WO2022100230A1 - 基于人工智能的锂离子电池充电曲线重构及状态估计方法 - Google Patents

基于人工智能的锂离子电池充电曲线重构及状态估计方法 Download PDF

Info

Publication number
WO2022100230A1
WO2022100230A1 PCT/CN2021/116036 CN2021116036W WO2022100230A1 WO 2022100230 A1 WO2022100230 A1 WO 2022100230A1 CN 2021116036 W CN2021116036 W CN 2021116036W WO 2022100230 A1 WO2022100230 A1 WO 2022100230A1
Authority
WO
WIPO (PCT)
Prior art keywords
charging
battery
charging curve
curve
complete
Prior art date
Application number
PCT/CN2021/116036
Other languages
English (en)
French (fr)
Inventor
熊瑞
田金鹏
段砚州
Original Assignee
北京理工大学
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 北京理工大学 filed Critical 北京理工大学
Priority to US17/916,041 priority Critical patent/US20230168304A1/en
Publication of WO2022100230A1 publication Critical patent/WO2022100230A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/378—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC] specially adapted for the type of battery or accumulator
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/392—Determining battery ageing or deterioration, e.g. state of health

Definitions

  • the present invention relates to the field of battery systems, and in particular, to state estimation of lithium-ion batteries.
  • the battery management system can only collect fragments of the battery's voltage, current, temperature and other signals, so its state can only be estimated by the measured signals.
  • Existing state estimation methods often only target a few specific states, and assume that other states are known, so there is a big limitation in the global estimation. For example, the estimation of battery capacity often only focuses on the establishment of the relationship between the capacity and the characteristics of the charging curve, while ignoring the estimation of other states. In fact, the battery charging curve (the relationship between the charging voltage and the charged capacity) reflects a large amount of battery state information, which can meet the requirements of fully and accurately characterizing the battery state.
  • the battery is often not fully charged and fully discharged, and the battery management system can often only collect part of the charging curve. Therefore, if a complete charging curve can be reconstructed by necessary technical means based on the obtained relatively accurate partial charging curve segments, it is of great significance for the improvement of the battery state estimation method and the improvement of the battery management function.
  • the present invention provides an artificial intelligence-based lithium-ion battery charging curve reconstruction and state estimation method, which specifically includes the following steps:
  • Step 1 Obtain the complete voltage/current charging curve of the battery under different aging states when different charging methods are used as training data;
  • Step 2 Divide the acquired charging curve into data segments through a suitable segmentation method, and perform discretization processing on the data segments and the charging curve;
  • Step 3 using the discretized data fragments obtained in step 2 to train the selected deep learning algorithm, and establish a mapping relationship between each data fragment and the complete charging curve;
  • Step 4 Apply the trained deep learning algorithm online, input the actual charging segment data collected by the battery management system into the deep learning algorithm, and output a complete charging curve;
  • Step 5 Extract the battery state parameters to be estimated from the complete charging curve.
  • the method also includes:
  • Step 6 After the battery management system collects a certain number of actual battery charging curves, retrain and update the deep learning algorithm.
  • the complete voltage/current charging curve of the battery under different aging states is obtained when different charging methods are used, which specifically includes: using common methods such as constant current charging, constant current and constant voltage charging, multi-stage constant current charging, Charging schemes such as pulse charging.
  • different charging methods which specifically includes: using common methods such as constant current charging, constant current and constant voltage charging, multi-stage constant current charging, Charging schemes such as pulse charging.
  • the daily charging curves of batteries in different aging states are obtained, including battery charging current, voltage, temperature and other signals under a given charging scheme.
  • the step 2 specifically includes: determining a segment length, sliding the segment length on the charging curve, thereby dividing the charging curve in step 1 into data segments of a certain length, and each segment includes sampling at each moment. Signals such as voltage, current, temperature, etc. At the same time, the obtained data segments are sampled at fixed time intervals or voltage intervals, thereby discretizing the complete charging curve.
  • the deep learning algorithm in the third step specifically adopts a convolutional neural network, a densely connected network, a recurrent neural network, and the like.
  • the above method provided by the present invention can reconstruct the complete charging curve of the battery through partial charging segments, and can simultaneously realize the estimation of the maximum capacity, maximum energy, state of charge, energy state and power state, and can be obtained through the derived capacity. Incremental curve, differential voltage curve, etc. to realize the analysis of battery aging. In long-term applications, the algorithm can be continuously updated according to the data output by the battery management system, which further improves the accuracy of charging curve reconstruction and state estimation.
