Observer Controller Identification of a Medium-Weight Co-axial Octocopter

In this work, system identification of the open-loop airframe characteristics of a medium-weight co-axial octocopter is performed using closed-loop simulated data. Data is generated from the closed-loop octocopter model at hover using moments and thrust as excitation signals. Observer/Controller Identification (OCID) is then used to estimate the observer, controller and system Markov parameters from the generated data. The Markov parameters are used to form the Hankel matrices, that form the basis of the Eigensystem Realization Algorithm (ERA) used to determine the minimal realization of the system matrices. The effect of noise on the system identification process is analyzed by computing the Mode Singular Values (MSV). The use of Eigensystem Realization Algorithm with Data Correlations (ERA/DC) is also explored to demonstrate the improvement in system identification in the presence of noise. The system identification process presented in this work can be used to determine the system’s modes that help in analyzing the handling qualities of the aircraft, which can be useful in improving simulation model fidelity and estimating higher order dynamics.