A spatiotemporal temperature modeling method for lithium battery packs

Abstract

The present disclosure provides a spatiotemporal temperature modeling method for lithium battery packs, which belongs to the field of data processing technology and specifically includes: preprocessing output data; selecting a corresponding scaling function based on the data change, and constructing an orthogonal wavelet basis function based on the scaling function; collecting spatiotemporal data and projecting it on the orthogonal wavelet basis function to obtain a time coefficient that characterizes the temporal dynamics of the temperature field; establishing a relationship vector between the time coefficient and the system input to describe the time series dynamics of the system; solving a support vector regression online model, using the loss coefficient corresponding to the loss function to represent the loss of the support vector, and obtaining a support vector regression time coefficient model that characterizes the nonlinear temporal dynamics of the temperature in the battery pack; integrating the orthogonal wavelet basis function with the support vector regression time coefficient model to obtain a spatiotemporal dynamic model of the lithium battery to dynamically model the battery pack temperature in time and space. The solution disclosed in the present disclosure improves prediction efficiency and safety during use.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of data processing technology, and in particular to a spatiotemporal temperature modeling method for a lithium battery pack. Background Art

[0002] Lithium-ion battery packs are currently being used in a wide range of applications, including hybrid buses, pure electric vehicles, underwater weapons and vehicles, and aerospace. However, the safety of lithium-ion batteries remains a significant concern. Overcharging and over-discharging can negatively impact battery performance and potentially lead to serious accidents. It is well known that the temperature of lithium batteries directly affects their capacity and lifespan, ultimately determining the battery pack's operational stability and safety.

[0003] It is generally believed that the heat generated by lithium-ion batteries can be divided into four parts, namely irreversible resistance heat, reversible reaction entropy heat, mixing heat, and phase change heat. Among them, the first two forms of heat are the key thermal behaviors that cause the battery temperature to rise. These processes make it difficult for lithium batteries to always be at a suitable temperature during operation. If lithium batteries are kept at high temperatures for a long time, they will accelerate aging and may even explode in severe cases. Usually, in order to achieve higher power or operating voltage, multiple single cells need to be combined in series or parallel to form a battery pack. The structure of the lithium battery pack and the sensor layout are as follows: Figure 1 As shown. However, inconsistent temperature of single cells will affect the heat transfer between the battery cells inside the battery pack, resulting in a shortened battery pack life and increased safety risks during vehicle driving. In addition, the temperature of the external environment also has a significant impact on the actual temperature of the lithium battery. According to the thermal radiation equation, the temperature difference between the inside and outside of the battery directly determines the efficiency of the battery heat exchange. Especially in extreme weather conditions such as summer and winter, the operating environment temperature of the battery pack is between -20 and 50 degrees Celsius, which has a great impact on the normal operation of the battery. The above complex internal and external factors have brought great challenges to the effective prediction of the temperature of the lithium battery pack.

[0004] It can be seen that there is an urgent need for a spatiotemporal temperature modeling method for lithium battery packs that can accurately predict the temperature of lithium battery packs and improve their safety. Summary of the Invention

[0005] In view of this, an embodiment of the present disclosure provides a spatiotemporal temperature modeling method for a lithium battery pack, which at least partially solves the problems of poor prediction efficiency and safety in the prior art.

[0006] The present disclosure provides a spatiotemporal temperature modeling method for a lithium battery pack, comprising:

[0007] Step 1: Using a series of spatially distributed sensors, collect input and temperature output data during the battery pack operation and pre-process the output data;

[0008] Step 2: Draw the change curve of the preprocessed output data, select the corresponding scaling function according to the change of the data, and construct the orthogonal wavelet basis function accordingly;

[0009] Step 3: collect spatiotemporal data and project them on orthogonal wavelet basis functions to obtain time coefficients that characterize the temporal dynamics of the temperature field;

[0010] Step 4: Establish the relationship vector between the time coefficient and the system input to describe the time series dynamics of the system;

[0011] Step 5: Solve the support vector regression online model, use the loss coefficient corresponding to the loss function to represent the loss of the support vector, and obtain the support vector regression time coefficient model that characterizes the nonlinear time dynamics of temperature in the battery pack;

[0012] Step 6: Integrate the orthogonal wavelet basis function with the support vector regression time coefficient model to obtain a spatiotemporal dynamic model of the lithium battery to dynamically model the battery pack temperature in time and space.

