System Analysis and Automated Control of Fruit Ripening Processes
Special Symposium - Innovations in Food Technology (LMC Congress)
Flexible Production, PAT & Modelling (Food-3a)
Keywords: banana ripening, model-based control, automation, simulation
The rising competition on the consumer market within the growing European Union creates higher demands on the products in the food industry. The demands of consumers and the industry on new engineering processes engineering developments are a cheap and homogeneous product, the increase and assurance of product quality and thus the reduction of energy and personal costs. A model based process control scheme should lead to moderate and efficient production techniques.
To date the ripening process of bananas has not been thoroughly analyzed or understood. The knowledge regarding this complicated process is basically of qualitative nature. Process control in ripening chambers is based on the longstanding experience of a few people. In 40 % of the cases manual corrections on the process are still necessary.
Reasons for that can be found in the complexity and variability of the banana ripening process. There is a variety in the initial state of ripeness of the input bananas due to different pre-treatment. In addition the course of the industrial banana ripening process is varying due to biological variables, like enzyme activity. The ripening time may change due to logistic prerequisites.
The objective of the work described here is to develop an automation strategy based on a model-driven quantitative estimation of the ripening state of bananas. This innovative automation strategy should enable an advanced process control of the banana ripening process that flexibly adapts to changing conditions and process values.
With the help of a quantitative analysis of the system dynamics the kinetic and metabolic processes in the bananas during ripening are analysed. Indicators for the ripening state and influencing variables can be found. Furthermore the technical process is analysed in terms of measurable and influencing state variables and parameters.
Within the scope of the project a pilot plant was constructed, closely adapted to the industrial ripening process but using additional measuring techniques and a process control system. The reliable and precise on-line measurements of significant process parameters under known conditions are used to estimate the state of the ripening process. In addition the measurements are the prerequisite for a quantitative system analysis and form the basis for the developed process model.
The mathematical model describes the dynamics of the banana ripening process in the form of a system of coupled differential equations. It will be used for the estimation of important state variables which then will be used to control the process. Additionally, the model allows the rapid simul¬ation of different scenarios and thus to find optimal ripening conditions according to the varying initial states, process parameters and logistic requirements.
The varying parameters in the banana ripening process require a high performance process control strategy that flexibly adapts to the variable and imprecisely described process. Adaptive process control with the OLFO-Method has been successfully applied in biotechnology [Luttmann 1985; Witte, 1996 and Frahm et al., 2003] and shall be applied, transferred and further developed to the banana ripening process, as a process in the food industry. It is expected that this model based adaptive and optimizing automation strategy could be transferred to other fruit ripening processes.
[1] Luttmann, R.; Munack, A.; Thoma, M.; “Mathematical Modelling, Parameter Identification and Adaptive Control of Single –Cell Protein Processes in Tower Bioreactors“, Advances in Biochemical Engineering/Biotechnology, vol. 32; (1985), p. 95-205.
[2] Witte, V.C.; “Mathematische Modellierung und adaptive Prozesssteuerung der Kultivierung von Cyathus striatus“, Dissertation, Fortschr.-Ber. VDI, Reihe 17, Nr. 144, VDI-Verlag, Düsseldorf, 1996.
[3] Frahm, B.; Hass, V. C.; Lane, P.; Munack, A.; Märkl, H.; Pörtner, R.; “Fed-batch Kultivierung tierischer Zellen - eine Herausforderung zur adaptiven, modellbasierten Steuerung“, Chemie Ingenieur Technik, vol. 4; (2003), p. 457-460.
Presented Wednesday 19, 16:20 to 16:35, in session Flexible Production, PAT & Modelling (Food-3a).
