Electrical-Electronics functionality in vehicles is increasing. Higher complexity control systems are needed for managing the new functions and their supervision requirements by means of suitable Energy, Power & Load Management algorithms. Forecasting function's behavior trajectory in state-space is nowadays a powerful tool for high-complex control systems design. An intelligent embedded time-series signal processing approach, based on Artificial Neural Networks (ANN) for vehicle status definition, to implement Energy Power & Load management algorithms is presented. Supported by an ANN algorithm, a time-series prediction strategy is implemented by means of function's control algorithm review. Such a new function control module reports to the master algorithm an intelligent processed frame where on-line function diagnosis, report of short-time probable future function behavior and its associated probability are included. These new intelligently processed inputs allow the system to distinguish clearly short-term function status and thus, self-inducing transient preventive work modes in order to well manage forthcoming functions events.
An embedded time-series prediction strategy based on ANN for vehicle status definition for energy, power & load management
Fahrzeug-Energiemanagement auf Basis künstlicher neuronale Netze
2003
13 Seiten, 6 Bilder, 8 Quellen
Aufsatz (Konferenz)
Englisch
British Library Conference Proceedings | 2003
|Adaptive energy management strategy for hybrid electric vehicle based on power prediction
SAGE Publications | 2023
|