| At present, small wind/light distributed generation has become an important development trend of new energy power technology, which connected to the power grid in the form of the micro-grid as a two-way schedulable unit by the low voltage distribution power grid. For distribution power grip or micro-grid cluster, micro-grid participates in the electric energy distribution as a whole two-way control unit, which is the new member of the distribution system. Although micro-grid partake the electric energy distribution as a whole controlled unit,generator power units in micro-grid are mostly the of renewable energy resources generator with strong randomness. Therefore, their overalls source/load characteristics for outside has not only the randomness but also is associated with the control of micro-grid. Thus, the external power grid scheduling micro-grid as a controlled unit has became a difficulty. That is to say, external power grid usually cannot directly schedule and control for internal generating and using power equipment of micro grid. Therefore, how to quantify and assess the random source/load characteristic of micro-grid, and how to schedule exchange energy for the entrance of micro-grid, are the problems to be solved in the industry.The theory and method of several aspects about random power and load forecast in the micro-grid, potential adjustable capability assessment (PACA) of micro-grid and real-time energy optimization scheduling of micro-grip cluster, are deeply researched in this paper. The main results are as follows:(1) For the problem synchronously considering the trend and fluctuation in the traditional forecast methods, a super short-term wind power forecast method based on mean generating function(MGF) and optimal subset regression(OSR) is proposed. This method implements two time differences for the original sequence, and generates the MGF sequence and the accumulative sequence. The method synchronously considers the trend and fluctuation of wind speed. Therefore, this method has higher forecast accuracy and speed. Similarly, this method can also be used to solve the trend and of fluctuation of super short-term load forecast. The simulation results verify the effectiveness of this method for the super short term wind power forecast and load forecast.(2) Four state variables describing adjustable capability of the micro grid are put forward based on statistical theory. Four state variables are risk reserve electrical energy, risk deficiency electrical energy, stored electrical energy and surplus electric energy storage space, respectively. They lay the foundation for the energy scheduling of the superior power grid for the micro-grid. Based on these state variables, the standard of PACA for micro-grid and calculation method are proposed in this paper. In the state space, offset functions of micro-grid for source/load randomness are simulated through the Monte Carlo method. The simulation experiments verify the correctness of this theory.(3) Taking into account the adjustable ability of micro-grid as a decision variable, the real-time power scheduling mathematical model considering the PCC exchange energy as a control variable is firstly constructed in micro-grid cluster. And the method changing this model as minimum cost maximum flow model is put forward. The execution effect of scheduling instruction is simulated in micro-grid cluster. The results of simulation experiment indicate that the renewable energy resources can be utilized maximum by scheduling model for micro-grid cluster. At the mode without grid, the scheduling method for micro-grid cluster can make minimum of total overflow and total outage of all micro-grids. At the mode of grid, it can make the exchange power with grid minimum.(4) The proposed method of PACA for micro-grid based on risk electrical energy and real-time scheduling method for micro-grid cluster are used for micro-grid cluster system of the first middle school of sanli, Guigang City, Guangxi Province. Accordingly, the formulation of PACA method for exemplary micro-grid and scheduling method for exemplary micro-grid cluster are achieved. Therefore, the above proposed theory and method are verified by this practical example. |
PDF Full Text Request