Advancement of the automatic simplification of discrete event material flow models by using continuous model elements.

点击翻译按钮获取中文摘要

Models in material flow simulation are of growing size and detail. Because of the increased complexity the simulation runtime is too big for conducting experiments in an efficient manner. One possibility to improve runtime is reducing the model complexity by using model simplification. The price of model simplification however is the creation of behavioral deviation in relation to the original model. A method of model simplification which takes complexity and behavioral deviation into account was developed by the applicant in his dissertation. The developed method simplifies models automatically and controls the complexity and behavioral deviation in order to create models of specific properties.This method shall be significantly improved and extended in this research project. The achievable complexity reduction shall be increased and the deviation shall be reduced when complexity is constant. This goal is to be reached by using hybrid simulation. The simplified discrete event model is coupled with continuous models. These continuous models are representing those model components of the original model on an abstract level which were substituted during simplification. By using hybrid simulation, components get substitutable which were not substitutable previously and the dynamics of substituted components is better approximated. Additionally, a function shall be developed, which calculates changes in the sequence of jobs based on the states of the discrete and the continuous models as they occur in the original model caused by overtaking and priority rules. All methods shall be developed in a matter, such that they are applicable by an algorithm for automatic simplification and are to be implemented prototypical and are thoroughly evaluated.