Both V-rep and Webots simulator says that they can simulate multiple robots. Is these simulators are capable for simulate swarm robots? Can I simulate the swarm intelligence algorithms like ACO, BeeClust algorithm, Stick pulling Algorithm, Occlusion Based Collective pushing etc. on these simulators? Which robot I have to select for implementing above algorithms? For implementing these algorithms which simulator will be better? Any suggestions is appreciated.
It's important to separate between simulation and artificial intelligence. An environment for simulating hundreds of robots on the screen is called a game engine. It can be realized with standard programming languages for example Python/pygame. The user creates an array of objects and each objects has a position and a speed value. The graphics engine displays all the objects on the screen. This is called a realtime simulation.
The other part is the artificial intelligence to control the objects. Without an AI, the robots in the simulation has to be controlled manually or by a random generator. They are displayed at the screen with smoothly 60 frames per second, but it's unknown how to move them right so that robots are solving a task. In general, a simulator for displaying hundreds of objects is easy to realize. The more advanced part is to create the AI which drives the objects autonomously.
Simulators like Webots, V-Rep, Gazebo, pygame or Unity3D are the environment in which a robotics controller is developed. They are providing the graphics and the underlying physics engine. They are working as a game construction set in which the user can script a game.
Any general purpose robot simulator like V-REP or Webots can get your task done, though with not the best support for swarm robots in particular. ARGoS is a good platform for swarm robot simulation, maintained well enough with active contributors and shall allow you to inherently have support for centralized/de-centralized control.
The most common swarm robots are the Kilobots and the Khepera and ARGoS supports both of them.
The ARGoS simulator specifically targets swarm robotics development. There is also a really awesome swarm robotics programming language, Buzz, that plugs into ARGoS. There is even a ROSBuzz version. Lastly, I recently discovered this fantastic repo named TensorSwarm for reinforcement learning of robot swarms that plugs into ARGoS.