No description available.

Call number: LD2668 .T4 1962 W36

Call number: LD2668 .T4 1966 H489 / Master of Science

Wireless sensor actuator networks (WSANs) are becoming a solution for the implementation of control applications. Sensors and actuators can be deployed forming a large or dense network to monitor and control physical parameters or systems. However, this comes with challenges. Reliable data transmission and real-time communication constraints are the most significant challenges in WSANs for control applications because wireless networks are characterised by harsh transmission conditions. The use of WSANs for critical control applications has not gained sufficient progress as wireless networks are perceived to be totally unreliable and hence unsuitable. This makes reliable data transmission a priority in this research. Control applications will have a number of quality of service (QoS) requirements, such as requiring a very low packet-loss rate (PLR), minimum delay and guaranteed packet delivery. The overall goal of this research is to develop a framework that ensures reliable and real-time communication within the sensor network. A totally reliable network design involves ensuring reliability in areas such as the medium access control, connectivity, scalability, lifetime, clustering and routing with trade-offs such as energy consumption, system throughput and computational complexity. In this thesis, we introduce a unique method of improving reliability and real-time communication for control applications using a link quality routing mechanism which is tied into the ZigBee addressing scheme. ZigBee routing protocols do not consider link quality when making routing decisions. The results based on common network test conditions give a clear indication of the impact on network performance for various path loss models. The proposed link quality aware routing (LQAR) showed a highly significant 20.5% improvement in network delays against the ZigBee hierarchical tree routing (HTR) protocol. There is also a 17% improvement in the PLR. We also investigate variable sampling to mitigate the effects of delay in WSANs using a neural network delay predictor and observer based control system model. Our focus on variable sampling is to determine the appropriate neural network topology for delay prediction and the impact of additional neural network inputs such as PLR and throughput. The major contribution of this work is the use of typical obtainable delay series for training the neural network. Most studies have used random generated numbers which are not a correct representation of delays actually experienced in a wireless network. In addition, results show that the use of network packet loss information improves the prediction accuracy of delay. Our results show that adequate prediction of the time-delay series using the observer based variable sampling model influences the performance of the control system model under the assumptions and stated conditions.

This master thesis contains modeling and nonlinear automatic control of an autonomous lawn mower. The vehicle shall be able to follow a magnetic field loop buried in the lawn on an arbitrary distance, including straddling. A unicycle model of the lawn mower has been derived based on previous work. The model holds for the assumption of non slip environment. Two diff erent control strategies have been implemented and evaluated in this thesis. One is based on feedback linearization  and with the feedback gain estimated using an LQ-controller. The other strategy is based on linearization at an equilibrium point  and also with the use of an LQ-controller. The only state that can be measured is the distance to the magnetic loop. In order to estimate the other states, an observer has been implemented using the linearized in an equilibrium point  model and a Kalman filter. The navigation of the autonomous lawn mower is performed using a magnetic loop buried in the lawn. The non-linearity of the magnetic field causes diffi culties when estimating the distance to the loop. One magnetic field strength corresponds to two distances to the magnetic loop. The nature of the magnetic field and the measurement noise causes the precision of the controller to be limited. The conclusion of this master thesis is that it is possible to navigate and follow the magnetic loop at a distance greater than 0.3 [m], but with overshoots during sharp turns. If the reference distance is smaller than that, the behavior of the robot will be oscillating when following the loop. The limitations of the controller performance is due to the lack of information which can be extracted from the measurements of the magnetic field.

The purpose of this research was to develop three real-time adaptive temperature controllers for hyperthermia heating systems. Each scheme is made adaptive by using a transient Gaussian estimation routine to estimate the tissue blood perfusion and by then using these estimated values either in an optimizing routine, or in an observer, or in both. The optimizing routine uses a steady-state Gaussian estimation technique to optimize the power distribution until the best possible match is obtained between the steady-state temperatures predicted by a treatment model and a prespecified ideal temperature distribution. The observer uses a treatment model to control unmeasured locations. The first adaptive control scheme uses the optimizing routine alone, the second uses the observer alone and the third uses both the optimzing routine and observer. The performance of each of the adaptive control schemes is compared to a standard proportional-integral-derivative (PID) control scheme for one-dimensional simulations of typical treatments. Results comparing the deviation of the controlled temperature distribution to the ideal desired temperature distribution for all locations and all times indicate that the adaptive schemes perform better than the PID scheme. It can be concluded that adaptive control yields improved performance if good a priori knowledge of the treated region tissue and perfusion region boundaries is available. While these control schemes were designed for eventual implementation on a scanned focused ultrasound hyperthermia treatment system, the techniques are applicable to any system with the capability to vary specific power with respect to location and with an unknown distributed energy sink proportional to the temperature elevation.

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

Submitted in compliance with the requirements for the Master's Diploma in Technology: Electronic Engineering, Technikon Natal, 1991. / This dissertation is primarily concerned with investigating and improving the degree of accuracy and precision that may be achieved from temperat~re measurements made utilising nickel-chromium/nickel-aluminium (Type K) thermocouples. The practice of heat treating extruded aluminium section creates specific metallurgical properties within section. Development of specialised aluminium alloys has necessitated the use of treatment temperatures,- close to the limit beyond which the alloy experiences undesirable, permanent, metallurgical change. This situation has demanded urgent attention to, in quality assurance terms, the, 'fitness for purpose', of primary temperature sensors. The most established of these sensors, the Type thermocouple, has known problems relating to calibration stability and drift. The substantial amount of furnace control instrumentation and cabling dedicated to measurement from Type K sensors precludes the simple conversion to an alternate sensor type. The more practical option of applying calibration correction factors to existing measuring systems is only feasible if sensor stability characteristics permit measurement traceability to' be established within required uncertainty limits. / M

Call number: LD2668 .T4 1958 G74 / Master of Science

M.Ing. / This research project is dedicated to the automation of environmental control within greenhouses. To create an optimal climate in the greenhouse, the main environmental parameters that need to be controlled are temperature, humidity and light intensity. As a result of process dead times and the extreme interdependence of these parameters, the control problem can be classified as non-linear and multi-variable. In the past, most greenhouse environmental control systems depended on the decision making of an experienced human operator. This often gave rise to trial and error, especially when new species were established. With the current advances in "intelligent" control systems and high accuracy sensors, more and more of the decisions involved in greenhouse control can be automated. In this way more emphasis can be placed on emulating the abilities of an expert operator, by means of a computerbased automatic control system. In this research project, "intelligent" as well as "non-intelligent" control techniques, for addressing the problem of automated climate control in a greenhouse, are investigated. These include PID-control as a "non-intelligent" technique, and rule-based fuzzy logic control and self-learning fuzzy logic control as two "intelligent" control techniques. These techniques are all applied to experimental greenhouse which is equipped with management mechanisms, such as fans, heaters, sprinklers and lights. The results of the experiments are evaluated according to two performance parameters: the Control Performance Index (CPI) and the Mean Square Error (MSE). The three techniques are not only assessed for their efficiency, but also for their applicability to the greenhouse environmental problem. Each of the control techniques has a unique characteristic response to the non-linear, non-stationary, multi-variable problem of environmental control and are subsequently addressed in the respective chapter.