  • Fig. 1 is the flow chart of the method provided by the present invention.
  • FIG. 3 is a schematic diagram of state estimation based on the charging curve reconstruction result of the present invention.
  • Figure 4 is a capacity increase curve derived from the reconstituted charging curve.
  • the charging curve reconstruction and state estimation method provided by the present invention is shown in FIG. 1, and specifically includes the following parts:
  • Step 1 Obtain the complete charging curve of the battery as training data, which specifically includes: charging by common charging schemes such as constant current charging, constant current and constant voltage charging, multi-stage constant current charging, and pulse charging.
  • common charging schemes such as constant current charging, constant current and constant voltage charging, multi-stage constant current charging, and pulse charging.
  • the daily charging curves of batteries in different aging states are obtained, including battery charging current, voltage, temperature and other signals under a given charging scheme.
  • Step 2 Divide the charging curve into data segments, and discretize the data segment and the charging curve, specifically including: determining a segment length, sliding the segment length on the charging curve, and dividing the charging curve in step 1 into A data segment of a certain length, each segment includes a sampled signal at each moment, such as voltage, current, temperature, etc. At the same time, the obtained data segments are sampled at fixed time intervals or voltage intervals to complete the complete charging curve for discretization processing.
  • Step 3 Use a deep learning algorithm to establish a mapping relationship between the data segment and the complete charging curve, which specifically includes: selecting a deep learning algorithm whose input is the discretized data segment in step 2, and the output is the discretized complete charging curve. Discretization results.
  • Step 4 During the actual application of the battery, collect the charging segment data as the input of the deep learning algorithm, and output the complete charging curve, which specifically includes: in the actual operation of the battery, through the battery management system, according to the preset segment segmentation rule in step 2. Collect charging clips and input the deep learning algorithm trained in step 3 to obtain the estimated complete charging curve.
  • a voltage window of 200mV is used to obtain the charging segment
  • a convolutional neural network is used to estimate the complete charging curve.
  • FIG. 2 shows the comparison between the charging curve reconstructed based on the present invention and the actual curve, which reflects the high accuracy that the method can achieve.
  • Step 5 Extract the battery state from the complete charging curve, which specifically includes: in the constant current charging curve as shown in FIG. 3 , the horizontal axis is the charged power, and the vertical axis is the battery voltage.
  • the power value corresponding to the complete charging process of the battery from the lower cut-off voltage to the upper cut-off voltage is the maximum capacity of the battery; the integral of the voltage to the charged power during the charging process is the maximum energy of the battery ( The sum of light and dark shades in the figure).
  • the state of charge (SOC) of the battery can be extracted based on the reconstructed complete charging curve, which is the ratio of the battery power corresponding to the current voltage to the maximum capacity.
  • the integration of the charged power from the lower cut-off voltage to the current voltage is the current battery energy (dark shaded in the figure), and the ratio of it to the maximum energy of the battery is the state of energy (SOE). Since this method can reconstruct a complete charging curve, the voltage change of the battery during the charging process can be predicted when the battery is not fully charged, so that the charging power of the battery can be evaluated, which is the state of power (SOP) . At the same time, by differentiating the reconstructed charging curve, this method can reconstruct the capacity increment curve of the battery (differential of electric quantity to voltage), differential voltage curve (differential of voltage to electric quantity), etc., which helps to understand the internal mechanism of the battery. Analysis, for example, the case of the capacity increment curve obtained by this method is shown in Figure 4.
  • Step 6 After collecting a large number of battery charging curves, update the algorithm, which specifically includes: after the battery has been running for a period of time, summarize the complete charging curve collected by the battery management system through the data platform, and use this part of the data as new training data.
  • the methods in steps 1 to 3 can be used to retrain the new deep learning algorithm, or use transfer learning and other means to fine-tune some parameters of the previously trained algorithm. In this way, the adaptive update of the deep learning algorithm with the working state of the battery can be realized.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Secondary Cells (AREA)
  • Charge And Discharge Circuits For Batteries Or The Like (AREA)

