ABSTRACT
With the vigorous development of wind power in recent years, most wind farms are incorporated into the grid scheduling plan (Chi 2007; Liu 2013). However, wind power due to its stochastic characteristic, makes large-scale of it merged into the grid on one hand bring great influence on grid’s scheduling and system planning and so on; on the other hand, the power system taking the pressure to peak load and frequency, has to limit its outputs to ensure grid’s safe, stable operation and the power balance, and abandoned wind power has become a common way adopted by most wind farms. So when large scale of wind cluster outputs are limited by the grid, dispatching needs to comprehensively consider many factors during its actual operation and the wind farm outputs also need to be re-planned (Bai et al., 2010; Han et al., 2010). In actual operation, wind power plan is decided by the scheming of total threshold power, and whether the scheduling is reasonable relates to the utilization of wind power and benefits of grid system. For correlation can give a more scientific explanation to the volatility of wind power outputs, and grasping regulars of correlation and volatility can help reduce the adverse impacts on grid scheduling, namely it will have an effect on the wind farm outputs plan. Therefore, how to implement the distribution with factors of wind power outputs reasonably and effectively, when not completely accepted by the grid, will have a great grid, will
Statements on cross-correlation coefficient can be described by the followed formula:
r y y
y y
N =
( )x x ( )
x x ( )
(1)
where x, y = active power sequences of two wind farm outputs; r = cross-correlation coefficient. r > 0 means a positive correlation and r < 0 a negative correlation. The larger the absolute value of r, the higher the degree of correlation, otherwise the lower the degree of correlation.
