| System short-term load forecasting predicts system load of future day or future week. Short-term load forecasting is an essential function in power system control centers.In the past three decades, many methods for short-term load forecasting have been proposed, for example, time series method, state space method, artificial neuron network method and so on, but these methods can not represent the electric load model accurately. The result of load data statistics shows that when electric load are scaled according to the below formula (1), the load scaled curves between similar day types are similar, but load scaled curves between different dayLoadscale = (Load - Loadmin ) /(Loadmax- Loadmin) (1)types are different, on basis of which short-term load forecasting can be divided into two steps which are the forecasting of load scaled curve and the forecasting of day maximal load and day minimal load. Based on the above analysis, this paper proposes one short-term load forecasting method with artificial neuron network and fuzzy logic strategy.Firstly day types are classified into five types: Monday, Weekday, Saturday, Sunday and Holiday.Secondly five Back-Propagation network models (one kind of artificial neuron network) are set up according to five different day types to forecast the scaled load curve. Thirdly day maximal load and day minimal load are partitioned into two parts: basic part and changed part. The basic part is predicted with statistical model according to the maximal load and the minimal load of the last days. The changed part is predicted with fuzzy logic strategy. The results of the scaled load curve and the day maximal load and minimal load are combined to get the final load of the future day.The practical test shows that the method this paper has proposed is more accurate than before.STLF(Short-Term Load Forecasting) software is developed for realizing the function of short-term load forecasting. With the development of IT technology and electrical power technology, STLF software is faced with new challenges. First STLF software needs to integrate with different EMS effectively and conveniently for efficiency and cost. Secondly STLF software is also required to process in Internet. In anyway, new STLF software should be open and distributed.As one kind of new software technology, component-oriented software development is another milestone of software engineering after object-oriented software development. During the process of component-oriented software development, one software system is divided into many modules which play different roles, these modules are designed for different components which communicate only by interfaces. If the external interfaces keep unchanged, any internal change of any component will not influence other components.IEC 61970 defines an application program interface (API) for an energy management system (EMS) for integration of different EMS effectively and conveniently, which includes Common Information Model (CIM) and Component Interface Specification (CIS). The CIM is an abstract model that represents all the major objects in an electric utility enterprise typically contained in an EMS information model. The CIM is partitioned into a set of packages. The CIS specifies the interfaces that a component (or application) should implement to be able to exchangeinformation with other components (or applications) and/or to access publicly available data in a standard way.In this paper, component-oriented software development and the standard IEC 61970 are adopted to develop new STLF software. STLF software is divided into three parts which are data interface part, algorithm logic part, man-machine part. The function of data interface part is to convert external load data which comes from database and has many kinds of structures to internal load data which is used for forecasting algorithm. The function of algorithm logic part is to execute forecasting process. The function of man-machine part is to receive input from users a... |
PDF Full Text Request