Past Research Projects

Research Interests and Projects

Dynamic Modelling and Intelligent Control of the Single Screw Extrusion Process

The dynamics of the extrusion process often necessitates both planned and/or unplanned changes in the state of the process, brought about by variations in rheological properties of the polymer materials, or in the operating conditions including system disturbances. Most of the analysis is based on correlating experimental data by means of curve fitting. Whilst these models are useful in their respective applications, they lack generality and may not be applicable to screws of a different design.

Modelling - A theoretical model of the dynamic behaviour of the extrusion operation is developed in state-space format. Learning fuzzy systems is used to identify the functional relationships (correlations) between the chosen state variables.

Development of process control and optimisation strategy - Soft computing technique is used to achieve a refinement of the dynamic model by defining the process variables as fuzzy variables, reasonable approximations of their functional relationships could be obtained. GA technique is also applied to optimise the multivariable extrusion process control.

Intelligent Identification of Polymers in Recycling Process

Design, develop and manufacture of an ultrasonic monitoring system for the identification, characterisation and automated separation of reclaimed plastic materials from mixed waste streams.

The main objective of the research work was to investigate the application of soft computing in the analysis of monitoring data with a view to develop intelligent on-line control capability. >>> more >>>

Automatic Inspection and Control of Surface Mounth Technology

An automatic inspection and control methodology was developed to reduce the number of board failures using an adaptive fuzzy controller. The fuzzy controller consists of four principle components of; fuzzy feature extraction, fuzzy defects classification, fuzzy rule-based modelling and controller.

The Automated Solder Paste Monitoring Inspection and Control (ASPMIC) was a project funded by the Engineering and Physical Sciences Research Council (EPSRC). It established an automatic closed-loop control system for the stencil printing stage of the Surface Mount Technology (SMT). >>> more >>>

Learning and Adaptive Fuzzy Systems

The main advantage of a fuzzy inference system is its ability to utilise information expressed in linguistic form. Up until quite recently,the design of a fuzzy inference system is very much an art. The designer, considers the knowledge accumulated and crafts the membership function(s) and/or the inference mechanism in the system. Often, these membership functions are fixed once the design process has terminated. It would be desirable to have a fuzzy inference system whose membership function parameters can be ``adapted'' or ``learned'' from input and output data. >>> more >>>

On-line Data Acquisition and Grid Computing

Intelligent Condition Monitoring

Multimedia and E-commerce

FISMAT - Fuzzy Inference Systems toolbox for MATLAB

A freeware simulation toolbox using MATLAB was developed to support the fuzzy inference system design. >>> more >>