A lithium battery health assessment method based on image processing

Abstract

A method for evaluating the health of lithium batteries based on image processing; the evaluation method is based on a lithium battery image processing platform, the platform comprising: a mechanism for collecting lithium battery electrical parameters, an image acquisition mechanism, an execution control mechanism, a local server, a cloud server, and a client access terminal; the local server generates judgment data after conversion based on lithium battery image information and transmits it to the cloud server via a wireless network; the cloud server receives the integrated data and calculates and outputs evaluation data for the lithium battery based on the temperature field edge distribution, the internal temperature field regional distribution, and the electrode column temperature; during the charging and discharging process of the lithium battery, the present invention detects the battery charging and discharging temperature distribution, determines the battery temperature change characteristics, determines the battery electrochemical conversion efficiency, estimates the battery life, and accurately estimates the battery replacement time.

Description

Technical field:

[0001] The present invention is applied to the fields of lithium battery pack temperature field detection and power supply electrical reliability detection; in particular, it relates to a lithium battery health assessment method based on image processing. Background technology:

[0002] Lithium batteries occupy the largest market share and have the widest application among power batteries and energy storage batteries.

[0003] Lithium batteries operate optimally between 20°C and 40°C, achieving peak charge and discharge efficiency within this temperature range. However, operating temperatures above 40°C or below 20°C can significantly reduce battery efficiency. Excessively low temperatures can significantly reduce battery lifespan, potentially causing irreversible damage or battery aging. Excessively high temperatures can lead to electrolyte decomposition and even fire. Therefore, accurately monitoring the internal temperature field of lithium battery packs is crucial to ensuring healthy battery operation.

[0004] Currently, battery pack temperature detection technology typically measures the external temperature of the battery pack as a basis for determining battery temperature. This method requires low detection accuracy and is effective when the battery pack temperature rises slowly. However, under high-power charging and discharging conditions, where the battery pack temperature rises rapidly and requires high battery pack temperature detection, there is currently no effective technology for detecting and calculating the battery pack temperature.

[0005] The ideal operating temperature of lithium batteries is 25℃±5℃. Only by accurately detecting the actual operating temperature of lithium batteries can we effectively ensure that lithium batteries operate in the optimal environment.

[0006] Lithium battery temperature field detection can only be achieved through sensors or image detection equipment, which collect the surface temperature distribution of the battery or battery pack to predict the internal temperature distribution and temperature gradient distribution. Battery temperature field distribution plays a vital role in estimating battery life and ensuring safe and reliable battery operation. Summary of the invention:

[0007] In order to solve the problems of the prior art, the present invention provides a lithium battery temperature field detection system and evaluation method based on image processing, which detects the temperature distribution of the battery pack by collecting the temperature field image of the outer surface of the battery pack. During the charging and discharging process, the battery voltage, current, internal resistance and other parameters, as well as the temperature field change gradient are measured to judge the battery charging efficiency and whether the battery has safety hazards. Furthermore, the health status of the battery pack is inferred through the algorithm, the performance of the battery pack is evaluated, and the health status of the battery pack is predicted. The battery surface temperature signal is collected by an infrared image camera to extract the battery surface temperature field distribution. According to the battery surface temperature field and the heating mechanism, the internal temperature field distribution of the battery is calculated. Based on the internal temperature field distribution of the battery, the charge and discharge current, the battery pack voltage, the ambient temperature, the battery internal resistance and other electrical parameters, the battery health status SOH is evaluated through a neural network.

[0008] The present invention solves the practical problem by adopting the following technical solutions:

[0009] A method for evaluating the health of lithium batteries based on image processing; the evaluation method is based on a lithium battery image processing platform, the platform comprising: a mechanism for collecting lithium battery electrical parameters, an image acquisition mechanism, an execution control mechanism, a local server, a cloud server, and a client access terminal; the local server generates judgment data after conversion based on lithium battery image information and transmits it to the cloud server via a wireless network; wherein:

[0010] The electrical parameter mechanism collects electrical information of the lithium battery temperature field through a Hall effect transmitter and an internal resistance detector;

[0011] The image acquisition mechanism acquires image information of the lithium battery temperature field through an 8k linear array camera;

[0012] The execution control mechanism processes the electrical information and image information through the PLC integrated circuit and transmits the integrated data to the local server;

[0013] The cloud server receives the integrated data and calculates and outputs evaluation data for the lithium battery according to the temperature field edge distribution, the internal temperature field regional distribution and the electrode column temperature;

[0014] The client accesses the lithium battery evaluation data presented to the human-machine interface by accessing the cloud server through a mobile device.

