市场环境下智能配用电系统分层协同优化运行:研究挑战、进展与展望

随着分布式资源在配电网中的比例不断提高,如何在市场化交易机制下实现配用电系统安全经济运行成为当下的研究热点。在市场环境下,配用电系统各层的运行管理面临着不确定性逐层加剧、市场规模快速扩展、市场交易与系统安全运行难以有效衔接等多重挑战。该文首先梳理市场环境下配用电系统运行优化的关键问题;其次,对传统解析优化方法的研究成果与研究中仍待解决的问题进行总结;然后,针对配用电系统市场交易、运行优化等问题特点,系统性地介绍深度强化学习技术,分析归纳深度强化学习在配用电系统中的研究现状。最后,提炼出贯穿配用电系统多层多主体协同优化问题中的三重研究需求,并对深度强化学习技术未来的应用路径与发展趋势进行展望。

电-碳耦合市场环境下可再生能源投资规划优化方法

中国面临着可再生能源高速发展和电力市场改革、碳市场建设稳步推进的多重机遇,市场化机制将科学引导可再生能源合理规划与有效投资。电力市场和碳市场虽分开运营,两者的出清结果却紧密耦合,且共同影响市场主体的运营与投资模型。在电-碳耦合市场的背景下,文中提出了一种最大化可再生能源接入容量并保障其投资成本回收的可再生能源投资规划双层优化模型。首先,基于动态碳排放特性推导了常规机组的电-碳耦合报价,提出了考虑碳配额交易影响的电力现货市场机制。然后,在“投资-运行”框架下建立了可再生能源投资规划双层优化模型:上层在能收回投资成本的前提下,最大化可再生能源的投资容量并对其进行定容和选址决策;下层分别模拟电力现货市场和国家核证自愿减排量市场的出清过程,以向上层模型的可再生能源发电商投资决策提供相应的价格信号。通过下层问题最优性条件和线性化方法,将双层优化模型转化为混合整数线性规划进行求解。最后,通过仿真算例验证了所提方法有效性,并探索了碳市场交易、网络容量与储能等不同因素对可再生能源最优投资容量的影响。

欧洲跨国电力市场的输电机制与耦合方式

中国电力市场改革已经取得初步成效,但发电侧市场集中度大、清洁能源消纳困难等现实矛盾亟待通过更大范围的资源优化配置解决,借鉴和参考欧洲跨国电力市场的运行经验对于中国下一步建设统一电力市场具有重要意义。首先,面向中国推进统一电力市场所面临的跨省区通道容量如何分配、输电成本如何回收以及省内外交易如何衔接等问题,针对性地研究了欧洲跨国电力市场的输电管理机制及市场耦合机制。其次,以欧洲电力交易所EPEX SPOT为例,描述了耦合市场的交易流程和交易衔接方式。最后,结合中国当前电网建设背景及市场运行实际,基于欧洲跨国电力市场运行经验,提出推进统一电力市场的意见和建议。

抑制柔性负荷过响应的微网分散式调控参数优化

基于价格的微网分散式调控相较于集中式调控具有扩展性和保密性优势,但存在柔性负荷在低电价时段大量聚集产生新峰值的过响应问题,影响系统安全高效运行。现有研究在主参数电价之外引入辅助参数以缓解过响应,但未进行辅助参数取值的优化,且未考虑微网网络约束对不同节点辅助参数最优取值的影响。为此该文提出缓解负荷过响应的微网分散式调控辅助参数优化方法,为不同节点的柔性负荷制定参数最佳取值以最小化微网总运行成本。首先建立微网分散式调控优化模型,以及电动汽车、智能家电的柔性负荷需求响应模型;进而提出基于深度强化学习的辅助参数优化方法,采用多维连续状态及动作空间学习各节点参数的取值。最后仿真结果验证了所提优化方法的有效性。

