and use the vibration damage index calculation formula The vibration damage index VDI is calculated; where A(t) represents the real-time vibration acceleration, A threshold is represented as the vibration tolerance threshold, k is represented as the material sensitivity coefficient, and dt is represented as the integral time element; a pre-set damage index threshold is obtained, and the portion of the current corresponding vibration damage index that exceeds the damage index threshold is marked as the vibration damage impact value; The capacity attenuation rate, equivalent full cycle number, vibration duration impact value and vibration damage impact value are normalized and their values are taken. The corresponding values of the capacity attenuation rate, equivalent full cycle number, vibration duration impact value and vibration damage impact value are calculated by weighted calculation to obtain the battery health impact index. The standard battery health index and the battery health impact index are subtracted to obtain the real-time battery health index.

[0012] The modal importance screening module obtains the current monitoring status level corresponding to the lithium battery, identifies the monitoring status level of the lithium battery to obtain specific level information, obtains the modal analysis data corresponding to each modal data based on the database, inputs the specific level information and modal analysis data into the importance screening unit, obtains the modal fusion instruction through the output of the importance screening unit, and sends the modal fusion instruction to the multimodal data fusion module.

[0013] As a further improvement of the present invention, the importance screening unit has the following specific analysis content: analyzing the specific level information of the lithium battery to obtain the importance screening level corresponding to the lithium battery, and analyzing the modal analysis data corresponding to each modal data to obtain the modal importance index; comparing the modal importance index corresponding to each modal data with the correlation index corresponding to the corresponding importance screening level; when the modal importance index is greater than the correlation index, marking the corresponding modal data as preferred modal data, and aggregating the preferred modal data to obtain a modal fusion instruction.

[0014] As a further improvement of the present invention, analysis is performed based on the specific grade information of the lithium battery, which is specifically: Identify the current specific level information of the lithium battery. When the specific level information is in a low-intensity monitoring state, generate a high-intensity screening level for the importance screening level. When the specific level information corresponds to the medium intensity monitoring state, the corresponding importance screening level generated is the medium intensity screening level; When the specific level information corresponds to a high-intensity monitoring state, the corresponding importance screening level generated is a low-intensity screening level; The importance screening levels include high-intensity screening level, medium-intensity screening level and low-intensity screening level; the preset correlation indicators corresponding to each importance screening level are obtained, and based on the order from high to low, the correlation indicators corresponding to the high-intensity screening level, medium-intensity screening level and low-intensity screening level decrease in turn.

[0015] As a further improvement of the present invention, analysis is performed using the modal analysis data corresponding to each modal data, specifically: Obtain the modal analysis data corresponding to each modal data, and identify the modal analysis data to obtain the physical correlation and data distribution characteristics corresponding to each modal data; obtain the physical association problem corresponding to each modal data based on the physical correlation, obtain the activation time corresponding to the physical association problem corresponding to each modal data, and divide the physical association problem into low-sensitivity physical association problem and high-sensitivity physical association problem based on the pre-set activation boundary time; count the number of high-sensitivity physical association problems and the total number of physical association problems corresponding to each modal data, and use The physical correlation index is calculated. For example, the physical correlation problems corresponding to the temperature mode are thermal effect, thermal cycle attenuation, and thermal runaway. That is, the total number of physical correlation problems corresponding to the temperature mode is 3, among which thermal runaway and thermal effect correspond to highly sensitive physical correlation problems, that is, the corresponding physical correlation index is 1.67.

[0016] Based on the data distribution characteristics, the number of abnormal signals and the total number of signals within the preset analysis period corresponding to each modal data are obtained. The ratio of the number of abnormal signals corresponding to each modal data to the total number of signals is calculated to obtain the abnormal density. The abnormal density is divided into multiple abnormal density intervals according to the pre-designed density interval. A risk coefficient is set for each abnormal density interval. The abnormal density corresponding to each modal data is matched with the corresponding multiple abnormal density intervals to obtain the corresponding risk coefficient. The abnormal density and the corresponding risk coefficient are multiplied by a multiplier to obtain the modal risk value F.

