Smart Rockets is a 0-player game that uses Genetic Alghoritms to optimize for the path between an Initial Point and a Target. There is a population of Rockets that can move freely in space during a time cycle, attempting to find the Target. The first generation starts with random DNA, which are instructions for the acceleration, therefore velocity and trajectory, that the rocket will perform during its life cycle. After the lifespan, a fitness calculating function is applied to the population in order to select the two rockets that finished closer (or, in case of a tie, faster) to the target, which will be selected for reproduction. The reproduction process is done by crossing over the DNA of the two progenitors, with the possibility of random mutations and genetic anomalies. With the assistance of an heuristic function, the new generation will likely increase its performance compared to the previous one, possibly leading to optimum performance after many generations.
On the left, there is the main canvas of the game, where the rockets can move freely. The rockets become translucent when they hit an obstacle or the wall. The target has a blinking effect in order to grab the user's attention. The rocket's propulsor has different sizes according to the rocket's velocity. The walls and obstacles are in greenish blue. There is a time bar in the bottom in orange that becomes shorter with time. The user can move an obstacle by clicking on an existing obstacle and holding while moving, can create an obstacle by pressing on an empty space and dragging the mouse and can remove obstacles by right clicking on an existing obstacle. The user can also change the location of the target. On the right, it’s the GUI. On the top it displays relevant information and on the bottom there are the buttons the user can interact with to change the options.
Buttons:
- Level: Other than moving\adding\removing obstacles and the target, there are 5 different pre-programmed levels you can try out.
- Fast Render: Each life cycle takes around 15 seconds, activate the fast render for instant generations.
- Anomaly Rate: The anomaly rate defines what is the percentage of rockets that will have random DNA mixed with their progenitors DNA. Options: 0.0, 0.01, 0.05, 0.1. Default: 0.05
- Mutation Rate: The Mutation rate defines how many rockets will have fully random DNA per generation. Options: 0.0, 0.01, 0.05, 0.1. Default: 0.05
- Rockets: The number of rockets. Options: 2, 10, 25, 50, 100, 250, 500, 1000. Default: 500
- Heuristic: Can be either Euclidean distance or Shortest Path using A* Path Finding Algorithm.
- Draw Path: Allows for the possibility of visualizing the shortest path between each rocket and the target. Warning: This is heavy on the CPU :) I implemented multicore processing to speed the calculations up since it can be quite processing intensive when the number of rockets its high.
- Reset: Re-initializes the rockets from generation 0 with random DNA instructions.
- 1X: Defines the speed of the game. Options: 1X, 2X, 3X. Default: 1X
Install Processing -> https://processing.org/
Download this Repository
Open with Processing
Press Play :)