Power battery multi-mode health assessment method, device, equipment and medium

Abstract

The invention discloses a multi-mode health assessment method, device and equipment for a power battery and a medium. The method is characterized by comprising the following steps: collecting multi-mode sensing data of the power battery; performing data processing on the battery image data and the battery sound information to determine battery standard image data and battery standard sound signals; extracting battery abnormal feature information corresponding to the battery standard image data and the battery standard sound signal; and performing health assessment on the power battery based on the battery abnormal feature information, and determining a comprehensive health assessment report. The early discovery capability of potential faults of the battery and the comprehensive study and judgment level of the health state are remarkably improved, and a visual and reliable decision basis is provided for safe operation and maintenance of the battery.

Description

Technical Field

[0001] This invention relates to the field of power battery testing technology, and in particular to a method, apparatus, equipment and medium for multimodal health assessment of power batteries. Background Technology

[0002] With the rapid development of the new energy vehicle industry, the safety and reliability of power batteries, as core components, are receiving increasing attention. During long-term use, power batteries are prone to failures such as casing deformation, internal short circuits, localized overheating, and insulation degradation due to factors such as overcharging, over-discharging, mechanical vibration, and changes in ambient temperature. In severe cases, this can lead to thermal runaway or even fire and explosion. Currently, traditional power battery testing methods mainly rely on monitoring electrical parameters such as voltage, current, and temperature provided by the battery management system. These methods have significant limitations: First, contact sensors are complex to install and cannot fully cover all critical parts of the battery pack; second, monitoring a single electrical parameter cannot effectively identify potential risks such as physical damage and partial discharge; and third, existing detection methods lack the ability to warn of early-stage faults, often only detecting them when they have progressed to a severe stage.

[0003] In recent years, non-contact detection technologies have been gradually applied to battery health monitoring. For example, infrared thermal imaging technology can be used to detect battery temperature distribution, ultraviolet imaging technology can be used to detect partial discharge phenomena, and acoustic fingerprint detection technology can be used to identify abnormal sounds inside the battery. However, most existing technical solutions are limited to single-modal detection and lack effective integration of multi-source information, making it impossible to achieve a comprehensive assessment of battery health status. Furthermore, the heterogeneity between different sensor data poses a significant challenge to data fusion and analysis, making the establishment of a unified assessment system a key technical difficulty.

[0004] Therefore, there is an urgent need to develop a system and method that can comprehensively utilize multimodal sensing information to achieve a comprehensive and accurate assessment of the health status of power batteries, so as to improve the safe operation level of new energy vehicles. Summary of the Invention

[0005] This invention provides a method, apparatus, device, and medium for multimodal health assessment of power batteries, in order to solve the technical problem that existing technologies cannot comprehensively and accurately assess the health status of power batteries.

[0006] According to one aspect of the present invention, a multimodal health assessment method for a power battery is provided, comprising:

[0007] Collect multimodal sensing data of the power battery, wherein the multimodal sensing data includes battery image data and battery sound signals;

[0008] The battery image data and the battery sound information are processed separately to determine the battery standard image data and the battery standard sound signal;

[0009] Extract the battery abnormality feature information corresponding to the battery standard image data and the battery standard sound signal; wherein, the battery abnormality feature information includes: physical abnormality feature information, thermal abnormality feature information, discharge abnormality feature information and sound abnormality feature information;

[0010] A health assessment of the power battery is performed based on the abnormal battery characteristics information, and a comprehensive health assessment report is generated.

[0011] According to another aspect of the present invention, a multimodal health assessment device for a power battery is provided, comprising:

[0012] The data acquisition module is used to acquire multimodal sensing data of the power battery, wherein the multimodal sensing data includes battery image data and battery sound signals;

[0013] The signal processing module is used to process the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal;

[0014] The signal recognition module is used to extract battery abnormality feature information corresponding to the battery standard image data and the battery standard sound signal; wherein, the battery abnormality feature information includes: physical abnormality feature information, thermal abnormality feature information, discharge abnormality feature information and sound abnormality feature information;

[0015] The health assessment module is used to perform a health assessment on the power battery based on the abnormal battery characteristic information and to determine a comprehensive health assessment report.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the power battery multimodal health assessment method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the power battery multimodal health assessment method according to any embodiment of the present invention.

[0021] The technical solution of this invention collects multimodal sensing data from a power battery, processes the battery image data and battery sound information to determine standard battery image data and standard battery sound signals, and effectively improves data accuracy and evaluation accuracy by processing multi-source data. It extracts abnormal battery feature information corresponding to the standard battery image data and standard battery sound signals, achieving comprehensive abnormal feature extraction of the power battery from physical structure, thermal state, insulation performance to operating sounds. Further health assessment is performed based on the extracted multi-dimensional abnormal features. A comprehensive health assessment report is generated based on the battery abnormal feature information, enabling quantitative evaluation of battery health status, effectively identifying different safety levels from early warning to high risk, and generating a health assessment report with a comprehensive health score and a clear risk level. This solves the technical problem of existing technologies being unable to comprehensively and accurately assess the health status of power batteries, significantly improving the early detection capability of potential battery faults and the comprehensive judgment level of battery health status, providing an intuitive and reliable decision-making basis for the safe operation and maintenance of batteries.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 A flowchart of a multimodal health assessment method for power batteries is provided as an embodiment of the present invention;

[0025] Figure 2 A flowchart of a multimodal health assessment method for power batteries provided in an embodiment of the present invention;

[0026] Figure 3 A flowchart of a multimodal health assessment method for power batteries provided in an embodiment of the present invention;

[0027] Figure 4 This is a schematic diagram of the structure of a multimodal health assessment device for power batteries provided in an embodiment of the present invention;

[0028] Figure 5 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. Detailed Implementation

[0029] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0031] Figure 1 This invention provides a flowchart of a multimodal health assessment method for power batteries, applicable to situations involving health assessments of power batteries installed in vehicles. This method can be executed by a multimodal health assessment device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0032] S110: Collects multimodal sensing data from the power battery.

[0033] The multimodal sensing data includes battery image data and battery sound signals; the battery image data includes a first visible light image, a first infrared image, and a first ultraviolet image; the first visible light image may be an image acquired by a high-resolution industrial camera of the multimodal sensor array; the first infrared image may be an image acquired by an infrared thermal imager of the multimodal sensor array; and the first ultraviolet image may be an image acquired by an acoustic sensor of the multimodal sensor array.

[0034] Optionally, the battery audio signal can be an audio signal acquired by a wideband microphone array of a multimodal sensor array.

[0035] Optionally, in this invention, the power battery is arranged on the chassis plane of the vehicle, and a multimodal sensor array can be arranged below the battery. The multimodal sensor array can be composed of a visible light camera, an infrared thermal imager, an ultraviolet imager, and an acoustic sensor. A unified synchronous trigger signal is sent to each sensor in the multimodal sensor array, enabling data acquisition to be started under a unified timing reference.

