A method and system for optimizing spatial temperature distribution

Abstract

The present invention discloses a method and system for optimizing spatial temperature distribution. The method includes the steps of: constructing a first spatial temperature distribution, optimizing the first node spatial distribution corresponding to the first spatial temperature distribution, and constructing a second spatial temperature distribution by using the temperature data of each detection node in the second node spatial distribution. By constructing the first spatial temperature distribution by using an interpolation function, the construction accuracy of the detection nodes is improved. By optimizing the first spatial node distribution in the first spatial temperature distribution, the detection nodes with low data reduction rate are removed. Under the condition of ensuring the construction accuracy, the second spatial temperature distribution of the lithium battery can be constructed by using fewer detection nodes, reducing the data complexity and the cost.

Description

Technical Field

[0001] The present invention relates to the field of temperature distribution, and particularly to a method and system for optimizing spatial temperature distribution. Background Art

[0002] Lithium-ion batteries have characteristics such as high energy density, high power density, high cycle times, and high voltage, and are one of the relatively ideal energy supplies at present. With the large-scale application of lithium-ion batteries, their safety has become particularly important, and the heat generation of lithium-ion batteries is one of the core contents of safety issues. Temperature has various effects on the performance of lithium-ion batteries, including the working conditions of the electrochemical system, cycle efficiency, capacity, power, safety, reliability, and lifespan, etc., thus affecting the safety and reliability of electric vehicles.

[0003] Lithium-ion batteries should operate within a reasonable temperature range, and high and low temperatures have a close impact on the battery's performance. The Karhunen-Loeve (KL) based spatio-temporal separation is a representative data-driven distributed thermal modeling method. It uses singular value decomposition to extract a set of main global spatial basis functions from the sampled data. Combining with the Galerkin projection method, the original infinite-dimensional thermal process is decomposed into several ordinary differential equations (ODEs). Finally, the spatio-temporal output can be reconstructed through spatio-temporal synthesis. However, for the KL algorithm model, a large number of sensors need to be deployed to achieve satisfactory modeling accuracy. Too many sensors will occupy space, reduce the power-volume ratio of the battery, which is not conducive to its compact design and heat dissipation. At the same time, excessive sensors will increase the sensing cost and data processing requirements of the battery management system, and the cost is high, and it cannot accurately reflect the overall thermal condition of the lithium battery. Summary of the Invention

[0004] In the prior art, for the temperature distribution of lithium batteries, excessive sensors will increase the sensing cost and data processing requirements of the battery management system, and the cost is high, and it cannot accurately reflect the overall thermal condition of the lithium battery.

[0005] To solve the above problems, a method and system for optimizing spatial temperature distribution are proposed to solve the above problems.

[0006] In a first aspect, a method for optimizing spatial temperature distribution, which is used to optimize the spatial temperature distribution of a lithium battery, includes:

[0007] Step 100: Obtain the spatial structure parameters of the lithium battery, construct an interpolation function using the spatial structure parameters, and construct a first spatial temperature distribution using the interpolation function;

[0008] Step 200: Optimize the first node spatial distribution corresponding to the first spatial temperature distribution: Calculate the data reduction rate of the eigenvalues of the data matrix of the first spatial temperature distribution, remove the detection nodes with a relatively small data reduction rate, and obtain the second node spatial distribution;

[0009] Step 300: Construct a second spatial temperature distribution using the temperature data of each detection node in the second node spatial distribution;

[0010] Wherein, the data reduction rate is the proportion of the original data contained in the eigenvector corresponding to the transformed eigenvalue.

[0011] Combined with the spatial temperature distribution optimization method of the present invention, in the first possible implementation manner, the step 100 includes:

[0012] Step 110: Arrange a plurality of detection nodes in the lithium battery;

[0013] Step 120: Obtain the spatial coordinates and temperature of the Nth detection node;

[0014] Step 130: Construct the interpolation function with the Nth detection node as the interpolation center.

[0015] Combined with the first possible implementation manner of the first aspect of the present invention, in the second possible implementation manner, the step 100 further includes:

[0016] Step 140: Obtain the spatial coordinates and temperature of each detection node;

[0017] Step 150: Interpolate the temperature detection data of the detection nodes using the interpolation function to obtain the first spatial temperature distribution.

