WO2023040486A1 - 一种新能源汽车电池管理系统sof估算方法 - Google Patents

一种新能源汽车电池管理系统sof估算方法 Download PDF

Info

Publication number
WO2023040486A1
WO2023040486A1 PCT/CN2022/109333 CN2022109333W WO2023040486A1 WO 2023040486 A1 WO2023040486 A1 WO 2023040486A1 CN 2022109333 W CN2022109333 W CN 2022109333W WO 2023040486 A1 WO2023040486 A1 WO 2023040486A1
Authority
WO
WIPO (PCT)
Prior art keywords
battery
voltage
current
value
parameter
Prior art date
Application number
PCT/CN2022/109333
Other languages
English (en)
French (fr)
Inventor
王禹
徐�明
栗顺
姜大威
王冬冬
张琦
张清扬
刘松柏
姜名勇
曲美玥
Original Assignee
一汽奔腾轿车有限公司
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 一汽奔腾轿车有限公司 filed Critical 一汽奔腾轿车有限公司
Publication of WO2023040486A1 publication Critical patent/WO2023040486A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/367Software therefor, e.g. for battery testing using modelling or look-up tables
    • YGENERAL 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
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00Road transport of goods or passengers
    • Y02T10/60Other road transportation technologies with climate change mitigation effect
    • Y02T10/70Energy storage systems for electromobility, e.g. batteries