[0013] According to a specific implementation of the embodiment of the present disclosure, step 2 specifically includes:

[0014] According to the temperature distribution curve of the lithium battery pack at each spatial sensing point, the corresponding scaling function is selected as the initial orthogonal basis;

[0015] The low-pass filter and the high-pass filter are solved according to the initial orthogonal basis and the orthogonal wavelet basis function is constructed according to the low-pass filter and the high-pass filter.

[0016] According to a specific implementation of the embodiment of the present disclosure, the expression of the low-pass filter is:

[0017] h(n)=<Φ(x),Φ(xk)>, the expression of the high-pass filter is g(n)=(-1) n h(-n+1), the expression of the orthogonal wavelet basis function is Where n is a non-negative integer, {Φ(xk)} k∈Z Represents the scaling function, which is used to construct the orthogonal wavelet basis function, x=1,2,…,N represents the spatial position point, (k∈Z)∩(k∈[0,N]), and N is the number of spatial sensing points.

[0018] According to a specific implementation of the embodiment of the present disclosure, before step 1, the method further includes:

[0019] Combining the time-space separation framework and wavelet transform to obtain the lithium battery pack temperature modeling formula Among them, α j,k (t k ) represents t k The time coefficient of the instantaneous system at point x represents the time nonlinear dynamics of the lithium battery pack temperature, ψ j,k (x) represents the wavelet basis function corresponding to the main feature of the spatial point x, which represents the spatial temperature feature of the battery pack.

[0020] According to a specific implementation of the embodiment of the present disclosure, step 3 specifically includes:

[0021] Multiply the lithium battery pack temperature modeling formula with the orthogonal wavelet basis function to obtain the time coefficient α that characterizes the time dynamics of the temperature field. i (t k )=<ψ i (x),y(x,t k )>.

[0022] According to a specific implementation of the embodiment of the present disclosure, the expression of the relationship vector is:

[0023] z(t k )=[α(u(t k-1 ));α(u(t k-2 ));…;α(u(t k-p ));u(t k );…;u(t k-q )], where p and q represent the time coefficients of the past p moments and the input step lengths of the past q moments, respectively.

[0024] According to a specific implementation of the embodiment of the present disclosure, the expression of the support vector regression time coefficient model is: Among them, ω(t k ) represents the weight coefficient, represents the training value of the bias term, ξ m (t k ) represents t k At this moment, the loss of the mth sample.

[0025] According to a specific implementation of the embodiment of the present disclosure, the expression of the spatiotemporal dynamic model is:

[0026] The spatiotemporal temperature modeling scheme for a lithium battery pack in an embodiment of the present disclosure includes: step 1, using a series of spatially distributed sensors to collect input and temperature output data during the operation of the battery pack and preprocessing the output data; step 2, drawing a change curve of the preprocessed output data, selecting a corresponding scaling function according to the change of the data, and constructing an orthogonal wavelet basis function based on this; step 3, projecting the collected spatiotemporal data on the orthogonal wavelet basis function to obtain a time coefficient that characterizes the temporal dynamics of the temperature field; step 4, establishing a relationship vector between the time coefficient and the system input to describe the time series dynamics of the system; step 5, solving a support vector regression online model, using the loss coefficient corresponding to the loss function to represent the loss of the support vector, and obtaining a support vector regression time coefficient model that characterizes the nonlinear temporal dynamics of temperature in the battery pack; step 6, integrating the orthogonal wavelet basis function with the support vector regression time coefficient model to obtain a spatiotemporal dynamic model of the lithium battery to dynamically model the battery pack temperature in time and space.

