风电的数字孪生模型使用与组合策略

Use and Combination Strategy of Digital Twin Model for Wind Power

为解决风电产电波动大、运维成本高、状态监测与故障诊断复杂、布局仿真计算效率低等问题,在风电数字孪生系统的框架下,总结并讨论基于人工智能技术的数据驱动模型与机理仿真模型的各自特点。具体分析2类建模方法在建模机理、数据需求、建模精度与计算效率、模型表现形式和典型模型算法等方面的优势与不足。进一步地,通过辨析二者关联,提出5种关于数据驱动模型与机理仿真模型的组合使用策略。最后,对风电预测、状态监测、故障诊断、优化控制和布局优化5项典型风电应用问题提出各自对应的模型组合使用方法,以综合提升风电数字孪生虚拟模型的精度及计算效率。该文的数字孪生模型使用与组合策略对与风机类似的装备有推广意义。

In order to solve the problems of large fluctuation of wind power generation,high cost of operation and maintenance,complexity of condition monitoring and fault diagnosis,and low efficiency of layout simulation,under the framework of wind power digital twin system,the respective characteristics of data-driven model and mechanism simulation model based on artificial intelligence technology are summarized and discussed.The advantages and disadvantages of the two kinds of modeling methods in modeling mechanism,data requirements,modeling accuracy and computational efficiency,model expression,typical model algorithm and so on are analyzed in detail.Furthermore,by analyzing the relationship between the two,five strategies for the combined use of data-driven model and mechanism simulation model are proposed.Finally,the corresponding model combination methods are proposed for five typical wind power application problems,such as wind power prediction,condition monitoring,fault diagnosis,optimal control and layout optimization,in order to comprehensively improve the accuracy and computational efficiency of wind power digital twin virtual model.The use and combination strategy of the digital twin model in this paper is of great significance to the equipment similar to the fan.