This thesis describes the design of a data structure for use with Digitised Terrain Elevation Data (DTED) in Terrain Reference Navigation (TRN) systems. The data structure is based on a variant of quad-tree and oct-tree data structures to provide an efficient representation of terrain in terms of storage requirements and acccss operations. These data structure are applied to flight path planning operations in mission management applications. The algorithms developed for flight path planning have becri implemented in the C programming language for a standard PC. Current research in TRN systems is reviewed and attention is given to the use of hierarchical data structures to cope with the potentially large data base needed for DTED files. Data structure combining quad-trees and oct-trees are developed with an emphasis on data reduction using pointerless trees and the use of locational codes to provide straightforward mapping between quad-trees and oct-trees, in other words, between two-dimensional co-ordinates and three-dimensional co-ordinates. Analysis of these algorithms is described for two DTED files to illustrate storage improvements and to verify a set of database access operations. These data structures are applied to problems of flight path planning where the navigation space comprises objects above a specific altitude and this three-dimensional space is searched for a flight path which avoids the obstacles and satisfies specific operational criteria. Algorithms are developed to extract a visibility graph from the terrain database and to determine the preferred flight path from a set of paths which satisfy defined constraints. Several search techniques are developed which exploit the efficiency of the quad-tree and oct-tree data structures. These methods are extended to real-time flight-path planning where predicted times for access operations are used to direct flight path extraction by varying the tree resolution during computation of the flight path. A comprehensive set of results are provided to illustrate: the storage efficiency of quad-tree and oct-tree data structures the application of pyramid structures to represent navigation space analysis of the time to compute the visibility graph and to extract flight paths integration of these methods with a real-time mission management simulation on a PC The thesis draws conclusions on the efficiency of these techniques for the represcntation of DTEDs and to access objects in TRN systems. It is observed that the use of hierarchical data structures in the form of quad-trees and oct-trees offers significant improvement in accessing DTEDS, for future use in TRN systems. The thesis concludes by outlining areas of further work where the techniques can be further &N, cloped for applications in mission management and navigation using DTED files.

In this thesis, energy optimal route of an unmanned solar powered air vehicle is obtained for the given mission constraints in order to sustain the maximum energy balance. The mission scenario and the constraints of the solar powered UAV are defined. Equations of motion are obtained for the UAV with respect to the chosen structural properties and aerodynamic parameters to achieve the given mission. Energy income and loss equations that state the energy balance, up to the position of the UAV inside the atmosphere are defined. The mathematical model and the cost function are defined according to the mission constraints, flight mechanics and energy balance equations to obtain the energy optimal path of the UAV. An available optimal control technique is chosen up to the mathematical model and the cost function in order to make the optimization. Energy optimal path of the UAV is presented with the other useful results. Optimal route and the other results are criticized by checking them with the critical positions of the sun rays.

No description available.