光伏监控中的故障诊断方法及应用

发布时间：2019-03-24 20:39

【摘要】：太阳能具有用之不竭、环境友好等优势。从长远来讲,光伏太阳能发电必将成为人类获取能源的主要渠道之一。由于太阳能电站有分布广、位置偏僻、维护费用高等特点,因此各种规模光伏电站的实时监控与故障诊断成为行业难点。因此设计一款适用于光伏电站的故障诊断专家系统是非常必要的。本文提出一种改进的概率故障树分析方法。针对太阳能行业监控与故障诊断技术落后的现状,应用改进的概率故障树分析方法,结合IF-THEN推理规则和太阳能系统专家经验,提出一类面向光伏监控与故障诊断的专家系统理论。将此方法应用到本团队自制微型太阳能电站系统进行实验验证,实验结果表明概率故障树分析法能够反映实际系统的不确定性,能够提升专家系统的故障诊断准确度。在文中建立了光伏电站中发电效率降低的概率故障树,最后将改进的概率故障树与专家系统相结合,设计了以正反向混合的推理方向、基于概率知识的模糊推理方法、改进的“滤波”启发式搜索的专家系统。这样的专家系统在最后给出的故障原因可能有多个,提高了专家系统的诊断正确率。最后通过编程实现了光伏故障诊断软件的开发。
[Abstract]:Solar energy has inexhaustible and environmentally friendly advantages. In the long run, photovoltaic solar power generation will be one of the main channels for human to obtain energy. Due to the wide distribution, remote location and high maintenance cost of solar power stations, the real-time monitoring and fault diagnosis of photovoltaic power stations of various sizes has become a difficult problem in the industry. Therefore, it is necessary to design a fault diagnosis expert system for photovoltaic power station. An improved probabilistic fault tree analysis method is proposed in this paper. In view of the backward monitoring and fault diagnosis technology in solar energy industry, the improved probabilistic fault tree analysis method, combined with IF-THEN reasoning rules and the experience of solar system experts, is applied. An expert system theory for photovoltaic monitoring and fault diagnosis is presented. The experimental results show that the probabilistic fault tree analysis method can reflect the uncertainty of the actual system and improve the fault diagnosis accuracy of the expert system. In this paper, a probabilistic fault tree is established to reduce the efficiency of photovoltaic power plants. Finally, the improved probabilistic fault tree is combined with the expert system, and a fuzzy reasoning method based on probabilistic knowledge is designed based on the mixed direction of both positive and negative directions. Improved "filter" heuristic search expert system. There may be more than one fault cause given by the expert system at the end of the paper, which improves the diagnostic accuracy of the expert system. Finally, the development of photovoltaic fault diagnosis software is realized by programming.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM615;TP277

【参考文献】