Abstract
In recent years, the advent of cyber-physical systems (CPS) has been witnessed. These systems are generally composed of collaborative computational elements (sensors, hardware and software) in order to control and manage systems or structures, preventing catastrophe and / or accidents from happening, ensuring people's lives and preventing economic losses.This research project presents the innovative proposal for detecting damage in structures according to the concept of cyber-physical systems, rooted in classic procedures from structural health monitoring (SHM), with a processing core and decision-making from the artificial intelligence (AI). In this way, the traditional SHM evolved into the cyber physical structural health monitoring system.From artificial intelligence, the artificial immune system (AIS) - applied through the negative selection algorithm (NSA) - and neural networks - applied through the autoencoders neural networks (DPAC-autoencoders) will be used. Each artificial intelligent method will have independent processing, acting in the decision making regarding structural integrity. Thus, it will also be the objective of this research to compare these two intelligent methods, observing efficiency, precision and processing time.The cyber-physical monitoring methodology will be applied to the experimental data obtained from the overpass Z-24 located in Switzerland. From the acquisition and processing of the obtained signals, artificial intelligence will be elaborated and applied acting in the processing nucleus of the structural state.Therefore, this project presents the evolution of the traditional structural integrity monitoring system to a cyber-physical system model with decision making based on artificial intelligence. Thus, early detection of damage will be more effective and autonomous.Therefore, this project presents the evolution of the traditional structural integrity monitoring system to a cyber-physical system model with decision making based on artificial intelligence. Thus, early damage detection will be more effective and autonomous, representing innovation in structural damage detection techniques.
|