Auto tuning adaptive I-PD control having an on-line system identification mechanism for industrial multi-rotor helicopters

Abstract

For unmanned helicopters to be used in industrial applications, autonomous flight control is necessary. This study aims to shorten the time required to tune a controller and design a stable control system for model variation of unmanned multi-rotor helicopters. Designing model based controllers requires a significant amount of time and effort for model identification and the tuning of controller parameters. Additionally, changes in the dynamics of the plant lead to deterioration of control performance. In this paper, we propose adaptive I-PD control having on-line system identification mechanism for change-of-altitude direction of a helicopter to solve the problems that have been presented. First, the calculation of the altitude model and system identification mechanism are described. In order to guarantee the minimum phase property of the altitude model, we use the delta operator for system identification. We show the accuracy of the altitude model and the validity of the proposed system identification algorithm using flight data. Next, we design an adaptive I-PD controller based on the identified parameters. I-PD gains are adjusted sequentially by a partial model matching method. Finally, we demonstrate the validity of the proposed control system by way of a flight experiment. The experimental result shows we were able to design a highly precise altitude controller in about 10 seconds.

Keywords

• moving robot
• autonomy
• adaptive algorithm
• adaptive control
• system identification
• multi-rotor
• uav