Search Results

It’s hard to achieve complete self-driving using hand-crafting generalized decision-making rules, while the end-to-end self-driving system is low in complexity, does not require hand-crafting rules, and can deal with complex situations. Modular-based self-driving systems require multi-task fusion and high-precision maps, resulting in high system complexity and increased costs. In end-to-end self-driving, we usually only use camera to obtain scene status information, so image processing is very important. Numerous deep learning applications benefit from multi-task learning, as the multi-task learning can accelerate model training and improve accuracy with combine all tasks into one model, which reduces the amount of calculation and allows these systems to run in real-time. Therefore, the approach of obtaining rich scene state information based on multi-task learning is very attractive. In this paper, we propose an approach to multi-task learning for semantics, geometry and motion.

A full drive-by-wire electric vehicle, named Urban Future Electric Vehicle (UFEV) is developed, where the four wheels' traction and braking torques, four wheels' steering angles, and four active suspensions (in the future) are controlled independently. It is an ideal platform to realize the optimal vehicle dynamics, the marginal-stability and the energy-efficient control, it is also a platform for studying the advanced chassis control methods and their applications. A centralized control system of hierarchical structure for UFEV is proposed, which consist of Sensor Layer, Identification and Estimation Layer, Objective Control Layer, Forces and Motion Distribution Layer, Executive Layer. In the Identification and Estimation Layer, identification model is established by utilizing neural network algorithms to identify the driver characteristics. Vehicle state estimation and road identification of UFEV based on EKF and Fuzzy Logic Control methods is also conducted in this layer.