WO2020051738A1 - 智能电池热管理方法及系统 - Google Patents

智能电池热管理方法及系统 Download PDF

Info

Publication number
WO2020051738A1
WO2020051738A1 PCT/CN2018/104846 CN2018104846W WO2020051738A1 WO 2020051738 A1 WO2020051738 A1 WO 2020051738A1 CN 2018104846 W CN2018104846 W CN 2018104846W WO 2020051738 A1 WO2020051738 A1 WO 2020051738A1
Authority
WO
WIPO (PCT)
Prior art keywords
battery
thermal
parameters
intelligent battery
control
Prior art date
Application number
PCT/CN2018/104846
Other languages
English (en)
French (fr)
Inventor
王玲
王�锋
柯瑞林
柯冰
Original Assignee
深圳市欧姆阳科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市欧姆阳科技有限公司 filed Critical 深圳市欧姆阳科技有限公司
Priority to PCT/CN2018/104846 priority Critical patent/WO2020051738A1/zh
Publication of WO2020051738A1 publication Critical patent/WO2020051738A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/10—Energy storage using batteries

Definitions

  • the invention belongs to the technical field of battery thermal management, and particularly relates to a method and system for intelligent battery thermal management.
  • the battery is the core component of a new energy vehicle, and its performance directly affects the working condition of the vehicle.
  • the battery needs to be charged and discharged with a large current, and a violent electrochemical reaction occurs inside the battery. During this process, a large amount of heat is generated, which causes a sharp change in the battery temperature.
  • the single battery has an optimal operating temperature range. Due to the difference in electrochemical performance, The optimal temperature range for different types of batteries is different. For example, the optimal temperature range for lead-acid batteries is 25 to 45 ° C, and the optimal operating temperature range for nickel-metal hydride batteries is 20 to 40 ° C. For every 10 ° C increase in battery temperature, the battery The internal electrochemical reaction rate increases exponentially.
  • the temperature rise inside the battery will also have a large impact on the charge and discharge capacity and power.
  • the reaction inside the battery is a change of the active material to an inert state, the degradation of the battery's charge and discharge capacity and power is unavoidable under any conditions, but the degradation is particularly severe under high temperature environments.
  • the thermal resistance will increase sharply, so the open circuit voltage (that is, the operating voltage) of the battery will be greatly reduced, and the work that can be output to the outside will be correspondingly reduced, which will cause a sharp degradation in charge and discharge capacity and power.
  • New energy battery packs are usually closely packed with a large number of battery cells, and the accumulated heat is difficult to discharge.
  • the battery temperature may exceed the normal range during long-term work, and changes in battery temperature directly affect battery safety and cycle life. Discharge capacity, charge and discharge efficiency, and other properties can cause thermal failure in severe cases, affecting the safety and reliability of the battery, especially the acceleration of the harmful internal chemical reaction rate, which will permanently destroy the battery structure and reduce the battery's working life. , Which in turn affects the performance of the vehicle.
  • the thermal management method of the battery system is the core content of power management.
  • the management method and the accuracy and performance of the control output parameters determine the performance of the new energy vehicle. It is necessary to add a thermal management system to the battery pack of the vehicle.
  • the present invention proposes an adaptive deep learning network and its application.
  • the adaptive deep learning network obtains the thermal energy parameter control output equation through an adaptive fitting method of forward thermal parameters and thermal performance distribution;
  • the configuration of the thermal control parameters is formed by the inverse control parameters, and different cooling methods are established, such as using wind, liquid, electricity, etc. to adjust the battery temperature.
  • the invention provides a battery thermal management method based on deep learning multilayer network modeling, which includes the following steps:
  • the thermal physical parameters include the density and specific heat capacity of the battery.
  • T is the temperature of the battery
  • t is the time
  • ⁇ is the average density of the material inside the battery
  • q is the heat generation rate per unit volume of the battery
  • c p is the constant pressure specific heat capacity of the battery
  • ⁇ x , ⁇ y , ⁇ z are The thermal conductivity in the three-dimensional orthogonal direction can be solved to obtain the temperature field distribution of the battery;
  • a thermal effect model of the battery pack under a deep learning multilayer network is formed.
  • the thermal energy parameter control output equation is obtained through the adaptive fitting method of the forward parameters, wherein the forward parameters include the heat generation, thermal performance distribution, and thermal physical parameters; through the reverse refrigeration control parameters, The configuration of the parameters of the thermal control equation is formed, and different cooling modes are established to regulate and control the temperature of the battery.
