پیش بینی مصرف انرژی و ریسک تقاضای بیش از حد در سیستم توزیع
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|2839||2005||8 صفحه PDF||سفارش دهید||2104 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Physica A: Statistical Mechanics and its Applications, Volume 355, Issue 1, 1 September 2005, Pages 46–53
An empirical model for prediction of energy consumption in a distribution system is described. The model resembles a normalized radial basis function neural network whose neurons contain prototype joint data about the consumption process and the environment. A set of prototype patterns of consumption and environmental variables is formed from a record of a multi-component time series by a self-organized process. Prediction of energy consumption is performed by a conditional average estimator based upon known prototype patterns and given future values of environmental variables. Importance of these variables for the prediction is determined by a genetic algorithm. Prediction performance of the model is tested on a one-year-long consumption record of a gas distribution system. Prediction error is determined by the difference between predicted and actually observed consumption. Its value depends on time and amounts to a few percent of the actual consumption. The probability distribution of prediction error is estimated from a properly selected time interval of prediction. This distribution can be used to estimate the risk of energy demand beyond some prescribed value. For an optimization of the distribution process, a cost function that includes operation and control costs of a distribution system as well as penalties related to excess energy demand is proposed. Its minimum corresponds to an economically optimal energy distribution.
Energy distribution systems exhibit appreciable fluctuations of energy consumption over shorter and longer time intervals. These fluctuations are a consequence of the changing needs of clients that are related to the population activity and environmental conditions. A distributor must satisfy the guaranteed energy demand of clients and should not exceed the contractually determined energy supply of providers. Thus the basic task of a distributor is to minimize the difference between requested and provided energy amounts by control actions in a way to achieve maximal economic profit. For this purpose, a reliable prediction  of energy consumption  is needed. One of the basic tasks of prediction is the estimation of excess energy demand risk, since it is related with an excess cost. Energy consumption is generally a high-dimensional complex process that depends on deterministic and random components. Deterministic components can be analytically described by characteristics of the distribution system and some variables that characterize the influence of the environment. Conversely, random components cannot be described analytically but only probabilistically. Reduction of the related uncertainty in system description can be achieved by properly accounting various environmental variables and their influence on energy consumption. A non-parametric regression  and genetic algorithm  are applied for this purpose.
نتیجه گیری انگلیسی
This paper presents a new empirical modeling approach to prediction and optimal control of energy distribution systems. A new method for optimal estimation of prediction error is proposed. Based on the prediction error, the accuracy of the predicted energy consumption can be estimated. This further enables the introduction of the system cost function on which an optimal system control can be determined. The quality and applicability of the model has been tested on a gas distribution system of a typical Slovenian city. The difference between predicted and actually observed consumption amounts to only a few percent of the actual consumption. A properly selected weighting function for the prediction error indicator can enhance the overall quality of the model considerably. The selection must be carried out in connection with a specific optimization goal for which the model is used.