[0027] The beneficial effects of the embodiments of the present disclosure are:

[0028] (1) According to the dynamic distribution characteristics of temperature in time and space during the operation of lithium battery packs, a spatiotemporal modeling strategy based on the spatiotemporal separation framework and wavelet transform is designed. The spatiotemporal data collected by spatially distributed sensors are used to effectively model the temperature dynamics of lithium batteries.

[0029] (2) According to the temperature distribution curve of the lithium battery pack at each spatial sensing point, a suitable wavelet basis function is selected. The original signal is decomposed into a superposition of a series of wavelet functions using the inner product of the wavelet basis function and the original signal, which characterizes the main spatial distribution characteristics of the temperature on the battery pack.

[0030] (3) The original spatiotemporal temperature data are projected onto the wavelet basis function to obtain the time coefficients representing the temporal dynamics of the temperature. Based on this, an online SVR method is developed, which reconstructs the nonlinear temporal dynamics of the lithium battery pack temperature.

[0031] (4) The constructed wavelet basis function is integrated with the SVR time coefficient model to obtain the spatiotemporal dynamic distribution model of the lithium battery surface temperature, thereby realizing accurate modeling of the spatiotemporal dynamics of the lithium battery pack temperature. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0033] Figure 1 A schematic flow chart of a spatiotemporal temperature modeling method for a lithium battery pack provided in an embodiment of the present disclosure;

[0034] Figure 2 A schematic diagram of the structure and sensor layout of a lithium battery pack provided in an embodiment of the present disclosure;

[0035] Figure 3 A schematic diagram of a temperature field spatiotemporal modeling strategy based on a spatiotemporal separation framework and wavelet transform ideas provided in an embodiment of the present disclosure;

[0036] Figure 4 A schematic diagram of a process for extracting spatial features of lithium battery pack temperature based on wavelet transform provided in an embodiment of the present disclosure;

[0037] Figure 5 A schematic diagram of a temperature field time dynamic modeling process based on SVR provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0038] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0039] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.

[0040] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this disclosure, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement an apparatus and / or practice a method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this apparatus and / or practice this method.

[0041] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The illustrations only show components related to the present disclosure and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0042] Additionally, in the following description, specific details are provided to provide a thorough understanding of the examples. However, one skilled in the art will appreciate that the aspects described can be practiced without these specific details.

[0043] An embodiment of the present disclosure provides a spatiotemporal temperature modeling method for a lithium battery pack, which can be applied to the temperature prediction process of lithium batteries in new energy application scenarios.

[0044] See also Figure 1 , is a flow chart of a method for spatiotemporal temperature modeling of a lithium battery pack provided by an embodiment of the present disclosure. Figure 1 As shown, the method mainly includes the following steps:

[0045] Step 1: Using a series of spatially distributed sensors, collect input and temperature output data during the battery pack operation and pre-process the output data;

[0046] Optionally, before step 1, the method further includes:

[0047] Combining the time-space separation framework and wavelet transform to obtain the lithium battery pack temperature modeling formula Among them, α j,k (t k ) represents t k The time coefficient of the instantaneous system at point x represents the time nonlinear dynamics of the lithium battery pack temperature, ψ j,k (x) represents the wavelet basis function corresponding to the main feature of the spatial point x, which represents the spatial temperature feature of the battery pack.

[0048] In the specific implementation, the structure of the lithium battery pack and the sensor layout are taken into consideration. Figure 2 As shown, specifically, based on the time-space separation framework, combined with the wavelet transform (WT) and support vector regression (SVR) algorithm, the effective reconstruction of the surface temperature of the lithium battery pack is achieved.