[0017] The physical correlation index and modal risk value corresponding to each modal data are normalized and the formula The modal importance index is calculated; wherein β represents the weight coefficient, and its value can be 0.65; the modal importance index corresponding to each modal data is compared with the correlation index corresponding to the corresponding importance screening level. When the modal importance index is greater than the correlation index, the corresponding modal data is marked as the preferred modal data, and the preferred modal data are aggregated to obtain the modal fusion instruction.

[0018] The multimodal data fusion module is used to receive modal fusion instructions and identify the modal fusion instructions to obtain preferred modal data, perform multimodal fusion on the preferred modal data to obtain a health status assessment model, obtain real-time health status information of the lithium battery based on the health status assessment model, and send the real-time health status information to the data storage and output module.

[0019] The data storage and output module includes a database and an information output unit; the database stores multimodal data such as electrochemical data, thermodynamic data, and environmental and operating condition data, as well as monitoring status levels, modal fusion instructions, and real-time health status information; the information output unit displays real-time health status information on pre-deployed displays and mobile terminals through data transmission for early warning.

[0020] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: 1. The present invention extracts the environmental and operating condition data corresponding to the lithium battery, inputs the environmental and operating condition data of the lithium battery into a pre-deployed monitoring status determination unit, outputs the monitoring status level through the monitoring status determination unit, performs monitoring adjustment according to the monitoring status level, adjusts the monitoring level, and optimizes the output of monitoring resources.

[0021] 2. The present invention obtains specific level information by identifying the monitoring status level of the lithium battery, obtains modal analysis data corresponding to each modal data based on the database, inputs the specific level information and modal analysis data into the importance screening unit, and obtains modal fusion instructions through the output of the importance screening unit. Screening of a large number of modal data reduces the system processing burden and reduces the computing load. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.

[0023] Figure 1 It is a principle block diagram of the present invention; Figure 2 is a schematic block diagram of a monitoring status analysis module of the present invention; Figure 3Schematic diagram of the modality importance screening module of the present invention. DETAILED DESCRIPTION

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0025] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1-3 In an embodiment of the present invention, an embodiment of a lithium battery health status monitoring system based on multimodal data fusion includes: The multimodal data acquisition module obtains the sensor group deployed corresponding to each lithium battery, obtains the data timestamp of each sensor corresponding to the sensor group, unifies the data timestamp corresponding to each sensor through a sliding window interpolation algorithm to obtain the baseline data timestamp of each sensor corresponding to the sensor group, obtains multimodal data based on the baseline data timestamp using each sensor in the sensor group, and inputs the multimodal data corresponding to the lithium battery into the data storage and output module.

[0026] Types of multimodal data include electrochemical data, thermodynamic data, and environmental and operating condition data; among them, the modal data corresponding to each type include but are not limited to voltage, current, incremental capacity analysis, temperature rise rate, temperature distribution, mechanical vibration, acoustic emission signals, ambient temperature and humidity, and number of charge and discharge cycles.

[0027] The monitoring status analysis module extracts the environmental and operating condition data corresponding to the lithium battery based on the database, inputs the environmental and operating condition data of the lithium battery into the pre-deployed monitoring status determination unit, outputs the monitoring status level through the monitoring status determination unit, and performs monitoring and adjustment according to the monitoring status level.

[0028] Furthermore, the specific analysis process of the monitoring status determination unit is as follows: Obtain the pre-set standard battery health index corresponding to the lithium battery and record it as α. At the same time, obtain the pre-set two health index separation indices and record them as α1 and α2 respectively. According to the health index separation indices, the standard battery health index is divided into three health status ranges. The three health status ranges correspond to α≥α1, α1>α>α2, and α≤α2 respectively. Analyze the input environment and working condition data to obtain the real-time battery health index α corresponding to the lithium battery. 实时 , change α 实时 Matching with the three health status ranges to obtain the corresponding monitoring status level; To α 实时For identification, when α 实时 When α≥α1, the corresponding monitoring state level is generated as low-intensity monitoring state; When α 实时 When α1>α>α2, the corresponding monitoring state level is medium intensity monitoring state; When α 实时 When α≤α2, the corresponding monitoring status level is generated as a high-light monitoring status.

