Motorcycle’s Trajectory Tracking Model Based on Polynomial Least-squares Approximation

Mardiati, Rina and Trilaksono, Bambang R. and Gondokaryono, Yudi Satri and Wibowo, Sony Sulaksoni (2017) Motorcycle’s Trajectory Tracking Model Based on Polynomial Least-squares Approximation. Advanced Science Letters, 23 (5). pp. 4237-4541.

Preview

Text Final Draft 1570326158_v01.pdf Download (617kB) | Preview

Abstract

This paper provides a simple method for modeling the trajectory of motorcycle movements in mixed traffic condition. The trajectories model has been built using Polynomial Least-squares Approximation based on the coordinates that earned from the motorcycle maneuvers. In order to collect the data, this research uses video cameras fixed at elevated positions to record traffic flow. There are three steps of modeling motorcycle's trajectory: 1) data preprocessing to obtained pixel coordinate which represents motorcycle's trajectory, 2) coordinates transformation to convert the pixel coordinate to real-world coordinate, and 3) build a trajectory model using polynomial least-square approximation. This method was implemented for tracking the trajectory of motorcycle's maneuver. The maneuvers are when they try to avoid the collision with another motorcycle, car, and pedestrian. Based on experimental and analysis result for twenty samples of data, using a comparison of MSE for each degree of the polynomial, the trajectory model of motorcycle maneuver for avoiding an object in front of them can be described properly by 4th order polynomial equation. This result shows that trajectory model using polynomial least-squares approximation give an accurate outcome and also has the simpler computation. Furthermore, the conclusion of this study about trajectory tracking model for motorcycle may be instructive as a prior knowledge for building mixed traffic model based on their trajectory behaviors. This model also provides the information necessary for safe behavior decision making or motion planning.

Actions (login required)

View Item