No description available.

Detta arbete har gjorts åt Scania som är ett av de ledande företagen inom lastbilstillverkning. Arbetet har utförts på plats I Södertälje på Scanias cylinderhuvudstillverkning. Området som arbetet behandlar är relativ nytt och därför upptäcks en del nya problem. Ett av dessa problem är att en befintlig 7 bitars buffert mellan två maskiner, OP70 och OP80, antingen står full eller tom, vilket leder till produktionsbortfall. Dessutom har OP80 många stopp vilket leder till att bufferten snabbt fylls upp och OP70 inte kan fortsätta arbeta. Därför ville Scania undersöka om en förändring på buffertstorlek skulle kunna lösa detta problem. För att kunna undersöka huruvida en större buffert skulle lösa problemet användes simulering som verktyg. I simuleringsprogrammet ExtendSim 10 byggdes två simuleringar upp. Den ena simuleringen som kallas “nuläge” är baserad på hur det såg ut då arbetet startades. Den andra kallas “målbild” och är uppbyggd med de cykeltider och stoppfrekvenser som Scania önskar att dessa maskiner har I framtiden. Under arbetets gång började man använda nya verktyg I OP80 maskinerna och det beslutades att dessa nya verktyg används i simuleringen av målbilden men inte I simuleringen av nuläget. När ett stort antal olika buffertstorlekar hade simulerats kunde grafer ställas upp som visar hur många bitar man får ut vid varje buffertstorlek samt hur stor ökning det motsvarar jämfört med den befintliga bufferten med 7 platser. Genom detta kunde även den optimala buffertstorleken för de två simuleringarna fastställas. Utifrån resultaten diskuteras tre olika handlingsalternativ, fördelar och nackdelar med dessa handlingsalternativ samt vilket av handlingsalternativen som de som utfört arbetet rekommenderar. / This project has been done for Scania, which is one of the leading companies in truck manufacturing. The work has been carried out on site in Södertälje at Scania’s cylinder head manufacturing. The production line the project has been carried out on is relatively new and has had a few problems during its start up. One of these problems is a buffer located between the machines OP70 and OP80 that, most of the time, is either full or empty. This leads to loss of production. That is why Scania wants to investigate whether a change in buffer size can solve this problem. To help with investigating whether a larger buffer would solve the problem simulation was used. In the simulation program ExtendSim 10 two simulation models were built. One model was named “nuläge” and is a reflection of how the production line looked at the start of the project. The other was named “målbild” and uses the cycle times and shut down frequencies Scania wishes to have in the future. During the course of the project Scania started using new tools in the OP80 machines and it was decided to use these new tools in the målbild simulation but not in the nuläge simulation. When all relevant buffer sizes had been simulated, graphs showing how many manufactured pieces you would get at different buffer sizes could be plotted. Graphs showing the percentage of which production was increased, compared to the existing buffer with 7 places, was also created in a similar way. Through these graphs the optimal buffer size for both simulations could be established. From these results three courses of action were discussed and compared by their pros and cons. Lastly a recommended course of action is chosen.

No / Traffic congestion in communication networks can dramatically deteriorate user-perceived Quality-of-Service (QoS). The integration of the Random Early Detection (RED) and priority scheduling mechanisms is a promising scheme for congestion control and provisioning of differentiated QoS required by multimedia applications. Although analytical modelling of RED congestion control has received significant research efforts, the performance models reported in the current literature were primarily restricted to the RED algorithm only without consideration of traffic scheduling scheme for QoS differentiation. Moreover, for analytical tractability, these models were developed under the simplified assumption that the traffic follows Short-Range-Dependent (SRD) arrival processes (e.g., Poisson or Markov processes), which are unable to capture the self-similar nature (i.e., scale-invariant burstiness) of multimedia traffic in modern communication networks. To fill these gaps, this paper presents a new analytical model of RED congestion control for prioritized multi-class self-similar traffic. The closed-form expressions for the loss probability of individual traffic classes are derived. The effectiveness and accuracy of the model are validated through extensive comparison between analytical and simulation results. To illustrate its application, the model is adopted as a cost-effective tool to investigate the optimal threshold configuration and minimize the required buffer space with congestion control.

