Gain Computation for Batch H2-FIR Filtering of Predictive Uncertain Disturbed Models using LMI

The gain for the receding horizon (RH) H2 finite impulse response (FIR) filter is derived using linear matrix inequality (LMI) under uncertainties, disturbances, initial, and measurement errors. The RH H2-FIR filter is developed by minimizing the squared Frobenius norm of the weighted error-to-error transfer function, where the weights are related to errors. The filter is tested by a harmonic model with an uncertain system matrix, and its higher accuracy is shown against the OFIR, Kalman, maximum likelihood FIR, and unbiased FIR (UFIR) filters.