rvo2.js

Multi-agent collision avoidance library

A JavaScript port of UNC's RVO2 library.

As stated in the RVO2 documentation: local collision avoidance as provided by RVO2 Library should in principle be accompanied by global path planning that determines the preferred velocity of each agent in each time step of the simulation.

Demo

Usage

Download the minified library and include it in your html.

<script src="js/rvo2.min.js"></script>

This standard collision avoidance simulation places agents around a ring at antipodal positions from their goals.

<canvas id="rvo-test" height="800" width="800" />
<script>
(function() {
  // parameter descriptions are given here: http://gamma.cs.unc.edu/RVO2/documentation/2.0/params.html
  var sim = new RVO.Simulator(2, 50, 5, 10, 10, 7, 1, [0, 0])
    , goals = []
    , center = [400, 400]
    , radius = 380
    , total = 100
    , timeStep = 13;

  // add 100 agents and specify goals for each
  for (var angle = Math.PI / total; angle < Math.PI * 2; angle += Math.PI / (.5 * total)) {
    var point = RVO.Vector.multiply([Math.sin(angle), Math.cos(angle)], radius);
    sim.addAgent(RVO.Vector.add(center, point));
    goals.push(RVO.Vector.subtract(center, point));
  }

  var ivl = setInterval(function() {
    if (reachedGoals(sim, goals)) {
      clearInterval(ivl);
    }
    else {
      updateVisualization(sim);
      setPreferredVelocities(sim, goals);
      sim.doStep();
    }
  }, timeStep);
})()

// checks to see if all agents have reached their goals
function reachedGoals(sim, goals) {
  for (var i = 0, len = sim.agents.length; i < len; i ++) {
    if (RVO.Vector.absSq(RVO.Vector.subtract(sim.agents[i].position, goals[i])) > 1) {
      return false;
    }
  }
  return true;
}

// assigns new velocities to all the agents
function setPreferredVelocities(sim, goals) {
  for (var i = 0, len = sim.agents.length; i < len; i ++) {
    var goalVector = RVO.Vector.subtract(goals[i], sim.agents[i].position);
    if (RVO.Vector.absSq(goalVector) > 1) {
      goalVector = RVO.Vector.normalize(goalVector);
    }
    sim.agents[i].prefVelocity = goalVector;
  }
}

// draws the agents
function updateVisualization(sim) {
  var ctx = document.getElementById('rvo-test').getContext('2d');
  ctx.clearRect(0, 0, 800, 800);
  ctx.fillStyle = 'rgba(200, 0, 0, .7)'

  for (var i = 0, len = sim.agents.length; i < len; i ++) {
    var agent = sim.agents[i];
    ctx.beginPath();
    ctx.arc(agent.position[0], agent.position[1], agent.radius * .9, 0, Math.PI * 2);
    ctx.closePath();
    ctx.fill();
  }
}
</script>