Predicting the performance of solar dish Stirling power plant using a hybrid random vector functional link/chimp optimization model

Abstract

Hybrid artificial intelligence models have become promising tools for soft computing and computational intelligence, as they can deal with complicated sustainable systems such as the prediction modeling of concentrated power systems. In these models, one or two artificial intelligence techniques are integrated with an optimization algorithm to develop a fine-tuned prediction modeling. In this paper, we develop a novel hybrid prediction model using an improved version of the Random Vector Functional Link (RVFL) network to predict the instantaneous output power and the monthly power production of a solar dish/Stirling power plant (SDSPP). A new metaheuristic algorithm called Chimp Optimization Algorithm (CHOA) has been combined with the RVFL network to effectively determine the optimal values of RVFL parameters. More so, the proposed RVFL-CHOA model is compared with four artificial-based models include the original RVFL, and three hybrid modified versions of the RVFL model using the Particle Swarm Optimization (PSO), Spherical Search Optimization (SSO), and Whale Optimization Algorithm (WOA). The prediction performance of the five models was compared using various statistical evaluation metrics. The statistical results prove the superiority and effectiveness of the proposed RFVL-CHOA method among the other investigated optimized models for performance prediction of the SDSPP. Based on the test data, the REVL-CHOA predicts the instantaneous output power and the monthly power production of the SDSPP with determination coefficient values of 0.9992, and 0.9108, and root mean square error values of about 0.00047, and 0.05995, respectively.