[0029] It is worth noting that, for example, if the standard battery health index α is set to 100%, where α1 and α2 correspond to 80% and 50% respectively, when the real-time health index α of the lithium battery is 实时 =80%, the corresponding monitoring status level is low-intensity monitoring status.

[0030] Analyze the input environment and working condition data, specifically: The environmental and working condition data are identified to obtain the battery usage time, charge and discharge cycle data and usage environment data; the battery usage time is identified to obtain the operating time and self-discharge time, the length of one day is set as a unit time, the operating days and self-discharge days corresponding to the operating time are counted, and the operating attenuation capacity and self-discharge attenuation capacity are obtained based on the capacity attenuation model of the lithium battery. The operating attenuation capacity and the self-discharge attenuation capacity are calculated and summed to obtain the total capacity attenuation, and the total capacity attenuation is calculated by ratioing the total capacity attenuation with the preset standard capacity to obtain the capacity attenuation rate.

[0031] Based on the charge and discharge cycle data, the single charge and discharge data corresponding to the lithium battery is obtained. According to the single charge and discharge data, the corresponding discharge capacity is obtained as a percentage of the rated capacity. According to the percentage of the rated discharge capacity corresponding to the single charge and discharge, the corresponding single charge and discharge depth is obtained. The single charge and discharge depth corresponding to each single charge and discharge is substituted into the calculation formula of the equivalent full cycle number. Calculate the equivalent full cycle number EFC; where DOD i Indicates the depth of single charge and discharge.

[0032] According to the usage environment data, the environmental vibration data corresponding to the lithium battery is obtained, and the amplitude and vibration duration of the environmental vibration are obtained based on the environmental vibration data of the lithium battery. A pre-given amplitude threshold is obtained, and the vibration duration corresponding to the amplitude of the environmental vibration exceeding the amplitude threshold is marked as abnormal vibration duration. The duration is divided into multiple abnormal vibration duration intervals according to the preset duration interval, and each abnormal vibration duration interval is set with a vibration duration impact value; the abnormal vibration duration is matched with the multiple abnormal vibration duration intervals to obtain the corresponding vibration duration impact value.

[0033] Get the real-time vibration acceleration based on the environmental vibration data, and obtain the vibration tolerance threshold and material sensitivity coefficient corresponding to the lithium battery stored in the database, and use the vibration damage index calculation formula The vibration damage index VDI is calculated; where A(t) represents the real-time vibration acceleration, A threshold is represented as the vibration tolerance threshold, k is represented as the material sensitivity coefficient, and dt is represented as the integral time element; a pre-set damage index threshold is obtained, and the portion of the current corresponding vibration damage index that exceeds the damage index threshold is marked as the vibration damage impact value; The capacity attenuation rate C1, the equivalent full cycle number N2, the vibration duration impact value T_v and the vibration damage impact value D_v are normalized and their values are taken. The corresponding values of the capacity attenuation rate C1, the equivalent full cycle number N2, the vibration duration impact value Tt and the vibration damage impact value TD are calculated by weighted calculation to obtain the battery health impact index S y , the standard battery health index S0 and the battery health impact index S y The difference calculation is performed to obtain the real-time battery health index (S0H), which is: S0H=S0-S y Where, are the attenuation coefficients corresponding to the capacity attenuation rate, equivalent full cycle number, and vibration duration impact value, respectively, and , whose size is obtained by regression analysis or machine learning fitting. are the attenuation coefficients corresponding to the capacity attenuation rate, equivalent full cycle number, vibration duration impact value, and vibration damage impact value, respectively. Their sizes are obtained by regression analysis or machine learning fitting, and .

[0034] The modal importance screening module obtains the current monitoring status level corresponding to the lithium battery, identifies the monitoring status level of the lithium battery to obtain specific level information, obtains the modal analysis data corresponding to each modal data based on the database, inputs the specific level information and modal analysis data into the importance screening unit, obtains the modal fusion instruction through the output of the importance screening unit, and sends the modal fusion instruction to the multimodal data fusion module.

