Master-Arbeit Univ. St. Gallen, 2005.

Prepared for the United States, Department of Energy, under contract no. EX-76-A-01-2295, task order 37, by Energy Laboratory, Massachusetts Institute of Technology.

No description available.

No description available.

Computer programs designed to converse with humans in natural language provide a framework against which to test supra-sentential theories of language production and interpretation. This thesis seeks to flesh out, in terms of a computer model, two basic assumptions concerning prosody-that speakers use intonation to convey intention, or attitude, and that prosodic prominence serves to convey conceptual prommence. A model of an information-providing agent in is proposed, based on an analysis of a corpus of spontaneous dialogues. This uses an architecture of communicating processes, which perform interpretation, application-specific planning, repair, and the production of output. Dialogue acts are then defined as feature bundles corresponding to significant events. A corpus of read dialogues is analysed in terms of these features, and using conventional intonational labelling. Correlations between the two are examined. Prosodic prominence is examined at three levels. At the level of surface encoding, re-use of substrings and structural parallelism can reduce processing for the speaker, and the listener. At the level of conceptual planning, similar benefits exist, given that speakers and listeners assume a common discourse model wherever possible. At these levels use is made of a short-term buffer of recent forms. A speaker may additionally use contrastive prominence to draw the listener's attention to disparities. Finally, at the level of intentions, a speaker wish to highlight certain information, regardless of accessibility. Prosodic focus is represented relationally, rather than via a simple binary-valued feature. This has the advantage of facilitating the mapping between levels; it also renders straightforward the notion of focus as the product of a number of potentially conflicting influences. Those parts of the theory concerned with discourse representation, language generation, and prosodic focus have been implemented as part of the Sundial dialogue system. In this system, discoursal and pragmatic decisions affecting prosody are converted to annotations on a text string, for realisation by a rule-based synthesizer.

As utilities become more interested in using renewable energy to power the grid, the problem becomes how to size and locate the generation facilities. This thesis approaches the idea of using distributed medium scale generation facilities at the distribution feeder level. We propose an algorithm to determine the optimum size of a photovoltaic(PV) array and an energy storage system for a distribution feeder. The cost of operating a feeder is quantified by considering the net load at the substation, voltage changes, load following, and the initial cost of implementing a photovoltaic system and a battery energy storage system. The PV inverter is utilized in order to improve the voltage on the circuit and is sized proportionally to the array size. The energy storage system operates in peak shaving and load following capacities in order to reduce stress on current generation facilities. The algorithm then operates to minimize the total cost of the feeder operation for a year by sizing these distributed generation resources utilizing particle swarm optimization. Optimization of a real-world system yielding results where the power at the substation (including all losses) is reduced by 5.39% over the course of a year and the average voltage drop on the circuit is improved by 50.17% using the proposed photovoltaic inverter control scheme.

Compilation encompasses many steps. Parsing turns the input program into a more manageable syntax tree. Verification ensures that the program makes some semblance of sense. Finally, code generation transforms the internal abstract program representation into an executable program. Compilers strive to produce the best possible programs. Optimizations are applied at nearly every level of compilation. Instruction Scheduling is one of the last compilation tasks. It is part of code generation. Instruction Scheduling replaces the internal graph representation of the program with an instruction sequence. The scheduler should produce some sequence that the hardware can execute quickly. Considering that Instruction Scheduling is an NP-Complete optimization problem, it is interesting that schedules are usually generated by a greedy, heuristic algorithm called List Scheduling. Given search-based algorithms' successes in other NP-Complete optimization domains, we ask whether search-based algorithms can be applied to Instruction Scheduling to generate superior schedules without unacceptably increasing compilation time. To answer this question, we formulate a problem description that captures practical scheduling constraints. We show that this problem is NP-Complete given modest requirements on the actual hardware. We adapt three different search algorithms to Instruction Scheduling in order to show that search is an effective Instruction Scheduling technique. The schedules generated by our algorithms are generally shorter than those generated by List Scheduling. Search-based scheduling does take more time, but the increases are acceptable for some compilation domains.

