Locomotion is a key characteristic of almost all forms of life and is often accomplished, whether on land, in water, or in the air, by reciprocal motion of two or more appendages. Among the zooplankton, many species propel themselves by rhythmically beating multiple pairs of closely spaced leg-like appendages in a back-to-front (metachronal) pattern. The focus of this study is to understand the mechanical design, kinematic operation, and hydrodynamic result of metachrony in the zooplankton. In the first part of this study, Antarctic krill (Euphausia superba) are investigated as an ecologically important model species that metachronally beats its swimming legs (pleopods) to perform drag-based propulsion. Based on high speed videos of freely swimming Antarctic krill, hovering, fast forward swimming, and upside down swimming are identified as three distinct swimming modes with significantly different stroke amplitudes and beat frequencies. When transitioning between hovering and fast forward swimming, Antarctic krill first increase beat amplitude and secondarily increase beat frequency. In considering the design components that contribute to metachrony being a successful swimming technique, a comparison among many different species shows that the ratio between the appendage separation distance and appendage length is limited to a narrow range of values (i.e. 0.2 - 0.65). In the second part of this study, metachrony is examined at smaller length and time scales by examining the impulsive escape jump of a calanoid copepod (Calanus finmarchicus). The wake generated by the copepod's metachronally beating swimming legs is experimentally measured using a novel (and newly developed) tomographic particle image velocimetry (PIV) system capable of making volumetric 3D velocity measurements with high temporal and spatial resolution using IR illumination. The flow generated by the escaping copepod consisted of a stronger posterior vortex ring generated by the metachronally stroking swimming legs and a weaker one generated anteriorly around the body by the impulsive start of the escape, both of which decayed over time. The experiments also revealed azimuthal asymmetry in the vortices caused by body yawing and the action of the swimming legs, flow features not considered in previous axisymmetric computational and theoretical models of copepod jumps. While not accounting for this asymmetry, an impulsive stresslet is nonetheless useful in modeling the flow created by the escaping copepod and represents the flow more accurately than an impulsive Stokeslet. In the final part of this study, the flow associated with metachronal hovering in Antarctic krill is experimentally and theoretically investigated in regards to the energy requirements of the pelagic lifestyle. Volumetric flow measurements of a hovering Antarctic krill show that each stroking pleopod drags flow behind it such that a downward stream develops medially. The lateral exopodites induce tip vortices which add to the lift force on each appendage. Furthermore, the flow beneath the hovering krill develops into a pulsed jet with a Strouhal number in the 'high-efficiency zone' of 0.2 < St < 0.4. Actuator disk theory is used to make theoretical estimates of the induced power necessary to hover, the results of which match induced power values calculated from measured flow gradients contributing to viscous energy dissipation.

ABSTRACT Micronekton and macrozooplankton were sampled from the Western Antarctic Peninsula (WAP) and eastern Ross Sea regions. Samples were collected over the course of six research cruises to the Southern Ocean. Four of those cruises were conducted in the Marguerite Bay region of the WAP during the austral fall and winter of 2001 and 2002. A fifth cruise sampled faunal assemblages at nine sites, ranging from Joinville Island at the northern tip of the WAP to Charcot Island near the southern extent of the WAP, during austral fall 2010. A sixth cruise was conducted in the pack ice within the offshore, continental slope, and continental shelf zones of the Eastern Ross Sea during austral summer 1999/2000. The purpose of this dissertation was to describe the macrozooplanktonic/ micronektonic faunal assemblages present in each of those regions and relate differences in species composition and distribution patterns to local bathymetry, hydrography, and physical conditions. A variety of multivariate techniques were used to identify unique multispecies assemblages and to quantify the contributions of both oceanic and neritic species to the assemblages within each study region. The invertebrate micronekton/macrozooplankton communities found in the Marguerite Bay region of the WAP during the 