[0018] Combined with the spatial temperature distribution optimization method of the first aspect of the present invention, in the third possible implementation manner, the step 200 includes:

[0019] Step 210: Perform eigen-decomposition processing on the data matrix to obtain a diagonal matrix of eigenvalues;

[0020] Step 220: Calculate the data reduction rate of the first t eigenvalues;

[0021] Step 230: If the data reduction rate is greater than the specified threshold, discard the detection data corresponding to the (t + 1)th eigenvalue when constructing the temperature distribution;

[0022] Step 240: Screen the detection nodes according to the comparison result, and use the spatial node arrangement of the remaining detection nodes as the second node spatial arrangement.

[0023] Combined with the third possible implementation manner of the first aspect of the present invention, in the fourth possible implementation manner, the step 300 includes:

[0024] Step 310, obtaining a detection matrix corresponding to the first t eigenvalues;

[0025] Step 320, constructing the second spatial temperature distribution by using the feature matrix, detection data, and the detection matrix.

[0026] In a second aspect, a spatial temperature distribution optimization system, adopting the detection method described in the first aspect, includes:

[0027] A first construction module;

[0028] An optimization module;

[0029] A second construction module;

[0030] The first construction module is configured to obtain the spatial structure parameters of the lithium battery, construct an interpolation function by using the spatial structure parameters, and construct a first spatial temperature distribution by using the interpolation function;

[0031] The optimization module is configured to optimize the first node spatial distribution corresponding to the first spatial temperature distribution: calculate the data reduction rate of the eigenvalues of the detected data matrix, remove the detection nodes with a smaller data reduction rate, and obtain a second node spatial distribution;

[0032] The second construction module constructs a second spatial temperature distribution by using the temperature data of each detection node in the second node spatial distribution;

[0033] Wherein, the data reduction rate is the proportion of the original data included in the eigenvector corresponding to the transformed eigenvalue.

[0034] Combined with the spatial temperature distribution optimization system described in the second aspect of the present invention, in the first possible implementation manner, the first construction module is further configured to:

[0035] Construct the interpolation function with the Nth detection node as the interpolation center, interpolate the detection data of the detection nodes by using the interpolation function, and obtain the first spatial temperature distribution.

[0036] Combined with the first possible implementation manner described in the second aspect of the present invention, in the second possible implementation manner, the optimization module is further configured to:

[0037] Perform eigen-decomposition processing on the data matrix of the detection node data, obtain a diagonal matrix of eigenvalues, and determine the detection nodes according to the data reduction rate of the first t eigenvalues and by comparing the data reduction rate with a specified threshold.

[0038] Implementing the spatial temperature distribution optimization method and system of the present invention, by using an interpolation function to construct the first spatial temperature distribution, the construction accuracy of the detection nodes is improved. By optimizing the first spatial node distribution in the first spatial temperature distribution, the detection nodes with low data reduction rate are removed. Under the condition of ensuring the construction accuracy, the second spatial temperature distribution of the lithium battery can be constructed with fewer detection nodes, reducing the data complexity and cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0040] Figure 1 It is the first schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention;

[0041] Figure 2 It is the second schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention;

[0042] Figure 3 It is the third schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention;

[0043] Figure 4 It is the fourth schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention;

[0044] Figure 5 It is the fifth schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention;

[0045] Figure 6 It is the schematic diagram of the composition of the spatial temperature distribution optimization system of the present invention;

[0046] Figure 7 It is the simulation diagram of the detection error of different numbers of detection nodes;

[0047] Figure 8 It is the second node spatial distribution diagram;

[0048] The parts referred to by the numbers in the drawings are: 10 - the first construction module, 20 - the optimization module, 30 - the second construction module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Next, the technical solutions in the present invention will be clearly and completely described in conjunction with the accompanying drawings in the invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, other embodiments obtained by those of ordinary skill in the art without creative efforts all fall within the scope of protection of the present invention.

[0050] In the prior art, for the temperature distribution of lithium batteries during charging, excessive sensors will increase the sensing cost and data processing requirements of the battery management system, and the cost is high, and it cannot accurately reflect the overall thermal condition of the lithium battery.

[0051] In view of the above problems, a method and system for optimizing the spatial temperature distribution are proposed to solve the above problems.

