Optimal solar COP prediction of a solar-assisted adsorption refrigeration system working with activated carbon/methanol as working pairs using direct and inverse artificial neural network

The aim of this work is to develop an ANN model to predict the solar COP (COPs) of a solar intermittent refrigeration system for ice production working with Activated carbon (AC)/methanol pair. A feedforward (FFBP) with one hidden layer, a Levenberg–Marquardt learning (LM) algorithm, hyperbolic tangent sigmoid transfer function and linear transfer function for the hidden and output layer respectively, were used. The best fitting training data was obtained with the architecture of (8 inputs, 8 hidden and 1 output neurons), Results of the ANN showed an excellent agreement R2 > 0.9985 between simulated and those obtained from literature with maximum root mean square error and RMSE = 0.0453%. A sensitivity analysis was also conducted using the inverse artificial neural network method to study the effect of all the inputs on the COPs. Results from the ANNi showed a good agreement in the case of the mass of activated (error less than 0.08%).

► Solar adsorption refrigeration system (SARS) was modeled using ANN.
► Performance of SARS may be influenced differently by the inputs.
► Sensitivity analysis was conducted by ANNi as a control strategy for the effective inputs.
► Graphical interface for the ANN model to make it convivial for users.