  • the invention also discloses a battery thermal management system based on deep learning multi-layer network modeling, which includes a front-end sensor, a battery management controller, and a cooling device.
  • the front-end sensor is arranged in a battery box for obtaining forward parameters.
  • the forward parameters include heat generation, thermal performance distribution, and thermal physical parameters;
  • the battery management controller includes a high-speed digital signal processing kernel and an ARM core, and the high-speed digital signal processing kernel is used to perform deep learning SAE fitting networks,
  • the ARM core outputs a cooling control parameter according to a fuzzy decision made by a battery thermal management model, and the cooling device is used for cooling and cooling the battery pack.
  • the front-end sensor includes a temperature sensor and a density sensor.
  • the front-end sensor is connected to the battery management controller through a signal conditioner.
  • the communication networking mode between the battery management controller and the unit management unit is a bus.
  • the cooling device used in the cooling method includes at least one of an air cooling device, a liquid cooling device, a heat pipe cooling device, and a phase change material cooling device.
  • the invention adopts the framework of a deep learning multilayer network, and through the SAE network fitting method, combines the heat output of the battery, the heat generation performance distribution, the thermophysical parameters and the governing equations to form the thermal effect of the battery pack under the deep learning multilayer network
  • the model is obtained by fitting according to the control equation of the thermal energy parameter.
  • the temperature of the battery is adjusted using adaptive reverse conduction thermal management to reduce damage to the battery life, discharge capacity, etc. , Effectively perform thermal control, maintain the stability and safety of battery operation, and then effectively ensure the safety of the operation of battery-powered systems such as new energy vehicles.
  • FIG. 1 is a schematic diagram of an embodiment of a battery thermal management method provided by the present invention
  • Figure 2 is a schematic diagram of a multilayer SAE (Stacked Autoencoder) network model
  • FIG. 3 is a schematic diagram of an embodiment of a battery thermal management method and corresponding hardware provided by the present invention.
  • FIG. 4 is a schematic diagram of a deep learning reverse parameter adjustment and implementation method architecture based on a multilayer SAE network.
  • a battery thermal management method based on deep learning multilayer network modeling includes the following steps:
  • T is the temperature of the battery
  • t is the time
  • ⁇ is the average density of the battery's internal material
  • q is the heat generation rate per unit volume of the battery
  • c p is the constant pressure specific heat capacity of the battery
  • ⁇ x , ⁇ y , ⁇ z are the battery
  • the thermal effect of the battery pack under the deep learning multilayer network is formed. model
  • the thermal energy parameter control output equation is obtained through the adaptive fitting method of the forward parameters, wherein the forward parameters include the heat generation, thermal performance distribution, and thermal physical parameters; through the reverse refrigeration control parameters, The configuration of the parameters of the thermal control equation is formed, and different cooling methods are established to regulate the temperature of the battery, thereby learning the SAE network parameter unsupervised control parameters.
  • step S2 the three-dimensional thermal effect model of the battery, that is, the three-dimensional thermal effect equation of the battery, is solved.
  • the solution obtained is the temperature field distribution of the battery.
  • the battery targeted in the embodiments of the present invention is preferably a lithium-ion battery.
  • the cooling device that can be used is at least one of an air cooling device, a liquid cooling device, a heat pipe cooling device, and a phase change material cooling device, and is preferably an air cooling device, that is, a gas cooling method.
  • the present invention discloses a battery thermal management system based on deep learning multilayer network modeling, which includes a front-end sensor, a battery management controller, a driving device, and a cooling device.
  • the front-end sensor is set in a battery case for obtaining forward parameters.
  • the battery management controller includes a high-speed digital signal processing kernel and an ARM core.
  • the high-speed digital signal processing kernel is used to perform unsupervised self-learning of the SAE network.
  • the control parameters are transmitted to the ARM core through the parallel bus, and the ARM core completes the precision control of the battery.
  • the sensor's feedback after the precision control achieves the balance of the battery temperature.
  • the high-speed digital signal processing kernel is used to perform a deep learning SAE fitting network in steps S2, S3, and S4, including calculating a thermal efficiency parameter fitting equation, establishing a battery thermal effect management model, and deep learning SAE reverse tuning parameters.
