On the residual based stochastic gradient algorithm for dual-rate sampled-data systems using the polynomial transform technique

This paper uses the polynomial transformation technique to transform an ARX model into a special model that can be identified with dual-rate input-output data, and presents the residual based stochastic gradient algorithm for dual-rate sampled-data systems, and studies convergence properties of the algorithm involved. The analysis indicates that the parameter estimation error consistently converges to zero under some proper conditions. Finally, we test the algorithms proposed in paper by a simulation example and show their effectiveness.