Short-term forecasting of electricity demand by decomposition analysis

A model for short-term forecasting of electricity demand is developed, which consists of an annual base load augmented by demand variations on a weekly, daily, and hourly or half-hourly basis. Each of these components is individually modelled and forecast, and the aggregate forecast is refined by inclusion of a stochastic component for characterization of variations not attributable to elements of time. The approach is consistent with the operational concept of power system management, and can be readily adopted for on-line forecasting and control.     

Short-term electricity demand forecasting using double seasonal exponential smoothing

This paper considers univariate online electricity demand forecasting for lead times from a half-hour-ahead to a day-ahead. A time series of demand recorded at half-hourly intervals contains more than one seasonal pattern. A within-day seasonal cycle is apparent from the similarity of the demand profile from one day to the next, and a within-week seasonal cycle is evident when one compares the demand on the corresponding day of adjacent weeks. There is strong appeal in using a forecasting method that is able to capture both seasonalities. The multiplicative seasonal ARIMA model has been adapted for this purpose. In this paper, we adapt the Holt–Winters exponential smoothing formulation so that it can accommodate two seasonalities. We correct for residual autocorrelation using a simple autoregressive model. The forecasts produced by the new double seasonal Holt–Winters method outperform those from traditional Holt–Winters and from a well-specified multiplicative double seasonal ARIMA model.     

Triple seasonal methods for short-term electricity demand forecasting

Online short-term load forecasting is needed for the real-time scheduling of electricity generation. Univariate methods have been developed that model the intraweek and intraday seasonal cycles in intraday load data. Three such methods, shown to be competitive in recent empirical studies, are double seasonal ARMA, an adaptation of Holt–Winters exponential smoothing for double seasonality, and another, recently proposed, exponential smoothing method. In multiple years of load data, in addition to intraday and intraweek cycles, an intrayear seasonal cycle is also apparent. We extend the three double seasonal methods in order to accommodate the intrayear seasonal cycle. Using six years of British and French data, we show that for prediction up to a day-ahead the triple seasonal methods outperform the double seasonal methods, and also a univariate neural network approach. Further improvement in accuracy is produced by using a combination of the forecasts from two of the triple seasonal methods.     

Modeling special-day effects for forecasting intraday electricity demand

We propose and apply a novel approach for modeling special-day effects to predict electricity demand in Korea. Notably, we model special-day effects on an hourly rather than a daily basis. Hourly specified predictor variables are implemented in the regression model with a seasonal autoregressive moving average (SARMA) type error structure in order to efficiently reflect the special-day effects. The interaction terms between the hour-of-day effects and the hourly based special-day effects are also included to capture the unique intraday patterns of special days more accurately. The multiplicative SARMA mechanism is employed in order to identify the double seasonal cycles, namely, the intraday effect and the intraweek effect. The forecast results of the suggested model are evaluated by comparing them with those of various benchmark models for the following year. The empirical results indicate that the suggested model outperforms the benchmark models for both special- and non-special day predictions.     

Development of improved adaptive approaches to electricity demand forecasting

This paper develops a short-term forecasting system for hourly electricity load demand based on Unobserved Components set up in a State Space framework. The system consists of two options, a univariate model and a non-linear bivariate model that relates demand to temperature. In order to handle the rapidly sampling interval of the data, a multi-rate approach is implemented with models estimated at different frequencies, some of them with ‘periodically amplitude modulated’ properties. The non-linear relation between demand and temperature is identified via a Data-Based Mechanistic approach and finally implemented by Radial Basis Functions. The models also include signal extraction of daily and weekly components. Both models are tested on the basis of a thorough experiment in which other options, like ARIMA and Artificial Neural Networks are also used. The models proposed compare very favourably with the rest of alternatives in forecasting load demand.     

Application of machine learning techniques for supply chain demand forecasting

Full collaboration in supply chains is an ideal that the participant firms should try to achieve. However, a number of factors hamper real progress in this direction. Therefore, there is a need for forecasting demand by the participants in the absence of full information about other participants’ demand. In this paper we investigate the applicability of advanced machine learning techniques, including neural networks, recurrent neural networks, and support vector machines, to forecasting distorted demand at the end of a supply chain (bullwhip effect). We compare these methods with other, more traditional ones, including naïve forecasting, trend, moving average, and linear regression. We use two data sets for our experiments: one obtained from the simulated supply chain, and another one from actual Canadian Foundries orders. Our findings suggest that while recurrent neural networks and support vector machines show the best performance, their forecasting accuracy was not statistically significantly better than that of the regression model.     

Compactness and balancing in scheduling

Summary We consider in this paper some graph-theoretical models for scheduling problems; we concentrate on 2 types of requirements which arise in real life situation: first a good schedule must be compact, i.e. there must not be too many interruptions of work for each machine and for each agent. Besides, a schedule should be balanced in the sense that for all periods the numbers of jobs which are processed simultaneously should be almost the same.We show how these requirements can be taken into account in some graph coloring models.

