Parameter Estimation for Hammerstein Systems with Colored Noise

An auxiliary model-based multi-innovation forgetting gradient parameter estimation algorithm is proposed for Hammerstein system with colored noise by combining the key-term separation technology, the multi-innovation identification theory, and the auxiliary model identification idea. Redundant parameter estimation is avoided by using key-term separation technique. The unknown inner variables are reconstructed by using the auxiliary model identification idea. Compared with single innovation forgetting gradient algorithm, the proposed algorithm has higher parameter estimation accuracy by extending innovation length. Simulation results show the effectiveness of the proposed algorithm.