Classification of steel using laser-induced breakdown spectroscopy combined with deep belief network.

The identification of steels is a crucial step in the process of recycling and reusing steel waste. Laser-induced breakdown spectroscopy (LIBS) coupled with machine learning is a convenient method to classify the types of materials. LIBS can generate characteristic spectra of various samples as input variable for steel classification in real time. However, the performance of classification model is limited to the complex input due to similar chemical composition in samples and nonlinearity problems between spectral intensities and elemental concentrations. In this study, we developed a method of LIBS coupled with deep belief network (DBN), which is suitable to deal with a nonlinear problem, to classify 13 brands of special steels. The performance of the training and validation sets were used as the standard to optimize the structure of DBN. For different input, such as the intensities of full-spectra signals and characteristic spectra lines, the accuracies of the optimized DBN model in the training, validation, and test set are all over 98%. Moreover, compared with the self-organizing maps, linear discriminant analysis (LDA), k-nearest neighbor (KNN) and back-propagation artificial neural networks (BPANN), the result of the test set showed that the optimized DBN model performed second best (98.46%) in all methods using characteristic spectra lines as input. The test accuracy of the DBN model could reach 100% and the maximum accuracy of other methods ranged from 62.31% to 96.16% using full-spectra signals as input. This study demonstrates that DBN can extract representative feature information from high-dimensional input, and that LIBS coupled with DBN has great potential for steel classification.

Carbon Sequestration Potential in Stands under the Grain for Green Program in Southwest China.

The Grain for Green Program (GGP) is the largest afforestation and reforestation project in China in the early part of this century. To assess carbon sequestration in stands under the GGP in Southwest China, the carbon stocks and their annual changes in the GGP stands in the region were estimated based on the following information: (1) collected data on the annually planted area of each tree species under the GGP in Southwest China from 1999 to 2010; (2) development of empirical growth curves and corresponding carbon estimation models for each species growing in the GPP stands; and (3) parameters associated with the stands such as wood density, biomass expansion factor, carbon fraction and the change rate of soil organic carbon content. Two forest management scenarios were examined: scenario A, with no harvesting, and scenario B, with logging at the customary rotation followed by replanting. The results showed that by the years 2020, 2030, 2040, 2050 and 2060, the expected carbon storage of the GGP stands in Southwest China is 139.58 TgC, 177.50-207.55 TgC, 196.86-259.65 TgC, 240.45-290.62 TgC and 203.22-310.03 TgC (T = 1012), respectively. For the same years, the expected annual change in carbon stocks is 7.96 TgCyr-1, -7.95-5.95 TgCyr-1, -0.10-4.67 TgCyr-1, 4.31-2.24 TgCyr-1 and -0.02-1.75 TgCyr-1, respectively. This indicates that the stands significantly contribute to forest carbon sinks in this region. In 2060, the estimated carbon stocks in the seven major species of GGP stands in Southwest China are 4.16-13.01 TgC for Pinus armandii, 6.30-15.01 TgC for Pinus massoniana, 11.51-13.44 TgC for Cryptomeria fortunei, 15.94-24.13 TgC for Cunninghamia lanceolata, 28.05 TgC for Cupressus spp., 5.32-15.63 TgC for Populus deltoides and 5.87-14.09 TgC for Eucalyptus spp. The carbon stocks in these seven species account for 36.8%-41.4% of the total carbon stocks in all GGP stands over the next 50 years.