[0036] For example, the visible light camera, infrared thermal imager, ultraviolet imager, and acoustic sensor can be selected from a high-resolution industrial camera, an uncooled infrared thermal imager, a solar-blind ultraviolet imager, and a wideband microphone array, respectively. The high-resolution industrial camera captures the deformation of the battery casing through its high dynamic range and low-light performance, and receives the visible light radiation energy reflected from the battery surface through its optical lens to obtain a first visible light image. The uncooled infrared thermal imager detects local overheating areas and abnormal temperature distribution on the surface of the power battery through its thermal sensitivity and temperature measurement range to obtain a first infrared image. The solar-blind ultraviolet imager detects the corona discharge phenomenon of the battery and high-voltage wiring through its ultraviolet band sensitivity to obtain a first ultraviolet image. The wideband microphone array collects abnormal sounds generated by gas production, arc discharge, and mechanical loosening inside the battery through its frequency response characteristics and noise suppression capabilities to obtain the battery sound signal.

[0037] Specifically, the visible light radiation energy, infrared radiation energy, and ultraviolet radiation energy of the battery area under test are collected, and the radiation energy is converted into corresponding electrical signals. The electrical signals are then converted into corresponding battery image data through signal processing, and the battery sound signals are collected through an acoustic sensor.

[0038] S120. Perform data processing on the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal.

[0039] Optionally, battery standard image data includes visible light standard images, infrared standard images, and ultraviolet standard images.

[0040] Optionally, in another optional embodiment of the present invention, the step of processing the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal includes:

[0041] Perform image processing on the first visible light image to determine a visible light standard image;

[0042] The first infrared image is processed to determine the infrared standard image;

[0043] Edge features are extracted from the visible light standard image and the first ultraviolet image respectively to determine the first edge pixel set and the second edge pixel set;

[0044] Pixel matching is performed on the first edge pixel set and the second edge pixel set to determine the coordinate transformation relationship, and the first ultraviolet image is resampled based on the coordinate transformation relationship to determine the ultraviolet standard image;

[0045] The battery sound information is processed to determine the standard sound signal of the battery.

[0046] Optionally, in another optional embodiment of the present invention, the step of performing image processing on the first visible light image to determine the visible light standard image includes:

[0047] For each pixel in the first visible light image, the weighted average of all pixels within the neighborhood window of the pixel is calculated based on the weighted average of neighboring pixels and a square filter window. The pixel is then replaced with the weighted average to obtain the second visible light image.

[0048] The contrast of the second visible light image is enhanced based on the brightness distribution histogram of the second visible light image to obtain a third visible light image;

[0049] Identify the apex corner pixel coordinates of the power battery in the third visible light image, and calculate perspective transformation parameters based on the apex corner pixel coordinates;

[0050] The third visible light image is resampled based on the perspective transformation parameters to obtain the visible light standard image.

[0051] Among them, the neighborhood pixel weighted average is used to accumulate and normalize the pixels and weights within the neighborhood window to obtain the weighted average of the pixels, which is then used to replace the current pixel; the square filter window can be a rectangular area of ​​the pixel's neighborhood, with the center of the window coinciding with the current pixel, to ensure that the neighborhood pixels are symmetrically distributed around the target pixel and to avoid positional shift.

[0052] The second visible light image can be a visible light image with noise suppressed from the first visible light image.

[0053] Optionally, for each pixel in the first visible light image, the weighted average of all pixels within the neighborhood window of the pixel is calculated based on the weighted average of neighboring pixels and a square filter window, and the pixel is replaced by the weighted average to obtain the second visible light image.

[0054] Optionally, the third visible light image is a visible light image with contrast enhancement after noise suppression. It should be noted that by statistically analyzing the brightness distribution histogram of all pixels in the second visible light image, and performing pixel-by-pixel brightness mapping based on the brightness distribution histogram, the original brightness values ​​concentrated in a narrow range are redistributed to a wider brightness range, thus obtaining the third visible light image.

[0055] Optionally, the vertex pixel coordinates can be the pixel position coordinates of the four vertex corners of the power battery casing in the image. It should be noted that the four vertex pixel coordinates of the power battery casing are identified in the third visible light image, and the perspective transformation parameters required to correct it to a standard rectangle are calculated based on the four vertex pixel coordinates.

[0056] Optionally, the perspective transformation parameters can be geometric transformation parameters, which can be used to correct the third visible light image. The third visible light image is resampled by the perspective transformation parameters to obtain a standard visible light image.

[0057] Optionally, in another optional embodiment of the present invention, the step of performing image processing on the first infrared image to determine the infrared standard image includes:

[0058] Obtain the original infrared digital value matrix, and convert the original digital values ​​of the original infrared digital value matrix into a first radiation intensity value;

[0059] The ambient temperature of the ambient temperature sensor is obtained, and background noise compensation is performed on each of the radiation intensity values ​​based on the ambient temperature to determine a second radiation intensity value.

[0060] For each of the second radiation intensity values, the surface emissivity parameter corresponding to the battery casing material of the power battery is obtained, and the second radiation intensity value is converted into a measured temperature value based on the surface emissivity parameter;

[0061] The measured temperature value is mapped to the first infrared image to determine the infrared standard image.

[0062] Optionally, the raw infrared digital value matrix can be the raw digital value matrix output by the infrared signal collected by the infrared sensor; each digital value in the raw infrared digital value matrix has a unique correspondence with a detection unit of the sensor, and the value is the raw response quantification result of the detection unit to infrared radiation.

[0063] Optionally, the first radiation intensity value can be a data value describing the infrared radiation intensity. It should be noted that after obtaining the original infrared digital value matrix, each original digital value in the original infrared digital value matrix is ​​converted into a relative first radiation intensity value based on the spectral response characteristics of the infrared sensor.

[0064] Optionally, the second radiation intensity value can be understood as the data value of infrared radiation intensity considering the influence of ambient temperature. It should be noted that the ambient temperature refers to the ambient temperature of the power battery. The second radiation intensity value is determined by compensating for background noise on each first radiation intensity value based on the ambient temperature.

[0065] Optionally, the battery casing material can be the same material that makes up the battery casing covering the power battery; the surface emissivity parameter can be the emissivity of the surface of the battery casing material; and the measured temperature value can be understood as the actual temperature value of the battery casing of the power battery.

[0066] Optionally, the measured temperature value converted from each second radiation intensity value is used as the document data for the corresponding pixel of each second radiation intensity value, so as to obtain a first infrared image with a measured temperature value for each image pixel, i.e., an infrared standard image.

[0067] Optionally, the first set of edge pixels can be a dataset composed of pixels with obvious corner or edge features of the power battery in the visible light standard image; the second set of edge pixels can be a dataset composed of pixels with obvious corner or edge features of the power battery in the first ultraviolet image.

[0068] Optionally, the coordinate transformation relationship can be the transformation relationship between each pixel of the visible light standard image and the first ultraviolet image; after obtaining the first edge pixel set and the second edge pixel set, the positions of each pixel of the first edge pixel set and the second edge pixel set are matched to obtain successfully matched pixel pairs, and then the corresponding coordinate transformation calculation is performed based on the successfully matched pixel pairs to determine the coordinate transformation relationship.