[0015] Furthermore, the server receives the integrated data and distributes the temperature field edge of the lithium battery, including the following steps:

[0016] 101. The cloud server uses the following formula to filter and denoise the image information in the integrated data to generate a denoised image f(x, y);

[0017] y=Med{x1,x2,…x n}

[0018]

[0019] Where: x1, x2, ... x n is the value of each point in the image sequence; Indicates the median of the image at that point when n is an odd number; Indicates the median of the image at that point when n is an even number.

[0020] 102. The cloud server extracts edges from the denoised image using the following formula to generate an image center edge point G(x,y);

[0021]

[0022] 103. The cloud server generates edge strength M(x,y) according to the central edge point using the following formula;

[0023]

[0024] 104. The cloud server obtains the edge position of the lithium battery temperature field according to the maximum intensity point of the edge intensity in the gradient direction of each point.

[0025] Furthermore, the cloud server receives the integrated data on the temperature distribution of the electrode column, which includes the following steps:

[0026] 201. The cloud server obtains the electrode column temperature model according to the following formula based on Fourier's law:

[0027]

[0028] Where i, j, k represent the x-axis, y-axis, and z-axis unit standard vectors respectively;

[0029] λ is the thermal conductivity of the object;

[0030] 202. The cloud server obtains the electrode column inflow temperature model according to the following formula based on Fourier's law:

[0031]

[0032]

[0033]

[0034] Where: Φ x Temperature flowing into the electrode column at any point in the x-axis direction; Φ y The temperature flowing into the electrode column at any point in the y-axis direction;

[0035] Φ z Temperature flowing into the electrode column at any point in the z-axis direction; xThe coefficient of temperature variation of the inflow in the x-axis direction at any point of the electrode column;

[0036] λ y Temperature variation coefficient of the flow in the y-axis direction at any point of the electrode column; λ z The coefficient of temperature variation of the flow in the z-axis direction at any point of the electrode column;

[0037] 203. The cloud server obtains the electrode column outflow temperature model according to the following formula based on Fourier's law:

[0038]

[0039]

[0040]

[0041] Where: Φ x+dx The outflow temperature at any point of the electrode column in the x-axis direction; Φ y+dy The outflow temperature in the y-axis direction at any point of the electrode column;

[0042] Φ z+dz The outflow temperature in the z-axis direction at any point of the electrode column;

[0043] 204. The cloud server obtains the internal temperature model of the battery pole according to the following formula based on Fourier's law

[0044]

[0045]

[0046]

[0047] Where: ΔT x Temperature rise in the x-axis direction at any point of the electrode column; ΔT y Temperature rise in the y-axis direction at any point on the electrode column; ΔT z Temperature rise in the z-axis direction at any point of the electrode column;

[0048] 205. The cloud server obtains the temperature distribution of the battery pole according to the following formula based on Fourier's law:

[0049]

[0050] Where, ρ is the battery pole density;

[0051] C is the heat capacity of the battery pole;

[0052] Φ V Generates heat at any point in the battery.

[0053] Furthermore, the cloud server receives the integrated data to distribute the temperature field inside the lithium battery; the process includes the following steps:

[0054] 301. The cloud server establishes the finite element equation through the three-dimensional steady state of the rectangular coordinate system:

[0055]

[0056] The thermal density boundary condition is

[0057]

[0058] Convective heat transfer boundary conditions:

[0059]

[0060] 302. The cloud server obtains the internal temperature field distribution of the battery according to the following formula using the finite element equation:

[0061]

[0062] Among them, T i is the internal temperature of the battery; Ω is the heat conduction area of ​​the battery; τ1 is the constant heat flow temperature; τ1 is the constant heat flow temperature; τ1 is the constant heat flow temperature; τ2 is the convection heat transfer temperature; τ2 is the convection heat transfer temperature; τ2 is the convection heat transfer temperature; τ2 is the convection heat transfer temperature; h is the convection heat transfer coefficient; N i is the weight function; Φ a is the initial temperature of the point; further, the cloud server outputs the lithium battery evaluation data process through the neural network; comprising the following steps:

[0063] 401. The cloud server extracts the internal temperature field data of the battery pack according to the differential equation and inputs it into the neural network;

[0064] 402. The neural network outputs evaluation battery data according to the following formula:

[0065] x i ={T i , ΔT k ,l,m,n,T h , U j , I r , R s}

[0066] y o ={SOH}.

