Evolutionary algorithms are inspired by Darwinian evolution by mimicking the mechanisms of natural selection. The most well-known type, namely genetic algorithms (GAs), uses populations of potential solutions represented as chromosomes, subjecting them to selection, crossover, and mutation operations. Tailored for specific problems and characteristics, they tend to be today's much murmured research. This chapter proposes the different EAs and their systematic workflow. EA, the process, begins with the initialization of a population of potential solutions. These solutions undergo evaluation based on a predefined fitness function. Crossover and mutation operations then generate new candidate solutions. This iterative process continues until convergence or a predefined stopping criterion is met. The performance of EA depends on parameter settings. Tuning parameters with crossover and mutation rates, and population size. EA's have been rooted in the elegance of nature's optimization strategies. They have evolved into indispensable tools for solving complex problems across domains. It has been changed as a valuable asset for many researchers. The overall perspective of EA in various ways is discussed in the chapter.
Top1. Introduction
A key idea in many exciting disciplines, including computer science, mathematics, engineering, economics, and more, is optimization. Optimization is the process of selecting an optimal solution from a range of workable options for a given problem. The goal is to maximize or reduce an objective function under a constraint, which is a particular criterion. Some terms are related to the goal function drives the optimization process. The quantity to be optimized is defined by the objective function, which is a mathematical expression. It could stand for any measurable criterion that needs to be met, such as maximizing profit, minimizing cost, minimizing distance, or any other. The parameters or variables that decision-makers can manipulate to affect how an optimization issue turns out are known as decision variables. Changing the values of these variables gives better results (Tantithamthavorn et al., 2019).
Limitations or requirements that must be met for a solution to be found valid are known as constraints. They serve as a representation of the choice variables' actual constraints. Constraints can be either inequality-related or equality-related. The feasible region is the set of all possible combinations of decision variable values that satisfy the constraints. The optimal solution must lie within this feasible region. There are local and global optima that influence the optimization performances. A local optimum is a solution that is optimal in its immediate neighborhood. A global optimum is the best solution across the entire feasible region (Jian et al., 2024).
1.1. Types of Optimization
• Linear Optimization: Involves linear relationships in the objective function and constraints.
• Nonlinear Optimization: Involves nonlinear relationships in the objective function and/or constraints.
• Integer Optimization: Requires decision variables to take on integer values.
• Combinatorial Optimization: Involves discrete decisions, often in the context of a finite set of possible solutions (Guo et al., 2023).
1.2. Optimization Techniques
• Analytical Methods-Closed-form mathematical solutions for simple problems.
• Numerical Methods- Iterative algorithms for complex problems.
• Heuristic Methods-Approximate algorithms that may not guarantee optimality but find good solutions in a reasonable time (Zhang et al., 2023).
• Nature Inspired Methods-forages the natural behavior exhibited in environment.
2. Motivation
The motivation behind Evolutionary Algorithms (EAs) lies in the quest to develop robust and effective optimization techniques inspired by the remarkable efficiency of natural evolution. The problems across diverse domains grew in complexity, traditional optimization methods faced challenges in providing global solutions or handling high-dimensional, non-linear, and multi-modal landscapes. Evolutionary Algorithms is drawing its inspiration from Darwinian principles introduce a paradigm shift by treating potential solutions as evolving entities within a population (Ahmad et al., 2022). The driving force behind EAs is the realization that the iterative processes of selection, crossover, and mutation, reminiscent of natural selection, can effectively explore solution spaces, adapt to dynamic environments, and discover optimal or near-optimal solutions. This adaptability coupled with the ability to handle complex, real-world problems with minimal assumptions, positions Evolutionary Algorithms as versatile tools for optimization challenges in fields ranging from engineering and computer science to finance and beyond. The motivation lies in harnessing nature's optimization strategies to address the intricacies of modern problem-solving landscapes (Zhou et al., 2022).