Development of Advance Emission Control using Machine Learning in SiL Environment

Job title: Development of Advance Emission Control using Machine Learning in SiL Environment

Introduction

Thesis work is an excellent way to get closer to Scania and build relationships for the future. Many of today's employees began their Scania career with their degree project.

Background

Scania develops advanced powertrain and exhaust aftertreatment systems to achieve high efficiency and low emissions across a wide range of operating conditions. Selective Catalytic Reduction (SCR) control requires accurate dosing decisions, robust emission prediction and effective use of signals available in the vehicle control system.

Objective

The objective of this master thesis is to develop and evaluate a methodology for AI-based, closed-loop SCR control. The controller shall use machine learning to determine dosing and support emission control over a representative operating cycle.

The work will use measured test data and additional data generated in a Software-in-the-Loop (SiL) environment. Inputs may include engine operating point, load, speed, temperatures, flow conditions and other control-system variables. The AI model will be benchmarked against an existing reference SCR model in SiL using a virtual truck powertrain named VTAB (Virtual Truck and Bus).

The project will establish a traceable workflow for data preparation, model training, validation and control integration. It will assess whether a data-driven model can reduce calibration effort and enable adaptable control while maintaining robust dosing and emission performance in a real powertrain.

Job description

The assignment includes the following activities:

• Literature study on SCR control, machine learning and AI-based control methods
• Collect, structure and quality-assure large volumes of test and SiL data
• Select relevant control-system signals and define model inputs, outputs and constraints
• Develop and train an AI model for SCR dosing and emission control
• Integrate and evaluate the model in a virtual truck powertrain SiL environment
• Compare performance with an existing reference SCR control model
• Analyse robustness, accuracy, limitations, calibration effort and implementation potential
• Document the methodology and present the results in a master thesis report

Education/program/focus

Machine learning, artificial intelligence, control engineering or applied mathematics.

Understanding of mathematical models and physical functions used in control systems

Experience in handling, processing and analysing large data sets. We value curiosity, structured problem solving, initiative and the ability to work both independently and collaboratively.

Contact persons and supervisors

Kim Petersson

Requisition ID:

33683

Number of Openings:

1.0

Part-time / Full-time:

Full-time

Permanent / Temporary:

Temporary

Country/Region:

SE

Location(s):

Södertälje, SE, 151 38

Required Travel:

0%

Workplace:

On-site