TOPAS Model Fitting enables simple, user-friendly digital twin creation
Digital twins are fundamental to advances in automation, making it possible to create virtual replicas of technological systems, test out different scenarios and optimize processes. But if reliable conclusions are to be drawn from the simulations, the digital twin must provide an accurate representation of reality. Precise calibration is key: the right values must be chosen for the underlying model’s parameters to ensure that it reflects the real system’s observed behavior as accurately as possible. Known as parameter identification, this often time-consuming step calls for specialist mathematical expertise, since it involves preprocessing measurement data, adjusting models and choosing the right solutions. In order to lower these barriers, the Steinbeis Enterprise TOPAS Industriemathematik Innovation has developed the TOPAS Model Fitting software.
The software automates substantial parts of the process and makes modern mathematical methods accessible even to users who do not have in-depth mathematical expertise. It focuses on dynamical models, i.e. systems whose behavior is described by time-dependent processes. To do this, TOPAS Model Fitting uses the mathematical language of differential equations, a tool proven in numerous technological applications.
A tool for creating reliable models
TOPAS Model Fitting combines established mathematical methods with a user-friendly interface that enables rapid implementation. Users can import and preprocess data, adjust models and instantly check the accuracy of the
results using visual feedback. Real application scenarios have shown that different parameter identification methods work best for different situations. “So one particularly useful aspect of TOPAS Model Fitting is that it lets you test different approaches and compare them directly against each other. This allows users to create digital twins efficiently and robustly”, says TOPAS CEO Dr.-Ing. Mitja Echim.
Automated parameter identification’s huge potential and diverse areas of application are illustrated by the following use cases, where TOPAS Model Fitting helped to create digital twins that reliably replicate reality. Be it in marine applications, energy systems or agriculture, precise model calibration is critical to a digital twin’s ability to reliably replicate a system’s real-world behavior.
Use case: An autonomous catamaran for marine exploration
This joint University of Bremen and DSI Aerospace Technologie GmbH project is funded by the Federal Ministry for Economic Affairs and Energy.
TOPAS Model Fitting is being used in the iMarEx project, which aims to develop new methods for autonomous navigation of marine robotic systems. As well as taking measurements in the ocean, the goal is for future systems to be capable of responding autonomously to unforeseen events such as pollutant dispersion. A specially equipped catamaran is being employed as a testing platform. A precise mathematical model of the catamaran’s motion behavior is needed to ensure that it can maneuver in a reliable and controlled manner. The challenge is to calibrate the model so that it replicates the real catamaran as accurately as possible. This is where TOPAS Model Fitting comes in. The model parameters are identified automatically based on position and speed measurements and the recorded control commands. This creates a digital twin that enables trajectory planning and can be used for automated real-time control and optimization of the catamaran.
Use case: Optimizing the use of energy storage systems
This joint project of the University of Bremen, the Steinbeis Innovation Center for Optimization, Control and Feedback Control, Q3 ENERGIE GmbH & Co. KG and nD Enerserve GmbH is funded by the Federal Ministry for Economic Affairs and Energy.
TOPAS Model Fitting is also being used in the SmartFarm2 project, which aims to identify opportunities to optimize self-consumption of renewable energy in rural areas and implement them in practice. The focus is on energy storage systems such as batteries or thermal systems like heat pumps, hot water tanks and milk cooling tanks. To enable optimal utilization of these storage systems, an intelligent energy management system requires accurate models of their behavior. These are produced with the aid of TOPAS Model Fitting. The models can be automatically calibrated based on measured operating data, ensuring that they accurately replicate real-world charging and discharging processes. This creates digital twins of the energy storage systems that can be used for planning, simulation and optimized control in real-time energy management systems.
Use case: Autonomous agricultural machinery
This joint project of the University of Bremen, Osnabrück University of Applied Sciences and NeXaT GmbH is funded through a grant from the Special-Purpose Fund of the Federal Government at Landwirtschaftliche Rentenbank.
The NeXaTauto project is developing an innovative farm vehicle distinguished by an extra wide span and four individually controllable track drives, allowing efficient driving on arable land while minimizing soil compaction. One of the project’s aims is to enable autonomous operation of the vehicle through the use of digital twins. Implementing this concept in practice calls for precise modeling of its driving behavior. And this is exactly what TOPAS Model Fitting does.
The models were automatically calibrated using sensor data from real field operations. This created a reliable digital twin that provides a basis for developing autonomous control strategies.
Bridging research and practice
The development of TOPAS Model Fitting is an example of knowledge transfer in action. Industrial mathematics methods are presented in a way that allows enterprises to start using them immediately. This
significantly lowers the entry barriers, especially for businesses with little previous experience of mathematical modeling and parameter identification. It makes the use of digital twins realistic and cost-effective, even for small and medium-sized enterprises.
TOPAS MODEL FITTING
The Steinbeis experts in Bremen use TOPAS Model Fitting software to develop digital twins that are precisely tailored to their customers’ requirements. This powerful solution enables accurate replication of even more complex physical processes. Its modern algorithms provide customers with reliable data that cuts development time and reduces costs while delivering a high level of accuracy and flexibility.
More information: https://topas.tech [1]