Multiple intelligent algorithms-based analysis and optimization of a CCHP Brayton system considering different turbomachinery designs

The operational performance of a supercritical carbon dioxide Brayton system is affected by the turbomachinery designs. Moreover, assessing the sensitivity of numerous operating parameters on the system performance is challenging due to high computational costs. This paper assesses a combined cooling, heat and power (CCHP) Brayton system based on metrics of energy, exergy, economic, and environmental impact (4E). Variable-efficiency models incorporating iterative calculation strategies for turbomachinery are developed. Based on this, the 4E system performance is compared under different turbomachinery designs. In this process, integrating with multiple intelligent algorithms (LHS sampling, SVM classification, and ANN surrogate), the SOBOL global sensitivity analysis is applied to the system evaluation. The optimization curve and decision stability regarding the system performance are analyzed. The results indicate that different turbomachinery designs can significantly affect the system exergy efficiency, with evaluation fluctuations exceeding 30% in the exergy destruction of compressor and solution pump. Furthermore, calculating first-order and total-order sensitivity indices with a sample size of 2×105 can take below 300 s, and the quantitative indices align well with the system performance evaluation. Finally, different turbomachinery designs lead to noticeable shifts in the Pareto front and decision points, consistent with the system performance analysis. Levelized cost of energy and environment emissions of the system can be reduced by up to 4.2% and 22.6%, respectively. These optimization results provide design guidance for the Brayton system.