[0069] Optionally, in this invention, pixels with obvious corner or edge features are extracted from the visible light standard image, and corresponding pixels are extracted from the first ultraviolet image to form matching pixel pairs.

[0070] Optionally, based on the coordinate transformation relationship, the first ultraviolet image is resampled to align it spatially with the visible light standard image, resulting in a fused ultraviolet standard image.

[0071] Optionally, a sound wave sensor receives mechanical waves targeting the area of ​​the power battery under test, converts the mechanical waves into electrical signals to obtain a time-domain signal of the sound wave, performs noise reduction processing on the time-domain signal to obtain a noise-reduced acoustic signature signal, and then calculates and transforms the noise-reduced acoustic signature signal to obtain a standard battery sound signal. The standard battery sound signal includes both time-domain and frequency-domain signals.

[0072] S130. Extract the battery standard image data and the battery abnormal feature information corresponding to the battery standard sound signal.

[0073] The battery anomaly information includes physical anomaly information, thermal anomaly information, discharge anomaly information, and acoustic anomaly information. It should be noted that physical anomaly information refers to anomalies in the physical structure of the power battery, which includes at least: the battery casing outline information, surface texture information, and geometric dimensions; physical anomaly information includes: casing deformation, damage, cracks, and leakage information. Thermal anomaly information refers to temperature anomalies in the power battery's temperature distribution field; thermal anomaly information is used to represent abnormally high-temperature and abnormally low-temperature regions in the power battery's temperature distribution field; discharge anomaly information can describe anomalies in abnormal discharge regions of the power battery; acoustic anomaly information can describe abnormal sounds from battery arc discharge and internal gas generation.

[0074] Optionally, for visible light standard images, infrared standard images, and ultraviolet standard images of battery standard image data, physical anomaly feature information is identified based on visible light standard images, thermal anomaly feature information is identified based on infrared standard images, and discharge anomaly feature information is identified based on ultraviolet standard images; for battery standard sound signals, sound anomaly feature information of the power battery is identified through battery standard sound signals.

[0075] Specifically, the abnormal battery feature information corresponding to the standard battery image data and the standard battery sound signal is extracted respectively.

[0076] S140. Based on the abnormal battery characteristic information, perform a health assessment on the power battery and determine a comprehensive health assessment report.

[0077] Optionally, the comprehensive health assessment report can be a report document that describes the health status of the power battery from multiple dimensions. It should be noted that the comprehensive health assessment report consists of three parts: the detected physical anomaly characteristics, thermal anomaly characteristics, discharge anomaly characteristics, and acoustic anomaly characteristics of the power battery; the safety risk level determined based on these characteristics; and the calculated comprehensive health score. The safety risk level can be data describing the risk status of the power battery, including low risk, medium risk, and high risk; the comprehensive health score can be a numerical value representing the overall health of the power battery.

[0078] Specifically, a health assessment of the power battery is performed based on the abnormal battery characteristic information, and a comprehensive health assessment report is generated.

[0079] The technical solution of this invention collects multimodal sensing data from a power battery, processes the battery image data and battery sound information to determine standard battery image data and standard battery sound signals, and effectively improves data accuracy and evaluation accuracy by processing multi-source data. It extracts abnormal battery feature information corresponding to the standard battery image data and standard battery sound signals, achieving comprehensive abnormal feature extraction of the power battery from physical structure, thermal state, insulation performance to operating sounds. Further health assessment is performed based on the extracted multi-dimensional abnormal features. A comprehensive health assessment report is generated based on the battery abnormal feature information, enabling quantitative evaluation of battery health status, effectively identifying different safety levels from early warning to high risk, and generating a health assessment report with a comprehensive health score and a clear risk level. This solves the technical problem of existing technologies being unable to comprehensively and accurately assess the health status of power batteries, significantly improving the early detection capability of potential battery faults and the comprehensive judgment level of battery health status, providing an intuitive and reliable decision-making basis for the safe operation and maintenance of batteries.

[0080] Figure 2 This is a flowchart illustrating a multimodal health assessment method for power batteries, provided as an embodiment of the present invention. The relationship between this embodiment and the previous embodiments is to explain a specific method for identifying abnormal battery characteristics of power batteries. Figure 2 As shown, the method includes:

[0081] S210: Collects multimodal sensing data from the power battery.

[0082] S220. Perform data processing on the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal.

[0083] S230: Extract the continuous contour point sequence and texture statistical features of the power battery surface from the visible light standard image based on the preset edge detection algorithm and the preset texture analysis algorithm, respectively; construct the battery structure model based on the continuous contour point sequence and texture statistical features through camera calibration parameters; and perform anomaly identification based on the battery structure model and the standard battery model to determine physical anomaly feature information.

[0084] Optionally, when identifying physical anomaly features of the power battery, a pre-built standard battery model is used to identify the area in the visible light standard image where the physical structure of the power battery extends beyond the standard battery model, in order to obtain physical anomaly features.

[0085] Optionally, the edge detection algorithm can be a pre-set algorithm for detecting the contour edges of the power battery in a visible light standard image. For example, the edge detection algorithm can be the Canny operator. The texture analysis algorithm can be a pre-set algorithm for extracting the surface texture features of the power battery. For example, the texture analysis algorithm can be a combination of the GLCM gray-level co-occurrence matrix method and LBP features.

[0086] Optionally, the continuous contour point sequence can be formed by arranging the edge points of the power battery in a certain direction in a visible light standard image; the texture statistical features can be the texture features obtained by collecting and statistically analyzing the surface of the power battery.

[0087] Optionally, the camera calibration parameters can be camera parameters used to establish the mapping relationship between pixel coordinates and physical coordinates. The battery structure model can be a three-dimensional structure model of the power battery obtained by fusing geometric structure modeling and texture features.

[0088] Optionally, based on a visible light standard image, a continuous contour point sequence of the battery casing is extracted using an edge detection algorithm, the texture statistical features of the surface image are calculated using a texture analysis algorithm, and the pixel coordinates are converted into actual physical dimensions using camera calibration parameters, thereby obtaining the battery casing contour information, surface texture information, and geometric dimensions to construct a battery structure model.

[0089] Optionally, spatial registration and comparison are performed between the constructed battery structure model and a standard battery model to identify sections where the contour offset exceeds the allowable tolerance, thus obtaining contour offset regions. The actual texture features of the battery structure model are compared with the baseline texture of the standard battery model to calculate similarity, and regions with similarity below a set threshold are marked, resulting in texture anomaly regions. The actual geometric dimensions are compared with the tolerance range in the model, and dimensions exceeding the allowable deviation are identified. For contour offset regions, their deformation area and deformation vector are calculated as deformation features. For texture anomaly regions, their boundaries are identified through morphological operations, and they are classified as damage or cracks based on the regularity of the region's shape. Corresponding area and length parameters are calculated. Color space analysis is used to detect abnormal color spots different from the normal outer shell, and their location and shape characteristics are combined to determine whether they are leakage traces. The type, location, geometric parameters, and severity of the anomaly regions are integrated to generate structured physical anomaly feature information, including the location and degree of shell deformation, the area and type of damage, the length and direction of cracks, and the range and location characteristics of leakage traces. The set threshold can be a data value indicating the degree of texture pixel identification.

