Scholarlink

It is shown here that by a simple data reordering and by a following data partition the problem of parametrical identification of Wiener and Hammerstein systems with PWL nonlinearity could be reduced to a multifold linear parametric estimation problem. Afterwards, we will represent the former regression function in ‘pieces’, using ordinary FIR filters and the well-known linear convolution. Consistency of parameter estimates were investigated here, too. Simulation results are given for the Wiener system with invertible as well as noninvertible PWL nonlinearities.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Parametric Identification of Systems with Piecewise-Linear Nonlinearities

  • 章节

  • 发表日期：

    2026年

查看全文
  • Rimantas Pupeikis,
  • Kazys Kazlauskas

摘要

It is shown here that by a simple data reordering and by a following data partition the problem of parametrical identification of Wiener and Hammerstein systems with PWL nonlinearity could be reduced to a multifold linear parametric estimation problem. Afterwards, we will represent the former regression function in ‘pieces’, using ordinary FIR filters and the well-known linear convolution. Consistency of parameter estimates were investigated here, too. Simulation results are given for the Wiener system with invertible as well as noninvertible PWL nonlinearities.