Multi Agent System: Fundamentals and Applications

By Fouad Sabry

What Is Multi Agent System

A multi-agent system is a type of computerized system that is made up of numerous intelligent agents that communicate with one another. It is conceivable for multi-agent systems to solve problems that a single agent or a monolithic system would have a difficult or impossible time resolving on their own. Methodical, functional, and procedural techniques, algorithmic search, and learning through reinforcement are all examples of possible types of intelligence.

How You Will Benefit

(I) Insights, and validations about the following topics:

Chapter 1: Multi-agent system

Chapter 2: Distributed artificial intelligence

Chapter 3: Software agent

Chapter 4: Intelligent agent

Chapter 5: Agent-based model

Chapter 6: Swarm intelligence

Chapter 7: Swarm robotics

Chapter 8: Consensus dynamics

Chapter 9: Agent-based social simulation

Chapter 10: Agent mining

(II) Answering the public top questions about multi agent system.

(III) Real world examples for the usage of multi agent system in many fields.

(IV) 17 appendices to explain, briefly, 266 emerging technologies in each industry to have 360-degree full understanding of multi agent system' technologies.

Who This Book Is For

Professionals, undergraduate and graduate students, enthusiasts, hobbyists, and those who want to go beyond basic knowledge or information for any kind of multi agent system.

• Language: English
• Publisher: One Billion Knowledgeable
• Release date: Jun 24, 2023

Book preview

Chapter 1: Multi-agent system

A multi-agent system, sometimes referred to as a self-organized system, is a computerized system that is made up of numerous intelligent agents that communicate with one another.

Even if there is a lot of overlap between the two, a multi-agent system and an agent-based model are not usually the same thing (ABM). In an agent-based model (ABM), the goal is not so much to solve specific practical or engineering problems as it is to search for explanatory insight into the collective behavior of agents (which do not necessarily need to be intelligent) obeying simple rules. These agents are typically found in natural systems. The language of ABM is more often used in the scientific community, but MAS is more common in engineering and technology.

Agents and the environments in which they operate make up multi-agent systems. Research on multi-agent systems often focuses on computer programs called agents. On the other hand, the agents that make up a multi-agent system might just as well be robots, people, or human groups working together. It's possible for a multi-agent system to have mixed human and computer agents on its roster.

Agents may be classified into a wide variety of categories, ranging from the simplest to the most complicated. Among the categories are also:

Agents who do nothing, often known as agents without aims (such as obstacle, apple or key in any simple simulation)

Active agents with straightforward objectives (such as birds flocking together or wolves and lambs in a scenario of prey and predator)

agents with a cognitive role (complex calculations)

Agent environments are able to be broken down into::

Virtual

Discrete

Continuous

Agent environments can also be organized according to properties such as accessibility (whether or not it is possible to gather complete information about the environment), determinism (whether or not an action causes a definite effect), dynamics (how many entities influence the environment at any given moment), discreteness (whether or not the number of possible actions in the environment is finite), and episodicity (whether or not the actions take place in discrete chunks) (whether agent actions in certain time periods influence other periods), The components that make up a multi-agent system each exhibit a number of essential qualities:

Agents that are at least partly independent, self-aware, and autonomous are said to have autonomy.

Local perspectives: either no agent possesses a complete global view, or the system is too complicated for any agent to effectively use such information

Decentralization is a kind of government in which no one person is identified as governing (or the system is effectively reduced to a monolithic system)

Even though the individual strategies of all of the multi-agent system's agents are straightforward, it is still possible for the system as a whole to exhibit self-organization, self-direction, and other forms of control paradigms, in addition to associated complicated behaviors. Within the confines of the communication protocol of the system, the method could result in a general improvement when agents are able to exchange their knowledge in any language that has been agreed upon by all parties. The Knowledge Query Manipulation Language (KQML) and the Agent Communication Language are two examples such languages (ACL).

A significant number of MAS are realized by the use of computer simulations, which advance the system through discrete time steps. The MAS components will often interact with one another via the use of a weighted request matrix, such as.

Speed-VERY_IMPORTANT: min=45 mph,  Path length-MEDIUM_IMPORTANCE: max=60 expectedMax=40,  Max-Weight-UNIMPORTANT

Contract Priority-REGULAR

in addition to a weighted response matrix like as.

Speed-min:50 but only if weather sunny,  Path length:25 for sunny / 46 for rainy

Contract Priority-REGULAR

note – ambulance will override this priority and you'll have to wait

It is standard practice in MAS systems to use a challenge-response contract arrangement.

To begin, a Who can? query is posed to the audience.

Only the components that are relevant will answer with I can, at this price..

In the end, a contract is established, which often takes place over the course of a number of very brief exchanges between the parties, in addition to this, take into account the other components, the constantly changing contracts, and the limitation sets of the component algorithms.

The pheromone paradigm is another one that is widely utilized with MAS. In this paradigm, components leave information for other components that are close. These pheromones may lose their potency or become more concentrated over time, which means that their values may drop (or increase).

MAS have a propensity to figure out the most effective answer to their difficulties without outside assistance. There are many parallels can be drawn between this and the behavior of physical phenomena, such as energy minimization, which refers to the tendency of physical things to achieve the lowest attainable energy level within