Optimization and prediction of pulsating heat pipe compound parabolic solar collector performances by hybrid deep belief network based bald eagle search optimizer

The compound parabolic collector (CPC) with pulsating heat pipe (PHP) is developed for enhance the heat transfer rate, thermal efficiency and heat losses, and so forth. and working fluid plays a major role in this process. The thermal resistance, temperature, and thermal efficiency have been experimented with under different conditions, and heating periods were analyzed. In this, cobalt oxide (Co3O4) and graphene oxide (GO) added distilled water (DW) is used as the working fluid in the filing ratio of 50%. Bald eagle search optimization (BES) algorithm is used for optimizing the experimented values, and the better-optimized values are used for hybrid BES based deep belief network (DBN) prediction. The maximum temperature obtained for experiment and optimization is 65 and 65.16161 degrees C. 59% of thermal efficiency was obtained as maximum for experimentation, and 59.1542% of thermal efficiency was obtained as maximum for optimization. The maximum thermal resistance obtained for experimentation and optimization is 0.08 and 0.06938 degrees C/W. In this, optimized outputs performed well than the experimental values. Besides, the hybrid DBN based BES algorithm is performed based on the optimized performances to predict the temperature, thermal efficiency and thermal resistance. Further, predicted outcomes are compared with the non-hybrid neural networks such as DBN, CNN and ANN. DBN-BES depicts low error values than the non-hybrid neural networks. Overall, the proposed hybrid solar collector model and the hybrid nanoparticles added water helps to enhance the thermal characteristics with minimum heat loss.