Jie Rong

Air traffic conflict negotiation and resolution using an onboard multi-agent system

Conflict detection and resolution is a critical capacity for realizing free flight. In this paper, a multi-agent system algorithm is proposed to solve the air traffic conflict detection and resolution problem involving multiple aircraft, where each aircraft may be regarded as an intelligent agent. The conflict detection and resolution problem for each agent is solved as a constraint satisfaction problem. A pair wise, argument-based negotiation approach is established for the aircraft to search for a multilaterally acceptable conflict resolution. The proposed algorithm guarantees to lead to an appropriate solution, except for some extreme cases. It also considers real-time constraints such as response time, negotiation deadline, etc. The ground based air traffic controller is always included in this system, and acts as a high-level supervisor and coordinator. It has authority to approve or override any proposal from any aircraft; force an aircraft to accept a proposal; and open or close a negotiation. It may also join the collaboration to represent those aircraft which lack the capacity to negotiate for themselves. However, in such cases it must obey the same negotiation protocol as other aircraft.

Preliminary Results of Adaptive-Reinforcement Learning Control for Morphing Aircraft

This paper applies an Adaptive-Reinforcement Learning Control methodology to the problem of aircraft morphing. The reinforcement learning morphing control function is integrated with an adaptive dynamic inversion control trajectory tracking function. An episodic unsupervised learning simulation using the Q-Learning method is developed to learn the optimal shape change policy, and optimality is addressed by cost functions representing optimal shapes corresponding to flight conditions. The methodology is demonstrated with a numerical example of a hypothetical 3-D smart aircraft that can morph in all three spatial dimensions, tracking a specified trajectory and autonomously morphing over a set of shapes corresponding to flight conditions along the trajectory. Results presented in the paper show that this methodology is capable of learning the required shape and morphing into it, and accurately tracking the reference trajectory in the presence of parametric uncertainties, unmodeled dynamics and disturbances.

AUTOMATION CAPABILITIES ANALYSIS METHODOLOGY FOR NON-CONTROLLED AIRPORTS

Advanced technologies will be needed to enhance and improve the current transportation capabilities of the nations small aircraft transportation network, and thus help relieve congestion at hub airports. To achieve this goal, current operation concepts and procedures will have to be changed and improved, or perhaps even replaced. This paper proposes a decentralized Air Traffic Management approach, centered on the design of an automated arrival/departure system for non-controlled airports. A hybrid system is introduced that is a fusion of distributed air traffic management and centralized control. A new functional description of the airport terminal area infrastructure named Multi-layer Air Traffic Space is introduced, and automated terminal operations and procedures are addressed. Several types of intelligent agents with negotiation functions are developed in the automation system, and simulation methodology is presented, containing a full hardware and software description.

SMALL AIRCRAFT PILOT ASSISTANT: ONBOARD DECISION SUPPORT SYSTEM FOR SATS AIRCRAFT

Distributed pilot decision-making plays a critical role in high volume operations at noncontrolled/non-radar airports. The High Volume Operation concept, proposed by the NASA Small Aircraft Transportation System program, is based on an environment within which the decisionmaking is distributed mostly to the pilot. Realization of this concept relies on advanced cockpit systems that assist pilots in both information-processing and decision-making. This paper introduces an onboard pilot decision aid system called the Small Aircraft Pilot Assistant, which is dedicated to Small Aircraft Transportation System aircraft conducting High Volume Operations. The goal is to increase cockpit decision-making capacity by automating part of the pilot decision-making process, especially in the early stages of information acquisition and analysis. The Small Aircraft Pilot Assistant uses Artificial Intelligence techniques to aid pilots in following the High Volume Operation procedures, identifying the current flight segment, monitoring performance, advising of possible pilot mis-operation; and warning of potential traffic related hazards. A human factors analysis was conducted to determine the appropriate level of automation, using a real-time, multi-aircraft, multi pilot-in-the-loop simulation system called the Multi-agent Intelligent Distributed Airspace Simulation. Preliminary pilot test results show that the Small Aircraft Pilot Assistant is a promising system to satisfy the cockpit system requirements for the High Volume Operations of the Small Aircraft Transportation System.

Onboard pilot decision aid for high volume operations in self-controlled airspace

Distributed pilot decision-making plays a critical part in high volume operation in non-controlled/non-radar airports, a concept proposed by the NASA Small Aircraft Transportation System program. Realization of the concept relies on advanced cockpit systems that assist pilots in both information-processing and decision-making. In this paper, we presented the design of an onboard pilot decision aid system, called the Small Aircraft Pilot Assistant, which is dedicated to help pilots perform the high volume operation flight tasks. It increases the cockpit decision-making capacity and pilot situation awareness by automating part of the pilot decision-making process, especially in its early stages of information acquisition and analysis. The pilot tests of the system are conducted using a real-time, multi-aircraft, pilot-in-the-loop simulation system that is presently capable of middle-fidelity HVO simulation.

Feasibility Analysis of Aircraft Landing Scheduling for Non-Controlled Airports

** † Advanced technologies are needed to enhance the transportation capabilities of the nation’s small aircraft transportation network, and thus relieve the congestion of the hub airports. The existing air transport system cannot fully meet the public demand for safety, higherspeed mobility, and increased accessibility since the current hub-and-spoke model concentrates a large percentage of the air traffic at a few hub airports. To alleviate this problem, current operational concepts and procedures will change and be improved to make practical use of the roughly five thousand existing public-use-landing facilities. A previous paper by the authors developed a basic approach/departure system for non-controlled airports, using an intelligent agent system. This paper extends that work, and addresses aircraft landing scheduling issues for non-controlled airports. The basic problem of aircraft landing scheduling is described, followed by an aircraft landing scheduling model for the single runway case at non-controlled airports. Two different scheduling algorithms, consisting of linear programming and job shop scheduling, are proposed to complete the feasibility analysis. Required simulation methodologies and capabilities are presented with a full description of software and hardware facilities. Finally, numerical results are presented which support the development.

COCKPIT SYSTEM DESIGN FOR GENERAL AVIATION FREE FLIGHT USING A COGNITIVE ENGINEERING APPROACH

The realization of general aviation Free Flight requires advanced cockpit systems to assist pilots in managing information and decision-making. In this paper, the application of cognitive engineering concepts to cockpit system design for general aviation is discussed. The design of an Aircraft Approach and Landing Assistant is presented as an example of this method. Its purpose is to enhance pilot situational awareness, aid pilot decision-making, and reduce pilot workload during the approach and landing phase in an environment with complex weather, traffic and terrain conditions. The ongoing development of the system is based on the cognitive model of general aviation pilots. It is implemented into existing flight software and a real-time, pilot-in-the-loop flight simulation system is developed for its validation. The proposed approach appears to be a promising candidate for designing intelligent cockpit systems and decision-aiding tools for future general aviation Free Flight pilots.