Geographic information system‐based prediction of solar power plant production using deep neural networks

Abstract

The study aims to predict solar energy generation to ensure the successful operation of solar power plants. This objective is crucial in light of the increasing energy demand, global warming concerns, and greenhouse gas emissions. To achieve this, the study employs multiple linear regression and feature selection techniques to calculate energy generation. Additionally, long short‐term memory (LSTM) is used to predict energy generation levels based on climate conditions. Furthermore, the spatial distribution of energy generation is analyzed using inverse distance weighting. The results of the study reveal that temperature, solar radiation, relative humidity, wind speed, wind direction, and vapor pressure deficit are the most significant parameters for predicting energy generation. The LSTM method proves to be highly accurate in predicting fluctuating energy generation patterns. Notably, the southern regions of the study area exhibit a greater potential for energy generation compared to the northern regions. Approximately 30% of the region generates over 1400 kWh, with the southern areas, characterized by hot and dry climates, producing around 1500 kWh, while the northern regions, with cold and humid climates, generate approximately 1100 kWh.

Estimation of future hydropower generation of reservoir systems used for different purposes in different regions

“…A simulationoptimization model was developed for the optimal design of hydropower generation systems. In various investigations, the hydropower potential under climate change has been investigated by combining the meteorological outputs with the ones from the hydrological modelling and the general circulation models (GCMs) [12][13][14][15][16].…”

“…Stochastic optimization is used in developing reservoir operation models and in dealing with the uncertainties in streamflow estimation. In this process, such deterministic optimization methods as genetic programming, particle swarm optimization and discrete differential dynamic programming for extracting the reservoir operation rules by using the linear regression [15][16][17][18][19][20][21][22][23][24][25]. The reservoir inflow is considered as a stochastic process described by probability distributions, and the streamflow uncertainty is integrated into the reservoir operation model [26][27][28].…”

Long-Term Solar Power Time-Series Data Generation Method Based on Generative Adversarial Networks and Sunrise–Sunset Time Correction

“…However, it is difficult to comprehensively and accurately obtain astronomical meteorological data when establishing a meteorological model of irradiance. Moreover, solar power is also influenced by meteorological factors such as temperature, relative humidity, wind speed, wind direction, atmospheric pressure, as well as equipment aging [13][14][15]. It is difficult to characterize these complex relationships using indirect modeling methods.…”