[0052] In the first aspect, as Figure 1 , Figure 1 is the first schematic diagram of the steps of the method for optimizing the spatial temperature distribution of the present invention; a method for optimizing the spatial temperature distribution includes:

[0053] Step 100: Obtain the spatial structure parameters of the lithium battery, construct an interpolation function using the spatial structure parameters, and construct the first spatial temperature distribution using the interpolation function.

[0054] Preferably, as Figure 2 , Figure 2 is the second schematic diagram of the steps of the method for optimizing the spatial temperature distribution of the present invention; Step 100 includes: Step 110: Arrange a plurality of temperature detection nodes on the lithium battery; Step 120: Obtain the spatial coordinates and temperature of the Nth temperature detection node; Step 130: Construct an interpolation function with the Nth detection node as the interpolation center.

[0055] The lithium battery can be in a cube shape, and the detection nodes are arranged on the square shell, and the detection nodes can be temperature sensors.

[0056] The interpolation function is defined as:

[0057]

[0058] where μ is the spatial structure parameter:

[0059]

[0060] where L is the maximum distance between the detection nodes, and K is the number of detection nodes.

[0061] Preferably, as Figure 3 , Figure 3 is the third schematic diagram of the steps of the method for optimizing the spatial temperature distribution of the present invention; Step 100 further includes:

[0062] Step 140: Obtain the spatial coordinates and temperature of each detection node; Step 150: Use the interpolation function to interpolate the temperature detection data of the detection nodes to obtain the first spatial temperature distribution.

[0063] The spatial coordinates of the Nth detection node are (x N , y N ), and the temperature detection data detected by the Nth detection node is After interpolation, the first spatial temperature distribution is:

[0064]

[0065] Among them, m, n are construction parameters, satisfying:

[0066]

[0067] Among them, H is the identity matrix,

[0068] Step 200: Optimize the first node spatial distribution corresponding to the first spatial temperature distribution: Calculate the data reduction rate of the eigenvalue of the first spatial temperature distribution data matrix, remove the detection nodes with a smaller data reduction rate, and obtain the second node spatial distribution.

[0069] The first node spatial distribution is the detection node distribution before optimization, and the first spatial temperature distribution is re-evaluated and constructed based on the detection data obtained from the first node spatial distribution.

[0070] Too many detection nodes not only increase the detection cost but also make the control circuit structure complex. When optimizing and screening the detection nodes, first obtain the detection data matrix K×W of the detection nodes in the first node spatial distribution. This data matrix represents the data samples of K detection nodes under W detection conditions.

[0071] The data samples are processed by mean and standardization to obtain the standardized data matrix Q corresponding to the detection data matrix K×W.

[0072] Preferably, as Figure 4 , Figure 4 is the fourth schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention; Step 200 includes: Step 210: Perform eigenvalue decomposition on the data matrix to obtain the diagonal matrix of eigenvalues; Step 220: Calculate the data reduction rate of the first t eigenvalues; Step 230: If the data reduction rate is greater than the specified threshold, discard the detection data corresponding to the (t + 1)th eigenvalue when constructing the temperature distribution; Step 240: Screen the detection nodes according to the comparison result, and use the spatial node arrangement of the remaining detection nodes as the second node spatial arrangement.

[0073] Perform eigen - decomposition on the data matrix Q:

[0074] C = QQ T = SΛS T (5),

[0075] where C is the covariance matrix of Q, represents the diagonal matrix formed by arranging the eigenvalues of the covariance matrix from large to small, represents the i - th eigenvalue, and S represents the eigen - matrix.

[0076] The data reduction rate represents how much of the original data is preserved by the eigen - vectors corresponding to the eigenvalues of the data matrix after conversion. The data reduction rate of the first t eigenvalues can be calculated as:

[0077]

[0078] Compare the data reduction rate ξ with the specified threshold M. If ξ≥M, it means that for the detection data matrix obtained by the first t detection nodes, after matrix conversion, the data reduction rate meets the detection requirements, and the data reduction rate of the (t + 1)-th eigenvalue is low. To reduce costs, discarding the detection data corresponding to the (t + 1)-th eigenvalue does not affect the re - estimation and construction of the temperature distribution. Therefore, the number of detection nodes can be reduced, and the spatial distribution of the detection nodes after screening is the second node spatial arrangement.

