Membership functions of standard (type-1) fuzzy sets are usually represented by the Guassian function. The mean and standard deviation of the Guassian function are frequently determined by 'learning' algorithms. One example is the Adaptive Network Based Fuzzy Inferencing System (ANFIS), which has successfully been applied to a variety of applications. Recently, there is more and more interest in type-2 fuzzy sets because they present a better representation of the 'fuzziness'. Type-2 fuzzy logic systems also provide better modelling accuracy with a smaller rule base. In view of the advantages offered by type-2 fuzzy sets, this project seeks to extend the learning algorithms for adapting a type-1 fuzzy controller to its type-2 counterpart.

Back to Top

Various methodologies are applied to the measurement of pressure. Among them, the capacitive pressure sensor is one of the most useful devices due to its high performance-to-price ratio, reliability and low power consumption. A typical capacitive pressure sensor element consists of two conductive plates with a vacuum between them. The plates are insulated from each other and behave like an electrical capacitor. An external fluid or gas pressure deflects the upper plate so that the capacitance of the element varies with pressure. This change in capacitance can be detected by a suitable electrical circuit. However, due to the need to resolve small sense capacitances, typically between 50fF to 1pF, capacitance meters may not be able to measure the capacitance of a micro capacitive pressure sensor. The main objective of the project is then to design and implement an ultra-low noise sensor readout circuit for a micro capacitive pressure sensor.

There are various ways to implement the sensor readout circuit. One approach is to convert. capacitance directly to charge. The possible area of research would perhaps be to construct a capacitance readout circuit that employ a switched capacitance integrating charge amplifier as its main component. The amplifier is able to act as a charge integrator that compensates the sensor’s electrical charge with a charge of equal magnitude and opposite polarity, and ultimately produces a voltage across a range capacitor. In effect, the purpose of the charge amplifier is to convert the high impedance charge input, q, into a usable output voltage, V. Another approach may be to introduce piezoelectric elements as another transducing level. The piezoelectric effect is a phenomenon resulting from a coupling between the electric and mechanical properties of a material. When mechanical stress is applied to a piezoelectric material, an electric potential will be produced. Therefore, by introducing piezoelectric elements as another transducing level, values of the pressure applied can be determined from the voltage values.

By varying the pressure, different voltage values can be obtained. From these two sets of values, a relationship can be found between the two quantities. Thus, one can then measure the pressure being applied directly by obtaining the voltage from the readout circuit.

Back to Top

In the semiconductor manufacturing industry, wafers undergo numerous thermal processing steps e.g. soft bake, hard bake, post-exposure bake and rapid thermal processing. The desire to print smaller features on larger substrates has necessitated more stringent wafer temperature control. It is becoming increasingly difficult to meet the tighter specifications using open loop control methodology, which is the current practice in the semiconductor manufactuirng industry. Consequently, there is a need to develop in-situ wafer temperature metrology in order to provide a feedback signal for performing closed-loop control. A problem hindering the wide-spread use of in-situ temperature measurement systems is that the accuracy of the measurement is highly dependent on the level of thermal contact between sensor and wafer. Estimation algorithms provide a means of overcoming the influence of varying contact level on the accuracy of a temperature measurement system. This project seeks to develop an embedded system that achieves accurate in-situ temperature measurement by using the output of two sensors to estimate the actual temperature.

Back to Top

Fuzzy systems have demonstrated their ability to solve different kinds of problems in various application domains. Currently, there is an interest in augmenting fuzzy systems with learning and adaptation capabilities. Two of the most successful approaches to hybridise fuzzy systems with learning and adaptation methods have been made in the realm of soft computing. Neural fuzzy systems and genetic fuzzy systems hybridise the approximate reasoning method of fuzzy systems with the learning capabilities of neural networks and evolutionary algorithms. Despite these advances, research has shown that the original (Type-1) fuzzy sets are unable to model and minimise the effect of uncertainties even though its name has the connotation of uncertainty. Recently, Type-2 fuzzy sets and a concept known as the footprint of uncertainty have been introduced to overcome the shortcomings of classical Type-1 fuzzy sets. This project seeks to hybridise type-2 fuzzy systems with the learning capabilities of evolutionary algorithms. The objective is to investigate if type-2 fuzzy controllers evolved by genetic algorithms are more robust than its type-1 counterpart.

Back to Top

Most MEMS gyroscopes are vibratory rate gyroscopes. The main mechanical component is a two degree-of-freedom vibrating structure, which is capable of oscillating in two directions on a plane. When the gyroscope is subjected to an angular velocity, the Coriolis effect transfers energy from one vibrating mode to another. The response of the second vibrating model provides information about the applied angular velocity. Ideally, the vibrating modes remain mechanically uncoupled. In practice, however, fabrication imperfections and environment variations are always present, resulting in a frequency of oscillation mismatch between the two vibrating. These imperfections degrade the gyroscope's performance and cause a false output. The project aims at developing an fuzzy control scheme for estimating the angular rate and, at the same time, identifies and compensates quadrature error.

Back to Top