A battery internal temperature estimation method and device based on ultrasonic timing characteristics

Abstract

The application discloses a battery internal temperature estimation method and device based on ultrasonic timing characteristics. The method comprises the following steps: acquiring an ultrasonic transmission waveform after an ultrasonic pulse penetrates a target battery; extracting multi-dimensional acoustic features from the ultrasonic transmission waveform frame by frame; constructing a feature vector based on the multi-dimensional acoustic features and the battery state of charge at the same time; constructing a timing input matrix based on a sliding time window and the feature vectors in a continuous preset number of sampling periods; inputting the timing input matrix into a pre-trained long short-term memory network model to obtain an internal temperature estimation value of the target battery output by the long short-term memory network model. In the application, the slow accumulation process of battery thermal inertia can be accurately described, the network can complete the nonlinear decoupling of the internal temperature and the SOC on the acoustic response, and the dependence on static linear compensation or empirical decoupling formula is eliminated, and the reliability is significantly better under the complex vehicle working conditions with frequent SOC fluctuations.

Description

Technical Field

[0001] The embodiments in this specification pertain to the field of battery temperature estimation, and specifically relate to a method and apparatus for estimating the internal temperature of a battery based on ultrasonic timing characteristics. Background Technology

[0002] Lithium-ion batteries, as the core power source for electric vehicles and energy storage systems, rely heavily on accurate sensing of their internal temperature for safety and thermal management efficiency. Due to the dense internal winding structure and poor thermal conductivity of the battery, the internal temperature is often much higher than the surface temperature, exhibiting a significant thermal hysteresis effect. Current mainstream methods for obtaining battery internal temperature all have limitations: implanted sensors, while able to directly measure the core temperature, require disrupting the sealed casing, introducing risks of leakage and short circuits; while the inverse estimation method based on a surface thermistor coupled with an electrothermal model is severely limited by the battery's thermal time constant. Under dynamic and aggressive charge and discharge conditions, the surface temperature response lags significantly behind the actual internal temperature rise, and the model estimation accuracy is highly dependent on empirical parameters such as thermal resistance and heat generation coefficient. These parameters drift with battery aging, leading to continuous divergence in estimation errors. Furthermore, existing technologies include direct ultrasonic detection, but current ultrasonic temperature measurement is mostly based on linear fitting of a single-frame snapshot or piecewise static mapping, lacking the time memory capability to describe the slowly changing thermal inertia process, yet still meeting the reliability requirements of real-time vehicle thermal management. Summary of the Invention

[0003] The embodiments of this disclosure provide a method and apparatus for estimating the internal temperature of a battery based on ultrasonic timing characteristics, aiming to solve one or more of the above-mentioned problems and other potential problems.

[0004] According to a first aspect of this disclosure, a method for estimating the internal temperature of a battery based on ultrasonic time-series features is provided. The method includes acquiring the ultrasonic transmission waveform after an ultrasonic pulse penetrates a target battery, extracting multidimensional acoustic features frame by frame from the ultrasonic transmission waveform, and constructing a feature vector based on the multidimensional acoustic features and the battery state of charge at the same time. The multidimensional acoustic features include flight time and signal amplitude.

[0005] Based on a sliding time window, a temporal input matrix is ​​constructed according to the feature vectors within a consecutive preset number of sampling periods;

[0006] The timing input matrix is ​​fed into a pre-trained Long Short-Term Memory (LSTM) network model to obtain the internal temperature estimate of the target battery output by the LSM network model. The LSM network model is used to nonlinearly decouple the internal temperature and the battery state of charge in the time dimension through a timing gating mechanism.

[0007] According to a second aspect of this disclosure, a battery internal temperature estimation device based on ultrasonic time-series features is provided. The device includes a feature vector construction module configured to acquire the ultrasonic transmission waveform after an ultrasonic pulse penetrates a target battery, extract multidimensional acoustic features frame by frame from the ultrasonic transmission waveform, and construct a feature vector based on the multidimensional acoustic features and the battery state of charge at the same time. The multidimensional acoustic features include flight time and signal amplitude.

[0008] The input matrix construction module is configured to construct a temporal input matrix based on a sliding time window and the feature vectors within a consecutive preset number of sampling periods.