[0079] Step 300: Use the detection data of each detection node in the second node spatial distribution to construct the second - space temperature distribution.

[0080] Preferably, as Figure 5 , Figure 5 is the fifth schematic diagram of the steps of the spatial temperature distribution optimization method of the present invention; Step 300 includes:

[0081] Step 310: Obtain the eigen - matrix corresponding to the first t eigenvalues; Step 320: Use the eigen - matrix, detection data, and detection matrix to construct the second - space temperature distribution.

[0082] The detection data D among them is the set of measured data, and the detection matrix is the standardized data matrix Q1. Then the second - space temperature distribution is:

[0083]

[0084] Simulation experiment

[0085] To cover as many positions as possible, the first - node spatial distribution arranges temperature sensors at equal intervals as the un - optimized detection node distribution. Figure 7It is a simulation diagram of the detection error for different numbers of detection nodes. The simulation results show that a larger number of detection nodes does not necessarily lead to better measurement accuracy. This is because a small number of detection nodes can already obtain sufficient information for fitting the temperature distribution, and increasing the number of detection nodes may result in data redundancy, that is, the generation of useless reference points.

[0086] The construction error d is:

[0087] where u is the temperature data to be constructed, u1 is the constructed temperature data, and the construction errors under different detection conditions are shown in Table 1.

[0088] Table 1 Construction errors under different detection conditions

[0089] 15℃ 20℃ 25℃ 30℃ 35℃ Before optimization 0.83 0.84 1.32 1.22 1.16 After optimization 0.71 0.72 1.17 1.02 0.91

[0090] As can be seen from Table 1, when the eigenvectors, detection data, and number of construction times are kept consistent, the optimized construction error is significantly smaller than the pre-optimization construction error, improving the construction accuracy.

[0091] Figure 8 It is the spatial distribution diagram of the second node. In the figure, there are more detection nodes in the central area of the lithium battery, where the temperature change in the lithium battery area is relatively large. The coordinates (6, 14) and (6, 10) are close to the charging circuit position, and the detection nodes here can effectively reflect the temperature state of the circuit board and obtain the positions of points during the optimization process. In addition, the optimized detection nodes are mainly distributed in the areas of coordinates (15, 13) and (19, 13). The overall temperature condition of the lithium battery can be described with fewer detection nodes.

[0092] In the second aspect, as Figure 6 , Figure 6 is a schematic diagram of the spatial temperature distribution optimization system of the present invention; a spatial temperature distribution optimization system includes a first construction module 10, an optimization module 20, and a second construction module 30; the first construction module 10 is used to obtain the spatial structure parameters of the lithium battery, construct an interpolation function using the spatial structure parameters, and construct a first spatial temperature distribution using the interpolation function; the optimization module 20 is used to optimize the first node spatial distribution corresponding to the first spatial temperature distribution: calculate the data reduction rate of the eigenvalues of the detected data matrix, remove the detection nodes with a smaller data reduction rate, and obtain the second node spatial distribution; the second construction module 30 constructs a second spatial temperature distribution using the temperature data of each detection node in the second node spatial distribution; where the data reduction rate is the proportion of the original data contained in the eigenvector corresponding to the transformed eigenvalue.

[0093] The first construction module 10 is further configured to: construct an interpolation function with the Nth detection node as the interpolation center, and use the interpolation function to interpolate the detection data of the detection nodes to obtain the first spatial temperature distribution.

[0094] The optimization module 20 is further configured to:

[0095] Perform eigenvalue decomposition processing on the data matrix of the detection node data to obtain a diagonal matrix of eigenvalues, and determine the detection nodes according to the data reduction rate of the first t eigenvalues and by comparing the data reduction rate with a specified threshold.

[0096] Implementing the spatial temperature distribution optimization method and system according to the present invention, by using an interpolation function to construct the first spatial temperature distribution, the construction accuracy of the detection nodes is improved. By optimizing the first spatial node distribution in the first spatial temperature distribution, the detection nodes with low data reduction rate are removed. Under the condition of ensuring the construction accuracy, the second spatial temperature distribution of the lithium battery can be constructed by using fewer detection nodes, reducing the data complexity and the cost.

