Implementation of the Harmonic Extreme Learning Machine for Clustering the Solar Irradiance in Bangkok, Thailand

Solar irradiance is renewable energy that can quickly produce energy for households and industries. On each day and event, the solar irradiance can be varied due to the change in weather. The solar irradiance clustering can evaluate the appropriate generation pattern of a solar photovoltaic system for analyzing daily energy generation. Therefore, this paper proposes a novel clustering method to evaluate the appropriate cluster of solar irradiance generation by utilizing the Harmonic Extreme Learning Machine model to improve the performance of solar irradiance analysis and management. The experimental results showed that the Harmonic Extreme Learning Machine model had given the minimum total sum of distances between the centroid cluster and the represented data. The plot of the result is evident for explaining and interpreting the behavior of solar irradiance data and managing the data to the appropriate cluster.