  • the ARM core makes a fuzzy decision according to the battery thermal management model and outputs the refrigeration control parameters.
  • the ARM core completes the accuracy control of the battery, and completes the balance of the battery temperature through the sensor feedback after the accuracy control.
  • the high-speed digital signal processing core and the ARM core both implement communication between the two cores by sharing a memory; the driving device receives a cooling control parameter instruction of the ARM core; and the cooling device drives the battery pack under the driving of the driving device. Cooling and cooling.
  • the sequence of steps in the high-speed digital signal processing kernel is shown in Figure 3.
  • the thermal efficiency parameter fitting equation is processed, and the result is passed to the thermal management model of the power supply.
  • the thermal management model reverse-tunes the deep learning SAE and passes the results to the deep learning SAE.
  • the above steps can be cyclically performed to complete the continuous learning and improvement process.
  • the battery thermal effect model is obtained by fitting.
  • the advantage is that when the temperature exceeds the maximum allowable operating temperature, the adaptive control method is adopted and the precise temperature feedback control method is adopted.
  • the prior knowledge of the temperature control is
  • the high-speed digital signal processing kernel is used to perform a deep learning SAE fitting network in steps S2, S3, and S4, including calculating a thermal efficiency parameter fitting equation and establishing a battery thermal effect management model.
  • the temperature feedback control can effectively adjust the ambient temperature of the battery at the working time, thereby extending the battery's reasonable time in the ambient temperature environment, thereby improving the effective working cycle and time of the battery, and thereby increasing the battery's service life.
  • the cycle of the charge and discharge capacity of the battery will increase, which will reduce the frequency of charge and discharge of the battery capacitor and reduce the damage to the capacity, that is, the reduction of the battery life and discharge capacity will cause damage.
  • cold air flows in from one side of the battery pack, flows out from the other side, and passes through the battery module in turn.
  • the temperature difference between the air and the battery gradually decreases, the heat exchange capacity decreases, and the cooling
  • the temperature of the air is gradually increasing, and the temperature difference between the front and the back of the air duct when using serial ventilation cooling is very large.
  • the average temperature of the battery is accurately controlled to obtain the air speed control parameters and average temperature of serial ventilation. Of balance.
  • Figure 2 shows a multilayer SAE (Stacked Autoencoder) network model.
  • SAE fitting network solves the thermal efficiency parameter fitting equation, so as to make the next fuzzy decision.
  • SAE uses the multilayer nonlinear function of the multilayer neural network and the sigmod function as the basic unit to fit the thermal effect model of the battery, accurately obtain the nonlinear characteristics of the thermal effect model, and then achieve precise control.
  • the front-end sensor includes a temperature sensor and a density sensor.
  • the front-end sensor is connected to the battery management controller through a signal conditioner.
  • the communication networking mode between the battery management controller and the unit management unit is a bus.
  • the communication networking mode between the battery management controller and the single management unit may also be in a daisy chain form.
  • the cooling device used in the cooling method includes an air cooling device, a liquid cooling device, a heat pipe cooling device, and a phase change material cooling device. It is preferably an air-cooled device or mainly an air-cooled device.
  • the embodiment of the present invention relates to an adaptive deep learning network.
  • the thermal energy parameter control output equation is obtained through an adaptive fitting method of forward thermal parameters and thermal performance distributions; the configuration of the thermal control parameters is formed by the inverse control parameters. , Establish different cooling methods.
  • An embodiment of the present invention also provides a framework using a deep learning multilayer network, and through the SAE network fitting method, combined with the heat generation of the battery, the heat generation performance distribution, the thermophysical property parameters, and the governing equations, a deep learning in the multilayer is formed.
  • the invention also provides a precise control method of refrigeration parameters, which detects the reverse conduction parameters and reconstructs the thermal effect equation of the battery based on the combination of the control equations and the definite solution conditions of the corresponding model to form air cooling
  • the cooling thermal model controls the wind speed and air inlet angle through reverse conduction management parameters to achieve accurate refrigeration control.