[0009] The temperature estimation module is configured to input the time-series input matrix into a pre-trained long short-term memory network model to obtain the internal temperature estimate of the target battery output by the long short-term memory network model. The long short-term memory network model is used to nonlinearly decouple the internal temperature and the battery state of charge in the time dimension through a time-series gating mechanism.

[0010] According to a third aspect of this disclosure, an electronic device is provided, including one or more processors and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform a method provided according to a first scheme.

[0011] According to a fourth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to the first aspect.

[0012] The solution provided in the embodiments of this specification can reflect the overall changes of the multilayered media inside the battery through ultrasonic transmission waveforms, providing internal temperature information without damaging the seal, thus compensating for the blind spot of surface NTC which can only measure boundary temperatures. Simultaneously, it incorporates the state of charge (SOC) as a known cooperative input into the multidimensional feature space and introduces a gating timing framework of a sliding time window and a long short-term memory (LSTM) network. Through forget gates, input gates, and output gates, it selectively memorizes and forgets the timing input matrix for multiple consecutive sampling periods, thereby accurately characterizing the gradual accumulation process of battery thermal inertia. This allows the network to achieve nonlinear decoupling between internal temperature and SOC in the acoustic response, eliminating reliance on static linear compensation or empirical decoupling formulas, and exhibiting significantly better reliability under complex automotive operating conditions with frequent SOC fluctuations. Attached Figure Description

[0013] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0014] Figure 1A flowchart illustrating a battery internal temperature estimation method based on ultrasonic timing features, representing some embodiments of this disclosure, is shown.

[0015] Figure 2 A schematic diagram showing the ultrasonic comparison at different temperatures under 100% SOC of some embodiments of the present disclosure is provided.

[0016] Figure 3 The following are schematic diagrams illustrating the LSTM operation process of some embodiments of this disclosure;

[0017] Figure 4 A schematic diagram of the structure of a battery internal temperature estimation device based on ultrasonic timing characteristics according to some embodiments of the present disclosure is shown;

[0018] Figure 5 A schematic block diagram of an electronic device according to some embodiments of the present disclosure is shown. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0020] The terms “comprising” and “having”, and any variations thereof, in this specification, claims, and the foregoing drawings are intended to cover a non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. Depending on the context, the word “if” as it applies herein may be interpreted as “when”, “in response to determination”, or “in response to detection”.

[0021] The elastic modulus of lithium-ion battery anode materials is also affected by temperature, which is the basis for using ultrasonic testing technology to estimate the internal temperature of the battery. The relationship is as follows:

[0022]

[0023] In the formula, The elastic modulus at that temperature. The elastic modulus of the negative electrode material at absolute zero. The temperature coefficient of elastic modulus. The internal temperature of the battery. The speed and impedance of ultrasound waves are both affected by the elastic modulus; therefore, changes in battery temperature will affect the propagation of ultrasound waves.

[0024] Different ultrasonic frequencies have varying impacts on detection results. Low-frequency ultrasound has a longer wavelength and stronger penetrating power, allowing it to penetrate thicker materials, but it may easily miss subtle changes inside the battery. High-frequency ultrasound has a shorter wavelength and higher resolution, making it more sensitive to details inside the battery, but its penetrating power is weaker. Considering the influence of these two factors, this application preferably uses a pair of piezoelectric probes with a frequency of 2.5MHz for verification in subsequent processing. Furthermore, to ensure the integrity of the ultrasonic signal recording during the experiment, the oscilloscope sampling frequency is preferably set to 20MHz.

[0025] Figure 1 A flowchart illustrating a battery internal temperature estimation method based on ultrasonic timing features, representing some embodiments of this disclosure, is shown. Method 100 can be executed, for example, by a terminal, which may include, but is not limited to, a mobile phone, tablet computer, desktop computer, server, etc. Figure 1 As shown, in method 100, step 102 can obtain the ultrasonic transmission waveform after the ultrasonic pulse penetrates the target battery, extract multidimensional acoustic features frame by frame from the ultrasonic transmission waveform, and construct a feature vector based on the multidimensional acoustic features and battery state of charge at the same time. The multidimensional acoustic features include flight time and signal amplitude.

