Distributed computing based on AI algorithms in battery early warning and SoH prediction of the intelligent connected ve
- PDF / 2,896,573 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 60 Downloads / 198 Views
(0123456789().,-volV)(0123456789(). ,- volV)
S.I. : BIO-INSPIRED COMPUTING FOR DLA
Distributed computing based on AI algorithms in battery early warning and SoH prediction of the intelligent connected vehicles Haibo Xiao1,2 • Yaonan Wang1 • Di Xiao2 • Yougui Zhou2 Received: 18 June 2020 / Accepted: 7 October 2020 Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract The foremost task of the battery management system is to estimate the end-of-life batteries, their capacity and internal resistance, which are commonly used to evaluate the State of Health (SoH) for the battery cells and packs. The SoH for batteries plays a vital role in intelligent connected vehicles. To evaluate the warning conditions of battery life units, this paper suggests the artificial intelligence-assisted particle swarm optimization for the SoH prediction of the intelligent connected vehicles. The key elements of this method include the battery ageing cycle, identifying the SoH and battery health forecasting based on the gradual changing processes of the battery. Besides, several accelerated test results are reported for intelligent connected vehicles, using battery mode packages. To validate the concept of the SoH prediction method, a simulation test-bed is developed and test results indicate that the idea is projected with a higher prediction rate. Keywords Intelligent connected vehicles SoH prediction method Artificial intelligence
1 Outline about battery life estimation Owing to the low energy capacities and higher internal resistance, batteries such as lithium-ion and platinum acid cells are degrading over time [1, 2]. For economical and efficiency purposes, it is essential to forecast their deterioration levels [3] and their remaining useful life (RULs) [4, 5]. For example, the driven range of an electric vehicle is linked explicitly with the battery capacity. The RUL‘s prediction during the design phase and operations is crucial for the evaluation of energy stock, the depreciation [6], the warranty, second-life value, insurance and preventive maintenance applications [7]. Unfortunately, battery breakdown is due to several complex chemical/mechanical and first-principle interactions, and physical modelling is very challenging [8, 9]. Batteries are sometimes over-sized and under-used to mitigate lifetime uncertainty, contributing to higher & Haibo Xiao [email protected] 1
College of Electrical and Information Engineering, Hunan University, Changsha 410082, Hunan, China
2
Hunan Tanzhou Education Network Technology Co., Ltd, Changsha 410205, Hunan, China
device costs and less effective performance [10]. Therefore, innovative methods are required to ensure accurate health forecasts and form an integral part of an advanced battery management system or energy management system [11, 12]. On the other hand, traditional SoH prediction models are adapted to laboratory data of a certainly undefined parametric method and electrochemical stimulation from the bottom-up of physical degradation [13–15]. First, it demands that
Data Loading...
