基于两层随机过程嵌套的光伏-储能系统容量优化研究

Research on Optimal Sizing of Photovoltaic-Energy Storage System Based on Two-Layer Nested Stochastic Processes

中文关键词:不确定性  光储系统  容量优化设计  马尔可夫模型  核密度估计

keywords:uncertainty  photovoltaic-energy storage system  capacity optimization design  Markov chain model  kernel density estimation

中文摘要:针对辐照度不确定性对含太阳能综合能源系统容量优化设计的影响，提出一种基于数据驱动的太阳辐照随机生成模型。该方法由两层随机过程嵌套组成，分别描述了日辐照强度水平与日内辐照随机变化规律。引入晴空指数来描述各时刻辐照度水平，根据日内晴空指数的分布构建日晴空状态；基于马尔可夫模型对日晴空状态随机转化进行建模，并通过卡方检验来验证其有效性。通过扩散核密度估计方法，拟合得到每种状态下的晴空指数分布；提出的生成模型可以快速生成大量的同时考虑了日辐照强度演化与日内辐照波动的辐照度序列；在对系统进行容量优化设计的同时，可对候选尺寸进行大量的技术经济分析，为系统设计提供数据支撑。基于住宅光伏-储能系统，对所提出的太阳辐照度生成模型进行了验证分析。

Abstract:To address the impact of irradiance uncertainty on the optimal sizing of solar-integrated energy systems，this paper proposes a data-driven stochastic generation model for solar irradiance. The model consists of two nested stochastic processes，which respectively characterize the daily irradiance intensity levels and the stochastic intraday irradiance variation patterns. The clearness index （CI） is introduced to quantify the irradiance level at each time step，and daily clearness states （DCS） are constructed based on the distribution of intraday CI statistics. A Markov chain model is employed to model the stochastic transitions of DCS，and the chi-square test is utilized to verify its validity. The diffusion Kernel density estimation （DKDE） method is applied to fit the CI distributions corresponding to each DCS. The proposed generation model can efficiently generate a large number of irradiance sequences that simultaneously account for both daily irradiance intensity evolution and intraday irradiance fluctuations. During the optimal sizing of the system，it enables extensive techno-economic analysis of candidate configurations，providing data support for system design. The proposed solar irradiance generation model is validated through a case study of a residential photovoltaic-energy storage system.