dlr.esy.grid.match is a grid matching & conversion tool primarily developed for the comparison of power grids.

Installation

dlr.esy.grid.match is available on PyPI and can be installed into your Python environment via pip:

pip install dlr.esy.grid.match

Command-line interface

dlr.esy.grid.match provides a command-line interface (see dlr.esy.grid.match.cli for more details):

$ dlr.esy.grid.match --help
usage: dlr.esy.grid.match [-h] {compare,example} ...
...
Command-line interface for grid comparisons.
...
positional arguments:
  {compare,example}
    compare          Compares two powergrids. `a` and `b` are expected to be
                     directories, each containing the files `bus.csv` and
                     `line.csv`.
    example          Creates an example datasets.
...

Usage

The command compare adds the columns center and map to the grids $a$ and $b$. The column center contains the index of the center of the cluster each node belongs to and the column map the index of the center of the corresponding cluster in the other grid. See the methodology section for details. The subcommand grids plots the two grids together and the subcommand plot shows or saves the plots.

The subcommand example can be used to create example datasets and write them into a directory, like the simple dataset:

$ dlr.esy.grid.match example simple example/simple
$ dlr.esy.grid.match compare --a=example/simple/a --b=example/simple/b grid plot --save=figures/grids.svg

The subcommand cluster plots the grids separately highlighting the convex hulls of each cluster (in orange for the hulls in the plotted grid and in grey for the ones in the other grid). The centers and the intercluster lines are plotted in red and the intracluster lines and the rest of the nodes are plotted in orange.

$ dlr.esy.grid.match compare --a=example/simple/a --b=example/simple/b cluster plot --save=figures/cluster.svg

The subcommand join produces a third grid with the nodes in gird $a$ which are in the interior of a region and the nodes in grid $b$ which are outside of a region. The information about the regions needs to be in a column called join in each grid (0 meaning inside the region and 1 meaning outside). If the column join is missing, it can be added to each grid with the subcommand max_res, as described in the methodology section.

$ dlr.esy.grid.match compare --a=example/simple/a --b=example/simple/b max_res join grid plot --save=figures/join.svg

Input and output

dlr.esy.grid.match supports reading and writing grid data from CSV files. Grid data may also be converted to geojson data.

dlr.esy.grid.match requires bus data to contain at least the fields name, x and y, while line data is required to contain at least the fields name, bus0 and bus1.

Input field names can be renamed as in the following example, which reads grid data from a non-conformative example in the directory example/nonconformative and stores the resulting grid in example/conformative:

$ dlr.esy.grid.match example nonconformative example/nonconformative
$ dlr.esy.grid.match compare --a=example/nonconformative/a --a_bus_fields=bus:name --a_line_fields=line:name --b=example/nonconformative/b --b_bus_fields=bus:name --b_line_fields=line:name csv example/conformative

The command geojson produces a geojson output, which can be used to visualize the grids with the clusters, for instance by importing it to https://geojson.io/.

$ dlr.esy.grid.match compare --a=example/simple/a --b=example/simple/b geojson
{"type": "FeatureCollection", ...}

Methodology

Comparison methodology

Given two grids $a$ and $b$ with geolocated nodes, clusters of nodes in each grid are produced and each cluster in $a$ is uniquely identified to a cluster in $b$. The method has the following steps:

1. For each node $n_a$ in $a$, define $f(n_a)$ to be its closest node in $b$ (using the Euclidean distance) and, for each node $n_b$ in $b$, define $g(n_b)$ to be its closest node in $a$.

2. Each node $n_a$ in $a$ such that $g(f(n_a))=n_a$ is chosen as center of a cluster in $a$ and each node $n_b$ in $b$ such that $f(g(n_b))=n_b$ is chosen as center of a cluster in $b$. Notice that, by construction, each grid has the same number of clusters and there is a one to one correspondence between them. Namely, the center $n_a$ is associated to the center $n_b = f(n_a)$ and the center $n_b$ is associated to $n_a = g(n_b)$.

3. Each node in $a$ not being a center of any cluster is assigned to a cluster using the shortest path (using Euclidean distance) to a center. The same is done for $b$.

The output associates to each node the center of its cluster and also the corresponding center in the other grid. This can be used to compare and transfer the data associated to the nodes in both grids. It can also be used to create a third grid with the highest spatial resolution of both, as explained in the next section.

Joining methodology

Given two grids $a$ and $b$ in which all nodes are geolocated and assigned to one of two regions $\Gamma$ or $\Omega$, a new grid is produced having the nodes of $a$ which belong to $\Gamma$ and the nodes of $b$ which belong to $\Omega$. The lines of the joined grid are the ones in the three following groups:

• Lines in $a$ connecting two nodes belonging to $\Gamma$;

• Lines in $b$ connecting two nodes belonging to $\Omega$; and

• for each line in $a$ connecting a node $n_a$ in $\Gamma$ to a node $m_a$ in $\Omega$ a line connecting $n_a$ to the closest node (using Euclidean distance) in $b$ which belongs to $\Omega$.

To get the joined grid with the maximal resolution, the output of the comparison methodology is used, where a node in $a$ is assigned to $\Gamma$ if it belongs to a cluster with more or equal number of elements than its corresponding cluster in $b$.

Design, Development & Contributing

Design and development notes are available in dlr.esy.grid.match.test.

We would be happy to accept contributions via merge requests, but due to corporate policy we can only accept contributions if you have send us the signed contributor license agreement.

License

dlr.esy.grid.match is licensed under the MIT license.

Contact

Please use the projects issue tracker to get in touch.

Team

dlr.esy.grid.match is developed cooperatively by the DLR Institute of Networked Energy Systems in the departement for Energy Systems Analysis (ESY) and by the Institute of Energy and Climate Research - Techno-economic System Analysis (IEK-3) of the Forschungszentrum Jülich.

Footnotes

Test console examples.

>>> import dlr.esy.grid.match.test
>>> dlr.esy.grid.match.test.doctest_console(__doc__)