Addressing the complexity of future power system dynamic behaviour

电力系统正在经历前所未有的变化，增加了复杂性和不确定性，主要是由实现净零运营和应对气候变化的脱碳目标驱动的。举个例子，朝着这个方向，英国政府制定了一个大胆的目标，即到2035年实现零碳电力。增加的复杂性来自于大量转换器接口器件（CID）的引入，这些器件表现出非常不同的动态行为，在很大程度上受控制。此外，由于可再生能源的间歇性行为，电力系统运行的不确定性正在增加，但电动汽车和潜在的供暖电气化以及复杂的市场和电力工业结构也越来越多地受到社会行为的影响。这导致可能的操作条件和突发事件的搜索空间爆炸式增长，这对计算密集型的稳定性评估和动态研究尤其具有挑战性。这一方面，再加上动态行为日益复杂，使得识别关键操作条件和突发事件变得具有挑战性。因此，这些发展需要改进动态现象的表示和理解，以及快速和信息丰富的动态安全性和稳定性评估。这两个方面都至关重要，以避免潜在的不稳定风险，在最坏的情况下，这些风险可能导致大范围的事件甚至停电。因此，本提案的目的是开发实现安全、有弹性和具有成本效益的电力系统运行所需的方法、工具和模型。在奖学金最初部分取得进展的基础上，扩展将继续侧重于两个主要方向。一方面，它将开发工具，方法和模型来表示和调查电力系统不断变化的动态行为，以捕捉新出现的动态现象，跨越传输和分配（例如海上/陆上风能，太阳能光伏，HVDC链接，电动汽车，热泵，电解槽等）。另一方面，它将开发新的基于机器学习和数据驱动的方法，用于快速和信息丰富的稳定性评估以及稳定性边界的估计。这个方向将使动态行为的独特理解，这将导致辅助服务和控制，以减轻或减轻干扰的影响，提高系统的安全性和弹性。此外，奖学金的扩展将继续并加强与工业合作伙伴的接触，以捕获实际方面并微调开发的方法，为现实世界的应用铺平道路。实际上，由于对系统稳定性限制有了更好的了解，该研究的结果将使电力系统更加安全、有弹性，并且可能更具成本效益。因此，可以实现可再生能源与具有各种技术和环境效益的新技术的更高整合，以便以安全、有弹性和成本效益的方式实现大胆的脱碳目标。

Electrical power systems are undergoing unprecedented changes that increase the levels of complexity and uncertainty, mainly driven by decarbonisation targets on the way to achieving net zero operation and addressing climate change. As an example, towards this direction the UK government has set a bold target for zero carbon electricity by 2035. Increasing complexity comes from the introduction of a large number of converter-interfaced devices (CID) that exhibit very different dynamic behaviour, governed to a large extent by control. In addition, uncertainty in power system operation is increasing, due to the intermittent behaviour of renewable sources but also increasingly by social behaviour through EVs and potential electrification of heating as well as complex market and power industry structures. This leads to an exploding search space of possible operating conditions and contingencies, which is particularly challenging for computationally intensive stability assessment and dynamic studies. This aspect coupled with the increasing complexity of dynamic behaviour, makes identifying critical operating conditions and contingencies challenging. Consequently, these developments raise the need for improved representation and understanding of dynamic phenomena as well as fast and informative dynamic security and stability assessment. Both aspects are crucial in order to avoid potentially hidden risks of instability that in the worst-case scenario can lead to widespread events and even blackouts. Consequently, the aim of this proposal is to develop methods, tools and models needed to achieve a secure, resilient and cost-effective power system operation.Building on progress made in the initial part of the fellowship, the extension will continue focusing on two main directions. From one hand, it will develop tools, methods and models to represent and investigate the changing dynamic behaviour of power systems in order to capture new arising dynamic phenomena, spanning both transmission and distribution (e.g. offshore/onshore wind, solar PVs, HVDC links, EVs, heat pumps, electrolysers, etc.). On the other hand, it will develop novel machine learning based and data-driven methods for the fast and informative stability assessment as well as the estimation of the stability boundary. This direction will enable unique understanding of the dynamic behaviour that will lead to ancillary services and control to mitigate or alleviate the impact of disturbances and improve system security and resilience. In addition, the fellowship extension will continue and ramp-up engagement with industrial partners to capture practical aspects and fine tune developed methodologies to pave the way for real world applications.In effect, the results of the fellowship will enable more secure, resilient and potentially more cost-effective operation of power systems due to better knowledge of system stability limits. Consequently, much higher integration of renewables and new technologies with various technical and environmental benefits can be achieved in order to meet bold decarbonisation targets in a secure, resilient and cost-efficient manner.