Name
Abstract | ||
|---|---|---|
Solar power is important for many scenarios of the Internet of Things (IoT). Resource-constrained devices depend on limited energy budgets to operate without degrading performance. Predicting solar energy is necessary for an efficient management and utilization of resources. While machine learning is already used to predict solar power for larger power plants, we examine how different machine learning methods can be used in a constrained sensor setting, based on easily available public weather data. The conducted evaluation resorts to commercial IoT hardware, demonstrating the feasibility of the proposed solution in a real deployment. Our results show that predicting solar energy is possible even with limited access to data, progressively improving as the system runs. |
Year | Venue | Field |
|---|---|---|
2017 | IOT | Software deployment,Computer science,Internet of Things,Computer network,Solar energy,Solar power,Real-time computing,Weather data,Data access |
DocType | Citations | PageRank |
Conference | 3 | 0.63 |
References | Authors | |
13 | 5 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Frank Alexander Kraemer | 1 | 262 | 21.13 |
| Doreid Ammar | 2 | 3 | 0.97 |
| Anders Eivind Braten | 3 | 27 | 3.87 |
| Nattachart Tamkittikhun | 4 | 24 | 2.10 |
| David Palma | 5 | 72 | 8.58 |