Landscapes

  • Secondary Cells (AREA)
  • Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Electrochemistry (AREA)
  • General Chemical & Material Sciences (AREA)

Abstract

本发明提供了一种智能电池热管理方法及系统，所述方法包括以下步骤：S1、测量电池的热物性参数，其中包括电池的密度、比热容；S2、根据电池热物性参数，建立电池的三维热效应方程，求解电池的三维热效应方程，得到的解就是电池的温度场分布情况；S3、建立电池热效应模型；S4、根据电池的热效应模型，通过前向参数的自适应拟合方法，其中前向参数包括产热量、热性能分布、热物性参数，获取热能参数控制输出方程；通过反向的制冷控制参数，形成热控制方程参数的配置，建立不同的冷却方式对电池进行温度调控。本发明采用自适应反向传导热管理方式对电池进行温度调控，可有效进行热控制，维持电池运行的稳定性和安全性。

Description

智能电池热管理方法及系统 技术领域
本发明属于电池热管理技术领域，特别涉及一种智能电池热管理方法及系统。
背景技术
电池是新能源汽车的核心部件，其性能的好坏直接影响到汽车的工作情况。电池需要进行大电流的充放电，其内部发生剧烈的电化学反应，此过程中会产生大量的热量，导致电池温度的急剧变化，单体电池有最佳工作温度范围，由于电化学性能的差异，不同种类电池的最佳温度范围不同，例如铅酸电池的最佳温度范围为25～45℃，而镍氢电池的最佳工作温度范围为20～40℃，电池温度每上升10℃，电池内部的电化学反应速率就会成倍增加。电池内部的温升亦会对充放电容量和功率造成大的影响。由于电池内部的反应是活性物质向惰性状态的变化，所以在任何条件下，电池的充放电容量和功率的降级是不可避免的，但是在高温环境下的降级尤其剧烈，此时电池内部的欧姆热阻将急剧增加，这样电池的开路电压(即工作电压)就会大幅度的降低，对外所能输出的功就相应的减少，进而导致充放电容量和功率的急剧降级。
通常新能源电池组由大量的电池单体紧密排布在一起，聚集的热量难以排出，长时间工作时电池温度可能超出正常范围，而电池温度的变化又直接影响着电池的安全性、循环寿命、放电容量及充放电效率等性能，严重时会导致热失效，影响电池的安全性和可靠性，尤其内部有害化学反应速度的加快，将永久性的破坏电池的结构，进而减少电池的工作寿命，进而影响整车的工作性能。
因此，电池系统的热管理方法是电源管理的核心内容，其管理方法及控制输出参数的精度及性能决定着新能源汽车整车的性能，在车用电池组中加入热管理系统很有必要。
发明内容
为解决以上问题，本发明提出一种自适应深度学习网络及其应用，该自适应深度学习网络通过前向的热参数，热性能分布的自适应拟合方法，获取热能 参数控制输出方程；还通过反向的控制参数形成热控制参数的配置，建立不同的冷却方式，如利用风，液体，电等来进行电池温度调节的一种基于深度学习多层网络建模的电池热管理方法及系统。
本发明提供的一种基于深度学习多层网络建模的电池热管理方法，包括以下步骤：
S1、测量电池的热物性参数，所述热物性参数包括电池的密度、比热容；
S2、根据电池热物性参数，建立电池的三维热效应方程，对应的公式如下，
Figure PCTCN2018104846-appb-000001
其中，T为电池的温度，t为时间，ρ为电池内部材料的平均密度，q为电池单位体积产热速率，c p为电池的定压比热容，λ x、λ y、λ z是电池在三维正交方向上的导热系数，求解得电池的温度场分布情况；
S3、建立电池热效应模型
根据电池的基本参数，包括标称容量、标称电压、放电截止电压、充电截止电压、内阻参数，结合电池的串联及并联热分布方式，形成在深度学习多层网络下电池组的热效应模型；