[0067] Among them, T i is the internal temperature of the battery pack; ΔT kis the temperature rise of the battery pole; l, m, n are the structural dimensions of the battery pack (length, width, height); T h is the ambient temperature; U j is the battery pack voltage; I r is the battery pack current; R s is the internal resistance of the battery;

[0068] Beneficial effects:

[0069] This method detects the battery pack's temperature distribution by capturing images of the battery pack's external surface temperature field. During the charge and discharge process, battery parameters such as voltage, current, and internal resistance, as well as the temperature field gradient, are measured to determine charging efficiency and whether there are safety hazards. Furthermore, an algorithm is used to infer the battery pack's health status, evaluate its performance, and predict its health.

[0070] During the charging and discharging process of a lithium battery, the present invention detects the battery charging and discharging temperature distribution, determines the battery temperature change characteristics, determines the battery electrochemical conversion efficiency, estimates the battery life, and accurately estimates the battery replacement time.

[0071] The present invention adopts edge cloud technology to collect local information in real time, analyze the data in a timely manner, remotely realize big data sample learning, and update the mathematical model parameters of the local server online to realize dynamic model updating.

[0072] In the application scenario of pure electric vehicle power batteries, the present invention monitors the temperature field distribution of lithium batteries during operation to prevent the temperature of the battery pack from getting out of control, which may cause serious hazards such as deflagration and explosion of the battery pack. Description of the drawings:

[0073] Figure 1 This is a schematic diagram of the structure of the lithium battery temperature field detection system based on image processing of the present invention;

[0074] Figure 2 This is a flow chart of the lithium battery temperature field detection based on image processing in the present invention. DETAILED DESCRIPTION

[0075] The present invention is described in detail below with reference to the accompanying drawings:

[0076] The present invention provides a method for evaluating the health of lithium batteries based on image processing. The method is based on a lithium battery image processing platform and includes: a mechanism for collecting lithium battery electrical parameters, an image acquisition mechanism, an execution control mechanism, a local server, a cloud server, and a client access terminal. The local server generates judgment data after converting lithium battery image information and transmits it to the cloud server via a wireless network.

[0077] The electrical parameter mechanism collects electrical information of the lithium battery temperature field through a Hall effect transmitter and an internal resistance detector;

[0078] The image acquisition mechanism acquires image information of the lithium battery temperature field through an 8k linear array camera;

[0079] The execution control mechanism processes the electrical information and image information through the PLC integrated circuit and sends the integrated data to the local server; wherein: the integrated data includes:

[0080] The cloud server receives the integrated data and calculates and outputs evaluation data for the lithium battery according to the temperature field edge distribution, the internal temperature field regional distribution and the electrode column temperature;

[0081] The client accesses the lithium battery evaluation data presented to the human-machine interface by accessing the cloud server through a mobile device.

[0082] The cloud server receives integrated data and calculates and outputs evaluation data for the lithium battery according to the temperature field edge distribution, the internal temperature field regional distribution and the electrode column temperature; comprising the following steps:

[0083] The cloud server receives the integrated data and calculates the temperature field edge distribution process of the lithium battery:

[0084] 101. The cloud server uses the following formula to filter and denoise the image information in the integrated data to generate a denoised image f(x, y);

[0085] y=Med{x1,x2,…x n}

[0086]

[0087] Where: x1, x2, ... x n is the value of each point in the image sequence.

[0088] Indicates the median of the image at that point when n is an odd number.

[0089] Indicates the median of the image at that point when n is an even number.

[0090] 102. The cloud server extracts edges from the denoised image using the following formula to generate an image center edge point G(x, y);

[0091]

[0092] 103. The cloud server generates edge strength M(x, y) according to the central edge point using the following formula;

[0093]

[0094] 104. The cloud server obtains the edge position of the lithium battery temperature field according to the maximum intensity point of the edge intensity in the gradient direction of each point.