2001/2002 cruises were a mixture of oceanic and neritic fauna: a direct result of local hydrographic conditions. Near the shelf break and in the outer reaches of the Marguerite Trough, a deep canyon transecting the shelf in a south-southeast direction, the communities were more diverse, dominated by oceanic species such as the euphausiid, Euphausia triacantha, the salp, Salpa thompsoni, and, an amphipod, Themisto gaudichaudii. The assemblages present in the nearshore fjords exhibited lower diversity and were dominated by neritic species such as the euphausiid, E. crystallorophias, and the mysid, Antarctomysis ohlinii. At the mid-shelf and mid-trough locations, the assemblages were composed of a variable mixture of oceanic and neritic fauna. The faunal mixing and overall species composition in those areas is the result of episodic Circumpolar Deep Water (CDW) intrusions onto the shelf via deep bathymetric features such as the Marguerite Trough. Distinct multispecies assemblages were identified at Joinville, Croker Passage, Marguerite Bay, Charcot Island, and from a region that included samples from sites near Anvers Island, Renaud Island and the Marguerite Trough. Assemblages at Joinville and Croker Passage were both dominated by E. superba and S. thompsoni, but hydrographic conditions at Joinville favored a neritic assemblage, underscored by the substantial numbers of the nototheniid fish, Pleuragramma antarcticum. The assemblage at Croker Passage was more oceanic in nature with major inputs from the myctophid fish, Electrona antarctica and the amphipod, T. gaudichaudii. Marguerite Bay and Charcot Island were well-mixed assemblages with strong representation by both neritic and oceanic fauna. The mid-peninsula assemblage was also oceanic in character, being overwhelmingly dominated by the euphausiid, Thysanoessa macrura, and T. gaudichaudii. Pleuragramma antarcticum were captured at five sites: Joinville, Croker Passage, Marguerite Bay, and the two sites near Charcot Island. They were completely absent at the two sites near Anvers Island, at Renaud Island, and in the Marguerite Trough. In the eastern Ross Sea, cluster analyses identified three primary groups, which were characterized as oceanic, mixed, and neritic assemblages due partly to their geographical location, but mostly to their faunal composition. The oceanic assemblage contained the highest number of taxa and was dominated by oceanic fauna, such as the hydrozoans Diphyes antarctica and Calycopsis borchgrevinki, and the scyphozoan Atolla wyvillei. Top contributors in the mixed assemblage included those species that dominated in the oceanic assemblage as well as substantial contributions from E. superba and the tunicate, Ihlea racovitzai. The neritic assemblage was overwhelmingly dominated by E. crystallorophias and E. superba. The physical environment strongly impacted micronektonic/ macrozooplanktonic distributions and densities in the eastern Ross Sea. Changes in faunal composition were directly related to temperature differences encountered at the shelf break and the subsequent filtering out of oceanic fauna from cold, shelf waters where endemic fauna were most prevalent.

Thesis (M.S.)--West Virginia University, 2009. / Title from document title page. Document formatted into pages; contains vii, 42 p. : ill. Includes abstract. Includes bibliographical references (p. 29-35).

For decades, humans have impacted marine ecosystems in a variety of ways including contamination by pollution, fishing, and physical destruction of habitats. Global change has, and will, lead to alterations in in a number of abiotic factors of our ocean in particular reduced oxygen saturation, salinity changes, elevated temperature (ocean warming or OW) and elevated carbon dioxide (ocean acidification or OA). Now and in the future, OA and OW will operate together with local anthropogenic drivers such as oil pollution. And yet, at present, very little is known about their potential combined interactive effects on physiological performance and tolerance of marine organisms. Therefore, multiple driver experiments are required if we are to understand and predict future vulnerability of species, populations and ecosystems. Early life stages of invertebrates are generally considered most vulnerable to environmental stress. However, few studies consider the combined effects OA and OW on survival and growth during early development of marine invertebrates, and to our knowledge, there is no information on the additional effects of oil pollution. Therefore, the aim of this thesis was to investigate the effects of combined exposure to OA, OW, and incorporating local drivers such as oil pollution on the development, morphology