Scania is a world-leading provider of transport solutions. Together with our partners and customers we are driving the shift towards a sustainable transport system. In 2017, we delivered 82 500 trucks, 8 300 buses as well as 8 500 industrial and marine engines to our customers. Net sales totalled nearly SEK 120 billion, of which about 20 percent were services-related. Founded in 1891, Scania now operates in more than 100 countries and employs some 49 000 people. Research and development are concentrated in Sweden, with branches in Brazil and India. Production takes place in Europe, Latin America and Asia, with regional production centres in Africa, Asia and Eurasia. Scania is part of Traton Group. For more information visit www.scania.com.
Scania is now undergoing a transformation from being a supplier of trucks, buses and engines to a supplier of complete and sustainable transport solutions.
Thesis project at Scania is an excellent way of making contacts for your future working life. Many of our current employees started their career with a thesis project.
Scania Eu6 diesel engines are equipped with an advanced aftertreatment system with the purpose of reducing soot and NOx emissions. In order to control and monitor the function of the system, a number of sensors and models are used. In the engine management system (EMS), several models run for the various components in the system. Modelling the engine and the aftertreatment components is a difficult task, involving thermodynamics, chemical kinetics etc. A complete physical model is often not even possible to create. An example of a difficult field of models are the exhaust emissions of the engine.
Future emission legislations require better and more accurate models. To meet these requirements, a new approach of creating the model is needed.
Several successful attempts has been made using different types of neural networks to produce models and predictions in various fields. As of today, a few frameworks exists for deep learning and development of neural networks. Particularly the TensorFlow framework from Google looks promising.Project Scope and Goal
The goal of the master thesis is to create a neural network using a deep learning framework, preferably using TensorFlow, to create a model of the emissions from a diesel engine. The project involves studying several types of networks and comparing their performance in terms of accuracy and in computation time. Also, to study the work of others in this field.
A second goal is to reduce the network in such a way that it can be converted/generated into C-code to run online in an engine management system.
The project can be divided into two parts.
Engine out NOx-model
Create a neural network model that predicts the NOx-rate as a function of engine sensor values, for example engine speed, inlet temperature and pressure, and combustion phasing. Using measurements from engine test bed and logs from trucks.
The model should have a few percent accuracy and should be able to run on an ordinary PC in at least real-time.
Engine soot model
Creating a soot model is more difficult than creating a NOx-model, both in terms of physics involved but also in terms of lack of available data for training the network. Only test bed measurements are available. The approach should otherwise be similar to the NOx-model above.
The model should have a few percent accuracy and should be able to run on an ordinary PC in at least real-time.Candidate profile
2 students are required.
Contact persons and supervisors
Henrik Flemmer, Manager Engine Aftertreatment , +46(0)855351928
Armin Hosseini, Development Engineer, +46 (08) 55351968 Mail: seyed.hossein_at_scania.com
Enclose CV, personal letter, and grades. Selection will be done continuously.
Ange nedan vart du önskar arbeta och glöm inte bort att ange din e-postadress!