Abstract

一种基于人工智能的锂离子电池充电曲线重构及状态估计方法，由此可以实现电池多种状态的估计；方法以充电片段数据作为输入，使用深度学习方法重构出完整的充电曲线，进而可从完整的充电曲线中提取电池的多种状态，包括电池的最大容量、最大能量、荷电状态、能量状态、功率状态和容量增量曲线等；电池充电曲线重构及状态估计方法可以随电池工作状态变化自适应更新。

Description

基于人工智能的锂离子电池充电曲线重构及状态估计方法 技术领域
本发明涉及电池系统领域，尤其涉及锂离子电池的状态估计。
背景技术
锂离子电池在电池实际运行过程中，由于电池管理系统仅能采集到电池的电压、电流、温度等信号的片段，电池内部状态无法直接测量，因此其状态仅能依靠测得信号进行估计。现有的状态估计方法往往只能针对某几个特定状态，而假设其他状态已知，因此在估计的全局性上存在较大的局限性。例如，对电池容量的估计往往只关注容量与充电曲线特征之间的关系建立，而忽视了其他状态的估计。事实上，电池充电曲线(充电电压与充入电量的关系)反映了大量的电池状态信息，能够满足全面精确表征电池状态的要求。但实际应用中电池往往不会进行满充、满放操作，电池管理系统往往仅能采集到部分充电曲线。因此如果能基于获取的较为精确的部分充电曲线片段，通过必要技术手段重构出完整的充电曲线，则对于电池状态估计手段的改进以及电池管理功能的提升，具有十分重要的意义。
发明内容
有鉴于此，本发明提供了一种基于人工智能的锂离子电池充电曲线重构及状态估计方法，具体包括以下步骤：
步骤一、获取采用不同充电方式时，不同老化状态下电池完整的电压/电流充电曲线作为训练数据；
步骤二、通过适合的分割方式，将获取的充电曲线分割为数据片段，并将数据片段和充电曲线进行离散化处理；
步骤三、利用步骤二得到的离散化数据片段，对选择的深度学习算法进行训练，建立各数据片段与完整充电曲线的映射关系；
步骤四、将训练好的深度学习算法进行在线应用，将电池管理系统采集的实际充电片段数据输入所述深度学习算法，输出完整充电曲线；
步骤五、从完整充电曲线中提取欲估计的电池状态参数。
进一步地，所述方法还包括：
步骤六、在电池管理系统采集到一定数量的实际电池充电曲线后，对深度学习算法重新训练并更新。
进一步地，所述步骤一中获取采用不同充电方式时，不同老化状态下电池完整的电压/电流充电曲线，具体包括：采用常见如恒流充电、恒流恒压充电、多阶恒流充电、脉冲充电等的充电方案。通过电池试验、电池管理系统采样等方法，取得不同老化状态下的电池的日常充电曲线，包括给定充电方案下的电池充电电流、电压、温度等信号。
进一步地，所述步骤二具体包括：确定一个片段长度，将片段长度在充电曲线上滑动，由此将步骤一中的充电曲线划分为某长度的数据片段，每个片段包含每个时刻的采样信号，例如电压、电流、温度等。同时，将获得的数据片段采用固定时间间隔或电压间隔对片段进行采样，从而将完整充电曲线离散化处理。
进一步地，所述步骤三中深度学习算法具体采用卷积神经网络、密集连接网络、循环神经网络等。
上述本发明所提供的方法，通过部分充电片段即能够重构出电池完整的充电曲线，可以同时实现最大容量、最大能量、荷电状态、能量状态以及功率状态的 估计，并可以通过导出的容量增量曲线、微分电压曲线等实现电池老化的分析。在长期的应用中，还可根据电池管理系统输出的数据对算法实现持续更新，进一步提高了充电曲线重构与状态估计的精确性。
附图说明
图1是本发明所提供方法的流程图；
图2是基于本发明的充电曲线重构优选实例；
图3是基于本发明的充电曲线重构结果进行状态估计的示意图；
