Chen is currently an undergraduate student at Beijing Normal University, School of Systems Science. He's interested in computer modeling and simulation. Chen has done some work on traffic flow modeling, ecosystem modeling, and human resource network modeling. It is his belief that everything is computable, and the best way to understand a complex system is to let it run. Influenced by The Matrix series, he doubts what's real and what is not. Wolfram's approach gives an insight to understanding the world of complexity, and Chen is fascinated by what these simple rules can create.Talking about what's beauty, he'll say, Bach's music and his girlfriend.
Project: Phase Diagram of One-Dimensional Boids Model
This project is about creating a self-stabilized artificial ecological system with preys and predators by simple rules and programming.
There are already many researches into the behavior of swarms, and one of them is called the Boids model. In my understanding, Boids simplifies the animal swarm into something like billiards: they are driven by simple forces from different directions, and the outer forces determinate the balls' locations. The Boids model describes the influences each agent receives as "forces": separation, alignment, and cohesion desires steer the Boids based on the positions and velocities of their nearby flock mates.
One problem about the Boids model is how to determine the combination of these kinds of "forces", say, the weights of the rules. In former studies, the weights were chosen experientially, and we can observe the different behavior of Boids when the weights are changed. It's OK when there is just one kind of Boids in the system, but when we include other species into the system and divide Boids into preys and predators, and we want to keep an ecological equilibrium (the numbers of both the preys and the predators fluctuate around some levels), troubles ensue. That is because the rights to determine the weights of the rules should be given to the agents, not kept in our hands.
My project is about finding a way to create a system to let the agents decide the best choice of the actions to be taken, to achieve the goal of keeping itself alive with care about the well-being of its kind. I hope it can finally be a self-stabilized ecological system with only some outside resources. I'm looking forward to some interesting patterns emerging from the system, and it might give a deeper understanding of the Boids model, and the tricks for creating artificial life.