Colorimetric patterns of wood pellets and their relations with quality and energy parameters
Colorimetric patterns of wood pellets and their relations with quality and energy parameters
复制标题
DOI：
10.1016/j.fuel.2014.07.080
复制
发表时间：
2014-12
期刊：
影响因子：
7.4
通讯作者：
A. Sgarbossa;C. Costa;P. Menesatti;F. Antonucci;F. Pallottino;M. Zanetti;S. Grigolato;R. Cavalli
中科院分区：
文献类型：
--
作者：
A. Sgarbossa;C. Costa;P. Menesatti;F. Antonucci;F. Pallottino;M. Zanetti;S. Grigolato;R. Cavalli
木屑颗粒的颜色主要受其生产所用原料的影响，其成分和特性影响最终产品的质量。由纯木材制成的颗粒颜色浅，灰分含量低，而由木材和树皮或树叶的不同混合物制成的颗粒通常颜色较深，矿物质含量更丰富。本研究旨在验证意大利市场上木质颗粒颜色和质量参数之间的相关性。所有样品都按照欧洲标准（EN）规定的程序进行了分析，用于固体生物燃料的水分，灰分，热值，耐久性，体积和固体密度。图像的采集是用两种技术完成的：CIEL的RGB颜色空间和RGB HSV颜色空间。分别用CIE-L色度B、RGB和HSV进行典型相关分析（CCA），显示所有颜色分量与球团灰分含量的良好相关程度。对两个主成分的主成分分析（总解释方差：64.2%）显示了从良好到中等或低质量参数的明显颜色梯度移动。这种模式通过最轻样本区域的认证颗粒聚集得到证实。Δ E和ΔRGB的计算结果表明，全球样品与木屑样品之间以及高灰分样品与低灰分样品之间具有良好的区分度。然而，当考虑具有相似颜色的样品时，基于其颜色的颗粒质量的视觉可预测性不是那么明显。考虑到执行特定的颜色校准，用于评估颗粒质量的这种方法的工业适用性对于在工作条件下更便宜且更可靠的RGB方法是期望的。
Color of wood pellets is mainly affected by the feedstock material used for their production and which composition and characteristic affect the final product quality. Pellets made from pure wood are light in color and have low ash content, while pellets made from different mixtures of wood and bark or foliage are generally darker and richer in minerals. This study aims to verify the correlation between color and quality parameters of wood pellets available on the Italian market. All the samples were analyzed following the procedures laid down by the European Norms (EN) on solid biofuels for moisture, ash, calorific value, durability, bulk and solid density. The acquisition of the images was done with two techniques: the CIEL∗a∗b∗color space and RGB-HSV color spaces. Canonical correlation analysis (CCA) was performed with CIE-L∗a∗b, RGB and HSV separately showing for all the color components good degree of correlation with ash content of pellets. The PCA analysis on two principal components (total explained variance: 64.2%) showed a clear color gradient moving form good to medium or low quality parameters. This pattern is confirmed by the clustering of certified pellets in the region of lightest samples. The calculation of ΔEand ΔRGB showed a good discrimination level between whole pellets samples and their sawdust, and between ones with high and low ash content. The visual predictability of pellets quality on the basis of their color is however not so sharp when considering samples with similar colors. The industrial applicability of such methods for the evaluation of pellets quality is desirable for RGB methodologies that are less expensive and more reliable in working condition, given that specific color calibration is performed.