SimArena Documentation

SimArena is a powerful toolkit for creating, configuring, and running simulations with multiple agents in various scenarios. It provides a robust foundation for AI research and development, with support for custom objectives, brains, weapons, and maps.

---

Project maintained by arbyun Hosted on GitHub Pages — Theme by mattgraham

Brains

Brains are the decision-making entities in SimArena. They are responsible for deciding what actions to take in response to the game state. Brains are attached to agents.

Index

• Brain Types
• Brain Creation
• Using Brains
• Brain Configuration

Brain Types

Brains can be of different types, each with its own set of parameters and behavior. We provide two example brain types, RandomBrain and SimpleBrain. Brains are intended to be extended by users to create custom brains, according to their needs.

Brain Creation

To create a new brain instance, your brain class must inherit from Brain and implement the correct constructor signatures:

public class MyCustomBrain : Brain {
    public MyCustomBrain(Agent agent, IMap map, int team, int tickIntervalMs = 500) 
            : base(agent, map, team, tickIntervalMs)
        {}
        
        public MyCustomBrain(IMap map, int team, int tickIntervalMs = 500) 
            : base(map, team, tickIntervalMs)
        {}
}

It’s recommended to also pass a Simulation instance to the constructor, so you have access to any data stored there.

The brain should use its constructor to receive all the information it will need to make its decisions.

The brain logic listened by the simulation is defined inside the ExecuteThink method.

protected override void ExecuteThink()
{
    // Your brain logic goes here!
}

This method will be called every time the simulation updates, through the method Think.

public virtual void Think()
{
    if (!_initialized)
    {
        throw new Exception("Brain not initialized.");
    }
    
    // In fast mode, always execute thinking logic without time checks
    if (FastMode)
    {
        ExecuteThink();
        return;
    }
    
    // In normal mode, check if tick interval has passed since last time we made a decision
    if ((DateTime.UtcNow - _lastDecisionTime).TotalMilliseconds < _tickIntervalMs)
        return;
    
    // If so then continue to this
    _lastDecisionTime = DateTime.UtcNow;
    
    // Execute brain-specific logic
    ExecuteThink();
}

Think can be overridden to add additional functionality before or after calling ExecuteThink, such as logging or other bookkeeping tasks. Full overrides of Think that remove other functionality is not recommended because it may interfere with the simulation’s internal timing mechanisms.

Using Brains

Once you’ve created a brain, you can attach it to an agent using the SetBrain method on the agent object.

For example:

var myAgent = new Agent(1);
myAgent.SetBrain(new MyCustomBrain(myAgent, map, 1));

Or alternatively, when creating the agent itself:

var myAgent = new Agent(2, new MyCustomBrain(map, 1));

You can also change the brain at runtime by simply setting a new one on the agent.

myAgent.SetBrain(new AnotherBrain(myAgent, map, 1));

You can also choose to create the agent and brain separately first, and then assign the brain later. In that case, after creating the brain, you should assign it to the agent like this:

var myAgent = new Agent(3);
var myBrain = new MyCustomBrain();
myAgent.SetBrain(myBrain);
myBrain.SetAgent(myAgent);

Note that the brain must be assigned to the agent before it can start making decisions.

Brain Configuration

Brains can be configured via JSON files. To set up configuration for your brain, see the brain configuration documentation.