Evaluate the next nanobubble movement with artificial intelligence

Abstract

This research supports solving the problem of how to know the next movement of the nanobubble, two methods will be used to represent the behavior of air nanobubbles in liquids, such as the correlation of data expressed in an equation and the backpropagation in learning its route. To obtain the data, technological tools were used to visualize the movement. of the air nanobubbles on a scale of 10-9 m in diameter. The images of the nanobubbles were obtained using high-power cameras. The algorithms used were, based on the processing of digital images to obtain the positions and follow them, the correlation of data to generate the equation and the neural network to learn their movements. In conclusion, the behavior of nanobubbles in water was identified, generating a specific movement pattern of y = 9E-06x3 - 0.0034x2 + 1.6831x + 299.25; with a correlation of: R² = 0.9976, and it was possible to learn these movements generating the appropriate synaptic weights with 99.8% certainty in their route.