Aerospace and Mechanical Engineering
Adjust Font Size: Normal Large X-Large

AEM faculty spotlight:

Yiyuan Zhao

At first glance, the idea of optimization may seem indigenous to aerospace or other transportation sciences. However, Yiyuan Zhao, an AEM professor and expert on the subject, argues that optimization is part of everyday life. Everything from planning the most efficient way to meet up with friends to the very language used today falls under this concept, Zhao says.

“There is a famous American slogan – be the best you can be,” Zhao says. “That’s a statement of optimization.”

Optimization has two parts. The first is to list things one wants to achieve and the second is the constraints placed on the situation. In the example ‘be the best you can be,’ ‘be the best’ is a goal while ‘you can be’ is a constraint, Zhao explains. “For ‘be the best,’ we must divide it further – be the best in terms of what? And ‘you can be in’ terms of…?" Zhao said. "These are all questions associated with optimization.”

The following is a Q&A with Professor Zhao regarding the current technologies and future problems associated with optimization.

Yiyuan Zhao

What is a major area of research interest for you?
I have several areas of interest. One focuses on finding a systematic way of classifying or categorizing trajectory prediction methods. In one day, there are 4,500-5,000 commercial airplanes in the air, or about one million people flying during any given day. With that many airplanes in the sky, people are trying to let every airplane save fuel and expedite arrival and departure process. A way to do that right now is through automation to predict where the aircraft will be in the future – that’s called trajectory prediction. There are many ways to predict a trajectory. By using an optimization framework, I have come up with a unified, comprehensive framework that can compare all trajectory prediction methods. By having that framework to compare different methods of trajectory prediction, you can easily go to the next step, which is to find ways to improve them.

UAVs are playing an ever-increasing role in aviation. Tell me a bit about the UAV research that you do.
For UAVs as opposed to standard transport aircraft, there is a main problem: how to stay in the air as long as possible. In order for you to be able to stay in the air for as long as possible, the vehicle has to consume the least amount of fuel per unit time. To do that, one may take advantage of wind in a UAV flight. This is an optimization problem. So far we have come up with interesting results. For some natural wind phenomena (like wind gradient or thermals), if you use them properly, you are able to fly with no or very little thrust. If you do that, you can sustain a UAV flight for a long time.

With so many aircraft, how are decisions made so no crashes occur?

Generally, if you are going to fly on your own, you will act differently than a group of aircraft.  But now, pilots need to plan an “optimal” way of flying by not only considering their objectives and constraints but also consider other vehicles. Needless to say, you need to “talk” a lot.  That talking is done through airborne networks. I send you my state periodically and vice-versa. Based on what you tell me, I will fly on an optimal trajectory not just for own performance but adjusted for your planned flight, as well.  Optimizing trajectories will also have noticeable impact for the average person. When individuals take a flight, sometimes a flight will be early, or more likely, late. If the system works well, over time the public will see less and less delays. 

Could you give an example of optimization we might see in everyday life?

If we plan a party, you might call letting me know that you are going to be a little late. I am still able to leave earlier than you, so to arrive at the best time, which is the same in this case, I can take a detour doing something else on my way there. In many optimization problems, like in the example I just gave, you know where you are going to start. But in today’s world, aerospace or everyday life, we experience more of a dynamic decision making, or distributed, model. We have a moving target, because where I’m gong to start is going to depend on where you are going to start and where you’re going to start will depend on where I’m going to start.

Last Modified: Wednesday, 05-Oct-2011 13:28:12 CDT -- this is in International Standard Date and Time Notation