jccaleroe / life Goto Github PK

This project is part of the Artificial Life course at the National University of Colombia, taught by Jonatan Gómez.

A video of the project can be found on YouTube

This project consists of two species, Fishes and Whales, Fishes search for food moving with a Boids Algorithm. This Algorithm considers different points of space as attractors to each single fish. Those point are the center of each fish's neighborhood, an avoidance point to stay a little saparate from other fishes, an energy point, which is the best food that a fish can see (each fish has a vision radius) according to a balance of distance, energy that it has and contamination on that point (each time a fish eats, it produces some contamination) and a scape point which is activated when a whale is close, this attractor is the strongest but is only activated when the fish feels is in danger. When a Fish is too close to a energy source (energy sources are fractal binary trees, made with a Lindenmayer System with a recursion limit of 3) it stops to it's minimum speed, but if it is part of a flock, the fish prefers to move with the average speed of the group. If the fish is searching for food or is scaping from a whale, it swims with it's fastest speed.

Each fish and whale have a metabolism rate, which tells the amount of energy they lose at each second as well as an age variable that increases per second. Some of those two variables will trigger the agent's death.

Fishes are always searching for food, but whales stop when they have plenty of energy.

Whales chase the closest fish, but if two are almost the same as far, it prefers the one he was hunting.

Fishes can have children, rewarding the ones that can survive longer, for this, male and female have to like each other according to their phenotype (the phenotype is the non-linear affine tranformation and skin texture (generated with turing-morph when the fish is born)) but from time on time they ignore this discrimination. It they agree to reproduce, a cross-over and mutation is applied to the variables of the parents and are inherited to the offspring, those variables are the energy it can take from a tree on a single bite, vision, expectedLife, capacity, metabolism, reproductionRate, affine transformation, alpha paramater in turing-morph and starting energy.

Whales can reproduce too, for this, there are reproduction seasons and females choose the male with the biggest amount of energy, then females approach to the male and when are close enough they reproduce. The variables that are inherited are the vision, expectedLife, capacity, metabolism, and starting energy. When a female is reproduced, she has to wait some time no matter if she is in a reproduction season.

Trees are generated on different points of space, but only some of those sources are activated from time on time, this allows to see how fishes migrate using boids.

When a fish or whale is born, they have to wait some time before they can reproduce, and when they reproduce, they'll have to wait another time (the reproduction rate) to be able to do that again.

To make the movement of the agents smoother, they don't change their direction at each frame, instead a Levy Walk is implemented so they have a probability of update their direction and speed at each frame.

To move the agents, a Rigidbody and physics were used (thre is no gravity), so fishes move differenly according to their affine tranformation and size, also the smaller the fish, the fewer the chance to be caught by a whale, but also the fewer the chances to eat.