Electric-Car-Racing

For Electric Race Car Simulations, simulating how much energy it takes to race an electric car around tracks

Usage

To use this simulation you must have python (3.X) installed on your machine Once installed simply navigate to the top level of this repo and run python3 main.py A UI should appear and results should be plotted, there will also be command line printout for debugging purposes.

This simulation uses NREL's FASTsim as validation. First results are calculated using this simulation's calculation methods. Then a custom drive cycle is made from results of this simulation and input to FASTsim to validate results.

NREL FASTsim: link FASTsim validation: link

Default Usage

The default car used in the simulation is ./cars/fastsim_car_test.csv The default track used in the simulation is ./tracks/high_plains_track.csv

Repository Layout

This repository uses the following file structure:

• Results are in the ./results/ folder
• Logging output is in the ./results/logging_output/ folder (if enabled)
• cProfile results are output to the ./results/cProfile-results folder (enabled by default)
• Car property files should be stored in the ./cars/ folder
• Track property files should be stored in the ./tracks/ folder

Present Status?

This is a next iteration at the new App architecture with QThreads and Qt.Core SIGNALs between the main window and the Race Sim Thread. It is builds on the the last working main from Matt AND merging in the concepts from branch "one-lock-to-rule-them-all" which implements a single read/write lock in the DataStore access methods to control access to the contained data from the MainWindow and SimulationThreads.

There is a QApplication which has a custom MainWindow class for hosting application controls and displaying the graphical results. The basic architecture utilizes Qthread and Signal/Slots coordinate activities between the MainWindow and worker threads.

At startup the MainWindow is responsible for starting the worker threads. For this version there are 2 worker threads - SimulationThread & PlotRefreshThread

• SimulationThread calculates and publishes calculations the actual scientific results to DataStore object
• PlotRefreshThread is a simple timer thread that periodically signals to the MainWindow to go retrieve SimulationThread results from the DataStore

For this version the MainWindow user "controls" are two push buttons that say "Run/Continue" and "Pause". Pressing "Run/Continue" causes the SimulationThread to emit a signal to be emitted from the Worker thread after "Run/Continue" is pressed.

This signal from the Worker thread emits is a signal back (with a string of data, actually a digit) to the mainWindow thread that is received on a main thread slot, that updates the status text widget "Status" with the signaled data.

A second periodic signal that is emitted from the PlotRefreshThread to the MainWindow is used as a trigger for the MainWindow to generate a plot. The MainWindow in turn plots the data in DataStore from 0 to the current simulation_index.

DataStore has been updated to use a single lock for coordinating writes/reads to the data store that is shared by the SimulationThread (reader/writer) and the MainWindow (thread) which is a read-only consumer of the data. In this intermediate version of the code, we're just playing with (sharing) the data store's simulation_index value.

This version of the code is run with the command: $ python3 simulation.py

And it works!

Vieweing cProfile simulation results

As of 12/15 the simulation will output a file named ./results/cprofile-results/profile-simulation.out and ./results/cprofile-results/profile-visualization.out using the cProfile module. There is a profile for each high use thread in the simulation. These results can be viewed using runsnake The things that I did to get cProfile and runsnake to work for me are documented here: https://kupczynski.info/2015/01/16/profiling-python-scripts.html

To view do the following things:

Install runsnake

On Ubuntu 20.04 Install the runsnake app and associated packages as follows:

$ sudo apt install runsnakerun

Verify the installation: which runsnake should return something in your $PATH

$ which runsnake

Run runsnake

call runsnake with profile.out as an argument: runsnake profile.out

$ runsnake profile.out

That should do it!

Car Variables

NEEDS UPDATE This simulation uses variables names that are named the same as FASTsim's variables for car related variables. Reference: link We also might use the TUM simulation for variable names...

Car Variables

Description of car variables and correlation to FASTsim's variables

Abbreviations used in FASTsim

Table of Variables

Track Variables

This simulation uses track data loaded from a designated, or default, csv file in the directory /tracks.

The format of the csv file follows that of the Technical University of Munich Institute for Automotive Technology's autonomous electric car racing simulation (https://github.com/TUMFTM/global_racetrajectory_optimization). There are two added variables, however, for elevation and air density.

The race trajectories in these files are created using TUM's simulation, with elevation added after the fact.

Track Variable Definitions

Initialization file (race_init.ini)

This file provides initial conditions for the simulation. These values do not change throughout its course.

Included variables