No abstract

An iterative constrained inversion technique is used to find the control inputs to the plant. That is, rather than training a controller network and placing this network directly in the feedback or feedforward paths, the forward model of the plant is learned, and iterative inversion is performed on line to generate control commands. The control approach allows the controllers to respond online to changes in the plant dynamics. This approach also attempts to avoid the difficulty of analysis introduced by most current neural network controllers, which place the highly nonlinear neural network directly in the feedback path. A neural network-based model reference adaptive controller is also proposed for systems having significant dynamics between the control inputs and the observed (or desired) outputs and is demonstrated on a simple linear control system. These results are interpreted in terms of the need for a dither signal for on-line identification of dynamic systems.

This paper presents an autonomous mission architecture for locating and tracking of harmful ocean debris with unmanned aerial vehicles (UAVs). Mission simulations are presented that are based on actual weather data, predicted icing conditions, and estimated UAV performance degradation due to ice accumulation. Sun position is estimated to orient search and observation maneuvers to avoid sun glare. The planning algorithms are based on evolutionary computation techniques combined with market-based cooperation strategies for multiple UAVs. Both single vehicle and multiple autonomously cooperating UAVs cases are demonstrated.

In order to operate in the national airspace, an aircraft system must have documentation and analysis to show that it can operate at a satisfactory level of safety. For traditional manned aircraft systems, this is equivalent to operating a reliable system. However with Unmanned Aerial Systems (UAS), a relatively unreliable system can safely be operated provided that the risk to bystanders on the ground is sufficiently low. This paper presents a set of design tools and methodologies which can be used to assess the risk associated with operating an UAS in a potentially populated area. The intended use of the tool is discussed and a risk assessment is provided for an existing UAS.

Evolutionary computation (EC) techniques have been successfully applied to compute near-optimal paths for unmanned aerial vehicles (UAVs). Premature convergence prevents evolutionary-based algorithms from reaching global optimal solutions. This often leads to unsatisfactory routes that are suboptimal to optimal path planning problems. To overcome this problem, this paper presents a framework of parallel evolutionary algorithms for UAV path planning, in which several populations evolve simultaneously and compete with each other. The parallel evolution technique provides more exploration capability to planners and significantly reduces the probability that planners are trapped in local optimal solutions.