Gray-box modeling of electric drives using recursive identification and radial basis functions

Modeling of electric drives coupled to complex mechanical loads may be a challenging task. Universal drives should have the capability to tune the drive control system to drive different loads and maintain system performance without knowing the details of the load. Gray-box modeling using neural networks is presented as a possible solution for the identification of the mechanical loads and the drive system. In the proposed gray-box modeling, the drive system is divided into the known part governed by the physical laws, which in our case is the electrical subsystem, and an unknown part, which in our case is the mechanical subsystem. The electrical subsystem is modeled using physical laws while the mechanical part is modeled a black box model using radial basis neural network. A two-stage parameter estimation algorithm based on linear I east squares for each stage is proposed. The tuning scheme is incorporated into the tunning of a feedback linearizing control scheme for a DC motor drive driving a static load. Simulation and experimental results are presented showing the potential of the approach.