Energy management for battery-powered embedded systems
ACM Transactions on Embedded Computing Systems (TECS)
Managing battery lifetime with energy-aware adaptation
ACM Transactions on Computer Systems (TOCS)
Managing battery lifetime with energy-aware adaptation
ACM Transactions on Computer Systems (TOCS)
On Dynamic Reconfiguration of a Large-Scale Battery System
RTAS '09 Proceedings of the 2009 15th IEEE Symposium on Real-Time and Embedded Technology and Applications
Scheduling of Battery Charge, Discharge, and Rest
RTSS '09 Proceedings of the 2009 30th IEEE Real-Time Systems Symposium
Dependable, efficient, scalable architecture for management of large-scale batteries
Proceedings of the 1st ACM/IEEE International Conference on Cyber-Physical Systems
Large-scale battery system modeling and analysis for emerging electric-drive vehicles
Proceedings of the 16th ACM/IEEE international symposium on Low power electronics and design
Scheduling battery usage in mobile systems
IEEE Transactions on Very Large Scale Integration (VLSI) Systems
| Hi-index | 0.00 |
A battery management system (BMS) is responsible for protecting the battery from damage, predicting battery life, and maintaining the battery in an operational condition. In this paper, we propose an efficient way of predicting the power requirements of electric vehicles (EVs) based on a history of their power consumption, speed, and acceleration, as well as the road information from a pre-downloaded map. The predicted power requirement is then used by the BMS to prevent the damage of battery cells that might result from high discharge rates. This prediction also helps BMS efficiently schedule and allocate battery cells in real time to meet an EV's power demands. For accurate prediction of power requirements, we need an accurate model for the power requirement of each given application. We generate this model in real time by collecting and using historical data of power consumption, speed, acceleration, and road information such as slope and speed limit. By using this information and the operator's driving pattern, the model extracts the vehicle's history of speed and acceleration, which, in turn, enables the prediction of the vehicle's (immediate) future power requirements. That is, the power requirement prediction is achieved by combining a real-time power requirement model and the estimation of the vehicle's acceleration and speed. The proposed approach predicts closer to the actual required power than a widely-used heuristic approach that uses measured power demand, by up to 69.2%.