S4、根据电池的热效应模型，通过前向参数的自适应拟合方法，获取热能参数控制输出方程，其中前向参数包括产热量、热性能分布、热物性参数；通过反向的制冷控制参数，形成热控制方程参数的配置，建立不同的冷却方式对电池进行温度调控。
本发明还公开了一种基于深度学习多层网络建模的电池热管理系统，包括前端传感器、电池管理控制器以及冷却装置，所述前端传感器设置于电池箱体中用于获取前向参数，所述前向参数包括产热量、热性能分布、热物性参数；所述电池管理控制器包括高速数字信号处理内核和ARM内核，所述高速数字信号处理内核用于进行深度学习SAE拟合网络，所述ARM内核根据电池热管理模型作出的模糊决策，输出制冷控制参数，所述冷却装置用于对电池组进行制冷散热。
其中，所述前端传感器包括温度传感器、密度传感器。
所述的前端传感器通过信号调节器与电池管理控制器连接。
具体地，所述电池管理控制器与单体管理单元之间的通讯组网方式为总线形式。
其中，所述的冷却方式采用的冷却装置包括风冷装置、液体冷却装置、热管冷却装置和相变材料冷却装置中的至少一种。
本发明的有益效果如下：
本发明采用深度学习多层网络的框架，通过SAE网络拟合方法，结合电池的产热量、产热性能分布、热物性参数和控制方程等，形成了在深度学习多层网络下电池组的热效应模型，该模型根据热能参数的控制方程拟合获取，当温度超过了最大允许的工作温度，采用自适应反向传导热管理方式对电池进行温度调控，降低对电池的寿命、放电容量等造成损害，有效进行热控制，维持电池运行的稳定性和安全性，进而有效保证电池供电系统例如新能源汽车的运行安全。
附图说明
下面结合附图，通过对本发明的具体实施方式详细描述，将使本发明的技术方案及其它有益效果显而易见。
图1为本发明提供的电池热管理方法实施例示意图；
图2为多层SAE(Stacked Autoencoder)网络模型示意图；
图3为本发明提供的电池热管理方法以及对应硬件的实施例示意图；
图4为基于多层SAE网络的深度学习反向调参及实施方法架构示意图。
具体实施方式
为更进一步阐述本发明所采取的技术手段及其效果，以下结合本发明的优选实施例及其附图进行详细描述。
请参考图1-4，本发明提供的一种基于深度学习多层网络建模的电池热管理方法，包括以下步骤：
S1、测量电池的热物性参数，包括电池的密度、比热容等；
S2、根据电池热物性参数，建立电池的三维热效应方程，公式如下，
Figure PCTCN2018104846-appb-000002
式中，T为电池的温度，t为时间，ρ为电池内部材料的平均密度，q为电 池单位体积产热速率，c p为电池的定压比热容，λ x、λ y、λ z是电池在三维正交方向上的导热系数，求解得电池的温度场分布情况；
S3、建立电池热效应模型
根据电池的基本参数，如标称容量、标称电压、放电截止电压、充电截止电压、内阻参数，同时结合电池的串联及并联热分布方式，形成在深度学习多层网络下电池组的热效应模型；
S4、根据电池的热效应模型，通过前向参数的自适应拟合方法，获取热能参数控制输出方程，其中前向参数包括产热量、热性能分布、热物性参数；通过反向的制冷控制参数，形成热控制方程参数的配置，建立不同的冷却方式对电池进行温度调控，由此进行SAE网络参数无监督控制参数学习。
上述步骤S2中，所建立的电池的三维热效应模型，即电池的三维热效应方程，对该三维热效应方程求解，得到的解就是电池的温度场分布情况。本发明实施例中针对的电池优选为锂离子电池。
上述冷却方式中，可以采用的冷却装置为风冷装置、液体冷却装置、热管冷却装置和相变材料冷却装置中的至少一种，优选为风冷装置，即气体冷却方式。在本发明更具体的实施方式中，请参考图3，本发明公开了一种基于深度学习多层网络建模的电池热管理系统，包括前端传感器、电池管理控制器、驱动装置以及冷却装置，所述前端传感器设置于电池箱体中用于获中的前向参数；所述电池管理控制器包括高速数字信号处理内核和ARM内核，采用高速数字信号处理内核，进行SAE网络的无监督自学习，通过并行总线将控制参数传到ARM内核，同时由ARM内核完成对电池的精度控制，通过精度控制后的传感器反馈，实现电池温度的平衡。
