Unmanned aerial vehicles (UAVs) are used in an increasing number of applications. Such applications may include navigating through heavy traffic and highly congested airways, where numerous static and dynamic obstacles impinge upon a UAV's flight. It is imperative that a UAV successfully avoids these obstacles, while improving its planned flight path according to certain criteria. We have modeled UAV path planning as a single objective optimization problem that utilizes a receding horizon approach, where the path is constrained to avoid obstacle collision and to account for flight aerodynamic constraints. The proposed method is gradient based, allowing for quick and robust convergence to a near optimal solution. This heuristic method converges closely to full-knowledge optimal solutions and will allow UAVs to be implemented in a greater amount of tasks and missions than before while lessening the risk to the safety of others and the safety of the UAV. Nomenclature i index for path segments j index for obstacles m number of planned path segments n number of static obstacles r radius, m t Bézier curve parametrization value v dynamic obstacle speed, m/s V UAV velocity, m/s x, y coordinates in domain path length, m f distance between UAV and final destination, m

Wind turbine design is a challenging multidisciplinary optimization problem, where the aerodynamic shapes, structural member sizing, and material composition must all be determined and optimized. Some previous blade design methods incorporate static loading with an added safety factor to account for dynamic effects. Others incorporate dynamic loading, but in general limit, the evaluation to a few design cases. By not fully incorporating the dynamic loading of the wind turbine, the final turbine blade design is either too conservative by overemphasizing the dynamic effects or infeasible by failing to adequately account for these effects. We propose an iterative method that estimates fatigue effects during the optimization process while quickly converging to the true solution. We also demonstrate an alternate approach where a surrogate model is trained to efficiently estimate the dynamic loading of the wind turbine in the design process. This surrogate model, once trained, was then incorporated in the optimization loop of the wind turbine blade. In contrast to the iterative method, there is significant upfront computational cost to construct the surrogate model. However, this surrogate model has been generalized to be used for different rated turbines and can predict the fatigue damage of a wind turbine with less than 5% error for baseline wind turbines of the same family. These methods can be used instead of the more computationally expensive method of calculating the dynamic loading of the turbine within the optimization routine.