[0090] Specifically, based on preset edge detection algorithms and preset texture analysis algorithms, continuous contour point sequences and texture statistical features of the power battery surface are extracted from the visible light standard image. A battery structure model is constructed based on the continuous contour point sequences and texture statistical features using camera calibration parameters. Anomaly identification is performed based on the battery structure model and the standard battery model to determine physical anomaly feature information.

[0091] S240. Obtain the temperature reference of the power battery, and determine the abnormal high temperature area and abnormal low temperature area in the infrared standard image based on the temperature reference; determine the thermal anomaly feature information corresponding to the infrared standard image based on the abnormal high temperature area and abnormal low temperature area.

[0092] Optionally, the abnormal high temperature zone and the abnormal low temperature zone can be understood as the pixel area with abnormally high temperature and the pixel area with abnormally low temperature on the power battery.

[0093] Optionally, infrared thermal image data under normal operating conditions can be collected in advance, and the typical temperature distribution range of each area on the battery surface can be statistically analyzed as a temperature reference. The temperature reference includes the upper and lower temperature limits of each area of ​​the power battery under normal operating conditions.

[0094] Optionally, the measured temperature value of each pixel in the infrared standard image is compared with the upper and lower temperature limits of each region in the temperature reference during normal operation. If the temperature value of a pixel continuously exceeds the upper temperature limit of the corresponding region, the pixel is marked as a high temperature anomaly; if the temperature value of a pixel continuously falls below the lower temperature limit of the corresponding region, the pixel is marked as a low temperature anomaly. Then, the marked anomaly pixels are spatially clustered, and adjacent high temperature anomalies are aggregated into high temperature anomaly regions, and adjacent low temperature anomalies are aggregated into low temperature anomaly regions.

[0095] Optionally, for each high-temperature or low-temperature anomaly region, calculate the minimum bounding rectangle of the region, using the coordinates of the rectangle's center point as the location coordinates, and count the total number of pixels within the region, converting this to the actual physical area. This yields the actual physical area of ​​each high-temperature and low-temperature anomaly region; the output includes temperature anomaly feature information containing location coordinates, region area, highest temperature value, and lowest temperature value.

[0096] Specifically, a temperature reference for the power battery is obtained, and abnormally high temperature and abnormally low temperature regions in the infrared standard image are determined based on the temperature reference; thermal anomaly feature information corresponding to the infrared standard image is determined based on the abnormally high temperature and abnormally low temperature regions.

[0097] S250: The brightness of each pixel in the ultraviolet standard image is identified by a preset brightness threshold to determine the discharge spot, and the discharge abnormality feature information is determined based on the discharge spot.

[0098] Optionally, the preset brightness threshold can be a pre-set brightness data value used to determine abnormal discharge. It should be noted that the preset brightness threshold is compared with the brightness of each pixel in the ultraviolet standard image to identify continuous pixel areas with brightness greater than the brightness threshold, and these continuous pixel areas are used as discharge spots.

[0099] Optionally, for each discharge spot, the total number of pixels covered by the discharge spot is calculated as the spot area, and the average brightness of all pixels within it is calculated as the spot intensity, thus obtaining the regional location and spot intensity of the discharge spot; the discharge anomaly feature information is obtained by combining the regional location and spot intensity of each discharge spot.

[0100] Specifically, the brightness of each pixel in the ultraviolet standard image is identified by using a preset brightness threshold to determine the discharge spot, and the discharge anomaly feature information is determined based on the discharge spot.

[0101] S260. Extract features from the time-domain and frequency-domain signals of the standard sound signal to obtain acoustic feature vectors. Identify the acoustic feature vectors based on the pre-trained voiceprint recognition model to determine the abnormal sound characteristics.

[0102] Optionally, the voiceprint recognition model can be a pre-trained model used to extract and establish a mapping relationship between acoustic features and the voice target.

[0103] Optionally, the acoustic feature vector can be a vector data that describes the quantized features of a standard sound signal.

[0104] Optionally, based on the time-domain and frequency-domain signals, nonlinear feature transformation is performed to extract acoustic features and obtain acoustic feature vectors; based on the acoustic feature vectors, a pre-trained voiceprint recognition model is used to generate sound anomaly feature information.

[0105] Specifically, features are extracted from the time-domain and frequency-domain signals of the standard sound signal to obtain acoustic feature vectors. Based on a pre-trained voiceprint recognition model, the acoustic feature vectors are identified to determine the abnormal sound characteristics.

[0106] S270. Based on the abnormal battery characteristic information, perform a health assessment on the power battery and determine a comprehensive health assessment report.

[0107] The technical solution of this invention collects multimodal sensing data from a power battery, processes the battery image data and battery sound information to determine standard battery image data and standard battery sound signals, and effectively improves data accuracy and evaluation accuracy by processing multi-source data. It extracts abnormal battery feature information corresponding to the standard battery image data and standard battery sound signals, achieving comprehensive abnormal feature extraction of the power battery from physical structure, thermal state, insulation performance to operating sounds. Further health assessment is performed based on the extracted multi-dimensional abnormal features. A comprehensive health assessment report is generated based on the battery abnormal feature information, enabling quantitative evaluation of battery health status, effectively identifying different safety levels from early warning to high risk, and generating a health assessment report with a comprehensive health score and a clear risk level. This solves the technical problem of existing technologies being unable to comprehensively and accurately assess the health status of power batteries, significantly improving the early detection capability of potential battery faults and the comprehensive judgment level of battery health status, providing an intuitive and reliable decision-making basis for the safe operation and maintenance of batteries.

[0108] Figure 3 This is a flowchart illustrating a multimodal health assessment method for power batteries provided in this embodiment of the invention. The relationship between this embodiment and the previous embodiments is that this specifically describes the detailed process of assessing the health of a power battery using abnormal battery characteristic information. Figure 3 As shown, the method includes:

[0109] S310 collects multimodal sensing data of the power battery.

[0110] S320. Perform data processing on the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal.

[0111] S330. Extract the battery standard image data and the battery standard sound signal corresponding to the battery abnormal feature information.

[0112] S340. Calculate the physical anomaly index value of the power battery based on the physical anomaly feature information.

[0113] Optionally, when evaluating the health of a power battery, the health assessment indicators shall include at least the following: structural safety, thermal safety, insulation safety, and operational safety indicators.

[0114] Optionally, physical anomaly indicators include deformation anomaly indicators, damage anomaly indicators, crack anomaly indicators, and leakage anomaly indicators. Specifically, deformation anomaly indicators can describe the degree of deformation of the battery casing; damage anomaly indicators comprehensively describe various damaged areas of the battery; crack anomaly indicators can describe the length, width, and distribution density of various cracks in the battery; and leakage anomaly indicators can describe the leakage situation of the battery.

