RAMP Introduction

An open-source bottom-up stochastic model for generating multi-energy load profiles (RAMP Website , RAMP Documentation)

What is RAMP

RAMP is an open-source software suite for the stochastic simulation of any user-driven energy demand time series based on few simple inputs.

The project aims to provide synthetic data wherever metered data does not exist, such as when designing systems in remote areas. Check out the documentation and learn more on the RAMP world from our website!

Recommended installation method

RAMP has been successfully installed and used on macOS, Windows and Linux.

The easiest way to make RAMP software working is to use the free conda package manager which can install the current and future RAMP dependencies in an easy and user friendly way.

To get conda, download and install “Anaconda Distribution”, or “miniconda” which is lighter. You can install RAMP using pip, conda or from source code.

Installing through pip

1. To install the RAMP software, we suggest to create a new environment by running the following command in the Anaconda prompt:

conda create -n ramp python=3.8

2. If you create a new environment for RAMP, you’ll need to activate it each time before using it, by writing the following line in the Anaconda Prompt:

conda activate ramp

3. Now you can use pip to install rampdemand on your environment as follow:

pip install rampdemand

Installing through the source code

You can also install RAMP from the source code! To do so, you first need to download the source code, which can be done in two ways:

• You can use git to clone the repository via:

git clone https://github.com/RAMP-project/RAMP.git

• Or, you may download the source code directly from:

“RAMP GitHub Repository”.

In this second case, the source code will be downloaded as a zip file, so you’ll need to extract the files.

After downloading the source code using any of abovementioned methods, you’ll need to use your anaconda prompt to install it. There are two options again:

• You may follow the first two steps mentioned in Installing through pip. Then, change the directory in the prompt to the folder where the source code is saved (where you can find the setup.py file). To install the RAMP software, you may then use:

python setup.py install

• Alternatively, without taking any prior action, simply change the directory in the prompt to the folder where the source code is saved and then use:

conda env create -f environment.yml

Quick start

There are different ways to build a model using RAMP! Here, we provide a first example but you can find more information in our documentation.

Example python input files

Three different input files are provided as example representing three different categories of appliances that can be modelled with RAMP. To have a look to the python files, you can download them using the download_example function:

from ramp import download_example

download_example("the specfic folder directory to save the files")

• input_file_1.py: represents the most basic electric appliances; it is an example of how to model lightbulbs, radios, TVs, fridges, and other electric appliances. This input file is based on the ones used for the first RAMP publication.

• input_file_2.py: shows how to model user-driven thermal loads, with the example of a “shower” appliance. The peculiarity of thermal appliances is that the nominal power can be provided as external input as a “.csv” file (in this case, shower_P.csv). For the example “shower” appliance, the varying nominal power accounts for the effect of groundwater temperature variation throughout the year. This input file is based on that used for this publication.

• input_file_3.py: represents an example of how to model electric cooking appliances. In this input file two different kind of meals are modelled: 1) short and repetitive meals (e.g. breakfast); and 2) main meals (e.g. lunch, dinner). Repetitive meals do not vary across days, whilst main meals do so. In particular, every household can randomly choose between 3 different types of main meal every day. Such variability in meal preferences is modelled by means of two parameters: the user preference and the preference index. The user preference defines how many types of meal are available for each user to choose every day (e.g. 3). Then, each of the available meal options is modelled separately, with a different preference index attached. The stochastic process randomly varies the meal preference of each user every day, deciding whether they want a “type 1” meal, or a “type 2”, etc. on a given day. This input file is used in this publication

You can execute python input files within an IDE, in your terminal with python command

python <path to .py input file>

or in your terminal with the ramp command, see Command line options below for more information.

Spreadsheet input files

It is also possible to use spreadsheets as input files. To do so, you need to run the ramp command with the option -i:

ramp -i <path to .xlsx input file>

Note

You can input several files, separated from each others by a single blank space you can also input python files

Command line options

In the command line you can also run .py input files If you already know how many daily profiles you want to simulate you can indicate it with the -n option:

ramp -i <path to .xlsx or .py input file> -n 10

will simulate 10 daily profiles. Note that if you do not provide this option you will being prompted for the number of daily profiles within the console.

