Predictive decision models for efficient energy usage in intralogistics by means of a DC voltage link

• Applicants: Professor Dr.-Ing. Wilfried Hofmann; Professor Dr.-Ing. Thorsten Schmidt
• Subject Area: Electrical Energy Systems, Power Management, Power Electronics, Electrical Machines and Drives ; Production Systems, Operations Management, Quality Management and Factory Planning
• Term: from 2020 to 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 430149671

• Final Report Year: 2025

Final Report Abstract

Coupled rotational speed variable multi-motor systems offer significant potential for the recovery of kinetic energy in intralogistics. A spatially distributed DC link is a previously untapped method for harnessing these reserves. Building on the development of prediction models for energy demand, predictive decision models were designed for energy-optimal transport. Using the application case of travel and lift drives of parallel stacker cranes (SC) in high-bay warehouses (HBW), a concept was developed and evaluated for cross-device energy coordination. Measurements were carried out on a multi-axis test bench to validate the power flow models. By measuring stationary operating points and comparing them with simulation results, the reliability of the calculation with respect to predicted power profiles and expected energy savings was assessed. Measurements during dynamic operation provided comparative data for different trajectories. Initially, a numerical approach was developed to determine energy-optimal trajectories, which was extended to account for jerk limitation of the SC. However, this formulation of the optimization problem led to high computation times and partially unstable solutions. These numerical issues prompted the development of a fundamentally new analytical approach. The maximization of energy recuperation between two drives (travel and lift mechanisms) was formulated as a variational problem with constraints, from which the optimality condition was derived in the form of a differential equation. A solver then computed the optimal trajectory between a given start and end point. A virtual HBW environment was created in Python. The combination of a storage and a retrieval operation is referred to as a dual-cycle. The goal of the scheduling optimization was to compose dual-cycles within a batch of orders in such a way that the energy demand for travel is minimized and cross-aisle recuperation is maximized. This combinatorial optimization problem was formulated to be solvable by a solver.

Publications

• „Leistungsflussmodelle und Trajektorienoptimierung für ein Regalbediengerät“. Technischer Bericht 07/2021. Dresden: Technische Universität, Professur Elektrische Maschinen und Antriebe.
  Weihrauch, Markus
  
• „Optimierte Bahnplanung für Regalbediengeräte in Hochregallagern“. In Elektromechanische Antriebssysteme 2021: Beiträge der ETG-Fachtagung 9.-10. November 2021, hrsg.v. VDE ETG, 49–55. ETG-Fachbericht 164. Berlin: VDE.
  Weihrauch, Markus, Karsten Turek & Wilfried Hofmann & Thorsten Schmidt
  
• „Prädiktive Trajektorienoptimierung und Speichersteuerung in Intralogistiksystemen zur Senkung der Netzbelastung“. In ETG-Kongress 2021: Das Gesamtsystem im Fokus der Energiewende 18.-19. Mai 2021, hrsg.v. VDE ETG, 234–39. ETG-Fachbericht 163. Berlin: VDE.
  Weihrauch, Markus; Chris Evers & Wilfried Hofmann
  
• „Leistungsmessung im stationären und dynamischen Betrieb“. Technischer Bericht 03/2022. Dresden: Technische Universität, Professur Elektrische Maschinen und Antriebe.
  Weihrauch, Markus
  
• “Predictive Control of Supercapacitors for Peak Power Reduction in Stacker Cranes in Intralogistics”. In pcim Europe: International Exhibition and Conference for Power Electronics, Intelligent Motion, Renewable Energy and Energy Management, 407–12. Nuremberg: VDE.
  Melkowski, Fabian & Wilfried Hofmann
  
• Predictive Decision Models for an Energy Efficient Operation of Stacker Cranes in a High-Bay Warehouse. 2024 Winter Simulation Conference (WSC), 1634-1644. IEEE.
  Zöllner, Rico; Handrich, Konrad; Schulze, Frank & Schmidt, Thorsten
  
• „Optimierungsmodelle in der dispositiven Ebene“. Technischer Bericht 01/2024. Dresden: Technische Universität, Professur für Technische Logistik.
  Turek, Karsten
  
• „Variational Approach to Trajectory Optimization w.r.t. Energy Recuperation for Stacker Cranes“. Logistics Journal: Proceedings, Nr. 20.
  Zöllner, Rico & Ella Jannasch & Thorsten Schmidt