Giardiello, Giovanni (2020) Internal combustion engine base calibration: computer aided tools and methodologies for the experimental effort reduction. [Tesi di dottorato]


Download (9MB) | Preview
[error in script] [error in script]
Item Type: Tesi di dottorato
Resource language: English
Title: Internal combustion engine base calibration: computer aided tools and methodologies for the experimental effort reduction
Date: 18 March 2020
Number of Pages: 125
Institution: Università degli Studi di Napoli Federico II
Department: Ingegneria Industriale
Dottorato: Ingegneria industriale
Ciclo di dottorato: 32
Coordinatore del Corso di dottorato:
Gimelli, AlfredoUNSPECIFIED
Muccillo, MassimilianoUNSPECIFIED
de Nola, FrancescoUNSPECIFIED
Date: 18 March 2020
Number of Pages: 125
Keywords: Engine base calibration, Electronic control Unit, ECU function validation, Neural Networks, 1D simulation, multivariable-multiobjective validation.
Settori scientifico-disciplinari del MIUR: Area 09 - Ingegneria industriale e dell'informazione > ING-IND/08 - Macchine a fluido
Date Deposited: 02 Apr 2020 08:32
Last Modified: 05 Nov 2021 12:52

Collection description

Over the last decades, internal combustion engines have undergone a continuous evolution to achieve better performance, lower pollutant emissions and reduced fuel consumption. The pursuit of these often-conflicting goals involved changes in engine architecture in order to carry out advanced management strategies. Therefore, Variable Valve Actuation, Exhaust Gas Recirculation, Gasoline Direct Injection, turbocharging and powertrain hybridization have found wide application in the automotive field. However, the effective management of such a complex system is due to the contemporaneous development of the on-board Engine Electronic Control Unit. In fact, the additional degrees of freedom available for the engine regulation highly increased the complexity of engine control and management, resulting in a very expensive and long calibration process. Indeed, the functions of the engine control units are calibrated trying to reduce the error between the quantity obtained from the ECU algorithms and the experimental quantity in a wide range of engine working conditions. To this aim, extensive experimental campaigns are carried out on the test bench, in which thousands of operating conditions are analyzed, resulting in high costs for the realization of the process. Figure - A schematizes the traditional base engine calibration process. Figure - A - main stages of the traditional base engine calibration process (on the left), and the proposed methodologies (on the center and on the right) With the aim to overcome the above issues, two methodologies have been investigated. The purpose of the proposed methodologies is to reduce the number of the experimental bench tests without reducing the performance of the control unit functions. A first effective methodology is based on the use of Neural Networks (NN) to overcome some critical issues concerning the calibration of engine control parameters. NN are adopted to provide a detailed engine data sheet starting from a reduced number of experimental data. The potential of the proposed methodology has been verified using this detailed data set as input to a specific Computer Aided Calibration algorithm developed in this work and evaluating the achievable calibration performance. In particular, the calibration performance has been assessed with reference to a specific ECU function. The research clearly demonstrates the effectiveness of the proposed approach since the calibration performance falls within acceptable limits even after a 60% cut of the experimental data usually acquired for calibration purposes, highlighting how the use of neural networks can allow a significant reduction of the experimental effort along with its related times and costs. A second methodology based on the adoption of 1D thermo-fluid dynamic analysis is proposed. In particular, starting from a complete experimental set of data currently used for the base calibration of a reference spark ignition engine, a novel procedure based on vector optimization approach is used to reliably calibrate a 1D engine model starting from a reduced experimental dataset. Once validated, the engine model is used as a virtual test bench to reproduce the experimental campaign numerically, thus obtaining a detailed and complete dataset exploitable for calibration purposes, here called numerical or virtual dataset. The potential of the proposed methodology has been verified by comparing the experimental and virtual dataset. The research clearly demonstrates the effectiveness of the proposed approach since the mean errors are comparable with the measurement errors. Therefore, the methodology shows promising results concerning the use of numerical dataset obtained from reliable 1D CFD engine models as input to computer aided calibration software. In this way, a significant cut to the experimental campaign required for calibration purposes is achieved, with their related times and costs.


Downloads per month over past year

Actions (login required)

View Item View Item