图4是从重构充电曲线中导出的容量增量曲线。
具体实施方式
上述仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，以下结合附图与具体实施方式对本发明作进一步的详细说明。
本发明所提供的充电曲线重构及状态估计方法如附图1所示，具体包括以下部分：
步骤一、获取电池的完整充电曲线作为训练数据，具体包括：采用常见的充电方案如恒流充电、恒流恒压充电、多阶恒流充电、脉冲充电等方式充电。通过电池试验、电池管理系统采样等方法，取得不同老化状态下的电池的日常充电曲线，包括给定充电方案下的电池充电电流、电压、温度等信号。
步骤二、将充电曲线分割为数据片段，并将数据片段和充电曲线进行离散化，具体包括：确定一个片段长度，将片段长度在充电曲线上滑动，由此将步骤一中的充电曲线划分为某长度的数据片段，每个片段包括每个时刻的采样信号，例如电压、电流、温度等。同时，将获得的数据片段采用固定时间间隔或电压间隔的采样，完成完整充电曲线进行离散化处理。
步骤三、使用深度学习算法建立数据片段与完整充电曲线的映射关系，具体包括：选取一种深度学习算法，其输入是步骤二中离散化的数据片段，输出则是离散化的完整充电曲线的离散化结果。
步骤四、在电池实际应用过程中，采集充电片段数据作为深度学习算法的输入，输出完整充电曲线，具体包括：在电池实际运行中，通过电池管理系统按步骤二中预设的片段分段规则采集充电片段，输入步骤三中训练好的深度学习算法，获得估计的完整充电曲线。本例中针对某款三元材料电池的恒流充电过程，采用200mV的电压窗口获得充电片段，并使用卷积神经网络估计完整充电曲线。图2示出了基于本发明重构出的充电曲线与实际曲线之间的对比，体现了该方法所能达到的较高精确性。
步骤五、从完整充电曲线中提取电池状态，具体包括：在如图3所示的恒流充电曲线中，其横轴为充入电量，纵轴为电池电压。在重构出完整充电曲线后，电池从下截止电压充电至上截止电压的完整充电过程所对应的电量值即为电池最大容量；该充电过程中电压对充入电量的积分即为电池最大能量(图中的浅色阴影与深色阴影之和)。此外，基于重构的完整充电曲线可以提取电池的荷电状态(State of charge,SOC)，即为当前电压所对应的电池电量与最大容量的比值。类似地，从下截止电压到当前电压对充入电量的积分即为当前电池能量(图中的深色阴影)，其与电池最大能量的比值即为能量状态(State of energy,SOE)。由于本方法可以重构出完整的充电曲线，在电池未充满的情况下可以预测电池在充电过程中的电压变化，由此可以评价电池的充电功率，即为功率状态(State of power,SOP)。同时，通过对重构的充电曲线进行微分，本方法可以重构出电池的容量增量曲线(电量对电压微分)，微分电压曲线(电压对电量微分)等，有助于对电池内部机理的分析，例如，本方法所获得的容量增量曲线案例如图4所示。
步骤六、在采集大量电池充电曲线后，对算法进行更新，具体包括：在电池运行一段时间后，通过数据平台汇总电池管理系统采集到的完整充电曲线，以此部分数据作为新的训练数据，对步骤三中的深度学习算法进行更新。可采用步骤一至三中的方法重新训练新的深度学习算法，或采用迁移学习等手段微调之前训练的算法的部分参数。由此可以实现深度学习算法随电池工作状态的自适应更新。
尽管已经示出和描述了本发明的实施例，对于本领域的普通技术人员而言，可以理解在不脱离本发明的原理和精神的情况下可以对这些实施例进行多种变化、修改、替换和变型，本发明的范围由所附权利要求及其等同物限定。

Claims (5)

