Optimizing Real-World Problems with Differential Evolution and Particle Swarm Optimization


Considering today’s competitive markets, the time for the development of technical sys- tems should be as short as possible. Simultaneously, systems become very complex due to manufacturing at the limits of technical feasibility. Furthermore, a high functionality is desired with low power consumption, small size and high reliability. However, the devel- opment of a system, e.g. in circuit design, is normally an iterative trial-and-error process. After the basic structure of the design has been established, parameters are repeatedly adjusted by the designer and the system is simulated to verify its performance. Due to the complex interactions of parameters, this is a time-consuming process. The success depends on the experience and knowledge of the designer. From this follows that the goal of a short time-to-market becomes hard to achieve. Therefore, the need arises for fast, easy-to-use optimization algorithms which can be integrated in design processes. Thus, the time-consuming iterative trial-and-error process can be substituted by a structured automated approach for finding optimal parameter settings.