Optimal Control Of Network Delay Systems In Noisy Environments

Network Control System(NCS), which is also known as network control, is the development and integration product of computer technology, communication technology and control technology. NCS is widely used, which can be applied to almost any need for data exchange occasions. NCS includes DCS, industrial Ethernet and fieldbus systems. It reflects that the control system is tend to the development of network, integration, distribution and intelligent nodes. It has an unrivaled superiority to compare with traditional control systems. However there are some problems in the communication process such as a network-induced delay, congestion and packet since the signal spread through the network.In this paper, considering a class of networked control systems with uncertain interference which includes a random noise interference term, propose a stochastic processing perturbation algorithms based on the Kalman filter method. Obtain the feedforward-feedback optimal control law by solving Riccati question and Stein question. Then deduce the feedforward-feedback optimal control law of the linear discrete systems with random interference according to the separation principle. In addition, according to the network control system has the characteristics of network- induced delay, add a state delay and a control delay to the system model, and then convert the corresponding systems by different Delay-free Conversions. First, propose a feedforward-feedback optimal controller design for linear discrete systems with random noise. Secondly, networked control system with control delay is modeled to obtain a linear discrete system model with control delay and random noise. Then transform the system into a linear discrete time systems without delays by delay-free conversion and a design feedforward- feedback controller. Chapter Four discuss the Network Delay Systems with state delays. Transform the system model by constructing a sensitivity function and the corresponding Maclaurin expansion, while increasing a companion vector to compensate the control law. Considering the difficulty of the actual operation, estimate the linear discrete system by using the M-order suboptimal control law. Finally, compare and analyze the stability of NCS in the noise environment under the feedforward- feedback optimal control and classic state feedback control respectively by studying the simulation. It can be seen from the simulation results that the control effect and the performance of the noise system are effectively reformed. And the improved system has a better ability of rejecting external perturbations and robustness which demonstrate the effectiveness and feasibility of this method.