Incremental Nonlinear Dynamic Inversion for the Apache AH-64 Helicopter Control

Abstract

Incremental nonlinear dynamic inversion (INDI)-based controller is a promising technique that can be adopted to control dynamic systems with high nonlinearities and extended cross-coupling effects. This paper presents the incremental strategy adopted as a sensor-based controller approach for the Apache AH-64D Longbow helicopter. The strength of the INDI sensor– based approach is that it does not require a detailed vehicle model as it uses only a control effectiveness model and estimates of the vehicle angular accelerations replacing the rest of the model. The weakness is its sensitivity to measurement and actuator delays, demanding additional estimations of model parameters and control effectiveness tuning. Using the Boeing FLYRT Apache model to simulate the vehicle dynamics, the paper has developed three adaptation schemes connecting the INDI sensor–based approach to the Apache partial control authority set as ±20% authority in all axes. The goal was to redesign the existing Apache's flight control systems and demonstrate handling qualities improvements for hover and low-speed flight in normal and degraded visual environment. Implementing the attitude command attitude hold and translational rate command response types in the system through the INDI controller, the paper demonstrates that improvements from Level 2 to Level 1 handling qualities can be achieved when flying the Aeronautical Design Standard-33 (ADS-33E-PRF) hover and pirouette maneuvers in a degraded visual environment.