Position and mass identification in nanotube mass sensor using neural networks

Carbon nanotube-based nanosensors have shown many promising potential applications in exploring the nano-world. To benefit the maximum capabilities of the nano mass sensor, it must be able to measure both of the augmented mass and its trapping position. Here, this requirement is fulfilled by employing artificial neural networks as an inverse tool. Accordingly, considering the single-walled carbon nanotube mass sensor as a vibrating Love shell, its equivalent characteristics are obtained by matching the shell response with the corresponding molecular dynamics simulation results. Then, the responses (natural frequencies) at different vibration modes are utilized for training a properly selected neural network. Afterward, the ability of the proposed neural networks to predict the mass and the trapping position of the augmented mass is investigated. The results indicated that the presented method can effectively predict the mass and position of an attached particle