Analyzing The Random Property Of The Node Voltage In The Distribution Network

In order to grasp the running condition of the power distribution system correctly which changes quickly every time, and guarantee the power distribution system operating safely and economically, we should analyze the random properties of various parameters of the power distribution system. Node voltage is one of the most important parameters of power distribution system, so that this paper mainly analyses the random property of node voltage.This paper firstly introduces three filtering approaches of white noise, which are least square method, filtering algorithm based on automatic regressive mode and Kalman filter separately, and carries out practical example emulations using these algorithms. And then studies the filtering algorithm of colored noise based on Kalman filter structure. Not all the colored noises can be transformed to white noises, and just only generalized Markov colored noise sequence can be transformed to white noise. This paper exclusively does research in how to transform this kind of colored noise to white noise. In this paper the dimension extension of observation equation and measurement equation is studied, but this method creates much more computations and extends the dimension of Kalman filter. In order to avoid the dimension extension and complicated computations, when emulating, I uses the estimated auto-correlation sequence of observation data to compute the transition function and the variance of the white noise transformed from colored noise. Through emulating, it is indicated that this method obtains better filtering effect than the above approaches of white noise.