Teoresi logo

Welcome Teoresi MedTech

Le nostre tesi
logo politecnico di torino test

Politecnico di Torino

Engineering

Master's degree

Autore

Soroush Arab

2020

Algorithm analysis for an automotive adaptive front light systems

logo politecnico di torino test

Politecnico di Torino

Engineering

Master's degree

Autore

Soroush Arab

C languageCAN-Bus
Relatori Accademici

Luciano Lavagno


Abstract

One of the crucial components of a car is headlights.  Lights are essential to safety on the road.  They allow the driver to see the road ahead, even at night or in obscure-light conditions.  They also ensure that others can see the vehicle.  Generally, front vehicle lights point only forward, and drivers turn them on/off and switch on the high-beam headlamps manually.  In this thesis, a model-based design approach was used to analyse and propose a prototype for an adaptive front lighting system that can automatically turn off high-beams and control the intensity of low-beams when detecting on coming traffic.  Furthermore, having a high visibility range provides safe driving conditions.  The fundamental concept behind adaptive headlights is the capability of swivelling from side to side and up and down.  This way, the driver can see throughout the turn. The  study  was  mainly  carried  out  on  STMicroelectronics  development  boards,  which were collected to develop and simulate an automotive adaptive front light system for development purposes. These are labelled as “AutoDevKit boards” in this outline, which includes the “AEKD-AFL001”, which represents logic and driving hardware for proto-typing,  testing,  and  development  plans.   The  assortment  involves  two  stepper  motor control boards, and a four-channel LED driver board, a command board with MCU, a connector board with a FAN switchboard, and an integrated connector board for wiring arrangement.  The configuration of boards for developing firmware has been done within“SPC5-STUDIO” as codes generator, expeditious source configurator, and Eclipse development  environment  for  “SPC5  MCUs”.   Debugging  and  programming  were  done through the “SPC5-UDESTK” as an interface to run and test the firmware.  The “Value CAN 4-2” was used to monitor and transmit on CAN networks to generate hardware simulations  for  analysis.   Additionally,  the  Vehicle  Spy  software  was  used  as  a  single tool for diagnostics data acquisition and testing, while the Adaptive Front Light System CAN bus monitoring by creating a graphical simulation panel.  The first step was 2to study each board for understanding, cognition design, functionality, communication protocols, and limitations.  Next, drivers and communication protocols such as SPI and CAN  were  configured  according  to  the  aspects  of  the  algorithm.   Further,  databases were created from the values of sensors and radars for transmitting and retransmitting the signal to simulate commands, and graphical panels were designed to demonstrate the results. This algorithm was designed to turn on daytime running lamp automatically whenever the  switch  mode  is  “Drive”,  and  it  is  dark  enough  to  require  lights.   The  automatic high-beam and low-beam lights help resolve two problems.  They turn off or reduce the intensity of bright lights as required to avoid blinding the occupants of oncoming cars. They  also  turn  the  high-beams  on  when  the  street  ahead  is  not  visible  enough;  they assist  drivers  who  do  not  use  them  by  turning  on  automatically  in  order  to  provide broader illumination.  Similar to automated lights, the system is driver-selectable.  The system uses a forward-facing radar which recognises lights—not just expected lights butalso the tail lights of vehicles ahead and streetlights or other light sources that indicate the driver is in an area that does not require high beams.  As soon as excessive lights are identified, the system turns the high-beam lights off, then switches them back on once the light fades.  The low-beam front lights closely follow the rotation of the steering wheel; they change their position according to the direction of the steering wheel. The Follow-Me-Home feature keeps headlights active for a few minutes after the engine is turned off and the doors are closed in order to illuminate the path.  The algorithm provides an alternative solution to predict and illuminate passengers’ intended paths. At the end of this research, one can understand that the vehicle headlight should not be a passive tool to switch on/ off.  It should be able to adapt to the environment to increase  safety  in  low  visibility  conditions.   I  have  illustrated  the  adaptability  of  the headlight for numerous duties; Allowing drivers to use high beams without glaring any other  driver  on  the  road.   Allowing  drivers  to  see  better  in  curvature,  and  allowing better illumination of road lanes, sidewalks and dividers.  It will be done through curve-adaptive  lights.   consequently,  provide  essential  safety  for  passengers  by  configurable and adaptive “Follow-Me-Home” .The main focus of this thesis was analysing an algorithm that operated with step mo-tors.  Technology is continuously moving forward,  and one of the aims in the area of lighting systems is to minimise moving parts.  One alternative is using adaptive lights in the form of LED pixels, which could reduce power consumption and motor delay.  An-other modification to improve the system could be done through faster communication protocol such as Automotive Ethernet or CAN – FD. The data over CAN is transmitted in frames.  The receiving nodes send an acknowledgement flag when they receive the frame.  As this acknowledgement is sent into the transmitted frame, the sender receives an in-frame response after successful transmission.  CAN – FD solves this challenge by applying two separate frames to transmit the real data and the acknowledgement data

Obiettivo Tesi

Analisi e sviluppo di algoritmi di gestione luci frontali esterne di autovettura tramite piattaforma Adaptive Font Light system di ST-Microelectronics.

Metodologia di ricerca

Sistemi intelligenti di gestione luci esterne autovetture basati su informazioni provenienti da rete vettura.

Sviluppi futuri

Integrazione con sistema rete vettura reale per verifica di nuove funzionalità. Integrazione con ambiente di simulazione per gestire in automatico la compilazione del FW e la verifica in processor-in-the-loop.