Embedded Files
In this research, interacting UAVs exchange their sensor measurements, i.e., LiDAR, GPS, Altimeter, and inertial measurement units, by using graph theories to achieve a common agreement on simultaneous localization and mapping (SLAM).
In this project, we are seeking to determine the correlation between the plantar foot pressure map and the human multibody dynamics using a motion capture system and a foot pressure mat. We seek to design a portable-affordable wearable rehabilitation device for diabetic patients that can be used in nonmedical environments.
In this project, we design and prototype a new special apparatus for testing adhesive joints. We use the theory of nonlocal operators to propose linear constitutive creeps models, instead of common nonlinear ones, to predict the deformation of a bonded joint over time.
In this project, we develop integrated research in nonlocal control laws revolving around the robust stabilization of uncertain dynamical systems. We interpret robustness in a probabilistic sense, allowing for uncertainty propagation control instead of the immunization against all possible uncertainty outcomes.
In this project, we enhance PDEs with nonlocal operators to model natural phenomena especially with infinite dimensions. In addition, we study their initialization, realization, and integration of them, which are still considered as open problems.
Optimal Stewart Platform for Lower Extremity Robotic Rehabilitation
Optimal Stewart Platform for Lower Extremity Robotic Rehabilitation
This project was done with Dr. Poursina's research group, where the objective was to design a LQR-based gain-scheduling controller for a Stewart robot designed for the ankle rehabilitation of the patients with diabetic neuropathy. For this purpose, the gain-scheduling technique was used to change the gains of the LQR controller to the proper desired ones along with the desired trajectory. Using this time-varying optimal controller guarantees the optimal performance of the controller along the desired trajectory. The advantages of the proposed technique wereshown in controlling a Stewart platform.
Neural-Fuzzy Inference System to Control Dynamical Systems
Neural-Fuzzy Inference System to Control Dynamical Systems
Adaptive neural fuzzy inference system (ANFIS)-based control technique can be used to stabilize dynamical systems. For this purpose, a classical control algorithm can be designed to train the ANFIS-based controller on the responses of the closed-loop system under different uncertainties. The number of fuzzy rules can be reduced using a fusion function.
As an illustrative example, the proposed technique is employed to stabilize an under-actuated double inverted pendulum on the cart.
AI for Intelligent Cube Counting Code
AI for Intelligent Cube Counting Code
Cube problems are part of DAT exams where participants must determine the number of cubes that has one, two, or three painted faces. In this study, we have developed a code based on fuzzy logic to design Cube problems based on the chosen level of difficulties. Next, we have used image processing algorithm along an AI algorithm to determine the number of painted faces, which is mainly similar to a human observation.
Cleanroom Trigger Module
Cleanroom Trigger Module
- This wireless module was built to log several sensors such as CO2, Oxygen, particle, temperature, humidity, and Nitrogen sensors, in clearrooms. The graphical-user-interface was developed in LabView.
Dynamics and Control of Hexapod Robots
Dynamics and Control of Hexapod Robots
In this project, we developed a code which can derive the dynamics equations of any multi-legged robots; found the suitable stable trajectory planning based on ZMP and HMS criteria. The proposed program generates the divided mass matrix, Centripetal and Coriolis matrix and gravity and external forces as m-files which can use as MEX level to increase the speed of calculation process.
Google Sites
Report abuse