The drivability of a vehicle describes all the perceptions of a driver, including perceived safety and comfort. The latter one is mostly affected by road irregularities, involving uncomfortable variations in vertical and longitudinal accelerations of sprung and unsprung masses. While the vertical dynamics has been already fully studied, the longitudinal dynamics has not been completely explored yet. However, its acceleration's variations cannot be neglected due to their high values and higher frequency than the vertical ones. The goal of this study is to develop and implement in real time a non-linear model predictive control (NMPC) based on road preview, whose aim is to attenuate the variations of longitudinal acceleration by changing the motor torque requested by the driver. In the end, the purpose is also to analyse the controller's behaviour under various conditions. In particular, three prediction models will be developed and implemented on a fully electric vehicle model with different powertrains, such as 4 in-wheel motors, 4 on-board motors and 2 on-board motors with open differential, which includes a realistic tire model for ride comfort simulation. The NMPCs’ performances will be evaluated and compared by considering four different key performance indicators (KPIs) along multiple manoeuvres, e.g. uneven road profile and step profile, at different speeds and required motor torques. Further analyses will be conducted by also changing some powertrain’s parameters of the 4 on-board configuration, such as motor time constant, inertia moments, shaft stiffness and angular backlash, to understand how the performance can be affected. In the end, the controllers' capability to be implemented in real time will be proved by using dSPACE MicroAutoBox II. The thesis' project has been developed in collaboration with the Centre of Automotive Engineering at the University of Surrey (UK). A colleague from "Politecnico di Torino", Silvio Santoro, and I were the engineers responsible for the development of this innovative research project. This activity has been supervised by Prof. Aldo Sorniotti and Prof. Umberto Montanaro, from "University of Surrey", Prof. Antonella Ferrara, from "Università degli Studi di Pavia", Prof. Alessandro Vigliani, from "Politecnico di Torino", and Pietro Stano, PhD student at "University of Surrey".
Controllo di una motorizzazione elettrica posizionata sul telaio per la compensazione delle oscillazioni di accelerazione longitudinale causate dalle irregolarità del terreno. Le irregolarità stradali possono influire sul comfort del veicolo, causando variazioni di accelerazione longitudinale e verticale di masse sospese e non sospese. Mentre la dinamica verticale è già stata completamente studiata, la dinamica longitudinale non è stata ancora completamente esplorata. Tuttavia, queste variazioni di accelerazione non possono essere trascurate a causa dei loro elevati valori e della frequenza più elevata rispetto a quelle verticali. L'obiettivo di questo studio è quello di capire, confrontando diversi propulsori, quali parametri possono influenzare le prestazioni di un NMPC, il cui scopo è quello di attenuare le variazioni di accelerazione longitudinale. Tre principali modelli di veicoli completamente elettrici con diversi propulsori vengono presi in considerazione: 4 motori sulle ruote, 4 motori sul telaio e 2 motori sul telaio con differenziali aperti. I parametri principali, che potrebbero influenzare le prestazioni del NMPC, sono costanti di tempo del motore e, in caso di motori sul telaio, la rigidezza dell'albero e il gioco equivalente dal motore alla ruota. Inoltre, la dinamica torsionale degli alberi può causare disturbi nell'accelerazione longitudinale.
On-board electric powertrain control for the compensation of the longitudinal acceleration oscillations caused by road irregularities
LAZZARINI, DAVIDE
2021/2022
Abstract
The drivability of a vehicle describes all the perceptions of a driver, including perceived safety and comfort. The latter one is mostly affected by road irregularities, involving uncomfortable variations in vertical and longitudinal accelerations of sprung and unsprung masses. While the vertical dynamics has been already fully studied, the longitudinal dynamics has not been completely explored yet. However, its acceleration's variations cannot be neglected due to their high values and higher frequency than the vertical ones. The goal of this study is to develop and implement in real time a non-linear model predictive control (NMPC) based on road preview, whose aim is to attenuate the variations of longitudinal acceleration by changing the motor torque requested by the driver. In the end, the purpose is also to analyse the controller's behaviour under various conditions. In particular, three prediction models will be developed and implemented on a fully electric vehicle model with different powertrains, such as 4 in-wheel motors, 4 on-board motors and 2 on-board motors with open differential, which includes a realistic tire model for ride comfort simulation. The NMPCs’ performances will be evaluated and compared by considering four different key performance indicators (KPIs) along multiple manoeuvres, e.g. uneven road profile and step profile, at different speeds and required motor torques. Further analyses will be conducted by also changing some powertrain’s parameters of the 4 on-board configuration, such as motor time constant, inertia moments, shaft stiffness and angular backlash, to understand how the performance can be affected. In the end, the controllers' capability to be implemented in real time will be proved by using dSPACE MicroAutoBox II. The thesis' project has been developed in collaboration with the Centre of Automotive Engineering at the University of Surrey (UK). A colleague from "Politecnico di Torino", Silvio Santoro, and I were the engineers responsible for the development of this innovative research project. This activity has been supervised by Prof. Aldo Sorniotti and Prof. Umberto Montanaro, from "University of Surrey", Prof. Antonella Ferrara, from "Università degli Studi di Pavia", Prof. Alessandro Vigliani, from "Politecnico di Torino", and Pietro Stano, PhD student at "University of Surrey".È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/14867