← 返回

一种电动汽车锂离子电池膝点直接预测的混合算法

A Hybrid Algorithm for Direct Knee Point Prediction of Lithium-Ion Battery in Electric Vehicle

语言:

中文摘要

针对现有锂电池膝点预测方法依赖早期循环数据且通用性差的问题,本文提出了一种直接预测算法。该方法仅需电池前10次循环的数据即可实现可靠的膝点预测,并结合迁移学习算法提升了模型的泛化能力。

English Abstract

The majority of knee point prediction approaches used today are indirect prediction approaches with poor generality and high input requirements for early battery cycle data. In order to solve this issue, the direct knee point prediction approach proposed in this study can reliably forecast the knee point with just the battery data from the first 10 cycles. Meanwhile, the transfer learning algorith...
S

SunView 深度解读

该技术对阳光电源的储能业务(PowerTitan、PowerStack及户用储能)具有极高价值。锂电池的“膝点”(容量衰减加速点)是评估储能系统寿命和安全性的关键指标。通过引入该直接预测算法,阳光电源的iSolarCloud智能运维平台可实现更精准的电池健康状态(SOH)监测与寿命预警,显著降低运维成本。此外,该算法对早期数据的高效利用,有助于优化BMS策略,提升储能系统在调峰调频等高频次应用场景下的资产管理效率,为客户提供更具竞争力的全生命周期储能解决方案。