Premium
Computationally efficient modeling for assessing the energy efficiency of electric drivetrains using convex formulations
Author(s) -
Vandeputte Matthias,
De Keyser Arne,
Crevecoeur Guillaume
Publication year - 2017
Publication title -
international journal of numerical modelling: electronic networks, devices and fields
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.249
H-Index - 30
eISSN - 1099-1204
pISSN - 0894-3370
DOI - 10.1002/jnm.2275
Subject(s) - computer science , convex optimization , control theory (sociology) , drivetrain , control engineering , optimal control , regular polygon , torque , mathematical optimization , engineering , mathematics , control (management) , physics , geometry , artificial intelligence , thermodynamics
High‐fidelity models capturing the dynamical behavior can be engaged for the analysis of complex mechatronic systems. Determining the optimal control parameters and design characteristics of such systems necessitates solving multiple interconnected models acting on their respective physical domains and time scales. In this paper, high‐fidelity physics‐based models are constructed for several electrical subsystems. Loss mechanisms in the various components are inferred because these are key when performing optimal design and control in terms of energy‐efficient conversion from power source to actuation. The complexity of the analyzed models is then reduced by introducing convex approximations for the occurring dissipation during power transfers, allowing abstracting the complicated dynamic behavior into a tractable convex formulation, specifically suited for time‐efficient numerical simulation. The effectiveness of the strategy is demonstrated on a case study originating from the field of all‐electric vehicles, embodying a series interconnection of a battery stack, a buck‐boost converter, a voltage source inverter, and an asynchronous electric motor. Results show that the dynamic simulation of the proposed system, composed of multiple time scales, can be reliably computed using the composed convex mappings, hereby reducing the computational time approximately by a factor 461, compromising only 1.8% accuracy regarding energy consumption assessment. The introduced convex formulation can therefore constitute the foundation for optimal control and design of complex mechatronic drives.