---

Zusammenfassung Wir betrachten in dieser Arbeit einige graphentheoretische Modelle für die Maschinenbelegungsplanung und konzentrieren uns dabei auf zwei Anforderungen, die in der Praxis auftreten: Zum einen muß ein guter Belegungsplan kompakt sein, d.h. die Arbeit ciner Maschine bzw. eines Arbeiters soll nicht zu häufig unterbrochen werden.Andererseits sollte der Plan ausgewogen sein in dem Sinne, daß für alle Perioden die Anzahl der gleichzeitig durchgeführten Arbeiten annähernd gleich bleibt. Wir zeigen, wie man diese beiden Anforderungen mit Hilfe Färbungsproblemen aus der Graphentheorie berücksichtigen kann.

     

Management of supply chain: an alternative modelling technique for forecasting

Forecasting is a necessity almost in any operation. However, the tools of forecasting are still primitive in view of the great strides made by research and the increasing abundance of data made possible by automatic identification technologies, such as radio frequency identification (RFID). The relationship of various parameters that may change and impact decisions are so abundant that any credible attempt to drive meaningful associations are in demand to deliver the value from acquired data. This paper proposes some modifications to adapt an advanced forecasting technique (GARCH) with the aim to develop it as a decision support tool applicable to a wide variety of operations including supply chain management (SCM). We have made an attempt to coalesce a few different ideas toward a ‘solutions’ approach aimed to model volatility and in the process, perhaps, better manage risk. It is possible that industry, governments, corporations, businesses, security organizations, consulting firms and academics with deep knowledge in one or more fields, may spend the next few decades striving to synthesize one or more models of effective modus operandi to combine these ideas with other emerging concepts, tools, technologies and standards to collectively better understand, analyse and respond to uncertainty. However, the inclination to reject deep-rooted ideas based on inconclusive results from pilot projects is a detrimental trend and begs to ask the question whether one can aspire to build an elephant using mouse as a model.     

Rule-based procedures in short-term electricity load forecasting

A short-term load-forecasting system has been developed to predictload demand for the Irish Electricity Supply Board from 1 to10 days in advance. Model output is acceptably accurate at mosttimes, but special days such as public holidays and the Easterweekend etc. require individual treatment. The basic model usedin everyday load forecasting is presented first, and specialdays when this model fails in accuracy are then examined andtreated separately using rule-based procedures. These proceduresare based on identified relationships between weather conditionsand prevailing daily load shapes. Comparisons are made betweenthe forecasts for these special days as given by the standardmodel and the ruled-based procedures.     

Design of electricity supply networks

The problem of designing a secure electricity supply network at minimal cost is formulated as a mathematical program. It is also shown how computationally convenient new constraints may be derived and these are added to the original set. The problem is dualized and solved approximately. It is indicated how this approach can be built into a Branch-and-Bound scheme for solving the original design problem, and an illustrative example is given.     

Intermittent demand: Linking forecasting to inventory obsolescence

The standard method to forecast intermittent demand is that by Croston. This method is available in ERP-type solutions such as SAP and specialised forecasting software packages (e.g. Forecast Pro), and often applied in practice. It uses exponential smoothing to separately update the estimated demand size and demand interval whenever a positive demand occurs, and their ratio provides the forecast of demand per period. The Croston method has two important disadvantages. First and foremost, not updating after (many) periods with zero demand renders the method unsuitable for dealing with obsolescence issues. Second, the method is positively biased and this is true for all points in time (i.e. considering the forecasts made at an arbitrary time period) and issue points only (i.e. considering the forecasts following a positive demand occurrence only). The second issue has been addressed in the literature by the proposal of an estimator (Syntetos-Boylan Approximation, SBA) that is approximately unbiased. In this paper, we propose a new method that overcomes both these shortcomings while not adding complexity. Different from the Croston method, the new method is unbiased (for all points in time) and it updates the demand probability instead of the demand interval, doing so in every period. The comparative merits of the new estimator are assessed by means of an extensive simulation experiment. The results indicate its superior performance and enable insights to be gained into the linkage between demand forecasting and obsolescence.     

A coordination system for seasonal demand problems in the supply chain

This article considers a single product coordination system using a periodic review policy, participants of the system including a supplier and one or more heterogeneous buyers over a discrete time planning horizon in a manufacturing supply chain. In the coordination system, the demand of buyer in each period is deterministic, the supplier replenishes all the buyers, and all participants agree to plan replenishment to minimize total system costs. To achieve the objective of the coordination system, we make use of small lot sizing and frequent delivery policies (JIT philosophy) to transport inventory between supplier and buyers. Moreover, demand variations of buyers are allowed in the coordination system to suit real-world situations, especially for hi-tech industries. Furthermore, according to the mechanisms of minimizing the total relevant costs, the proposed method can obtain the optimal number of deliveries, shipping points and shipping quantities in each order for all participants in the coordination system.     

