The term Industry 4.0 was created as part of a strategic document by the German government aimed at integrating the internet of things and services into the production environment (i.e., the formation of cyber-physical systems). One of the basic tools of artificial intelligence is search and optimization. The task of optimization is to search and find the best possible solution from the solution space. When defining product form and structure during product design, structure optimization is among the essential stages. Believing that the product must be optimized, modern CAD systems have begun to implement structure optimization as an option during product design. Based on the previously mentioned in this chapter, heuristic optimization methods will be presented with special reference to genetic algorithms, particle swarm optimization, ant colony optimization, and their application in structure optimization. Some software that offer the option of optimization, either as a separate package or as an option of a CAD system, are given at the end of this work.
TopIntroduction
Optimization at the moment is one of the most popular areas of research, especially in the fields of mechanical engineering, civil engineering, aeronautics, mining, nuclear engineering and some other engineering disciplines. The question is what is the causative agent, that is, what made this area of research so interesting and research intensive. There are many reasons for this, but one of the main reasons that has its place when is talked about any research today is the market. It looks for a variety, that is, in various variants, functionally, easily and in a short time available, and especially that it is cheap, whether it is a product or a service. These demands are directly related to the demand on which Industry 4.0 try to answer. The market directly places these demands on a designer who designs mechanical or other systems in such a way that it forces him to think about systems for example with the lowest possible mass, as this will lead him to the lowest possible costs, or the maximum rigidity of the system, which will lead to maximum possible functionality or maximum possible productivity of the designed system.
Finding a minimum or maximum is in domain of optimization. Optimization itself is a separate scientific discipline that is highly applicable to a variety of engineering problems, and its rapid development is achieved in World War II.Although optimization had its abrupt rise in the Second World War and a little later, the first works related to the engineering optimization problem can be found at the beginning of the twentieth century. The first scientist to deal with problems involving structural optimization was the Australian engineer Antony Michell, who published in his paper “The Limits of the Economics of Material in Frame-Structures” back in 1904, so this paper is also considered crucial by working in this scientific field-discipline. This paper also aimed at satisfying the market, as can be seen from the title, but resources were very limited, which made it impossible for the optimization of resilient systems to evolve significantly at that time. By resources in this case we mean both the equipment used for the calculations and the very development of optimization methods.
In the Second World War and then in the 1950s and 1960s, the development of optimization discipline based on linear and nonlinear programming methods. These methods required a great deal of numerical computation, which significantly slowed down the optimization process, as systems were not currently developed that could process data relatively quickly (Arora, 1989;Kirsch, 1981).
The development of microcomputers in the 1970s, as well as their widespread use in the 1980s, renewed interest in optimization in general.In the period up to the 1990s, the methods mentioned above were mostly used, which can be summarized under the common name of the mathematical programming method. They were deterministic in nature and can be regarded as classical methods.
In the early 1990s, the so-called optimization modern or contemporary methods appeared. Most of these methods are based on specific characteristics and behaviors of biological systems, molecular systems, neurobiological systems, swarms and insect systems, etc., and are mostly stochastic in nature. Although the rapid growth of these methods began in the 1990s, they can be traced back to the 1970s, as is the case with the genetic algorithm method proposed by the American scientist John Henry Holland (Holland, 1975). This method is based on the principles of natural genetics and natural selection. An ant colony optimization method based on the cooperative behavior of real ant colonies was proposed by Italian scientists Alberto Colorni, Marco Dorigo and Vittorio Maniezzo in 1991. The particle swarm optimization method is based on the behavior of sets of living things such as swarms of insects, flocks of birds and fish, etc. This method was proposed in 1995 by James Kennedy and R. C. Eberhart. Although this method was introduced relatively late, tens of thousands of papers in various areas of optimization have been published. In addition to the mentioned methods, there are also annealing simulation methods, tabu search, neural network based methods and fuzzy logic, etc. (Singiresu, 2009;Colorni, Dorigo & Maniezzo, 1991)