Performance investigation of the cross-flow closed-type heat-source tower using experiments and an adaptive neuro-fuzzy inference system model

Abstract In this study, a cross-flow closed-type heat-source tower (CCHT) was designed and experiments were carried out under low ambient temperature conditions. Here, a robust heuristic approach for an analysis of the performance characteristics of the CCHT was proposed. The suggested approach consisted of an Adaptive Neuro-Fuzzy Inference System (ANFIS), which was developed and validated by conducting numerous experiments. To address the problem of balancing the predictive accuracy of the model against the number of input variables, the Non-dominated Sorted Genetic Algorithm II (NSGA II) was used. In addition, by considering the optimized number of input variables, a comparative study was conducted with two other models. The integration of the ANFIS and NSGA II was proposed as the foremost model for evaluating the performance characteristics of CCHTs. Ultimately, a sensitivity analysis was carried out by utilizing the suggested predictive model to provide a comprehensive vision of the CCHT for improved understanding. For instance, it was observed that with the increase in the air inlet temperature, air inlet humidity ratio, and solution inlet temperature, the air outlet humidity ratio gradually increased, while the increase in the solution concentration led to a decrease in the air outlet humidity ratio.