To overcome the limitations of traditional state estimation in dealing with noise measurement

a general model of robust state estimation (RSE) is proposed based on maximum correntropy criterion (MCC)

which is able to unify several existing RSE models in theory and derive new RSE models. An adaptive robust state estimator (ARSE) for measuring noise is proposed

which makes it possible to obtain the distribution type of noise measurement through statistical learning. Then the corresponding optimal RSE model is selected online to adapt to the variation of noise measuring distribution type. ARSE is able to yield optimal estimation results closer to the true values of the state variables for common noise measuring distribution types. Simulation results show the effectiveness of the proposed ARSE. This work is supported by National High Technology Research and Development Program of China (863 Program) (No. 2012AA050208)

National Natural Science Foundation of China (No. 51407069) and the Fundamental Research Funds for the Central Universities (No. 2014QN02).