Definitions

  • the invention relates to a method for estimating SOF of a battery management system of a new energy electric vehicle, in particular to a method for estimating SOF of a battery management system of a new energy vehicle based on a recursive least square method.
  • Battery SOF State of Function, charge and discharge capacity
  • BMS main battery management system
  • the mainstream SOF estimation method for new energy electric vehicles is the table look-up method, that is, the test data is obtained through a large number of tests, and the maximum charge and discharge capacity of the current battery is obtained by checking the two-dimensional power ammeter considering the temperature and the current SOC state.
  • this method relies too much on big data, and if the experimental data is insufficient, it will affect the accuracy estimation of SOF.
  • the current SOF estimation method does not consider the impact of battery protection, safety, failure, and power-on and off modes. The driving range and power performance of the vehicle will be affected, and the battery life will also be reduced.
  • the present invention provides a new energy electric vehicle battery management system SOF estimation method, which establishes a lithium battery according to the characteristics of the battery.
  • the second-order RC model generates six battery control parameters, and uses the recursive least squares method to estimate SOF online in real time to improve estimation accuracy.
  • the method of the invention can effectively estimate battery SOF with high precision, and can properly adjust control parameters according to battery characteristics, and is suitable for lithium battery SOF estimation.
  • a new energy electric vehicle battery management system SOF estimation method comprising the following steps:
  • Step 1 Establish a second-order RC discrete-time model of the lithium battery to generate six battery control parameters:
  • the second-order RC discrete-time model of lithium battery is:
  • V(k) is the measured terminal voltage of the battery; is the measurable parameter matrix, including terminal voltage and current; ⁇ is the parameter to be estimated;
  • the parameter to be estimated ⁇ contains six parameter vectors:
  • Step 2 Estimate the value of ⁇ by recursive least squares method:
  • step 1 generates six battery control parameters, specifically including:
  • the measurable parameter matrix is:
  • V and I are the battery terminal voltage and current collected by the voltage and current sensor respectively;
  • V V oc +I B R ⁇ +V s +V m
  • V s is the concentration polarization voltage
  • V m is the electrochemical polarization voltage
  • V oc is the open circuit voltage of the battery
  • R ⁇ is the internal resistance of the battery
  • I B is the input and output current of the battery pack collected by the current sensor; Collect the battery terminal voltage for the voltage sensor;
  • the concentration polarization voltage V s and the electrochemical polarization voltage V m are discretized to obtain:
  • ⁇ t is the voltage and current sampling period
  • V s (k) a 1 V s (k-1)+b 1 I B (k-1)
  • V m (k) a 2 V m (k-1)+b 2 I B (k-1)
  • the second-order discrete model can be transformed into a difference equation:
  • V(k) (a 1 +a 2 )V(k-1)-a 1 a 2 V(k-2)+R ⁇ I(k)+[b 1 -b 2 -R ⁇ (a 1 + a 2 )]I(k-1)+(a 1 a 2 R ⁇ -b 1 a 2 -b 2 a 1 )I(k-2)+[1-(a 1 +a 2 )+a 1 a 2 ] V oc
  • V(k) ⁇ 1 V(k-1)+ ⁇ 2 V(k-2)+ ⁇ 3 I(k)+ ⁇ 4 I(k-1)+ ⁇ 5 I(k-2)+ ⁇ 6
  • V oc ⁇ 6 /(1 ⁇ 1 ⁇ 2 ).
  • the initial value of the adaptive parameter ⁇ and the initial value of the relative matrix P are obtained from the data stored in the NVM at the previous moment.
  • the adaptive parameter ⁇ update formula is:
  • ⁇ (k) ⁇ (k-1)+G(k) ⁇ .
  • the relative matrix P upgrade calculation formula is:
  • V t is the cut-off charge and discharge voltage of the battery
  • V oc is the open circuit voltage of the battery in the current state
  • V r is the ohmic voltage drop of the battery
  • V s is the electrochemical polarization voltage drop of the battery
  • V m is the concentration difference of the battery Polarization voltage drop.
  • the invention provides a new energy electric vehicle battery management system SOF estimation method, which establishes a lithium battery second-order RC model according to the battery characteristics, generates six battery control parameters, and uses the recursive least square method to perform online real-time estimation of SOF to improve estimation accuracy .
  • the method of the invention can effectively estimate battery SOF with high precision, and can properly adjust control parameters according to battery characteristics, and is suitable for lithium battery SOF estimation.
  • Figure 1 is a circuit diagram of a second-order RC model of a lithium battery
  • Fig. 2 is a flow chart of the SOF estimation method of the new energy vehicle battery management system according to the present invention.
  • Lithium batteries are widely used in new energy vehicles. In order to better control the battery management system to make the battery have a longer service life and economy, battery parameters such as open circuit voltage V oc , ohmic internal resistance and battery capacity need to be estimated.
  • the maximum charge and discharge capacity of the battery in the current state is affected by the battery open circuit voltage, battery internal resistance, temperature and polarization effects.
