Energy- and Cost-Efficient Pumping Station Control
Authors: Timon Kanters, Frans Oliehoek, Michael Kaisers, Stan van den Bosch, Joep Grispen, Jeroen Hermans
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | An empirical comparison with the current control algorithms indicates that substantial cost, and thus peak load, reduction can be attained. |
| Researcher Affiliation | Collaboration | Timon V. Kanters Informatics Institute University of Amsterdam tvkanters@tvkdevelopment.com Frans A. Oliehoek Informatics Institute, Univ. of Amsterdam Dept. of CS, University of Liverpool fao@liverpool.ac.uk Michael Kaisers Centrum Wiskunde & Informatica michael.kaisers@cwi.nl Stan R. van den Bosch Nelen & Schuurmans stan.vandenbosch@nelen-schuurmans.nl Joep Grispen Nelen & Schuurmans joep.grispen@nelen-schuurmans.nl Jeroen Hermans HH Hollands Noorderkwartier j.hermans@hhnk.nl |
| Pseudocode | No | Pseudocode for CN is described by Kanters [2015]. |
| Open Source Code | No | The paper does not mention providing open-source code for the described methodology. |
| Open Datasets | Yes | In the current simulations, we use historical data from Meteobase (STOWA 2015)...The experiments in this article thus use historical data of imbalance prices (Tenne T 2015). |
| Dataset Splits | No | The paper does not explicitly describe training/validation/test dataset splits for the UCT model. The evaluation of simulation realism section validates the simulator, not the model's training. |
| Hardware Specification | No | The paper mentions running simulations on 'a typical work station' but does not provide specific hardware details such as CPU/GPU models, memory, or other specifications. |
| Software Dependencies | No | The paper does not specify version numbers for any software dependencies or libraries used. |
| Experiment Setup | Yes | Based on our experiments (Kanters 2015), we found 65, 000 planning simulations (which takes about about 5 seconds on a typical work station) and a search depth of 32 decision epochs (which equates to 8 hours) to be good settings. |