Study On The Methods Of Mid-Long Term Load Forecasting For Distribution Network

Mid-long term load forecasting is the precondition and foundation of distribution networks planning, the accuracy of load forecasting directly affects the quality of distribution networks planning and the safe economical operation of the distribution network; Load forecasting is also an important part of power system modernizing management. Distribution network planning demands not only the gross of future load, but also its position in the planning area, namely gross load forecasting and spatial load forecasting, and which have great guiding significance for distribution network planning.Based on analysis of each gross load forecasting method, a new gross load weights changeable combination forecasting method for distribution networks based on hierarchical structure is presented. On the choice of single models for combination forecasting, an optimal selection method based on area correlation degree is presented. On the determination for combined weight of combination model, according to the idea of analytic hierarchy process, a hierarchical structure for determining combined weight is constructed: entropy method is introduced to ascertain the weights of model-evaluation indexes; variance-covariance optimization combination method and grey relational analysis are used to ascertain the weights of each single model under each evaluation index respectively, and determine the combined weights of combination forecasting method finally. In the whole process of combination forecasting, equal dimension information is introduced according to the"inertia principle"of load development, thus achieves weights changeable combination forecasting, and forecasting results much better reflect the development law of gross load. At last, the reliability and applicability are validated by an actual example.Based on analysis of each spatial load forecasting method, a new spatial load forecasting method for distribution networks based on rough set theory and cellular automata is presented. In the process of dynamic development of small area land types, cellular automata theory is introduced to simulate the process of urban land-use dynamic development. The iteration time and adjusting time of transition rules of CA are determined with the practical situation of urban development, and the forecasting results will be more reasonable and consistent with the actual development law of spatial load in planning area. In the process of obtaining small area land-use decision-making rules, rough set theory is introduced to obtain future small area land-use decision-making by carry out attributes reduction to potential influencing factors. In the whole process of spatial load forecasting, according to the idea of"from top to bottom"load distribution of land-use based method, gross load will be distributed into each small-area, thus accomplish spatial load forecasting. At last, the applicability and credibility of this method is validated by an actual example, and a novel and practical spatial load forecasting method for distribution networks is presented.