[0049] The specific principles of the modeling method proposed in this invention are as follows: Figure 3 As shown, the technical solutions of the present invention are as follows:

[0050] First, based on the actual operating conditions of lithium-ion battery packs, the physical mechanism of the temperature distribution of lithium-ion battery packs is studied, and the influence of unknown boundary conditions inside and outside the battery pack, the energy transfer process, and the spatiotemporal coupling dynamics on the temperature distribution of the battery pack are analyzed. Based on the above characteristics, a layout strategy for lithium-ion battery pack surface temperature sensors in a series-parallel mixed arrangement mode is designed. Combining the basic framework of spatiotemporal separation modeling, a lithium-ion battery pack temperature modeling method based on wavelet transform (WT) and support vector regression (SVR) algorithms is proposed. This model combines the actual sensor layout strategy and uses wavelet basis functions and time coefficient models respectively to reconstruct the nonlinear dynamic behavior of the temperature field in time and space during the operation of the lithium-ion battery pack, providing important support for the temperature prediction and control of the lithium-ion battery pack.

[0051] The idea of ​​lithium battery pack temperature modeling based on the time-space separation framework is:

[0052] The spatiotemporal temperature dynamics of lithium-ion battery packs during operation are typical distributed parameter systems (DPS). This system is characterized by significant nonlinear dynamic variations in both spatial and temporal state. Compared to traditional lumped parameter systems, this spatiotemporal distribution presents significant challenges for accurate temperature field modeling. An effective approach is to employ spatiotemporal separation modeling to effectively model the temperature field. Specifically:

[0053]

[0054] Among them, ψ(x)={ψ(x1),ψ(x2),…,ψ(x D )} represents the spatial basis function, which represents the distribution characteristics of the system in space; D is the dimension of the spatial basis function, Indicates t k At the moment the system is at x i The time coefficient of the position point is used to characterize the dynamic characteristics of the system in time.

[0055] According to the idea of ​​wavelet transform, all signals can be represented by a linear combination of a series of wavelet basis functions. Therefore, combining the time-space separation framework and wavelet transform, a lithium battery pack temperature modeling method based on the time-space separation framework is proposed here, as follows:

[0056]

[0057] Among them, α j,k (t k ) represents t k The time coefficient of the instantaneous system at point x represents the time nonlinear dynamics of the lithium battery pack temperature; ψ j,k(x) represents the wavelet basis function corresponding to the main feature of the spatial point x, which represents the spatial temperature characteristics of the battery pack. Then, a series of spatially distributed sensors are used to collect the input and temperature output data of the battery pack during operation. And preprocess the data.

[0058] Step 2: Draw the change curve of the preprocessed output data, select the corresponding scaling function according to the change of the data, and construct the orthogonal wavelet basis function accordingly;

[0059] Furthermore, the step 2 specifically includes:

[0060] According to the temperature distribution curve of the lithium battery pack at each spatial sensing point, the corresponding scaling function is selected as the initial orthogonal basis;

[0061] The low-pass filter and the high-pass filter are solved according to the initial orthogonal basis and the orthogonal wavelet basis function is constructed according to the low-pass filter and the high-pass filter.

[0062] Furthermore, the expression of the low-pass filter is h(n)=<Φ(x),Φ(xk)>, and the expression of the high-pass filter is g(n)=(-1) n h(-n+1), the expression of the orthogonal wavelet basis function is Where n is a non-negative integer, {Φ(xk)} k∈Z Represents the scaling function, which is used to construct the orthogonal wavelet basis function, x=1,2,…,N represents the spatial position point, (k∈Z)∩(k∈[0,N]), and N is the number of spatial sensing points.

