Cognitive Sensor Technology for Structural Health Monitoring
Alexander Serov · Procedia Structural Integrity · 2017
Current paper presents artificial neural network architecture based on Cognitive Sensor technology which may be used for development of intelligent SHM systems. Dynamic Artificial Neural Network (DANN) has time dependent structure which is defined by experience of processing input data streams. Advantages of proposed model include ability to learn both linear and non-linear patterns on the basis of processing data streams. Evolution of DANN architecture includes stage of autonomous growth of subnets developed by separate Cognitive Sensors and stage of cooperative growth of resulting network. Example application of proposed cognitive technology for solution of Structural Health Monitoring problems is discussed.