Towards Intelligent Optimization of Design Strategies of Cyber-Physical Systems: Measuring Efficacy Through Evolutionary Computations

Designing of effective cyber-physical system (CPS) encompassing different vertical applications solicits different components of design. Most of the components are uncertain and dynamic in nature. They either could be in the form of hardware sensors, optimization process and their scheduling nature. In this chapter, we investigate various levels of CPS formulation driven by machine learning and evolutionary algorithms with their strategic similarities. We argue that how far intelligent optimization in the level designing a CPS should be viable? Thus, suitability of appropriate evolutionary and machine learning algorithms is discussed in the context of different design uncertainty of CPS. The efficacy of auto-adaptive or self-organization principle is also discussed.