[0026] In this embodiment, a single-transmitter, single-receiver ultrasonic transmission configuration can be used. The ultrasonic transmitting probe and receiving probe are respectively positioned on opposite sides of the target battery. The transmitting probe emits ultrasonic pulses into the target battery. The ultrasonic pulses penetrate the multi-layered winding structure inside the battery along its thickness direction and reach the receiving probe, which then receives the transmitted ultrasonic waveform. For each frame of the received ultrasonic transmission waveform, multi-dimensional acoustic features are extracted. These features include at least the time of flight (ToF) and signal amplitude (SA). The time of flight is the total propagation time of the ultrasonic pulse from the transmitting probe to the receiving probe receiving the transmitted waveform; the signal amplitude is the peak amplitude or effective amplitude of the received waveform. Since the propagation speed and energy attenuation of ultrasonic waves inside the battery are affected by the elastic modulus of the internal battery materials, and the elastic modulus changes with temperature, ToF and SA can reflect the temperature changes inside the battery. Simultaneously, the state of charge (SOC) of the battery at various times is also acquired. The SOC can be provided in real-time by the battery management system or estimated using conventional methods such as voltage and current integration. Thus, based on the multidimensional acoustic characteristics and battery state of charge at the same time, a feature vector can be constructed that simultaneously contains both the battery's acoustic response information and its state of charge information. As an example, the feature vector at time k could be... .

[0027] In method 100, step 104 can be based on a sliding time window to construct a time-series input matrix according to the feature vectors within a consecutive preset number of sampling periods.

[0028] In this embodiment, a time-series input matrix is ​​constructed from feature vectors within a consecutive preset number of sampling periods using a sliding time window. Where L is the window length (i.e., the number of time steps) and d is the feature dimension (d=3 in this embodiment). The sliding time window setting allows the input data in the Long Short-Term Memory Network model to include not only the information of the current moment, but also the historical information of the past moments, which facilitates the construction of the time series input matrix and allows the model to better understand the temporal changes of the data, thereby making more accurate predictions.

[0029] In method 100, step 106 can input the time-series input matrix into the pre-trained long short-term memory network model to obtain the internal temperature estimate of the target battery output by the long short-term memory network model. The long short-term memory network model is used to nonlinearly decouple the internal temperature and the battery state of charge in the time dimension through a time-series gating mechanism.

[0030] In this embodiment, a Long Short-Term Memory (LSTM) network model can be pre-trained. During training, the LSM model uses time-of-flight, signal amplitude, and state of charge (STC) as input features and measured internal temperature as a supervisory signal. A time-series gating mechanism is used to achieve nonlinear decoupling of internal temperature and STC in the time dimension. Specifically, the LSM model's ability to remember historical sequences can be utilized to automatically learn the temporal evolution patterns of internal temperature and STC during training. For example, changes in STC are usually accompanied by rapid fluctuations during charging and discharging, while temperature changes are influenced by thermal inertia and exhibit a slow, cumulative characteristic, thus separating the two in the time dimension. Inputting the time-series input matrix into the pre-trained LSM model yields the model's predicted internal temperature estimate, achieving an effective and accurate estimation of the internal temperature.

[0031] In one possible implementation, the window length of the sliding time window is determined based on the battery's thermal time constant, such that the number of timing frames contained in a single sliding time window covers the thermal inertial response cycle of battery temperature changes.

[0032] In this embodiment, the thermal time constant of the battery characterizes the speed at which the battery's internal temperature responds to thermal excitation, and its value depends on factors such as the battery's material, size, and heat dissipation conditions. The window length L of the sliding time window is determined based on the battery's thermal time constant, ensuring that the number of timing frames contained within a single window is sufficient to cover the thermal inertia response period of the battery's temperature change. As an example, the selection rule for the window length could be: ,in, The sampling period is Let be the thermal time constant of the battery. When the window length meets the above conditions, the LSTM network can observe the complete evolution of the battery temperature from being thermally stimulated to generating a response and then tending to stabilize within a window, thereby accurately capturing the slow-varying accumulation characteristics of temperature changes.

