Parameter Estimate and Adaptive Control of DARMA Systems With Uniform Quantized Output Data

Abstract

This paper is concerned with parameter estimate and adaptive control problems of deterministic autoregressive moving average (DARMA) systems on the basis of quantized data of system output signals which are generated by a kind of uniform quantizer. By designing system input signals, the extended least-squares (ELS) algorithm with uniform output observations is proved to yield bounded estimation errors under some mild assumptions. Moreover, the adaptive tracking controller under inaccuracy observations is also designed. To analyse the properties of tracking error, we use the expanded form of the ELS and research the boundedness of quantization noise. In addition, we give the expression of tracking error and show how it depends on the size of quantization step when the quantization step satisfies some conditions. A numerical example is supplied to demonstrate the theoretical results.