ACQUISITION AND PRIMARY PROCESSING OF DATA FROM THE ARDUPILOT SYSTEM FOR IDENTIFICATION OF THE QUADROCOPTER DYNAMICS MODEL

Abstract

The necessity of the quadrocopter flight experimental data received from the Ardupilot system or a similar system primary processing is substantiated. The sequential nature of signals polling and registration from sensors, as well as the different measurement principles used in them are the main reasons for the emergence of this necessity. The processing purpose is to synchronize the samples in the signal records and bring the measurement results of the synchronized data to the coordinate system associated with the object. On the basis of literature sources and as a result of the experiment, it is shown that it is possible to consider the signal vectors characterizing the motion of the quadcopter in hovering mode as vectors of stationary random processes. An algorithm for applying the method of cross-correlation function to synchronize the counts of a set of experimental data has been developed. The essence of the algorithm is to determine the delay of one signal in relation to another and use the delay to determine the numbers of synchronous samples in the records. It is proved that the Ardupilot hardware allows you to obtain experimental data that are necessary to identify the quadrocopter dynamics model, which characterizes its dynamics relative to a linked coordinate system, since it allows you to measure the coordinates of one vector relative to two coordinate systems. An algorithm for the unambiguous determination of the transformation matrix using the vector of relative to the associated coordinate system quadrocopter center mass accelerations and the vector of relative to the Earth's surface quadrocopter center mass velocities is presented. The algorithm is based on the method for determining the quaternions of instantaneous rotation of the aircraft as an eigenvector corresponding to the maximum eigenvalue of a specially defined numerical matrix.