Autonomous semi-trucks (tractor-trailer combinations) call for dedicated path planners for on-road driving due to their unique articulated structures and large dimensions, which are often ignored in the current path/trajectory planner design for normal autonomous cars, and which could lead to off-tracking without proper modification. While numerical optimization approaches such as model predictive control-based path planning algorithms do exist, these complex planners usually require large computation resources to meet real-time requirements. In this paper, we propose a novel path planning approach based on semi-supervised learning. We train an encoder-decoder type of deep neural network to generate / plan paths with the objective to minimize the off-track of the tractor-trailer swept area. The encoder encodes input information such as lane markings, static obstacles, and potentially other features, and pass it to the decoder to generate a planned path. The key to our approach is the construction of a path cost function that scores / penalizes each network-generated path based on its deviation from the lane center, the path smoothness and collision with any static obstacles, and backpropagates the cost of the paths through the encoder-decoder network to train it. As the path cost function acts as a critic of the path quality, our approach requires no collected data from expert driving for training, but only randomly generated samples of many possible combinations of lane shapes and obstacles arrangements. The proposed approach is finally verified in simulation environment.
A Path Planning Approach for Tractor-Trailer System based on Semi-Supervised Learning
2022-10-08
850099 byte
Conference paper
Electronic Resource
English
Semi-trailer tractor main vehicle and semi-trailer tractor
European Patent Office | 2023
|Flow guider special for semi-trailer tractor and semi-trailer tractor
European Patent Office | 2020
|Tractor-semi-trailer operation
Engineering Index Backfile | 1930
|A LEARNING-BASED METHOD AND SYSTEM FOR PATH PLANNING OF AN AUTONOMOUS TRACTOR-TRAILER
European Patent Office | 2023
|