Contents
- Introduction
- Swarm based algorithms
Introduction
It all started by studing behaviors of natural swarms such as Ants, Termites, Honey bees, Birds, Fishes, Locusts, etc. Swarm robotics is about how large number of simple agents designed such that desired collective behavior emerges from local interaction between agents and enviroment. The main specifications of swarms, they show common goals and distributed network and decentralised control where goals are much more complex. Swarm robotics is new age technology in which swarm (group of bots) show interesting features such as
- Self-organisation (without centralalized entity to control),
- Flexibility (performing wide variety of tasks)
- Stigmergy (indirect communication) ,
- Self-assembly (autonomous self organization) , and Cooperation among each other.
Swarm has different purpose such for example,
- Biological swarm such as birds, ants show collective behavior for common goal.
- Swarm behavior demonstrates collective behavior for different goals.
- Swarm intelligence has ability to learn or decide as one identity.
- Swarm engineering is expression of problem,conditions on set of individual agents and to perform overall behaviors.
The ways in which swarms can be classified
- Size of swarm
- Communication range
- Communication Topology
- Communication Bandwidth
- Reorganization rate of swarms
- Ability of individual members
- Whether Homogeneous or Hetrogeneous
Question here arises is that how population of individuals that optimizes a function or a goal by collectively adapting to enviroment(local/global)?
Swarm based algorithms
- Ant Colony Optimization (ACO)
- The source of inspiration of ACO is Ants finding shortest path from their nests to food.
- Particle Swarm Optimization (PSO)
- Inspiration of PCO comes from birds moving togather for long distances in search of food.
- Artificial Fish Swarm Algorithm (AFSA)
- Inspiration of AFSA comes from collective movement observed in behaviors exhibited by fishes in search of food, following other fishes and protecting group against dangers.
- Bee Based Algorithms
- Honeybee foraging behavior based Algorithm
- Mating behavior in honeybee
- Queenbee evolution process based algorithm