rdoddanavar / hpr-sim Goto Github PK

*.h vs *.hpp --> header files in project are strictly C++ compatible via pybind11

Discussion: https://stackoverflow.com/questions/152555/h-or-hpp-for-your-class-definitions?rq=1

Commit initial environment.yml style file to root directory. Freeze python3 & dependency versions.

• https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#sharing-an-environment

Proposed preproc_model.py that defines parsing routines specific to model input files; ex. aero model, prop model, etc.

Model file format TBD; ideally minimal work needed to condition RASAero for sim input

Module should create data suitable for initializing aero model

How can I release the memory used by Model derived objects after their associated flight completes? Do I call their destructor from the python side?

Provide more flexibility w/ interpolation options; currently, interpolation method is hard-coded to gsl_interp_steffen

Considerations for launch rail effects on EOM:

• force vector --> dot product w/ launch rail unit vector
• sum of moments --> 0 while on rail

When are rail interactions over?

• When CG clears?
• Or aft end?
• Or location(s) of launch lugs / rail buttons?

Geodetic model for gravitational acceleration, as a function of altitude and global location. Keeping it simple for now....

Probably going to use basic WGS-84 w/ altitude dependence as described in Eq. (4-3) in NGA.STND.0036 v1.0.0: https://earth-info.nga.mil/GandG/publications/NGA_STND_0036_1_0_0_WGS84/NGA.STND.0036_1.0.0_WGS84.pdf

In the future think about upgrading to Eq. (4-4) to better model altitude dependence (up to 20e3 m)

Define a class function similar to Param::check_value() to validate the string value against a valid set of strings defined in config_param.yaml

• Unit conversion routine
• Unit configuration YML

Implement class that defines utilities for managing the internal state of a physical model:

Possible methods/functions:

• State initialization
• " " update --> the model routine itself
• " " reset
• process to call & update models it depends on

Bash script to update directory index in README.md as project develops

Implement test script that finds program execution time

https://help.github.com/en/articles/ignoring-files

Simplify load_input.py by removing nested getattr() calls with one of these solutions: https://stackoverflow.com/questions/31174295/getattr-and-setattr-on-nested-objects

Function convert(*args):

• Unit multiplier 1.0 returned for unspecified or nondimensional units
• Need exception for invalid dict keys --> nonexistent units

Reference: https://docs.python.org/3/tutorial/errors.html

Telem ("telemetry") class

Develop build settings and test routine to profile C++ pybind11 modules for runtime efficiency & memory usage. Use gprof in some way (or sprof)?

Implement ODE utilities using solvers provided by GNU GSL

Documentation: https://www.gnu.org/software/gsl/doc/html/ode-initval.html

Initial build against Eigen library. Should not need a separate CMake target, simply link to header directories:

We use the CMake build system, but only to build the documentation and unit-tests, and to automate installation. 
If you just want to use Eigen, you can use the header files right away. 
There is no binary library to link to, and no configured header file. 
Eigen is a pure template library defined in the headers

• https://pyinstaller.readthedocs.io/en/stable/operating-mode.html

Control which fields are accessible during input parameter assignment. Forces user to modify config_param.yml when changing hidden fields

User input:

• value
• unit
• dist & related

Config only:

• min
• max
• quantity
• path (for Name class)

Target specific release of dependencies using corresponding SHA tag(s)

Define a parameter class similar to Param or Name that handles boolean input

Class data:

• bool value

Class functions:

• bool check_value()

Class derived from model to provide mass props during runtime:

• Overall mass
• CG (X,Y,Z) position
• Inertia tensor terms (Ixx, Iyy, Izz, Ixy, Ixz, Iyz)

Proposed preproc_model.py that defines parsing routines specific to model input files; ex. aero model, prop model, etc.

Engine model file format: http://www.thrustcurve.org/raspformat.shtml

Module should create data suitable for initializing prop model

Error handling against check_value() and check_path(); consider moving routines to python away from C++ member functions

Master python script that executes python & cpp subroutines. Takes command line argument(s) of input file paths

ex. (CLI) $ python3 exec.py myinput.yaml

Integrate Travis CI service with .travis.yml in root directory.

Reference: https://docs.travis-ci.com/user/languages/cpp/

Revisit:

• Importing user packages
• Use of __init.py__ files
• Adding src/ subdirectories to system path

Eigen linear algebra library (C++)

• Choose one of these: https://gitlab.com/libeigen/eigen/-/releases

Update Geodetic::update() routine to reflect more detailed models described in NGA.STND.0036_1.0.0_WGS84

Figure out the binding necessary to make Param a template class, to handle additional input types beyond float:

• pybind/pybind11#199
• https://stackoverflow.com/questions/47487888/pybind11-template-class-of-many-types

Implement U.S. Standard Atmosphere 1976

Quick check w/ MATLAB's ode45 to make sure my solver in GSL is working; 1DOF, no atmos/drag, just thrust & gravity

• https://stackoverflow.com/questions/1186789/what-is-the-best-way-to-call-a-script-from-another-script
• https://stackoverflow.com/questions/436198/what-is-an-alternative-to-execfile-in-python-3

Extend YAML capabilities:

• Internal references: ref( ... )
• Arithmetic expressions: eval( ... ) or expr( ... )

Implement lambda functions in config_unit.yml w/ corresponding routine in util_unit.py to handle more complex unit conversions

CPP routine that generates random sequence given seed (int) and desired distribution (string), using the <random> library: http://www.cplusplus.com/reference/random/

Multiprocessing capability w/ RNG