  • the open circuit voltage V oc and other battery internal parameters cannot be directly measured. Therefore, it is necessary to estimate the internal parameters of the battery through the directly measurable terminal voltage and the input and output current of the battery.
  • V(k) is the measured terminal voltage of the battery; is the measurable parameter matrix, including terminal voltage and current; ⁇ is the parameter to be estimated.
  • the estimated parameter ⁇ includes six parameter vectors:
  • the measurable parameter matrix is:
  • V and I are the battery terminal voltage and current collected by the voltage and current sensor respectively.
  • the second-order RC discrete-time model of the lithium battery established in this example considers the internal resistance and polarization reaction during charging and discharging of the battery.
  • the circuit diagram is shown in Figure 1.
  • the polarization reaction includes concentration polarization and electrochemical polarization.
  • the equivalent expression is:
  • V V oc +I B R ⁇ +V s +V m
  • V s is the concentration polarization voltage
  • V m is the electrochemical polarization voltage
  • V oc is the open circuit voltage of the battery
  • R ⁇ is the internal resistance of the battery
  • I B is the input and output current of the battery pack collected by the current sensor; Acquire battery terminal voltage for the voltage sensor.
  • the concentration polarization voltage V s and the electrochemical polarization voltage V m are discretized to obtain:
  • ⁇ t is the sampling cycle of voltage and current.
  • V s (k) a 1 V s (k-1)+b 1 I B (k-1)
  • V m (k) a 2 V m (k-1)+b 2 I B (k-1)
  • the second-order discrete model can be transformed into a difference equation:
  • V(k) (a 1 +a 2 )V(k-1)-a 1 a 2 V(k-2)+R ⁇ I(k)+[b 1 -b 2 -R ⁇ (a 1 + a 2 )]I(k-1)
  • V(k) ⁇ 1 V(k-1)+ ⁇ 2 V(k-2)+ ⁇ 3 I(k)+ ⁇ 4 I(k-1)+ ⁇ 5 I(k-2)+ ⁇ 6
  • V oc ⁇ 6 /(1- ⁇ 1 - ⁇ 2 )
  • the value of ⁇ is estimated by the recursive least square method, and the specific method is as follows:
  • the new recursive parameter estimates are based on the old recursive estimates.
  • a forgetting factor a is introduced, where a is a number between 0 and 1, so as to maintain the new data’s correctness to the parameter.
  • the adaptive parameter ⁇ at the current moment is updated by the estimated error ⁇ and the adaptive parameter ⁇ (k-1) at the previous moment:
  • the parameters R ⁇ , R m , C m , R s , C s , and V oc of the battery are calculated according to the value of the adaptive parameter ⁇ .
  • V t(t) V oc(t) -V r(t) -V s(t) -V m(t)
  • V t is the charge and discharge cut-off voltage of the battery, which is determined by the characteristics of the lithium battery (discharge cut-off voltage 3.2V ⁇ 3.5V, charge cut-off voltage 4.1V ⁇ 4.3V);
  • V oc is the open circuit voltage of the battery in the current state;
  • V r is The ohmic voltage drop of the battery;
  • V s is the electrochemical polarization voltage drop of the battery;
  • V m is the concentration polarization voltage drop of the battery.
  • the ohmic voltage drop V r is obtained by multiplying the battery charge and discharge current by the battery ohmic internal resistance:
  • V r(t) I B *R ⁇
  • the electrochemical polarization voltage drop V s is determined by the electrochemical polarization voltage drop at the previous sampling moment and the battery charge and discharge current:
  • V s(t) I B* R s (1-e -t/RsCs )-V s(t-1) e -t/RsCs
  • V s(t-1) is the electrochemical polarization voltage drop at the previous sampling moment.
  • the concentration polarization voltage drop V m is determined by the concentration polarization voltage drop at the previous sampling moment and the battery charge and discharge current:
  • V m(t) I B* R m (1-e -t/RmCm )–V m(t-1) e -t/RmCm
  • V s(t-1) is the concentration polarization pressure drop at the previous sampling time.
  • the maximum charge and discharge current of the battery in the current state is estimated as:
  • V t(t) V oc(t) -I B *R ⁇ –(I B* R s (1-e -t/RsCs )-V s(t-1) e -t/RsCs )–(I B* R m (1-e -t/RmCm )–V m(t-1) e -t/RmCm )
  • I B_max (V oc(t) +V s(t-1) e -t/RsCs +V m(t-1) e -t/RmCm -V t(t) )/(R ⁇ +R s ( 1-e -t/RsCs )+R m (1-e -t/RmCm ))
  • I B_max is the maximum charging and discharging current estimated by the battery according to the state parameters.
  • the power battery fault management system divides battery faults into three levels, and the fault handling scheme is as follows: After the SOF estimation module receives the fault information reported by the fault diagnosis, it multiplies the maximum charging and discharging current of the battery by a fault-related correction coefficient ⁇ 4 , The first-level fault ⁇ 4 is 0.7, the second-level fault ⁇ 4 is 0.5, and the third-level fault ⁇ 4 is 0.
  • the maximum charge and discharge current of the charging pile is limited to 0A, and the maximum discharge current and maximum braking energy recovery current capacity are determined by the above estimated maximum charge and discharge values;
  • the maximum discharge current and the maximum braking energy recovery battery are limited to 0A, and the maximum charging current capacity of the charging pile is determined by the estimated maximum charging current above.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Secondary Cells (AREA)
  • Tests Of Electric Status Of Batteries (AREA)