[0033] In one possible implementation, multidimensional acoustic features are extracted frame by frame from the ultrasonic transmission waveform, including:

[0034] The arrival time window of the main wave of the ultrasonic signal is determined based on the propagation speed of ultrasonic waves in the medium inside the battery and the thickness of the battery. Multidimensional acoustic features are extracted frame by frame from the ultrasonic transmission waveform within the arrival time window of the main wave.

[0035] In this embodiment, because ultrasonic waves undergo multiple reflections between the multi-layer interfaces inside the battery, the received waveform may contain subsequent reflected waves and clutter in addition to the main wave. These reflected waves and clutter do not carry effective information directly related to the battery's internal temperature and instead interfere with the accurate extraction of Time-of-Flight (ToF) and Automatic Sensing (SA). Therefore, the arrival time window of the main wave of the ultrasonic signal can be determined based on the propagation speed of the ultrasonic wave in the battery's internal medium and the battery thickness. Only the waveform data within this time window is extracted for subsequent feature extraction, while noise and reflected clutter outside the window are discarded. As an example, assuming the battery thickness is d and the typical propagation speed of ultrasonic waves in the battery material is v, the theoretical time required for the ultrasonic wave to penetrate the battery is approximately... The arrival time window of the main wave will be set accordingly. In other possible implementations, a fixed time can also be set directly based on experience, for example... .

[0036] like Figure 2 The image shows a comparison of ultrasonic waves at 100% SOC: -10℃, 20℃, and 50℃. As temperature increases, the Time-of-Flight (ToF) gradually increases, while the Speed-of-Flight (SA) gradually decreases. This is because the thermal expansion and contraction of the battery causes changes in the ultrasonic wave characteristics. At lower temperatures, the battery volume decreases, resulting in a shorter propagation path for the ultrasonic waves. Assuming the propagation speed of the ultrasonic waves remains constant, the propagation path lengthens with increasing temperature, leading to an increase in ToF. With the lengthening of the propagation path, the attenuation of ultrasonic energy also increases, resulting in a decrease in SA.

[0037] In one possible implementation, the start time of the arrival time window of the main wave is determined based on the emission time and the shortest propagation path of the ultrasonic wave inside the battery, and the window width of the arrival time window of the main wave is determined based on the battery thickness and the attenuation characteristics of the ultrasonic wave in the battery material.

[0038] In this embodiment, the width of the arrival time window of the main wave is not a fixed value set empirically, but is determined jointly by the battery thickness and the attenuation characteristics of the ultrasound in the battery material. This ensures that the window, while fully accommodating the main energy of the main wave waveform, strictly terminates on the time axis before the arrival of the first reflected wave, achieving single main wave interception. The propagation path of the ultrasound is a straight line along the battery thickness from the transmitting probe to the receiving probe, which is the shortest propagation path. The determination of the window width needs to comprehensively consider the duration of the main wave itself (which depends on the width of the transmitted pulse and the attenuation characteristics of the battery) and the time interval between adjacent reflected waves and the main wave (this interval is approximately equal to the time it takes for the ultrasound to travel one round trip inside the battery). To ensure that only one main wave is included in the window and subsequent reflected waves are excluded, the window width W should satisfy W < 2d / v, and W should be sufficient to cover the main energy concentration area of ​​the main wave. In addition, an attenuation coefficient can be determined by looking up a table based on the attenuation characteristics of the ultrasound in the battery material. The window width can be compensated according to the attenuation coefficient to achieve adaptive optimization of the window parameters, ensuring that a pure single main wave effective signal is always intercepted.

[0039] In one possible implementation, the method further includes:

[0040] In a temperature chamber environment, a stepped temperature rise operation was performed with different preset battery charge states. The ultrasonic transmission waveform and the actual internal temperature value were recorded after each temperature step reached thermal equilibrium. A training set was constructed, and a long short-term memory network model was trained based on the training set.