基于深度强化学习的居民实时自治最优能量管理策略

随着居民分布式资源的普及,如何考虑用户多类型设备的运行特性,满足实时自治能量管理需求以达到用户侧经济性最优成为亟待解决的课题。传统基于模型的最优化方法在模型精准构建和应对多重不确定性等方面存在局限性,为此提出一种无模型的基于深度强化学习的实时自治能量管理优化方法。首先,对用户设备进行分类,采用统一的三元组描述其运行特性,并确定相应的能量管理动作;接着,采用长短期记忆神经网络提取环境状态中多源时序数据的未来走势;进而,基于近端策略优化算法,赋能在多维连续-离散混合的动作空间中高效学习最优能量管理策略,在最小化用电成本的同时提升策略对不确定性的适应性;最后,通过实际情境对比现有方法的优化决策效果,验证所提方法的有效性。

基于参数共享机制多智能体深度强化学习的社区能量管理协同优化

智能电网背景下社区和端对端电能交易有助于挖掘利用产消者分布式能源的灵活性并最大化其价值。尽管多智能体深度强化学习提供了合适的无模型框架以实现多个产消者间能量管理策略的协同优化,该方法仍存在环境状态不稳定、产消者隐私保护和计算复杂度高等局限。该文提出一种将参数共享与优先深度确定性策略梯度法相结合的多智能体强化学习方法,通过智能体间的策略与经验共享以提升学习效率,并降低训练难度。接着构建端对端交易平台以协同社区市场内产消者的电能交易;执行奖励修正以避免产生新的负荷/发电高峰,从而保护本地配网的安全运行;作为可信任第三方向产消者提供有关社区市场的全局信息,在保护产消者隐私的同时减轻环境不稳定性,并提升算法的可扩展性。最后,通过算例验证所提方法能够有效降低社区总运行成本,保证产消者的利益,且较现有算法提高了训练速率与可扩展性。

Optimal Offering of Energy Storage in UK Day-Ahead Energy and Frequency Response Markets

The offering strategy of energy storage in energy and frequency response(FR) markets needs to account for country-specific market regulations around FR products as well as FR utilization factors, which are highly uncertain. To this end, a novel optimal offering model is proposed for stand-alone price-taking storage participants, which accounts for recent FR market design developments in the UK, namely the trade of FR products in time blocks, and the mutual exclusivity among the multiple FR products. The model consists of a day-ahead stage, devising optimal offers under uncertainty, and a real-time stage, representing the storage operation after uncertainty is materialized. Furthermore, a concrete methodological framework is developed for comparing different approaches around the anticipation of uncertain FR utilization factors(deterministic one based on expected values, deterministic one based on worst-case values, stochastic one, and robust one), by providing four alternative formulations for the real-time stage of the proposed offering model, and carrying out an out-of-sample validation of the four model instances. Finally, case studies employing real data from UK energy and FR markets compare these four instances against achieved profits, FR delivery violations, and computational scalability.

Communication Resources Allocation for Time Delay Reduction of Frequency Regulation Service in High Renewable Penetrated Power System

The high renewable penetrated power system has severe frequency regulation problems.Distributed resources can provide frequency regulation services but are limited by com-munication time delay.This paper proposes a communication resources allocation model to reduce communication time delay in frequency regulation service.Communication device resources and wireless spectrum resources are allocated to distributed resources when they participate in frequency regulation.We reveal impact of communication resources allocation on time delay reduction and frequency regulation performance.Besides,we study communication resources allocation solution in high renewable energy penetrated power systems.We provide a case study based on the HRP-38 system.Results show communication time delay decreases distributed resources'ability to provide frequency regulation service.On the other hand,allocating more communication resources to distributed resources'communica-tion services improves their frequency regulation performance.For power systems with renewable energy penetration above 70%,required communications resources are about five times as many as 30%renewable energy penetrated power systems to keep frequency performance the same.Index Terms-Communication resources allocation,commun-ication time delay,distributed resource,frequency regulation,high renewable energy penetrated power system.

Resilience Enhancement of Urban Energy Systems via Coordinated Vehicle-to-grid Control Strategies

Coordinated vehicle-to-grid(V2G)control strategies can sustain essential loads of an energy system during islanding,thereby increasing resilience.In this context,this paper investigates the resilience enhancement benefits of smart V2G control,the value of electric vehicle(EV)owner cooperation on system resilience,as well as the complementary effects of PV and EV interaction in an urban multi-energy microgrid(MEMG).By using a rolling horizon approach to optimize day-ahead operation of the MEMG and subsequently dispatching EVs,uncertainties in outage start time,EV arrival/departure times,and initial state of charge(SOC)are mitigated.Results show that smart V2G control can provide a substantial essential load curtailment reduction compared to a non-EV scenario,meanwhile,non-coordinated grid-to-vehicle(G2V)operation was shown to slightly burden the system with a slight increase in non-essential load curtailment.Investigations into the influence of EV cooperation on resilience showed that a high percentage of system-prioritized(SP)EVs could help greatly further reduce essential load curtailment compared to individual-prioritized(IP)EVs.Finally,the complementary benefits of smart V2G control and PV were demonstrated,showing a reduction in both PV and essential load curtailments with increasing numbers of EVs.Overall,the application of smart V2G control,especially with cooperation of EV owners,can drive significant resilience enhancement during islanding,while further benefits can be obtained through having a sufficient number of EVs to utilize high PV penetration.

Risk-based method to secure power systems against cyberphysical faults with cascading impacts: a system protection scheme application

The utilization levels of the transmission network can be enhanced by the use of automated protection schemes that rapidly respond to disturbances. However,such corrective systems may suffer from malfunctions that have the potential to exacerbate the impact of the disturbance. This paper addresses the challenge of jointly optimizing the dispatch of generators and protection settings in this context. This requires a holistic assessment of the cyber(protection logic) and physical(network) systems,considering the failures in each part and their interplay.Special protection schemes are used as a prototypical example of such a system. An iterative optimization method is proposed that relies on power system response simulations in order to perform detailed impact assessments and compare candidate solutions. The candidate solutions are generated on the basis of a security-constrained dispatch that also secures the system against a set of cyber failure modes. A case study is developed for a generation rejection scheme on the IEEE reliability testsystem(RTS): candidate solutions are produced based on a mixed integer linear programming optimisation model, and loss-of-load costs are computed using a basic cascading outage algorithm. It is shown that the partial security approach is able to identify solutions that provide a good balance of operational costs and loss-of-load risks, both in a fixed dispatch and variable dispatch context.

Business cases for energy storage with multiple service provision

Energy storage(ES)has been considered as the key source of flexibility to support the integration of renewable energy.Previous studies have demonstrated the substantial system cost savings by the deployment of ES,including both investment and operation of generation,transmission and distribution infrastructure.However,this societal benefit may not be realized if industry actors do not have a viable business case to appropriately capture these multiple value streams.In this context,this paper investigates the value that ES may deliver to its owner over two specific business cases in a 2030 UK system.Firstly,the application of large-scale ES in the wholesale market is analysed.It is demonstrated that the optimal allocation of ES to provide multiple services is the key element for ES to become competitive in the electricity market.In the second business case,this paper analyses the value of kilowattscale ES combined with roof photovoltaic(PV)system in the household and community level.The study shows that multiple service provision of ES through advanced pricing schemes,for example time-of-use(ToU)tariff and dynamic distribution use of system(DUoS),lead to higher value and the coordination in the community level could further justify the application of domestic ES.

Selecting decision trees for power system security assessment

Power systems transport an increasing amount of electricity,and in the future,involve more distributed renewables and dynamic interactions of the equipment.The system response to disturbances must be secure and predictable to avoid power blackouts.The system response can be simulated in the time domain.However,this dynamic security assessment(DSA)is not computationally tractable in real-time.Particularly promising is to train decision trees(DTs)from machine learning as interpretable classifiers to predict whether the systemwide responses to disturbances are secure.In most research,selecting the best DT model focuses on predictive accuracy.However,it is insufficient to focus solely on predictive accuracy.Missed alarms and false alarms have drastically different costs,and as security assessment is a critical task,interpretability is crucial for operators.In this work,the multiple objectives of interpretability,varying costs,and accuracies are considered for DT model selection.We propose a rigorous workflow to select the best classifier.In addition,we present two graphical approaches for visual inspection to illustrate the selection sensitivity to probability and impacts of disturbances.We propose cost curves to inspect selection combining all three objectives for the first time.Case studies on the IEEE 68 bus system and the French system show that the proposed approach allows for better DT-selections,with an 80%increase in interpretability,5%reduction in expected operating cost,while making almost zero accuracy compromises.The proposed approach scales well with larger systems and can be used for models beyond DTs.Hence,this work provides insights into criteria for model selection in a promising application for methods from artificial intelligence(AI).