The spot and balancing markets for electricity: open- and closed-loop equilibrium models

Computational Management Science - The increasing penetration of inflexible and fluctuating renewable energy generation is often accompanied by a sequential market setup, including a day-ahead spot...     

Using imperfect advance demand information in forecasting

Email: t.tan{at}tue.nl Received on 4 January 2007. Accepted on 11 January 2008. In this paper, we consider the demand-forecasting problem ofa make-to-stock system operating in a business-to-business environmentwhere some customers provide information on their future orders,which are subject to changes in time, hence constituting imperfectadvance demand information (ADI). The demand is highly volatileand non-stationary not only because it is subject to seasonalityand changing trends but also because some individual clientdemands have significant influence on the total demand. In suchan environment, traditional forecasting methods may result inhighly inaccurate forecasts, since they are mostly developedfor the total demand based only on the demand history, not makinguse of demand information and ignoring the effects of individualorder patterns of the customers. We propose a forecasting methodologythat makes use of individual ordering pattern histories of theproduct–customer combinations and the current build upof orders. Moreover, we propose making use of limited judgementalupdates on the statistical forecasts prior to the use of ADI.     

iGreen: green scheduling for peak demand minimization

Home owners are typically charged differently when they consume power at different periods within a day. Specifically, they are charged more during peak periods. Thus, in this paper, we explore how scheduling algorithms can be designed to minimize the peak energy consumption of a group of homes served by the same substation. We assume that a set of demand/response switches are deployed at a group of homes to control the activities of different appliances such as air conditioners or electric water heaters in these homes. Given a set of appliances, each appliance is associated with its instantaneous power consumption and duration, our objective is to decide when to activate different appliances in order to reduce the peak power consumption. This scheduling problem is shown to be NP-Hard. To tackle this problem, we propose a set of appliance scheduling algorithms under both offline and online settings. For the offline setting, we propose a constant ratio approximation algorithm (with approximation ratio \(\frac{1+\sqrt{5}}{2}+1\)). For the online setting, we adopt a greedy algorithm whose competitive ratio is also bounded. We conduct extensive simulations using real-life appliance energy consumption data trace to evaluate the performance of our algorithms. Extensive evaluations show that our schedulers significantly reduce the peak demand when compared with several existing heuristics.     

Short-term electricity procurement: A rolling horizon stochastic programming approach

This paper addresses the problem faced by a large electricity consumer in determining the optimal procurement plan over a short-term time horizon. The inherent complexity of the problem, due to its dynamic and stochastic nature, is dealt by means of the stochastic programming modeling framework. In particular, a two-stage problem is formulated with the aim of establishing the optimal amount of electricity to be purchased through bilateral contracts and in the Day-Ahead Electricity Market. Recourse actions are used to hedge against uncertainty related to future electricity prices and consumer’s needs. The optimal plan is defined so to minimize the overall cost and to control risk, which is measured in the form of violation of budget constraints. The stochastic model is dynamically solved in a rolling horizon fashion by iteratively considering more and more recent information and a planning horizon of decreasing length. Extensive numerical experiments have been carried out to assess the performance of the proposed dynamic decision approach. The results collected considering a real test case are very encouraging and provide evidence of the superiority of the approach also in comparison with other alternative procurement strategies.     

Short-term work scheduling with job assignment flexibility for a multi-fleet transport system

In recent years, planning and management of operations of transportation companies has become increasingly complex as tighter financial constraints affect the ability to respond to changing demands for travel. This paper investigates how a firm can reduce its total labor cost, enhance the flexibility of its operations and improve worker productivity and utilization by determining the right mix of jobs in work schedules. Using a method we have developed to generate low cost work schedules for bus drivers of an inter-city transport system in India, we study the changes in worker productivity and utilization when a mix of primary and secondary jobs is built in work schedules. We also investigate other factors that can strongly influence worker productivity and utilization such as the amount of overtime in work schedules, the level of job assignment flexibility, and staff size. Test results and discussions of managerial implications are presented.     

优化交货期窗口的两阶段供应链排序问题

研究一类优化交货期窗口的两阶段供应链排序问题. 优化交货期窗口是指交货期窗口的开始与结束时刻是决策变量, 不是输入常量. 两阶段是指工件先加工, 后运输: 加工阶段是一台加工机器逐个加工工件;运输阶段是无限台车辆分批运输完工的工件. 工件的开始运输时刻与完工时刻之差定义为工件的储存时间, 且有相应的储存费用. 若工件的运输完成时刻早于(晚于)交货期窗口的开始(结束)时刻, 则有相应的提前(延误)惩罚费用. 目标是极小化总提前惩罚费用、总延误惩罚费用、总储存费用、总运输费用以及与交货期窗口有关的费用之和. 针对单位时间的延误惩罚费用不超过单位时间的储存费用、单位时间的储存费用不超过单位时间的提前惩罚费用的情形, 给出了时间复杂性为O(n^{8})的动态规划算法.