[0041] In this embodiment, considering that the current application of ultrasonic nondestructive testing in lithium-ion batteries mainly revolves around state of charge (SOC) estimation, state of health (SOH) diagnosis, and internal defect identification, research on directly using ultrasonic signals for quantitative estimation of battery internal temperature is relatively limited, and ultrasonic feature observations exhibit strong coupling relationships. Existing ultrasonic temperature measurement methods often degenerate into static mapping or segmented calibration of single-dimensional features, failing to consider the state of charge (SOC) and internal temperature. The coupling relationship is considered. Therefore, this application employs a Long Short-Term Memory (LSTM) network model to enable the prediction results to comprehensively consider ultrasonic features and state of charge. Specifically, the target battery can be placed in a temperature chamber and uniformly heated according to a set temperature step, maintaining a sufficient duration after each temperature increase to ensure that the internal temperature of the battery reaches thermal equilibrium. Initially, the battery's state of charge can be, for example, 0%, and the temperature chamber temperature is -10°C. The temperature can be increased in 10°C increments every 90 minutes until it reaches 50°C, during which time the ultrasonic signal is recorded every 30 seconds. Then, the battery's SOC is sequentially changed to 20%, 40%, 60%, 80%, and 100%, and the above-mentioned step-by-step temperature increase process is repeated at each SOC. In this way, characteristic data (including flight time and signal amplitude) of ultrasonic transmission waveforms under different battery states of charge and different temperature steps, as well as the true internal temperature value, can be collected, and a training set can be constructed based on this. Within the training set, time of flight, signal amplitude, and state of charge will serve as input samples, while the actual internal temperature will be used as the output sample. In addition to the training set, one-third of the data can be reserved as a validation set. The LSTM model trained using the training set in this application can contain one input layer, one hidden layer, and one output layer. The input layer contains three units: SOC, SA, and TOF; the hidden layer contains 100 hidden nodes; and the output layer contains one unit, representing the battery's internal temperature.

[0042] Long Short-Term Memory (LSTM) neural networks are a special type of recurrent neural network primarily used for processing and predicting sequential data. Lithium-ion battery state of charge (SOC) and temperature are both significantly correlated with ultrasonic characteristics, and changes in both battery temperature and ultrasonic waves are typical time-series problems, closely related to battery behavior over a past period. LSTM can effectively capture this time dependence, thereby estimating the internal temperature of the battery. Figure 3 As shown, each LSTM unit controls the output state of information through a forget gate, an input gate, and an output gate. The forget gate determines the previous memory state. How much information should be retained? The sigmoid function outputs a vector with values ​​between 0 and 1 to control how much information is retained. The calculation formula is as follows:

[0043]

[0044] In the formula, The input at time k, This is the output at time k-1. , Here is the forget gate weight matrix. Let be the bias matrix of the forget gate, and sigmoid be the activation function. The output of the forget gate at time k.

[0045] This application uses ultrasonic features (SA), Time-of-Flight (ToF), and the battery's State of Charge (SOC) as inputs, along with the battery's internal temperature. As output, we get:

[0046]

[0047] The input gate determines the current input. What impact will this have on memory state? The sigmoid function determines how much information enters, and the tanh function generates a new candidate memory. .

[0048]

[0049] In the formula, Here is the weight matrix of the input gate. Let be the bias matrix of the input gate, and tanh be the activation function. The output of the input gate at time k. This is a candidate memory for the current input.

[0050] After the input gate completes its output, calculate the cell state at that moment. .

[0051]

[0052] In the formula, Let k be the cell state at time k.

[0053] The output gate is responsible for controlling the state of the control unit. How much output is given to the current output value of the LSTM? .

[0054]

[0055] In the formula, Here is the forget gate weight matrix. Let be the bias matrix of the forget gate. The output of the gate at time k. This is the final output of the LSTM.

[0056] In one possible implementation, the holding time of the temperature step is determined based on the thermal time constant of the target battery and the response delay of the temperature sensor.

[0057] In this embodiment, the holding time of the temperature gradient can be determined based on the thermal time constant of the target battery and the response delay of the temperature sensor. As an example, the holding time can, for instance, meet the following requirements: ,in The thermal time constant, This includes the response delay of the temperature sensor. This ensures that the holding time is sufficient for the battery's internal temperature to approach the chamber's set temperature, while also guaranteeing that the temperature measurement accurately reflects the battery's internal temperature due to the added temperature sensor response delay. For example, for a battery with a thermal time constant of approximately 20 minutes, the holding time can be set to 60 to 90 minutes.

