Robust Control for Non-Minimum Phase Systems with Actuator Faults: Application to Aircraft Longitudinal Flight Control

This study is concerned with developing a robust tracking control system that merges the optimal control theory with fractional-order-based control and the heuristic optimization algorithms into a single framework for the non-minimum phase pitch angle dynamics of Boeing 747 aircraft. The main control objective is to deal with the non-minimum phase nature of the aircraft pitching-up action, which is used to increase the altitude. The fractional-order integral controller (FIC) is implemented in the control loop as a pre-compensator to compensate for the non-minimum phase effect. Then, the linear quadratic regulator (LQR) is introduced as an optimal feedback controller to this augmented model ensuring the minimum phase to create an efficient, robust, and stable closed-loop control system. The control problem is formulated in a single objective optimization framework and solved for an optimal feedback gain together with pre-compensator parameters according to an error index and heuristic optimization constraints. The fractional-order integral pre-compensator is replaced by a fractional-order derivative pre-compensator in the proposed structure for comparison in terms of handling the non-minimum phase limitations, the magnitude of gain, phase-margin, and time-response specifications. To further verify the effectiveness of the proposed approach, the LQR-FIC controller is compared with the pole placement controller as a full-state feedback controller that has been successfully applied to control aircraft dynamics in terms of time and frequency domains. The performance, robustness, and internal stability characteristics of the proposed control strategy are validated by simulation studies carried out for flight conditions of fault-free, 50%, and 80% losses of actuator effectiveness.

The Adaptive Dynamic Programming Toolbox

The paper develops the adaptive dynamic programming toolbox (ADPT), which is a MATLAB-based software package and computationally solves optimal control problems for continuous-time control-affine systems. The ADPT produces approximate optimal feedback controls by employing the adaptive dynamic programming technique and solving the Hamilton–Jacobi–Bellman equation approximately. A novel implementation method is derived to optimize the memory consumption by the ADPT throughout its execution. The ADPT supports two working modes: model-based mode and model-free mode. In the former mode, the ADPT computes optimal feedback controls provided the system dynamics. In the latter mode, optimal feedback controls are generated from the measurements of system trajectories, without the requirement of knowledge of the system model. Multiple setting options are provided in the ADPT, such that various customized circumstances can be accommodated. Compared to other popular software toolboxes for optimal control, the ADPT features computational precision and time efficiency, which is illustrated with its applications to a highly non-linear satellite attitude control problem.

Finite Sample Properties for Closed Loop System Identification

In this paper, after closed loop system identification is reviewed, asymptotic analysis and finite sample analysis for closed loop system identification are studied respectively, corresponding to the infinite data and finite data. More specifically, within the framework of infinite data, the cost function is modified to its simplified form, and one optimal feedback controller is obtained based on our own derivations. The simplified cost function and optimal feedback controller are benefit for practical application. Furthermore, the asymptotic variance of that optimal feedback controller is also yielded from the point of asymptotic analysis. In the case of finite data, finite sample properties are constructed for closed loop system identification, then one difference between the sampled identification criterion and its corresponding expected criterion is derived as an explicit form, which can bound one guaranteed interval for the sampled identification criterion. Finally, one simulation example is used to prove the efficiency of our proposed theories.