A photovoltaic system's configuration specification can be inferred by an evaluative process that searches through a space of candidate values for the variables in the specification. Each variable is selected in a specific ordering that narrows the field of candidate values. A constant horizon is assumed to account for diffuse irradiance insensitive to specific obstruction locations relative to the photovoltaic system's geographic location. Initial values for the azimuth angle, constant horizon obstruction elevation angle, and tilt angle are determined, followed by final values for these variables. The effects of direct obstructions that block direct irradiance in the areas where the actual horizon and the range of sun path values overlap relative to the geographic location are evaluated to find the exact obstruction elevation angle over a range of azimuth bins or directions. The photovoltaic temperature response coefficient and the inverter rating or power curve of the photovoltaic system are determined.

To provide a cloud observation device capable of reducing calculation cost and predicting sunshine probability by a simple method. A cloud observation device includes an image acquisition module which acquires an image in which a camera photographs the sky, a cloud extraction module which extracts clouds in the image, a sun position determination module which determines a sun position in the image, a sunshine probability calculation area setting module which sets a sunshine probability calculation area having the sun position as a base point in the image, and a sunshine probability calculation module which calculates a sunshine probability after a predetermined time has elapsed based on the sunshine probability calculation area and the extracted clouds.

To provide a cloud observation device capable of reducing calculation cost and predicting sunshine probability by a simple method. A cloud observation device includes an image acquisition module which acquires an image in which a camera photographs the sky, a cloud extraction module which extracts clouds in the image, a sun position determination module which determines a sun position in the image, a sunshine probability calculation area setting module which sets a sunshine probability calculation area having the sun position as a base point in the image, and a sunshine probability calculation module which calculates a sunshine probability after a predetermined time has elapsed based on the sunshine probability calculation area and the extracted clouds.

A power generation prediction system using a first and second neural networks is provided, and the first neural network is connected to the second neural network. The first neural network receives first input data, and generates the amount prediction data according to the first input data. The first input data is used to determine amount prediction data, and the amount prediction data is used to determine power generation prediction data. The second neural network receives the amount prediction data, and calculates the power generation prediction data according to the amount prediction data. When a device in a selected area is deteriorated or reinstalled, the second neural network is fine-tuned and trained again. The power generation prediction data is a power generation prediction bound having a maximum and minimum power generation prediction values, and thus the power deployment terminal in a power grid can deploy power more precisely.

Certain aspects pertain to a combination sensor comprising a set of physical sensors facing different directions proximate a structure, and configured to measure solar radiation in different directions. The combination sensor also comprises a virtual facade-aligned sensor configured to determine a combi-sensor value at a facade of the structure based on solar radiation readings from the set of physical sensors.

Certain aspects pertain to a combination sensor comprising a set of physical sensors facing different directions proximate a structure, and configured to measure solar radiation in different directions. The combination sensor also comprises a virtual facade-aligned sensor configured to determine a combi-sensor value at a facade of the structure based on solar radiation readings from the set of physical sensors.

Provided are an insolation correction method, an insolation correction device, a recording medium, a model, a model generating method, and a model providing method in which insolation data having a small deviation from an actually measured value can be provided. An insolation correction method includes: acquiring insolation data; acquiring weather data; and correcting the acquired insolation data, on the basis of the acquired weather data.

Provided are an insolation correction method, an insolation correction device, a recording medium, a model, a model generating method, and a model providing method in which insolation data having a small deviation from an actually measured value can be provided. An insolation correction method includes: acquiring insolation data; acquiring weather data; and correcting the acquired insolation data, on the basis of the acquired weather data.

The calculation of the variance of a correlation coefficient matrix for a photovoltaic fleet can be completed in linear space as a function of decreasing distance between pairs of photovoltaic plant locations. When obtaining irradiance data from a satellite imagery source, irradiance statistics must first be converted from irradiance statistics for an area into irradiance statistics for an average point within a pixel in the satellite imagery. The average point statistics are then averaged across all satellite pixels to determine the average across the whole photovoltaic fleet region. Where pairs of photovoltaic systems are located too far away from each other to be statistically correlated, the correlation coefficients in the matrix for that pair of photovoltaic systems are effectively zero. Consequently, the double summation portion of the calculation can be simplified to eliminate zero values based on distance between photovoltaic plant locations, substantially decreasing the size of the problem space.

The calculation of the variance of a correlation coefficient matrix for a photovoltaic fleet can be completed in linear space as a function of decreasing distance between pairs of photovoltaic plant locations. When obtaining irradiance data from a satellite imagery source, irradiance statistics must first be converted from irradiance statistics for an area into irradiance statistics for an average point within a pixel in the satellite imagery. The average point statistics are then averaged across all satellite pixels to determine the average across the whole photovoltaic fleet region. Where pairs of photovoltaic systems are located too far away from each other to be statistically correlated, the correlation coefficients in the matrix for that pair of photovoltaic systems are effectively zero. Consequently, the double summation portion of the calculation can be simplified to eliminate zero values based on distance between photovoltaic plant locations, substantially decreasing the size of the problem space.