73C-21 |
Modeling and optimization of constant retort temperature (CRT) thermal processing of foods using coupled neural networks and genetic algorithms |
C. CHEN and H. S. Ramaswamy. Dept. of Food Science, McGill Univ., Macdonald Campus, 21111 Lakeshore Rd., Sainte-Anne-de-Bellevue, QC H9V 3V9, Canada Computer simulation and systematic optimization routines are essential for efficient food thermal process development. The artificial intelligent technologies including artificial neural networks (ANNs) and genetic algorithms (GAs), have been found the more advantages than conventional methods to deal with the system modeling and optimization problems. The objective of this study was to evaluate the application of ANNs and GAs for modeling and optimization of constant retort temperature thermal processing. ANN prediction models were developed for processing time, quality retention, surface cooking value, final temperature difference at can center, and lethality ratio. The processing conditions as inputs for ANN models were as follows: retort temperature (110-140oC), thermal diffusivity (1.1-2.14*10-7m2/s), volume of can (164-655cm3), ratio of height to diameter of can (0.2-1.8), the total desired lethality value (4.5-10.5 min) at can center and quality kinetic parameters: temperature dependence (15-40oC) and decimal destruction time (150-300 min). The data for training and testing ANN were obtained by a finite difference computer simulation program. ANN model linked Genetic Algorithms (GA) were employed for searching the optimal quality retention and corresponding retort temperature, and for investigating the effects of main processing factors. The results indicated that ANN based prediction models (correlation coefficients > 0.98; relative errors < 3%) could successfully describe the various outputs of CRT thermal processing. The coupled ANN-GA models, verified by computer simulation, could be effectively used for optimization of CRT thermal processing. The main processing conditions and their interactions in the order of their importance with respect to the optimal quality retention and corresponding retort temperature were determined using ANNs and GAs. These results confirmed that it is reliable for ANNs and GAs to be used for modeling and optimization of food thermal processing, demonstrating the potentials for developing advanced model based control systems for food industries using AI technologies.
Session 73C, Food Engineering: Transport Processes and Kinetics
|