Performance predication of a solar assisted desiccant air conditioning system using radial basis function neural...

Area of Science:

• Renewable Energy Systems
• HVAC Technology
• Artificial Intelligence in Engineering

Background:

• Solar desiccant air conditioning (Sol-DAC) systems offer a sustainable alternative for cooling by utilizing solar energy for regeneration.
• Integration with Maisotsenko cycle (M-cycle) indirect evaporative coolers enhances latent and sensible load handling capabilities.
• Accurate performance prediction is crucial for optimizing Sol-DAC system design and operation.

Purpose of the Study:

• To develop and validate a predictive model for a novel Sol-DAC system incorporating an M-cycle indirect evaporative cooler.
• To investigate the system's performance under various operating parameters using experimental data.
• To utilize Radial Basis Function Neural Network (RBF-NN) for predicting key system performance metrics.

Main Methods:

• A solar desiccant air conditioning system, featuring a cross-flow M-cycle indirect evaporative cooler and a desiccant wheel (DW), was experimentally evaluated.
• A solar evacuated tube electric heater provided regeneration temperature for the DW.
• A Radial Basis Function Neural Network (RBF-NN) was trained using nine input parameters and four output parameters (temperature, humidity, Cooling Capacity (CC), Coefficient of Performance (COP)) under transient conditions.

Main Results:

• The RBF-NN model demonstrated high accuracy in predicting system performance, with excellent Mean Squared Error (MSE) and Regression coefficient (R) values for all output parameters.
• Optimal predictions were achieved with a regeneration temperature of 70°C across various inlet humidity conditions.
• Experimental and predicted performance parameters showed close agreement, with minimal deviation, confirming the model's reliability.

Conclusions:

• The trained RBF-NN model effectively predicts the performance of the integrated Sol-DAC and M-cycle system under dynamic operating conditions.
• The predictive capability of the RBF-NN model is validated, offering a valuable tool for system analysis and optimization.
• The study highlights the potential of combining solar desiccant cooling with M-cycle technology for efficient and sustainable air conditioning.