Doctor of Philosophy / Department of Psychological Sciences / Richard Jackson Harris / In recent years, Baddeley (2010) has added a new component, the episodic buffer, to his Working Memory (WM) model. The episodic buffer binds information from long-term memory (LTM) to the central executive but has been researched very little, especially with respect to its use with a second language. In fact, Juffs and Harrington (2011) stated, “To date there has been no research on the possible role of the episodic buffer in L2 learning and use” (p. 140). One goal of this study was to do just that. Domain knowledge (DK) in baseball (Experiment 1) and English proficiency levels (Experiment 2) were used as proxies for difficulty level to study how inference processing under different conditions of domain knowledge and cognitive load in native (L1) and non-native (L2) English readers contribute to understanding the episodic buffer. In Experiment 1, 67 participants varying in domain knowledge about baseball read stories related to baseball with or without a concurrent cognitive load task of responding to tones while reading; they then answered comprehension questions of varying degrees of inferential difficulty. In Experiment 2, three groups varying in English reading proficiency, split into groups based on their lexical decision task scores (72 native, 40 intermediate, 40 beginner readers) read general stories with or without cognitive load and answered comprehension questions requiring varying degrees of inferential difficulty. Accuracy and Reaction Time (RT) were differentially affected by working memory (OSpan), cognitive load, and inferential complexity. In Experiment 1, greater DK explained variance in effectiveness (accuracy) and efficiency (RT) as inferential complexity increased. In Experiment 2 OSpan was needed even at lower levels of inferential complexity for beginning readers. Surprisingly, for both experiments, participants responded faster under cognitive load conditions, although not at the expense of accuracy. This suggests that the episodic buffer is important for different levels of DK and proficiency, especially as the task becomes more difficult. Theoretical and practical implications of these findings are discussed.

This thesis discusses and demonstrates the benefits of simulating and optimising a manufacturing control system in order to improve flow of production material through a system with high variety low volume output requirements. The need for and factors affecting synchronous flow are also discussed along with the consequences of poor flow and various solutions for overcoming it. A study into and comparison of various planning and control methodologies designed to promote flow of material through a manufacturing system was carried out to identify a suitable system to model. The research objectives are; • Identify the best system to model that will promote flow, • Identify the potential failure mechanisms within that system that exist and have not been yet resolved, • Produce a model that can fully resolve or reduce the probability of the identified failure mechanisms having an effect. This research led to an investigation into the main elements of a Drum-Buffer-Rope (DBR) environment in order to generate a comprehensive description of the requirements for DBR implementation and operation and attempt to improve the limitations that have been identified via the research literature. These requirements have been grouped into three areas, i.e.: a. plant layout and kanban controls, b. planning and control, and c. DBR infrastructure. A DBR model was developed combined with Genetic Algorithms with the aim of maximising the throughput level for an individual product mix. The results of the experiments have identified new knowledge on how DBR processes facilitate and impede material flow synchronisation within high variety/low volume manufacturing environments. The research results were limited to the assumptions made and constraints of the model, this research has highlighted that as such a model becomes more complex it also becomes more volatile and more difficult to control, leading to the conclusions that more research is required by extending the complexity of the model by adding more product mix and system variability to compare results with the results of this research. After which it will be expected that the model will be useful to enable a quick system response to large variations in product demand within the mixed model manufacturing industry.

No description available.

