Evolutionary Algorithms (EAs) are general-purpose search procedures based on the mechanisms of natural selection and population genetics. These algorithms are increasingly used in applications as diverse as architectural design, engineering, factory job scheduling, electronic circuit design, signal processing, network configuration, robotic control, etc. In recent years, a series of optimization techniques inspired by biological systems have been developed. Most of them can be referred as the term of “Evolutionary Algorithms” (EA). Evolutionary algorithms have strongly adaptive search ability by mimicking the evolution process of biological systems. They can usually find the best solution for most complex problems. The solutions maybe are not the theoretical or global optimum, but they are extremely satisfying to the practical applications, generally. Therefore, evolutionary computation has become the most popular optimization tool to researchers and engineers.

• What Are Evolutionary Algorithms?
• Why to Use Evolutionary Algorithms?
• Evolutionary Algorithms Basics

What Are Evolutionary Algorithms?

Evolutionary Algorithms are a branch of Artificial Intelligence (AI) that performs optimization or learning tasks with the ability to evolve. In evolution, each member of the population faces is one of searching for beneficial adaptations to a complicated and changing environment. EAs have three main characteristics:

• Population-based. EAs maintain a group of solutions, called a population, to optimize or learn the problem in a parallel way. The population is a basic principle of the evolutionary process.
• Fitness-oriented. Every solution in a population is called an individual. Every individual has its gene representation, called its code, and performance evaluation called its fitness value. EAs prefer fitter individuals, which is the foundation of the optimization and convergence of the algorithms.
• Variation-driven. Individuals will undergo a number of various operations to mimic genetic gene changes, which is fundamental to search the solution space.

Since the 1960s, many algorithms with population-based, fitness-oriented, and variation-driven properties have been proposed. The relationship between Operations Research (OR), Artificial Intelligence (AI)

Why to Use Evolutionary Algorithms?

There are various complex optimization problems in engineering, economics, biological and sociological areas in a real world application. For small spaces, classical exhaustive methods usually suffice; for larger spaces special heuristic techniques must be employed. Evolutionary algorithms are among such techniques; they are stochastic algorithms whose search methods model some natural phenomena. larger space problems usually cannot be solved by general well-structured optimization methods such as function extreme methods and/or Operations Research methods.

Another important goal of research on evolutionary algorithms is to understand the class of problems for which EAs are most suited, and, in particular, the class of problems on which they outperform other optimization algorithms.

Evolutionary Algorithms Basics?

The following presentation defines basic EA terms and concepts:

• A structure or individual is an encoding solution to some problem. Typically, an individual is represented as a string corresponding to a biological genotype (Search Space). This genotype defines an individual organism when it is expressed (decoded) into a phenotype (Objective Space).
• A genotype is composed of one or more chromosomes or in other words individuals, where each chromosome is composed of separate genes which take on certain values from some genetic alphabet.

Fig. 2 Generalized EA Data Structure and Terminology

• Finally, a given set of chromosomes is termed a population.

basic EA terms are illustrated by Fig.3

Fig. 3 Key EA Components