Artificial Neural Network Modeling for Predicting Wood Moisture Content in High Frequency Vacuum Drying Process

Abstract

The moisture content (MC) control is vital in the wood drying process. The study was based on BP (Back Propagation) neural network algorithm to predict the change of wood MC during the drying process of a high frequency vacuum. The data of real-time online measurement were used to construct the model, the drying time, position of measuring point, and internal temperature and pressure of wood as inputs of BP neural network model. The model structure was 4-6-1 and the decision coefficient R2 and Mean squared error (Mse) of the training sample were 0.974 and 0.07355, respectively, indicating that the neural network model had superb generalization ability. Compared with the experimental measurements, the predicted values conformed to the variation law and size of experimental values, and the error was about 2% and the MC prediction error of measurement points along thickness direction was within 2%. Hence, the BP neural network model could successfully simulate and predict the change of wood MC during the high frequency drying process.

Hybrid Signal-Processing Method Based on Neural Network for Prediction of NO3, K, Ca, and Mg Ions in Hydroponic Solutions Using an Array of Ion-Selective Electrodes

“…To solve the interference issue, the ANN, which would possibly compensate for the interfering responses by training the various backgrounds, was employed and improved the performance of the actual test (Figure 5a,c). It supports the theory that ANN would be effective for the non-linear interference by adjusting the relationship as reported in the previous studies [9,15,28,31,42]. In addition, we applied the ANN to predict the Mg concentration because we expected the ANN would extract the signals from the Mg ions through the training with defined background samples.…”

“…The numbers of neurons in the input layer and the output layer were 9 (signals from eight ISEs and one conductivity probe) and 4 (NO 3 , K, Ca, and Mg), respectively. Although ANNs with multiple hidden layers and neurons have a stronger generalization ability, the training time is usually increased and more samples are required to avoid an over-fitting issue [31]. Therefore, for the application of the ANN, the parameters of ANN such as the number of hidden layers or hidden neurons should be determined carefully.…”

“…Three replicate results were then obtained for each layer number and their root mean square errors (RMSEs) were calculated and compared to select the optimal number of hidden layers. Similarly, the number of hidden neurons was tested using ranges of 8 to 16 with an interval of 2 because the neuron number is highly related to the predictability of the ANN model [31]. The model performance was evaluated based on RMSEs of three replicate training results.…”

Drying Stress and Strain of Wood: A Review

“…This model can be applied to wood drying of other tree species under the same conditions. Chai et al input four parameters of drying time, measuring point location, internal temperature and pressure of wood into the BP neural network model, and predicted the MC of wood during the high-frequency/vacuum drying process [117]. Afterwards, Fu inspected the distribution of drying stress from the pith to the bark during the conventional drying process and predicted wood elastic strain and mechano-sorptive creep strain, respectively [118].…”

Artificial Neural Networks To Distinguish Charcoal from Eucalyptus and Native Forests Based on Their Mineral Components

“…ANN has been applied as a modeling tool to overcome various challenges in a number of timber forestry sectors. Some studies have developed ANN models to estimate the wood density (Leite et al, 2016;Demertzis et al, 2017), wood stiffness (García-Iruela et al, 2016), wood strenght (Zanuncio et al, 2017), to assess the surface quality of wood (Hazir and Koc 2018) and to predict the moisture content of wood during drying (Chai et al, 2018;Zanuncio et al, 2016).…”