DATA DRIVEN INTELLIGENT AGENT NETWORKS FOR ADAPTIVE MONITORING AND CONTROL

To analyze the characteristics and predict the dynamic behaviors of complex systems over time, comprehensive research to enable the development of systems that can intelligently adapt to the evolving conditions and infer new knowledge with algorithms that are not predesigned is crucially needed. This dissertation research studies the integration of the techniques and methodologies resulted from the fields of pattern recognition, intelligent agents, artificial immune systems, and distributed computing platforms, to create technologies that can more accurately describe and control the dynamics of real-world complex systems. The need for such technologies is emerging in manufacturing, transportation, hazard mitigation, weather and climate prediction, homeland security, and emergency response. Motivated by the ability of mobile agents to dynamically incorporate additional computational and control algorithms into executing applications, mobile agent technology is employed in this research for the adaptive sensing and monitoring in a wireless sensor network. Mobile agents are software components that can travel from one computing platform to another in a network and carry programs and data states that are needed for performing the assigned tasks. To support the generation, migration, communication, and management of mobile monitoring agents, an embeddable mobile agent system (Mobile-C) is integrated with sensor nodes. Mobile monitoring agents visit distributed sensor nodes, read real-time sensor data, and perform anomaly detection using the equipped pattern recognition algorithms. The optimal control of agents is achieved by mimicking the adaptive immune response and the application of multi-objective optimization algorithms. The mobile agent approach provides potential to reduce the communication load and energy consumption in monitoring networks. The major research work of this dissertation project includes: (1) studying effective feature extraction methods for time series measurement data; (2) investigating the impact of the feature extraction methods and dissimilarity measures on the performance of pattern recognition; (3) researching the effects of environmental factors on the performance of pattern recognition; (4) integrating an embeddable mobile agent system with wireless sensor nodes; (5) optimizing agent generation and distribution using artificial immune system concept and multi-objective algorithms; (6) applying mobile agent technology and pattern recognition algorithms for adaptive structural health monitoring and driving cycle pattern recognition; (7) developing a web-based monitoring network to enable the visualization and analysis of real-time sensor data remotely. Techniques and algorithms developed in this dissertation project will contribute to research advances in networked distributed systems operating under changing environments.

Estimation of vertical groundwater fluxes into a streambed through continuous temperature profile monitoring and the relationship of groundwater fluxes to coaster brook trout spawning habitat

We hypothesized that the spatial distribution of groundwater inflows through river bottom sediments is a critical factor associated with the selection of coaster brook trout (a life history variant of Salvelinus fontinalis,) spawning sites. An 80-m reach of the Salmon Trout River, in the Huron Mountains of the upper peninsula of Michigan, was selected to test the hypothesis based on long-term documentation of coaster brook trout spawning at this site. Throughout this site, the river is relatively similar along its length with regard to stream channel and substrate features. A monitoring well system consisting of an array of 27 wells was installed to measure subsurface temperatures underneath the riverbed over a 13-month period. The monitoring well locations were separated into areas where spawning has and has not been observed. Over 200,000 total temperature measurements were collected from 5 depths within each of the 27 monitoring wells. Temperatures within the substrate at the spawning area were generally cooler and less variable than river temperatures. Substrate temperatures in the non-spawning area were generally warmer, more variable, and closely tracked temporal variations in river temperatures. Temperature data were inverted to obtain subsurface groundwater velocities using a numerical approximation of the heat transfer equation. Approximately 45,000 estimates of groundwater velocities were obtained. Estimated velocities in the spawning and non-spawning areas confirmed that groundwater velocities in the spawning area were primarily in the upward direction, and were generally greater in magnitude than velocities in the non-spawning area. In the non-spawning area there was a greater occurrence of velocities in the downward direction, and velocity estimates were generally lesser in magnitude than in the spawning area. Both the temperature and velocity results confirm the hypothesis that spawning sites correspond to areas of significant groundwater influx to the river bed.

INTERACTIVE EFFECTS OF CLIMATE CHANGE AND FUNGAL COMMUNITIES ON THE DECOMPOSITION OF WOOD-DERIVED CARBON IN FOREST SOILS

Soils are the largest sinks of carbon in terrestrial ecosystems. Soil organic carbon is important for ecosystem balance as it supplies plants with nutrients, maintains soil structure, and helps control the exchange of CO2 with the atmosphere. The processes in which wood carbon is stabilized and destabilized in forest soils is still not understood completely. This study attempts to measure early wood decomposition by different fungal communities (inoculation with pure colonies of brown or white rot, or the original microbial community) under various interacting treatments: wood quality (wood from +CO2, +CO2+O3, or ambient atmosphere Aspen-FACE treatments from Rhinelander, WI), temperature (ambient or warmed), soil texture (loamy or sandy textured soil), and wood location (plot surface or buried 15cm below surface). Control plots with no wood chips added were also monitored throughout the study. By using isotopically-labelled wood chips from the Aspen-FACE experiment, we are able to track wood-derived carbon losses as soil CO2 efflux and as leached dissolved organic carbon (DOC). We analyzed soil water for chemical characteristics such as, total phenolics, SUVA254, humification, and molecular size. Wood chip samples were also analyzed for their proportion of lignin:carbohydrates using FTIR analysis at three time intervals throughout 12 months of decomposition. After two years of measurements, the average total soil CO2 efflux rates were significantly different depending on wood location, temperature, and wood quality. The wood-derived portion soil CO2 efflux also varied significantly by wood location, temperature, and wood quality. The average total DOC and the wood-derived portion of DOC differed between inoculation treatments, wood location, and temperature. Soil water chemical characteristics varied significantly by inoculation treatments, temperature, and wood quality. After 12 months of decomposition the proportion of lignin:carbohydrates varied significantly by inoculation treatment, with white rot having the only average proportional decrease in lignin:carbohydrates. Both soil CO2 efflux and DOC losses indicate that wood location is important. Carbon losses were greater from surface wood chips compared with buried wood chips, implying the importance of buried wood for total ecosystem carbon stabilization. Treatments associated with climate change also had an effect on the level of decomposition. DOC losses, soil water characteristics, and FTIR data demonstrate the importance of fungal community on the degree of decomposition and the resulting byproducts found throughout the soil.