[0095] The cloud server receives the integrated data on the temperature distribution of the electrode column:

[0096] 201. The cloud server obtains the electrode column temperature model according to the following formula based on Fourier's law:

[0097]

[0098] Where i, j, k represent the x-axis, y-axis, and z-axis unit standard vectors respectively;

[0099] λ is the thermal conductivity of the object;

[0100] 202. The cloud server obtains the electrode column inflow temperature model according to the following formula based on Fourier's law:

[0101]

[0102]

[0103]

[0104] Where: Φ x The temperature flowing into the electrode column at any point in the x-axis direction;

[0105] Φ y The temperature flowing into the electrode column at any point in the y-axis direction;

[0106] Φ z Temperature flowing into the electrode column at any point in the z-axis direction;

[0107] λ x The coefficient of temperature variation of the inflow in the x-axis direction at any point of the electrode column;

[0108] λ y The temperature variation coefficient of the flow in the y-axis direction at any point of the electrode column;

[0109] λ z The coefficient of temperature variation of the flow in the z-axis direction at any point of the electrode column;

[0110] 203. The cloud server obtains the electrode column outflow temperature model according to the following formula based on Fourier's law:

[0111]

[0112]

[0113]

[0114] Where: Φ x+dx The outflow temperature in the x-axis direction at any point of the electrode column;

[0115] Φ y+dy The outflow temperature in the y-axis direction at any point of the electrode column;

[0116] Φ z+dz The outflow temperature in the z-axis direction at any point of the electrode column;

[0117] 204. The cloud server obtains the internal temperature model of the battery pole according to the following formula based on Fourier's law

[0118]

[0119]

[0120]

[0121] Where: ΔT x Temperature rise in the x-axis direction at any point of the electrode column;

[0122] ΔT y Temperature rise in the y-axis direction at any point of the electrode column;

[0123] ΔT z Temperature rise in the z-axis direction at any point of the electrode column;

[0124] 205. The cloud server obtains the temperature distribution of the battery pole according to the following formula based on Fourier's law:

[0125]

[0126] Where, ρ is the battery pole density;

[0127] C is the heat capacity of the battery pole;

[0128] Φ V Generates heat at any point in the battery.

[0129] The cloud server receives the integrated data to analyze the internal temperature field distribution of the lithium battery:

[0130] 301. The cloud server establishes the finite element equation through the three-dimensional steady state of the rectangular coordinate system:

[0131]

[0132] The thermal density boundary condition is

[0133]

[0134] Convective heat transfer boundary conditions:

[0135]

[0136] 302. The cloud server obtains the internal temperature field distribution of the battery according to the following formula using the finite element equation:

[0137]

[0138] Among them, T i is the internal temperature of the battery;

[0139] Ω is the battery heat conduction area;

[0140] τ1 is the constant heat flow temperature;

[0141] τ2 is the convective heat transfer temperature;

[0142] h is the convective heat transfer coefficient;

[0143] N i is the weight function;

[0144] Φ a is the initial temperature of the point;

[0145] The cloud server outputs the lithium battery evaluation data process through the neural network:

[0146] 401. The cloud server extracts the internal temperature field data of the battery pack (the temperature at the center of the positive terminal and the negative terminal of the battery pack, T i , temperature rise ΔT k ), combined with the battery pack structure size (l, m, n), ambient temperature (T h ), battery pack voltage (U j ), current (I r ), internal resistance (R s ) as the neural network input.

[0147] 402. The neural network outputs evaluation battery data according to the following formula:

[0148] x i ={T i , ΔT k ,l,m,n,T h , U j , I r , R s}

[0149] y o ={SOH}

[0150] Among them, T i is the internal temperature of the battery pack; ΔT k is the temperature rise of the battery pole; l, m, n are the structural dimensions of the battery pack (length, width, height); T his the ambient temperature; U j is the battery pack voltage; I r is the battery pack current; R s is the internal resistance of the battery.

[0151] The present invention is not limited to the embodiments described above. The above description of the specific embodiments is intended to describe and illustrate the technical solutions of the present invention. The above specific embodiments are merely illustrative and not restrictive. Without departing from the scope of the present invention and the scope of protection of the claims, those skilled in the art may make various specific modifications based on the teachings of the present invention, all of which fall within the scope of protection of the present invention.

Claims

1. A method for evaluating the health of lithium batteries based on image processing; characterized by: The evaluation method is based on a lithium battery image processing platform, which includes: a lithium battery electrical parameter acquisition mechanism, an image acquisition mechanism, an execution control mechanism, a local server, a cloud server, and a client access terminal. The local server generates judgment data after conversion based on lithium battery image information and transmits it to the cloud server via a wireless network; wherein: The electrical parameter mechanism collects electrical information of the lithium battery temperature field through a Hall effect transmitter and an internal resistance detector; The image acquisition mechanism acquires image information of the lithium battery temperature field through an 8k linear array camera; The execution control mechanism processes the electrical information and image information through the PLC integrated circuit and transmits the integrated data to the local server; The cloud server receives the integrated data and calculates and outputs evaluation data for the lithium battery according to the temperature field edge distribution, the internal temperature field regional distribution and the electrode column temperature; The client accesses the lithium battery evaluation data presented to the human-machine interface by accessing the cloud server through a mobile device.

