MY RESEARCH
My research interests lie in the study of motion control of dynamic systems such as robot locomotion and manipulation, (non)linear control, autonomous vehicle systems.
3D Modelling and Simulation of Warehouse Robot
A physical model of pick and drop warehouse robot is designed in SolidWorks and completely simulated in MATLAB SimScape environment. The robot is a telescopic fork, and its motion is controlled by a state machine.
Parameter Estimation of a Two-Mass System
To enable model-based automatic tuning of the motion controllers, the mechanical parameters should be automatically identified during the start-up of a drive or during the drive operation
Parameter tuning of a model-based speed or position controller typically relies on the knowledge of mechanical parameters and some performance specifications (e.g., closed-loop bandwidth). However, datasheets of the mechanical components are not often available or the calculation of the mechanical parameters can be a highly complex task.
In this project,
a benchmark two-mass system is considered to estimate two of its
physical parameters. A torsion test bench made by Mitsubishi Heavy Industries
is used for the emulation of a two-mass system. Velocities and torques of drive,
and load sides are assumed known, whereas, spring coefficient of the shaft, and
inertia of the load side are considered unknowns. After mathematical modeling
of the two-mass system, a least square estimation algorithm is developed to
estimate the spring coefficient of the shaft, and inertia of the load motor. In
last, LSE algorithm is successfully validated through simulations and
experiment.
Modeling and Control of a Self-Balancing Bicycle
Self-balancing bicycles could form the basis for environment friendly and space efficient vehicles
In this study, a bicycle is modeled and statically stabilized using Control Moment Gyroscope as an actuator. Gyroscopes have been used for stability and navigation control of sea vessels and spacecrafts, however, it has rarely been applied in small ground based vehicles. To stabilize the bicycle, a so called polynomial approach has been adopted for designing the PD controller configuration with low-pass filter. The clear physical meaning of polynomial method's design parameters helps to reach a good compromise among different trade-offs during controller design. A stabilizing torque is produced by controlling the precession speed of the gyroscope relative to the body of bicycle.
Furthermore, an inverted pendulum based experimental setup has been used to emulate the dynamics of a rolling bicycle. The simulation and experimental results demonstrate the effectiveness of the polynomial method in designing low-order controllers for highly nonlinear systems.