具体地，所述高速数字信号处理内核用于在步骤S2、S3和S4中进行深度学习SAE拟合网络，包括计算热效能参数拟合方程，建立电池热效应管理模型、深度学习SAE反向调参，所述ARM内核根据电池热管理模型作出模糊决策，输出制冷控制参数，ARM内核完成电池的精度控制，通过精度控制后的传感器反馈，完成电池温度的平衡。所述高速数字信号处理内核和ARM内核两者通过共享内存的方式实现双核之间的通讯；所述驱动装置接收ARM内核的制冷控制参数指令；所述冷却装置在驱动装置的驱动下对电池组进行制冷散热。
在高速数字信号处理内核中各步骤的顺序如图3所示，深度学习SAE拟合网络接收到上游传递的信息后进行热效能参数拟合方程的处理，结果传递给电源热管理模型，该电源热管理模型对深度学习SAE进行反向调参，把结果传递给深度学习SAE。上述步骤可以循环进行，以完成不断的学习完善的过程。
根据热能参数的控制方程拟合获取电池热效应模型，其优势在于，当温度超过了最大允许的工作温度，通过自适应控制方式，采用精确的温度反馈控制方式，该温度控制的先验知识是通过高速数字信号处理内核用于在步骤S2、S3和S4中进行深度学习SAE拟合网络，包括计算热效能参数拟合方程，建立电池热效应管理模型获取的。该温度反馈控制可以有效调整电池在工作时刻的环境温度，进而将电池在合理的温度环境时间延长，从而提升电池的有效工作周期及时间，进而提高电池的使用寿命。由于环境工作的温度较为合理，因此，电池的电容充放电量周期增加，进而降低电池电容的充放电的频率，降低对电容量的损害，即降低对电池的寿命、放电容量等将造成损害，并且采用ARM内核自适应反向传导热管理方式对电池进行温度调控。
根据一个实施例，采用串行通风风冷时，冷空气从电池组的一侧流入，从另一侧流出，依次通过电池模块，空气和电池的温差逐渐减小，换热能力减小，冷却空气的温度逐渐升高，采用串行通风冷却时风道前后端的温度差异性很大，通过SAE的无监督热效应模型学习，精确控制电池平均温度，从而获取串行通风的风速控制参数与平均温度的平衡。
图2显示了多层SAE(Stacked Autoencoder)网络模型。其中，SAE拟合网络通过求解热效能参数拟合方程，由此来进行下一步的模糊决策。SAE以多层神经网络的多层非线性函数，以sigmod函数为基础单元，来拟合电池热效应模型，精确获取热效应模型的非线性特征，进而实现精确控制。其中，所述前端传感器包括温度传感器、密度传感器。
所述的前端传感器通过信号调节器与电池管理控制器连接。
具体地，所述电池管理控制器与单体管理单元之间的通讯组网方式为总线形式。
当然，所述电池管理控制器与单体管理单元之间的通讯组网方式还可为菊链形式。
其中，所述的冷却方式采用的冷却装置包括风冷装置、液体冷却装置、热 管冷却装置和相变材料冷却装置。优选为风冷装置或者以风冷装置为主。
本发明实施例涉及一种自适应深度学习网络，通过前向的热参数，热性能分布的自适应拟合方法，获取热能参数控制输出方程；通过反向的控制参数，形成热控制参数的配置，建立不同的冷却方式。
本发明实施例还提供一种采用深度学习多层网络的框架，通过SAE网络拟合方法，结合电池的产热量、产热性能分布、热物性参数和控制方程等，形成了在多层深度学习网络下电池组的热效应模型，该模型根据热能参数的控制方程拟合获取，其优势在于，当温度超过了最大允许的工作温度，通过自适应控制方式，降低对电池的寿命、放电容量等造成的损害，并且采用自适应反向传导热管理方式对电池进行温度调控。
本发明还提供一种制冷参数精确控制方法，该方法通过检测反向传导参数，进行电池的热效应方程重构为基础，结合相应模型的控制方程和定解条件，分别形成了以风冷为主冷却的热模型，通过反向传导的管理参数对风速度和进风角度等因素进行控制，由此实现精确制冷控制。以上所述，对于本领域的普通技术人员来说，可以根据本发明的技术方案和技术构思作出其他各种相应的改变和变形，而所有这些改变和变形都应属于本发明权利要求的保护范围。

