Aircraft fault-tolerant trajectory control using Incremental Nonlinear Dynamic Inversion

Highlights

• A novel aircraft trajectory controller with four control loops is proposed.

• The approach does not require designing parameter update laws or online model identification.

• The approach does not require additional effort for designing the aerodynamic model structure.

• The approach is robust to model uncertainties as well as structural and actuator faults.

• The performance of the proposed approach is compared to existing approaches.

Abstract

This paper deals with aircraft trajectory control in the presence of model uncertainties and actuator faults. Existing approaches, such as adaptive backstepping and nonlinear dynamic inversion with online model identification, can be applied. However, since there are a number of unknown aerodynamic derivatives, the tuning of parameter update law gains is time-consuming. Methods with online model identification require excitation and the selection of a threshold. Furthermore, to deal with highly nonlinear aircraft dynamics, the aerodynamic model structure needs to be designed. In this paper, a novel aircraft trajectory controller, which uses the Incremental Nonlinear Dynamic Inversion, is proposed to achieve fault-tolerant trajectory control. The detailed control law design of four loops is presented. The idea is to design the loops with uncertainties using the incremental approach. The tuning of the approach is straightforward and there is no requirement for excitation and selection of a threshold. The performance of the proposed controller is compared with existing approaches using three scenarios. The results show that the proposed trajectory controller can follow the reference even when there are model uncertainties and actuator faults.

Introduction

Civil aircraft are usually required to follow trajectories in three-dimensional space, such as those imposed by air traffic control (Kaminer, Pascoal, Hallberg, & Silvestre, 1998). In the design phase of civil aircraft, safety is of critical concern. Many techniques (Almeida and Leißling, 2010, Alwi et al., 2010, Castaldi et al., 2014, Dobrokhodov et al., 2011, Patton, 1997, Shtessel et al., 2002, Yang et al., 2001, Zhang and Jiang, 2008, Zolghadri, 2012, Zolghadri et al., 2014) have been proposed to improve the safety level and reduce critical risks. Conventionally, since the aircraft model is nonlinear, flight control systems are designed based on a number of linearized models around certain operating points (Zhang & Jiang, 2008). Next, a gain scheduling method has to be used to blend the gains in different operating points into one controller.

Using nonlinear control approaches, design of different operating points can be avoided. Linearization-based methods, such as Nonlinear Dynamic Inversion (NDI) and Backstepping (BS), are nonlinear control methods which can handle nonlinearities in the model. This paper also makes use of nonlinear control methods to design a trajectory controller. Aircraft trajectory control has been considered by several researchers (Farrell et al., 2005, Kaminer et al., 1998, Lombaerts et al., 2011, Maximilian et al., 2013, Ren and Beard, 2004, Singh et al., 2003, Sonneveldt et al., 2009). To design the trajectory controller, the uncertainties of aerodynamic derivatives have to be considered since they can degrade the performance of the nonlinear control approaches. Singh et al. (2003) use a sliding mode adaptive controller to reduce the influence of uncertainties. In Ren and Beard (2004), a control Lyapunov function approach is applied. In Farrell et al. (2005), a Command Filtered Backstepping approach which uses adaptive function approximation is applied to design the trajectory controller. Maximilian et al. (2013) deal with model uncertainties using the NDI method. The influence of model uncertainties is decreased by making use of a concurrent learning approach (Maximilian et al., 2013). Sonneveldt et al. (2009) apply Adaptive Backstepping (ABS) (Krstić et al., 1992, Krstic et al., 1995) with parameter update laws. However, the computational load of the ABS is intensive and the tuning of the parameter update law gains is time-consuming (Sonneveldt, Chu, & Mulder, 2007). In Lombaerts et al. (2011) and Tol, Visser, Sun, Kampen, and Chu (2016), NDI is used to deal with the nonlinearities and the model uncertainties influence is reduced by identifying the unknown parameters online. The method is validated on the SIMONA (SImulation, MOtion and NAvigation) research simulator (Lombaerts et al., 2011). However, this method is based on parameter identification and it requires excitation (Lombaerts et al., 2011) which could limit its performance when there is no excitation. Furthermore, to deal with highly nonlinear aircraft dynamics, an aerodynamic model structure needs to be designed (Lombaerts, Van Oort, Chu, Mulder, & Joosten, 2010).

The present paper proposes a novel nonlinear controller for aircraft trajectory control. The NDI-based approach is used to deal with nonlinearities in the model. The solution to cope with uncertain aerodynamic derivatives is to make use of Incremental Nonlinear Dynamic Inversion (INDI) (Bacon et al., 2001, Sieberling et al., 2010, Simplício et al., 2013). A control structure with four loops is designed: position control, flight path control, attitude and angular rate control. Through analysis, it is found that there are only model uncertainties in the flight path and angular rate control loops. Therefore, these two control loops are designed based on the INDI control law while the remaining two control loops are designed based on the NDI control law. An additional benefit of using the proposed approach is that there is no need to design the aerodynamic model structure. The overall control architecture of the trajectory controller and the detailed design are presented in the paper.

The performance of our approach is compared to the approach proposed in Lombaerts et al. (2011). The model-based approach in Lombaerts et al. (2011) requires sufficient excitation, which may be difficult to obtain during failure situations. The performance comparison is performed using three scenarios: no fault, model uncertainties, and actuator faults. All three scenarios demonstrate the performance of the proposed trajectory controller.

The faults considered in this paper only include actuator faults. For fault-tolerant control (FTC) in the presence of sensor faults, the reader is referred to Marzat, Piet-Lahanier, Damongeot, and Walter (2012), Castaldi, Geri, Bonfe, Simani, and Benini (2010), Freeman, Seiler, and Balas (2013), and Lu et al., 2015, Lu et al., 2016.

The structure of this paper is as follows: Section 2 presents the aircraft model which is used for designing the trajectory controller. In Section 3, the four control loops and the distribution of model uncertainties are introduced. The detailed control law design is given in Section 4. The position control and attitude control loops are designed based on the NDI control approach while the flight path control and angular rate control loops are designed based on the INDI control approach. In Section 5, the aircraft model and the fault scenario are presented. The baseline controller, Nonlinear Dynamic Inversion with model identification (NDI-MI), is also presented. In Section 6, the performance of the proposed trajectory controller is compared to the existing method which is NDI-MI. Their performances are compared under three scenarios. Finally, the conclusions are given in Section 7.