Actuated Mobile Sensing in Distributed, Unstructured Environments
Mobile sensor networks present opportunities for improved in situ sensing in complex hydrodynamic environments such as estuarial deltas. This dissertation considers the design and implementation of the mobile sensor network system that was built as part of the Floating Sensor Network project for use in the Sacramento-San Joaquin Delta in California over the 2007-2012 time period. Individual Lagrangian sensor units collect hydrodynamic state information, which is then transmitted to a centralized server and assimilated to produce a state estimate for the entire hydrodynamic system. Physical obstacles, including the shoreline and natural vegetation, present a major challenge to operating mobile sensors in estuarial environments. Actuated mobile sensors are shown to be a viable solution; appropriate control techniques allow these sensors to avoid obstacles, meet navigational goals, and still collect the Lagrangian data necessary for the sensing objective. Issues addressed include physical design, communication techniques for mobile sensor networks, control schemes for fleets of underactuated vehicles in unstructured flow environments, assimilation techniques for mobile Lagrangian data, and field experiments to validate and demonstrate the actuated mobile Lagrangian sensor concept.