Detecting coating thickness distribution in paint coating involves converting real input parameters into input parameters for model of quasi-3D spray figure in trained neural network

摘要

The method involves generating a model of a quasi-3D spray figure, entering fixed model parameters, converting additional real parameters into model input parameters in neural network, generating spray images in the model for points on a movement path depending on sprayer movement data, producing copies of spray images on a virtual surface depending on a scaling factor, integrating the copies for the coating and outputting the distribution. The method involves generating a model of a quasi-3D spray figure, entering specific fixed model parameters, entering additional real parameters to a trained neural network to convert them into model input parameters, generating spray images in the model for selected points on a movement path depending on sprayer movement data held as a polygonal series, producing individual copies of spray images on a virtual surface depending on a scaling factor, integrating the copies for the entire coating and outputting the thickness distribution.