Optimal Dubins Path Generation, System Identification and Control of a Paramotor

This paper presents a system identification, control and guidance framework for a small scale paramotor. The paramotor’s nonlinear state space model is initially introduced. Subsequently, a nonlinear model system identification procedure based on the Trust Region Reflective (TRR) least squares algorithm is used to obtain the model’s coefficients from actual recorded flight data. A Model Predictive Controller (MPC) is then designed for both the lateral and longitudinal control aspects of the paramotor based on the aforementioned model coefficients with simulations carried out to test the validity of the designed control scheme. A Dubins curve based guidance technique is then developed for generating a 2D in-flight trajectory for the paramotor’s autonomous flight navigation. The path planning algorithm is developed and simulated on the nonlinear model to verify its performance.