面向智能电网多元储能系统的信息综合利用及自学习研究

The integration of large scale renewable power has significant influence on the operation economy、security、stability and reliability of the power systems, and it has been widely recognized as one of the technical challenges in smart grid applications. The energy storage system (ESS) is a key technology to address this problem. In recent years, ESSs employed in the power system are evolving from conventional single-type energy storage technology to modern hybridized ESS that consists of multi-type energy storage mediums interconnected in a multi-layer structure, the scale is increasing rapidly, and its monitoring, control and optimal operation is becoming more and more challenging. This research project proposes a novel ‘ESS hologram’ technology, based on which novel algorithms for ESS control and dispatch optimization will be developed. The research will address the following key issues: 1) modelling of the complex ESS system; 2) ESS predictive control; and 3) multi-objective dispatch optimization algorithms. Firstly, the physical mechanism of the multi-layer multi-type ESS and the fault evolving logics will be studied. Then, we propose to use advanced modelling techniques to estimate the key internal states of the ESS with incomplete measurements, aiming to develop the hologram of the ESS by combining the prior knowledge of system physical mechanism, and data-driven system analysis algorithms. Further, advanced meta-heuristic optimization methods will be used for ESS real-time protection and dispatch optimization. This research project will lay a solid foundation for the theoretical and technological development to ensure high operation efficiency of ESSs in the smart grid applications.

大规模可再生能源的发电并网对于网络运行的经济性、安全性、稳定性及可靠性影响巨大，是智能电网应用中公认的技术难题。储能系统是解决这一难题的关键技术之一。近年来，储能系统的应用正从单一类型走向多种储能类型分层混合接入的模式，规模不断扩大，其监控及优化利用更加具有挑战性。本项目研究复杂储能系统全息技术及实时监控与优化控制问题，就复杂储能系统的建模、预测控制以及多目标优化自愈调控等关键问题展开研究。首先从机理的角度研究储能系统的动态过程和演化规律，解决信息不完全量测条件下储能系统的内部状态估计。结合机理分析与数据驱动技术，对储能系统的全息状态进行估计，进一步结合启发式优化方法及预测控制对储能系统实施控制及优化调度。本项目研究成果将为复杂储能系统在智能电网中的高效运行奠定初步的理论与技术基础。

集成大规模可再生能源对经济、安全、稳定以及电力系统的可靠性等方面至关重要，同时也被广泛认识到其在智能电网中的技术挑战性。能量存储系统ESS是一个解决此类问题的关键技术。本项目主要面向智能电网多元储能系统的信息综合利用及自学习研究，研究了复杂储能系统全息技术及实时监控与优化控制问题，就复杂储能系统的建模、预测控制以及多目标优化自愈调控等关键问题展开研究。首先从机理的角度研究储能系统的动态过程和演化规律，解决了信息不完全量测条件下储能系统的内部状态估计。结合机理分析与数据驱动技术，对储能系统的全息状态进行估计，进一步结合启发式优化方法及预测控制对储能系统电网集成实施控制及优化调度。本项目研究成果将为复杂储能系统在智能电网中的高效运行奠定初步的理论与技术基础。