[0115] Optionally, for power batteries, based on the location and degree of shell deformation in the physical anomaly feature information, the deformation anomaly index value is obtained by calculating the weighted sum of the area and depth parameters of the deformed area; the damage anomaly index value is obtained by calculating the comprehensive evaluation value of the geometric features of the damaged area by measuring the damaged area and type; the crack anomaly index value is obtained by analyzing the length, width and distribution density of the cracks; and the morphological characteristics of power battery leakage traces are identified, and the leakage anomaly index value is obtained by evaluating the area, location and morphological characteristics of the leakage traces.

[0116] Specifically, the physical anomaly index value of the power battery is calculated based on the physical anomaly characteristic information.

[0117] S350. Calculate the high temperature anomaly index value and low temperature anomaly index value of the power battery based on the thermal anomaly characteristic information.

[0118] Optionally, the high temperature anomaly index value can be an index value that comprehensively describes the various abnormal high temperature zones of the power battery; the low temperature anomaly index value can be an index value that comprehensively describes the various abnormal low temperature zones of the power battery.

[0119] Optionally, based on the location coordinates, area, and highest temperature value of each abnormal high-temperature zone in the thermal anomaly feature information, the high-temperature anomaly index value of each abnormal high-temperature zone is obtained by calculating the fusion value of the temperature extreme value, area, and temperature gradient of each abnormal high-temperature zone; and based on the location coordinates, area, and highest temperature value of each abnormal low-temperature zone, the fusion value of the temperature extreme value, area, and temperature distribution of the abnormal low-temperature zone is calculated to obtain the low-temperature anomaly index value of each abnormal low-temperature zone.

[0120] Specifically, the high-temperature anomaly index value and low-temperature anomaly index value of the power battery are calculated based on thermal anomaly characteristic information.

[0121] S360. Calculate the discharge anomaly index value of the power battery based on the discharge anomaly characteristic information.

[0122] Optionally, the discharge anomaly index value can be a comprehensive index value that describes the discharge anomaly of the power battery. It should be noted that the discharge anomaly index value is calculated by comprehensively analyzing the number of discharge spots, spatial distribution density, and brightness extreme values ​​of the discharge spots based on the discharge anomaly characteristic information.

[0123] Specifically, the discharge anomaly index value of the power battery is calculated based on the discharge anomaly characteristic information.

[0124] S370. Calculate the internal gas generation abnormality index value and arc discharge abnormality index value of the power battery based on the sound abnormality feature information.

[0125] Optionally, the internal gas generation anomaly index value can be an index value that comprehensively describes the persistence and energy of the bubble noise component; the arc discharge anomaly index value can be an index value that comprehensively describes the energy and frequency of the high-frequency burst pulse component.

[0126] Optionally, based on the arc discharge sound information in the sound anomaly feature information, the arc discharge anomaly index value is obtained by calculating the energy and frequency of the high-frequency burst pulse component in the soundprint signal; and the internal gas generation anomaly index value is obtained by calculating the persistence and energy of the low-frequency bubble noise component in the soundprint signal.

[0127] S380: Identify abnormal states of physical abnormality index values, high temperature abnormality index values, low temperature abnormality index values, discharge abnormality index values, internal gas generation abnormality index values, and arc discharge abnormality index values ​​through a preset set of index safety thresholds, and determine a set of state identifiers; based on the set of state identifiers, determine the risk of the power battery and determine the safety risk level.

[0128] Optionally, the status identifier set can be a status identifier dataset obtained by identifying all different types of abnormal indicator values ​​and security thresholds.

[0129] Optionally, two safety thresholds are set for each different type of abnormal indicator value: a first-level safety threshold and a second-level safety threshold. The first-level safety threshold for each type of abnormal indicator value is used to determine whether the abnormal indicator value of that type is within the normal range; the second-level safety threshold for each type of abnormal indicator value is used to determine whether the abnormal indicator value of that type poses an abnormal risk. For an abnormal indicator value of a certain type, if the abnormal indicator value of that type does not exceed the first safety threshold, it is considered to be within a safe and normal range, and the status identifier of the abnormal indicator is set to normal. If the abnormal indicator value is greater than the first safety threshold, it is necessary to compare the abnormal indicator value of the abnormal indicator with the second safety threshold. If the abnormal indicator value of the abnormal indicator is greater than the second safety threshold, it is considered to pose an abnormal risk, and the status identifier is set to risk. If the abnormal indicator value of the abnormal indicator is not greater than the second safety threshold, it is considered to pose no abnormal risk, and an early warning is required for the abnormal indicator, and the status identifier is set to early warning. For example, taking the discharge anomaly index value as an example, if the discharge anomaly index value is not greater than the first safety threshold, the status of the discharge anomaly index value is set to normal; if the discharge anomaly index value is at the first safety threshold, the discharge anomaly index value needs to be compared with the second safety threshold. If the discharge anomaly index value is not greater than the second safety threshold, the status of the discharge anomaly index value is set to warning; if the discharge anomaly index value is greater than the second safety threshold, the status of the discharge anomaly index value is set to risk; and so on, the same judgment process is performed on physical anomaly index values, high temperature anomaly index values, low temperature anomaly index values, internal gas generation anomaly index values, and arc discharge anomaly index values.

[0130] Specifically, by using a preset set of safety threshold values, abnormal states of physical anomalies, high temperature anomalies, low temperature anomalies, discharge anomalies, internal gas generation anomalies, and arc discharge anomalies are identified to determine a set of state identifiers; based on the set of state identifiers, the power battery is risk-assessed to determine its safety risk level.

[0131] Optionally, in another optional embodiment of the present invention, the step of determining the safety risk level of the power battery based on the set of status identifiers includes:

[0132] The status identifier set is traversed. If at least one risk identifier is identified in the status identifier set, the subsequent judgment process is terminated and the security risk level is determined to be high risk. If at least one warning identifier is identified in the status identifier set when there is no risk identifier, the security risk level is determined to be medium risk. If all the status identifiers in the status identifier set are identified as normal identifiers, the security risk level is determined to be low risk.

[0133] Specifically, the status identifier set is traversed. If at least one risk identifier is identified in the status identifier set, the subsequent judgment process is terminated and the security risk level is determined to be high risk. If no risk identifier is identified in the status identifier set, and at least one warning identifier is identified in the status identifier set, the security risk level is determined to be medium risk. If all status identifiers are identified as normal identifiers, the security risk level is determined to be low risk.

[0134] S390. Conduct a health assessment of the power battery based on battery abnormality characteristic information, and determine the physical abnormality score, thermal abnormality score, discharge abnormality score, and sound abnormality score; perform weighted fusion of the physical abnormality score, thermal abnormality score, discharge abnormality score, and sound abnormality score to determine the comprehensive health score; and determine the health assessment report based on the safety risk level and the comprehensive health score.

[0135] Optionally, the physical anomaly score can be a health score that comprehensively evaluates the physical anomaly characteristics of the power battery; the thermal anomaly score can be a health score that comprehensively evaluates the thermal anomaly characteristics of the power battery; the discharge anomaly score can be a health score that comprehensively evaluates the discharge anomaly characteristics of the power battery; and the sound anomaly score can be a health score that comprehensively evaluates the sound anomaly characteristics of the power battery.

