Study On Power Forecasting Of Distributed Wind And Photovoltaic Integration Generation System

With the current traditional energy sources are depleted, new energy and renewable energy technologies are gradually rose. Their solar and wind energy have abundant resources and broad prospects. But solar and wind with no regular change of seasons, weather, environment, etc., and the instability of its power generation is the bottlenecks in the system safe and stable operation of power constraints. Therefore, only power predicting quickly and accurately, in order to make the power grid safe and orderly, continuous supply and stable operation.However, the present study of predicted power concentrated mostly reflected in the power station, with the application of new energy generation more widely, especially in distributed generation systems, wind energy and solar energy as the main source of power generation more and more people’s attention and attention. Therefore, accurate prediction of distributed generation power system is safe and stable operation of the grid and reasonable arrangements for scheduling a powerful protection and the top priority of the current study. In this context, the problems of related paper in distributed power generation system small wind generators, photovoltaic power generation as well as the integration of short-term wind power prediction launched an in-depth thesis.Firstly, this paper presents a Phase Space Reconstruction neural network after clustering analysis based on the establishment of small wind turbine short-term power prediction model, namely clustering data preprocessing, the principle of chaos phase space reconstruction for training the neural network model to achieve the expected output and close to forecast power, finally using rough set theory to identify and assess the predictive value of the difference between the actual value and to achieve the optimal solution.Secondly, the proposed adaptive disadvantage chaos search space amendment BP was established by distributed PV short-term power forecasting model, this method can optimize BP neural network into local minima, improve prediction accuracy and convergence speed of the network.Next, in order to make the distributed generation system total power output is more accurate predictive value, established on the basis of small wind and photovoltaic short-term power generation predictive model, the establishment of integrated scenery total short-term power prediction model uses a combination forecasting method to obtain high precision power prediction value for distributed generation systems.Finally, we design a plan about integration of distributed generation systems scenery power prediction software. It include data acquisition system, the forecasting software functional design; structural design system design block diagram, a communication system, interfaces and interface. Power prediction by the design of distributed power generation system integration scenery, the initial realization of small wind turbines and photovoltaic power generation system combining wind power prediction integration theory and practical application, it provides a certain reference value basis and foundation with distributed generation system and new energy power prediction fields.