  1. 一种基于人工智能的锂离子电池充电曲线重构及状态估计方法，其特征在于：具体包括以下步骤：
    步骤一、获取采用不同充电方式时，不同老化状态下电池完整的电压/电流充电曲线作为训练数据；
    步骤二、通过适合的分割方式，将获取的充电曲线分割为数据片段，并将数据片段和充电曲线进行离散化处理；
    步骤三、利用步骤二得到的离散化数据片段，对选择的深度学习算法进行训练，建立各数据片段与完整充电曲线的映射关系；
    步骤四、将训练好的深度学习算法进行在线应用，将电池管理系统采集的实际充电片段数据输入所述深度学习算法，输出完整充电曲线；
    步骤五、从完整充电曲线中提取欲估计的电池状态参数。
  2. 如权利要求1所述的方法，其特征在于：所述方法还包括：
    步骤六、在电池管理系统采集到一定数量的实际电池充电曲线后，对深度学习算法重新训练并更新。
  3. 如权利要求1所述的方法，其特征在于：所述步骤一中获取采用不同充电方式时，不同老化状态下电池完整的电压/电流充电曲线，具体包括：采用恒流充电、恒流恒压充电、多阶恒流充电、脉冲充电等方式进行充电；通过电池试验、电池管理系统采样方法，取得不同老化状态下的电池的日常充电曲线，包括对应各充电方案下的电池充电电流、电压、温度信号。
  4. 如权利要求1所述的方法，其特征在于：所述步骤二具体包括：确定一个片段长度，使该片段长度在充电曲线上滑动，由此将步骤一中获取的充电曲线划分为某长度的数据片段，每个片段包含每个时刻的采样信号；将获得的数据片段采用固定时间间隔或电压间隔对各数据片段进行采样，从而将完整充电曲线离散化处理。
  5. 如权利要求1所述的方法，其特征在于：所述步骤三中深度学习算法具体采用卷积神经网络或者密集连接网络或者循环神经网络。
PCT/CN2021/116036 2020-11-16 2021-09-01 基于人工智能的锂离子电池充电曲线重构及状态估计方法 WO2022100230A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/916,041 US20230168304A1 (en) 2020-11-16 2021-09-01 Artificial intelligence (ai)-based charging curve reconstruction and state estimation method for lithium-ion battery

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011281459.5A CN112379273B (zh) 2020-11-16 2020-11-16 基于人工智能的锂离子电池充电曲线重构及状态估计方法
CN202011281459.5 2020-11-16

Publications (1)

Publication Number Publication Date
WO2022100230A1 true WO2022100230A1 (zh) 2022-05-19

Family

ID=74584812

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/116036 WO2022100230A1 (zh) 2020-11-16 2021-09-01 基于人工智能的锂离子电池充电曲线重构及状态估计方法

Country Status (3)

Country Link
US (1) US20230168304A1 (zh)
CN (1) CN112379273B (zh)
WO (1) WO2022100230A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114720882A (zh) * 2022-05-20 2022-07-08 东南大学溧阳研究院 基于神经网络和迁移模型的锂离子电池最大容量衰退曲线的重构方法
CN116068449A (zh) * 2023-03-23 2023-05-05 中国科学技术大学 一种物理模型约束深度神经网络的锂电池评估系统及方法

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112379273B (zh) * 2020-11-16 2021-07-30 北京理工大学 基于人工智能的锂离子电池充电曲线重构及状态估计方法
KR102395182B1 (ko) * 2021-11-30 2022-05-10 모나일렉트릭 주식회사 배터리 진단 방법 및 그 장치
CN117289141A (zh) * 2023-11-22 2023-12-26 深圳市麦迪瑞科技有限公司 基于人工智能的电动自行车充电状态监测方法

