Iterative Learning Model Predictive Control Approaches for Trajectory Based Aircraft Operation with Controlled Time of A
- PDF / 501,950 Bytes
- 9 Pages / 594.77 x 793.026 pts Page_size
- 31 Downloads / 181 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
Iterative Learning Model Predictive Control Approaches for Trajectory Based Aircraft Operation with Controlled Time of Arrival Gaoyang Jiang* and Zhongsheng Hou Abstract: In this work, iterative learning model predictive control approaches are presented to resolve the control problem of trajectory based aircraft operation with time-of-arrival constraint. Firstly, we formulated this problem being an output tracking issue with along-track wind disturbance. Next, P-type point-to-point iterative learning control algorithms are designed to overcome the influence of repetitive along-track wind. In addition, a point-topoint iterative learning model predictive control algorithm with variable prediction step is also designed to eliminate the interference of non-repetitive gusts on the air route. Finally, numerical simulations are provided to verify the effectiveness of the proposed method. Keywords: Air traffic control, iterative learning control, model prediction control, trajectory based operation.
1.
INTRODUCTION
Currently, air traffic control is mainly handled at a strategic level by spacing air routes in order to minimize the potential conflict, and at a tactical level by giving instructions to aircraft to eliminate the incoming flight conflict. Trajectory based operation (TBO), as one of the main solution to current issue, is proposed by SESAR and NextGen program. TBO is defined as the ability to precisely fly a 3D trajectory while meeting the specified times of arrival at the predesigned waypoints in the airspace, which is called controlled time of arrival (CTA). It has been recognized as one of the most promising operational concepts in several future air traffic plans [1–3]. Extensive research has been carried out on the concept of CTA, such as computing the time-of-arrival according to basic flight models [4], quantifying the feasible CTA window at the metering fix [5], designing tailored arrival program to satisfy arrival time constraint [6], and adjusting aircraft speed according to the longitudinal position uncertainty in different time periods [7, 8]. Similar efforts have also been devoted to the application of CTA, enabling time and energy managed operations [9–11], time based sequence [12, 13], and merging procedures [14] in a 4D trajectory context. Although many studies have been performed, little research makes use of the historical data generated by flight.
The value of data has attracted more and more attention, especially in this era of big data. Thus, we attempt to find a simple and effective method, so that aircraft can learn from the historical data and change its state to meet the constraints of arrival time. It is worth noting that air traffic has a daily similarity between city pairs. For an aircraft, the airline will arrange for it to fly between the origin-destination city pairs repetitively along the predesigned route. A large amount of flight data will be produced during repeated operation, and the valuable information contained
Data Loading...
