WSEAS Transactions on Circuits and Systems
Print ISSN: 1109-2734, E-ISSN: 2224-266X
Volume 17, 2018
Adaptive Driver Model for Velocity Profile Prediction
Authors: ,
Abstract: Modern driver assistant systems are responsible for maintaining safe and reliable operation and reducing the energy consumption in electric vehicles since these systems have to possess the capability to predict the expected load. Drive cycles can not fully coincide with real driving behaviour and a one-time test does not reflect the overall traffic and road conditions. The Interval-Type-2 (IT2) Fuzzy System is proved to be a higly efficient tool for modeling uncertainties. In contrast to conventional Type-1 fuzzy modeling an IT2 Fuzzy System has the ability to deal with flexible the various types of uncertainties and modeling errors simultaneously and approximates better real-life systems. This paper presents an Adaptive IT2 Fuzzy System for velocity profile forecasting from the measured velocity and acceleration data. The adaptive driver model is based on Interval Type-2 fuzzy sets. Histograms of input features are used for generating membership functions which parameters are adaptiveley tuned according to the driver’s behaviour. Simulation results validate the efficiency and demonstrate that the proposed method is a viable alternative of conventional time series prediction.
Search Articles
Keywords: Interval Type-2 Fuzzy System (IT2FS), adaptive fuzzy model, driver model, driving cycle, velocity profile, driver assistant system, time series prediction, intermittent operation, electric vehicles, intelligent systems
Pages: 138-145
WSEAS Transactions on Circuits and Systems, ISSN / E-ISSN: 1109-2734 / 2224-266X, Volume 17, 2018, Art. #17