Modelling the microbial quality and safety of foods.

摘要: Quality and safety of foods are often influenced by the presence growth microorganisms. Microorganisms in can be divided into two groups: pathogenic organisms, causing illness, microorganisms that not harmful to health, but spoil a product. Presence organisms should avoided as much possible. Growth spoilage is allowed certain extent. largely influences food safety, whereas generally determines shelf life product, Food quality assumed both organisms. A method predict microbial presented. The construction product from its ingredients simulated, following recipe. processing heuristics combined with models developed predictive microbiology. Parameter values foods, such water activity acidity, for decay used prediction. these parameters collected present databases. If required information lacking, methods make reliable guesses developed. Furthermore, expert knowledge production development applied improve Shelf calculated function fluctuating temperature time. Several distribution chains simulated assess influence on quality. described implemented computerised decision support system. Mathematical growth, system, given more attention to. Microbiological examined modelling bacterial bacterium; Lactobacillus curvatus. modelled assessing behaviour Listeria monocytogenes. Models describe effect temperature, variables curvatus validated. (lag time, specific rate, maximum population density) data at various temperature-acidity combinations. acidity monitored several constant values. set up fitted data. same procedure model temperature. For lag density, multiplied acidity. equipped suggesting cease growing minimal or maximal controlling (Temperature, pH, w ). Evidence presented existence lower upper boundary value growth. rate an extended model. based two, earlier models, one mentioned It combinatory effects between do exist. Therefore, result resulting sets measured earlier, determined. new outside range where developed, validate found very well able range. Bacterial monocytogenes activity, which equations first equation predicts sub optimal values, temperatures activities, second throughout entire validated statistically comparing predictions reported literature. Finally, system identification bacteria key lactic acid bacteria. carried out steps. step distinguishes classes tree general tests. distinction species within class basis biochemical fermentation patterns. During group classification, probabilities test failure used. These answer. also select most probable answer case inconclusive result. determined experts potential users identification, similarity indices all class. "learn" different sessions step, improving speed accuracy. Because versatile way up, it easily expanded keys other Structured changes foods. This thesis shows even complex problems prediction through combination models. Model systems giving insight processes importance determination safety.