[0063] In specific implementation, temperature space feature extraction based on wavelet basis function

[0064] In order to accurately model the temperature of lithium battery packs, sensors are needed to collect the temperature of key nodes of the battery pack and identify key locations. A key issue is how to select an appropriate method to effectively characterize the spatial characteristics of the battery pack. To solve this problem, the present invention proposes a method for extracting spatial characteristics of lithium battery pack temperature based on wavelet transform. Figure 4 It selects a suitable mother wavelet according to the distribution characteristics of the data, and then scales and translates the mother wavelet to obtain the wavelet basis function ψ suitable for lithium battery packs j,k (x) collection, so that the dynamic characteristics of the lithium battery pack in space are effectively characterized. The specific process is as follows:

[0065] According to the temperature distribution curve of the lithium battery pack at each spatial sensing point, select the appropriate scaling function Φ(xk) k∈Z , such as Haar wavelet, as the initial orthogonal basis;

[0066] The low-pass filter h(x) is obtained by solving the following formula:

[0067] h(x)=<Φ(x),Φ(xk)> (3)

[0068] The high-pass filter g(x) is obtained by solving the following formula:

[0069] g(x)=(-1) x h(-x+1) (4)

[0070] Construct orthogonal wavelet basis function ψ(x) according to h(x) and g(x)

[0071]

[0072] By analogy, we can get the wavelet basis function set ψ(x)={ψ j,k (x)}, the wavelet basis functions are orthogonal to each other and satisfy ψ j,k (x) = 2 j / 2 ψ(2 j x-kN).

[0073] Step 3: collect spatiotemporal data and project them on orthogonal wavelet basis functions to obtain time coefficients that characterize the temporal dynamics of the temperature field;

[0074] Furthermore, the step 3 specifically includes:

[0075] Multiply the lithium battery pack temperature modeling formula with the orthogonal wavelet basis function to obtain the time coefficient α that characterizes the time dynamics of the temperature field. i (t k )=<ψ i (x),y(x,t k )>.

[0076] In specific implementation, the collected spatiotemporal data can be projected onto the wavelet basis function ψ(x) constructed above to obtain the time coefficient that characterizes the nonlinear temperature dynamics of the lithium battery pack. Then, a time system model based on online SVR is developed to achieve effective modeling of the temperature and time dynamics of the lithium battery pack. Figure 5 As shown in Figure 2, the proposed temporal dynamic modeling process is as follows:

[0077] By multiplying formula (2) and formula (5), we can get the temperature-time coefficient corresponding to each moment:

[0078] α i (t k )=<ψ i (x),y(x,t k )> (6)

[0079] Step 4: Establish the relationship vector between the time coefficient and the system input to describe the time series dynamics of the system;

[0080] Based on the above embodiment, the expression of the relationship vector is z(t k )=[α(u(t k-1 ));α(u(t k-2 ));…;α(u(t k-p ));u(t k );…;u(t k-q )], where p and q represent the time coefficients of the past p moments and the input step lengths of the past q moments, respectively.

[0081] In specific implementation, we can establish α i (t k ) and system input u(t k ) of the relationship vector z(t k ) to describe the time series dynamics of the system.

[0082] z(t k )=[α(u(t k-1 ));α(u(t k-2 ));…;α(u(t k-p ));u(t k );…;u(t k-q )] (7)

[0083] Among them, p and q represent the time coefficient of the past p moments and the input step size of the past q moments respectively.

[0084] In order to accurately model the time coefficient, an online modeling strategy based on SVR is designed here. Generally, support vector regression has a good fitting ability for linear relationships, but traditional SVR cannot accurately handle nonlinear modeling problems, and in the solution process, due to the involvement of dual problems and sequential minimum optimization processes, the solution process has a high computational complexity. To address this situation, kernel technology is used to handle nonlinear dynamics by using nonlinear mapping functions. In addition, this method uses the loss coefficient derived from the loss function to represent the loss of the support vector, effectively ensuring the sparsity of the modeling process. By converting the weight coefficient into the loss coefficient, the SVR solution process is converted into the calculation of the loss coefficient, and the loss coefficient can be calculated by the SGD algorithm, which can greatly reduce the computational complexity.

[0085] Introducing kernel techniques into traditional SVR yields the following regression equation:

[0086]

[0087] Among them, ω(t k) and b(t k ) represent the SVR model at t k The weight coefficient and bias term at the moment. And,ω(t k )=[ω1(t k ),ω2(t k ),…,ω N-1 (t k )] T .