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107769335A (zh) * 2017-12-01 2018-03-06 深圳市森树强电子科技有限公司 一种多模式锂电池智能充电管理方法及装置
CN108072845A (zh) * 2017-12-20 2018-05-25 上海理工大学 基于不完整充电电压曲线的锂电池容量估计方法
CN108398652A (zh) * 2017-05-26 2018-08-14 北京航空航天大学 一种基于多层特征融合深度学习的锂电池健康状态评估方法
CN109934955A (zh) * 2019-02-28 2019-06-25 深圳智链物联科技有限公司 充电模式识别方法、装置、终端设备及存储介质
CN110058175A (zh) * 2019-05-05 2019-07-26 北京理工大学 一种动力电池开路电压-荷电状态函数关系的重构方法
US20200011932A1 (en) * 2018-07-05 2020-01-09 Nec Laboratories America, Inc. Battery capacity fading model using deep learning
CN110673039A (zh) * 2019-11-11 2020-01-10 安徽优旦科技有限公司 一种基于大数据的磷酸铁锂电池soc充电在线校正方法
CN111781504A (zh) * 2020-08-03 2020-10-16 北京理工大学 一种锂离子动力电池老化状态识别与开路电压重构方法
CN112379273A (zh) * 2020-11-16 2021-02-19 北京理工大学 基于人工智能的锂离子电池充电曲线重构及状态估计方法

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030184307A1 (en) * 2002-02-19 2003-10-02 Kozlowski James D. Model-based predictive diagnostic tool for primary and secondary batteries
JP5043777B2 (ja) * 2007-08-22 2012-10-10 パナソニック株式会社 非水電解質二次電池の充電方法
TW200933181A (en) * 2008-01-30 2009-08-01 Inventec Appliances Corp A method for estimating lithium battery lifetime of portable electronic products
FR3010532B1 (fr) * 2013-09-11 2017-06-09 Commissariat Energie Atomique Procede, dispositif et systeme d'estimation de l'etat de charge d'une batterie
KR102221756B1 (ko) * 2014-07-18 2021-03-02 삼성전자주식회사 배터리의 상태를 추정하는 방법 및 장치
KR102332399B1 (ko) * 2015-02-06 2021-11-29 삼성전자주식회사 배터리 상태 추정 장치 및 방법
KR20160101506A (ko) * 2015-02-17 2016-08-25 삼성전자주식회사 배터리 신호 세그먼트 데이터의 확률 추론을 기반으로 한 배터리 상태 추정 방법 및 장치
US11691518B2 (en) * 2017-07-21 2023-07-04 Quantumscape Battery, Inc. Predictive model for estimating battery states
WO2018019313A2 (zh) * 2017-11-17 2018-02-01 深圳市恒翼能科技有限公司 基于电池部分充放电数据重建完整充放电数据的方法及系统
JP7106362B2 (ja) * 2018-06-15 2022-07-26 大和製罐株式会社 蓄電池の充放電曲線推定装置および充放電曲線推定方法
US20200124675A1 (en) * 2018-10-23 2020-04-23 Samsung Electronics Co., Ltd. Method and system for direct estimation of exact power in real-time
JP7204938B2 (ja) * 2020-03-17 2023-01-16 株式会社東芝 蓄電池の状態を評価する情報処理装置、情報処理方法、コンピュータプログラム及び情報処理システム
DE102020117609B8 (de) * 2020-07-03 2022-11-03 TWAICE Technologies GmbH Verarbeitung von Zustandsdaten einer Batterie zur Alterungsschätzung

