How to Win Using Modeling and Simulation

@/Katherine there’s a link to fill out for us to get free MatLab licenses - https://mathworks.com/academia/student-competitions/american-solar.html

Open

Open

Equation Based Modelling:

• Easiest and fastest

For power train components:

Not much information needed to input to get useful info from graphs and a check if parts will work

• Powertrain components are built using smaller components

• Can look in each model for details about how it is being modeled

• Can override specific parts (ex: specific analysis of batteries or motors)

• Can generate code and put them in simulator for hardware testing

• Made with MatLab project and helps organize everything

• inside the motor component

• can also switch to more complicated motor model

• can consider effects during certain maneuvers

• don’t care about pitch or roll of vehicle

• start considering yaw, turns

• Everything being considered - 14 degrees of freedom

• Models suspension in lots of detail

• Can’t model custom vehicle rn in 3D but can use a different car model to see

• red dots = target path

• Can make lidar point cloud generation

• Get camera output

• For multidomain systems: combination of systems (mech, electrical, etc)

• There’s also a simscape language to create custom blocks/components

• Those blocks can use same equations as earlier

• Automatically creates direction between components

Simple multidomain system

• Includes friction and load torque in system

• Simscape will derive equations

• Lots of add-ons

• Multi-body simulation

• Starter models for solar components

• Plot from above slide!

• can use data from battery cells and develop Simscape model

• Mechanical Systems

• Can import CAD geometry

• 4 post test rig

• Uses simplified CAD geometry

• Similar to what’s used in industry

• Can get 3D representation of force on test rig

  • Can combine the different models/toos

• Can import and export C/C++ code from simulink

• https://www.mathworks.com/academia/student-competitions/tutorials-video.html

Useful tutorial links

• They also have Onramp courses for more in depth areas, signal processing, control design, machine learning, etc.

• more resources in discord

Question: Is there a way to work collaboratively on the cloud

Answer: For smaller stuff, can use MatLab drive. For larger stuff, not really, but using projects is really helpful and can incorporate version control.