Yoshikazu Miyazawa

Flight Trajectory Optimization for Modern Jet Passenger Aircraft with Dynamic Programming

Future Air Transportation Systems (ATS) would eventually have to treat the ever increasing demands in aviation industry. Fuel consumption and flight time could be significantly reduced by modifying the current sector based system into a more relaxed user priority based ATS. Trajectory Based Operations (TBO) is considered as one of the key technologies in this transition. This paper concentrates on a proposed flight trajectory optimization tool based on Dynamic Programming (DP), meteorological data and an aircraft performance model to obtain fuel minimum 4D-optimal flight trajectories for a single jet passenger aircraft considering the effect of wind conditions. Major drawbacks such as “Curse of Dimensionality” in DP are overcome with a unique method called “Moving Search space Dynamic Programming (MS-DP)” method. A quantitative evaluation has revealed that an average reduction of 9% in fuel consumption could be achieved with a tradeoff of flight time which exceeds at an average of 7% compared to a series of flight data measured by a commercial GPS receiver in an airliner cabin.

Dynamic Programming Applications to Flight Trajectory Optimization

Abstract Rapid increase of computational capability with high performance processors and parallel processing and continual cost-down of hardware make overlooked numerical methods revive for their own favorable nature. Dynamic Programming is one of the cases in the trajectory optimization. This paper pursues possibility of Dynamic Programming as a numerical method for flight trajectory optimization of aerospace vehicles. It examines two application examples of winged space vehicles reentry flight and fuel efficient flight of a jet passenger aircraft to demonstrate advantages of Dynamic Programming. It also discusses how to overcome disadvantages of Dynamic Programming, such as a large amount of computational time and quantization errors.

Arrival Time Assignment by Dynamic Programming Optimization

Japanese airspace capacity must expand in order to accommodate the increased air traffic expected in the near future. Efficient air congestion management is a promising approach for achieving this goal. Arrival management for inbound flights to Tokyo International Airport, the busiest airport in Japan, is considered to be the most demanding challenge for efficient air congestion management. In this paper, a concept of arrival management based on multiple aircraft trajectory optimization is proposed and examined using the actual flight track data. At first, free-flight trajectory optimization is applied to the inbound flights landing on one of the two runways to simulate the most efficient ideal flights. Next, time separation constraints at a merging point on the boundary of the terminal area are imposed in order to avoid conflicts among the aircraft in the terminal area. As a result of the optimization, optimal sequencing and flight time adjustments are generated. Benefits of the proposed concept are evaluated by comparing the optimal trajectories with the corresponding actual flight trajectories. Dynamic programming is used for the optimization of each flight trajectory and scheduling of the arrival times. The obtained results reveal that in addition to safe arrival time separation, trajectory optimization with arrival time assignment produces substantial benefits in terms of fuel consumption and flight time.

Quantitative Analysis of Conflict Between Aircraft by Using Radar Track Data

As the demand for air traffic increases, new air traffic management systems are needed. In the system, it is expressly necessary to avoid conflict between aircraft. Efficient methods to prevent conflict are currently based on the expertise of air traffic controllers. Hence, a two-step analysis method is described in this study. First, it is shown using a proposed method with en route radar track data that no conflicts occurred in the airspace over Japan during the period in question. Second, instructions provided to pilots by the air traffic controllers to solve conflicts are estimated from the same radar data, and the results allow some groups to be rearranged in terms of avoidance procedures. These results will be useful references when developing an automated system of conflict detection and resolution in the future.

A study on input design method for nonlinear non-minimum phase systems

This paper proposes a simple and direct input calculation method for the precise output tracking control of nonlinear non-minimum phase systems. This approach is based on receding horizon control, though it does not require any iterative calculation. The control inputs that eliminate the error between the desired trajectory and the estimated output trajectory are generated by using the preview command at each sampling time. In order to prevent control input divergence, an internal state condition at a finite preview time ahead is added to the command. Performance of this approach is demonstrated in some numerical simulations by using a nonlinear F-16 flight simulation model. The result shows that our proposed method can achieve an almost perfect tracking performance. Although the realtime use of the control may be difficult due to the considerable amount of control computation required, this method can also be utilized to assess the feasibility of guidance commands.