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108398652A (zh) * 2017-05-26 2018-08-14 北京航空航天大学 一种基于多层特征融合深度学习的锂电池健康状态评估方法
CN107769335A (zh) * 2017-12-01 2018-03-06 深圳市森树强电子科技有限公司 一种多模式锂电池智能充电管理方法及装置
CN108072845A (zh) * 2017-12-20 2018-05-25 上海理工大学 基于不完整充电电压曲线的锂电池容量估计方法
US20200011932A1 (en) * 2018-07-05 2020-01-09 Nec Laboratories America, Inc. Battery capacity fading model using deep learning
CN109934955A (zh) * 2019-02-28 2019-06-25 深圳智链物联科技有限公司 充电模式识别方法、装置、终端设备及存储介质
CN110058175A (zh) * 2019-05-05 2019-07-26 北京理工大学 一种动力电池开路电压-荷电状态函数关系的重构方法
CN110673039A (zh) * 2019-11-11 2020-01-10 安徽优旦科技有限公司 一种基于大数据的磷酸铁锂电池soc充电在线校正方法
CN111781504A (zh) * 2020-08-03 2020-10-16 北京理工大学 一种锂离子动力电池老化状态识别与开路电压重构方法
CN112379273A (zh) * 2020-11-16 2021-02-19 北京理工大学 基于人工智能的锂离子电池充电曲线重构及状态估计方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
LI CHAORAN;XIAO FEI;FAN YAXIANG;TANG XIN;YANG GUORUN: "Joint Estimation of the State of Charge and the State of Health Based on Deep Learning for Lithium-ion Batteries", PROCEEDINGS OF THE CSEE, 10 April 2020 (2020-04-10), pages 681 - 692, XP055929392, ISSN: 0258-8013, DOI: 10.13334/j.0258-8013.pcsee.191867 *
ZHENG YUEJIU; OUYANG MINGGAO; HAN XUEBING; LU LANGUANG; LI JIANQIU: "Investigating the error sources of the online state of charge estimation methods for lithium-ion batteries in electric vehicles", JOURNAL OF POWER SOURCES, vol. 377, 1 February 2018 (2018-02-01), AMSTERDAM, NL, pages 161 - 188, XP085322261, ISSN: 0378-7753, DOI: 10.1016/j.jpowsour.2017.11.094 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114720882A (zh) * 2022-05-20 2022-07-08 东南大学溧阳研究院 基于神经网络和迁移模型的锂离子电池最大容量衰退曲线的重构方法
CN114720882B (zh) * 2022-05-20 2023-02-17 东南大学溧阳研究院 锂离子电池最大容量衰退曲线的重构方法
CN116068449A (zh) * 2023-03-23 2023-05-05 中国科学技术大学 一种物理模型约束深度神经网络的锂电池评估系统及方法

Also Published As

Publication number Publication date
CN112379273A (zh) 2021-02-19
CN112379273B (zh) 2021-07-30
US20230168304A1 (en) 2023-06-01

Similar Documents

Publication Publication Date Title
WO2022100230A1 (zh) 基于人工智能的锂离子电池充电曲线重构及状态估计方法
WO2021185308A1 (zh) 一种电动汽车动力电池组健康状态在线确定方法和系统
CN110221225B (zh) 一种航天器锂离子电池循环寿命预测方法
CN108896914B (zh) 一种锂电池健康状况的梯度提升树建模与预测方法
CN111624494B (zh) 基于电化学参数的电池分析方法和系统
CN112798960B (zh) 一种基于迁移深度学习的电池组剩余寿命预测方法
CN110658459B (zh) 基于双向循环神经网络的锂离子电池荷电状态估计方法
CN107219463B (zh) 基于电瓶车状态数据的电池使用寿命预测方法
CN103278777B (zh) 一种基于动态贝叶斯网络的锂电池健康状况估计方法
WO2022100229A1 (zh) 一种基于人工智能的锂离子电池系统soc估计方法
CN110703114A (zh) 一种基于电-热-神经网络耦合模型的动力电池soc和sot联合状态估计方法
CN105676138A (zh) 一种预测电池的剩余电量的方法和系统
CN104569844A (zh) 阀控密封式铅酸蓄电池健康状态监测方法
CN112731161B (zh) 基于少量数据混插的非线性数据特征提取和分类预测方法
CN111426957A (zh) 一种模拟车辆工况动力电池荷电状态soc估算优化方法
CN110658462A (zh) 一种基于数据融合与arima模型的锂电池在线寿命预测方法
CN116224085A (zh) 一种基于数据驱动的锂电池健康状态评估方法
CN114839540A (zh) 一种基于深度学习算法的锂离子电池剩余寿命在线预测方法
CN114200333A (zh) 一种铅酸蓄电池容量预测方法
CN116796623A (zh) 锂电池soc的预测方法、系统、设备和介质
CN113255199A (zh) 一种基于粒子滤波的电池剩余寿命预测方法
CN116593899A (zh) 一种基于XGBoost算法预测退役电池容量的方法
Dong et al. State of health estimation and remaining useful life estimation for Li-ion batteries based on a hybrid kernel function relevance vector machine
CN117272783A (zh) 一种基于云边协同多模型融合的动力电池寿命预测方法
CN115598546A (zh) 锂离子电池soh、soc和rul的联合估计方法

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: 21890758

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21890758

Country of ref document: EP

Kind code of ref document: A1