Abstract
This paper presents the application of Independent Component Analysis (ICA) in power system. ICA is a method for finding unknown signals from observed ones. Given that observations are linear mixture of independent signal sources, then we can estimate signal sources from just only observations using ICA method. ICA is based on a linear-mixture model, in contrast, power system is a nonlinear system. However, applying DC power flow model, branch active power flows are represented as a linear mixture of active power injected into bus. DC power flow model matches ICA model. Therefore, in power systems, we can estimate bus injections from observed branch power flows using ICA. In this paper, we extend this idea to power estimation of Distributed Generation (DG) connected to radial distribution system. Note that ICA has two indeterminacies, scaling and permutation. To solve these matters, we use preliminary knowledge, DG maximum output power, connected node and network topology. We demonstrate the validity of this approach on a 69-bus distribution system, and then discuss the results with error analysis and spectrum analysis.