and physiology of three economically and ecologically important marine invertebrates. These are Northern shrimp Pandalus borealis, Northern krill Meganyctiphanes norvegica, and the green sea urchin Strongylocentrotus droebachiensis. All are cold-water species, assumed to have a narrower tolerance than more temperate species, and so could be particular sensitive to combined stressor affects. Both Northern krill and to a lesser extent Northern shrimp larvae survived experimental conditions, mirroring those predicted under a future global change scenario (combined OA and OW exposure). Neither was hatching success affected. Both shrimp and krill larvae exhibited accelerated developmental rates and incurred greater maintenance costs as a result of exposure to these stressors. Shrimp larvae showed accelerated developmental rates (-9 days), increased metabolic rates (+20 %), and increased feeding rates (+20 %), but reduced growth (- 9 %) when exposed to OW compared with the control. OA increased development rate but only at the control temperature. Although juvenile mortality of krill was not affected by predicted OA/OW conditions, metabolic rate increased significantly (+ 36 %), as did larval developmental rate, while number of moults, feeding rate and growth (- 67 %) decreased significantly (- 67 %, - 60 % and -8 % respectively). Accelerated development was accompanied by greater maintenance costs possibly due to experience a mismatch between energy supply and demand. Both species had an excess of food, and so growth reduction was more likely to be associated with higher metabolic demands in the future global change treatments. Food shortage in situ, due to variable food availability in the sea and/or mismatch with key prey species (algae and zooplankton) could result in more negative effects on growth and ultimately survival. Green sea urchins were also able to survive OA exposure, without detectable effects on hatching success. However, at day 44 post-fertilization, larval body length in the OA treatment was 9 % lower compared to the control. Furthermore, there was a significant tendency of urchin larvae to increase swimming activity in the OA conditions that might indicate compensatory feeding. Elevated maintenance and repair costs as a result of exposure to multi-stressors affected the energy budget of all the three species studied here resulting in reduced growth. Global drivers (OA and OW) resulted in trade-offs with more energy reallocated to swimming activity and metabolism, rather than growth. Exposure to oil reduced the acquisition of energy by reduced feeding which in turn resulted in less energy being available for growth. Both shrimp and sea urchin larvae showed reduced activity and feeding when exposed to oil. It is possible that the reduced swimming activity observed may be due to a narcotic effect of the oil. Furthermore, early stage sea urchin larvae showed increased mortality when exposed to oil while the older larvae did not, indicating a stage specific toxicity to oil for sea urchin larvae. The combination of global drivers and oil pollution acted additively on growth for both sea urchin and shrimp larvae. The impact of combined drivers on the size of shrimp larvae was equal the sum of the negative impacts observed for each driver: a 5 % reduction when exposed to OA and OW, a 9 % reduction when exposed to oil, and a cumulative 15 % reduction when exposed to all stressors. Similarly, the impact of combined drivers on the size of sea urchin larvae was equal to the sum of the negative impacts observed for each driver: a 14 % reduction when exposed to OA, a 9 % reduction when exposed to oil, and a 21 % reduction when exposed to all drivers. Therefore, the study demonstrated the additive physiological effects of OA, OW and a contaminant, and indicated that larval (sea urchin and shrimp) resilience to future changes (i.e. pollution) could be greatly reduced if larvae were already energy limited and severely stressed (reduced development) as a result of exposure to the global drivers. This study therefore shows the importance that the effective management of local drivers such as oil pollution could have against the backdrop of OA and OW, and emphasises that it is important to study impacts of toxicants, such as an oil pollution, in the context of predicted changes in the environment, as OW and OA are becoming major concerns. Finally, the fact that some local and global drivers seem to act additively should encourage local managers to act on local driver regulations, to obtain positive effects on local populations and environment and thereby rendering them more resilient to the negative impacts of future global drivers.