[0088] Step 5: Solve the support vector regression online model, use the loss coefficient corresponding to the loss function to represent the loss of the support vector, and obtain the support vector regression time coefficient model that characterizes the nonlinear time dynamics of temperature in the battery pack;

[0089] Furthermore, the expression of the support vector regression time coefficient model is: Among them, ω(t k ) represents the weight coefficient, represents the training value of the bias term, ξ m (t k ) represents t k At this moment, the loss of the mth sample is,

[0090] In specific implementation, the Lagrange multiplier method is used for equation (8) and a regularization factor is introduced to further obtain:

[0091]

[0092] Where λ represents the regularization factor; is an insensitive loss function, defined as follows:

[0093]

[0094] Using gradient descent (SGD), assuming n is in t k The maximum number of iterations to collect training data at any time; (z p (t k ),α p (t k )) is a test sample randomly selected in the pth (p = 1, 2, 3, ..., n) iteration, and then the following objective function can be obtained:

[0095]

[0096] Among them, ω p (t k ) represents the coefficient of the SVR model at the pth iteration. ε (·) is an insensitive loss function.

[0097]

[0098] Furthermore, the subgradient of equation (11) can be derived as follows:

[0099]

[0100] Definition η p =1 / (λp) as the step coefficient in the negative direction of the gradient We can get:

[0101]

[0102] Assumptions You can get:

[0103]

[0104] Define the loss function ξ m (t k ) represents the total loss of sample m being selected in the past k moments. Formula (15) can be transformed into

[0105]

[0106] Among them, ξ m (t k )satisfy

[0107]

[0108] represents the modeling error of the pth iteration. From formula (17), we can see that ξ m (t k ) and δ p (t k ) is directly related: when δ p (t k )>0,ξ m (t k )=ξ m (t k-1 )+δ p (t k ); On the contrary, m (t k )=ξ m (t k-1 ). On this basis, we can get t k Time weight coefficient ω(t k ) and the training value of the bias term b is:

[0109]

[0110]

[0111] Substituting equations (18) and (19) into equation (8), we obtain α(t k ) are as follows:

[0112]

[0113] Step 6: Integrate the orthogonal wavelet basis function with the support vector regression time coefficient model to obtain a spatiotemporal dynamic model of the lithium battery to dynamically model the battery pack temperature in time and space.

[0114] Based on the above embodiment, the expression of the spatiotemporal dynamic model is:

[0115]

[0116] In specific implementation, we can finally combine formula (5) and (20) to get

[0117]

[0118] As can be seen from formula (21), the spatiotemporal model effectively models the dynamics of the battery pack temperature in time and space through wavelet basis functions and time coefficient models. Therefore, the reconstructed model can well simulate the nonlinear spatiotemporal dynamics of the temperature field during the operation of the lithium battery pack.

[0119] The spatiotemporal temperature modeling method for lithium battery packs provided in this embodiment takes into account the dynamic distribution characteristics of the temperature distribution of lithium battery packs in both time and space. Based on the existing spatiotemporal separation architecture, a spatiotemporal temperature distribution model of the lithium battery pack is established to characterize the dynamic distribution characteristics of the battery pack temperature field in time and space. Based on the temperature distribution curve of the lithium battery pack at each spatial sensing point, appropriate wavelet basis functions are selected. The wavelet functions at different scales are then inner-producted with the original signal to perform wavelet decomposition on the signal. This decomposition of the original signal into a superposition of a series of wavelet functions is performed, and the main spatial distribution characteristics of the temperature on the battery pack are characterized by the set of wavelet basis functions. Projecting the original spatiotemporal temperature data onto the wavelet basis functions reveals the temporal dynamics of the temperature. This nonlinear dynamics is manifested in the superposition coefficients of the wavelet functions, i.e., the time coefficients. Using the obtained time coefficients, an online SVR method is developed to construct a time coefficient model to characterize the nonlinear temporal dynamics of the lithium battery pack temperature. The constructed wavelet basis functions are integrated with the SVR time coefficient model to obtain a spatiotemporal dynamic distribution model of the lithium battery surface temperature, achieving accurate modeling of the lithium battery pack temperature.

