The rapid development of power system monitoring technology provides a solid data foundation for system analysis based on sample information

which is a breakthrough for the power system operating analysis theory. The injection shift factor (ISF) is an important linear sensitivity factor in the power system. Other linear factors such as line outage distribution factor (LODF)

power transfer distribution factor (PTDF)

and outage transfer distribution factor (OTDF)

can all be directly derived from the ISFs. Therefore

ISFs play an important role in the power system congestion management

security evaluation

optimal power flow and so on. In this context

a novel ISF probabilistic estimation approach based on the Bayesian linear regression model and Gibbs sampling solution is proposed. The proposed approach not only has the advantages of the data-based deterministic ISF estimation approaches

but also provides probabilistic estimation results of the ISFs while quantifying the ISF deviation resulting from the linearization assumption and the random measurement noise. The ISF deviation information provides necessary decision basis for the robust power system security evaluation and dispatch technologies. Actual measurement data of a local transmission network in Henan Province are satisfactorily used to test the performance of the proposed approach. This work is supported by National Basic Research Program of China (973 Program) (No. 2013CB228205)

National Natural Science Foundation of China (No. 51007047

No. 51077087) and Shandong Provincial Natural Science Foundation of China (No. ZR2014EEM022).