This article focuses on cleaning scheduling for photovoltaic (PV) systems in the field. A method to design cleaning schedules for PV strings is proposed to optimize the profit of a PV system over an extended period. Previous studies have usually focused solely on the cleaning frequency on experimental systems. Our method designs specific cleaning schedules for PV systems in the field based on the forecasting of environmental conditions, PV power generation, and dust deposition. We propose that the cleaning decision is an infinite scheduling problem. A problem model in the infinite time horizon is developed and a periodic scheduling method based on forecasting is proposed to transform the infinite process to an optimizable problem. The cleaning schedules are amended with time as the forecasting accuracy for future times increases with time. The performance of our proposed method that operates under various conditions is evaluated in a case study and compared with two classical scheduling methods.

中文翻译：

---

基于发电量和粉尘沉积预测的现场光伏系统优化清洁调度

本文重点介绍现场光伏 (PV) 系统的清洁调度。提出了一种设计光伏组串清洁计划的方法，以优化光伏系统在较长时期内的利润。以前的研究通常只关注实验系统的清洁频率。我们的方法基于对环境条件、光伏发电和灰尘沉积的预测，为现场光伏系统设计特定的清洁计划。我们提出清洁决策是一个无限调度问题。建立了无限时间范围内的问题模型，提出了基于预测的周期性调度方法，将无限过程转化为可优化问题。随着未来时间的预测准确性随着时间的推移而增加，清洁计划会随着时间的推移而修改。