[0035] The specific analysis content of the importance screening unit is: The importance screening level corresponding to the lithium battery is obtained by analyzing the specific level information of the lithium battery. Specifically, the current specific level information of the lithium battery is identified. When the specific level information is in a low-intensity monitoring state, the importance screening level is generated as a high-intensity screening level. When the specific level information corresponds to the medium intensity monitoring state, the corresponding importance screening level generated is the medium intensity screening level; When the specific level information corresponds to a high-intensity monitoring state, the corresponding importance screening level generated is a low-intensity screening level; The importance screening levels include high-intensity screening level, medium-intensity screening level and low-intensity screening level; the preset correlation indicators corresponding to each importance screening level are obtained, and based on the order from high to low, the correlation indicators corresponding to the high-intensity screening level, medium-intensity screening level and low-intensity screening level decrease in turn.

[0036] The modal importance index is obtained by analyzing the modal analysis data corresponding to each modal data, which is specifically as follows: the modal analysis data corresponding to each modal data is obtained, and the modal analysis data is identified to obtain the physical correlation and data distribution characteristics corresponding to each modal data; the physical correlation problem corresponding to each modal data is obtained based on the physical correlation, the activation time corresponding to the physical correlation problem corresponding to each modal data is obtained, and the physical correlation problem is divided into low-sensitivity physical correlation problem and high-sensitivity physical correlation problem based on the pre-set activation boundary time; among them, the low-sensitivity physical correlation problem corresponds to the physical correlation problem with a longer triggering time, and the high-sensitivity physical correlation problem corresponds to the physical correlation problem with a lower triggering time; the number of high-sensitivity physical correlation problems corresponding to each modal data is counted Total number of physics-related questions ,pass The physical correlation index is calculated. For example, the physical correlation problems corresponding to the temperature mode are thermal effect, thermal cycle attenuation, and thermal runaway. That is, the total number of physical correlation problems corresponding to the temperature mode is 3, among which thermal runaway and thermal effect correspond to highly sensitive physical correlation problems, that is, the corresponding physical correlation index is 1.67.

[0037] Based on the data distribution characteristics, the number of abnormal signals and the total number of signals within the preset analysis period corresponding to each modal data are obtained. The ratio of the number of abnormal signals corresponding to each modal data to the total number of signals is calculated to obtain the abnormal density. The abnormal density is divided into multiple abnormal density intervals according to the pre-designed density interval. A risk coefficient is set for each abnormal density interval. The abnormal density corresponding to each modal data is matched with the corresponding multiple abnormal density intervals to obtain the corresponding risk coefficient. The abnormal density and the corresponding risk coefficient are multiplied by a multiplier to obtain the modal risk value F.

[0038] The physical correlation index and modal risk value corresponding to each modal data are normalized and the formula The modal importance index is calculated; wherein β represents the weight coefficient, and its value can be 0.65; the modal importance index corresponding to each modal data is compared with the correlation index corresponding to the corresponding importance screening level. When the modal importance index is greater than the correlation index, the corresponding modal data is marked as the preferred modal data, and the preferred modal data are aggregated to obtain the modal fusion instruction.

[0039] The multimodal data fusion module receives the modal fusion instruction and identifies the modal fusion instruction to obtain the preferred modal data, performs multimodal fusion on each preferred modal data to obtain a health status assessment model, obtains the real-time health status information of the lithium battery based on the health status assessment model, and sends the real-time health status information to the data storage and output module.

[0040] The data storage and output module includes a database and an information output unit; the database stores multimodal data such as electrochemical data, thermodynamic data, and environmental and operating condition data, as well as monitoring status levels, modal fusion instructions, and real-time health status information; the information output unit displays real-time health status information on pre-deployed displays and mobile terminals through data transmission for early warning.