[0136] Optionally, the comprehensive health score can be a health score that combines physical abnormality scores, thermal abnormality scores, electrical discharge abnormality scores, and sound abnormality scores.

[0137] Optionally, when comprehensively scoring the physical anomaly characteristics of the power battery, battery casing deformation is used as the evaluation standard. The battery casing deformation score is pre-set to a full score, and quantitative evaluation is performed based on the deformation depth parameter and area of ​​the battery casing deformation in the physical anomaly characteristic information. Specifically, when the deformation depth parameter is lower than a first threshold and the affected area is less than a first area ratio, a deduction score is calculated based on a first deduction ratio; when the deformation depth parameter is between the first and second thresholds or the affected area reaches a second area ratio, a deduction score is calculated based on a second deduction ratio; when the deformation depth parameter exceeds the second threshold or the affected area exceeds a third area ratio, a deduction score is calculated based on a third deduction ratio. The final physical anomaly score is obtained by subtracting the average of all deduction scores from the battery casing deformation score. The affected area ratio is calculated by dividing the projected area of ​​the deformed region by the total surface area of ​​the battery casing.

[0138] Optionally, when comprehensively scoring the thermal anomaly characteristic information of the power battery, a comprehensive score is calculated based on the location coordinates, area, highest temperature value, and lowest temperature value of the high-temperature anomaly area and the low-temperature anomaly area in the thermal anomaly characteristic information. The thermal anomaly score is set as the full score. The temperature-affected area ratio is calculated by dividing the area of ​​the anomaly temperature region by the total surface area of ​​the battery, and a deduction is made based on the temperature-affected area ratio. When the temperature anomaly value is lower than the preset temperature threshold, a deduction is calculated based on the deduction ratio. When the temperature anomaly value exceeds the preset temperature threshold, a deduction is calculated based on the deduction ratio. The thermal anomaly score is obtained by subtracting all deductions from the thermal anomaly score.

[0139] Optionally, when comprehensively scoring the discharge anomaly characteristic information of the power battery, a quantitative evaluation is performed based on the number of discharge spots and the discharge intensity in the discharge anomaly characteristic information, and the discharge anomaly score is set as the full score; when the number of discharge spots is lower than a first quantity threshold and the discharge intensity is lower than a first intensity threshold, a deduction score is calculated based on the deduction ratio; the discharge spot quantity ratio is calculated by the ratio of the number of discharge spots to the standard discharge spot quantity benchmark value; when the number of discharge spots is between the first quantity threshold and the second quantity threshold or the discharge intensity is between the first intensity threshold and the second intensity threshold, a deduction score is calculated based on the deduction ratio; when the number of discharge spots exceeds the second quantity threshold or the discharge intensity exceeds the second intensity threshold, a deduction score is calculated based on the deduction ratio; the discharge anomaly score is obtained by subtracting all deduction scores from the discharge anomaly score.

[0140] Optionally, when comprehensively scoring the abnormal sound characteristics of the power battery, the frequency and intensity of arc discharge and the duration of gas production sound in the abnormal discharge characteristics are quantitatively evaluated separately. For the frequency and intensity of arc discharge, the arc discharge abnormality score is set to full. When the arc discharge frequency is lower than the first frequency threshold and the intensity is lower than the first intensity threshold, a deduction score is calculated based on the deduction ratio. The arc discharge frequency ratio is calculated by the ratio of the arc discharge frequency to the standard discharge frequency benchmark value. When the arc discharge frequency is between the first frequency threshold and the second frequency threshold or the intensity is between the first intensity threshold and the second intensity threshold, a deduction score is calculated based on the deduction ratio. When the arc discharge frequency exceeds the second frequency threshold or the intensity exceeds the second intensity threshold, a deduction score is calculated based on the deduction ratio. The arc discharge abnormality score is obtained by subtracting the deduction score from the arc discharge abnormality score.

[0141] For the internal gas production anomaly score, the maximum score is set. The gas production sound duration ratio is calculated by comparing the gas production sound duration with the standard duration baseline. If the gas production sound duration is lower than the first time threshold and the intensity is lower than the first intensity threshold, a deduction score is calculated based on the deduction ratio. If the gas production sound duration is between the first time threshold and the second time threshold, or the intensity is between the first intensity threshold and the second intensity threshold, a deduction score is calculated based on the deduction ratio. If the gas production sound duration exceeds the second time threshold or the intensity exceeds the second intensity threshold, a deduction score is calculated based on the deduction ratio. The internal gas production anomaly score is obtained by subtracting the deduction score from the internal gas production anomaly score.

[0142] Optionally, the sound abnormality score can be obtained by calculating the average of the arc discharge abnormality score and the internal gas generation abnormality score.

[0143] Optionally, a comprehensive health score for the battery can be calculated by weighted fusion based on physical anomaly scores, thermal anomaly scores, discharge anomaly scores, and sound anomaly scores; wherein the weighting coefficients of the anomaly scores are preset according to the power battery.

[0144] Optionally, the anomaly scores are multiplied by their corresponding weighting coefficients and then summed to obtain a comprehensive health score; the comprehensive health score is positively correlated with the battery's health status. Based on the safety risk level and the comprehensive health score, a battery health assessment report is generated.

[0145] The technical solution of this invention collects multimodal sensing data from a power battery, processes the battery image data and battery sound information to determine standard battery image data and standard battery sound signals, and effectively improves data accuracy and evaluation accuracy by processing multi-source data. It extracts abnormal battery feature information corresponding to the standard battery image data and standard battery sound signals, achieving comprehensive abnormal feature extraction of the power battery from physical structure, thermal state, insulation performance to operating sounds. Further health assessment is performed based on the extracted multi-dimensional abnormal features. A comprehensive health assessment report is generated based on the battery abnormal feature information, enabling quantitative evaluation of battery health status, effectively identifying different safety levels from early warning to high risk, and generating a health assessment report with a comprehensive health score and a clear risk level. This solves the technical problem of existing technologies being unable to comprehensively and accurately assess the health status of power batteries, significantly improving the early detection capability of potential battery faults and the comprehensive judgment level of battery health status, providing an intuitive and reliable decision-making basis for the safe operation and maintenance of batteries.

[0146] Figure 4 This is a schematic diagram of the structure of a multimodal health assessment device for a power battery provided in an embodiment of the present invention. Figure 4 As shown, the device includes: a data acquisition module 410, a signal processing module 420, a signal recognition module 430, and a health assessment module 440; wherein,

[0147] The data acquisition module 410 is used to acquire multimodal sensing data of the power battery, wherein the multimodal sensing data includes battery image data and battery sound signals;

[0148] Signal processing module 420 is used to process the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal;

[0149] The signal recognition module 430 is used to extract battery abnormality feature information corresponding to the battery standard image data and the battery standard sound signal; wherein, the battery abnormality feature information includes: physical abnormality feature information, thermal abnormality feature information, discharge abnormality feature information and sound abnormality feature information;

[0150] The health assessment module 440 is used to perform a health assessment on the power battery based on the abnormal battery characteristic information and to determine a comprehensive health assessment report.

