Coding Scheme Design And Stability Analysis Of Quantized Control Systems

Under communication constraints, it is significant to take into account theInteraction between control and communication performance for the analysis andsynthesis of the system. Information is needed to be encoded and decoded fortransmission through finite data rate channels. Coding scheme design has importanteffect on the system stability. The coding schemes adopted in the literature aredivided into two classes: coding scheme of uniform quantization and coding schemeof logarithmic quantization. This paper mainly introduces the application oflogarithmic coding scheme. In such a framework, since feedback information istransmitted through digital network, we will be facing new challenges in manycontrol problems. Thus the control design and state estimation are very importantamong them.Meanwhile, the paper briefly discusses a range of feedback control problems ofquantization and estimation. Although quantization in signal processing and digitalcommunications is a research topic, in these areas the direct application of controltechnology is not appropriate. The main reason is related to feedback control system,which has two major implications:1) The quantized signal re-enters the system;2)The input signal to the quantizer is not known a prior to be bounded. Both of theseimplications make the analysis and design of quantizers much more difficult.Considering these effects, in order to simplify system design and analysis process,quantizing error is often ignored or viewed as simple noise model. But this methodis effective when the quantization error is small. To overcome the above problem s,we put forward the method of logarithmic quantization. It can be applied in manyprograms and can evaluate the quantization error. These advantages don’t appear inother quantizers.Finally, we study the linear systems under the non-ideal network condition andanalyze the stability of system using logarithm quantization method. At the sametime, simulation examples show that the proposed method is effective. This methodcan stabilize the control system under the joint action of non-ideal networkenvironment and quantization.