The understanding of what properties the fibre should have in order to give the right end-product paper properties, along with the type of processing, is a subject for a lot of research and development. Today the ability to measure fibre properties on-line is widely used for pulps. It is often necessary to measure many properties and variables in a process. The data collected is therefore almost always multivariate. It is very hard to analyse process data due to a lot of noises. Correlations between fibre and paper properties are hard to find, but this does not mean that correlations do not exist. Fibre properties, measured by the pulp analyser PulpEye, were investigated and correlations to paper properties were studied. The work was divided into three different studies. Study 1 was an investigation of historical process data, in study 2 pulp samples from the production was analysed and study 3 was a refiner setting trial, were different refiner segments, flows through refiners and intensities were studied. Both the group-plots and MVDA’s based on the historical process data in study 1, showed that the Scott Bond was increased with increased amount of kinks and curl for the unbleached pine pulp (softwood pulp). Coarseness measurements, made in the study of historical data, indicated that the coarseness was varying in such a large extent that it was believable that it had effects in the papermaking process. Another interesting fibre property, investigated in the refiner setting trial, was crill. The amount of crill is said to have strong correlation to paper strength. The analysis showed that the incoming pulp had different amount of crill and that the amount of crill after the refiners also was varying for the different samples. The development of crill at different kappa numbers and for pulps refined with different segments and refiner strategies should be further investigated. In this work it has been difficult to find correlations between fibre properties and paper properties in the refiner setting trial. This could have been due to small variations of the different parameters. This work showed that the normal production can be handled very well and variations are rather small. It can be seen though, that problems do appear when parameters are deviating from the normal case. An efficient way to work is to do measurements when the incoming pulp parameters are deviating. It should also be more investigated how the most common deviating pulp parameters should be handled in the refining process and at the board machine. The communication between the pulp production and the board machines is recommended to be further developed, especially when the pulp production have disturbances that can be affecting the refining and further the board production.

京都大学 / 0048 / 新制・論文博士 / 博士(農学) / 乙第13066号 / 論農博第2841号 / 新制||農||1046(附属図書館) / 学位論文||H28||N5020(農学部図書室) / (主査)教授 安達 修二, 教授 谷 史人, 教授 保川 清 / 学位規則第4条第2項該当 / Doctor of Agricultural Science / Kyoto University / DFAM

Le krill est un crustacé macrozooplanctonique qui joue un rôle clé dans l'écosystème marin pélagique. Il représente une plateforme trophique entre les producteurs primaires planctoniques et les niveaux supérieurs de la chaine alimentaire. Dans un contexte de changements environnementaux, la compréhension des impacts de la variabilité naturelle du forçage physique sur la physiologie du krill et la dynamique des populations apparait essentielle pour mieux appréhender la dynamique, l’évolution et la gestion des écosystèmes marins subarctiques. Dans l'estuaire et le golfe du Saint-Laurent (EGSL), les communautés de krill sont dominées par deux espèces sympatriques, Meganyctiphanes norvegica et Thysanoessa raschii. L'objectif principal de ce projet de recherche était d'étudier, à travers le développement de modèles biophysiques, les effets de la variabilité environnementale sur la distribution, la reproduction et la croissance de ces deux espèces. Le projet nous a permis i) de construire un modèle physiologique spécifique à l'espèce (IBM) pour le stade adulte des deux espèces de krill et de le valider avec les données disponibles ; ii) d'identifier spatiotemporellement les zones de croissance et de reproduction potentielles de l'EGSL pour les deux espèces iii) d'étudier le rôle de la variabilité interindividuelle sur la dynamique de production de T. raschii face à la variabilité environnementale. Dans le premier chapitre, nous avons construit un modèle physiologique basé sur l'individu et spécifique à chaque espèce représentant de manière réaliste le cycle de vie annuel des deux espèces à la station Rimouski située dans l'estuaire du Saint-Laurent. Nous avons montré que la dynamique de production des individus de M. norvegica et de T. raschii était fortement liée à l'environnement alimentaire. Cependant, les deux espèces ont montré des trajectoires de croissance et de reproduction bien différentes, T. raschii montrant une saison de production estivale, alors que M. norvegica montrait une saison de production plutôt automnale. Le modèle a permis de mettre en évidence comment la compréhension et la mise en oeuvre des processus d'alimentation étaient essentielles pour une représentation précise de la dynamique de croissance de chacune des espèces. Dans le deuxième chapitre, nous avons couplé le modèle physiologique développé au chapitre 2 à un modèle de circulation générale 3D et à un modèle biogéochimique de type NPZD, afin d'étudier la dynamique de production à l'échelle de l'EGSL dans son ensemble. Les résultats ont montré une forte saisonnalité du potentiel de croissance et de reproduction des deux espèces, étroitement liée à la dynamique du phytoplancton et du zooplancton, représentant leur champ de proies. Bien que les connaissances soient limitées au nord-est du golfe, les résultats de notre modèle ont montré l'émergence de zones de forte production, dans l'estuaire, dans la région du détroit de Belle Isle, le long de la côte ouest de Terre-Neuve et le long de la côte nord. Notre