This thesis describes the design, construction, operation, and characterisation of an experimental apparatus that produces a source of internally cold, slow molecules that can be used for studying ion-molecule reactions at low temperatures. The apparatus combines buffer-gas cooling with a bent quadrupole velocity selector to cool both the translational and rotational degrees of freedom of the molecules. A cold cell (6 K) is filled with a buffer gas, such as helium, that exhibits sufficiently high vapour pressure for cryogenic applications. Hot molecules (150 to 300 K) enter the cell and thermalise with the buffer gas through collisions. Molecules are subsequently loaded into an electrostatic quadrupole guide, which acts as a velocity filter; only translationally cold polar molecules are guided around the bend. Using a buffer-gas-cooled source of molecules for electrostatic velocity selection, rather than a 300 K effusive source, yields a rotationally cold sample, with J ≤ 3. This rotational selectivity will enable the dependence of reaction cross sections on the reactant rotational state to be examined. Mass spectrometry is used to characterise cold molecular beams of ND3 and CH3F, while (2+1) REMPI spectra are recorded for the ammonia isotopologues. The peak velocity of guided ND3 is 75.86(0.70) ms-1 for standard conditions in a 6 K helium buffer gas cell (1.0 sccm ND3 flow rate, 0.6 mbar helium inlet pressure, ± 5 kV voltage). This corresponds to a peak kinetic energy of 6.92(0.13) K. (2+1) REMPI spectroscopy of the B1E''(v2'=5) ← X(1) transition enabled the rotational state distribution of guided ammonia molecules to be established. PGOPHER simulations of the experimental spectra suggest a rotational temperature of 10 K for ND3 molecules (from a 6 K helium buffer gas cell). The extent of translational and rotational cooling can be controlled by varying the molecular and buffer gas densities within the cell, by changing the temperature of the buffer gas cell (we can operate at 6 K or 17 K), or by changing the buffer gas. The translational temperature of guided ND3 is similar in a 6 K helium and 17 K neon buffer gas cell (peak kinetic energies of 6.92(0.13) K and 5.90(0.01) K, respectively) because the heavier neon gas has a slightly lower thermal velocity at 17 K than helium does at 6 K. Despite similar translational temperatures, the rotational temperature of guided ND3 is lower for molecules exiting the 6 K helium cell compared to the 17 K neon buffer gas cell (10 K and 15 K, respectively). The 6 K helium and 17 K neon buffer gas cells provide an excellent opportunity to investigate the effect of rotational cooling on branching ratios and reaction rates in low temperature ion-molecule reactions. The buffer gas cell and velocity guide described in this work will be combined with a linear Paul ion trap, to facilitate the study of cold ion-molecule reactions.

Buffer Overflows are a common type of network intrusion attack that continue to plague the networked community. Unfortunately, this type of attack is not well detected with current data mining algorithms. This research investigated the use of Random Forests, an ensemble technique that creates multiple decision trees, and then votes for the best tree. The research Investigated Random Forests' effectiveness in detecting buffer overflows compared to other data mining methods such as CART and Naïve Bayes. Random Forests was used for variable reduction, cost sensitive classification was applied, and each method's detection performance compared and reported along with the receive operator characteristics. The experiment was able to show that Random Forests outperformed CART and Naïve Bayes in classification performance. Using a technique to obtain Buffer Overflow most important variables, Random Forests was also able to improve upon its Buffer Overflow classification performance.

<p>Tanzania is famous for it’s beautiful nature and rich wildlife. Proud of it’s natural heritage, Tanzania has dedicated over 20% of it’s territory as protected areas to shield the wildlife from human interference. But the wildlife is regarded as a menace by the local communities that lives close to the impressive national parks. At the same time, the increasing human population threatens the survival of the large migratory species in the parks by blocking vital dispersal areas.</p><p>This essay describes the human/wildlife conflict around Tarangire National Park and focuses on communities close to park borders. The intention is to examine if a buffer zone could solve the conflict in this area. By studying the political ecology of wildlife conservation in Tanzania, from local to global scale and through a historical perspective, it is concluded that the poor state-society relation as experienced in local communities is a crucial factor for the diminishing wildlife.</p>

We introduce the tractable buffer stock savings setup of Carroll (2009 NBER Working Paper) into an otherwise conventional New-Keynesian dynamic stochastic general equilibrium model with financial frictions. The introduction of a precautionary saving motive arising from an uninsurable risk of permanent income loss, affects the model's properties in a number of interesting ways: it produces a more hump-shaped reaction of consumption in response to both supply (technology) and demand (monetary) shocks, and more pronounced reactions in response to demand shocks. Adoption of the buffer stock savings setup thus offers a more microfounded way, compared to, e.g., habit preferences in consumption, to introduce Keynesian features into the model, serving as a device to curbing excessive consumption smoothing, and to attributing a higher role to demand driven fluctuations. We also discuss steady state effects, determinacy properties as well as other practical issues. (authors' abstract) / Series: Department of Economics Working Paper Series