[0120] It should be understood that various parts of the present disclosure can be implemented in hardware, software, firmware, or a combination thereof.

[0121] The above description is merely a specific embodiment of the present disclosure, but the scope of protection of the present disclosure is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this disclosure should be included in the scope of protection of the present disclosure. Therefore, the scope of protection of the present disclosure should be based on the scope of protection of the claims.

Claims

1. A spatiotemporal temperature modeling method for lithium battery packs, characterized in that: include: Step 1: Using a series of spatially distributed sensors, collect input and temperature output data during the battery pack operation and pre-process the output data; Step 2: Draw the change curve of the preprocessed output data, select the corresponding scaling function according to the change of the data, and construct the orthogonal wavelet basis function accordingly; The step 2 specifically includes: According to the temperature distribution curve of the lithium battery pack at each spatial sensing point, the corresponding scaling function is selected as the initial orthogonal basis; The low-pass filter and the high-pass filter are solved according to the initial orthogonal basis and the orthogonal wavelet basis function is constructed according to the low-pass filter and the high-pass filter, wherein the expression of the low-pass filter is h(n)=<Φ(x),Φ(xk)>, and the expression of the high-pass filter is g(n)=(-1) n h(-n+1), the expression of the orthogonal wavelet basis function is Where n is a non-negative integer, {Φ(xk)} k∈Z Represents the scaling function, which is used to construct the orthogonal wavelet basis function, x=1,2,…,N represents the spatial position point, (k∈Z)∩(k∈[0,N]), N is the number of spatial sensing points; Step 3: collect spatiotemporal data and project them on orthogonal wavelet basis functions to obtain time coefficients that characterize the temporal dynamics of the temperature field; Step 4: Establish the relationship vector between the time coefficient and the system input to describe the time series dynamics of the system; Step 5: Solve the support vector regression online model, use the loss coefficient corresponding to the loss function to represent the loss of the support vector, and obtain the support vector regression time coefficient model that characterizes the nonlinear time dynamics of temperature in the battery pack; Step 6: Integrate the orthogonal wavelet basis function with the support vector regression time coefficient model to obtain a spatiotemporal dynamic model of the lithium battery to dynamically model the battery pack temperature in time and space.

2. The method according to claim 1, characterized in that Before step 1, the method further includes: Combining the time-space separation framework and wavelet transform to obtain the lithium battery pack temperature modeling formula Among them, α j,k (t k ) represents t k The time coefficient of the instantaneous system at point x represents the time nonlinear dynamics of the lithium battery pack temperature, ψ j,k (x) represents the wavelet basis function corresponding to the main feature of the spatial point x, which represents the spatial temperature feature of the battery pack.

3. The method according to claim 2, characterized in that , the step 3 specifically includes: Multiply the lithium battery pack temperature modeling formula with the orthogonal wavelet basis function to obtain the time coefficient α that characterizes the time dynamics of the temperature field i (t k )=<ψ i (x),y(x,t k )>.

4. The method according to claim 3, characterized in that , the expression of the relationship vector is z(t k )=[α(u(t k-1 ));α(u(t k-2 ));…;α(u(t k-p ));u(t k );…;u(t k-q )], where p and q represent the time coefficients of the past p moments and the input step lengths of the past q moments, respectively.

5. The method according to claim 4, characterized in that , the expression of the support vector regression time coefficient model is Among them, ω(t k ) represents the weight coefficient, represents the training value of the bias term, ξ m (t k ) represents t k At this moment, the loss of the mth sample.

6. The method according to claim 5, characterized in that , the expression of the spatiotemporal dynamic model is

Images