Abstract

一种新能源电动汽车电池管理系统SOF估算方法,根据电池特性建立锂电池二阶RC模型,生成六个电池控制参数,采用递推最小二乘法对SOF进行在线实时估算:读取电压值及电流值,设定自适应参数θ及相对矩阵P初始值,根据电压值及电流值计算参数矩阵及增益矩阵,计算估算误差,对当前时刻自适应参数θ进行更新,对相对矩阵P进行升级;根据自适应参数θ值计算电池的实时参数;并根据电池实时参数,估算电池当前状态最大充放电电流值。可以有效的估计电池SOF,精度较高,根据电池特性可适当调整控制参数,适用于锂电池SOF估计。

Description

一种新能源汽车电池管理系统SOF估算方法 技术领域
本发明涉及一种新能源电动汽车电池管理系统SOF估算方法,具体涉及一种基于递推最小二乘法的新能源汽车电池管理系统SOF估算方法。
背景技术
现阶段,新能源汽车高速发展,其中以电能为动力源,电机为驱动装置的新能源汽车发展显著。动力电池及电池管理系统是电动汽车的关键部件。电池SOF(State of Function,充放电能力)是电池管理系统状态估算的重要参数,简单的说就是电池状态下的最大充放电电流,如果没有精确的SOF,BMS(主电池管理系统)将无法正常工作,因为SOF估算精度不高会对电动车产生两种不良影响:1、SOF估算过高,电池容易过放,影响电池使用寿命;2、SOF估算过低,整车动力性能会受到影响,达不到整车设定动力性目标。SOF精度越高,对于相同容量的电池,可以在任何行驶工况下充分发挥电池的性能,满足驾驶需求,在一定程度上提高电池的使用寿命。准确的估算电池的最大充放电能力一直是国内外研究的重点和难点。
目前新能源电动汽车主流的SOF估算方法为查表方法,即经过大量的试验得到试验数据,考虑温度和当前SOC状态,查二维功率电流表得到当前电池的最大充放电能力。但这种方法过分依赖大数据,如果试验数据不充分将会影响SOF的精度估算。目前的SOF估算方法也没有考虑电池保护、安全、故障及上下电模态的影响整车的续驶里程和动力性能会受到影响,电池的寿命也会降低。
发明内容
为了解决现有电动汽车电池管理系统SOF估算方法存在累计误差大,需要大量标定数据,不能实时修正等问题,本发明提供一种新能源电动汽车电池管理系统SOF估算方法,根据电池特性建立锂电池二阶RC模型,生成六个电池控制参数,采用递推最小二乘法对SOF进行在线实时估算,提高估算精度。本发明方法可以有效的估计电池SOF,精度较高,根据电池特性可适当调整控制参数,适用于锂电池SOF估计。
本发明的目的是通过以下技术方案实现的:
一种新能源电动汽车电池管理系统SOF估算方法,包括以下步骤:
步骤一、建立锂电池二阶RC离散时间模型,生成六个电池控制参数:
锂电池二阶RC离散时间模型为:
Figure PCTCN2022109333-appb-000001
其中,V(k)为电池的测量终端电压;
Figure PCTCN2022109333-appb-000002
为可测量参数矩阵,包括终端电压和电流;θ为待估参数;
待估参数θ包含六个参数向量:
θ=[θ 123456] T
步骤二、通过递推最小二乘法估算θ值:
2.1)读取两组电压值V(1)、V(2),及两组电流值I(1)、I(2),设定自适应参数θ及相对矩阵P初始值;
2.2)定义遗忘因子,0<a<1,以保持新数据对参数估计值的修正能力;
2.3)读取一组新的电压值V(k)及电流值I(k);
2.4)根据电压值V(k)、电流值I(k),上一时刻读取的电压值V(k-1)、电流值I(k-1),再上一时刻读取的电压值V(k-2)、电流值I(k-2),计算本时刻的参数矩阵
Figure PCTCN2022109333-appb-000003
2.5)由上一时刻计算得到的参数矩阵
Figure PCTCN2022109333-appb-000004
和相对矩阵P(k-1)计算本时刻增益矩阵G(k);
2.6)计算估算误差α,由估算误差α和前一时刻自适应参数θ(k-1)对当前时刻自适应参数θ进行更新;
2.7)对相对矩阵P进行升级,预留下一时刻使用;
2.8)根据自适应参数θ值计算电池的实时参数,包括欧姆内阻R Ω、电化学极化内阻R m、电化学极化电容C m、浓差极化内阻R s、浓差极化电容C s、开路电压V oc
2.9)根据上述电池实时参数,估算电池当前状态最大充放电电流值。
进一步地,所述步骤一生成六个电池控制参数,具体包括:
可测参数矩阵为:
Figure PCTCN2022109333-appb-000005
其中,V、I分别为电压电流传感器采集的电池端电压和电流;
所述锂电池二阶RC离散时间模型的等效表达式为:
V=V oc+I BR Ω+V s+V m
其中,V s为浓差极化电压;V m为电化学极化电压;V oc为电池的开路电压;R Ω为电池的内阻;I B为电流传感器采集的电池包输入输出电流;V为电压传感器采集电池端电压;
将浓差极化电压V s和电化学极化电压V m离散化,可得:
Figure PCTCN2022109333-appb-000006
Figure PCTCN2022109333-appb-000007
其中,Δt为电压、电流采样周期;
令:
Figure PCTCN2022109333-appb-000008
Figure PCTCN2022109333-appb-000009
Figure PCTCN2022109333-appb-000010
Figure PCTCN2022109333-appb-000011
则:
V s(k)=a 1V s(k-1)+b 1I B(k-1)
V m(k)=a 2V m(k-1)+b 2I B(k-1)
根据Z函数转换,二阶离散模型可转化成差分方程:
V(k)=(a 1+a 2)V(k-1)-a 1a 2V(k-2)+R ΩI(k)+[b 1-b 2-R Ω(a 1+a 2)]I(k-1)+(a 1a 2R Ω-b 1a 2-b 2a 1)I(k-2)+[1-(a 1+a 2)+a 1a 2]V oc
