Genetic Bike¶
In Genetic Bike, you will create a simulation in Python to evolve an optimal bicycle frame.
The project factorizes into separate tasks quite naturally, but it is entirely up to you to decide how to split up the work. The whole task can be done in Python alone, using matplotlib for the visualization. Consider using numpy, too. The following sections will describe the problem in some more detail, and will give some background information.
Task description¶
You can break most of these constraints if you have a good reason. The only rule that you must stick to is: Have something to show on Friday!
- Geometry
The simulation is restricted to two dimensions: distance x and height h (plus time t of course).
- Bicycle
The bike consists of four vertices. Two of them represent wheels of an arbitrary size and small mass. The other two are more massive points that simulate the load of the rider on the frame. One of the wheels drives the frame. The four vertices are fully connected by six springs, with variable rest length and spring constants.
- Ground
A height profile of a road surface. To make the problem interesting, it must not be completely flat! Gravity exists and points to –h.
- Physics
Do not hardcode any kinematic behaviour, all movement should come from the laws of mechanics. You can assume that the driving wheel provides a constant acceleration and does not slip.
- Evolution
The final shape of the bike must not be hardcoded! Starting from random initial configurations, the final shape should emerge through an evolutionary process of competitive selection.
- Visualization
Each bike’s movement along the road in the evolution process should be shown as a matplotlib animation.