Adaptive Parameter Estimation for MISO System Using Decomposition Principle

In the paper, we discuss the identification of the multiple-input and single-output (MISO) system. Based on the internal relation of the linear subsystem and nonlinear element, the estimation model of the system considered is recasted as the entirety estimation equation. For the equation, here the single parameter item and bilinear parameter item coexist. Because the bilinear item includes compound parameters, the calculative burden of the identification method will be high. Then, to cut down the amount of calculation, the matrix conversion technology is utilized to reconstruct two estimation models. In parameter estimation process, for each model, an adaptive parameter estimation scheme is submitted to interactively identify the estimated parameters by virtue of hierarchical identification idea. Via the stochastic theory and martingale theorem, the convergence of parameter estimation is provided. The numerical simulation verifies the usefulness of the presented estimator.