Computers and Electronics in Agriculture 69 (2009) 177–184 Contents lists available at ScienceDirect Computers and Electronics in Agriculture journal homepage: www.elsevier.com/locate/compag Real-time, volatile-detection-assisted control for microwave drying Zhenfeng Li a,∗ , G.S. Vijaya Raghavan a , Ning Wang b , Yvan Gariepy a a Department of Bioresource Engineering, McGill University, 21111 Lakeshore Road, Ste-Anne-de-Bellevue, QC, Canada H9X 3V9 b Biosystems and Agricultural Engineering, Division of Agricultural Sciences and Natural Resources, Oklahoma State University, 111 Ag Hall, Stillwater, OK 74078, USA article info Article history: Received 11 November 2008 Received in revised form 4 August 2009 Accepted 5 August 2009 Keywords: Food aroma Microwave drying Fuzzy logic control Electronic nose abstract A real-time, volatile-detection-assisted control system was designed for microwave drying. Detected volatile signals were integrated to a fuzzy logic algorithm to determine the drying temperature. A phase controller was used to automatically and continuously adjust the microwave power. A data acquisition unit with developed program was employed to integrate the entire control. Carrot samples were used in system tests. The results showed that the designed system could successfully achieve the desired temperature, power, and volatiles control and lead to acceptable product’s quality. © 2009 Elsevier B.V. All rights reserved. 1. Introduction Aroma is one of the most important quality aspects of food prod- ucts. This characteristic, to a great extent, determines consumer’s preference and acceptance. However, in most food processing, aroma is often lost, changed, distorted, or even destroyed leading to unacceptable products quality. How to preserve the preferred aroma and limit the development of unacceptable aroma is a crucial task. To fulfill the task, accurate and rapid detection of food volatiles is needed. Unfortunately, most available instruments, either electronic noses or conventional analytical equipments, are time-consuming, expensive, and complex (Li et al., 2006a), which make it almost impossible to detect food volatiles in real-time and online. Consequently, no literature is so far available regarding online detection of food volatiles, and only a few offline measure- ments have been reported (Kompany and Rene, 1993; Krokida and Philippopoulos, 2006; Samuelsson et al., 2006). The new thermal processing technique of microwave drying, developed in recent decades, has the advantages of a higher dry- ing rate, lower energy consumption, and better quality of the dried products (Gerard and Roberts, 2004; Mullin, 1995; Orsat et al., 2006; Raghavan et al., 2005; Sanga et al., 2000; Vadivambal and Jayas, 2007). In the applications of this technique, microwave power and drying temperature were proven to have significant influence on the quality of final products. To optimize microwave drying, researchers have studied various microwave power control ∗ Corresponding author. Tel.: +1 514 398 4400x7632; fax: +1 514 398 8387. E-mail address: zhenfeng.li@mail.mcgill.ca (Z. Li). profiles, including intermittent methods and continuous meth- ods (Venkatachalapathy and Raghavan, 2000; Cheng et al., 2006). Temperature feedback combined with phase control of microwave power achieved a precise and effective control of the drying pro- cess (Li et al., 2006b). Despite numerous attempts at power and temperature control, no volatiles evaluation of food products dur- ing microwave drying process has hitherto been reported. How to improve microwave drying effects based on volatiles detection is still a challenge to microwave researchers. Fuzzy logic control, which was first introduced by Mamdani (1974) and is based on Zadeh’s fuzzy sets theory (Zadeh, 1965), is a precise, quick, and effective control strategy. It is usually applied in processes where a mathematical model is deficient or compli- cated, or if the process is nonlinear or probabilistic. As there is no available description to directly relate volatile signals with human senses, it is better to express these signals as linguistic terms and applied in a fuzzy logic controller. In this study, a microwave drying system with temperature and power control was developed. A fast GC (zNose TM ) was used to detect volatiles emanated from food samples. Detected volatile sig- nals were interpreted with a fuzzy logic controller to determine the drying temperature. The performance of the system was evaluated by drying carrot samples. 2. Materials and methods The developed system consisted of a volatiles detection unit, a microwave drying unit, a power and temperature control unit, and a PC-based data acquisition unit (DAQ) (Fig. 1). The volatiles detection unit included a zNose TM , a two-position pneumatic valve, an air flow installation, a moisture condenser, and a flow meter. 0168-1699/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.compag.2009.08.002