由于:
V(k)=θ 1V(k-1)+θ 2V(k-2)+θ 3I(k)+θ 4I(k-1)+θ 5I(k-2)+θ 6
则根据对应关系得:
a 1+a 2=θ 1
-a 1a 2=θ 2
R Ω=θ 3
b 1-b 2-R Ω(a 1+a 2)=θ 4
a 1a 2R Ω-b 1a 2-b 2a 1=θ 5
[1-(a 1+a 2)+a 1a 2]V oc=θ 6
则待估算参数为:
θ=[a 1+a 2,-a 1a 2,R Ω,b 1-b 2-R Ω(a 1+a 2),a 1a 2R Ω-b 1a 2-b 2a 1,[1-(a 1+a 2)+a 1a 2]V oc] T
电池特性参数R Ω、R m、C m、R s、C s、V oc通过计算θ值后求得,其中:
V oc=θ 6/(1-θ 12)。
进一步地,所述步骤2.1)中,自适应参数θ初始值和相对矩阵P初始值由上一时刻存入NVM中的数据获得。
进一步地,所述步骤2.4)中,参数矩阵
Figure PCTCN2022109333-appb-000012
的计算公式为:
Figure PCTCN2022109333-appb-000013
进一步地,所述步骤2.5)中,增益矩阵G(k)的计算公式为:
Figure PCTCN2022109333-appb-000014
进一步地,所述步骤2.6)中,估算误差α的计算公式为:
Figure PCTCN2022109333-appb-000015
自适应参数θ更新公式为:
θ(k)=θ(k-1)+G(k)α。
进一步地,所述步骤2.7)中,相对矩阵P升级计算公式为:
Figure PCTCN2022109333-appb-000016
进一步地,所述步骤2.9)中,根据V t=V oc–V r-V s-V m公式求得电池当前状态下最大的充放电电流,
其中,V t为电池的截止充放电电压;V oc为当前状态电池的开路电压;V r为电池的欧姆压降;V s为电池的电化学极化压降;V m为电池的浓差极化压降。
本发明具有以下有益效果:
本发明提供一种新能源电动汽车电池管理系统SOF估算方法,根据电池特性建立锂电池二阶RC模型,生成六个电池控制参数,采用递推最小二乘法对SOF进行在线实时估算,提高估算精度。本发明方法可以有效的估计电池SOF,精度较高,根据电池特性可适当调整控制参数,适用于锂电池SOF估计。
附图说明
图1为锂电池二阶RC模型电路图;
图2为本发明新能源汽车电池管理系统SOF估算方法流程图。
具体实施方式
以下结合附图和实施例详细描述本发明的技术方案:
锂电池大量应用在新能源汽车上,为了更好的控制电池管理系统以使电池具有更长的使用寿命和经济性,电池参数如开路电压V oc、欧姆内阻和电池容量等参数需要估算。
当前状态下电池最大的充放电能力受电池开路电压、电池内阻、温度及极化效应的影响,然而开路电压V oc和其它电池内部参数都无法直接测量。因此需要通过可直接测量的端电压和电池输入输出电流对电池内部参数进行估算。
锂电池的特征参数包括端电压V、欧姆内阻R Ω、开路电压V oc、时间常数τ 11=R sC s)、时间常数τ 21=R mC m),其中R s为浓差极化内阻,C s为浓差极化电容,R m为电化学极化内阻,C m为电化学极化电容。
通过实时监测电池的端电压、电流值估算电池的特性参数,从而估算电池当前状态下允许的最大的充放电电流。
本实施例提供一种动力电池的SOF估算方法:
一、建立锂电池二阶RC离散时间模型,其方程表达式为:
Figure PCTCN2022109333-appb-000017
其中V(k)为电池的测量终端电压;
Figure PCTCN2022109333-appb-000018
为可测量参数矩阵,包括终端电压和电流;θ为待估参数。
本实施例中建立的锂电池二阶RC离散时间模型中,待估参数θ包含六个参数向量:
θ=[θ 123456] T
可测参数矩阵为:
Figure PCTCN2022109333-appb-000019
其中,V、I分别为电压电流传感器采集的电池端电压和电流。
本实施例中建立的锂电池二阶RC离散时间模型,考虑电池充放电过程中内阻和极化反应,电路图如图1所示,极化反应包括浓差极化和电化学极化,其等效表达式为:
V=V oc+I BR Ω+V s+V m
其中,V s为浓差极化电压;V m为电化学极化电压;V oc为电池的开路电压;R Ω为电池的内阻;I B为电流传感器采集的电池包输入输出电流;V为电压传感器采集电池端电压。
将浓差极化电压V s和电化学极化电压V m离散化,可得:
Figure PCTCN2022109333-appb-000020
Figure PCTCN2022109333-appb-000021
其中,Δt为电压、电流采样周期。
令:
Figure PCTCN2022109333-appb-000022
Figure PCTCN2022109333-appb-000023
Figure PCTCN2022109333-appb-000024
Figure PCTCN2022109333-appb-000025
则:
V s(k)=a 1V s(k-1)+b 1I B(k-1)
V m(k)=a 2V m(k-1)+b 2I B(k-1)
实际应用中,电压、电流的采集计算都是离散的,进行电池荷电状态估计需对数学公式进行离散化处理。
根据Z函数转换,二阶离散模型可转化成差分方程:
V(k)=(a 1+a 2)V(k-1)-a 1a 2V(k-2)+R ΩI(k)+[b 1-b 2-R Ω(a 1+a 2)]I(k-1)
+(a 1a 2R Ω-b 1a 2-b 2a 1)I(k-2)+[1-(a 1+a 2)+a 1a 2]V oc
由于:
V(k)=θ 1V(k-1)+θ 2V(k-2)+θ 3I(k)+θ 4I(k-1)+θ 5I(k-2)+θ 6
则根据对应关系得:
a 1+a 2=θ 1
-a 1a 2=θ 2
R Ω=θ 3
b 1-b 2-R Ω(a 1+a 2)=θ 4
