Application of the Hopfield network in robust estimation of parametric membership sets for linear models

High computation rates can be achieved using artificial neural networks. Optimization problems can be solved by neural networks with feedback connections by employing a massive number of simple processing elements with high degree of connectivity between these elements. In this paper, an application of Hopfield neural networks in robust parametric estimation with unknown-but-bounded disturbance is presented. The internal parameters of the Hopfield neural network are obtained using the valid-subspace technique. These parameters are explicitly computed to assure the network convergence. A comparative analysis with other robust estimation methods is carried out by a simulation example.