We focus on application oriented robotic systems: bio-inspired robots, construction robots, marine robots, and other novel robotic applications such as exoskeleton technology, security inspection robots or structural health monitoring robots etc, and related control, navigation and motion planning methods. A track-based robot with a novel quasi-zero-stiffness suspension system is developed; Several bio-inspired robots are developed either with multiple octopus-like tails,  undulating fins of omni-directional motion capability, or quadruped  mechanisms, and a crossing-discipline team is formed for several benchmark applications with marine robots currently supported by strategic focus area funds; We are also developing novel field and construction robots for special tasks in industry.

Currently, we have several PhD, Postdoc or Research Assistant positions opening for robot control, navigation, motion planning, sensor systems, and novel driving mechanisms etc. 

We focus on development of new generation of vibration isolators/absorbers or vibration suppression by employing nonlinear benefits, active/passive sound and vibration control, bio-inspired approach, vibration sensors based on quasi-zero/zero stiffness concepts, nonlinear energy harvesting, vibration based fault detection and structural health monitoring etc;

I initiated, proposed and have been leading my team to establish a bio-inspired limb-like or X-shaped structure/mechanism approach to vibration control (especially for passive vibration control), e.g., bio-inspired limb-like, human body inspired anti-vibration structures/mechanisms, while quadrilateral or polygon mechanisms are actually sharing the same nonlinear mechanism. This approach can overcome many obvious drawbacks of existing other methods for exploring nonlinear benefits in design of equivalent nonlinear stiffness, damping and inertia/mass characteristics. There are a series of applications and benchmark technical innovations in nonlinear stiffness, nonlinear damping and nonlinear inertia. 

The bio-inspired limb-like or X-shaped structures and applications

The quasi-zero or zero-stiffness based vibration sensors:

Nonlinear Energy Harvesting by employing nonlinear properties and structural benefits

Employing nonlinear benefits in nonlinear active control systems has been studied in our group and it is revealed that much more robust control can be achieved together with significant energy saving performance. This series of research results present a novel insight into robust control of nonlinear systems and benchmark applications can be found in vehicle suspension systems. 

We focus on a recently developed GFRF and nonlinear characteristic output spectrum (NCOS) based method and applications (vibration control, bio-systems, mechanical systems, fault detection etc); A systematic characteristic analysis approach has been established for the analysis and design of nonlinear systems in the frequency domain with a series of applications and pioneering work in vibration control, fault detection and energy harvesting etc. 

We have interests in developing vibration signals based fault detection or diagnosis methods. We proposed a fault diagnosis method for complex structures with least prior knowledge, less testing data and other practical requirements. A recent work is done for bolt loosening of a satellite-like structure with a “virtual beam” approach, which actually presents a nonlinear feature dynamics based method under further investigation now. We are also developing some new fault-related nonlinear features by using our research results in the frequency-domain nonlinear analysis and design.

We have interest in development of robust control theory or methods in general to solve critical theoretical and/or practical issues focusing more on application oriented control methods [D.1-5]. For system identification, we developed a new robust control approach, which casts the traditional identification problem with input output data into a robust control or robust output feedback control problem and thus achieve more powerful and robust identification performance [D.6-10]. We focus on kernel learning based approach, robust control approach, robust learning methods with NARX models, block-oriented nonlinear models (Wiener/ Hammerstein models), Volterra/Wiener kernel methods, and Neural networks for distributed/lumped parameter systems and stochastic systems, and their applications (mechanical systems, biological systems, neuronal systems, time-series data analysis, fault detection, structural health monitoring etc); Novel intelligent computing or optimization methods are also of interest to us.