Our research areas mainly include:
Despite widely used in industry, PID controller poses many drawbacks in modern control systems. Our research is to develop an advanced controller which is able to overcome the limitations of the conventional PID controller, e.g. trading-off between regulation and response time and particularly, to provide optimal control which usually cannot be obtained with PID.
Adaptive control employs various techniques into the conventional PID controller in order to provide knowledge and information of the controlled system, by thus, improving the performance of the controller particularly with non-linearity, uncertainty and/or time-varying parameters.
System models and estimation theory are commonly used in adaptive controller, e.g. Linear Quadratic Gaussian (LQG), Neural Network (NN), Model Reference Adaptive System (MRAS) and Learning Feed-Forward Controller (LFFC). Our research has been applied to various industrial control systems in diverse fields such as load sharing of DC motors, electromechanical motion, and robotics, etc.
Machine learning, a part of Artificial Intelligence (AI), has got many scientific attentions today. The core of machine learning is dealing with the representation of data instances and the generalization which performs on unseen data instances. Several approaches for machine learning considered include artificial Neural Network, support vector machines, clustering, etc. Our research on machine learning has been applied in various problems of engineering, e.g. selection and recognition systems, etc.
Power electronics which are mainly based on semiconductor device are used ubiquitously in the modern life, ranging from small electric appliances such as computer, TV, fluorescent lamp, etc. to very large applications in power systems: DC/DC converter, DC/AC inverter in renewable energy systems or power quality device, e.g. Static Var Compensator (SVC) and Static Compensator (STATCOM), etc. Managing and controlling these devices to mitigate its consequence, particularly related to the quality and problems are of significant concern.