Abstract: In an electrical grid, several electricity meters may be associated with the same transformer, and may measure electricity sold to respective customers. In an example, one electricity meter executes manager software, while the other electricity meters execute agent software. One or more applications operated on each electricity meter, and control metrology devices, data processing, operation of radio(s) and/or a powerline communications modem, aspects of electrical phase determination, etc. The agent software operating on each electricity meter may relay messages from applications operating on that electricity meter to other (possibly similar) applications operating on other electricity meters. Each message sent by each instantiation of agent software may include a networking score of that electricity meter. The manager software may additionally communicate with a data collector and/or main office server(s).

Abstract: In an electrical grid, several electricity meters may be associated with the same transformer, and may measure electricity sold to respective customers. In an example, one electricity meter executes manager software, while the other electricity meters execute agent software. One or more applications operated on each electricity meter, and control metrology devices, data processing, operation of radio(s) and/or a powerline communications modem, aspects of electrical phase determination, etc. The agent software operating on each electricity meter may relay messages from applications operating on that electricity meter to other (possibly similar) applications operating on other electricity meters. Each message sent by each instantiation of agent software may include a networking score of that electricity meter. The manager software may additionally communicate with a data collector and/or main office server(s).

Abstract: In an electrical grid, several electricity meters may be associated with the same transformer, and may measure electricity sold to respective customers. In an example, one electricity meter executes manager software, while the other electricity meters execute agent software. One or more applications operated on each electricity meter, and control metrology devices, data processing, operation of radio(s) and/or a powerline communications modem, aspects of electrical phase determination, etc. The agent software operating on each electricity meter may relay messages from applications operating on that electricity meter to other (possibly similar) applications operating on other electricity meters. Each message sent by each instantiation of agent software may include a networking score of that electricity meter. The manager software may additionally communicate with a data collector and/or main office server(s).

Abstract: Techniques for predicting solar power generation include a node measuring, using one or more sensors, a first time series of voltage readings for a first power line located at a first location, wherein a portion of power on the first power line is generated from solar irradiance on one or more first solar panels located at the first location; generating a first cross-correlation between the first time series of voltage readings and a second time series of voltage readings for a second power line located at a second location, wherein a portion of power on the second power line is generated from solar irradiance on one or more second solar panels located at the second location; and computing a wind vector based on the first location, the second location, and the first cross-correlation, wherein the wind vector is usable to forecast solar power generation at one or more other locations.

Abstract: A wireless mesh network includes a group of nodes configured to predict cloud movements based on voltage time series data. A node residing in the wireless mesh network records voltage fluctuations at a site where solar power is generated. The voltage fluctuations occur when an advancing cloud reduces solar irradiance at the site, thereby reducing solar power generation. The node correlates these voltage fluctuations with other voltage fluctuations recorded by other nodes at other sites where solar power is generated. The node computes a time offset between these voltage fluctuations that corresponds to the time needed for the cloud to advance between the different sites. Based on this time offset and the locations of the various nodes, the node estimates a wind vector. The wind vector can be used to perform near-term solar forecasting by predicting when the cloud will advance to other sites and reduce solar power generation.

Abstract: A wireless mesh network includes a group of nodes configured to predict cloud movements based on voltage time series data. A node residing in the wireless mesh network records voltage fluctuations at a site where solar power is generated. The voltage fluctuations occur when an advancing cloud reduces solar irradiance at the site, thereby reducing solar power generation. The node correlates these voltage fluctuations with other voltage fluctuations recorded by other nodes at other sites where solar power is generated. The node computes a time offset between these voltage fluctuations that corresponds to the time needed for the cloud to advance between the different sites. Based on this time offset and the locations of the various nodes, the node estimates a wind vector. The wind vector can be used to perform near-term solar forecasting by predicting when the cloud will advance to other sites and reduce solar power generation.