Transition Matrices Estimation of Power Supply Based on Periodic Markov Chain

Proshyn S.

In this paper we research the power supply taking into account its Markov’s properties and its stochastic periodicity. As the mathematical model it is suggested to use the periodic Markov chain. Algorithm of finding of estimation of transition matrices is described. In its basis lays the separation of homogeneous elements out from the periodic chain, which allows to use existing methods of statistical processing. We conduct the definition of states of the power supply system and the period of chain with the account of peculiarity of power supply during the day at workdays and weekends. The software on the basis of developed algorithm is created and the estimation of matrices of transition is conducted. Received estimations of matrices reflect the probable peculiarities of behavior of electric power supply graphs and can be used in problems of prognosis.

Full text (pdf)