Driving cycle recognition neural network algorithm based on the sliding time window for hybrid electric vehicles

Driving cycles greatly influence the fuel economy and exhaust of the vehicle, especially in hybrid electric vehicles. The purpose of this study is to develop a method to identify the type of driving cycle with better accuracy and less sampling time than other driving cycle recognition algorithms. A...

Full description

Saved in:
Bibliographic Details
Published in:International journal of automotive technology 2015-08, Vol.16 (4), p.685-695
Main Authors: Wang, J., Wang, Q. N., Zeng, X. H., Wang, P. Y., Wang, J. N.
Format: Article
Language:English
Subjects:
Accuracy
Algorithms
Automobile driving
Automotive Engineering
Computer simulation
Consumption
Control algorithms
Dynamic programming
Electric vehicles
Energy consumption
Energy efficiency
Energy management
Engineering
Fuel consumption
Global positioning systems
GPS
Hybrid vehicles
Matlab
Methods
Neural networks
Principal components analysis
R&D
Recognition
Research & development
Simulation
Studies
Windows (intervals)
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Driving cycles greatly influence the fuel economy and exhaust of the vehicle, especially in hybrid electric vehicles. The purpose of this study is to develop a method to identify the type of driving cycle with better accuracy and less sampling time than other driving cycle recognition algorithms. A driving cycle recognition algorithm based on Learning Vector Quantization neural network is first created to analyze four selected representative standard driving cycles. Micro-trip extraction and box-and-whisker plots are then applied to ensure the diversity and magnitude of training samples. Finally, a sample training simulation is conducted to determine the minimum neuron number of learning vector quantization network, using the simulation platform of Matlab/Simulink. Afterwards, we simplify the structure of the recognition model to reduce data convergence time. Simulation results show the feasibility and efficiency of the proposed algorithm, which decreases the time window length from 120 s to 60 s with acceptable accuracy. Furthermore, the driving cycle recognition algorithm is used in a series-parallel hybrid vehicle model to improve the fuel economy by about 6.29%.
ISSN:1229-9138
1976-3832
DOI:10.1007/s12239-015-0069-3