Artificial neural networks in variable process control: application in particleboard manufacture

Abstract

Artificial neural networks are an efficient tool for modelling production control processes using data from the actual production as well as simulated or design of experiments data. In this study two artificial neural networks were combined with the control process charts and it was checked whether the data obtained by the networks were valid for variable process control in particleboard manufacture.The networks made it possible to obtain the mean and standard deviation of the internal bond strength of the particleboard within acceptable margins using known data of thickness, density, moisture content, swelling and absorption. The networks obtained met the acceptance criteria for test values from non-standard test methods, as well as the criteria for using these values in statistical process control.Key words: Artificial neural networks (ANN), statistical process control (SPC), internal bond strength, wood based panels. ResumenRedes neuronales artificiales en el control de procesos por variables: aplicación en la fabricación de tableros de partículas Las redes neuronales artificiales son una herramienta eficaz para el modelado de los procesos de control de producción, tanto partiendo de datos de la propia producción como de datos simulados o procedentes de diseños de experimentos. En este estudio se han combinado dos redes neuronales artificiales con los gráficos de control de procesos y se ha comprobado si los datos obtenidos con ellas eran válidos para el control de producción por variables en la fabricación de tableros de partículas.Las redes han permitido obtener valores de la media y la desviación típica de la cohesión interna del tablero de partículas dentro de unos márgenes aceptables a partir de datos conocidos de espesor, densidad, contenido de humedad, hinchazón y absorción. Las redes obtenidas han cumplido con los requisitos de aceptación de valores de ensayo por métodos alternativos al normalizado y con los requisitos impuestos para su utilización en el control estadístico de procesos.Palabras clave: Redes neuronales artificiales (RNA), control estadístico de procesos (CEP), resistencia a la tracción interna, tableros derivados de madera.

Using Artificial Neural Networks to Model the Surface Roughness of Massive Wooden Edge-Glued Panels Made of Scotch Pine (Pinus sylvestris L.) in a Machining Process with Computer Numerical Control

“…Artificial neural networks (ANNs) have been widely used in wood science, such as in the recognition of wood species (Esteban et al 2009;Khalid et al 2008), the drying process of wood (Wu and Avramidis 2006;Ceylan 2008), the prediction of some mechanical properties in wood and wood products (Mansfield et al 2007;Fernández et al 2012;Tiryaki and Aydin 2014), the optimization of process parameters in the manufacturing process of wood products (Cook and Whittaker 1993;Cook et al 2000), the classification of wood and wood veneer defects (Drake and Packianather 1998;Nordmark 2002;Packianather and Drake 2005;Castellani and Rowlands 2008;Kurdthongmee 2008), the calculation of wood thermal conductivity (Avramidis and Iliadis 2005), the analysis of moisture in wood (Zhang et al 2006;Avramidis and Wu 2007), and the prediction of fracture toughness of wood (Samarasinghe et al 2007).…”

Prediction of mechanical properties of wood fiber insulation boards as a function of machine and process parameters by random forest

“…Moreover, the performance of the RF algorithm is comparable or even better than models primarily based on ANNs used by other authors for predictive modeling of forest product characteristics (André et al 2008, Barnes 2001, Cook et al 1997, Cook et al 2000, Esteban et al 2009a, Esteban et al 2009b, Esteban Luis et al 2011, Gupta et al 2007. In this context, it is also worth to mention that the authors of the mentioned references used considerably smaller data sets with only a few hundred samples.…”

Prediction of plywood bonding quality using an artificial neural network

“…(2)x (Cook and Chiu 1997;Malinov et al 2001;García Fernández et al 2008a;Esteban et al 2009a), considering an error of no more than 15% in a production process as acceptable and errors of 20-30% as unacceptable (Cook and Chiu 1997;Malinov et al 2001).…”