Haugen J E
Norwegian Food Research Institute, As.
Adv Exp Med Biol. 2001;488:43-57. doi: 10.1007/978-1-4615-1247-9_4.
Gas sensor array technology combined with multivariate data processing methods as artificial neural network has been demonstrated to have a promising potential for rapid non-destructive analysis of food quality. It may be applicable in quality control of raw material, food processing or products. This technique cannot completely replace reference methods like the use of sensory panels as the technique requires a frequent calibration against some valid reference method. As with all new techniques there remain some basic problems to be solved concerning sample handling and instrumental performance. The emerging research activity in the development of chemical sensors including hardware and software combined with applied research makes it realistic to expect applications with this technique implemented on-line in the food industry in near future. In particular, promising applications on meat seem to be within the field of spoilage, off-flavor, sensory analysis and fermentation processes.
气体传感器阵列技术与多元数据处理方法（如人工神经网络）相结合，已被证明在食品质量快速无损分析方面具有广阔的应用前景。它可应用于原材料、食品加工或产品的质量控制。由于该技术需要经常与一些有效的参考方法进行校准，因此不能完全取代如使用感官评定小组等参考方法。与所有新技术一样，在样品处理和仪器性能方面仍存在一些基本问题有待解决。化学传感器（包括硬件和软件）开发方面新兴的研究活动以及应用研究，使得在不久的将来在食品工业中在线应用该技术成为现实。特别是，在肉类方面，该技术在变质、异味、感官分析和发酵过程领域似乎有着广阔的应用前景。
