Thomas Bernard

Kinetic modeling of the time course of N-butyryl-homoserine lactone concentration during batch cultivations of Pseudomonas aeruginosa PAO1.

Quorum sensing affects the regulation of more than 300 genes in Pseudomonas aeruginosa, influencing growth, biofilm formation, and the biosynthesis of several products. The quorum sensing regulation mechanisms are mostly described in a qualitative character. Particularly, in this study, the kinetics of N-butyryl-homoserine lactone (C4-HSL) and rhamnolipid formation in P. aeruginosa PAO1 were of interest. In this system, the expression of the rhamnolipid biosynthesis genes rhlAB is directly coupled to the C4-HSL concentration via the rhl system. Batch cultivations in a bioreactor with sunflower oil have been used for these investigations. 3-oxo-dodecanoyl-homoserine lactone (3o-C12-HSL) displayed a lipophilic character and accumulated in the hydrophobic phase. Degradation of C4-HSL has been found to occur in the aqueous supernatant of the culture by yet unknown extracellular mechanisms, and production was found to be proportional to biomass concentration rather than by autoinduction mechanisms. Rhamnolipid production rates, as determined experimentally, were shown to correlate linearly with the concentration of autoinducer C4-HSL. These findings were used to derive a simple model, wherein a putative, extracellular protein with C4-HSL degrading activity was assumed (putative C4-HSL acylase). The model is based on data for catalytic efficiency of HSL-acylases extracted from literature (kcat/Km), experimentally determined basal C4-HSL production rates (qC4 - HSLbasal), and two fitted parameters which describe the formation of the putative acylase and is therefore comparatively simple.

Kinetic modeling of rhamnolipid production by Pseudomonas aeruginosa PAO1 including cell density-dependent regulation

The production of rhamnolipid biosurfactants by Pseudomonas aeruginosa is under complex control of a quorum sensing-dependent regulatory network. Due to a lack of understanding of the kinetics applicable to the process and relevant interrelations of variables, current processes for rhamnolipid production are based on heuristic approaches. To systematically establish a knowledge-based process for rhamnolipid production, a deeper understanding of the time-course and coupling of process variables is required. By combining reaction kinetics, stoichiometry, and experimental data, a process model for rhamnolipid production with P. aeruginosa PAO1 on sunflower oil was developed as a system of coupled ordinary differential equations (ODEs). In addition, cell density-based quorum sensing dynamics were included in the model. The model comprises a total of 36 parameters, 14 of which are yield coefficients and 7 of which are substrate affinity and inhibition constants. Of all 36 parameters, 30 were derived from dedicated experimental results, literature, and databases and 6 of them were used as fitting parameters. The model is able to describe data on biomass growth, substrates, and products obtained from a reference batch process and other validation scenarios. The model presented describes the time-course and interrelation of biomass, relevant substrates, and products on a process level while including a kinetic representation of cell density-dependent regulatory mechanisms.

Nonlinear model predictive control of a glass forming process based on a Finite Element model

A complex glass forming process as an example of a complex, nonlinear distributed parameter system is investigated. The system is modeled by four coupled and strongly nonlinear partial differential equations (Trouton model) which are numerically solved by Finite Element method (FEM). As large steps in the setpoint can hardly be controlled with linear controllers and the the controlled variables can only be measured with dead time we investigate nonlinear model predictive control (NMPC) as control methodology for the forming process. The use of FEM models in NMPC has not attracted much attention so far although it has a huge potential for process optimization. One reason for the absence of FEM models in control engineering is that the calculation of FEM models in many cases are very time consuming. Hence the NMPC scheme is designed in a way that a trade-off between computational effort and control performance is met. A large step in the setpoint of the two controlled variables (diameter and cross section area of the tube) is investigated as a realistic scenario. The impact of important design parameters of the NMPC is worked out. The good performance of the control and its potential for industrial application is shown.

