In today’s globalized automotive market, the ability to quickly and flexibly adapt to the technical approval requirements of different countries has become increasingly important. This challenge is even more critical for high-performance sports cars, where the entire design is mainly driven by performance rather than efficiency and day by day conventional use. In this context, the thesis presents the development of an advanced simulation tool for the complete electric drive (e-drive) of a high-performance electric vehicle. The aim is to improve the accuracy of inverter efficiency evaluations to enable the prediction of inverter behaviour over homologation driving cycles already during early design stages, reducing the necessity of early bench testing and significantly saving development time and costs. Also, a highly accurate simulation model is essential not only to reliably estimate inverter losses but also to support the conceptual development of new control strategies and modulation techniques. A fundamental element of the e-drive is the Surface-Mounted Permanent Magnet Synchronous Motor (SMPM), whose electromechanical architecture and conventional electromagnetic modelling equations are initially reviewed. Subsequently, a more advanced dynamic model is introduced, which accounts for spatial harmonics and magnetic saturation effects by leveraging Finite Element Analysis (FEA) flux maps. An innovative implementation of this model is proposed, using Look-Up Tables and offline data processing in MATLAB to reconstruct the FEA-based inductance, thus avoiding numerical instability issues during real-time PLECS simulations. The conventional and proposed dynamic models are then compared and validated through experimental data collected from no-load tests conducted at the test bench, providing a first verification of the improvements introduced. The second key component analysed is the Two-Level Three-Phase Voltage Source Inverter. Basic concepts and modulation schemes are briefly presented before discussing its implementation in PLECS and validation through experimental data. The validation is performed by matching simulated and measured waveforms at a given operating point, defined in terms of electric motor’s torque and speed, and therefore corresponding d- and q-axis current references. Moreover, to further enhance simulation precision, a thermal model of the inverter’s power modules is developed and integrated into the simulation environment. This model is based on test bench measurements obtained through Double Pulse Tests (DPTs) and allows accurate evaluation of inverter losses under various operating conditions. Finally, the complete E-Drive Simulation Tool is presented, including its workflow and use. A novel method is proposed to unify the simulation environments (PLECS and MATLAB) using JSON-RPC communication, enabling multiple simulations to be executed in parallel and allowing full vehicle mission profiles analysis while maximising CPU utilization. The methods, results, and conclusions of this work are detailed in the following chapters. Throughout the thesis, the modelling choices made will be clearly presented and justified, and the corresponding results will be illustrated. Moreover, simulation outcomes will be systematically compared against experimental data obtained from test bench characterizations, validating the developed models and methodologies.

A COMPREHENSIVE CIRCUITAL MODEL OF INVERTER AND ELECTRICAL MACHINE FOR ACCURATE LOSSES ESTIMATION IN ELECTRIC VEHICLES POWERTRAIN

BIGLIARDI, ANDREA
2024/2025

Abstract

In today’s globalized automotive market, the ability to quickly and flexibly adapt to the technical approval requirements of different countries has become increasingly important. This challenge is even more critical for high-performance sports cars, where the entire design is mainly driven by performance rather than efficiency and day by day conventional use. In this context, the thesis presents the development of an advanced simulation tool for the complete electric drive (e-drive) of a high-performance electric vehicle. The aim is to improve the accuracy of inverter efficiency evaluations to enable the prediction of inverter behaviour over homologation driving cycles already during early design stages, reducing the necessity of early bench testing and significantly saving development time and costs. Also, a highly accurate simulation model is essential not only to reliably estimate inverter losses but also to support the conceptual development of new control strategies and modulation techniques. A fundamental element of the e-drive is the Surface-Mounted Permanent Magnet Synchronous Motor (SMPM), whose electromechanical architecture and conventional electromagnetic modelling equations are initially reviewed. Subsequently, a more advanced dynamic model is introduced, which accounts for spatial harmonics and magnetic saturation effects by leveraging Finite Element Analysis (FEA) flux maps. An innovative implementation of this model is proposed, using Look-Up Tables and offline data processing in MATLAB to reconstruct the FEA-based inductance, thus avoiding numerical instability issues during real-time PLECS simulations. The conventional and proposed dynamic models are then compared and validated through experimental data collected from no-load tests conducted at the test bench, providing a first verification of the improvements introduced. The second key component analysed is the Two-Level Three-Phase Voltage Source Inverter. Basic concepts and modulation schemes are briefly presented before discussing its implementation in PLECS and validation through experimental data. The validation is performed by matching simulated and measured waveforms at a given operating point, defined in terms of electric motor’s torque and speed, and therefore corresponding d- and q-axis current references. Moreover, to further enhance simulation precision, a thermal model of the inverter’s power modules is developed and integrated into the simulation environment. This model is based on test bench measurements obtained through Double Pulse Tests (DPTs) and allows accurate evaluation of inverter losses under various operating conditions. Finally, the complete E-Drive Simulation Tool is presented, including its workflow and use. A novel method is proposed to unify the simulation environments (PLECS and MATLAB) using JSON-RPC communication, enabling multiple simulations to be executed in parallel and allowing full vehicle mission profiles analysis while maximising CPU utilization. The methods, results, and conclusions of this work are detailed in the following chapters. Throughout the thesis, the modelling choices made will be clearly presented and justified, and the corresponding results will be illustrated. Moreover, simulation outcomes will be systematically compared against experimental data obtained from test bench characterizations, validating the developed models and methodologies.
2024
Power Electronics
Electric Drives
Electric Vehicles
Inverter
Automotive
File in questo prodotto:
File Dimensione Formato  
Bigliardi.Andrea.pdf

accesso aperto

Dimensione 13.98 MB
Formato Adobe PDF
13.98 MB Adobe PDF Visualizza/Apri

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14251/3229