Leveraging Biologically Inspired Models for Cyber–Physical Systems Analysis

Abstract

Cyber–physical systems (CPS) are systems composed of distributed sensors, physical actuators, and controlling computers that are interconnected through a computer network. Notable examples include: electric utility “smart grids” that can sense and optimize power distribution, transportation systems, and healthcare and medical systems. As CPS complexity increases, system-level trade studies become more challenging due to the combined interaction of the computing, network communications, and physical sensor and actuator elements. Although fundamentally different, complex CPS and biological systems share common attributes that suggest the use of similar modeling approaches. This research investigated if CPS multiobjective optimization methods using mathematical models of a CPS that incorporate biologically inspired CPS system performance objective functions are better than those methods that do not use biologically inspired objective functions. The research methodology employed computer modeling and simulation. A multiobjective optimization study, using a commercially available genetic algorithm-based optimizer, was conducted on four slightly different versions of a dynamic model of a simple distributed robotic sensor system to assess the effect of incorporating a biological mathematical model on the quality of the Pareto fronts generated by the optimizer. An improvement in Pareto front quality was found after incorporating a biological mathematical model.