[0041] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A lithium battery health status monitoring system based on multimodal data fusion, including a multimodal data acquisition module and a data storage and output module, characterized in that: Also includes: The monitoring status analysis module extracts the environmental and operating condition data corresponding to the lithium battery based on the database, inputs the environmental and operating condition data of the lithium battery into the pre-deployed monitoring status determination unit, outputs the monitoring status level through the monitoring status determination unit, and performs monitoring and adjustment according to the monitoring status level; The modal importance screening module obtains the current monitoring status level of the lithium battery, identifies the monitoring status level of the lithium battery to obtain specific level information, obtains the modal analysis data corresponding to each modal data based on the database, inputs the specific level information and modal analysis data into the importance screening unit, obtains the modal fusion instruction through the output of the importance screening unit, and sends the modal fusion instruction to the multimodal data fusion module; The multimodal data fusion module is used for multimodal data fusion and generation of real-time health status information, and sends the real-time health status information to the data storage and output module.

2. The lithium battery health status monitoring system based on multimodal data fusion according to claim 1 is characterized in that: The multimodal data acquisition module obtains the sensor group deployed corresponding to each lithium battery, obtains the data timestamp of each sensor corresponding to the sensor group, unifies the data timestamp corresponding to each sensor through a sliding window interpolation algorithm to obtain the baseline data timestamp of each sensor corresponding to the sensor group, obtains multimodal data based on the baseline data timestamp using each sensor in the sensor group, and inputs the multimodal data corresponding to the lithium battery into the data storage and output module; the types of multimodal data include electrochemical data, thermodynamic data, and environmental and operating condition data.

3. The lithium battery health status monitoring system based on multimodal data fusion according to claim 1 is characterized in that: The monitoring status determination unit is specifically: Obtain the pre-set standard battery health index corresponding to the lithium battery and record it as α. At the same time, obtain the pre-set two health index separation indices and record them as α1 and α2 respectively. According to the health index separation indices, the standard battery health index is divided into three health status ranges. The three health status ranges correspond to α≥α1, α1>α>α2, and α≤α2 respectively. Analyze the input environment and working condition data to obtain the real-time battery health index α corresponding to the lithium battery. 实时 , change α 实时 Matching with the three health status ranges obtains the corresponding monitoring status level.

4. The lithium battery health status monitoring system based on multimodal data fusion according to claim 3 is characterized in that: The input environment and working condition data are analyzed, specifically: The environmental and working condition data are identified to obtain the battery usage time, charge and discharge cycle data, and usage environment data. The battery usage time is identified to obtain the operating time and self-discharge time. The length of one day is set as a unit time. The operating days and self-discharge days corresponding to the operating time are counted. Based on the capacity decay model of the lithium battery, the operating decay capacity and self-discharge decay capacity are respectively obtained. The operating decay capacity and self-discharge decay capacity are calculated and summed to obtain the total capacity decay. The total capacity decay is calculated by comparing it with the preset standard capacity to obtain the capacity decay rate. Based on the charge and discharge cycle data, the single charge and discharge data corresponding to the lithium battery is obtained. According to the single charge and discharge data, the corresponding discharge capacity is obtained as a percentage of the rated capacity. According to the percentage of the rated discharge capacity corresponding to the single charge and discharge, the corresponding single charge and discharge depth is obtained. The single charge and discharge depth corresponding to each single charge and discharge is substituted into the calculation formula of the equivalent full cycle number. Calculate the equivalent full cycle number EFC; where DOD i Indicates the depth of single charge and discharge; Obtaining environmental vibration data corresponding to the lithium battery according to the usage environment data, obtaining the amplitude and vibration duration of the environmental vibration based on the environmental vibration data of the lithium battery, obtaining a pre-designed amplitude threshold, marking the vibration duration corresponding to the amplitude of the environmental vibration exceeding the amplitude threshold as abnormal vibration duration, dividing the abnormal vibration duration into multiple abnormal vibration duration intervals according to preset duration intervals, and setting a vibration duration impact value for each abnormal vibration duration interval; The real-time vibration acceleration is obtained based on the environmental vibration data, and the vibration tolerance threshold and material sensitivity coefficient corresponding to the lithium battery stored in the database are obtained, and the vibration damage index calculation formula is used to calculate The vibration damage index VDI is obtained; where A(t) represents the real-time vibration acceleration, A threshold is represented as the vibration tolerance threshold, k is represented as the material sensitivity coefficient, and dt is represented as the integral time element; a pre-set damage index threshold is obtained, and the portion of the current corresponding vibration damage index that exceeds the damage index threshold is marked as the vibration damage impact value; The capacity attenuation rate, equivalent full cycle number, vibration duration impact value and vibration damage impact value are normalized and their values are taken. The corresponding values of the capacity attenuation rate, equivalent full cycle number, vibration duration impact value and vibration damage impact value are calculated by weighted calculation to obtain the battery health impact index. The standard battery health index and the battery health impact index are subtracted to obtain the real-time battery health index.

