The investigation of dimensionality reduction method for state evaluation of oil-immersed transformer in 220kV substation

The state evaluation of power equipment is an important basis for maintenance scheduling. This paper starts from the performance state data of power transformer and aims to study the matrix model and mathematical analysis method that is applicable to the state evaluation parameter system of power equipment in 220 kV substations. Different from the traditional threshold deduction method, this paper is more inclined to propose a closed loop model by finding the correlation among the internal key parameters of the system by reducing the dimension of the state matrix and increasing the dimension of iterations. This paper presents the basic state evaluation parameter system from various aspects which are based on the multi-source information of 220 kV substations. The dimensionality reduction of the matrix is carried out for the criterion matrix of the nature.