2/18: Meeting with Brian
we will be notified by Brian
mid semester demo?
2/18: Meeting with Phil and Noah about Vehicle Dynamics
Noah is not mad about our ideas
Phil was not there
2/19 Meeting with Prof Neuman:
keeping it simple is very important.
no need for a kalman filter just yet - use transfer functions and whatnot. Can also take a weighted average of sensor measurements instead of kalman filter for starters. Only use kalman filters if you know what the noise is
desired radius minus actual radius is the control variable
State space seems to make more sense - how do we characterise the car as a system though?
may need two transfer functions - one for each wheel
2/20 Meeting with prof Kantor
current sensing for wheel spin - if there is a sharp drop in current, cut wheel torque
look into angular rate for steering
keep it simple
try multiple algorithms
2/23 Samantha and Pronoy meeting
Pronoy got batch read working on the IMU and refined the UM6 communication library so it better takes advantage of object oriented programming
Samantha began work on interfacing the Arduinos with the DACs
We are hoping to finish these tasks by thursday
2/25 meeting with Brian Flores
we need to work on our website
make sure videos are taken of your working project for the midsemester demo just in case
submit parts order for missing screen programmer (we forgot to order it)
post meeting notes on piazza
we will look into live streaming car racing at the final demo at competiton
no presentation for class tomorrow (Probably)
3/5 meeting with Brian Flores
*Missed - did not need to meet with him anyway. Working on Midsemester demo
*targeting a simple simulation of traction control where the car travels in a straight line
*have one Arduino due functioning as a main controller already with a simple traction control algorithm. It can also listen to data from the UM6 IMU
*also developed a simulator module which simulates the rest of the world. Accounts for motor drive force, air drag force, acceleration, velocity, position, front and rear wheel speeds, and time.
outputs data over serial port to computer with a rudimentary model of a car and road made in Processing programming language.
3/16 meeting with phil
* driver asks for yaw moments
* open loop control - deviation from center = torque split by a certain difference.
* compare acceleration and throttle position for traction control
* plain limit on tractive torque for traction control
* different profiles for different drivers
* launch control could try to cancel yaw moments
* for launch control, you just mash and once you reach 100% throttle
* for regen, apply brakes before a certain threshold (60%) and leave regen constant after that point.
* work on open loop controls
* measure derivative of driver steering wheel to determine torque split
* coarse and find adjustment
* turn-in is a short period
* jacking a wheel by comparing large differentials in rear wheel speed
* turn in should be open loop
3/24 meeting with professor Kantor
* having simulator sampling rate 10x of the controller may be a bit overkill
* using simulator to rougly dial in parameters is okay
* try to make a block diagram of control system so it is easier to understand
* position controller or speed controller? Something we have to work out. For steering, and torque vectoring, this is definitely in terms of angular position and rate
3/25 meeting with Brian Flores
covarage testing is white box - make sure thr whole thing is working the way it should be. it could be thr entire system woring œn a table with everything showing
fuctional testing is black box - did you meet the requirements?
presentation should be 5 minutes long or so.
need to prove the project is working. have tests to show that the project workng.