OpenMAS

A research and simulation library for multi-agent systems. This library is concerned to provide a set of easy-to-use simulation tools for researchers in fields such as multi-agent systems, multi-robot systems, wireless sensor networks, economic networks and ecological networks.

Level-0 interface is now avaiable

usage: Firstly, regist dynamic variables using global variable state. For example:

global state
state.x=unifrnd(-1,1,[2,10]);
state.theta=unifrnd(-pi,pi,[1,10]);

Then you have regist 10 nonholonomic particles with initial random values.

Next, you could define a global variable named 'para' to contain parameters. For example:

global para
para.c1=1;
para.c2=2;

Then, you should define a function to describe the differential equations. In this function, you need to claim global variables named 'state' and 'grad'. In addtion, you also need the 'para' if you need parameters. Once you regist a variable such as x and theta, then there will be automatically a grad_x and grad_theta and you have to assign values to them.

function myAlgorithm(t)
global grad state para
grad.dot_x = %some values
grad.dot_theta = %some_values;
end

After doing that, you need to start the simulation by a simple function 'go'. For example: go(@myAlgorithm);

When finished, the evolution of state is recorded into a global variable 'result' and you can take the data directly by name+His. For example:

global result
result.xHis
result.thetaHis

Features

agent-model support

single-integrator particles
double-integrator particles
self-propelled particles
general-linear models (The above models will provide in the first version of this library.)
user-defined nonlinear models
two-wheels cars
four-wheels cars
quad-rotors aircraft
fixed-wing aircraft
ships
satellites
walking humans
smart grids
sellers and consumers

communication capability support

sensors
wireless networks
random communication topologies
preset fixed communication topologies
switching communication topologies
time-varying communication topologies

visualization support

2D animation
3D animation
GIF files support
video files support

simulation methods support

Runge-Kutta methods
Fixed-step methods

multi-levels interface

Level-0 interface. In this level, we provide a template including a ode45 simulation template and a fixed-step template and the user can control all the variables and to build a novel system.
Level-1 interface. In this level, we provide a serious of common systems and the users just take care of the algorithm parts and get rid of the details of simulation details.
level-2 interface. In this level, we provide a set of common agents and users can only concentrating on the distributed protocols.

Coding style To be continued

luohc2004/openMAS