étude a également mis en évidence l'existence d'un circuit cyclonique de production du krill longeant les rives de l'ensemble de l'EGSL, de la péninsule d'Avalon, au sud de Terre-Neuve, à la rive nord du Québec en traversant le détroit de Belle Isle, durant la saison de production (avril à octobre). Dans le troisième chapitre, nous nous sommes intéressés à quantifier la contribution relative de la variabilité environnementale interannuelle (2003 et 2006) et celle de la variabilité interindividuelle sur la dynamique de production de T. raschii dans le système EGSL. Nos résultats ont montré qu'environ 15% de la variabilité totale des variables d'état du modèle IBM pouvait être expliquée par la variabilité interindividuelle simulée ajoutée au modèle IBM. Les résultats montrent également que l'inclusion de la variabilité interindividuelle n'a pas modifié significativement la réponse de la population en réponse à l’environnement : les mesures de centralité des variables d'état obtenues entre 2003 et 2006 sont restées similaires dans les simulations avec et sans variabilité interindividuelle. Cependant, il en est résulté un élargissement des distributions de fréquence des variables d'état. De plus, une augmentation des réponses positives des variables d'états lors des simulations avec variabilité interindividuelle a également émergé des résultats du modèle pour l'année 2003, évoquant une résistance de la population face à la variabilité de l'environnement. Les résultats de cette thèse fournissent des données importantes pouvant être intégrées dans un plan de gestion écosystémique et constituent une base pour l'étude de la variabilité interannuelle et la compréhension de l'évolution future de la dynamique de production pour les deux espèces. / Krill are macrozooplanktonic crustaceans play a key role in the pelagic marine ecosystem. They represent a crucial trophic platform between planktonic primary producers and the upper food-web levels. In a context of environmental changes, understanding the impacts of the natural variability of physical forcing on krill physiology and population dynamics appears essential to better apprehend the dynamics, evolution and management of subarctic marine ecosystems. In the Estuary and Gulf of St. Lawrence (EGSL), krill communities are dominated by two sympatric species, names Meganyctiphanes norvegica and Thysanoessa raschii. The main objective of this research project was to study, through the development of biophysical models, the effects of environmental variability on the distribution, reproduction and growth of these two species. The project allowed us i) to build a species-specific physiological based model (IBM) for adult stage of both krill species and to validate it with available data; ii) to identify spatio-temporally the potential growth and reproduction areas in the EGSL for both species iii) to study the role of the intraspecific variability on T. raschii population dynamics facing the environmental variability. In the first chapter we built a physiological species-specific based model representing realistically the annual adult life cycle for both dominant species at the Rimouski station located in the St. Lawrence River Estuary. We showed that the production dynamics of M. norvegica and T. raschii individuals were strongly linked to the feeding environment. However, both species exhibited well different growth and reproductive trajectories, T. raschii showing a summer production season, while M. norvegica an autumnal production season. The model highlighted how understanding and implementing feeding processes was essential for accurate representation of the growth dynamics of each species. In the second chapter, we coupled the physiological model developed in Chapter 2 with a 3D general circulation model and a biogeochemical model of the NPZD type, in order to study the production dynamics at the EGSL scale. The results showed a strong seasonality of the growth and reproduction potential of both species, closely related to the dynamics of phytoplankton and zooplankton, representing their prey field. Although knowledge was limited in the northeastern Gulf, model results showed the emergence of high production areas in the Estuary, in the Strait of Belle Isle, along the western coast of the Newfoundland and along the North Shore. Our study also highlighted the existence of a larger cyclonic circuit of krill production and transport that runs along the shores of the whole EGSL, from the Avalon Peninsula in the south of Newfoundland to the North Shore of Québec across the Strait of Belle Isle, and back to the Estuary upstream during the production season (April to October). In the third chapter, we are interested in quantifying the relative contribution of interannual environmental variability (2003 and 2006) and inter-individual variability on the production dynamics of T. raschii in the EGSL system. Our results showed that about 15% of the variability of IBM model state variables could be explained by the level of simulated inter-individual variability. Results also showed that inclusion of inter-individual variability did not significantly modify the system's response at the population level in response to the environment: the centrality measures of the state variables obtained between 2003 and 2006 remained similar in the simulations with and without inter-individual variability. However, this has resulted in an expansion of the frequency distributions of the state variables. Moreover, an increase in the positive responses of the state variables during simulations with inter-individual variability also emerged from the results of the model for the year 2003, evoking a resistance of the population to the variability of the environment. These results provide important data to integrate in fully ecosystem-based management plan. They represent a baseline to the study of the inter-annual variability and the understanding of the future production dynamics evolution for the both species.