a 1a 2R Ω-b 1a 2-b 2a 1=θ 5
[1-(a 1+a 2)+a 1a 2]V oc=θ 6
则待估算参数为:
θ=[a 1+a 2,-a 1a 2,R Ω,b 1-b 2-R Ω(a 1+a 2),a 1a 2R Ω-b 1a 2-b 2a 1,[1-(a 1+a 2)+a 1a 2]V oc] T
电池特性参数R Ω、R m、C m、R s、C s、V oc可通过计算θ值后求得,其中:
V oc=θ 6/(1-θ 12)
二、本实施例通过递推最小二乘法估算θ值,具体方法如下:
递推辨识算法的原理为:
新的参数估计值=旧的参数估计值+修正项
即新的递推参数估计值是在旧的递推估计值的基础上而成。
首先,读取两组电压值及两组电流值,V(1)、V(2)、I(1)、I(2),设定自适应参数θ及相对矩阵P初始值,自适应参数θ初始值和相对矩阵P初始值由上次下电存入NVM中的数据获得。
在辨识计算过程中,防止算法增益矩阵G(k)急剧衰减,使得新数据失去对参数估计值的修正能力,引进遗忘因子a,a为0~1之间的数,以保持新数据对参数估计值的一定的修正能力,使得能得到更准确的参数估计值,或能适应对慢时变参数的辨识。
读取一组新的电压值V(k)及电流值I(k),根据电压值V(k)、电流值I(k),上一时刻读取的电压值V(k-1)、电流值I(k-1),再上一时刻读取的电压值V(k-2)、电流值I(k-2),计算本时刻的参数矩阵
Figure PCTCN2022109333-appb-000026
Figure PCTCN2022109333-appb-000027
由上一时刻计算得到的参数矩阵φ(k-1)和相对矩阵P(k-1)计算本时刻增益矩阵:
Figure PCTCN2022109333-appb-000028
计算相应估算误差:
Figure PCTCN2022109333-appb-000029
由估算误差α和前一时刻自适应参数θ(k-1)对当前时刻自适应参数θ进行更新:
θ(k)=θ(k-1)+G(k)α
对相对矩阵P进行升级:
Figure PCTCN2022109333-appb-000030
根据自适应参数θ值计算电池的参数R Ω、R m、C m、R s、C s、V oc
根据上述电池实时参数,估算电池当前状态最大充放电电流值:
V t(t)=V oc(t)–V r(t)-V s(t)-V m(t)
其中,V t为电池的充放电截止电压,由锂电池特性决定(放电截止电压3.2V~3.5V,充电截止电压4.1V~4.3V);V oc为当前状态电池的开路电压;V r为电池的欧姆压降;V s为电池的电化学极化压降;V m为电池的浓差极化压降。
欧姆压降V r由电池充放电电流乘以电池欧姆内阻得到:
V r(t)=I B*R Ω
电化学极化压降V s由前一采样时刻的电化学极化压降和电池充放电电流决定:
V s(t)=I B*R s(1-e -t/RsCs)-V s(t-1)e -t/RsCs
其中,V s(t-1)为前一采样时刻电化学极化压降。
浓差极化压降V m由前一采样时刻的浓差极化压降和电池充放电电流决定:
V m(t)=I B*R m(1-e -t/RmCm)–V m(t-1)e -t/RmCm
其中,V s(t-1)为前一采样时刻浓差极化压降。
根据电池状态参数估算当前状态电池的最大充放电电流为:
V t(t)=V oc(t)-I B*R Ω–(I B*R s(1-e -t/RsCs)-V s(t-1)e -t/RsCs)–(I B*R m(1-e -t/RmCm)–V m(t-1)e -t/RmCm)
I B_max=(V oc(t)+V s(t-1)e -t/RsCs+V m(t-1)e -t/RmCm-V t(t))/(R Ω+R s(1-e -t/RsCs)+R m(1-e -t/RmCm))
其中,I B_max为电池根据状态参数估算的最大充放电电流。
三、考虑电池保护、故障、安全及模态估算电池当前最大的充放电电流:
考虑电池保护对电池最大充放电电流的限制:
1、当电池温度过高或者过低时,将I B_max乘以一与电池温度相关的修正系数α 1当电池温度在[0℃40℃]之间时,α T为1,当电池温度在此区间之外,随着温度升高或降低,α T逐渐减小到0,温度区间为标定量,与电池特性相关。
2、估算电池最大放电电流时,当电池单体最低电压过低,将I B_max乘以一与电池电压相关的修正系数α 2,当电池最低单体电压在[3.7V 4.2V]时,α 2为1,随着单体电压降低,α 2逐渐减小到0,电压区间为标定量,与电池特性相关。
3、估算电池最大充电电流时,当电池单体最高电压过高,将I B_max乘以一与电池电压相关的修正系数α 3,当电池最高单体电压在[3.2V 3.7V]时,α 3为1,随着单体电压升高,α 3逐渐减小到0,电压区间为标定量,与电池特性相关。
考虑电池故障对电池最大充放电电流的限制:
动力电池故障管理系统将电池故障分三个等级,故障处理方案为:SOF估算模块接收到故障诊断上报的故障信息后,将电池的最大充放电电流乘以一与故障相关的修正系数α 4,一级故障α 4为0.7,二级故障α 4为0.5,三级故障α 4为0。
考虑电池系统安全对电池最大充放电电流的限制:
当电池系统发生安全故障时,对最大充放电电流进行限制,将I B_max乘以一与电池系统安全相关的修正系数α 5
1、发生环路互锁故障或电池系统绝缘故障时,同时考虑车速对系统安全行驶的影响,当故障发生且车速高于20km/h时,α 5为0.5,当车速低于20km/h时,α 5为0;
2、电池系统发生短路、短路或者碰撞安全故障时,α 5为0。
考虑电池系统模态对电池最大充放电电流的限制:
1、当电池系统处于放电模态时,将充电桩充电最大充放电电流限为0A,最大放电电流、最大制动能量回收电流能力由上述估算最大充放电值决定;
2、当电池系统处于充电模态时,将放电最大电流、最大制动能量回收电池限为0A,充电桩充电最大电流能力由上述最大充电电流估算值决定。

