Among the processes that make up automotive painting, drying/curing in ovens is characterized as one of the main steps to guarantee the final quality of the painting. At this stage, the ideal operating conditions for the stoves must be rigorously maintained, especially the temperature profile of body parts. Obtaining a representative model allows prediction and better control of process behavior. With the great advancement of technology, new strategies for model identification are being developed, among which stands out the use of artificial neural networks (ANN) for the identification and control of non-linear dynamic processes. This study presents a methodology for the development and implementation of a model using ANN to represent the process that occurs in an automotive paint drying/curing oven used during the electrodeposition painting step (Elpo). For the prediction of future values of temperatures at measurement positions on the bodywork (parts of the bodywork), a global neural model was developed, composed of a set of ANNs from the 13 zones that form the greenhouse. After evaluating the performance of the global neural model, it was verified that the model was able to predict, in a satisfactory way, the temperature value of the bodywork parts throughout the entire process, which was evidenced in the obtained values of coefficients of R2 fits and mean absolute percentage errors (MAPE). Thus, it is concluded that the methodology proposed in this work can be applied to develop innovative strategies for modeling and predicting the conditions of the drying and curing process of automotive paint in greenhouses, and that the global neural model obtained can be used as a Soft Sensor based on the application of the ANN technique.
