摘要本课题的研究内容是对动力电池的容量估算 SOC（Status of Charge）和健康估算（State of Health）进行建模预测。现有的估算模型一般是将电池状态的非线性模型作 简化处理和线性化变换近似处理，因此不可避免地会对 SOC 和 SOH 的估算引入大量误差。77203
本项目的核心算法是使用局部敏感哈希算法和 K 近邻算法对电池状态建立估算模型。 首先对数据用局部敏感哈希算法做一次数据的拟合，将电池历史数据中的特征数减少， 再使用 K 近邻算法对电池的数据进行分类处理。本项目的云监控技术的主要流程是先建 立数据模型，再根据模型对电池寿命进行预测。
本课题研究的重点是对电池数据进行大数据分析和建模，建立电池系统 SOC 和 SOH 的精确估算模型，从而对电池的运行状态进行估算，提高电池系统的使用寿命和安全性。 通过云监控和大数据分析技术，可以较为精准地预测出电池的 SOC 和 SOH 值，为电池安 全的检测作出了贡献。
毕业论文关键词 容量估算 局部敏感哈希 K 近邻 数据拟合
毕 业 设 计 说 明 书 外 文 摘 要
Title The research on a new intelligent cloud monitoring technology for electric vehicle batteries
Abstract The research task of this thesis is to model and estimate the SOC(State of Charge) and SOH (State of Health) of electric vehicle battery。 The existing estimation methods in general relies on the simplification and linear approximation to the non-linear model of battery status, which will inevitably introduce considerable errors in the estimation of SOC and SOH values。
The fundamental algorithms employed in this project are locality-sensitive hashing algorithm and K-nearest neighboring algorithm。 The implementation flow first uses locality sensitive hashing algorithm to perform a fitting procedure on the batteries’ historical data, for the purpose of reducing the dimension of characteristic features。 Subsequently, K-nearest neighbor algorithm is applied to conduct a classification analysis on battery data。 The main flow of the presented cloud monitoring technology includes establishing the data fitting models and estimating the battery life based on the established models。
This research focuses on big data analysis and modeling on a large amount of battery data, and establishes accurate estimation models for SOC and SOH of battery system。 The established models can be applied to estimate the battery life for the purpose of extending battery life and guaranteeing the security of battery system。 By employing cloud computing and big data analysis techniques, the presented system is capable of accurately predicting SOC and SOH values, making contributions to the detection of battery safety。
Keywords State of Charge，Locality-sensitive Hashing，K-nearest neighbor，Data fitting
本科毕业设计说明书 第 I 页
目 次
1 引言 1
1。1 研究背景 1
1。2 研究现状 1
1。3 本文主要工作 2
1。4 论文结构 2
2 理论知识 2
2。1局部敏感哈希（Locality-Sensitive Hashing）算法 新能源电动汽车动力电池的智能云监控技术研究:http://www.youerw.com/jisuanji/lunwen_88703.html