The project has been developed during a period of stage of six months in Robert Bosch GmbH Branch in Italy in Turin. The area of interest was the automotive sector and in particular the validation of the dataset and the software installed on the vehicle engine control units. The aim of the thesis is: developing a prognostic algorithm that is able to predict a component malfunction during its lifetime starting from data already available. The activity concerned data collection and its analysis, the study of the component and the development of a machine learning algorithm. The obtained results are positive but at the same time are far from what we wanted because the data were collected for another purpose. This in uenced our analysis because we did not have all the necessary information. At the moment we are only able to classify if a measure shows anomalies, but we cannot say anything about its future behavior.
Il progetto è stato sviluppato durante un periodo di stage di 6 mesi in Bosch branch in Italy a Torino. L'area di interesse era il settore automotive e, in particolare, la validazione di software e dataset installati sulle centraline motore. Lo scopo della tesi è: sviluppare un'algoritmo di prognostica che è in grado di predire il comportamento futuro di un componente partendo da dati già disponibili. L'attività riguardava la raccolta dati e la loro analisi, lo studio del componente e lo sviluppo di un algoritmo di machine learning. I risultati ottenuti sono positivi ma, allo stesso tempo, sono lontani da quello che volevamo perchè i dati sono stati raccolti per un altro progetto. Questo ha influenzato la nostra analisi perchè non avevamo tutte le informazioni necessarie. Al momento siamo in grado di classificare se una misura mostra anomalie, ma non possiamo dire nulla riguardo il suo comportamento futuro.
Prognostic modeling of a metering unit in heavy Diesel engines
CROCE, LORENZO
2017/2018
Abstract
The project has been developed during a period of stage of six months in Robert Bosch GmbH Branch in Italy in Turin. The area of interest was the automotive sector and in particular the validation of the dataset and the software installed on the vehicle engine control units. The aim of the thesis is: developing a prognostic algorithm that is able to predict a component malfunction during its lifetime starting from data already available. The activity concerned data collection and its analysis, the study of the component and the development of a machine learning algorithm. The obtained results are positive but at the same time are far from what we wanted because the data were collected for another purpose. This in uenced our analysis because we did not have all the necessary information. At the moment we are only able to classify if a measure shows anomalies, but we cannot say anything about its future behavior.File | Dimensione | Formato | |
---|---|---|---|
Tesina_Croce_Lorenzo.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
3.12 MB
Formato
Adobe PDF
|
3.12 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/147438