[0058] Figure 4 Schematic diagrams of a battery internal temperature estimation device based on ultrasonic timing characteristics according to some embodiments of this disclosure are shown. The various embodiments in this specification are described in a progressive manner; similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments. Figure 4 As shown, the device 400 includes a feature vector construction module 401, which is configured to acquire the ultrasonic transmission waveform after the ultrasonic pulse penetrates the target battery, extract multi-dimensional acoustic features frame by frame from the ultrasonic transmission waveform, and construct a feature vector based on the multi-dimensional acoustic features and the battery state of charge at the same time. The multi-dimensional acoustic features include flight time and signal amplitude.

[0059] The input matrix construction module 402 is configured to construct a temporal input matrix based on the feature vectors within a consecutive preset number of sampling periods using a sliding time window.

[0060] The temperature estimation module 403 is configured to input the timing input matrix into a pre-trained long short-term memory network model to obtain the internal temperature estimate of the target battery output by the long short-term memory network model. The long short-term memory network model is used to nonlinearly decouple the internal temperature and the battery state of charge in the time dimension through a timing gating mechanism.

[0061] In one possible implementation, the flight time is the total propagation time of the ultrasonic pulse from the transmitting probe to the receiving probe receiving the ultrasonic transmitted waveform; the signal amplitude is the peak amplitude or effective amplitude of the ultrasonic transmitted waveform.

[0062] In one possible implementation, the window length of the sliding time window is determined based on the battery's thermal time constant, such that the number of timing frames contained in a single sliding time window covers the thermal inertial response period of battery temperature changes.

[0063] In one possible implementation, the feature vector construction module 401 is further configured to determine the arrival time window of the main wave of the ultrasonic signal based on the propagation speed of the ultrasonic wave in the medium inside the battery and the thickness of the battery, and extract multidimensional acoustic features frame by frame from the ultrasonic transmission waveform within the arrival time window of the main wave.

[0064] In one possible implementation, the start time of the arrival time window of the main wave is determined based on the emission time and the shortest propagation path of the ultrasonic wave inside the battery, and the window width of the arrival time window of the main wave is determined based on the battery thickness and the attenuation characteristics of the ultrasonic wave in the battery material.

[0065] In one possible implementation, the device further includes a training module configured in a temperature chamber environment to perform stepped temperature rise operations with different preset battery states of charge, record the ultrasonic transmission waveform and the actual internal temperature value after each temperature step reaches thermal equilibrium, construct a training set, and train a long short-term memory network model based on the training set.

[0066] In one possible implementation, the holding time of the temperature step is determined based on the thermal time constant of the target battery and the response delay of the temperature sensor.

[0067] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this specification is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0068] Figure 5 A block diagram of an electronic device 500 that can implement various embodiments of the present disclosure is shown. For example... Figure 5As shown, the electronic device 500 includes a processor 510, a disk drive 520, an input / output interface 530, a network interface 540, and a memory 550. The processor 510, disk drive 520, input / output interface 530, network interface 540, and memory 550 can communicate with each other via a communication bus.

[0069] The processor 510 can be implemented using a general-purpose CPU, microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits to execute relevant programs and implement the technical solution provided in this application.

[0070] The memory 550 can be implemented in the form of ROM (Read Only Memory), RAM (Read Access Memory), static memory, dynamic storage devices, etc. The memory 550 can store the operating system 551 used to control the operation of the electronic device 500, and the basic input / output system (BIOS) 552 used to control the low-level operations of the electronic device 500. Additionally, it can store a web browser 553, a data storage management system 554, etc. In summary, when implementing the technical solution provided in this application through software or firmware, the relevant program code is stored in the memory 550 and is called and executed by the processor 510.

[0071] Input / output interface 530 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, vibrators, indicator lights, etc.

[0072] Network interface 540 is used to connect a communication module (not shown in the figure) to enable communication and interaction between the device and other devices. The communication module can communicate via wired means (e.g., USB, Ethernet cable) or wireless means (e.g., mobile network, Wi-Fi, Bluetooth).

