Search

A research has been carried out finalized to the definition of a methodology useful for the diagnosis and prediction of the correct evolution state of physical systems. In this paper we present a related model and a specific network topology for the considered problem. In particular, the prediction procedure is based on a 'Self Organizing Map'(SOM) and an 'Error Back-Propagation'(EBP) networks combined to form a hierarchical architecture. The network system has been developed and tested using data furnished by Alenia and consisting in sensorial data (FBG, Fiber Bragg Grating) and multi-format descriptive data regarding evaluation (SB). The obtained results have shown that the developed methodology is a promising tool for the diagnosis activity.

Mixing studies in the horizontal rotating tubular bioreactor (HRTB) were done to explore the influences of the liquid level (HM = 0.05 - 0.08 m) and the distance between the partition walls (DS = 0.02 - 0.07 m) on the mixing performance in the bioreactor described by the "spiral flow" model. The optimised adjustable parameters of the model were correlated with the process parameters of the bioreactor expressed as dimensionless numbers: Reynolds rotation number (ReN) and Reynolds axial flow number (ReD). The polynomial coefficients of the correlations were correlated further with the liquid level in the bioreactor (HM) and the distance between the partition walls (DS). In that way, three modified prediction systems (SC-2A, SC-6A and SC-9A) were established. The analysis based on different criteria selected the prediction system SC-9A as the most suitable to describe the mixing performance of HRTB.

In previous investigations on mixing in a horizontal rotating tubular bioreactor (HRTB) the structured "spiral flow" model was developed which contained four adjustable parameters. In order to incorporate the mixing model in a semifundamental scale-up procedure it was necessary to make a relation between the adjustable model parameters and process parameters of the bioreactor expressed as dimensionless numbers. Mathematical equations which relate adjustable model parameters with dimensionless numbers were developed by non-linear and surface regression methods. These equations were applied to develop the prediction systems for adjustable model parameters. In total, nine systems of equations for the prediction of the adjustable model parameters were established and examined by simulation. Three of them (SC-2, SC-6 and SC-9) were selected as adequate to describe the mixing performance of HRTB in a wide range of process conditions.

Recently, the structured "spiral flow" model was developed to describe the mixing performance of the horizontal rotating tubular bioreactor (HRTB). The adjustable parameters of the model were optimised by Monte-Carlo method. In order to apply the mixing model for a semifundamental scale-up procedure it was necessary to make a relation between the adjustable model parameters and bioreactor process parameters usually expressed as dimensionless numbers. The sets of mathematical equations which relate adjustable model parameters with dimensionless numbers were developed by using non-linear and surface regression methods. These sets of equations were applied for developing the prediction systems of adjustable model parameters. In total, nine systems of equations for the prediction of the adjustable model parameters were established and examined by simulation. Three of them were selected as adequate to describe the mixing performance of HRTB in a wide range of process conditions.

Investigations in the horizontal rotating tubular bioreactor (HRTB) were done to explain the influences of the liquid levels (H_M = 0.05 - 0.08 m) and the distances between the partition walls (D_S = 0.02 - 0.07 m) on the mixing performance of bioreactor. The mixing in HRTB was described by a "spiral flow" model. The adjustable model parameters were optimised by the Monte-Carlo method. The three sets of mathematical equations were selected which related the adjustable model parameters with dimensionless numbers. These prediction systems were modified by incorporating the influence of the liquid level in the bioreactor (H_M) and the distance between partition walls (D_S) as additional parameters. Such modified prediction systems (SC-2A, SC-6A and SC-9A) were used for mixing simulation. Prediction system SC-9A was distinguished as the most suitable to describe the mixing performance of HRTB. This prediction system was incorporated as a part of the scale-up procedure of HRTB, which was based on physiological, hydrodynamic and geometric similarity.