Datasets with measurements of both solar electricity production and domestic electricity consumption separated into the major loads are interesting for research focussing on (i) local optimization of solar energy consumption and (ii) non-intrusive load monitoring. To this end, we publish the iHomeLab RAPT dataset consisting of electrical power traces from five houses in the greater Lucerne region in Switzerland spanning a period from 1.5 up to 3.5 years with a sampling frequency of five minutes. For each house, the electrical energy consumption of the aggregated household and specific appliances such as dishwasher, washing machine, tumble dryer, hot water boiler, or heating pump were metered. Additionally, the data includes electric production data from PV panels for all five houses, and battery power flow measurement data from two houses. Thermal metadata is also provided for the three houses with a heating pump.
In the light of the current respiratory disease pandemic, air quality gains relevance in cities around the world. Continuously monitoring trends of pollutants with a higher spatial and time resolution is desired to support the information from high quality stations. Given the availability of low-cost sensors and data transmission via wireless networks, it is possible to envision such monitoring with high quality at large scales. Facilitating that smart cities actively influence the distribution of pollutants, for instance, by active traffic control. Thus, offering citizens a better environment to live in. Our sensor-box prototype aims at monitoring air quality and other environmental parameters with lower cost sensors and components. Several challenges exist to achieve sensible measurements from metal oxide semiconductor (MOS) or from electro-chemical (EC) sensors such as: reproducibility, stability, uncertainty, drift, cross sensitivities. Here, the focus is on comparing nitrogen dioxide (NO2) measurements from these two families of low-cost sensors. Thus, we report on a side-by-side evaluation of the MOS sensor from a commercial system, and the EC sensor in our sensor-box. Recalibration of our EC sensors is demonstrated with a linear fit to the commercial system.
Figure 1: We propose an automatic feature selection method to find near-optimal auxiliary feature subsets that maximize the denoising quality of neural denoisers. In this figure, we demonstrate our approach by comparing the quality obtained with our automatically selected feature set to not using any volumetric features (KP-V18 [VRM * 18]), or to the hand-crafted feature sets proposed by the authors (DP-H20 [HMES20]). Compared to baseline feature sets, our selected feature sets can lead to much improved results with the same underlying neural network architecture on different types of volumetric effects.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.