The team behind this project consists of 6 students at Linköping University, all studying the final year on the masters program Mechatronics. Todays rapid changes in demand from the outside world puts a lot of pressure on the automotive industry, especially to reduce pollution. To cope with the pressure, new and inovative solutions are required. The scope of this project is to create a model of the EVAP-system and thereby allow better control and prevent the emission of hazardous gases in the environment.
The EVAP Team
Our team consists of 6 engineering student with different academic backgrounds at Linköpings University
Martin Ling (M)
Project leader
Dylan Patterson (Y)
Quality
Jonna Jämte (M)
Testing
Karl Asklund (M)
Documentation
Lucas Sevelin (M)
Software
Theodor Vallgren (M)
Design
Results
The project is divided into the following results
Hardware
The project have resulted in a data collection hardware that can be used to collect data from the EVAP system.
Model
A model to calculate the mass in the canister was created, based on the data collected from the EVAP system.
Software
All code to collect data from the EVAP system is available on our GitLab.
Hardware setup
The hardware setup consists of a Raspberry Pi 3B+ and an Arduino, the Arduino connects to all the sensors and the Raspberry Pi 3B+ collects the data from the Arduino.
The hardware setup consists of a Raspberry Pi 3B+ and an Arduino, the Arduino connects to all the sensors and the Raspberry Pi 3B+ collects the data from the Arduino. The Rapsberry Pi is set up as a local network hotspot which you can connect to a computer to transmit information and move files. The Arduino is connected to the Raspberry Pi via a USB cable. The Arduino logs data from the sensors at approximately 1.25 Hz and sends the data to the Raspberry Pi. The Raspberry Pi then saves the data to a csv files. The Arduino communicates with the flow and temperature sensors using I2C, and the pressure sensor is read using one of the analog inputs on the arduino.
For most measurements the weight has also been recorded, this was done using a scale with a USB port, which transmitted data to the connected computer at 1 Hz.
Model for mass in canister
A model for the mass in the canister was created, with the goal to have as few sensors and parameters as possible.
The model is based on a physical model, and is extended with a term from a black-box to better capture the dynamics of the system. To find the two variables which has to be found through experimentation, the model also includes a linear regression. Data is used from at least one temperature sensor, one flow meter and a scale to verify the model. In the image above the measured mass change is shown together with the model, which has been divided into four intervals for parameter optimization.
Software
The code for this project is written i C++, Python and Matlab.
All code written follows the Google code standard, and is stored in a private GitLab repository. The Arduino runs on C++, the Raspberry Pi runs on Python and the data analysis is done in Matlab.