The North Pacific humpback whale (Megaptera novaeangliae) population has been increasing at an average annual rate of ~6% since the early 1990s. In northern Southeast Alaska alone, there are now more whales than estimated for the entire North Pacific several decades ago. An understanding of how this growing population is repopulating traditional foraging grounds will benefit from detailed investigations of their prey preferences and trends in whale abundance and distribution relative to those prey. This dissertation examines these issues from late May until early September 2008 in Frederick Sound and Stephens Passage, a Southeast Alaskan feeding area historically used by humpback whales. The foundation for the study is an analysis of the life histories and abundance patterns of euphausiids, the principal prey of humpbacks in the area, during late spring and summer. Four species, Thysanoessa raschii, T. longipes, T. spinifera, and Euphausia pacifica, were identified in plankton net samples collected at random locations throughout the study site (n = 49) and in locations where a strong scattering layer was observed on a 120 kHz echosounder (n = 48). Both sample types varied in euphausiid species composition. Abundance patterns of immature euphausiids coupled with observations of females carrying spermatophores indicated differences between species in spawning schedules. Thysanoessa spp. began spawning in early April with the spring phytoplankton bloom and continued until late June, whereas E. pacifica began spawning in early June and continued until late August. This protracted recruitment of immature euphausiids was geographically widespread throughout the summer in contrast to adults, which, although present all summer, were found primarily in slope and shallow (< 100 m) areas. To determine if humpback whales preferred one euphausiid species or life-stage over another, net sample and hydroacoustic data collected in the vicinity of whales were compared to similar data collected in random locations throughout the study site. This revealed that whales targeted dense aggregations of adult euphausiids, but did not discriminate between the various species, which was surprising because of presumed differences in the energy density linked to their different spawning schedules. Additionally, whales did not spend time in areas with concentrations of immature euphausiids, which were likely not large enough during the study period to be suitable prey. With this preference for adult euphausiids, the abundance and distribution patterns of humpbacks were examined in relation to prey availability. Whale abundance was lowest at the beginning of the study in late May at ca. 68 whales and peaked in late July at ca. 228 animals – approximately 12% of the region’s estimated abundance for the study year. This study did not detect a concomitant increase in the availability of adult euphausiids, which is unsurprising since immature euphausiids would not recruit into the adult population until after the end of the study, and post-spawning mortality and predation pressure is presumably high during this time. Instead, whales clustered increasingly around comparatively fewer prey as the summer progressed. These observations, combined with a plateau in whale abundance after July, suggest that their abundance in the area was limited by euphausiid availability. Estimates of whales using the study site during the summer have remained similar over several decades despite a dramatic increase in humpback numbers in Southeast Alaska and elsewhere in the North Pacific. The results from this study suggest that, although the study site remains important seasonally to some whales, it is not a significant source of prey responsible for regional population growth in general. More likely, it is part of a network of feeding areas that has influenced the population trend. Further insight into these and the other issues raised in this dissertation could come from several additional analyses. An extended sampling season that captures the recruitment of immature euphausiids into the adult population would reveal whether a given year's prey cohort represents an important resource to whales in that same year, which has potential implications for interpreting mid-late season whale abundance patterns. As well, a photo-identification study would be useful in characterizing whale residency patterns and determining whether the abundance trends reflect a relatively small subset of the regional population using the area for most of the season or a continuous flow of a larger portion of the population. Finally, similar analyses as those outlined here but conducted in other areas within the region would provide additional insight into the network’s capacity to support the recovering whale population. / Graduation date: 2012

