Open Access

In this paper, we evaluate the application of Bayesian Optimization (BO) to discrete event simulation (DES) models. In a first step, we create a simple model, for which we know the optimal set of parameter values in advance. We implement the model in SimPy, a framework for DES written in Python. We then interpret the simulation model as a black box function subject to optimization. We show that it is possible to find the optimal set of parameter values using the open source library GPyOpt. To enhance our evaluation, we create a second and more complex model. To better handle the complexity of the model, and to add a visual component, we build the second model in Simio, a commercial off-the-shelf simulation modeling tool. To apply BO to a model in Simio, we use the Simio API to write an extension for optimization plug-ins. This extension encapsulates the logic of the BO algorithm, which we deployed as a web service in the cloud. The fact that simulation models are black box functions with regard to their behavior and the influence of their input parameters makes them an apparent candidate for Bayesian Optimization (BO). Simulation models are multivariable and stochastic, and their behavior is to a large extent unpredictable. In particular, we do not know for sure which input parameters to adjust to maximize (or minimize) the model’s outcome. In addition, the complex models can take a substantial amount of time to run. Bayesian Optimization is a sequential and self-learning algorithm to optimize black box functions similar to as we find them in simulation models: they contain a set of parameters for which we want to identify the optimal set, they are expensive to evaluate, and they exhibit stochastic noise. BO has proven to efficiently optimize black box functions from varius disciplines. Among those, and most notably, it is successfully applied in machine learning algorithms to optimize hyperparameters.

Semi-solid metal alloys, as used in industrial thixoforming, have a special microstructure of globular grains suspended in a liquid metal matrix. The complex rheological properties are strongly influenced by the local solid fraction, particle shape, particle size and state of agglomeration. It was analysed how the microstructure develops in dependence of the shear rate and cooling rate during the solidification and it was observed that the average particle size increased with increasing shear rate and decreasing cooling rate. In order to account for those phenomena, the rate of crystal growth and the relationship between average particle diameter and viscosity was modelled by applying the Sherwood two-film model for the mass transport. The dependence of the viscosity from the particle size were modelled with a modified Krieger–Dougherty model. Based on the rheological and microstructural observations an evaluation method was elaborated that allows for the construction of objective master curves that are independent of the particle growth during the experimentation. The isothermal experiments for the characterisation of the rheological behaviour consisted of step-change of shear-rate and yield-stress experiments. From the experimental data the steady-state flow curves could be determined as well as the time-dependent relaxation of the shear stress after a change of shear rate. The steady-state rheological behaviour was found to be shear thinning. Nevertheless, immediately after a shear-rate change an overshoot was observed that resulted from a short-time shear-thickening behaviour. The yield stress was found to strongly depend on the microstructure and the degree of agglomeration of the solid phase. With increasing rest time the yield stress was increasing strongly, because of the agglomeration of the solid particles. Based on the step-change of shear-rate experiments a single-phase flow has been developed that consists of a modified Herschel–Bulkley approach and accounts for the thixotropic as well as for the yield-stress behaviour of the alloys.

Semi-solid metal alloys, as used in thixoforming, have a special microstructure of globular grains suspended in a liquid metal matrix. The material under investigation is a tin–lead alloy (Sn–15% Pb) which exhibits a similar microstructure as aluminum alloys. The experiments were performed with concentric cylinder rheometers. Initially, the liquid alloy is cooled down to the semi-solid range under constant shearing and then kept under isothermal conditions for further experimentation. The microstructure is characterized in dependence of the shearing time. The rheological techniques consisted of step change of shear rate and shear stress ramp experiments for different solid fractions (40–50%). Based on the experimental data a single phase model has been derived, where the semi-solid alloy is regarded as a homogeneous material with thixotropic properties and the microstructure is characterized by a structural parameter. The model consists of two parts: the equation of state, including a finite yield stress, and a rate equation for the structural parameter. The model equations are employed into numerical software and used for the simulation of a characteristic thixocasting process. The results are compared to real experiments.

Background: Midwifery care in Germany is a legal right for every woman (SGB V). Midwives work employed or freelance in hospitals or in community services, providing maternal care from pregnancy until the end of breastfeeding (Sayn-Wittgenstein 2007). Increasingly, a shortage of midwifery care has been observed, forcing hospitals to understaff or to close their birth units, leaving women and their families without care (Sander et al. 2018). At the same time, birth rates are rising, thus leading to an increasing demand of midwifery care (Destatis 2019). As off today there is no central register for midwives across Germany’s 16 states. Therefor the exact number of registered midwives as well as the scope of services provided by midwives are not known (Niedersächsisches Landesgesundheitsamt 2019). Given the present situation, it seems to be imperative to establish effective midwifery workforce planning. The aim of this poster is to identify already existing health workforce planning approaches and to determine the extent to which those can be transferred to the German system of midwifery care. Methods: Health workforce planning approaches, already being used on a national and international level, have been analysed, focusing their applicability to midwifery services in Germany. Results: Particular elements of the workforce planning approaches already being used in Germany for registered physicians seem to be adoptable. However, they need to be adjusted and enhanced to ensure the characteristics of midwifery in the German public health services. Internationally used approaches are not readily transferable due to systemic differences in health care systems. Conclusions: The development of new specific workforce and service planning approaches for midwifery care in Germany is crucial to meet present and future needs of women and their families during the childbirth period.