Adaptive recursive least squares algorithm for joint FIR filtering and pre-delay tracking and its application in the chemical industry process modeling

In this paper, the joint FIR filtering and pre-delay tracking system identification problem is considered. The input signal to the unknown system is first delayed then filtered. An adaptive recursive least squares algorithm based on fast transversal filters is developed and applied to the field data from the chemical industry process of some synthetic ammonia plant. In the simulation of 25 hours field running, it gives a relative prediction error no more than 3.6% and the modeling results are reasonable.