The Classification of Log Decay Classes and an Analysis of Their Physical and Chemical Characteristics Based on Artificial Neural Networks and K-Means Clustering

Most existing methods for determining log decay levels normally use variations in log surface characteristics, and the results are subject to human subjectivity, which is uncertain and inaccurate. In order to investigate a novel method for the quantitative determination of log decay levels, we randomly selected log samples from four species (Pinus tabulaeformis, Larix principis-ruprechtii, Betula albosinensis and Quercus aliena var. acuteserrata) with different levels of decay and determined their basic physicochemical characteristics in the laboratory. An artificial neural network (ANN) model was used to predict the hardness values of the log samples with different levels of decay at different moisture contents. The hardness was then used as a clustering factor to quantify the decay levels of the log via K-means clustering analysis. The variations in and correlations between the basic physicochemical factors of the log specimens were investigated between the different decay classes and between the different tree species, and then ANOVA and correlation analysis were used to verify the reliability of the clustering results. The results showed that the prediction of the hardness of the decayed log by the ANN was very effective and that the highly significant variability in the dry matter content, basic density and some basic chemical element contents between the log samples that were classified into different decay grades confirmed the reliability of the clustering results. This study explores an innovative method for the quantitative determination of log decay classes.