Support apparatus and design support method for photovoltaic power generation facilities

A design support method for photovoltaic power generation facilities, the design support method comprising the steps of:

• input data acquisition, by an input data acquisition unit, the input data acquisition including acquiring input data including site data on a site for a photovoltaic power generation facility that is an object to be designed, panel specifications data that includes data on power generation performance of a photovoltaic panel used in the photovoltaic power generation facility, sun direction data that shows a direction of a sun on a time-series basis in a position of the site, and time-series weather data on an area corresponding to the site;
• and
• calculation, by a calculation unit that includes a learned model that has been created by learning a relation between a plurality of sets of configuration data composing reference data, the reference data including site data on sites for a plurality of reference photovoltaic power generation facilities, panel layout data that includes data on installation angles of reference photovoltaic panels installed in the sites, panel specifications data that includes data on power generation performance of the reference photovoltaic panels, sun direction data that shows a direction of the sun on a time-series basis in positions of the sites, time-series weather data in areas corresponding to the sites, and time-series power generation data equivalent to actual amounts of power generated by the reference photovoltaic power generation facilities, the learned model estimating data corresponding to the power generation data from data corresponding to the reference data excluding the power generation data, the calculation including creating a plurality of pieces of provisional panel layout data on the photovoltaic panels having different installation angles from each other, calculating a value equivalent to an amount of power generation on a time-series basis by processing, for each piece of the provisional panel layout data, the relevant piece of the provisional panel layout data and the input data, by the learned model, calculating a value equivalent to a cumulative amount of power generation for a predetermined period by adding up the amounts of power generation on a time-series basis, and extracting a piece of the provisional panel layout data in which the cumulative amount of power generation for the predetermined period is maximum,
• wherein the reference photovoltaic power generation facilities are existing power generation facilities that are actually used or power generation facilities that were actually used in the past,
• the panel layout data included in the reference data includes data on inclination angles and azimuth angles of the reference photovoltaic panels,
• the learned model has been created by learning a relation between the time-series power generation data equivalent to actual amounts of power generated by the plurality of reference photovoltaic power generation facilities, and the data on inclination angles and azimuth angles of the reference photovoltaic panels, and
• a plurality of pieces of provisional panel layout data on the photovoltaic panels having different inclination angles and different azimuth angles from each other are created in the calculation step.