基于有偏t分布ARIVIAX模型的短期电价预测

基于正态分布假设的时间序列分析模型不能有效地处理电价的有偏厚尾性,在对电力市场现货电价的影响因素和波动规律综合分析的基础上,提出了一种基于有偏学生t分布ARMAX模型的短期电价预测方法。该方法可同时考虑电价分布的有偏厚尾性、多重周期性及其与负荷之间的非线性相关性。对PJM电力市场历史数据的算例研究表明,该方法计算量小,待估参数少。     

基于有偏t分布ARMAX模型的短期电价预测

基于正态分布假设的时间序列分析模型不能有效地处理电价的有偏厚尾性,在对电力市场现货电价的影响因素和波动规律综合分析的基础上,提出了一种基于有偏学生t分布ARMAX模型的短期电价预测方法.该方法可同时考虑电价分布的有偏厚尾性、多重周期性及其与负荷之间的非线性相关性.对PJM电力市场历史数据的算例研究表明,该方法计算量小,待估参数少.     

PJM电力市场交易价格分布特征及波动风险度量

根据PJM电力市场披露的逐时电力交易价格信息,采用假设检验方法,对日前市场、实时市场及辅助服务市场部分交易品种的电价分布特征进行了统计建模,并针对电价分布的尖峰厚尾特性,结合VaR、CVaR尾部风险度量指标,对各市场中的电价波动风险进行精确度量。研究结果表明：PJM电力市场负荷与日前市场电价、实时市场电价具有强相关性,日前市场及实时市场电价服从对数正态分布,辅助服务价格差异较大,分布拖尾特征显著;在给定置信水平下,负荷、日前市场电价和实时市场电价波动风险逐渐增大,但显著小于辅助服务价格波动风险。     

美国加州电力市场电价的分形特征

以美国加州电力市场长期电价序列为研究对象，并由此引入了分数布朗运动模型和R／S分析法，计算出了不同时间区间内电价的赫斯特指数，发现加州电力市场符合分形市场假设，电价遵循有偏的随机游走．具有长期相关性和统计自相关性等性质。这些发现为描述市场环境下的电价分布提供了一种新模式。     

基于SQP法的电厂日前合约市场报价策略研究

在电力市场环境下,发电企业为了追求经济效益的最大化,必须在优化机组负荷分配的基础上掌握企业运行成本,并在电力市场上报出具有竞争力的上网电价。针对日前市场的特殊性,并建立了机组经济性曲线拟合的数学模型和基于序列二次规划法(SQP法)的机组负荷优化分配模型,指出了完全成本加成法的报价方法。用算例验证了一个发电企业在4个时段日前市场负荷的耗量成本。     

考虑风电并网的分时段短期电价预测

考虑了并网风电量对电价影响,并将相关系数作为选取电价影响因素的标准,考虑了历史电价、负荷、并网风电量与负荷的比值等影响电价的因素。分别将负荷与历史清算电价,等效负荷与历史清算电价,负荷、并网风电量与负荷的比值及历史清算电价作为神经网络的输入因子对市场清算电价进行分时段预测。算例采用丹麦电力市场的历史数据,分别对其2010年并网风电量所占比例较大和较小的日期进行预测,验证了选择负荷、并网风电量与负荷的比值及历史清算电价作为预测神经网络的输入变量是恰当的,其预测精度能够满足电力市场实际运行的需要。     

基于数据挖掘与支持向量机的现货市场出清价预测方法

现货市场环境下,市场出清价对于电力市场的发、用两侧参与者和市场管理者都是极为重要的信息。因此,市场出清价的预测研究越发重要。首先分析大多数传统电价预测方法采用的连续序列与该文选取的同时段的电价序列表现出的变化特征差异,给出选取同时段电价序列作为输入的原因。然后基于数据挖掘相似性理论,通过欧氏距离和角度距离2个维度识别历史电价相似序列,得到模型所需训练集数据。以支持向量机（SVM）为预测工具,并利用遗传算法对SVM的关键参数进行寻优。最后将模拟预测结果与不考虑历史相似状态的SVM模型、BP神经网络模型进行对比,通过误差分析证明了所提模型具备更高的预测精度。     

LS-SVM在电力系统负荷预测中的应用

针对支持向量机(SVM)法预测电力负荷存在空间划分参数率定人为因素影响的缺陷,采用谱分析法进行周期分析,比较了二次函数趋势的年负荷序列与周期性月负荷序列,并采用最小二乘支持向量机(LS-SVMlabl.5)法预测负荷.实例结果表明,周期性的月负荷序列实测值与预测值拟合度较好,预测精度高、简捷、合理、实用.     