If you want to save ramp results to a custom file, you can provide it with the option -o

ramp -i <path to .xlsx input file> -o <path where to save RAMP outputs>

Note

You can provide a number of output files, separated from each others by a single blank space, matching the number of input files.

Other options are documented in the help of ramp, which you access with the -h option

ramp -h

If you have existing python input files, you can convert them to spreadsheet. To do so, go to the \ramp folder and run

python ramp_convert_old_input_files.py -i <path to the input file you wish to convert>

For other examples of command lines options, such as setting date ranges, please visit the dedicated section of the documentation.

Building a model with a python script

# importing functions
from ramp import UseCase, User

# Create a user category
household_1 = User(
 user_name = "Household type 1", # an optional feature for the User class
 num_users = 10, # Specifying the number of specific user category in the community
)

You can add appliances to a user category by:

# adding some appliances for the household
radio = household_1.add_appliance(
 name = "Small Radio", # optional feature for the appliance class
 number = 1, # how many radio each household type 1 has
 power = 10, # RAMP does not take care of units of measure (e.g., Watts), you must be consistent
 func_time = 120, # Total functioning time of appliance in minutes
 num_windows = 2, # how many time-windows the appliance is used in
)

The use time frames can be specified using the ‘window’ method for each appliance of the user category:

# Specifying the functioning windows
radio.windows(
 window_1 = [480,540], # from 8 AM to 9 AM
 window_2 = [1320,1380], # from 10 PM to 11 PM
)

You can also add another, different user to the simulation. In this case, we use a more compact formulation:

# Create a second user category
household_2 = User(
 user_name = "Household type 2", # an optional feature for the User class
 num_users = 13, # Specifying the number of specific user category in the community
 )

# adding some appliances for the new household type in compact form, with windows specified directly and random variability
light_bulbs = household_2.add_appliance(
 name = "Light bulbs", # optional feature for the appliance class
 number = 5, # how many light bulbs each household type 2 has
 power = 7, # RAMP does not take care of units of measure (e.g., Watts), you must be consistent
 func_time = 120, # total functioning time of appliance in minutes
 time_fraction_random_variability=0.2, # 20% random variability associated to the total functioning time
 num_windows = 2, # how many time-windows the appliance is used in
 window_1 = [390,480], # from 6.30 AM to 8 AM
 window_2 = [1020,1440], # from 5 PM to 12 PM
 random_var_w=0.35 # 35% randomness assigned to the size of the functioning windows
 )

At this point, we can group our different users into a “use case” and run the simulation, for instance for a whole year.

use_case = UseCase(users=[household_1,household_2], date_start="2020-01-01", date_end="2020-12-31")
whole_year_profile = use_case.generate_daily_load_profiles()

Here is your first load for a community including two types of housholds, for a total of 23 individual users. Of course, more variations and many more features are possible! For instance, you can simulate loads even for an individual appliance or user. In addition, you can use in-built plotting functionalities to explore your results. Check out the documentation for all the possibilities.

Contributing

This project is open-source. Interested users are therefore invited to test, comment or contribute to the tool. Submitting issues is the best way to get in touch with the development team, which will address your comment, question, or development request in the best possible way. We are also looking for contributors to the main code, willing to contribute to its capabilities, computational-efficiency, formulation, etc.

To contribute changes:

1. Fork the project on GitHub

2. Create a feature branch (e.g. named “add-this-new-feature”) to work on in your fork

3. Add your name to the AUTHORS file

4. Commit your changes to the feature branch

5. Push the branch to GitHub

6. On GitHub, create a new pull request from the feature branch

When committing new changes, please also take care of checking code stability by means of the qualitative testing functionality.

How to cite

Please cite the original Journal publication if you use RAMP in your research:

F. Lombardi, S. Balderrama, S. Quoilin, E. Colombo, Generating high-resolution multi-energy load profiles for remote areas with an open-source stochastic model, Energy, 2019, https://doi.org/10.1016/j.energy.2019.04.097

More information

Want to know more about the possible applications of RAMP, the studies that relied on it and much more? Then take a look at the RAMP Website!

License

Copyright 2019-2023 RAMP, contributors listed in Authors

Licensed under the European Union Public Licence (EUPL), Version 1.2-or-later; you may not use this file except in compliance with the License.

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Note

This project is under active development!