Motion profile optimization for enhanced energy efficiency in industrial positioning applications : contributions towards global optimal reciprocating point-to-point motions
In recent years, environmental concerns have driven researchers to devise techniques to reduce the energy consumption of industrial machinery. More specifically, electric motor-driven systems (EMDSs) are found to be responsible for approximately 40% to 50% of the total industrial electricity consumption, thereby unveiling a substantial prospect for energy reduction. Among the various EMDSs, position-controlled systems have garnered substantial attention as they play a pivotal role in contemporary automation processes and machinery. Consequently, motion profile or trajectory optimization has emerged as a solution. Given that in numerous industrial contexts, only the start and end points are dictated by the user, optimizing the position function in between those points allows us to optimize the necessary motor torque and its associated energy consumption. Since this approach necessitates merely the modification of position setpoints in the drive settings, it can be readily implemented to virtually any point-to-point motion, thereby offering a very accessible and economical means of enhancement. Nevertheless, efficiently optimizing this motion profile function proves to be challenging. For instance, the current state-of-the-art frequently relies on complex, machine-specific methods to minimize the energy consumption of a motion task, rendering it inapplicable for many machine builders. Moreover, the literature often resorts to piecewise or polynomial position functions, which require a large number of (unbounded) design parameters. The latter not only complicates the pursuit of a computationally efficient technique but also heightens the risk of converging to locally optimal solutions. To this end, the purpose of this doctorate is twofold. On the one hand, a generic, easy-applicable motion profile optimization method is introduced, which utilizes the CAD models that are already available for manufacturing and is applicable to industrial motion control platforms. On the other hand, the utilization of Chebyshev polynomials for the position functions is conceived and developed. This approach results in a minimal number of design parameters and leads to a bounded design space, both of which contribute to enabling the search for a globally optimal solution. The significance of this is underscored by the relative supplementary savings potential of -17.5% compared to state-of-the-art polynomial optimization strategies. Moreover, the feasibility is validated experimentally, resulting in measured savings of up to -63% and -48%, respectively, compared to the standard reference profiles. In addition, a method is presented that re-optimizes the motion profile online based on system properties that are tracked in real-time, thus demonstrating its effectiveness even under varying machine conditions.
Antwerpen : Universiteit Antwerpen, Faculteit Toegepaste Ingenieurswetenschappen , 2023
xvi, 198 p.
Supervisor: Derammelaere, S. [Supervisor]
Supervisor: Cuyt, A. [Supervisor]
Full text (open access)
The publisher created published version Available from 17.11.2024
Research group
Publication type
Publications with a UAntwerp address
External links
Creation 13.11.2023
Last edited 04.03.2024
To cite this reference