A Methodology of Making a Real World Driving Pattern (Case Study on Major Indian Cities)

Automotive manufacturers are striving to improve the real-world driving experience for customers. To provide the better overall performance like fuel economy, we must have the real world driving behavior of our customers. The driving behavior of customers may change from city to city or region to region. Thus to make a real representative driving behavior of customer for each different region becomes tedious task. The present paper describes a detailed methodology to make a representative driving pattern quickly based on a Matlab script. The vehicles were driven in major Indian cities (Delhi, Mumbai, Nasik and Pune). The data was recorded with a frequency of 1 Hz. This data is segregated into different short trips. These short trips are grouped together based on maximum vehicle velocity in each short trip & velocity acceleration frequency distribution. Velocity acceleration frequency distribution of individual short trip is compared with each other quantitatively using Matlab script. The best short trip is selected from each group which best represents the group. The selected short trip is further modified based on trip time frequency analysis & reducing difference between velocity acceleration frequency distribution of whole group and selected trip. In the end all the modified trips are joined together along with idle time to make a complete driving pattern.