UG Projects 2014-15
Control-Based Projects
[HKL01] Fuzzy control of virtual ball-and-plate system
The aim of this project is to control the ball-and-plate system using fuzzy-model-based control method. A fuzzy model will be developed to facilitate the design of the fuzzy controller. Stability analysis will then be carried out. By solving numerically the solution to the stability conditions, a fuzzy controller can be obtained. To verify the analysis result, the control system will be implemented using a Lego Mindstorm NXT kit/Arduino microcontroller and a smart phone/tablet. Lego Mindstorm NXT kit/Arduino microcontroller will be used to implement the control platform. The smart phone/tablet will be used to implement the virtual ball-and-pate system, which is mounted on the Lego Mindstorm/Ardunio platform. The smart phone/tablet will show a ball on the screen. The information of the ball, e.g., position, velocity, etc, will be fed back to the Lego Mindstorm/Ardunio platform. By adjusting the angle of the platform (which controls the angle of the smart phone/tablet), the ball position is driven to the origin. Student taking this project is required to be hardworking and self-motivated. Good Matlab and smart phone programming skills are essential to this project.
[HKL02] Fuzzy control of virtual inverted pendulum system
Please refer to the description of [HKL01]. Instead of ball-and-plate system, an inverted pendulum system is used.
[HKL03] Fuzzy control of ball-and-beam system
This project aims at the implementation of a physical ball-and-beam system using Lego Mindstorm NXT kit or Arduino microcontroller. An image processing technique will be developed to locate the ball position in the beam or an infrared sensor is used instead. Various controllers will be employed to stabilize the ball position or drive the ball to keep track of a pre-defined trajectory. Good programming and hardware building skills are essential to the project.
Neural network-based projects
[HKL04] Voice command recogniser
This project will implement a calculator, which performs calculation according to voice commands. Speech recognition will be done by the following steps: 1) Student will collect speech data of himself/herself as the command input of calculator. 2) Feature extraction will be done for all collected speech data. 3) Neural networks/support vector machines/fuzzy logic/other classifiers will be employed to implement the voice command recogniser. The overall systems will be implemented in Matlab or other platforms such as IOS/Android platform. Recognition performance will be investigated and compared with existing methods. Student taking this project is required to be hardworking and self-motivated. Good Matlab programming and smart phone programming skills are essential to this project.
[HKL05] Gesture command recogniser
Instead of using voice, gestures are used to issue commands. For example, a web camera/smart phone/Kinect can be used to capture gestures. Project details is more or less the same as described in [HKL04].
[HKL06] Time-series predictor
This project is to create a predictor which is to forecast the time series using historictal data. One of an examples for time series is stock market prices. Taking stock markert price as an example, an application (e.g., mobile application) will collect historial data online (say from Yahoo Finance), implement and train the predictor, and perform prediction. Predictor will be constructed by neural networks/support vector machines/fuzzy logic/other traditional methods. Comparison and evaluation will be done by considering different sets of data. Good Matlab programming and smart phone programming skills are essential to this project.