Non-reduced order method for state estimation of quaternion-valued inertial neural networks

The state estimation problem of delayed quaternion-valued inertial neural networks(QVINNs) is investigated via quaternion direct method. Based on the non-reduced order method, some algebraic criteria are derived to realize the state estimation for the QVINNs. Moreover, the algorithm for calculating the estimator gain parameter
$k_{i}$
is provided as well. Finally, one numerical simulation example is presented to verify the efficacy of theoretical results.