2. The method for lithium battery health assessment based on image processing according to claim 1, characterized in that: The cloud server receives the integrated data and distributes the lithium battery according to the edge of the temperature field, including the following steps: The cloud server 101 uses the following formula to filter and denoise the image information in the integrated data to generate a denoised image f(x, y); y=Med{x1,x2,…x n }; Where: x1, x2, ... x n is the value of each point in the image sequence; Indicates the median of the image at that point when n is an odd number; Indicates the median of the image at that point when n is an even number; 102 The cloud server uses the following formula to extract the edge of the denoised image to generate the image center edge point G(x, y); 103 The cloud server generates edge strength M(x, y) according to the central edge point using the following formula; 104 The cloud server obtains the edge position of the lithium battery temperature field according to the maximum intensity point of the edge intensity in the gradient direction of each point.

3. The method for lithium battery health assessment based on image processing according to claim 2, characterized in that: The cloud server receives integrated data on the temperature distribution of the electrode column; the process includes the following steps: The 201 cloud server obtains the electrode column temperature model according to the following formula based on Fourier's law: Where i, j, k represent the unit standard vectors of x, y, and z axes respectively; λ is the thermal conductivity of the object; The 202 cloud server obtains the electrode column inflow temperature model according to the following formula based on Fourier's law: Where: Φ x Temperature flowing into the electrode column at any point in the x-axis direction; Φ y The temperature flowing into the electrode column at any point in the y-axis direction; Φ z Temperature flowing into the electrode column at any point in the z-axis direction; x The coefficient of temperature variation of the inflow in the x-axis direction at any point of the electrode column; λ y Temperature variation coefficient of the flow in the y-axis direction at any point of the electrode column; λ z The coefficient of temperature variation of the flow in the z-axis direction at any point of the electrode column; The 203 cloud server obtains the electrode column outflow temperature model according to the following formula based on Fourier's law: Where: Φ x+dx The outflow temperature at any point of the electrode column in the x-axis direction; Φ y+dy The outflow temperature in the y-axis direction at any point of the electrode column; Φ z+dz The outflow temperature in the z-axis direction at any point of the electrode column; The 204 cloud server obtains the internal temperature model of the battery pole according to the following formula based on Fourier's law: Where: ΔT x Temperature rise in the x-axis direction at any point of the electrode column; ΔT y Temperature rise in the y-axis direction at any point on the electrode column; ΔT z Temperature rise in the z-axis direction at any point of the electrode column; The 205 cloud server obtains the regional distribution of battery pole temperature using the following formula based on Fourier's law: Where ρ is the battery pole density; C is the battery pole heat capacity; Φ V Generates heat at any point in the battery.

4. The method for lithium battery health assessment based on image processing according to claim 3, characterized in that: The cloud server receives the integrated data to distribute the internal temperature field of the lithium battery; the process includes the following steps: The 301 cloud server establishes the finite element equation through the three-dimensional steady state of the rectangular coordinate system: The thermal density boundary condition is Convective heat transfer boundary conditions: The 302 cloud server obtains the internal temperature field distribution of the battery according to the following formula through the finite element equation: Among them, T i is the internal temperature of the battery; Ω is the heat conduction area of ​​the battery; τ1 is the constant heat flow temperature; τ2 is the convection heat transfer temperature; h is the convection heat transfer coefficient; N i is the weight function; Φ a is the initial temperature at this point.

5. The method for lithium battery health assessment based on image processing according to claim 4, characterized in that: The cloud server outputs the lithium battery evaluation data through the neural network; the process includes the following steps: The 401 cloud server extracts the internal temperature field data of the battery pack based on the differential equation and inputs it into the neural network; 402 The neural network outputs the evaluation battery data according to the following formula: x i ={T i ,ΔT k ,l,m,n,T h ,U j ,I r ,R s }; and o ={SOH}; Among them, T i is the internal temperature of the battery pack; ΔT k is the temperature rise of the battery pole; l, m, n are the structural dimensions of the battery pack (length, width, height); T h is the ambient temperature; U j is the battery pack voltage; I r is the battery pack current; R s is the internal resistance of the battery.

Images