Hierarchical Recursive Least Squares Estimation Algorithm for Second-order Volterra Nonlinear Systems

This paper considers the parameter identification problems of a Volterra nonlinear system. In order to overcome the excessive calculation amount of the Volterra systems, a hierarchical least squares algorithm is proposed through combining the hierarchical identification principle. The key is to decompose the Volterra systems into three subsystems with a smaller number of parameters and to estimates the parameters of each subsystem, respectively. The calculation analysis indicates that the proposed algorithm has less computational cost than the recursive least squares algorithm. Finally, the simulation results indicate that the proposed algorithm are effective for identifying Volterra systems.

Automated summary

This paper proposes a hierarchical least squares algorithm for parameter identification problems of a Volterra nonlinear system. By decomposing the Volterra system into three subsystems with a smaller number of parameters and estimating the parameters of each subsystem separately, the proposed algorithm overcomes the excessive calculation amount of the Volterra systems. The calculation analysis shows that the proposed algorithm has lower computational cost compared to the recursive least squares algorithm, and simulation results demonstrate its effectiveness in identifying Volterra systems.