[0097] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A spatial temperature distribution optimization method for optimizing the spatial temperature distribution of a lithium battery, characterized in that: include: Step 100, obtaining spatial structural parameters of the lithium battery, constructing an interpolation function using the spatial structural parameters, and constructing a first spatial temperature distribution using the interpolation function; The step 100 comprises: Step 110, arranging a plurality of detection nodes on the lithium battery; Step 120, obtaining the spatial coordinates and temperature of the Nth detection node; Step 130, constructing the interpolation function with the Nth detection node as the interpolation center; The interpolation function is defined as: Among them, μ is the spatial structure parameter: Among them, L is the maximum distance between detection nodes, and K is the number of detection nodes; The step 100 further includes: Step 140, obtaining the spatial coordinates and temperature of each detection node; Step 150: interpolate the temperature detection data of the detection node using the interpolation function to obtain the first spatial temperature distribution; Assume that the spatial coordinates of the Nth detection node are (x N ,y N ), the temperature detection data detected by the Nth detection node is After interpolation, the first spatial temperature distribution for: in, To construct the parameters, satisfy: in, H is the identity matrix, Step 200, optimizing the first node spatial distribution corresponding to the first spatial temperature distribution: calculating the data restoration rate of the eigenvalues ​​of the data matrix of the first spatial temperature distribution, and comparing the data restoration rate with a specified threshold to determine the detection node based on the data restoration rate of the first t eigenvalues, and obtaining the second node spatial distribution; The step 200 includes: Step 210, perform eigendecomposition on the data matrix to obtain a diagonal matrix of eigenvalues; Let the data matrix be Q, and perform eigendecomposition on the data matrix Q: C=QQ T =SΛS T (5), Where C is the covariance matrix of Q, Represents a diagonal matrix composed of the covariance matrix eigenvalues ​​arranged from large to small. represents the i-th eigenvalue, S represents the characteristic matrix; Step 220, calculating the data restoration rate of the first t eigenvalues; The data restoration rate of the first t eigenvalues ​​is calculated as: Step 230: If the data restoration rate is greater than a specified threshold, the detection data corresponding to the t+1th eigenvalue is discarded when constructing the temperature distribution; Step 240: Screen the detection nodes according to the comparison result, and use the spatial node arrangement of the remaining detection nodes as the second node spatial arrangement; Step 300: construct a second spatial temperature distribution using the temperature data of each detection node in the second node spatial distribution; The step 300 includes: Step 310, obtaining the detection matrix corresponding to the first t eigenvalues; Step 320: construct the second spatial temperature distribution using the feature matrix, the detection data and the detection matrix; Assume that the detection data is D, the detection data is D is the measured data set, and the detection matrix is ​​the standardized data matrix Q1, then the second spatial temperature distribution for: The data restoration rate is the ratio of the original data contained in the eigenvector corresponding to the converted eigenvalue.

2. A spatial temperature distribution optimization system, using the optimization method according to claim 1, characterized in that: include: First building block; Optimization module; The second building block; The first construction module is used to obtain spatial structural parameters of the lithium battery, construct an interpolation function using the spatial structural parameters, and construct a first spatial temperature distribution using the interpolation function; The optimization module is used to optimize the first node spatial distribution corresponding to the first spatial temperature distribution: calculate the data restoration rate of the eigenvalues ​​of the detected data matrix, and compare the data restoration rate with a specified threshold to determine the detection node to obtain the second node spatial distribution; The second construction module constructs a second spatial temperature distribution using the temperature data of each detection node in the second node spatial distribution; The data restoration rate is the ratio of the original data contained in the eigenvector corresponding to the converted eigenvalue.

3. The spatial temperature distribution optimization system according to claim 2, characterized in that: The first building block is further used for: The interpolation function is constructed with the Nth detection node as the interpolation center, and the detection data of the detection node is interpolated using the interpolation function to obtain the first spatial temperature distribution.

4. The spatial temperature distribution optimization system according to claim 3, characterized in that: The optimization module is further used for: The data matrix of the detection node data is subjected to eigendecomposition processing to obtain a diagonal matrix of eigenvalues, and the detection node is determined based on the data restoration rate of the first t eigenvalues ​​and the data restoration rate is compared with a specified threshold.

Images