Modeling and control of a gyrodyne aircraft using neural-network based control allocation

As vertical lift vehicles gain prominence in industry, there is an increased need for studying different configurations. Although an infinite number of configurations can be designed based on different requirements, only a few make it past the drawing board. In this thesis, a gyrodyne configuration is modelled and simulated, and an appropriate control system is developed in a simulation-based environment. This work comprises of developing mathematical models based on fundamental aerodynamic principles. The model is further validated and control laws are then designed based on the implemented model depending on the requirements set early in the design process. The thesis focuses on modelling, trim procedure, linearisation, development of control system. A novel approach to control allocation is taken to map the most efficient method to map desired moments to effector commands. A proof of concept simulation of the aircraft and comparisons based on multiple approaches are presented.