Over the past decade, solving complex optimization problems with metaheuristics has received considerable attention among practitioners and researchers. As a result, many metaheuristic algorithms have been developed in recent years, and a large majority of these algorithms are often inspired by nature. Today, bio-inspired methods are becoming more and more popular. This popularity and success stems mainly from the fact that these algorithms were developed by mimicking nature's most efficient processes, including biological systems and physical and chemical processes. In this chapter, we will talk about metaheuristics and bio-inspired algorithms in general. From their basic concepts to their applications.
Top1. Introduction And Background
The algorithms play an important role in IT and for several real applications. It delivers an objective is to find answers, which are probably best likened to the runtime environment. In late years, engineers and decision makers are confronted daily with complex problems (NP-hard) that generally involve all sectors. Historically, researchers have attempted to resolve these problems as efficiently as possible. For many years ago, research has been conducted towards the proposition of exact algorithms for polynomial special problems. Afterwards, the appearance of heuristic algorithms allowed generally finding solutions with good quality but often solutions for small instances, so why the need to find new types of algorithms that can lead to a major breakthrough for the practical resolution these problems became paramount.
Today, a huge success was achieved through modelling of organic and natural intelligence resulting in what is called “computational intelligence algorithms”. This class of algorithms (include artificial neural networks, evolutionary computations, collective intelligence, artificial immune systems, human organ systems and fuzzy systems) constitutes a part of meta-heuristic and bio-mimicry areas. They have demonstrated their strength face to different complex issues where they are even attempting to determine the optimal solution from a finite number of existing solutions and offer a high performance results in experimental studies. It is frequently hard to understand why they perform well in a particular context. Another significant advantage is that these algorithms can often be applied without much knowledge about the problem, which makes them very suitable for various applications.
The conception of such algorithm requires the presence of the three characteristics to facilitate the implementation of these algorithms on a new problem:
- •
Choose a representation of possible solutions.
- •
Determine a function to measure the quality of a solution.
- •
Define the operators producing from a current set of solutions a new set of solutions.
It is no overstatement to state that this type of algorithm is everywhere, from design engineering to business planning and from routing of network to travel planning. In all these actions, we strain to reach some goals or optimize something like the quality of performance and the execution time. The delegation of this paper is a very important way to consolidate a number of new algorithms inspired by nature and have been offered in the literature. It is composed from more than 20 algorithms that were unionised in 2 parts classified by the biological source of inspiration of each one of them as illustrated in the next figure 1.
Figure 1. Taxonomy of different bio-inspired algorithms existed in literature
1.2 The Organization of Our Work
We describe the algorithm following the next structure:
It describes the start and the first appearance of this algorithm with references regarding the creator or developer of the algorithm.
- -
The Inspiration Source:
It traces the inspiration phenomenon of the algorithm that can be natural, biological, physical or social.
It summarizes the general cognitive process and the primary goal of the algorithm.
It describes in more detail the various steps of the algorithm and the architecture, in the form of the figure.
It is the structured image of the algorithm which describes its arrangement in terms of the necessary parameters, the input/output, the stopping criterion, and the intercourse between the different calculations performed by the constituents of the algorithm.