Claims (8)

  1. 一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,包括以下步骤:
    步骤一、建立锂电池二阶RC离散时间模型,生成六个电池控制参数:
    锂电池二阶RC离散时间模型为:
    Figure PCTCN2022109333-appb-100001
    其中,V(k)为电池的测量终端电压;
    Figure PCTCN2022109333-appb-100002
    为可测量参数矩阵,包括终端电压和电流;θ为待估参数;
    待估参数θ包含六个参数向量:
    θ=[θ 123456] T
    步骤二、通过递推最小二乘法估算θ值:
    2.1)读取两组电压值V(1)、V(2),及两组电流值I(1)、I(2),设定自适应参数θ及相对矩阵P初始值;
    2.2)定义遗忘因子,0<a<1,以保持新数据对参数估计值的修正能力;
    2.3)读取一组新的电压值V(k)及电流值I(k);
    2.4)根据电压值V(k)、电流值I(k),上一时刻读取的电压值V(k-1)、电流值I(k-1),再上一时刻读取的电压值V(k-2)、电流值I(k-2),计算本时刻的参数矩阵
    Figure PCTCN2022109333-appb-100003
    2.5)由上一时刻计算得到的参数矩阵
    Figure PCTCN2022109333-appb-100004
    和相对矩阵P(k-1)计算本时刻增益矩阵G(k);
    2.6)计算估算误差α,由估算误差α和前一时刻自适应参数θ(k-1)对当前时刻自适应参数θ进行更新;
    2.7)对相对矩阵P进行升级,预留下一时刻使用;
    2.8)根据自适应参数θ值计算电池的实时参数,包括欧姆内阻R Ω、电化学极化内阻R m、电化学极化电容C m、浓差极化内阻R s、浓差极化电容C s、开路电压V oc
    2.9)根据上述电池实时参数,估算电池当前状态最大充放电电流值。
  2. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤一生成六个电池控制参数,具体包括:
    可测参数矩阵为:
    Figure PCTCN2022109333-appb-100005
    其中,V、I分别为电压电流传感器采集的电池端电压和电流;
    所述锂电池二阶RC离散时间模型的等效表达式为:
    V=V oc+I BR Ω+V s+V m
    其中,V s为浓差极化电压;V m为电化学极化电压;V oc为电池的开路电压;R Ω为电池的内阻;I B为电流传感器采集的电池包输入输出电流;V为电压传感器采集电池端电压;
    将浓差极化电压V s和电化学极化电压V m离散化,可得:
    Figure PCTCN2022109333-appb-100006
    Figure PCTCN2022109333-appb-100007
    其中,Δt为电压、电流采样周期;
    令:
    Figure PCTCN2022109333-appb-100008
    Figure PCTCN2022109333-appb-100009
    Figure PCTCN2022109333-appb-100010
    Figure PCTCN2022109333-appb-100011
    则:
    V s(k)=a 1V s(k-1)+b 1I B(k-1)
    V m(k)=a 2V m(k-1)+b 2I B(k-1)
    根据Z函数转换,二阶离散模型可转化成差分方程:
    V(k)=(a 1+a 2)V(k-1)-a 1a 2V(k-2)+R ΩI(k)+[b 1-b 2-R Ω(a 1+a 2)]I(k-1)+(a 1a 2R Ω-b 1a 2-b 2a 1)I(k-2)+[1-(a 1+a 2)+a 1a 2]V oc
    由于:
    V(k)=θ 1V(k-1)+θ 2V(k-2)+θ 3I(k)+θ 4I(k-1)+θ 5I(k-2)+θ 6
    则根据对应关系得:
    a 1+a 2=θ 1
    -a 1a 2=θ 2
    R Ω=θ 3
    b 1-b 2-R Ω(a 1+a 2)=θ 4
    a 1a 2R Ω-b 1a 2-b 2a 1=θ 5
    [1-(a 1+a 2)+a 1a 2]V oc=θ 6
    则待估算参数为:
    θ=[a 1+a 2,-a 1a 2,R Ω,b 1-b 2-R Ω(a 1+a 2),a 1a 2R Ω-b 1a 2-b 2a 1,[1-(a 1+a 2)+a 1a 2]V oc] T
    电池特性参数R Ω、R m、C m、R s、C s、V oc通过计算θ值后求得,其中:
    V oc=θ 6/(1-θ 12)。
  3. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤2.1)中,自适应参数θ初始值和相对矩阵P初始值由上一时刻存入NVM中的数据获得。
  4. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤2.4)中,参数矩阵
    Figure PCTCN2022109333-appb-100012
    的计算公式为:
    Figure PCTCN2022109333-appb-100013
  5. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤2.5)中,增益矩阵G(k)的计算公式为:
    Figure PCTCN2022109333-appb-100014
  6. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤2.6)中,估算误差α的计算公式为:
    Figure PCTCN2022109333-appb-100015
    自适应参数θ更新公式为:
    θ(k)=θ(k-1)+G(k)α。
  7. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤2.7)中,相对矩阵P升级计算公式为:
    Figure PCTCN2022109333-appb-100016
  8. 如权利要求1所述的一种新能源电动汽车电池管理系统SOF估算方法,其特征在于,所述步骤2.9)中,根据V t=V oc–V r-V s-V m公式求得电池当前状态下最大的充放电电流,
    其中,V t为电池的截止充放电电压;V oc为当前状态电池的开路电压;V r为电池的欧姆压降;V s为电池的电化学极化压降;V m为电池的浓差极化压降。
PCT/CN2022/109333 2021-09-17 2022-07-31 一种新能源汽车电池管理系统sof估算方法 WO2023040486A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111089974.8 2021-09-17
CN202111089974.8A CN113835033A (zh) 2021-09-17 2021-09-17 一种新能源汽车电池管理系统sof估算方法

Publications (1)

Publication Number Publication Date
WO2023040486A1 true WO2023040486A1 (zh) 2023-03-23

Family

ID=78959692

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/109333 WO2023040486A1 (zh) 2021-09-17 2022-07-31 一种新能源汽车电池管理系统sof估算方法

