Particle Evolver is an interactive application which uses genetic algorithms (GA) to evolve a 3-D particle system. Just as in biological evolution, methods such crossover, mutation and natural selection are used to evolve the system. Properties of the particle system such as speed, color and number of particles are used as genetic material. The user is the selective agent, working as what’s know in GA parlance as the fitness function. All that a user of this system has to do is designate whether he or she likes what they see. If they do, the system continues to evolve toward the predetermined targets, if not the population is killed off and a new one is randomly generated.

GitHub