基于电价波动率的现货电价风险度量研究

风险度量因子的选取是风险度量的核心工作之一,统计性质好的指标通常能降低建模复杂度,提高精度。根据电价序列的特征及电价风险度量模型的特点,提出以电价波动率替代电价作为风险度量因子,避开电价序列的非平稳性,建立GARCH-VaR模型用于现货电价风险度量,以北欧电力市场的电价风险度量为例,对模型的可行性和有效性进行检验,并将所提出的电价风险度量方法与电价波动率正态分布法、电价ARMA-GARCH模型度量的电价风险进行比较。结果表明,所提方法不仅能有效降低电价风险度量的模型复杂度,还可提高风险度量的准确性。     

基于大数据分析的上网电价预测模型研究与应用

通过对电力市场短期电价及长期电价预测进行大数据分析建模研究,结果表明指数平滑法、ARMA模型、最大信息熵模型及“关联规则+ARMA”等模型适合短期电价预测;状态空间模型及神经网络模型较适合长期电价预测。通过预测模型,对中长期电力市场以及现货电能量市场提供交易辅助决策有一定的指导作用。     

A combined modeling approach for wind power feed-in and electricity spot prices

Wind power generation and its impacts on electricity prices has strongly increased in the EU. Therefore, appropriate mark-to-market evaluation of new investments in wind power and energy storage plants should consider the fluctuant generation of wind power and uncertain electricity prices, which are affected by wind power feed-in (WPF). To gain the input data for WPF and electricity prices, simulation models, such as econometric models, can serve as a data basis.This paper describes a combined modeling approach for the simulation of WPF series and electricity prices considering the impacts of WPF on prices based on an autoregressive approach. Thereby WPF series are firstly simulated for each hour of the year and integrated in the electricity price model to generate an hourly resolved price series for a year. The model results demonstrate that the WPF model delivers satisfying WPF series and that the extended electricity price model considering WPF leads to a significant improvement of the electricity price simulation compared to a model version without WPF effects. As the simulated series of WPF and electricity prices also contain the correlation between both series, market evaluation of wind power technologies can be accurately done based on these series.     

The large-scale integration of wind generation: Impacts on price,reliability and dispatchable conventional suppliers

This work examines the effects of large-scale integration of wind powered electricity generation in a deregulated energy-only market on loads (in terms of electricity prices and supply reliability) and dispatchable conventional power suppliers. Hourly models of wind generation time series, load and resultant residual demand are created. From these a non-chronological residual demand duration curve is developed that is combined with a probabilistic model of dispatchable conventional generator availability, a model of an energy-only market with a price cap, and a model of generator costs and dispatch behavior. A number of simulations are performed to evaluate the effect on electricity prices, overall reliability of supply, the ability of a dominant supplier acting strategically to profitably withhold supplies, and the fixed cost recovery of dispatchable conventional power suppliers at different levels of wind generation penetration. Medium and long term responses of the market and/or regulator in the long term are discussed.     

Derivatives of the nodal prices in market power screening

This paper proposes a novel method for market power screening. This method is developed for horizontally and vertically consolidated power markets, and is based on the optimal power flow (OPF) model properties. It undertakes the calculation and analysis of a matrix of derivatives of the nodal prices with respect to generating unit offer prices. The analysis takes into consideration the influence of particular partakers and energy groups on the nodal prices. This paper presents a theoretical analysis of the partakers' and groups' profits as well as a method for the market power screening. Moreover, the issue wherein the LMP model generates prices above the highest bidding price is discussed. Illustrative case studies and case studies based on the Polish wholesale balancing power market model are also presented.     

Cross-border effects in interconnected electricity markets - an analysis of the Swiss electricity prices

Electricity markets in Europe become increasingly interconnected due to new grid connections and market coupling regulations. This paper examines the interdependencies between the Swiss electricity market and those of neighbouring countries. The Swiss market serves as a good example for a smaller electricity market which is increasingly affected by developments in the large neighbouring countries. To study these cross-border effects, especially those on Swiss electricity prices, we apply two different methodologies, an econometric and a Nash-Cournot equilibrium model.The analyses show that the Swiss electricity price correlates strongly with the German electricity price in the summer, but tends to follow the French electricity price in the winter. Another finding is that gas prices and the electricity load of neighbouring countries have a significant influence on prices. In particular, the load of France and Italy is driving up Swiss prices in the winter, while the German electricity demand and renewable energy generation have a larger influence on Swiss prices in the summer.     

