考虑积尘污染效应的光伏电站选址：一种基于不确定性与可靠性的新型混合框架及最优清洗调度

Abstract One of the main challenges faced by solar plants is soiling issues. Establishing solar plants in suitable locations can significantly reduce the soiling effects. In this study, Bayesian Best-Worst method is proposed as a probabilistic model , along with the COCOSO based on uncertainty and reliability ( Z -numbers). For robust evaluation, the results are compared with five other decision-making methods (R 2 > 0.98). Finally, the soiling effects at the selected site were assessed, and the optimal cleaning schedule was determined based on three types of cleaning: manual, semi-automatic, and fully automatic. For the case study , the Khuzestan region, one of the areas heavily affected by dust storms in the Middle East, was examined. The results showed that the frequency of dust events has the highest impact, and the Mahshahr site is the most suitable location for solar plant. Additionally, rainfall was found to be an effective natural cleaning process during the cold season, but in the warm seasons, the soiling ratio decreased to 90 %. The optimization results for cleaning solar panels indicated 2, 4, and 13 cleanings per year are required for manual, semi-automatic, and automatic technologies, respectively. This assessment showed the proposed framework have strong impact on mitigation of soiling issues.