Unmanned aerial vehicles (UAV) have been gaining momentum in recent years because of their vast application areas in the defence and civilian sectors. Their use can result in future missions to be more effective while also be conducted in a safer manner. UAVs, due to their large operational potential may be required to travel over long distances and prespecified way points thus requiring an effective and efficient decision making mechanism which allow the UAV to start and complete its mission. These missions may start and terminate at different or identical topological locations (or nodes). In the current paper a node-to-node graph theory energy costs modelling method is developed and presented. The modelling method requires as a priori the node-to-node energy costs. Furthermore, several propositions were presented to allow for the energy matrix to be perturbable, thus representing possible atmospheric variations which may occur during the UAVs mission. The UAV is modelled using energy graphs which allow a topological optimum to be obtained via suitable optimization algorithms. The energy costs implicitly contain time, propulsion force, and velocity information thus producing realistic results.

In this article, a hybrid electric propulsion system which consists of a fuel cell (FC) and a battery is proposed for an unmanned aerial vehicle (UAV) propulsion application. Based on the UAV propulsion power requirements during take-off, climb, endurance, and maximum velocity, the hybrid electric power plant specifications are defined to respond to any propulsion power demand. A power and energy management system is introduced to control the hybrid system power flow while optimizing the FC system performances. The power and energy management system consists of three sections called, power management system, power electronic interface, and energy management system. The power management system decides the operating power of the each power source based on the propulsion power demand and the battery state of charge. The power electronic interface is an implementation protocol of the power management decisions through a unidirectional power converter and a bidirectional power converter which are connected to the FC, the battery, and the DC bus. Based on the FC current decided by the power management system, the energy management system controls the air supply system of the FC to maximize the FC system net power output. A referenced model is used to obtain the optimum inlet air pressure of the FC and a neuro-fuzzy-based adaptive control architecture adapts the FC air compressor power to the optimum value. The results show that the optimum compressor power configuration is superior than the constant power configuration.

No abstract