placement

Production code for Sensor Placement Optimization

Optimal Sensor Placement

We have developed an intelligent and adaptable system designed to optimize the placement of continuous monitoring sensors on oil and gas sites for methane emissions detection. Further, our approach addresses a larger problem scale compared to previous studies and can be customized for various sensor placement objectives.

Read the paper associated with this work for more detail.

Installation & Usage

To run and engage with the code, the simplest approach is to follow these steps to clone and work from the placement directory within the gp environment:

0. Setup the gp environment for FastGaussianPuff

1. Clone placement locally:

git clone https://github.com/Hammerling-Research-Group/placement.git

2. Move into the placement directory:

cd placement

3.1 If a first-time user, create the placement environment:

conda env create -f root_environment.yml

3.2 If previously created, update the placement environment:

conda env update -f root_environment.yml

4. Activate the gp environment (from FastGaussianPuff):

conda activate gp

5. Pip install to work with the latest version of FastGaussianPuff:

pip install .

Of note: As noted above, placement relies heavily on FastGaussianPuff, and Step 1 assumes users have set up the associated gp environment already. If not done, users must first setup the gp environment for FastGaussianPuff prior to running the placement code per Step 0. Follow the steps to do so here.

When finished and input data are either developed or ingested (see the following section for a clearer understanding of the directory structure), users may run each of the three core scripts in sequence (as well as the unit testing suite, each prefixed by test_*):

• simulate_concentrations.py
• evaluate_detection.py
• optimization.py

In the current architecture, output data from running this process will be sent to ./demo/output_data/. Users may easily update these paths if desired.

Testing

Users are encouraged to test the placement codebase as well. Our testing framework leverages pytest, so be sure to call pytest instead of python when testing. See the following examples for more.

To test all main step scripts (3 in total), run:

pytest tests/

To test individual scripts, run, e.g.:

pytest tests/test_optimization.py

(Evolving) Package Structure

Note: structuring according to requirements for submission to pypi, as well as in line with package best practices.

placement/
│
├── placement/
│   ├── __init__.py
│   ├── simulate_concentrations.py
│   ├── evaluate_detection.py
│   └── optimization.py
│   └── PORSS.py
│
├── tests/
│   ├── __init__.py           
│   ├── test_simulate_concentrations.py
│   ├── test_evaluate_detection.py
│   └── test_optimization.py
│
├── docs/                   # Documentation (Note: only for pypi)
│   ├── index.rst           # Main documentation index for Sphinx
│   └── api.rst             # API docs for each module
│
├── demo
│   ├── input_data
│   ├── output_data
│   ├── results_fenceline_locations
│   ├── step1_generate_emission_scenarios.ipynb
│   ├── step2_specify_sensor_locations.ipynb
├── LICENSE
├── README.md
├── setup.py                # Setup script for packaging and metadata
├── MANIFEST.in             # Manifest file to include non-code files in the package
├── .gitignore
└── pyproject.toml