Claims (9)

  1. 一种智能电池热管理方法，其特征在于，包括以下步骤：
    S1、测量电池的热物性参数，所述热物性参数包括电池的密度、比热容；
    S2、根据电池热物性参数，建立电池的三维热效应方程，对应的公式如下，
    Figure PCTCN2018104846-appb-100001
    式中，T为电池的温度，t为时间，ρ为电池内部材料的平均密度，q为电池单位体积产热速率，c p为电池的定压比热容，λ x、λ y、λ z是电池在三维正交方向上的导热系数，求解得电池的温度场分布情况；
    S3、建立电池热效应模型
    根据电池的基本参数，包括标称容量、标称电压、放电截止电压、充电截止电压、内阻参数，结合电池的串联及并联热分布方式，形成在深度学习多层网络下电池组的热效应模型；
    S4、根据电池的热效应模型，通过前向参数的自适应拟合方法，获取热能参数控制输出方程，其中前向参数包括产热量、热性能分布、热物性参数；通过反向的制冷控制参数，形成热控制方程参数的配置，建立不同的冷却方式对电池进行温度调控。
  2. 根据权利要求1所述的电池热管理方法，其特征在于，所述冷却方式采用的冷却源包括风、液体和电。
  3. 根据权利要求1所述的电池热管理方法，其特征在于，步骤S1中采用前端传感器获取前向参数，步骤S4中采用高速数字信号处理内核处理器进行深度学习SAE拟合网络，包括计算热效能参数拟合方程，建立电源热管理模型、深度学习SAE反向调参。
  4. 根据权利要求1所述的电池热管理方法，其特征在于，还包括以下步骤：ARM内核根据电池热管理模型作出的模糊决策，输出制冷控制参数。
  5. 一种智能电池热管理系统，其特征在于，包括：
    前端传感器，所述前端传感器设置于电池箱体中用于获取前向参数，所述前向参数包括产热量、热性能分布、热物性参数；
    电池管理控制器，所述电池管理控制器包括高速数字信号处理内核和ARM内核，所述高速数字信号处理内核用于进行深度学习SAE拟合网络，所述ARM 内核根据电池热管理模型作出的模糊决策，输出制冷控制参数，以及冷却装置，所述冷却装置用于对电池组进行制冷散热。
  6. 根据权利要求5所述的电池热管理系统，其特征在于，所述前端传感器包括温度传感器、密度传感器。
  7. 根据权利要求5所述的电池热管理系统，其特征在于，所述前端传感器通过信号调节器与电池管理控制器连接。
  8. 根据权利要求5所述的电池热管理系统，其特征在于，所述电池管理控制器与单体管理单元之间的通讯组网方式为总线形式。
  9. 根据权利要求2所述的电池热管理系统，其特征在于，所述冷却装置包括风冷装置、液体冷却装置、热管冷却装置和相变材料冷却装置中的至少一种。
PCT/CN2018/104846 2018-09-10 2018-09-10 智能电池热管理方法及系统 WO2020051738A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/104846 WO2020051738A1 (zh) 2018-09-10 2018-09-10 智能电池热管理方法及系统