[0151] The technical solution of this invention collects multimodal sensing data from a power battery, processes the battery image data and battery sound information to determine standard battery image data and standard battery sound signals, and effectively improves data accuracy and evaluation accuracy by processing multi-source data. It extracts abnormal battery feature information corresponding to the standard battery image data and standard battery sound signals, achieving comprehensive abnormal feature extraction of the power battery from physical structure, thermal state, insulation performance to operating sounds. Further health assessment is performed based on the extracted multi-dimensional abnormal features. A comprehensive health assessment report is generated based on the battery abnormal feature information, enabling quantitative evaluation of battery health status, effectively identifying different safety levels from early warning to high risk, and generating a health assessment report with a comprehensive health score and a clear risk level. This solves the technical problem of existing technologies being unable to comprehensively and accurately assess the health status of power batteries, significantly improving the early detection capability of potential battery faults and the comprehensive judgment level of battery health status, providing an intuitive and reliable decision-making basis for the safe operation and maintenance of batteries.

[0152] Optionally, the signal processing module 420 is specifically used for: the battery image data including a first visible light image, a first infrared image, and a first ultraviolet image;

[0153] The step of processing the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal includes:

[0154] Perform image processing on the first visible light image to determine a visible light standard image;

[0155] The first infrared image is processed to determine the infrared standard image;

[0156] Edge features are extracted from the visible light standard image and the first ultraviolet image respectively to determine the first edge pixel set and the second edge pixel set;

[0157] Pixel matching is performed on the first edge pixel set and the second edge pixel set to determine the coordinate transformation relationship, and the first ultraviolet image is resampled based on the coordinate transformation relationship to determine the ultraviolet standard image;

[0158] The battery sound information is processed to determine the standard sound signal of the battery.

[0159] Optionally, the signal processing module 420 is further configured to: for each pixel of the first visible light image, calculate the weighted average of all pixels in the neighborhood window of the pixel based on the weighted average of neighboring pixels and a square filter window, and replace the pixel with the weighted average to obtain the second visible light image;

[0160] The contrast of the second visible light image is enhanced based on the brightness distribution histogram of the second visible light image to obtain a third visible light image;

[0161] Identify the apex corner pixel coordinates of the power battery in the third visible light image, and calculate perspective transformation parameters based on the apex corner pixel coordinates;

[0162] The third visible light image is resampled based on the perspective transformation parameters to obtain the visible light standard image.

[0163] Optionally, the signal processing module 420 is further configured to: acquire the original infrared digital value matrix, and convert the original digital values ​​of the original infrared digital value matrix into a first radiation intensity value;

[0164] The ambient temperature of the ambient temperature sensor is obtained, and background noise compensation is performed on each of the first radiation intensity values ​​based on the ambient temperature to determine the second radiation intensity value.

[0165] For each of the second radiation intensity values, the surface emissivity parameter corresponding to the battery casing material of the power battery is obtained, and the second radiation intensity value is converted into a measured temperature value based on the surface emissivity parameter;

[0166] The measured temperature value is mapped to the first infrared image to determine the infrared standard image.

[0167] Optionally, the signal recognition module 430 is specifically used to: extract the continuous contour point sequence of the power battery and the texture statistical features of the power battery surface in the visible light standard image based on a preset edge detection algorithm and a preset texture analysis algorithm, respectively;

[0168] A battery structure model is constructed based on the continuous contour point sequence and the texture statistical features using camera calibration parameters. Anomaly identification is performed based on the battery structure model and a standard battery model to determine the physical anomaly feature information.

[0169] Obtain the temperature reference of the power battery, and determine the abnormal high temperature region and abnormal low temperature region in the infrared standard image based on the temperature reference; determine the thermal anomaly feature information corresponding to the infrared standard image based on the abnormal high temperature region and abnormal low temperature region;

[0170] The brightness of each pixel in the ultraviolet standard image is identified by using a preset brightness threshold to determine the discharge spot, and the discharge abnormality feature information is determined based on the discharge spot.

[0171] The time-domain and frequency-domain signals of the standard sound signal are extracted to obtain acoustic feature vectors. The acoustic feature vectors are then identified based on a pre-trained voiceprint recognition model to determine abnormal sound characteristics.

[0172] Optionally, the health assessment module 440 is specifically used to: calculate the physical abnormality index value of the power battery based on the physical abnormality feature information; wherein, the physical abnormality index value includes deformation abnormality index value, damage abnormality index value, crack abnormality index value and leakage abnormality index value.

[0173] The high-temperature anomaly index value and low-temperature anomaly index value of the power battery are calculated based on the thermal anomaly characteristic information.

[0174] The discharge anomaly index value of the power battery is calculated based on the discharge anomaly characteristic information.

[0175] Based on the sound abnormality feature information, calculate the internal gas generation abnormality index value and the arc discharge abnormality index value of the power battery.

[0176] The abnormal status of the physical anomaly index value, the high temperature anomaly index value, the low temperature anomaly index value, the discharge anomaly index value, the internal gas generation anomaly index value, and the arc discharge anomaly index value is identified by a preset set of index safety thresholds, and a set of status identifiers is determined.

[0177] Based on the set of status identifiers, the power battery is risk-assessed to determine its safety risk level.

[0178] A health assessment of the power battery is performed based on the battery abnormality feature information to determine physical abnormality scores, thermal abnormality scores, discharge abnormality scores, and sound abnormality scores.

[0179] The comprehensive health score is determined by weighted fusion of the physical anomaly score, the thermal anomaly score, the discharge anomaly score, and the sound anomaly score.

[0180] The health assessment report is determined based on the safety risk level and the comprehensive health score.

[0181] Optionally, the health assessment module 440 is further configured to: traverse the set of status identifiers, and if at least one risk identifier is identified in the set of status identifiers, terminate the subsequent judgment process and determine the safety risk level as high risk.

[0182] If no risk identifier is found in the set of status identifiers, but at least one warning identifier is found in the set of status identifiers, then the safety risk level is determined to be medium risk.

[0183] If all the status identifiers are identified as normal, the security risk level is determined to be low risk.

[0184] The power battery multimodal health assessment device provided in this embodiment of the invention can execute the power battery multimodal health assessment method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0185] Figure 5 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their patterns are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0186] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0187] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of monitors, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer grids such as the Internet and / or various telecommunications grids.

[0188] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the multimodal health assessment method for power batteries.

[0189] In some embodiments, the power battery multimodal health assessment method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the power battery multimodal health assessment method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the power battery multimodal health assessment method by any other suitable means (e.g., by means of firmware).

[0190] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0191] Computer programs used to implement the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the patterns / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0192] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0193] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0194] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or grid browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication grid). Examples of communication grids include local area networks (LANs), wide area networks (WANs), blockchain grids, and the Internet.

[0195] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS servers, such as high management difficulty and weak business scalability.