Country Status (2)

Country Link
CN (1) CN113835033A (zh)
WO (1) WO2023040486A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116387653A (zh) * 2023-06-05 2023-07-04 钛玛科(北京)工业科技有限公司 一种用于锂电池的低损耗管理方法及装置

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113835033A (zh) * 2021-09-17 2021-12-24 一汽奔腾轿车有限公司 一种新能源汽车电池管理系统sof估算方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8880367B1 (en) * 2011-11-10 2014-11-04 Energy Pass Incorporation Method for accurately performing power estimation on a battery of an electronic device, and associated apparatus
CN107064811A (zh) * 2017-03-01 2017-08-18 华南理工大学 一种锂电池soc在线估计方法
CN107390127A (zh) * 2017-07-11 2017-11-24 欣旺达电动汽车电池有限公司 一种soc估算方法
CN108693472A (zh) * 2017-04-12 2018-10-23 上海蓝诺新能源技术有限公司 电池等效模型参数在线辨识方法
CN109061520A (zh) * 2018-10-25 2018-12-21 杭州神驹科技有限公司 一种动力电池健康与功率状态在线估算方法及系统
CN113835033A (zh) * 2021-09-17 2021-12-24 一汽奔腾轿车有限公司 一种新能源汽车电池管理系统sof估算方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111208432A (zh) * 2020-01-16 2020-05-29 北方工业大学 用于电池的二阶rc等效电路模型参数实时辨识方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8880367B1 (en) * 2011-11-10 2014-11-04 Energy Pass Incorporation Method for accurately performing power estimation on a battery of an electronic device, and associated apparatus
CN107064811A (zh) * 2017-03-01 2017-08-18 华南理工大学 一种锂电池soc在线估计方法
CN108693472A (zh) * 2017-04-12 2018-10-23 上海蓝诺新能源技术有限公司 电池等效模型参数在线辨识方法
CN107390127A (zh) * 2017-07-11 2017-11-24 欣旺达电动汽车电池有限公司 一种soc估算方法
CN109061520A (zh) * 2018-10-25 2018-12-21 杭州神驹科技有限公司 一种动力电池健康与功率状态在线估算方法及系统
CN113835033A (zh) * 2021-09-17 2021-12-24 一汽奔腾轿车有限公司 一种新能源汽车电池管理系统sof估算方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116387653A (zh) * 2023-06-05 2023-07-04 钛玛科(北京)工业科技有限公司 一种用于锂电池的低损耗管理方法及装置
CN116387653B (zh) * 2023-06-05 2023-08-22 钛玛科(北京)工业科技有限公司 一种用于锂电池的低损耗管理方法及装置

Also Published As

Publication number Publication date
CN113835033A (zh) 2021-12-24

Similar Documents

Publication Publication Date Title
WO2023040486A1 (zh) 一种新能源汽车电池管理系统sof估算方法
EP3764114A1 (en) Method, device and system for estimating remaining charging time and a storage medium
US7324902B2 (en) Method and apparatus for generalized recursive least-squares process for battery state of charge and state of health
CN103616647B (zh) 一种用于电动汽车电池管理系统的电池剩余电量估计方法
KR100818520B1 (ko) 전기화학셀(cell)의 현재 상태와 현재 파라미터를추정하는 장치, 방법 및 시스템 및 기록매체
CN103744030B (zh) 电池组健康状态和荷电状态在线估算装置及估算方法
WO2017119393A1 (ja) 状態推定装置、状態推定方法
CN109669131B (zh) 一种工况环境下动力电池soc估算方法
CN103424710A (zh) 用于监测电池组中的老化单体的性能变化的方法和系统
Wei et al. Lyapunov-based state of charge diagnosis and health prognosis for lithium-ion batteries
US8823326B2 (en) Method for determining the state of charge of a battery in charging or discharging phase
CN109752660B (zh) 一种无电流传感器的电池荷电状态估计方法
EP4152022B1 (en) Method for determining full-charge capacity of battery pack, method for determining state of health of battery pack, system, and apparatus
CN110716141B (zh) 一种基于扩展卡尔曼滤波的电池荷电状态估计方法
CN113785209B (zh) 用于检测异常电池单体的方法
JP2016161448A (ja) 内部状態推定装置
CN111027203B (zh) 一种超级电容soc计算方法
WO2019184879A1 (zh) 动力电池组的rc网络参数获取方法、装置和电动汽车
JP2021157926A (ja) 二次電池装置および二次電池システム
Baba et al. State of charge estimation of HEV/EV battery with series Kalman filter
Zhao et al. An online model identification for state of charge estimation of lithium-ion batteries using extended kalman filter
Liu et al. SOH diagnostic and prognostic based on external health indicator of lithium-ion batteries
Kumar et al. Non-Linear Kalman Filter Based SOC Estimation of a Lithium-Ion Battery Pack Considering Cell Balancing
Liu et al. Li-ion battery state of charge estimation of multi-type working conditions using a segmented multiple independent forgetting factors recursive least squares method
CN111624491A (zh) 一种确定电池剩余电量的方法、装置及电池管理系统

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: 22868863

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE