Distributed optimization of electricity-Gas-Heat integrated energy system with multi-agent deep reinforcement learning

The coordinated optimization problem of the electricity-gas-heat integrated energy system(IES)has the characteristics of strong coupling,non-convexity,and nonlinearity.The centralized optimization method has a high cost of communication and complex modeling.Meanwhile,the traditional numerical iterative solution cannot deal with uncertainty and solution efficiency,which is difficult to apply online.For the coordinated optimization problem of the electricity-gas-heat IES in this study,we constructed a model for the distributed IES with a dynamic distribution factor and transformed the centralized optimization problem into a distributed optimization problem in the multi-agent reinforcement learning environment using multi-agent deep deterministic policy gradient.Introducing the dynamic distribution factor allows the system to consider the impact of changes in real-time supply and demand on system optimization,dynamically coordinating different energy sources for complementary utilization and effectively improving the system economy.Compared with centralized optimization,the distributed model with multiple decision centers can achieve similar results while easing the pressure on system communication.The proposed method considers the dual uncertainty of renewable energy and load in the training.Compared with the traditional iterative solution method,it can better cope with uncertainty and realize real-time decision making of the system,which is conducive to the online application.Finally,we verify the effectiveness of the proposed method using an example of an IES coupled with three energy hub agents.

Optimal decision fusion given sensor rules

When all the rules of sensor decision are known, the optimal distributeddecision fusion, which relies only on the joint conditional probability densities, can be derivedfor very general decision systems. They include those systems with interdependent sensorobservations and any network structure. It is also valid for m-ary Bayesian decision problems andbinary problems under the Neyman-Pearson criterion. Local decision rules of a sensor withcommunication from other sensors that are optimal for the sensor itself are also presented, whichtake the form of a generalized likelihood ratio test. Numerical examples are given to reveal someinteresting phenomena that communication between sensors can improve performance of a senordecision, but cannot guarantee to improve the global fusion performance when sensor rules were givenbefore fusing.

Distributed control and optimization of process system networks:A review and perspective

Large-scale and complex process systems are essentially interconnected networks.The automated operation of such process networks requires the solution of control and optimization problems in a distributed manner.In this approach,the network is decomposed into several subsystems,each of which is under the supervision of a corresponding computing agent(controller,optimizer).The agents coordinate their control and optimization decisions based on information communication among them.In recent years,algorithms and methods for distributed control and optimization are undergoing rapid development.In this paper,we provide a comprehensive,up-to-date review with perspectives and discussions on possible future directions.

Multicriteria Methods for Distributed Generation Resources Optimization

The optimization process of embedded, or DG （distributed generation） is a very complex task, and it should be evaluated and compared by means of multi-criteria methods of analysis. The classical method for selection is usually based only on a single criterion analysis, and it is defined by thermal or economic aspects. The problem of optimal dispatch of DG is typical example of optimization, because it differs from the classical problem of generation dispatch in the power system, due to the specific criteria related to the DG interconnection. The most important goals are to maximize the renewable production and to minimize the total cost, while satisfying additional constraints related to the operation of a distribution network. As there are many DGs in a distribution network, it is very complicated to decide the optimal DG outputs to satisfy all the criteria and constraints imposed by the distribution network. Another problem is the lack of the dispatcher control over DGs, and very often, the only available action is to switch on or off the generator. Finally, network operator and DG owner perspective are often opposed regarding appropriate control action in the network. In this paper, a multicriteria decision support based on AHP （analytical hierarchical processes） method is proposed for the choice of the dispatching action. The method is illustrated on the choice of the DG to be switched off in the case or reverse power flow.

Distributed tasks-platforms scheduling method to holonic-C2 organization

To solve the problem of distributed tasks-platforms scheduling in holonic command and control(C2) organization,the basic elements of the organization are analyzed firstly and the formal description of organizational elements and structure is provided. Based on the improvement of task execution quality,a single task resource scheduling model is established and the solving method based on the m-best algorithm is proposed. For the problem of tactical decision-holon cannot handle tasks with low priority effectively, a distributed resource scheduling collaboration mechanism based on platform pricing and a platform exchange mechanism based on resource capacities are designed. Finally,a series of experiments are designed to prove the effectiveness of these methods. The results show that the proposed distributed scheduling methods can realize the effective balance of platform resources.

Two phase decision algorithm of replica allocation

In distributed parallel server system, location and redundancy of repficas have great influence on availability and efficiency of the system. In order to improve availability and efficiency of the system, two phase decision algorithm of replica allocation is proposed. The algorithm which makes use of auto-regression model dynamically predicts the future count of READ and WRITE operation, and then determines location and redundancy of replicas by considering availability, CPU and bands of the network. The algorithm can not only ensure the requirement of availability, but also reduce the system resources consumed by all the operations in a great scale. Analysis and test show that communication complexity and time complexity of the algorithm satisfy O（n）, resource optimizing scale increases with the increase of READ count.

Conjunctive Use of Engineering and Optimization in Water Distribution System Design

Water Distribution Systems (WDSs) design and operation are usually done on a case-by-case basis. Numerous models have been proposed in the literature to solve specific problems in this field. The implementation of these models to any real-world WDS optimization problem is left to the discretion of designers who lack the necessary tools that will guide them in the decision-making process for a given WDS design project. Practitioners are not always very familiar with optimization applied to water network design. This results in a quasi-exclusive use of engineering judgment when dealing with this issue. In order to support a decision process in this field, the present article suggests a step-by-step approach to solve the multi-objective design problem by using both engineering and optimization. A genetic algorithm is proposed as the optimization tool and the targeted objectives are: 1) to minimize the total cost (capital and operation), 2) to minimize the residence time of the water within the system and 3) to maximize a network reliability metric. The results of the case study show that preliminary analysis can significantly reduce decision variables and computational burden. Therefore, the approach will help network design practitioners to reduce optimization problems to a more manageable size.

考虑高温超导电缆接入的配电网变电站及网架协同规划

城市配电网面临着用电需求增长而土地等规划资源受限的困境,随着高温超导电缆(简称超导电缆)成本下降及运行技术的逐渐成熟,为提升城市配电网的供电能力、保障供电需求提供了解决措施。超导电缆系统与常规输电线路在应用场景与技术经济特性上存在较大差异,需考虑超导电缆纳入的差异性,才能充分发挥其输电能力强、损耗低和截面小的优势,为此提出了一种考虑超导电缆接入的配电网变电站及网架协同规划方法。基于主从决策和全寿命周期成本的理念,通过分析超导电缆接入配电网后系统变电站配置和网架布局间的耦合关系,综合考虑经济性要求与可靠性约束,建立了考虑超导电缆接入的变电站-网架协同规划模型。采用基于多种策略改进后的鲸鱼优化算法对模型进行求解,得到含超导电缆的配电网规划方案,并通过算例验证了所提方法的有效性。

基于可解释强化学习的智能虚拟电厂最优调度

随着电动汽车的不断普及,能源系统日益复杂。虚拟电厂(VPP)可以通过物联网和人工智能技术,将分布式电源、储能系统、可控负荷以及EV等分布式能源进行聚合和协调优化,有助于提升能源的使用效率,并促进非可再生能源的消纳,增强电网稳定性。现阶段人工智能技术在电力系统等安全要求较高的应用领域缺乏可靠性和透明度,可能导致用户和运营商难以理解算法如何做出特定的能源调配决策。针对人工智能技术下的VPP实现最优调度并兼顾解释其决策过程的平衡问题,提出一种可解释强化学习的交互式框架,使用近端策略优化算法实现VPP的最优调度,并使用决策树建立一种可解释性强化学习框架,用于提供透明的决策支持,使非专业用户能够理解人工智能在调节能源系统方面的决策过程。试验表明,与传统强化学习优化方法相比,该方法不仅提高了能源分配的效率,而且通过增强模型的可解释性,加强了用户对智能VPP管理系统的信任。

考虑光伏电源可靠性的新能源配电网数据驱动无功电压优化控制

充分挖掘分布式光伏电源的无功支撑能力,有助于解决光伏高比例接入带来的配电网电压波动、电压越限以及新能源消纳等问题,但光伏电源无功输出会造成其功率器件结温越限或剧烈波动,严重威胁到光伏电源的可靠运行。为此,提出考虑光伏电源可靠性的新能源配电网数据驱动无功电压优化控制策略。首先,提出一种基于数据驱动的光伏电源可靠性评估方法,该方法采用XGBoost机器学习模型计算IGBT结温,提高了IGBT结温计算效率,避免了评估精度对IGBT参数的依赖;进而建立考虑光伏电源可靠性的配电网无功电压优化模型,将IGBT结温均值和结温波动引入模型优化目标;然后,将该模型进行马尔可夫决策过程转化,并基于深度确定性策略梯度强化学习算法完成智能体训练;最后,通过IEEE33节点系统验证所提策略在无功电压快速优化和光伏电源可靠性提升方面的优势。

A New Proof for the Tight Range of Optimal Order Quantities for the Newsboy Problem with Mean and Standard Deviation

In the classical Newsboy problem, we provide a new proof for the tight range of optimal order quantities for the newsboy problem when only the mean and standard deviation of demand are available. The new proof is only based on the definition of the optimal solution therefore it is the most straightforward method. It is also shown that the classical Scarf’s rule is the mid-point of the range of optimal order quantities. This provides an additional understanding of Scarf’s order rule as a distribution free decision.

大型炼化一体化过程全流程低碳运行分析与决策优化综述

石化工业是国民经济支柱产业,也是高能耗高排放行业。面向双碳背景下传统石化行业低碳化、智能化转型需求,针对碳排与工艺耦合机理复杂的炼化一体化全流程碳足迹表征精度低问题、耦合互联多装置生产流程碳排异常环节难追溯问题以及实际工业环境下全流程优化方案不收敛问题,本文以大型炼化一体化全流程为研究对象,描述了利用过程机理、工业大数据和人工智能技术,研究炼化一体化全流程碳足迹建模理论、全流程碳排异常环节溯源理论以及全流程装置分布式协同决策优化理论,实现炼化一体化全流程低碳优化。在为流程工业碳足迹建模、溯源和优化提供方法的同时,也为实际炼化一体化企业高效、低碳、智能生产提供指导建议。

基于模糊结构元的新零售末端配送网络优化研究

新零售模式下,高时效性的物流末端配送对提高顾客满意度具有重要作用,而对末端配送中心位置和数量的确定,是优化末端配送网络的关键。首先,综合考虑影响配送中心位置与数量的模糊因素,构建以配送时间最少、配送中心覆盖范围最大以及系统成本最低为目标的模糊多目标规划模型;其次,采用结构元加权序准则法求解模型,生成三角模糊数简化求解过程;最后,以算例检验方法的有效性。结果表明,所提出的模型能够将模糊因素融入模糊决策过程中,且结构元加权序准则法有助于决策者灵活调整方案和决策,从而对新零售末端配送网络布局进行优化。

考虑风电不确定性的电力系统在线动态分区恢复优化方法

规模风电的接入给电力系统运行与控制带来了很大不确定性,也给大停电后的系统恢复带来挑战。根据恢复过程中变化的风电出力场景动态调整分区恢复方案有助于提升恢复效率。在计及初始停电场景中风电不确定性的基础上,为进一步考虑恢复过程中风电出力的不确定性,提出了一种电力系统在线动态分区恢复优化方法。首先,建立风电出力的不确定场景集合,基于Wasserstein距离构建分布之间的测度,采用核密度估计求取风电出力预测误差的不确定集合。然后,刻画恢复模型约束、分区模型约束、动态分区约束,分别从系统网架和运行状态两个角度设立两阶段优化目标,建立两阶段动态分区恢复分布鲁棒优化模型,并采用对偶理论等实现模型的转化与求解。最后,新英格兰10机39节点系统和实际电网算例的仿真结果表明所提动态分区恢复方法能有效应对风电出力不确定性和提高系统恢复效率。

量子计算技术在新型电力系统决策优化中的应用

新型电力系统的规划、运行和市场运营等决策优化过程呈现变量激增、约束繁杂等特点,而量子计算具有运算并行和状态叠加等特性,为高效解决此类“维数灾难”难题提供了新的技术路径。文中围绕量子计算技术赋能新型电力系统决策优化的原理可行性及实现思路展开探析。首先,梳理分析量子计算应用于新型电力系统决策优化过程的先进性与局限性,构建量子-经典计算混合的变分量子决策优化框架。在此基础上,提炼新型电力系统典型优化问题的共性,推导统一的问题结构,形成可利用量子比特系统描述的能量模型。随后,提出基于量子近似优化算法的求解流程,寻找能量模型的极值,并映射得到原优化问题的最优解。最后,从软硬件、算法框架以及行业发展等角度提出思考与展望。

考虑移动电氢资源多区互济的负荷恢复方法

电-氢-交通协同可以提高灵活性并增强能源系统韧性。提出了一种考虑移动电氢资源多区互济的负荷恢复方法。针对分布在多个区域的电-氢耦合系统,提出在停电期间协同利用本地的分布式电源、储氢和移动电氢资源,快速恢复关键用户需求。通过基于交通地图系统的移动电氢资源调度,提出了负荷恢复和移动电氢资源调度集成的故障恢复决策框架。进而构建了移动电氢资源交通调度与负荷恢复的耦合约束,建立了移动电氢资源参与恢复的混合整数二阶锥规划模型,可以进行恢复与交通调度的统一决策。所提方法从时间和空间维度协调和分配电氢资源,最大限度地恢复关键用户的需求。通过构建的多区域电-氢耦合系统和相对应的交通系统,验证了所提协同恢复方法的有效性和优越性。

决策者期望下虚拟孵化生态系统的群体最优利益分配模型

基于生态系统的群体决策者对其孵化项目所获得的收益预期(分为乐观和悲观两类),构建了决策者期望下虚拟孵化生态系统的群体最优利益分配模型,并给出了考虑群体加入意愿的最优利益分配约束条件。最后,通过洪泰智能制造案例分析了群体最优利益分配对其乐观(和悲观)期望值的敏感性。结果表明:在某群体持乐观期望情形下,其最优利益分配为孵化项目所得总利益的线性增函数,为联盟所有成员乐观期望值总和的线性减函数;在边际贡献相同情况下,群体最优利益分配从大到小依次为孵化器、高校或科研机构、创投机构和中介机构。结论不但弥补了“互联网+”下虚拟孵化器的模型分析理论,而且对虚拟孵化器的创新决策具有现实指导意义。

地震灾害下两阶段多目标应急避难救援策略研究

为了系统化应急避难救援活动,并不断完善应急避难场所救援功能和提高灾后物资供应效率,以地震灾害为现实背景,综合灾前准备和灾后响应两阶段,通过多准则决策方法评价应急避难场所服务质量。构建以最小化总救援成本期望值、总受灾群众疏散距离和总物资运输距离以及最大化应急避难服务质量、总物资需求满足率为目标的两阶段决策模型。以2013年雅安地震为背景,验证两阶段决策模型的有效性,完成应急避难场所和属地众储点预选址以及应急物资预储工作,确定灾后不同类型紧急避难场所位置并对需求点动态分配应急物资。研究结果表明,灾前配置的应急避难资源为震后展开避难救援活动奠定基础,加快灾后响应速度和提高工作质量,体现对避难人员救援的重视并起到受灾群众生命安全保障的重要作用。

基于深度确定性策略梯度算法的配电网最优电压实时控制

随着光伏发电在配电网中的渗透率逐渐增大,在降低系统网损和全社会的碳排放量的同时,也导致了电压出现时段性越限等问题,而电压安全对配电网的稳定运行有重要意义。提出了一种基于深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法的电压控制策略。研究了太阳能光伏逆变器在配电网无功电压优化中的作用;以配电网有功损耗最小化为目标函数,同时考虑到逆变器的无功补偿能力,提出了一种基于深度确定性策略梯度算法的配电网电压控制策略;利用修改后的IEEE33节点算例对所提策略的有效性进行验证,仿真结果表明:DDPG算法学习所得策略可以动态调节各光伏逆变器的无功输出,从而实现控制电压安全的目标,并且与调控前相比系统网络耗损减少了13.5%。

基于矩不确定性的虚拟发电厂分布鲁棒机组组合

由于分布式能源输出的随机性使得虚拟发电厂难以向独立系统运营商提供具体物理参数。为此,本文建立基于高阶矩模糊集的分布鲁棒优化模型来评估虚拟发电厂的物理特性。首先采用虚拟净负荷的矩信息(例如均值和协方差)来描述风电输出功率和负荷需求的不确定性;然后,在第1阶段优化虚拟发电厂的机组组合,在第2阶段优化机组出力和储能充放电,建立虚拟发电厂的两阶段分布鲁棒优化模型;最后,结合对偶变换将虚拟发电厂分布鲁棒机组组合模型转化为二次二阶锥规划模型,便于采用两阶段迭代求解。通过案例分析验证了本文所建立的虚拟发电厂分布鲁棒机组组合模型的有效性。