A novel hybrid approach based on self-organizing maps, support vector regression and particle swarm optimization to forecast solar irradiance. (15th March 2015)

• Record Type: Journal Article
• Authors: Dong, Zibo; Yang, Dazhi; Reindl, Thomas; Walsh, Wilfred M.
• Is Part Of: Energy. Volume 82(2015)
• Page Start: 570
• Page End: 577
• Publication Date: 2015-03-15
• Subjects: Hourly solar irradiance forecasting -- Self-organizing maps -- Support vector regression -- Particle swarm optimization; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2015.01.066
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: We forecast hourly solar irradiance time series using a novel hybrid model based on SOM (self-organizing maps), SVR (support vector regression) and PSO (particle swarm optimization). In order to solve the noise and stationarity problems in the statistical time series forecasting modelling process, SOM is applied to partition the whole input space into several disjointed regions with different characteristic information on the correlation between the input and the output. Then SVR is used to model each disjointed regions to identify the characteristic correlation. In order to reduce the performance volatility of SVM (support vector machine) with different parameters, PSO is implemented to automatically perform the parameter selection in SVR modelling. This hybrid model has been used to forecast hourly solar irradiance in Colorado, USA and Singapore. The technique is found to outperform traditional forecasting models. Highlights: SOM (self-organizing maps) is applied to partition the whole input space into several disjointed regions. SVR (support vector regression) is used to model each disjointed regions to identify the characteristic correlation. PSO (particle swarm optimization) is implemented to automatically perform the parameter selection in SVR modelling. The hybrid technique is found to outperform traditional forecasting models.