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/104846 WO2020051738A1 (zh) 2018-09-10 2018-09-10 智能电池热管理方法及系统

Publications (1)

Publication Number Publication Date
WO2020051738A1 true WO2020051738A1 (zh) 2020-03-19

Family

ID=69776926

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/104846 WO2020051738A1 (zh) 2018-09-10 2018-09-10 智能电池热管理方法及系统

Country Status (1)

Country Link
WO (1) WO2020051738A1 (zh)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106450388A (zh) * 2016-09-26 2017-02-22 中国计量大学 一种水冷型燃料电池温度优化与控制方法
CN107134604A (zh) * 2017-03-29 2017-09-05 南京航空航天大学 一种基于工况特性的动力电池热管理方法
CN108417906A (zh) * 2018-01-10 2018-08-17 中山大学 带有智能热管理系统控制功能的电池管理系统

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106450388A (zh) * 2016-09-26 2017-02-22 中国计量大学 一种水冷型燃料电池温度优化与控制方法
CN107134604A (zh) * 2017-03-29 2017-09-05 南京航空航天大学 一种基于工况特性的动力电池热管理方法
CN108417906A (zh) * 2018-01-10 2018-08-17 中山大学 带有智能热管理系统控制功能的电池管理系统

Similar Documents

Publication Publication Date Title
CN109271700B (zh) 基于深度学习多层网络建模的电池热管理方法及系统
CN108172930B (zh) 电池包冷却控制方法、装置和电池包
Vaideeswaran et al. Battery management systems for electric vehicles using lithium ion batteries
CN112152280B (zh) 一种动力电池多变量均衡控制方法
CN112092676A (zh) 一种利用虚拟温度传感器对电池包温度场的估算修正方法
CN113224412B (zh) 动力电池的控温方法、ampc控制器、热管理系统及介质
CN114335648B (zh) 全钒液流电池系统的控制方法和控制系统
CN110492185B (zh) 一种锂电池组均衡方法及系统
CN115020874A (zh) 一种动力电池热管理控制方法
WO2020051738A1 (zh) 智能电池热管理方法及系统
CN116872795A (zh) 电池热管理方法、装置、车辆和存储介质
CN107590617B (zh) 一种基于可信度推理模型的电池组均衡方法
CN116072437A (zh) 一种储能热管理系统、控制方法及控制装置
CN115848174A (zh) 一种动力电池快速加热充电方法、系统、装置及电动车辆
CN108808790A (zh) 基于温度补偿的镍氢电池包充电控制保护方法
Zhu et al. Analysis of the structure arrangement on the thermal characteristics of Li‐ion battery pack in thermoelectric generator
Wan et al. Design of balanced charging circuit for lithium ion battery
CN112165130A (zh) 一种充放自主可控的伺服动力电源及其实现方法
CN112002853A (zh) 防爆电池
CN112639496A (zh) 用于确定与第二电蓄能单元复合的第一电蓄能单元的周围环境温度的方法以及相对应的设备、计算机程序和机器可读存储介质
Chen et al. Phased control reciprocating airflow cooling strategy for a battery module considering stage of charge and state of health inconsistency
TWI842258B (zh) 用於電動載具之混合電池電能輸出分配控制方法
Ren et al. The charging and discharging balance control strategy of power batteries for hybrid energy storage
CN210723282U (zh) 一种锂离子电池组装置
Pengcheng et al. Thermal optimization of lithium-ion battery modules based on PCM under extreme cycle operation

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18933267

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 13/08/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 18933267

Country of ref document: EP

Kind code of ref document: A1