In the past, the emission of SO[2], NO[x] and CO[2] gasses, as a result of the combustion of fossil fuels within the power industry has been a major contributor to the pollution of the environment. Controls over the last few years have been implemented to combat the effects of air pollution on the environment. One such control is the reduction of sulphur dioxide (SO[2]) from power stations, which is achieved by fitting desulphurisation systems to new and existing installations. At present there are many types of desulphurisation processes available, which produce a variety of wastes that vary in quality and quantity. The majority of processes fit into three main categories, wet, semi dry and dry desulphurisation processes. Desulphurisation systems work by introducing alkaline sorbents, such as limestone, to the SO[2] gasses, either during or after combustion. The limestone reacts with the SO[2] gasses to form new insoluble materials such as gypsum (CaSO[4].2H[2]O). However, semi-dry and dry processes produce wastes that are usually a blend of fly ash and some form of calcium sulphate. The wide variation in chemical, physical and mineralogical properties of the wastes produced hinders their utilisation, because general experimental investigations are only relevant to specific types of wastes. In addition, the wastes can contain large quantities of sulphates, which are normally limited to a few percent in plain cements to regulate setting, and therefore, their initial potential appears limited. The current investigation aims to evaluate the performance of paste, mortar and concrete that contain large quantities of actual and simulated desulphurised wastes. Members of the Copernicus project (1999) provided various desulphurised wastes from Eastern European installations for investigation. Fly ash and gypsum from UK sources were blended in different proportions to simulate desulphurised wastes based on a range of SO[3] contents typically associated with actual desulphurised wastes produced throughout Eastern Europe. Several tests were conducted on paste, mortar and concrete containing actual and simulated desulphurised wastes to investigate basic physical, mechanical and durability properties. These include chemical shrinkage, porosity and pore size distribution, water absorption, length change, compressive strength, and sulphate resistance. The outcomes of investigation were positive in that cement based materials containing high levels of desulphurised wastes were produced, which exhibit physical, mechanical, and durability properties equal or superior to reference cements, and cements containing more common replacement materials such as fly ash and slag. Attempts were made to correlate specific properties such as chemical composition, porosity and pore size distribution, and chemical shrinkage with strength in order to determine the key factors influencing strength development of cement-based materials containing desulphurised wastes. The relationship between SO[3] content and compressive strength of pastes containing simulated desulphurised wastes (FA-G blends) was used to predict the theoretical strength of cement-based materials containing actual desulphurised wastes with reasonable accuracy. Thus reinforcing the decision to evaluate simulated desulphurised wastes due to the diverse nature of desulphurised wastes currently available across Eastern Europe.

This study has investigated the use of exergy analysis when applied to the UK steel sector. The aim has been twofold: to assess whether exergy analysis provides greater insight into physical processes at this level than conventional thermodynamic analyses; and whether it can be used to identify ways that this sector has reduced its impact on the environment in the past, and how this might continue in the future. An overview of the issues surrounding resource consumption and waste generation and their relationship to the steel sector are given [Chapter 1]. The theory behind exergy analysis is then reviewed and its relevance to reducing resource use and waste minimisation examined [Chapter 2]. An exergy analysis was performed on all the processes within the steel sector [Chapters 3 & 4] and the reasons for thermodynamic inefficiencies discussed. Based on these analyses, exergy analysis was found to be superior to other methods in describing some, but not all, processes when compared with two forms of energy analysis [Chapter 5]. This is due to fact that exergy includes an account of entropy which quantifies thermodynamic quality. It was shown that reducing the exergy consumption of the steel sector leads to a reduction in resource use and to some extent will lead to a reduction in the emission of pollutants. Two different systems which could represent the UK steel sector were investigated to determine which best fitted the aims of the study [Chapter 6]. The first system was based on all processes within the UK which were linked to the use of steel; the second was based on all processes, regardless of location, which were linked to the use of steel in the UK. It was found that for the steel sector at least, there was little difference in the exergy consumption of each system, although these were structurally quite different. Finally a system was chosen to represent the UK steel sector which best suited the needs of the study. This system was used to examine the evolution of the steel sector from 1954-1994, detailing all the main steel and scrap flows, the exergy consumption in; mining, transport, steel production and scrap processing, the effects of trade and the role of manufacture and product use in generating scrap [Chapter 7]. It was found that the exergy consumption of the UK steel sector has fallen twofold in this time period, indicating a large reduction in resource consumption. Based on past evolution, scenarios for possible futures of the steel sector were constructed [Chapter 8]. These showed that from 1994 to 2019 substantial reductions in the exergy consumption of the sector are possible (between 15% and 72% of 1994 levels by 2019) through changes to: the overall demand for steel goods, the technology of steel production and the mix of steel production methods.

Two basic types of photovoltaic (PV) array simulator have been designed and tested. The first involves the use of a pilot panel and variable light source. It is implemented with analogue circuits. A stability analysis based on Popov's method is presented for this simulator with resistance-inductance (R-L) loads. In the second, characteristic array curves are stored in the memory of a microprocessor-based simulator. The design of both simulators is based on the transfer function method. By using the computing facility available, a stability study for the Type I simulator and some dynamic simulations are carried out. Both simulators are capable of driving a special load, namely, an experimental solar pumping system. The experimental results for both types of' simulator are satisfactory in terms of steady state precision and dynamic behaviour when used with this load. Compared with previously-reported PV array simulator designs [6,7,8,9,18], the two simulators described here have the following distinctive features: 1. A new method of sample curve generation for the Type II simulator results in relatively short sampling period and small memory size. 2. The sample curves of the type II simulator are based directly on the real PV array to be simulated. They are more accurate than the sample curves in references [6,7,9]. 3. Different loads (R, R-L and an experimental solar pumping system) have been considered in the design and have been tested in laboratory. 4. A stability analysis and some dynamic simulations are presented for the type I simulator. An analysis of this type has not been reported in previous studies [6,7,8,9,18]. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate