Real-Time Hybrid Simulation of an Unmanned Aerial Vehicle

Abstract

This work presents a real-time hybrid simulation for the analysis and optimization of the electronic control unit of a quadcopter. Therefore, the existing physical microcontroller hardware is coupled to a real-time computer model used to simulate the flight. This requires the numerical solution of nonlinear equations of motion including coordinate transformations. Knowing the flight dynamics, the simulated measurements of a virtual inertial measurement unit are determined and fed back to the physical flight control unit to calculate the required actuator response for a desired behavior, thereby closing the control loop. This type of hybrid simulation is currently the most efficient method to obtain a desired system performance before carrying out experimental tests with the entire physical system. Furthermore, a virtual reality module for real-time flight visualization was developed for better analysis of different flight scenarios. Since all results show excellent agreement with real flight testing, the work confirms the efficiency of the proposed system. During the tests it was e.g. possible to determine the effect of different inertia measurement unit sensors with specific noise characteristics on the overall flight dynamics and consequently, find the reason for rarely occurring engine failures. In addition, the project shows that complex real-time hybrid simulations on industrial level are possible even with low investment costs.