No abstract

A new Modelica library is presented that is used to model safe hierarchical state machines in combination with any Modelica model, e.g., controllers, logical blocks, and physical systems described by differential-algebraic equations. It has been designed to simplify usage, improve safety aspects and to harmonize with the design of the new Modelica_EmbeddedSystems library. Furthermore, new blocks are introduced to define actions in a visual way, and not textually. The library is inspired by Statecharts, Sequential Function Charts, Safe State Machines (SSM) and Mode-Automata. It has been designed so that only small extensions to Modelica 3.1 are needed. The algorithms are sketched that are used to guarantee consistent graphs that give a limited number of event iterations. Furthermore, it is shown how a symbolic verifier can be used to guarantee additional properties of state machines.

No abstract

No abstract

A new strategy for controlling substrate feed in the exponential growth phase of aerated fed-batch fermentations is presented. The challenge in this phase is typically to maximize specific growth rate while avoiding the accumulation of overflow metabolites which can occur at high substrate feed rates. In the new strategy, regular perturbations to the feed rate are applied and the proximity to overflow metabolism is continuously assessed from the frequency spectrum of the dissolved oxygen signal. The power spectral density for the frequency of the external perturbations is used as a control variable in a controller to regulate the substrate feed. The strategy was implemented in an industrial pilot scale fermentation set up and calibrated and verified using an amylase producing Bacillus licheniformis strain. It was shown that a higher biomass yield could be obtained without excessive accumulation of harmful overflow metabolites. The general applicability of the strategy was further demonstrated by implementing the controller in another process using a Bacillus licheniformis strain currently used in industrial production processes. In addition, in this case a higher growth rate and decreased accumulation of overflow metabolites in the exponential growth phase was achieved in comparison to the reference controller.