A Wireless Sensor Network (WSN) is a group of spatially dispersed sensors that monitor the physical conditions of the environment and collect data at a central location. Sensor deployment is one of the main design aspects of WSNs as this a ffects network coverage. In general, WSN deployment methods fall into two categories: planned deployment and random deployment. This thesis considers planned sensor deployment of a Mobile Wireless Sensor Network (MWSN), which is defined as selectively deciding the locations of the mobile sensors under the given constraints to optimize the coverage of the network. Metaheuristic algorithms are powerful tools for the modeling and optimization of problems. The Krill Herd Algorithm (KHA) is a new nature-inspired metaheuristic algorithm which can be used to solve the sensor deployment problem. A Multi-Agent System (MAS) is a system that contains multiple interacting agents. These agents are autonomous entities that interact with their environment and direct their activity towards achieving speci c goals. Agents can also learn or use their knowledge to accomplish a mission. Multi-agent systems can solve problems that are very difficult or even impossible for monolithic systems to solve. In this work, a modification of KHA is proposed which incorporates MAS to obtain a Multi-Agent Krill Herd Algorithm (MA-KHA). To test the performance of the proposed method, five benchmark global optimization problems are used. Numerical results are presented which show that MA-KHA performs better than the KHA by finding better solutions. The proposed MA-KHA is also employed to solve the sensor deployment problem. Simulation results are presented which indicate that the agent-agent interactions in MA-KHA improves the WSN coverage in comparison with Particle Swarm Optimization (PSO), the Firefly Algorithm (FA), and the KHA. / Graduate

The understanding of what properties the fibre should have in order to give the right end-product paper properties, along with the type of processing, is a subject for a lot of research and development. Today the ability to measure fibre properties on-line is widely used for pulps. It is often necessary to measure many properties and variables in a process. The data collected is therefore almost always multivariate. It is very hard to analyse process data due to a lot of noises. Correlations between fibre and paper properties are hard to find, but this does not mean that correlations do not exist. Fibre properties, measured by the pulp analyser PulpEye, were investigated and correlations to paper properties were studied. The work was divided into three different studies. Study 1 was an investigation of historical process data, in study 2 pulp samples from the production was analysed and study 3 was a refiner setting trial, were different refiner segments, flows through refiners and intensities were studied. Both the group-plots and MVDA’s based on the historical process data in study 1, showed that the Scott Bond was increased with increased amount of kinks and curl for the unbleached pine pulp (softwood pulp). Coarseness measurements, made in the study of historical data, indicated that the coarseness was varying in such a large extent that it was believable that it had effects in the papermaking process. Another interesting fibre property, investigated in the refiner setting trial, was crill. The amount of crill is said to have strong correlation to paper strength. The analysis showed that the incoming pulp had different amount of crill and that the amount of crill after the refiners also was varying for the different samples. The development of crill at different kappa numbers and for pulps refined with different segments and refiner strategies should be further investigated. In this work it has been difficult to find correlations between fibre properties and paper properties in the refiner setting trial. This could have been due to small variations of the different parameters. This work showed that the normal production can be handled very well and variations are rather small. It can be seen though, that problems do appear when parameters are deviating from the normal case. An efficient way to work is to do measurements when the incoming pulp parameters are deviating. It should also be more investigated how the most common deviating pulp parameters should be handled in the refining process and at the board machine. The communication between the pulp production and the board machines is recommended to be further developed, especially when the pulp production have disturbances that can be affecting the refining and further the board production.