Research Of Quantization Control For Networked Control Systems With Communication Constraints

With the rapid technological development of network, communication andcomputer and extensive use of smart sensors and actuators, the structure andoperation mode of the control system are changed, leading to a novel networkedcontrol systems (NCSs) that transmit measuring and control signals through thenetwork. It has been gradually used in metallurgy, power plants, and aerospaceand other fields. However, the introduction of the network will inevitably bringnetworked uncertainties factors/networked performance parameters (i.e.,network-induced delay, data packet loss, data packet out of order, etc.). Inspecially, networked uncertainty factors of multi-channel homogeneous (wire orwireless) and hybrid (wire and wireless) network may significantly deterioratethe system performance and potentially lead to system instability. Therefore,considering networked uncertainty factors and the plant with uncertaintyparameters, the model, quantizer, controller, observer and system stability areinvestigated. The main work is summarized as follows1) Under the homogeneous network environment, quantization control ofsingle-channel NCSs is studied. Signal is quantized by the logarithmic quantizerin both forward and feedback channel, and two Bernoulli distributions are usedto describe data packet loss. An observer-based state feedback controller isdesigned, and the closed-loop system augmented model is given. Furthermore,the sufficient condition for NCSs to be the exponentially mean-square stable andcontroller design method are presented.2) Under the homogeneous network environment, quantization control ofmulti-channel NCSs is investigated. To quantize the Measurement data in eachchannel, the logarithmic quantizer is employed, and quantization error is treatedas sector-bounded uncertainty. Multi-channel communication constrainsincluding network-induced delay and data packet disorder and loss are treated as packet disordering, which are then described by an analytical formulation.Markov jump systems model that contains multi-channel packet disorder andquantization is then established. A stochastically stable condition has beenderived, and the relationship among the quantization density, networkperformance parameters and system stability is presented. Finally, the mainresults are further extended to multi-channel NCSs with parameter uncertainty.3) Under the hybrid network environment, quantization control ofmulti-channel NCSs is studied. To reduce the burden of data transmission, themeasurement signal is quantized firstly before entering each network channel,and two different Markov chains are employed to describe network-induceddelay. Markov jump systems model that integrates multi-channelnetwork-induced delay and quantization is established. The results are furtherextended to multi-channel control system with parameter uncertainty.Simulation results confirm the effectiveness of the proposed method.