Optimal Management of Regional Water Supply Systems Using a Reduced Finite-Element Groundwater Model (Optimierte Bewirtschaftung von regionalen Wasserversorgungssystemen unter Nutzung eines reduzierten Finite-Elemente Grundwassermodells).

Abstract The sustainable management of water resources and a reliable supply of safe drinking water will play a key role for the human prosperity in the following decades. This paper presents an optimal control approach for the management of water resources in a rapid developing region with critical water shortage. A model of the water allocation system is developed which considers both surface and groundwater resources and the distribution network. A reduced model of the complex spatially distributed 3D Finite Element groundwater model is applied. Hence the optimization problem, which is formulated as a large scale structured non-linear programming problem, can be solved with appropriate computation effort. The performance of the proposed concept is analyzed using realistic optimization scenarios.

Extraction of optimal control patterns in industrial batch processes based on Support Vector Machines

In general industrial processes are complex systems that have to be optimized due to several performance criteria. As control optimization based on the development of physical models is in many cases very time consuming and cost intensive or not even feasible an alternative consists in analyzing historical process data, as lots of it is usually available. Hereby an approach is to use computational intelligent methods as these can identify characteristic control patterns and classify them according to their contribution for process optimization. As data coming from these processes is in general only available as time series a crucial question is how to generate and find these significant features from the measured data concerning a prior defined performance index. Traditional ways of finding these features need plenty of knowledge about the underlying process. In this paper we propose an automated feature extraction and ranking methodology based on Support Vector Machines for Regression (SVR). Furthermore the methodology is used to derive a model of the performance index in terms of the relevant features within the algorithm. The found model can finally be used to find a suitable control pattern of the complex system. The proposed concept is applied to a real world industrial batch process.

Konferenz ML4CPS – Machine Learning for Cyber Physical Systems and Industry 4.0 (29./30.9.2016 in Karlsruhe, Fraunhofer IOSB)

Cyber-physische Systeme (CPS) zeichnen sich durch Anpassungsund Lernfähigkeit aus: Sie analysieren ihre Umgebung und lernen auf Basis ihrer Beobachtungen Muster, Zusammenhänge und prognosefähige Modelle. Typische Anwendungen für solche Systeme sind das Condition Monitoring, Predictive Maintenance, Bildverarbeitung und Diagnose. Als Schlüsseltechnologie für die Entwicklung von CPS gilt das maschinelle Lernen. Die zweite Konferenz ML4CPS – Machine Learning for Cyber Physical Systems and Industry 4.0 findet statt vom 29.–30.9.2016 in Karlsruhe und wird ausgerich-

SAFEWATER – Innovative Tools for the Detection and Mitigation of CBRN Related Contamination Events of Drinking Water

Abstract The safety and or security of drinking water can be threatened by natural disasters, accidents or malevolent attacks. The European FP7 project SAFEWATER aims at developing a comprehensive event detection and event management solution for drinking water security management and mitigation against major deliberate, accidental or natural CBRN related contaminations. New cost-effective C, B, and RN sensors will be developed. An innovative concept with a broad network of low-cost sensors – “domestic sensors” (complementary to a set of sensors in strategic locations) will be developed. A technology platform will be provide which is able to capture and analyze the data collected by the sensors and from other information systems and give a full overview of the crisis to the responders by means of online look-ahead simulations to efficiently manage potential crises. For testing the SAFEWATER solution it will be integrated with on utility-partners’ information systems.

Model based sustainable management of regional water supply systems

Sustainable management of water resources and a safe supply of drinking water will play a key role for the development of the human prosperity in the following decades. This paper presents an optimal control approach for the management of the total water resources in a fast developing region under a critical water shortage. A model of the water allocation system is developed which considers both surface and groundwater resources and the distribution network. The spatially distributed groundwater model is considered as a reduced model of the complex 3D Finite Element model. Hence the optimization problem, which is formulated as large scale structured non-linear programming problem, can be solved in an appropriate computation time. The performance of the proposed concept is demonstrated by close to reality optimization scenarios.