Vehicle Following in Intelligent Multi-Vehicle Systems Based on SSD-MobileNet
Weiqi Sun, Sicong Chen, Liangren Shi, Yuanlong Li, Zongli Lin · 2019
In this paper, we propose a vehicle following control algorithm based on object detection for intelligent multi-vehicle systems. We use the SSD-MobileNet as the object detection model, and collect more than 3,000 pictures of the target signs as the training dataset. By detecting the target signs from the camera on-board, we obtain the relative position between the front and follower vehicles. Based on the relative position, the following control of the intelligent vehicle is carried out. In particular, we use the discrete incremental PID control method for the longitudinal control of the vehicle, and apply the pure pursuit algorithm to calculate the vehicle steering angles. Experimental results show the effectiveness of the control algorithms.