[0073] Bus 560 includes a pathway for transmitting information between various components of the device, such as processor 510, disk drive 520, input / output interface 530, network interface 540, and memory 550.

[0074] It should be noted that although the above-described device only shows the processor 510, disk drive 520, input / output interface 530, network interface 540, memory 550, bus 560, etc., in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the method of this application, and does not necessarily include all the components shown in the figures.

[0075] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may 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.

[0076] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. 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 fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0077] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics, characterized in that, The method includes: The ultrasonic transmission waveform after the ultrasonic pulse penetrates the target battery is obtained. Multidimensional acoustic features are extracted frame by frame from the ultrasonic transmission waveform. A feature vector is constructed based on the multidimensional acoustic features and the battery state of charge at the same time. The multidimensional acoustic features include flight time and signal amplitude. Based on a sliding time window, a time-series input matrix is ​​constructed according to the feature vectors within a consecutive preset number of sampling periods; The time-series input matrix is ​​input into a pre-trained long short-term memory network model to obtain the internal temperature estimate of the target battery output by the long short-term memory network model. The long short-term memory network model is used to nonlinearly decouple the internal temperature and the battery state of charge in the time dimension through a time-series gating mechanism.

2. The method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics according to claim 1, characterized in that, The flight time is the total propagation time of the ultrasonic pulse from the transmitting probe to the receiving probe receiving the ultrasonic transmitted waveform; the signal amplitude is the peak amplitude or effective amplitude of the ultrasonic transmitted waveform.

3. The method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics according to claim 1, characterized in that, The window length of the sliding time window is determined based on the battery's thermal time constant, so that the number of timing frames contained in a single sliding time window covers the thermal inertial response period of battery temperature changes.

4. The method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics according to claim 1, characterized in that, The step of extracting multidimensional acoustic features frame by frame from the ultrasonic transmission waveform includes: The arrival time window of the main wave of the ultrasonic signal is determined based on the propagation speed of the ultrasonic wave in the medium inside the battery and the thickness of the battery. Multidimensional acoustic features are extracted frame by frame from the ultrasonic transmission waveform within the arrival time window of the main wave.

5. The method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics according to claim 4, characterized in that, The start time of the arrival time window of the main wave is determined based on the emission time and the shortest propagation path of the ultrasonic wave inside the battery, and the width of the arrival time window of the main wave is determined based on the battery thickness and the attenuation characteristics of the ultrasonic wave in the battery material.

6. The method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics according to claim 1, characterized in that, The method further includes: In a temperature chamber environment, a stepped temperature rise operation is performed with different preset battery charge states. The ultrasonic transmission waveform and the actual internal temperature value are recorded after each temperature step reaches thermal equilibrium. A training set is constructed, and a long short-term memory network model is trained based on the training set.

7. The method for estimating the internal temperature of a battery based on ultrasonic time-series characteristics according to claim 6, characterized in that, The holding time of the temperature step is determined based on the thermal time constant of the target battery and the response delay of the temperature sensor.

8. A battery internal temperature estimation device based on ultrasonic time-series characteristics, characterized in that, The device includes: The feature vector construction module is configured to acquire the ultrasonic transmission waveform after the ultrasonic pulse penetrates the target battery, extract multidimensional acoustic features frame by frame from the ultrasonic transmission waveform, and construct a feature vector based on the multidimensional acoustic features and the battery state of charge at the same time. The multidimensional acoustic features include flight time and signal amplitude. The input matrix construction module is configured to construct a temporal input matrix based on the feature vectors within a consecutive preset number of sampling periods using a sliding time window. The temperature estimation module is configured to input the time-series input matrix into a pre-trained long short-term memory network model to obtain the internal temperature estimate of the target battery output by the long short-term memory network model. The long short-term memory network model is used to nonlinearly decouple the internal temperature and the battery state of charge in the time dimension through a time-series gating mechanism.

9. An electronic device, characterized in that, include: One or more processors, and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the battery internal temperature estimation method based on ultrasonic timing characteristics according to any one of claims 1-7.

10. A computer program product, characterized in that, The method includes a computer program that, when executed by a processor, implements a battery internal temperature estimation method based on ultrasonic timing characteristics according to any one of claims 1-7.

Images