Artificial neural network for aspect ratio prediction of lignocellulosic micro/nanofibers

Abstract

In this work a wide sample analysis, under similar conditions, has been carried out and a calibration strategy based on a careful selection of input variables combined with sensitivity analysis has enabled us to build accurate neural network models, with high correlation (R > 0.99), for the prediction of the aspect ratio of micro/nanofiber products. The model is based on cellulose content, applied energy, fiber length and diameter of the pre-treated pulps. The number of samples used to generate the neural network model was relatively low, consisting of just 15 samples coming from pine pulps that had undergone thermomechanical, kraft and bleached kraft treatments to produce a significant range of aspect ratio. However, the ANN model, involving 4 inputs and 4 hidden neurons and calibrated on the basis of pine dataset, was accurate and robust enough to predict the aspect ratio of micro/nanofiber materials obtained from other cellulose sources including very different softwood and hardwood species such as Spruce, Eucalyptus and Aspen (R = 0.84). The neural network model was able to capture the nonlinearities involved in the data providing insight about the profile of the aspect ratio achieved with further homogenization during the fibrillation process.

Analysis and Prediction of Electrospun Nanofiber Diameter Based on Artificial Neural Network

“…Once trained on this data, the ANN can predict the properties of PAN nanofibers produced under new process conditions by taking in the process parameters as input and generating a prediction of the resulting fiber properties. By analyzing the output of the ANN, researchers can determine which process parameters have the greatest impact on fiber properties and identify the optimal values for these parameters [17][18][19][20]. Faridi-Majidi et al utilized the ANN method to predict the impact of different needleless electrospinning parameters on the diameter of PAN nanofibers by considering parameters such as polymer solution concentration, applied voltage, and nozzle-to-collector distance [17].…”

Prediction of cellulose micro/nanofiber aspect ratio and yield of nanofibrillation using machine learning techniques

“…Therein, four cellulosic and lignocellulosic pulps, following different treatments, were employed to produce mechanical (L)CMNFs with varying nanofibrillation yields. For the sake of comparison, it should be noted that this experimental dataset, explicitly displayed in Table 1, partially overlaps that of our previous work on the ANN-based prediction of the aspect ratio (Santos et al 2022).…”

“…When it comes to ML techniques in the cellulose/ nanocellulose fields, some work worthy of mention are: Aguado et al (2016), using a support regression vector approach to predict paper strength from morphological characteristics; Pennells et al (2022), with a similar approach on nanopaper; Özkan et al (2019), who evaluated three different algorithms (LR, RF and ANN) in the prediction of mechanical properties of three-component nanocomposite films; Almonti et al (2019), who made use of ANNs to predict fiber length based on the pulp refining process parameters. Recently, another article of ours provided ANN predictions of the aspect ratio of (L)CMNFs, using pulps from aspen, eucalyptus and spruce to validate the performance of the model with other types of biomass (Santos et al 2022). This work widens the scope of said attempts, including the yield towards nanostructured cellulose and other alternative methods.…”

“…A simplified scheme is depicted in Fig. 2 and more detailed descriptions can be found elsewhere (Haykin 1999;Torrents-Barrena and Pellicer 2021;Santos et al 2022).…”

Assessment of machine learning methods predicting the axial vibration frequencies of microbars

“…Santos at al. [51] worked on building an artificial neural network (ANN) model for predicting the aspect ratio of micro/nanofiber items. They successfully predicted the nonlinearities in data by achieving a good correlation ( R > 0.99).…”