[0196] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0197] This embodiment provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the program implements the steps of the power battery multimodal health assessment method provided in any embodiment of the present invention. The method includes:

[0198] Collect multimodal sensing data of the power battery, wherein the multimodal sensing data includes battery image data and battery sound signals;

[0199] The battery image data and the battery sound information are processed separately to determine the battery standard image data and the battery standard sound signal;

[0200] Extract the battery abnormality feature information corresponding to the battery standard image data and the battery standard sound signal; wherein, the battery abnormality feature information includes: physical abnormality feature information, thermal abnormality feature information, discharge abnormality feature information and sound abnormality feature information;

[0201] A health assessment of the power battery is performed based on the abnormal battery characteristics information, and a comprehensive health assessment report is generated.

[0202] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0203] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0204] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0205] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of mesh, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0206] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a grid of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computing device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0207] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0208] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A multimodal health assessment method for power batteries, characterized in that, include: Collect multimodal sensing data of the power battery, wherein the multimodal sensing data includes battery image data and battery sound signals; The battery image data and the battery sound information are processed separately to determine the battery standard image data and the battery standard sound signal; Extract the battery abnormality feature information corresponding to the battery standard image data and the battery standard sound signal; wherein, the battery abnormality feature information includes: physical abnormality feature information, thermal abnormality feature information, discharge abnormality feature information and sound abnormality feature information; A health assessment of the power battery is performed based on the abnormal battery characteristics information, and a comprehensive health assessment report is generated.

2. The method according to claim 1, characterized in that, The battery image data includes a first visible light image, a first infrared image, and a first ultraviolet image; The step of processing the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal includes: Perform image processing on the first visible light image to determine a visible light standard image; The first infrared image is processed to determine the infrared standard image; Edge features are extracted from the visible light standard image and the first ultraviolet image respectively to determine the first edge pixel set and the second edge pixel set; Pixel matching is performed on the first edge pixel set and the second edge pixel set to determine the coordinate transformation relationship, and the first ultraviolet image is resampled based on the coordinate transformation relationship to determine the ultraviolet standard image; The battery sound information is processed to determine the standard sound signal of the battery.

3. The method according to claim 2, characterized in that, The step of processing the first visible light image to determine a visible light standard image includes: For each pixel in the first visible light image, the weighted average of all pixels within the neighborhood window of the pixel is calculated based on the weighted average of neighboring pixels and a square filter window. The pixel is then replaced with the weighted average to obtain the second visible light image. The contrast of the second visible light image is enhanced based on the brightness distribution histogram of the second visible light image to obtain a third visible light image; Identify the apex corner pixel coordinates of the power battery in the third visible light image, and calculate perspective transformation parameters based on the apex corner pixel coordinates; The third visible light image is resampled based on the perspective transformation parameters to obtain the visible light standard image.

4. The method according to claim 2, characterized in that, The step of processing the first infrared image to determine the infrared standard image includes: Obtain the original infrared digital value matrix, and convert the original digital values ​​of the original infrared digital value matrix into a first radiation intensity value; The ambient temperature of the ambient temperature sensor is obtained, and background noise compensation is performed on each of the first radiation intensity values ​​based on the ambient temperature to determine the second radiation intensity value. For each of the second radiation intensity values, the surface emissivity parameter corresponding to the battery casing material of the power battery is obtained, and the second radiation intensity value is converted into a measured temperature value based on the surface emissivity parameter; The measured temperature value is mapped to the first infrared image to determine the infrared standard image.

5. The method according to claim 2, characterized in that, The extraction of battery anomaly feature information corresponding to the battery standard image data and the battery standard sound signal includes: The continuous contour point sequence of the power battery and the texture statistical features of the power battery surface in the visible light standard image are extracted based on the preset edge detection algorithm and the preset texture analysis algorithm, respectively. A battery structure model is constructed based on the continuous contour point sequence and the texture statistical features using camera calibration parameters. Anomaly identification is performed based on the battery structure model and a standard battery model to determine the physical anomaly feature information. Obtain the temperature reference of the power battery, and determine the abnormal high temperature region and abnormal low temperature region in the infrared standard image based on the temperature reference; determine the thermal anomaly feature information corresponding to the infrared standard image based on the abnormal high temperature region and abnormal low temperature region; The brightness of each pixel in the ultraviolet standard image is identified by using a preset brightness threshold to determine the discharge spot, and the discharge abnormality feature information is determined based on the discharge spot. The time-domain and frequency-domain signals of the standard sound signal are extracted to obtain acoustic feature vectors. The acoustic feature vectors are then identified based on a pre-trained voiceprint recognition model to determine abnormal sound characteristics.

6. The method according to claim 1, characterized in that, The step of performing a health assessment on the power battery based on the abnormal battery characteristic information and determining a comprehensive health assessment report includes: The physical anomaly index values ​​of the power battery are calculated based on the physical anomaly feature information; wherein, the physical anomaly index values ​​include deformation anomaly index values, damage anomaly index values, crack anomaly index values ​​and leakage anomaly index values. The high-temperature anomaly index value and low-temperature anomaly index value of the power battery are calculated based on the thermal anomaly characteristic information. The discharge anomaly index value of the power battery is calculated based on the discharge anomaly characteristic information. Based on the sound abnormality feature information, calculate the internal gas generation abnormality index value and the arc discharge abnormality index value of the power battery. The abnormal status of the physical anomaly index value, the high temperature anomaly index value, the low temperature anomaly index value, the discharge anomaly index value, the internal gas generation anomaly index value, and the arc discharge anomaly index value is identified by a preset set of index safety thresholds, and a set of status identifiers is determined. Based on the set of status identifiers, the power battery is risk-assessed to determine its safety risk level. A health assessment of the power battery is performed based on the battery abnormality feature information to determine physical abnormality scores, thermal abnormality scores, discharge abnormality scores, and sound abnormality scores. The comprehensive health score is determined by weighted fusion of the physical anomaly score, the thermal anomaly score, the discharge anomaly score, and the sound anomaly score. The health assessment report is determined based on the safety risk level and the comprehensive health score.

7. The method according to claim 6, characterized in that, The step of determining the safety risk level of the power battery based on the set of status identifiers includes: The status identifier set is traversed. If at least one risk identifier is found in the status identifier set, the subsequent judgment process is terminated and the security risk level is determined to be high risk. If no risk identifier is found in the set of status identifiers, but at least one warning identifier is found in the set of status identifiers, then the safety risk level is determined to be medium risk. If all the status identifiers are identified as normal, the security risk level is determined to be low risk.

8. A multimodal health assessment device for power batteries, characterized in that, include: The data acquisition module is used to acquire multimodal sensing data of the power battery, wherein the multimodal sensing data includes battery image data and battery sound signals; The signal processing module is used to process the battery image data and the battery sound information respectively to determine the battery standard image data and the battery standard sound signal; The signal recognition module is used to extract battery abnormality feature information corresponding to the battery standard image data and the battery standard sound signal; wherein, the battery abnormality feature information includes: physical abnormality feature information, thermal abnormality feature information, discharge abnormality feature information and sound abnormality feature information; The health assessment module is used to perform a health assessment on the power battery based on the abnormal battery characteristic information and to determine a comprehensive health assessment report.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the power battery multimodal health assessment method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the power battery multimodal health assessment method according to any one of claims 1-7.

Images