5. The lithium battery health status monitoring system based on multimodal data fusion according to claim 1, characterized in that: The importance screening unit specifically analyzes the following: analyzing the specific level information of the lithium battery to obtain the importance screening level corresponding to the lithium battery, and analyzing the modal analysis data corresponding to each modal data to obtain the modal importance index; comparing the modal importance index corresponding to each modal data with the correlation index corresponding to the corresponding importance screening level; when the modal importance index is greater than the correlation index, marking the corresponding modal data as preferred modal data, and aggregating the preferred modal data to obtain a modal fusion instruction.

6. The lithium battery health status monitoring system based on multimodal data fusion according to claim 5, characterized in that: The analysis is performed based on the specific grade information of the lithium battery, which is specifically: Identify the current specific level information of the lithium battery. When the specific level information is in a low-intensity monitoring state, generate a high-intensity screening level for the importance screening level. When the specific level information corresponds to the medium intensity monitoring state, the corresponding importance screening level generated is the medium intensity screening level; When the specific level information corresponds to a high-intensity monitoring state, the corresponding importance screening level generated is a low-intensity screening level; Importance screening levels include high intensity screening level, medium intensity screening level and low intensity screening level; The preset correlation index corresponding to each importance screening level is obtained. Based on the order from high to low, the correlation index corresponding to the high intensity screening level, the medium intensity screening level and the low intensity screening level decreases in turn.

7. The lithium battery health status monitoring system based on multimodal data fusion according to claim 6, characterized in that: The analysis is performed on the modal analysis data corresponding to each modal data, which is specifically: Obtaining modal analysis data corresponding to each modal data, and identifying the modal analysis data to obtain the physical correlation and data distribution characteristics corresponding to each modal data; Based on the physical correlation, the physical correlation problems corresponding to each modal data are obtained, the activation time corresponding to the physical correlation problems corresponding to each modal data is obtained, and the physical correlation problems are divided into low-sensitivity physical correlation problems and high-sensitivity physical correlation problems based on the pre-set activation boundary time; the number of high-sensitivity physical correlation problems and the total number of physical correlation problems corresponding to each modal data are counted, and the activation time is calculated by The physical correlation index is calculated; Based on the data distribution characteristics, the number of abnormal signals and the total number of signals within a preset analysis period corresponding to each modal data are obtained, the ratio of the number of abnormal signals corresponding to each modal data to the total number of signals is calculated to obtain the abnormal density, the abnormal density is divided into multiple abnormal density intervals according to a pre-designed density interval, a risk coefficient is set for each abnormal density interval, the abnormal density corresponding to each modal data is matched with the corresponding multiple abnormal density intervals to obtain the corresponding risk coefficient, and the abnormal density and the corresponding risk coefficient are multiplied by a multiplier to obtain the modal risk value F; The physical correlation index and modal risk value corresponding to each modal data are normalized and the formula The modal importance index is calculated, where β represents the weight coefficient.

8. The lithium battery health status monitoring system based on multimodal data fusion according to claim 1, characterized in that: The multimodal data fusion module receives a modal fusion instruction and identifies the modal fusion instruction to obtain preferred modal data, performs multimodal fusion on each preferred modal data to obtain a health status assessment model, obtains real-time health status information of the lithium battery according to the health status assessment model, and sends the real-time health status information to the data storage and output module.

9. The lithium battery health status monitoring system based on multimodal data fusion according to claim 1, characterized in that: The data storage and output module includes a database and an information output unit; the database stores multimodal data including electrochemical data, thermodynamic data, and environmental and working condition data, as well as monitoring status levels, modal fusion instructions, and real-time health status information; the information output unit displays the real-time health status information on a pre-deployed display and a mobile terminal for early warning through data transmission.

Images