Dynamic model for market-based capacity investment decision considering stochastic characteristic of wind power

This paper proposes a decentralized market-based model for long-term capacity investment decisions in a liberalized electricity market with significant wind power generation. In such an environment, investment and construction decisions are based on price signal feedbacks and imperfect foresight of future conditions in electricity market. System dynamics concepts are used to model structural characteristics of power market such as, long-term firms’ behavior and relationships between variables, feedbacks and time delays. For conventional generation units, short-term price feedback for generation dispatching of forward market is implemented as well as long-term price expectation for profitability assessment in capacity investment. For wind power generation, a special framework is proposed in which generation firms are committed depending on the statistical nature of wind power. The method is based on the time series stochastic simulation process for prediction of wind speed using historical and probabilistic data. The auto-correlation nature of wind speed and the correlation with demand fluctuations are modeled appropriately. The Monte Carlo simulation technique is employed to assess the effect of demand growth rate and wind power uncertainties. Such a decision model enables the companies to find out the possible consequences of their different investment decisions. Different regulatory policies and market conditions can also be assessed by ISOs and regulators to check the performance of market rules. A case study is presented exhibiting the effectiveness of the proposed model for capacity expansion of electricity markets in which the market prices and the generation capacities are fluctuating due to uncertainty of wind power generation.     

Impact of wind farm integration on electricity market prices

Wind generation is considered one of the most rapidly increasing resources among other distributed generation technologies. Recently, wind farms with considerable output power rating are installed. The variability of the wind output power, and the forecast inaccuracy could have an impact on electricity market prices. These issues have been addressed by developing a single auction market model to determine the close to real-time electricity market prices. The market-clearing price was determined by formulating an optimal power flow problem while considering different operational strategies. Inaccurate power prediction can result in either underestimated or overestimated market prices, which would lead to either savings to customers or additional revenue for generator suppliers.     

Impact of wind generation control strategies, penetration level and installation location on electricity market prices

Integration of wind facilities into power system grids have several impact on power system related issues including; transmission congestion, optimum power flow, system stability, power quality, system economics and load dispatch. Consequently, wind farm control strategy, location across the distribution network and its penetration level could have an impact on electricity market prices. This paper addresses these issues, by developing a single auction market model. An optimal power flow problem was formulated for determining the close to real time electricity market-clearing price and the total cost of generation. Simulation results, considering different operational cases, are presented to highlight the impact.     

Optimal investment strategies in decentralized renewable power generation under uncertainty

This paper presents a method for evaluating investments in decentralized renewable power generation under price un certainty. The analysis is applicable for a client with an electricity load and a renewable resource that can be utilized for power generation. The investor has a deferrable opportunity to invest in one local power generating unit, with the objective to maximize the profits from the opportunity. Renewable electricity generation can serve local load when generation and load coincide in time, and surplus power can be exported to the grid. The problem is to find the price intervals and the capacity of the generator at which to invest. Results from a case with wind power generation for an office building suggests it is optimal to wait for higher prices than the net present value break-even price under price uncertainty, and that capacity choice can depend on the current market price and the price volatility. With low price volatility there can be more than one investment price interval for different units with intermediate waiting regions between them. High price volatility increases the value of the investment opportunity, and therefore makes it more attractive to postpone investment until larger units are profitable.     

Market power in the European electricity market—The impacts of dry weather and additional transmission capacity

This paper uses a static computational game theoretic model of a fully opened European electricity market and can take strategic interaction among electricity-producing firms into account. The model is run for a number of scenarios: first, in the baseline under perfect competition, the prices differ due to the presence of various generation technologies and a limited ability to exchange electricity among countries. In addition, when large firms exercise market power, the model runs indicate that prices are the highest in countries where the number of firms is low. Second, dry weather would increase the prices in the hydro-rich Nordic countries followed by the Alpine countries. The price response would be about 20% higher with market power. Third, more transmission capacity would lower the prices in countries with high prices and it also reduces the impact of market power. Hence, more transmission capacity can improve market competitiveness.