Our Research Topics
We would like to welcome any student who is interested in the below research explanation exciting areas and is willing to push the boundaries of these scientific areas, and as well as the boundaries of his or her own potential.
Advanced driver-assistance systems (ADAS) are developed to automate/adapt/enhance vehicle systems for safety and better driving. Based on intelligent sensor technology (camera, radar, lidar, ultrasound,), constantly monitor the vehicle surroundings as well as the driving behavior to detect potentially dangerous situations at an early stage.
​
We have developed several vision-based ADAS, such as lane departure warning system (LDWs), forward collision warning system (FCWs), rear safety assist system (RSAS), trailer truck around view monitor (TAVM), blind spot detection (BSD), Traffic Sign Recognition (TSR), and so on, in order to alert the driver to any upcoming hazards or road conditions in time.
​
Currently our Lab is mainly focus on deep learning technologies which needs training artificial neural networks on lots of data. The Nvidia GPUs and TensorFlow frameworks are used to build and train deep neural networks.
We specialize in advancing the technology of autonomous and collaborative mobility systems. Our Lab major research focus on control algorithm development and deployment including various methods from nonlinear control theory, robust optimal control, predictive control, machine learning, artificial intelligence, reinforcement learning for vehicle dynamic control system, adaptive cruise control system and adaptive collaborative vehicle control system. Currently we have developed a novel intelligent eco-cruise control systems for running a vehicle on roads with up-down slopes and continuous curves, considering both energy-saving and safety.
​
Based on the vehicle longitudinal and lateral dynamic models and fuel consumption model, we presents an efficient way of using finite horizon dynamic programming and taking a penalty function as the soft constraints with pre-specified hard limitations for the design of nonlinear model-based predictive controller; it is aimed at providing the most appropriate velocity, perfect time of accelerating and decelerating, successfully achieving economy speed and feasible steering angle on curves, simultaneously. Finally, the approximate real car dynamic modelling and driving simulators are created in vehicle dynamics software MSE CarSim for virtual testing and quantitative analysis.
Control systems are a key enabling technology for the increase in functionality and safety of numerous critical applications such as automation systems, aircraft systems, transportation systems, medical equipment, diagnosis systems, manufacturing systems, and embedded systems. Our Lab focuses its research on innovative control strategies, methodologies, theories as well as prototypes to enable intelligent, robustness, decision making and control in various systems or applications. We are conducting research on advanced control methods, such as optimal control, robust control, adaptive fuzzy-neural networks (FNNs), moving sliding mode control, model-based predictive control, reinforcement learning, genetic algorithm and so on, for uncertain time-delay nonlinear systems.
We have proposed an optimal control approach for solving the general robust control problem of active pantograph suspension systems with actuator delays and time-varying contact force such that both the stabilization and optimal performance. Based on Bellman's optimality principle and Razumikhin theorem, the general robust control design problem can be equivalently transformed into an optimal control problem with the amount of matched uncertainties involved in the performance index. A suitable linear state feedback control law is characterized via Lyapunov stability theory to ensure quadratic stability and performance robustness of the closed-loop systems.
​
Furthermore, a novel cooperative adaptive cruise control (CACC) approach is presented which concerns with not only safety and riding comfort but also the enhancement of fuel efficiency of the vehicle. We presented an adaptive neuro-fuzzy predictive control (ANFPC) approach for CACC systems, which can assists each vehicle in the platoon in following its predecessor within a desired distance such that the tracking error is minimized thereby ensuring vehicle safety, string stability and fuel economy, simultaneously.