Combined Optimization and Virtual Commissioning of Production Systems with a High Volume of Material Flow using Multiscale Network Models (OptiPlant)

在物流密集型生产系统中，机器的时间确定性硬件在环仿真(HILS)很有前途，例如在饮料和包装技术(特殊机械)等领域。每台机器的单独设计和调试都容易出错，这往往需要在现场对控制代码进行昂贵且耗时的更正。将新机器集成到现有生产系统会导致整个系统的停机时间耗费大量成本。此外，对更广泛和快速变化的产品范围的需求不断增加，导致控制系统的复杂性更高。在特种机械领域，机器是根据客户的要求设计和制造的，基于仿真的方法可以加快工程进程，降低成本，并显著降低错误率。然而，基于仿真的方法涉及单个机器以及与实际控制系统相关的物流，目前还没有应用于这一领域。其目的是将通过数控机床的HILS采集的用于测试服务性能和控制验证的结果转移到具有特定要求的饮料技术中。然而，使用HILS描述虚拟机需要实现时间确定性和计算效率高的算法。目前，还没有一种物流模型来保证大量移动物体(饮料技术中每小时大约20.000-50.000瓶)的物流动力学的时间确定性计算。因此，使用多尺度网络模型来模拟HILS内的物质流动是有前景的和新的。这将在这个研究项目中进行调查。此外，还可以利用流量模型对运输系统进行数学吞吐量优化，以求出最优流量。目前，只在实施阶段进行吞吐量优化，而不考虑可行性。此外，在这里，大多数情况下只发现了事件离散的物流模拟，它没有考虑系统布局。在本研究项目中，利用实际控制系统上的多尺度网络模型，在虚拟调试期间，基于流量模型和服务性能，建立了半实物仿真系统之前的数学吞吐量优化方法。

The time-deterministic hardware-in-the-loop simulation (HiLS) of machines in a material flow-intensive production system is promising, as for example in areas like beverage and packaging technology (special machinery). The individual design of each machine and the commissioning is error-prone, which often requires costly and time-consuming corrections of the control code on site. The integration of new machines into existing production systems leads to cost-intensive downtimes of the complete system. Further, the increasing demand for more extensive and rapidly changing product ranges results in a higher complexity of the control system. In the area of special machinery, where machines are designed and manufactured according to customer requirements, simulation-based methods could accelerate the engineering process, reduce costs and lower the error rate significantly. Nevertheless, simulation-based methods concerning individual machines as well as the material flow in connection with the real control system are not applied today in this area. The objective is to transfer the results collected with the HiLS of CNC machines for testing the service performance and control validation to the beverage technology with its specific requirements. However, the description of virtual machines with HiLS the implementation of time-deterministic and computationally efficient algorithms. At present, there is no material flow model that ensures a time-deterministic computation of the material flow dynamics with a large number of moving objects (in beverage technology approximately 20.000 - 50.000 bottles per hour). The use of a multi-scale network model for simulating the material flow within a HiLS is, therefore, promising and new. This will be investigated within this research project. Furhermore, a mathematical throughput optimization of the transport system for finding the optimal flow rate can be performed with the flow model. Currently, throughput optimizations are only conducted disregarding the feasibility in the implementation phase. In addition, here for the most part only event-discrete material flow simulations are found, which do not consider the system layout. In this research project, a mathematical throughput optimization preceding the HiLS and based on the flow model in combination with a service performance is to be developed during the virtual commissioning, using the multi-scale network model on the real control system.