Decomposition algorithm for power system state estimation by the test equation technique and its implementation on the basis of multi-agent approach

State estimation in the modern power systems of large dimensionality is a sophisticated problem which is accompanied by the difficulties related to inhomogeneity of the calculated schemes, a large volume of information to be processed and the requirement for high-speed software. Distributed data processing in decomposition of the state estimation problem is an efficient method of tackling the difficulties. The paper addresses decomposition algorithm of state estimation by the test equation technique. The algorithm is based on the structural and functional decomposition of state estimation problem. The structural decomposition suggests dividing the calculated schemes into subsystems. A two-level algorithm is proposed to divide the calculated network into subsystems. Application of the test equation technique which makes it possible to fix the values of measured variables and set zero variances for them, as well as placement of phasor measurement units (PMU) at boundary nodes make it possible to essentially simplify the procedure of coordinating the solutions obtained for separate subsystems. The functional decomposition is performed in accordance with the problems solved within state estimation: a priory detection of bad data, state estimation on the basis of quadratic and robust criteria. The multiagent system proposed to implement the suggested state estimation algorithm is described. The example of calculating the estimates while dividing the calculated scheme into subsystems with PMU placed at boundary nodes is presented.