Data Analysis and Visualization Platform Design for Batteries Using Flask-Based Python Web Service

Beichen Liang,Zongwei Liang,Linfeng Zheng,Yubin Zheng,Zuyi Liang

doi:10.3390/wevj12040187

Beichen Liang, Zongwei Liang + Show 3 more

Open Access

https://doi.org/10.3390/wevj12040187

Copy DOI

Abstract

Battery operating data of electric vehicles is becoming increasingly quantified and complicated. A data analysis platform is necessary to excavate high-value battery status information for more efficient battery management. This paper proposes a Flask framework and Pyecharts-based lithium-ion data analysis and visualization platform. The design processes including the front-end and back-end frameworks, data preprocessing, data visualization, and data storage are elaborated. In the proposed data platform, a case study of battery state of charge estimation using different machine learning methods is demonstrated, and most of the estimation errors are less than 2.0%, highlighting the effectiveness of the platform.

Highlights

The development of electric vehicles (EVs) has become a global consensus to solve the problems of environmental pollution and resource scarcities [1]
It is essential to introduce new and efficient data analysis technologies into battery systems of EVs and build a corresponding platform that can dig out high-value battery status information, significantly improving battery management technology [9], as well as being helpful for EVs companies
This paper proposes a Python-based battery data analysis and visualization platform design scheme, and applies the platform to estimate the state of charge (SOC) of lithium-ion batteries

Summary

Introduction

The development of electric vehicles (EVs) has become a global consensus to solve the problems of environmental pollution and resource scarcities [1]. It is essential to introduce new and efficient data analysis technologies into battery systems of EVs and build a corresponding platform that can dig out high-value battery status information, significantly improving battery management technology [9], as well as being helpful for EVs companies. This paper proposes a Python-based battery data analysis and visualization platform design scheme, and applies the platform to estimate the state of charge (SOC) of lithium-ion batteries. The technical principle is that Jinja redirects the URL login path to the large-screen template “index” that has been rendered, for automatic escape. Only the internal personnel with the password can log in to the large visualization screen to view the battery status information, which improves the security and reliability of the platform to a certain extent

Data Preprocessing

Data Visualization

Findings

Data Storage